JP5534045B2 - Road shape estimation device - Google Patents

Description

  The present invention relates to a road shape estimation apparatus, and more particularly to a road shape estimation apparatus that is mounted on a vehicle and estimates the shape of a road on which the vehicle is traveling.

  2. Description of the Related Art Conventionally, an in-vehicle system has been developed that estimates the shape of a road on which a vehicle travels, controls the steering of the vehicle according to the road shape, and notifies the danger of a vehicle collision. In the in-vehicle system as described above, it is necessary to accurately estimate the shape of the road in order to appropriately control the vehicle. An apparatus for estimating such a road shape is disclosed in, for example, Patent Document 1.

  The vehicle road shape recognition device disclosed in Patent Document 1 detects the position and shape of a roadside stationary object such as a guardrail based on a plurality of detection points obtained from a radio wave radar. Specifically, the vehicle road shape recognition device extracts a detection point group composed of adjacent detection points among a plurality of detection points obtained by a radio wave radar, and the detection point group is used as a roadside stationary object such as a guardrail. Detect as. Generally, a roadside stationary object such as a guardrail is arranged along a road on which a vehicle travels, and has a shape similar to the road. Therefore, the vehicle road shape recognition device can estimate and calculate the road shape based on the shape of the detection point group representing the roadside stationary object such as the guardrail detected as described above.

  In a vehicle road shape recognition apparatus such as that disclosed in Patent Document 1, detection points that exist within a predetermined distance range within a predetermined direction are simply connected and grouped. In such a simple process, when the guard rail is curved in a curve or the like, the shape indicated by the group of detection points may be different from the shape of the guard rail.

  Therefore, the inventor of the present application invented a method for grouping detection points as shown in Patent Document 2. Specifically, the road shape estimation apparatus shown in Patent Document 2 sets a rectangular connection range around a detection point that is a connection source. Then, the road shape estimation apparatus connects the detection point of the connection source and the detection point of the connection destination by a line segment with the detection point in the connection range as the connection destination. The road shape estimation device calculates a connection line indicating a road shape by repeating such connection processing. When setting the connection range, the road shape estimation device rotates and sets the connection range according to the angle (azimuth) indicated by the line segment connecting the previous detection points. By rotating the connection range in this way, it is possible to obtain a connection line in which detection points are continuously connected along a guard rail that is curved in a curve or the like.

JP 2007-161162 A PCT / JP2010 / 006489

  However, the conventional road shape estimation apparatus may not be able to accurately calculate connection lines along the road shape.

  For example, as shown in FIG. 21, a situation is assumed in which roadside stationary objects such as guardrails are intermittently arranged. FIG. 21 is a diagram illustrating a situation in which the road shape cannot be estimated well by the conventional technique. In FIG. 21, detection points are represented by circles. The detection points Ps2, Pt2, Pv2, and Pw2 are detection points that indicate the guard rail 401 that exists along the road. The detection point Pu is a detection point that indicates a house 600 that is located farther away from the guard rail 401 when viewed from the road side end. In such a situation, when the detection point is connected by the method according to Patent Document 2, the connection range At2 set based on the detection point Pt2 is indicated by a line segment Qs2 connecting the detection points Ps2 and Pt2. It is set according to the direction. At this time, as shown in FIG. 21, the connection range At2 may be rotated in a direction not along the running path. As a result, the detection point Pt2 indicating the guard rail 401 and the detection point Pu2 indicating the house 600 may be connected by the line segment Qt2. That is, the connection line may be calculated in a shape that does not follow the road shape. Furthermore, when the connection range Au2 is set based on the detection point Pu2 according to the direction indicated by the line segment Qt2, there is no detection point indicating the guard rail 401 inside the connection range Au2, and the connection line is interrupted. was there. As described above, in the conventional technique, the detection points cannot be properly connected, and it may be difficult to accurately estimate the shape of the road.

  The present invention has been made in view of the above problems, and an object of the present invention is to provide a road shape estimation device capable of accurately estimating a road shape.

  In order to solve the above problems, the present application adopts the following configuration. In other words, a first aspect of the present invention is a road shape estimation device that is mounted on a host vehicle and estimates the shape of the road on which the host vehicle travels, in the traveling direction of another vehicle that travels in front of the host vehicle. A traveling direction information calculating unit that calculates traveling direction information indicating a change, an object detecting unit that detects positional information of an object around the host vehicle, and the positional information of the roadside stationary object detected by the traveling direction information and the object detecting unit. A road shape estimation device comprising: a road shape estimation unit that estimates a shape of a road on which the host vehicle travels.

  According to a second aspect, in the first aspect, the object detection unit detects position information of roadside stationary objects around the host vehicle as a plurality of stationary object detection points, and the road shape estimation unit includes a plurality of stationary object detection points. Are sequentially connected based on the traveling direction information to form a connection line, and the shape of the road is estimated based on the connection line.

  According to a third aspect, in the second aspect, the road shape estimation unit includes a connection range setting unit that sets a connection range based on a stationary object detection point that is a connection source, and a stationary object detection point that exists within the connection range. A connection range setting unit including a line segment connection unit that connects a stationary object detection point as a connection destination and a stationary object detection point as a connection source with a line segment to form a connection line Is characterized in that the connection range is rotated and set according to the traveling direction information.

  According to a fourth aspect, in any one of the second and third aspects, the object detection unit detects the presence position of the other vehicle as the other vehicle detection point separately from the stationary object detection point, and the traveling direction information calculation unit is A road calculation unit that calculates a travel trajectory of the other vehicle based on the other vehicle detection point, and a trajectory approximation unit that approximates the travel trajectory of the other vehicle to a bent straight line that bends at the folding point. The shape estimation unit is characterized in that the stationary object detection point is connected using the direction indicated by the straight line constituting the bent straight line as the traveling direction information.

  According to a fifth aspect, in the fourth aspect, the bending approximate straight line is a nearby locus straight line existing on the vehicle vicinity side with respect to the bending point, and a distant side existing on the vehicle far side with reference to the bending point. The road shape estimation unit is composed of a trajectory straight line, and (A) when connecting a stationary object detection point that exists in the vicinity of the bending point when viewed from the own vehicle, the direction indicated by the vicinity trajectory line is used as the traveling direction information. (B) When a stationary object detection point existing far from the bending point as viewed from the host vehicle is connected, the direction indicated by the first far locus line is used as the traveling direction information.

  In a sixth aspect according to any one of the fourth and fifth aspects, the trajectory approximating unit calculates a near trajectory line and a far trajectory line by using a least square method for coordinates indicating a position of another vehicle detection point. It is characterized by.

  In a seventh aspect according to any one of the fourth to sixth aspects, the road shape estimation unit determines whether or not the travel locus of the other vehicle is highly likely along the road on which the host vehicle travels. (C) If it is determined that there is a high possibility that the travel locus of the other vehicle is along the road on which the host vehicle travels, a connection line is formed based on the traveling direction information, and (D) the travel of the other vehicle. When it is determined that there is a low possibility that the trajectory is along a road on which the host vehicle is traveling, the connection line is formed without using the traveling direction information.

