CN107943066B - Method for supervising and controlling obstacle avoidance of unmanned aerial vehicle by using human - Google Patents
Method for supervising and controlling obstacle avoidance of unmanned aerial vehicle by using human Download PDFInfo
- Publication number
- CN107943066B CN107943066B CN201710553624.XA CN201710553624A CN107943066B CN 107943066 B CN107943066 B CN 107943066B CN 201710553624 A CN201710553624 A CN 201710553624A CN 107943066 B CN107943066 B CN 107943066B
- Authority
- CN
- China
- Prior art keywords
- unmanned aerial
- aerial vehicle
- obstacle
- action
- obstacle avoidance
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 26
- 230000009471 action Effects 0.000 claims abstract description 167
- 230000007613 environmental effect Effects 0.000 claims abstract description 35
- 238000004891 communication Methods 0.000 claims abstract description 30
- 230000000875 corresponding effect Effects 0.000 claims description 27
- 230000003993 interaction Effects 0.000 claims description 16
- 230000008569 process Effects 0.000 claims description 13
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 3
- 230000035945 sensitivity Effects 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 3
- 238000011084 recovery Methods 0.000 abstract description 2
- 238000007726 management method Methods 0.000