  In an eighth aspect, in the seventh aspect, when the angle formed by the near locus line and the distant locus line is less than a predetermined angle threshold, the road shape estimation unit travels on the traveling locus of another vehicle. When it is determined that there is a high possibility that the vehicle is along the road, and the angle formed by the near locus line and the distant locus line is equal to or greater than a predetermined angle threshold, the traveling locus of the other vehicle is along the road on which the host vehicle is traveling. It is characterized that it is determined that there is a low possibility of being

  According to a ninth aspect, in the eighth aspect, the road shape estimation unit sets the angle threshold value to a smaller value as the absolute speed of the other vehicle increases.

  According to a tenth aspect, in any one of the seventh to ninth aspects, the road shape estimation unit determines that the traveling locus of the other vehicle is the own vehicle when the other vehicle detection point is updated within a predetermined time. If the other vehicle detection point is not updated within a predetermined time, the travel locus of the other vehicle is along the road on which the host vehicle is traveling. It is characterized that it is determined that there is a low possibility of being

  According to the first aspect, the shape of the road on which the host vehicle travels can be accurately estimated. It is considered that the traveling direction of other vehicles traveling in front of the host vehicle changes along the road on which the host vehicle travels. Therefore, by estimating the road shape based on the traveling direction of the other vehicle traveling in front of the host vehicle, the shape of the road on which the host vehicle is traveling can be estimated more accurately than in the past.

  According to the second aspect, the road shape can be estimated with a relatively small amount of processing by the process of sequentially connecting the detection points.

  According to the third aspect, the shape of the road can be accurately estimated by a simple process of rotating the connection range according to the traveling direction of the other vehicle.

  According to the fourth aspect, information indicating a change in the traveling direction of the other vehicle can be easily and accurately calculated by a simple process.

  According to the fifth aspect, it is possible to select appropriate information on the traveling direction of the other vehicle according to the position of the stationary object detection point to be connected. Therefore, the shape of the road can be estimated more accurately.

  According to the sixth aspect, a straight line indicating the traveling direction of the other vehicle can be calculated as a plausible line by approximating the trajectory of the other vehicle using a least square method.

  According to the seventh aspect, the road shape can be estimated with higher accuracy. When the traveling direction information of other vehicles not along the road of the own vehicle is used, there is a possibility that the road shape cannot be estimated accurately. Therefore, the road shape can be estimated more accurately by estimating the road shape using the traveling direction information of the other vehicle only when the travel locus of the other vehicle is likely to be along the road on which the host vehicle is traveling. Can be estimated.

  According to the eighth aspect, it is possible to determine with a simple process whether or not there is a high possibility that the travel locus of the other vehicle is along the road on which the host vehicle travels. When the traveling direction of the other vehicle changes extremely, there is a high possibility that the other vehicle is not traveling along the road. Therefore, when the angle of the broken line forming the trajectory line is relatively large, it can be easily determined that the travel trajectory of the other vehicle is unlikely to be along the road on which the host vehicle travels.

  According to the ninth aspect, it can be more accurately determined whether or not there is a high possibility that the travel locus of the other vehicle is along the road on which the host vehicle travels. When the other vehicle is moving at a relatively high speed, it is considered that there is a low possibility that an operation that greatly changes the traveling direction is performed. In other words, when the traveling direction is greatly changed even when the other vehicle is traveling at high speed, there is a possibility that normal traveling is not performed. Therefore, when the traveling direction is greatly changed even when the other vehicle is traveling at high speed, the road shape is estimated more accurately by estimating the road shape without using the traveling direction information of the other vehicle. be able to.

  According to the tenth aspect, it is possible to determine with a simple process whether or not there is a high possibility that the travel locus of the other vehicle is along the road on which the host vehicle travels. If the detection point of the other vehicle is not updated within the predetermined time, it is considered that the calculated trail information of the preceding vehicle is not accurate. That is, it is considered that there is a low possibility that the travel locus is along the road on which the host vehicle travels. Therefore, when the detection point of the other vehicle is not updated, it can be easily determined that there is a low possibility that the travel locus of the other vehicle is along the road on which the host vehicle travels.

FIG. 1 is a block diagram illustrating an example of a configuration of a road shape estimation device 1 according to an embodiment of the present invention. FIG. 2 is a diagram illustrating an example of detection points detected by the radar device 11. FIG. 3 is a block diagram illustrating an example of a functional configuration of the ECU 12. FIG. 4 is an example of a flowchart showing details of processing executed by each functional unit of the ECU 12. FIG. 5 is an example of a flowchart showing details of the forward vehicle trajectory calculation process executed by the trajectory calculation unit 122. FIG. 6 is a diagram illustrating a manner in which the travel locus of the preceding vehicle 200 is approximated to a bent straight line. FIG. 7 is an example of a flowchart showing details of the locus approximation process executed by the locus approximation unit 123. FIG. 8 is an example of a flowchart showing details of instantaneous inflection point calculation processing executed by the locus approximation unit 123. FIG. 9 is a diagram illustrating a region in which the other vehicle detection point is selected as an instantaneous inflection point candidate. FIG. 10 is a diagram showing parameters used for calculation of instantaneous inflection points. FIG. 11 is a diagram illustrating a state in which an instantaneous inflection point is calculated from the preceding vehicle detection point illustrated in FIG. FIG. 12 is an example of a flowchart showing details of the inflection point calculation process executed by the locus approximation unit 123. FIG. 13 is an example of a graph representing a function for calculating the angle threshold φth. FIG. 14 is an example of a flowchart showing details of the connection range setting process executed by the connection range setting unit 132. FIG. 15 is a diagram illustrating how the connection range is set. FIG. 16 is an example of a graph representing a function for calculating the vertical dimension Ly of the connection range. FIG. 17 is an example of a graph representing a function for calculating the horizontal dimension Lx of the connection range. FIG. 18 is an example of a graph representing a function for calculating the coefficient ε. FIG. 19 is an example of a graph representing a function for calculating the coefficient γ. FIG. 20 is a diagram illustrating a situation where the road shape is estimated well by the road shape estimation device 1 according to the present invention. FIG. 21 is a diagram illustrating a situation in which the road shape cannot be estimated well by the conventional technique.

  Hereinafter, a road shape estimation apparatus 1 according to an embodiment of the present invention will be described. The road shape estimation device 1 according to the present invention is a device that is mounted on the host vehicle 100 and estimates the shape of the road on which the host vehicle 100 travels. Hereinafter, the road on which the host vehicle 100 travels is referred to as the host vehicle traveling road.

  First, the hardware configuration of the road shape estimation apparatus 1 will be described with reference to FIG. FIG. 1 is a block diagram showing an example of the configuration of the road shape estimation apparatus 1 according to the embodiment of the present invention. As shown in FIG. 1, the road shape estimation device 1 includes a radar device 11 and an ECU 12. The ECU 12 is electrically connected to a vehicle control device 50 mounted on the host vehicle 100.

  For example, the radar device 11 is a device that detects the positions of objects existing around the host vehicle 100 as a plurality of detection points as shown in FIG. That is, the radar device 11 corresponds to the object detection unit described in the claims. FIG. 2 is a diagram illustrating an example of detection points detected by the radar apparatus 11. In the present embodiment, the radar device 11 is mounted on the front end of the host vehicle 100 and detects an object existing in front of the host vehicle 100. The radar device 11 is typically an FM-CW radar device that transmits and receives electromagnetic waves in the millimeter wavelength band. For example, the radar device 11 irradiates a detection wave signal such as an electromagnetic wave in front of the host vehicle 100. Then, the position of the reflection point is detected as a detection point based on the reflected wave of the detection wave signal reflected by the object. As shown in FIG. 2, the radar device 11 has the host vehicle 100 as the origin, the axis indicating the traveling direction of the host vehicle 100 as the Y axis, and the Y axis perpendicular to the Y axis and the horizontal plane as shown in FIG. The position information of the detection point is acquired in the XY coordinate system with the axis intersecting with X as the X axis. Hereinafter, the X coordinate of the detection point is denoted as Px, and the Y coordinate is denoted as Py.

  The radar device 11 also determines whether each detection point indicates a roadside stationary object, a preceding vehicle, or an oncoming vehicle. The radar device 11 may determine whether each detection point indicates a stationary object, a preceding vehicle, or an oncoming vehicle using a conventionally known method. Then, the radar device 11 transmits position information (XY coordinates) of the detection points to the ECU 12 together with flag data indicating which of the detection points indicates a roadside stationary object, a preceding vehicle, or an oncoming vehicle. In the following, detection points representing roadside stationary objects are referred to as stationary object detection points, detection points representing preceding vehicles are referred to as preceding vehicle detection points, and detection points representing oncoming vehicles are referred to as oncoming vehicle detection points. The preceding vehicle detection point and the oncoming vehicle detection point are collectively referred to as other vehicle detection points. In FIG. 2, Pd is shown from stationary object detection points Pa indicating the guardrail 300 which is a roadside stationary object. FIG. 2 also shows a preceding vehicle detection point P6 indicating the current position of the preceding vehicle 200, and preceding vehicle detection points P1 to P5 indicating the position of the previous preceding vehicle 200.

  The ECU 12 is typically an electronic control device that includes an information processing device such as a CPU (Central Processing Unit), a storage device such as a memory, an interface circuit, and the like. The ECU 12 is electrically connected to the vehicle control device 50.

  Hereinafter, the functional configuration of the ECU 12 will be described with reference to FIG. FIG. 3 is a block diagram illustrating an example of a functional configuration of the ECU 12. As shown in FIG. 3, the ECU 12 functionally includes a traveling direction information calculation unit 121 and a road shape estimation unit 131. The traveling direction information calculation unit 121 includes a locus calculation unit 122 and a locus approximation unit 123. The road shape estimation unit 131 includes a connection range setting unit 132 and a line segment connection unit 133. For example, the ECU 12 causes the CPU to execute a control program stored in advance in a memory or the like included in the ECU 12, thereby causing the traveling direction information calculation unit 121, the trajectory calculation unit 122, the trajectory approximation unit 123, the road shape estimation unit 131, the connection range. It operates as the setting unit 132 and the line segment connection unit 133.

  The traveling direction information calculation unit 121 is a functional unit that calculates traveling direction information indicating changes in the traveling direction of other vehicles (a preceding vehicle and an oncoming vehicle) traveling in front of the host vehicle 100. The trajectory calculation unit 122 is a functional unit that calculates a travel trajectory of another vehicle. The trajectory approximating unit 123 is a functional unit that approximates a traveling trajectory of another vehicle to a bent straight line that bends at an inflection point.

  The road shape estimation unit 131 is a functional unit that estimates the shape of the road on which the host vehicle travels based on the traveling direction information and the position information of the roadside stationary object detected by the radar device 11. The connection range setting unit 132 is a functional unit that sets a connection range with reference to a stationary object detection point that is a connection source. In addition, the connection range setting part 132 rotates and sets the said connection range according to the direction which the straight line which comprises a bending straight line shows. The line segment connecting unit 133 selects a stationary object detection point existing within the connection range as a connection destination, and connects the stationary object detection point as the connection destination and the stationary object detection point as the connection source by a line segment. It is a functional part that forms a connection line. The road shape estimation unit 131 estimates the shape of the vehicle traveling road based on the shape of the connection line.

  Returning to the description of FIG. 1, the vehicle control device 50 is a control device such as a brake control device, a steering control device, and an alarm device. Each of the brake control device, the steering control device, and the warning device, depending on the shape of the own vehicle traveling road acquired from the ECU 12, each of the traveling direction and traveling of the own vehicle 100 so that the own vehicle 100 does not deviate from the own vehicle traveling road. The speed is controlled, the deviation is predicted, and a warning is issued to the driver of the host vehicle 100. When the ECU 12 correctly detects the shape of the traveling road of the host vehicle, the vehicle control device 50 can execute appropriate vehicle control.

Next, a process executed by the ECU 12 will be described with reference to FIG. FIG. 4 is an example of a flowchart showing details of processing executed by each functional unit of the ECU 12. For example, the ECU 12 starts the processing of the flowchart of FIG. 4 when the IG power supply of the host vehicle 100 is set to the on state. When the process of the flowchart of FIG. 4 is started, the ECU 12 first executes the process of step S1. The ECU 12 repeatedly executes a series of processes from step S1 to step S12 shown below as one cycle.

  In step S <b> 1, the traveling direction information calculation unit 121 acquires information on other vehicle detection points from the radar device 11. Further, the road shape estimation unit 131 acquires information on stationary object detection points from the radar device 11. When the traveling direction information calculation unit 121 and the road shape estimation unit 131 complete the process of step S1, the process proceeds to step S2.

  In step S2, the trajectory calculation unit 122 performs a forward vehicle trajectory calculation process. The forward vehicle trajectory calculation process is a process of calculating the travel trajectories of the preceding vehicle and the oncoming vehicle that travel in front of the host vehicle 100. Hereinafter, the forward vehicle trajectory calculation process will be described with reference to FIG. FIG. 5 is an example of a flowchart showing details of the forward vehicle locus calculation process executed by the locus calculator 122. When the trajectory calculation unit 122 starts the processing of the flowchart of FIG. 5, first, the trajectory calculation unit 122 executes processing of step S <b> 21.

  In step S21, the trajectory calculation unit 122 determines whether a preceding vehicle has been detected. Specifically, it is determined whether or not position information about the preceding vehicle detection point is received from the radar device 11. The trajectory calculation unit 122 determines that the preceding vehicle has been detected when position information about the preceding vehicle detection point is received from the radar device 11, and advances the process to step S22. On the other hand, when it determines with the locus | trajectory calculation part 122 not detecting the preceding vehicle, a process is advanced to step S23.

  In step S22, the trajectory calculation unit 122 calculates a preceding vehicle trajectory. Specifically, the trajectory calculation unit 122 first stores the current position of the preceding vehicle detection point in the storage device of the ECU 12. Then, the locus calculation unit 122 maps the current position of the preceding vehicle detection point in the past to the XY coordinate system as shown in FIG. 2 together with the position information of the previous preceding vehicle detection point stored in the storage device of the ECU 12 in the past. . The trajectory calculation unit 122 calculates the trajectory line of the preceding vehicle by connecting the mapped preceding vehicle detection points with line segments in time series. When the trajectory calculation unit 122 completes the process of step S22, the process proceeds to step S23.

  In step S23, the trajectory calculation unit 122 determines whether an oncoming vehicle has been detected. Specifically, it is determined whether or not position information on the oncoming vehicle detection point is received from the radar device 11. The trajectory calculation unit 122 determines that an oncoming vehicle has been detected when position information about the oncoming vehicle detection point is received from the radar device 11, and advances the processing to step S24. On the other hand, when determining that the oncoming vehicle is not detected, the trajectory calculation unit 122 advances the process to step S3 in FIG.

  In step S24, the trajectory calculation unit 122 calculates an oncoming vehicle trajectory. Specifically, first, the locus calculation unit 122 stores the current position of the oncoming vehicle detection point in the storage device of the ECU 12. Then, the current position of the oncoming vehicle detection point is mapped to the XY coordinate system as shown in FIG. 2 together with the past position information of the oncoming vehicle detection point stored in the storage device of the ECU 12 in the past. The trajectory calculation unit 122 calculates the trajectory line of the oncoming vehicle by connecting the mapped oncoming vehicle detection points with line segments in time series. When the trajectory calculation unit 122 completes the process of step S24, the process proceeds to step S3 of FIG.

  In step S3, the trajectory approximating unit 123 executes trajectory approximating processing. The trajectory approximation process is a process for approximating the travel trajectory of the other vehicle calculated in the above-described forward vehicle trajectory calculation process to a bent straight line. As shown in FIG. 6, the bent straight line is a straight line bent at the folding pole HP. FIG. 6 is a diagram illustrating a manner in which the travel locus of the preceding vehicle 200 is approximated to a bent straight line. Details of the locus approximation process will be described below with reference to FIG. FIG. 7 is an example of a flowchart showing details of the locus approximation process executed by the locus approximation unit 123. When the trajectory approximating unit 123 starts the process of the flowchart of FIG. 7, first, the trajectory approximating unit 123 executes the process of step S31. In addition, when both the preceding vehicle and the oncoming vehicle are detected, the locus approximating unit 123 executes the following locus approximation process for each of the preceding vehicle locus and the oncoming vehicle locus.

  In step S31, the trajectory approximation unit 123 executes instantaneous inflection point calculation processing. The instantaneous inflection point calculation process is a process for calculating an instantaneous inflection point. The instantaneous inflection point is a detection point that will be closest to the bending point among the other vehicle detection points indicating the trajectory of the other vehicle at the present time. The instantaneous inflection point is calculated for each cycle of the process of FIG. Hereinafter, the instantaneous inflection point calculation process will be described with reference to FIG. FIG. 8 is an example of a flowchart showing details of instantaneous inflection point calculation processing executed by the locus approximation unit 123. When the trajectory approximating unit 123 starts the processing of the flowchart of FIG.

  In step S311, the trajectory approximating unit 123 selects one of the other vehicle detection points mapped in the above-described forward vehicle trajectory calculation process. Hereinafter, the other vehicle detection point selected in this step is referred to as a selected vehicle detection point. When the trajectory approximating unit 123 completes the process of step S311, the process proceeds to step S312.

  In step S312, the locus approximation unit 123 determines whether the selected vehicle detection point exists between the locus start point and the locus end point on the X axis. The locus start point is the other vehicle detection point detected the oldest. The trajectory end point is the most recently detected other vehicle detection point. If the trajectory approximation unit 123 determines that the selected vehicle detection point exists between the trajectory start point and the trajectory end point on the X axis, the process proceeds to step S313. On the other hand, when the trajectory approximating unit 123 determines that the other vehicle detection point does not exist between the trajectory start point and the trajectory end point on the X axis, the process proceeds to step S315.

  In step S313, the trajectory approximation unit 123 determines whether the selected vehicle detection point exists between the trajectory start point and the trajectory end point on the Y axis. If the trajectory approximating unit 123 determines that another vehicle detection point exists between the trajectory start point and the trajectory end point on the Y axis, the process proceeds to step S314. On the other hand, when the trajectory approximating unit 123 determines that the other vehicle detection point does not exist between the trajectory start point and the trajectory end point on the Y axis, the process proceeds to step S315.

  In step S314, the trajectory approximating unit 123 selects the selected vehicle detection point as an instantaneous inflection point candidate. When the locus approximating unit 123 completes the process of step S314, the locus approximating unit 123 proceeds with the process to step S316.

  In step S315, the locus approximation unit 123 excludes the selected vehicle detection point from the instantaneous inflection point candidates. When completing the process of step S315, the trajectory approximating unit 123 advances the process to step S316.

  In step S316, the trajectory approximation unit 123 determines whether all other vehicle detection points have been selected. If the trajectory approximating unit 123 determines that all other vehicle detection points have been selected, the process proceeds to step S317. On the other hand, when determining that there is another vehicle detection point that has not yet been selected, the trajectory approximation unit 123 returns the process to step S311.

  According to the processing from step S312 to step S316, the other vehicle detection point existing within the hatched area in FIG. 9 is selected as the instantaneous inflection point candidate, and the other vehicle detection point existing outside the area is instantaneously changed. Excluded from music point candidates. FIG. 9 is a diagram illustrating a region where the other vehicle detection point is selected as an instantaneous inflection point candidate. According to the processing from step S312 to step S315, other vehicle detection point information such as noise that cannot be an instantaneous inflection point in the processing of step S317 described later can be excluded from the calculation target. Therefore, the trajectory approximating unit 123 can calculate an instantaneous inflection point in a short time with an accurate and small processing amount.

  In step S317, the locus approximation unit 123 calculates an instantaneous inflection point based on the instantaneous inflection point candidate. Hereinafter, a method for calculating the instantaneous inflection point will be described with reference to FIG. In addition, FIG. 10 is a figure which shows the parameter used for calculation of an instantaneous inflection point.

  Specifically, the trajectory approximating unit 123 first sets any one point of the instantaneous inflection point candidates excluding the trajectory start point and the trajectory end point as a temporary instantaneous inflection point. FIG. 9 illustrates a case where P2 is set as a temporary instantaneous inflection point among instantaneous inflection point candidates P1 to P6. Next, the trajectory approximation unit 123 connects the trajectory start point P1 and the temporary instantaneous inflection point P2 with a straight line, and sets a first reference line LA # temp. Further, the trajectory approximating unit 123 sets the second reference line LB # temp by connecting the trajectory terminal point P6 and the temporary instantaneous inflection point P2 with a straight line. The trajectory approximating unit 123 then generates a first reference line LA # temp or a second reference line LB # temp from each of the instantaneous inflection point candidates excluding the trajectory start point P1, the trajectory end point P6, and the temporary instantaneous inflection point P2. Draw a perpendicular to the nearer one. In FIG. 10, a perpendicular line V3 is drawn from the instantaneous inflection point candidate P3, a perpendicular line V4 is drawn from the instantaneous inflection point candidate P4, and a perpendicular line V5 is drawn from the instantaneous inflection point candidate P5 to the first reference line LA # temp. . The trajectory approximating unit 123 calculates the sum of the lengths of the drawn perpendicular lines as a perpendicular sum ΣV, and stores the ΣV in the storage device of the ECU 12. The trajectory approximating unit 123 executes the above processing for all instantaneous inflection point candidates excluding the trajectory start point and trajectory end point, and calculates the perpendicular sum ΣV for each instantaneous inflection point. Then, the locus approximation unit 123 calculates the instantaneous inflection point candidate for which the perpendicular sum ΣV is the smallest value as the instantaneous inflection point. The trajectory approximating unit 123 stores the calculated instantaneous inflection point position in the storage device of the ECU 12.

  When the locus of the preceding vehicle 200 is calculated as shown in FIG. 10, the preceding vehicle detection point P4 is finally set as an instantaneous inflection point as shown in FIG. FIG. 11 is a diagram illustrating a state in which an instantaneous inflection point is calculated from the preceding vehicle detection point illustrated in FIG. When the trajectory approximating unit 123 completes the process of step S317, the trajectory approximating unit 123 advances the process to step S32 of FIG.

  In step S32, the trajectory approximating unit 123 executes an inflection point calculation process. The inflection point calculation process is a process of calculating the current inflection point based on the instantaneous inflection point and the past inflection point. Hereinafter, the inflection point calculation process will be described with reference to FIG. FIG. 12 is an example of a flowchart showing details of the inflection point calculation process executed by the locus approximation unit 123. When the trajectory approximating unit 123 starts the process of the flowchart of FIG. 12, first, the trajectory approximating unit 123 executes the process of step S321.

  In step S321, the trajectory approximation unit 123 determines whether an inflection point has been calculated in the previous cycle of the process illustrated in FIG. Specifically, it is determined whether or not data indicating an inflection point is stored in the storage device of the ECU 12. When the data indicating the inflection point is stored in the storage device, the trajectory approximation unit 123 determines that the inflection point has been calculated, and advances the process to step S322. On the other hand, when data indicating the previous inflection point is not stored, the trajectory approximation unit 123 determines that the inflection point has not been calculated, and advances the process to step S326.

  In step S322, the trajectory approximation unit 123 determines whether the previous inflection point is a part of the current trajectory line. If the trajectory approximating unit 123 determines that the previous inflection point is a part of the current trajectory line, the trajectory approximating unit 123 advances the process to step S323. On the other hand, when the trajectory approximation unit 123 determines that the previous inflection point is not a part of the current trajectory line, the trajectory approximation unit 123 proceeds with the process to step S325.

  In step S323, the trajectory approximation unit 123 determines whether the inflection point is the start point or the end point of the trajectory line. If the trajectory approximation unit 123 determines that the inflection point is the start point or the end point, the trajectory approximation unit 123 proceeds with the process to step S330. On the other hand, when the locus approximation unit 123 determines that the inflection point is not the start point or the end point, the process proceeds to step S324.

  In step S324, the trajectory approximation unit 123 sets the inflection point calculated in the previous cycle as the inflection point of the current cycle. When the trajectory approximating unit 123 completes the process of step S324, the process proceeds to step S33 of FIG.

  In step S325, the trajectory approximation unit 123 initializes inflection point data stored in the storage device of the ECU 12. When the trajectory approximating unit 123 completes the process of step S325, the trajectory approximating unit 123 advances the process to step S326.

  In step S326, the trajectory approximation unit 123 determines whether or not the instantaneous inflection point of the current cycle is the same as the previous instantaneous inflection point. If the trajectory approximating unit 123 determines that the instantaneous inflection point of the current cycle is the same as the previous instantaneous inflection point, the process proceeds to step S328. On the other hand, when the locus approximation unit 123 determines that the instantaneous inflection point of the current cycle is not the same as the previous instantaneous inflection point, the process proceeds to step S327.

  In step S327, the trajectory approximation unit 123 resets the counter N. The counter N is a value indicating that instantaneous inflection points are continuously calculated. The locus approximation unit 123 sets the value of the counter N stored in the storage device of the ECU 12 to an initial value (for example, zero). When the trajectory approximating unit 123 completes the process of step S327, the trajectory approximating unit 123 advances the process to step S328.

  In step S328, the locus approximation unit 123 counts up the counter N by a predetermined value (for example, 1). When the locus approximating unit 123 completes the process of step S328, the locus approximating unit 123 proceeds with the process to step S329.

  In step S329, the trajectory approximation unit 123 determines whether or not the counter N is equal to or greater than the counter threshold value Nth. The counter threshold Nth is a constant value (for example, 4) stored in advance in the storage device of the ECU 12. If the trajectory approximating unit 123 determines that the counter N is greater than or equal to the counter threshold value Nth, the trajectory approximating unit 123 advances the process to step S330. On the other hand, when the trajectory approximating unit 123 determines that the counter N is less than the counter threshold Nth, the process proceeds to step S33 in FIG.

  In step S330, the locus approximation unit 123 sets the instantaneous inflection point as the inflection point of the current cycle. When the trajectory approximation unit 123 completes the process of step S330, the trajectory approximation unit 123 proceeds with the process to step S33 of FIG.

  According to the above inflection point calculation process, when an inflection point calculated in the past is part of the current trajectory line, the position of the past inflection point continues as the current inflection point position. Retained. If no inflection point has been calculated in the past, when the same instantaneous inflection point is continuously detected, the instantaneous inflection point is set as the inflection point. Such a process can prevent the position of the inflection point from fluctuating greatly every cycle, so that a stable track shape recognition result can be obtained.

  In step S33, the trajectory approximating unit 123 calculates the trajectory straight line LA before bending and the trajectory straight line LB after bending using the least square method. The pre-bending trajectory straight line LA is a straight line existing on the vicinity side of the host vehicle 100 with reference to the inflection point among straight lines constituting the bent straight line (see FIG. 6). The post-bending locus straight line LB is a straight line existing on the far side of the host vehicle 100 with reference to the bending point among the straight lines constituting the bending straight line (see FIG. 6).

なお、添え字のｋは各検出点を示すものとする。同様に、軌跡近似部１２３は、変曲点から軌跡終端点までの軌跡線を構成する他車両検出点を直線近似することによって折曲前軌跡直線ＬＢを算出する。具体的には、軌跡近似部１２３は、変曲点から軌跡終端点までの軌跡線を構成する他車両検出点について上記数式１から数式３を適用することによって折曲前軌跡直線ＬＢを算出する。軌跡近似部１２３は、ステップＳ３３の処理を完了すると、処理をステップＳ３４へ進める。The trajectory approximating unit 123 calculates the trajectory straight line LA before bending by linearly approximating the other vehicle detection points constituting the trajectory line from the trajectory start point to the inflection point. Specifically, when the expression representing the trajectory straight line LA before bending in the XY coordinate system as shown in FIG. 6 is Expression 1, the parameters A and B shown in Expression 1 are expressed by Expression 2 and Expression based on the least square method. It is represented by 3. Note that N is the total number of other vehicle detection points that form a trajectory line from the trajectory start point to the inflection point.
(Equation 1)
y = Ax + B

Note that the subscript k indicates each detection point. Similarly, the trajectory approximating unit 123 calculates the trajectory straight line LB before bending by linearly approximating the other vehicle detection points constituting the trajectory line from the inflection point to the trajectory end point. Specifically, the trajectory approximating unit 123 calculates the trajectory straight line LB before bending by applying the above formulas 1 to 3 to the other vehicle detection points constituting the trajectory line from the inflection point to the trajectory end point. . When the trajectory approximating unit 123 completes the process of step S33, the process proceeds to step S34.

  According to the process of step S33, a straight line indicating the traveling direction of the other vehicle can be calculated as being plausible by linearly approximating the trajectory of the other vehicle using the least square method.

  In step S34, the trajectory approximating unit 123 calculates the near straight line angle α and the far straight line angle β. The near straight line angle α is an angle formed by the trajectory straight line LA before bending and the Y axis (see FIG. 6). The far linear angle β is an angle formed by the trajectory straight line LB after bending and the Y axis (see FIG. 6). Specifically, the trajectory approximation unit 123 calculates the near straight line angle α and the far straight line angle β based on the expressions of the pre-bending trajectory straight line LA and the post-bending trajectory straight line LB obtained in the process of step S33. When the trajectory approximation unit 123 completes the process of step S34, the process proceeds to step S35.

In step S35, the trajectory approximating unit 123 calculates the angle difference φ. The angle difference φ is a difference value between the near straight line angle α and the far straight line angle β, and is calculated by Equation 4.
(Equation 4)
φ = | α−β |
When the trajectory approximating unit 123 completes the process of step S35, the process proceeds to step S36.

  In step S36, the trajectory approximation unit 123 calculates an angle threshold value φth. The angle threshold φth is a threshold for determining whether or not to calculate a bending point. Specifically, the trajectory approximating unit 123 calculates the angle threshold value φth based on the absolute speed of the other vehicle and a predetermined function as shown in FIG. FIG. 13 is an example of a graph representing a function for calculating the angle threshold φth. In FIG. 13, the horizontal axis represents the absolute speed of the other vehicle, and the vertical axis represents the angle threshold value φth. It is more preferable that the locus approximation unit 123 uses a different function depending on whether the other vehicle is a preceding vehicle or an oncoming vehicle. More specifically, it is preferable that the trajectory approximating unit 123 calculates the angle threshold value φth to be larger in the low speed region when the other vehicle is a preceding vehicle than when the other vehicle is an oncoming vehicle. When the process of step S36 is completed, the locus approximation unit 123 advances the process to step S37.

  In step S37, the trajectory approximation unit 123 determines whether the angle difference φ is less than the angle threshold φth. If the trajectory approximating unit 123 determines that the angle difference φ is less than the angle threshold φth, the process proceeds to step S38. On the other hand, when the trajectory approximating unit 123 determines that the angle difference φ is equal to or greater than the angle threshold φth, the process proceeds to step S39.

  In step S38, the trajectory approximating unit 123 calculates an intersection point between the pre-bending straight line LA and the post-bending straight line LB as a bending point. When the process of step S38 is completed, the locus approximation unit 123 advances the process to step S4 in FIG.

  In step S <b> 39, the trajectory approximating unit 123 initializes bending point data stored in the ECU 12. When the process of step S38 is completed, the locus approximation unit 123 advances the process to step S4 in FIG.

  According to the processing from step S34 to step S39, the bending point is not calculated when the angle difference φ is an extremely large value. When the angle difference φ is an extremely large value, it is considered that there is a high possibility that the other vehicle is not moving along the road, for example, by turning right or left. Therefore, in such a case, it is possible to recognize the shape of the traveling road of the own vehicle with high accuracy by not calculating the bending point and not using the traveling information of other vehicles in the processing described later.

  Further, the road shape can be estimated with higher accuracy by changing the angle threshold value φth according to the speed at which the other vehicle is traveling. Specifically, when the other vehicle is moving at a relatively high speed, it is considered unlikely that an operation that greatly changes the traveling direction is performed. In other words, when the traveling direction is greatly changed even when the other vehicle is traveling at high speed, there is a possibility that normal traveling is not performed. Therefore, when the traveling direction is greatly changed even when the other vehicle is traveling at high speed, the road shape is estimated without calculating the bending point and using the traveling direction information of the other vehicle. The road shape can be estimated more accurately.

  In step S4, the road shape estimation unit 131 calculates a connection start point. The road shape estimation unit 131 calculates, for example, a stationary object detection point present at a position closest to the host vehicle 100 in the Y-axis direction as a connection start point. The road shape estimation unit 131 may calculate the connection start point using any conventionally known method. More preferably, the road shape estimation unit 131 calculates the connection start point using the method described in PCT / JP2010 / 006489. When the road shape estimating unit 131 completes the process of step S1, the process proceeds to step S5.

  In step S5, the road shape estimation unit 131 sets the connection start point as the connection source. When the road shape estimating unit 131 completes the process of step S5, the process proceeds to step S6.

In step S6, the connection range setting unit 132 performs a connection range setting process. The connection range setting process is a process for setting a connection range. Hereinafter, the connection range setting process executed by the connection range setting unit 132 will be described with reference to FIG. FIG. 14 is an example of a flowchart showing details of the connection range setting process executed by the connection range setting unit 132. When the connection range setting unit 132 starts the process of the flowchart of FIG. 14 , first, the connection range setting part 132 executes the process of step S61.

  In step S61, the connection range setting unit 132 calculates the size of the connection range. As shown in FIG. 15, the connection range is, for example, a rectangular area that is set based on the connection source. FIG. 15 is a diagram illustrating how the connection range is set.

  The connection range setting unit 132 calculates the vertical dimension Ly such that the vertical dimension Ly of the connection range increases as the distance from the host vehicle 100 to the detection point that is the connection source increases. More specifically, the connection range setting unit 132 calculates the vertical dimension Ly based on the Y coordinate py of the connection source and a predetermined function as shown in FIG. FIG. 16 is an example of a graph representing a function for calculating the vertical dimension Ly of the connection range. In FIG. 16, the horizontal axis indicates the Y coordinate py of the detection point that is the connection source, and the vertical axis indicates the vertical dimension Ly.

  Similarly, the connection range setting unit 132 calculates the lateral dimension Lx so that the lateral dimension Lx increases as the distance from the host vehicle 100 to the detection point that is the connection source increases. More specifically, the connection range setting unit 132 calculates the lateral dimension Lx based on the Y coordinate py of the connection source and a predetermined function as shown in FIG. FIG. 17 is an example of a graph representing a function for calculating the lateral dimension Lx of the connection range. In FIG. 17, the horizontal axis indicates the Y coordinate py of the detection point that is the connection source, and the vertical axis indicates the vertical dimension Lx. When the connection range setting unit 132 completes the process of step S61, the process proceeds to step S62.

  In step S62, the connection range setting unit 132 determines whether an inflection point exists. Specifically, it is determined whether data indicating the inflection point is stored in the storage device of the ECU 12. If the connection range setting unit 132 determines that an inflection point exists, the process proceeds to step S63. On the other hand, if the connection range setting unit 132 determines that no inflection point exists, the process proceeds to step S72.

  In step S63, the connection range setting unit 132 determines whether the other vehicle trajectory has been updated within the past T time. For example, the connection range setting unit 132 determines whether or not the trajectory has been updated depending on whether or not the position and number of other vehicle detection points constituting the trajectory line calculated in the above-described forward vehicle trajectory calculation process have changed within T time. Determine whether. If the connection range setting unit 132 determines that the vehicle trajectory has been updated within the past T time, the process proceeds to step S64. On the other hand, if the connection range setting unit 132 determines that the vehicle trajectory has not been updated within the past T time, the process proceeds to step S72.

  In step S64, the connection range setting unit 132 determines whether or not a preceding vehicle is detected in the same manner as in step S21 described above. If the connection range setting unit 132 determines that the preceding vehicle is detected, the process proceeds to step S65. On the other hand, if the connection range setting unit 132 determines that the preceding vehicle is not detected, the process proceeds to step S68.

  In step S65, the connection range setting unit 132 determines whether or not the connection source exists on the vicinity side of the host vehicle 100 with the inflection point as a reference. If the connection range setting unit 132 determines that the connection source exists on the vicinity side of the host vehicle 100 with the inflection point as a reference, the connection range setting unit 132 proceeds with the process to step S66. On the other hand, if the connection range setting unit 132 determines that the connection source exists on the far side of the host vehicle 100 with the inflection point as a reference, the process proceeds to step S67.

  In step S66, the connection range setting unit 132 calculates the connection range rotation angle θ based on the vicinity locus line angle α of the preceding vehicle. The connection range rotation angle θ is a rotation amount used when rotating the connection range. The connection range setting unit 132 calculates the connection range rotation angle θ according to the direction indicated by the line segment connected to the connection source. For example, in FIG. 15, it is assumed that θc is calculated as the connection range rotation angle θ when the detection point Pc is the connection source and Ac is set as the connection range.

The detection point Pc is connected in advance to the detection point Pb via the line segment Qb by the process of step S9 described later. First, the connection range setting unit 132 determines the angle ωbc formed by the line segment Qb and the Y axis based on the coordinates (cx, cy) of the detection point Pc, the coordinates (bx, by) of the detection point Pb, and Equation 5. calculate.
(Equation 5)
ωbc = arctan {(cx−bx) / (cy−by)}
Next, assuming that the connection range rotation angle θ calculated when the connection range Ab is set based on the detection point Pb last time is θb, the connection range setting unit 132 calculates ψ based on ωbc, θb, and Equation 6. calculate.
(Equation 6)
ψ = ωbc × ε + (1−ε) × θb
Note that the coefficient ε in Equation 2 is a coefficient of 0 or more and 1 or less. The connection range setting unit 132 calculates the value of the coefficient ε as a smaller value as the number of detection points connected to reach the connection source (hereinafter referred to as the connection hierarchy number J) increases. More specifically, the connection range setting unit 132 calculates the value of the coefficient ε based on the connection hierarchy number J and the function shown in FIG. FIG. 18 is an example of a graph representing a function for calculating the coefficient ε. In FIG. 18, the horizontal axis indicates the Y coordinate py of the connection source, and the vertical axis indicates the vertical dimension Ly. When the connection source is the start point, the connection range setting unit 132 calculates the connection range rotation angle θ (θb in the above example) as 0.
Next, the connection range setting unit 132 calculates the connection range rotation angle θ based on the ψ obtained above, the neighborhood locus line angle α, and Equation 7.
(Equation 7)
θ = ψ × γ + (1-γ) × α
The connection range setting unit 132 calculates the coefficient γ of Expression 7 based on the function shown in FIG. FIG. 19 is an example of a graph representing a function for calculating the coefficient γ. In FIG. 19, the vertical axis represents the coefficient γ, and the horizontal axis represents the detection point distance ΔPy. The detection point distance ΔPy represents the distance in the Y-axis direction from the connection source to the connection source destination (see FIG. 15). When the connection range setting unit 132 completes the process of step S66, the process proceeds to step S73.

In step S67, the connection range setting unit 132 calculates the connection range rotation angle θ based on the far locus line angle β of the preceding vehicle. Specifically, after calculating ψ in the same manner as in step S66, the connection range rotation angle θ is calculated using Equation 8 instead of Equation 7.
(Equation 8)
θ = ψ × γ + (1-γ) × β
When the connection range setting unit 132 completes the process of step S67, the process proceeds to step S73.

  In step S68, the connection range setting unit 132 determines whether an oncoming vehicle is detected as in step S23. If the connection range setting unit 132 determines that an oncoming vehicle is detected, the process proceeds to step S69. On the other hand, if the connection range setting unit 132 determines that no oncoming vehicle is detected, the process proceeds to step S72.

  In step S <b> 69, the connection range setting unit 132 determines whether or not the connection source exists on the vicinity side of the host vehicle 100 with the inflection point as a reference. If the connection range setting unit 132 determines that the connection source is present in the vicinity of the host vehicle 100 with the inflection point as a reference, the connection range setting unit 132 proceeds with the process to step S70. On the other hand, if the connection range setting unit 132 determines that the connection source exists on the far side of the host vehicle 100 with the inflection point as a reference, the process proceeds to step S71.

  In step S <b> 70, the connection range setting unit 132 calculates the connection range rotation angle θ based on the near-track line angle α of the oncoming vehicle. Specifically, the connection range setting unit 132 calculates the connection range rotation angle θ using the near locus line angle α of the oncoming vehicle instead of the near locus line angle α of the preceding vehicle in the process of step S66. When the connection range setting unit 132 completes the process of step S70, the process proceeds to step S73.

  In step S71, the connection range setting unit 132 calculates the connection range rotation angle θ based on the far locus line angle β of the oncoming vehicle. Specifically, the connection range setting unit 132 calculates the connection range rotation angle θ using the far locus line angle β of the oncoming vehicle in place of the far locus line angle β of the preceding vehicle in the process of step S67. The connection range setting part 132 will advance a process to step S73, if the process of step S71 is completed.

  In step S72, the connection range setting unit 132 calculates the connection range rotation angle θ without using the locus line angle. Specifically, after setting the value of the coefficient γ to 1, the connection range rotation angle θ is calculated in the same manner as the method described in step S66 and step S67 described above. When the connection range setting unit 132 completes the process of step S72, the process proceeds to step S73.

  In step S73, the connection range setting unit 132 sets a connection range. When the connection range setting unit 132 completes the process of step S73, the process proceeds to step S7 of FIG.

  As described above, according to the connection range setting process, the connection range rotation angle θ is calculated based on the nearby locus line angle α when the detection point that is the connection source exists in the region RA of FIG. When the detection point that is the connection source exists in the region RB of FIG. 6, the connection range rotation angle θ is calculated based on the near locus line angle β. Further, when the traveling direction information of both the preceding vehicle and the oncoming vehicle is detected, the traveling direction information of the preceding vehicle is preferentially used.

  According to such a connection range setting process, it is possible to appropriately set the connection range in consideration of both the direction indicated by the previous connection line and the traveling direction of the other vehicle. In addition, when the bending point is not calculated or when the trajectory of the other vehicle is not updated within a predetermined time, the connection range is set without using the near trajectory line angle α and the far trajectory line angle β. . That is, when it is unlikely that the travel locus of the other vehicle is along the road on which the host vehicle travels, the shape of the host vehicle traveling road is estimated without using information on the traveling direction of the other vehicle. . Therefore, the shape of the traveling road can be estimated with high accuracy.

  Returning to the description of FIG. 4, in step S <b> 7, the line segment connecting unit 133 determines whether or not a detection point exists within the connection range. If the line segment connecting unit 133 determines that a detection point exists within the connection range, the line segment connecting unit 133 proceeds with the process to step S8. On the other hand, if the line segment connecting unit 133 determines that the detection point does not exist within the connection range, the process proceeds to step S11.

  In step S8, the line segment connecting unit 133 selects a detection point having the shortest distance to the connection source among the detection points existing within the connection range as the connection destination. When completing the process of step S8, the line segment connecting unit 133 advances the process to step S9.

  In step S9, the line segment connecting unit 133 connects the connection source and the connection destination with a line segment. The line segment connecting unit 133 may connect the connection source and the connection destination with a curve. When the ECU 12 completes the process of step S9, the process proceeds to step S10.

  In step S10, the line segment connecting unit 133 sets the connection destination connected in step S8 as a new connection source. When completing the process of step S9, the line segment connecting unit 133 returns the process to step S6.

  As described above, according to the processing from step S6 to step S10, the connection range can be appropriately set according to the change in the traveling direction of the other vehicle. In various situations, such as when the roadside stationary object is intermittently disposed, a connection line along the own vehicle traveling road can be suitably obtained. That is, the road shape can be estimated more accurately than in the past. Moreover, the shape of the road can be estimated with a relatively small amount of processing by a simple process of sequentially connecting stationary object detection points.

In step S11, the ECU 12 outputs the connection line as a running road shape, and then in step S12, the ECU 12 determines whether or not the IG power source of the host vehicle 100 is set to OFF. When the ECU 12 determines that the IG power source of the host vehicle 100 is set to OFF, the ECU 12 ends the process of the flowchart of FIG. On the other hand, if the ECU 12 determines that the IG power supply of the host vehicle 100 is maintained in the ON state, the ECU 12 returns the process to step S1 and repeats the process of each step described above.

  The manner in which the shape of the traveling road of the vehicle is well estimated by the road shape estimation device 1 described above will be described with reference to FIG. In addition, FIG. 20 is a figure which illustrates the condition where a road shape is estimated favorably by the road shape estimation apparatus 1 which concerns on this invention. In FIG. 20, following the preceding vehicle 200, the host vehicle 100 is traveling along a gentle curve. In addition, the guardrail 400 is intermittently arrange | positioned so that the own vehicle travel road may follow a curve.

  First, the road shape estimation device 1 calculates information indicating a change in the traveling direction of the preceding vehicle 200 based on the preceding vehicle detection points P11 to P14 (steps S1 to S4 described above). Next, the stationary object detection point indicating the guard rail 400 is connected according to the traveling direction information of the preceding vehicle 200 (from step S5 to step S11 described above). Here, when setting the connection range At1 with the stationary object detection point Pt1 as a reference, the road shape estimation device 1 rotates the connection range At1 according to the advance direction of the preceding vehicle 200 calculated in advance (steps described above). S6). Thus, the road shape estimation device 1 rotates the connection range At1 in consideration of not only the direction indicated by the previously connected line segment Qs1, but also the traveling direction of the preceding vehicle 200. Therefore, for example, even in a location where the guard rail 400 is interrupted as shown in FIG. 20, the road shape estimation device 1 can appropriately set a connection range and form a connection line along the actual road shape.

  As described above, according to the road shape estimation apparatus 1 according to the embodiment of the present invention, the shape of the own vehicle traveling road can be accurately calculated according to the change in the traveling direction of the other vehicle.

  In the above embodiment, an example in which the road shape estimation device 1 rotates the connection range according to the traveling direction information of the other vehicle has been described. However, the road shape estimation device 1 estimates the road shape according to the traveling direction information. If possible, the above process may be modified into various modes. For example, the road shape estimation device 1 may change the size of the connection range according to the traveling direction information.

  In the above embodiment, the example in which the road shape estimation device 1 estimates the road shape using the travel locus of the preceding vehicle has been described. However, the present invention is based on the traveling of the preceding vehicle that travels further ahead of the preceding vehicle. You may apply so that a road shape may be estimated using a locus | trajectory.

  The road shape estimation apparatus according to the present invention is useful as a road shape estimation apparatus that can accurately estimate the road shape.

1 Road shape estimation device 11 Radar device 12 ECU
DESCRIPTION OF SYMBOLS 50 Vehicle control apparatus 121 Travel direction information calculation part 122 Trajectory calculation part 123 Trajectory approximation part 131 Road shape estimation part 132 Connection range setting part 133 Line segment connection part 100 Own vehicle 200 Predecessor vehicle 300, 400 Guardrail 600 House

Claims (8)

A road shape estimation device that is mounted on a host vehicle and estimates a shape of a road on which the host vehicle travels,
A traveling direction information calculating unit that calculates traveling direction information indicating a direction in which another vehicle traveling in front of the host vehicle travels;
An object detection unit for detecting position information of roadside stationary objects around the vehicle as a plurality of stationary object detection points;
A road shape estimation unit that forms a connection line by sequentially connecting the plurality of stationary object detection points based on the traveling direction information, and estimates a shape of a road on which the host vehicle travels based on the connection line; Prepared,
The road shape estimation unit
A connection range setting unit for setting a connection range on the basis of the stationary object detection point as a connection source;
The stationary object detection point existing within the connection range is selected as a connection destination, and the stationary object detection point as the connection destination and the stationary object detection point as the connection source are connected by a line segment to form a connection line A line segment connecting part,
The said connection range setting part rotates and sets the said connection range according to the said advancing direction information, The road shape estimation apparatus characterized by the above-mentioned. The object detection unit detects the presence position of the other vehicle as an other vehicle detection point in distinction from the stationary object detection point;
The traveling direction information calculation unit includes:
A trajectory calculation unit that calculates a travel trajectory of the other vehicle based on the other vehicle detection point;
A trajectory approximating unit approximating a travel trajectory of the other vehicle to a bent straight line that bends at a folding point;
The road shape estimation device according to claim 1, wherein the road shape estimation unit connects the stationary object detection points using a direction indicated by a straight line constituting the bent straight line as the traveling direction information. . The bending approximate straight line consists of a near locus straight line existing on the vehicle vicinity side with respect to the bending point and a distant locus straight line existing on the vehicle far side with reference to the bending point,
The road shape estimation unit (A), when connecting the stationary object detection point that is present in the vicinity of the bending point when viewed from the host vehicle, uses the direction indicated by the vicinity locus line as the traveling direction information, (B) When connecting the stationary object detection point that exists farther from the bending point when viewed from the host vehicle, the direction indicated by the far-distance locus line is used as the traveling direction information. Item 5. The road shape estimation apparatus according to Item 4.   The said locus | trajectory approximation part calculates the said vicinity locus line and the said far locus line by using the least squares method about the coordinate which shows the position of the said other vehicle detection point, Either of Claim 4 and 5 characterized by the above-mentioned. The road shape estimation apparatus according to item 1.   The road shape estimation unit determines whether or not the travel locus of the other vehicle is likely along the road on which the host vehicle travels, and (C) the travel locus of the other vehicle is the host vehicle. If it is determined that the vehicle is likely to be along a road on which the vehicle travels, the connection line is formed based on the traveling direction information, and (D) the road on which the host vehicle travels is a travel locus of the other vehicle. The road shape according to any one of claims 4 to 6, wherein the connection line is formed without using the traveling direction information when it is determined that the possibility of being along the road is low. Estimating device.   The road shape estimation unit is configured such that when the angle formed by the near locus line and the far locus line is less than a predetermined angle threshold, the traveling locus of the other vehicle is along the road on which the host vehicle is traveling. When it is determined that the possibility is high, and the angle formed by the near locus line and the far locus line is equal to or greater than a predetermined angle threshold, the traveling locus of the other vehicle is along the road on which the host vehicle is traveling. The road shape estimation apparatus according to claim 7, wherein it is determined that there is a low possibility.   The road shape estimation apparatus according to claim 8, wherein the road shape estimation unit sets the angle threshold value to a smaller value as the absolute speed of the other vehicle increases.   When the other vehicle detection point is updated within a predetermined time, the road shape estimation unit is likely to have a travel locus of the other vehicle along a road on which the host vehicle travels. If the other vehicle detection point is not updated within a predetermined time, it is determined that there is a low possibility that the travel locus of the other vehicle is along the road on which the host vehicle travels. The road shape estimation device according to any one of claims 7 to 9, wherein

Images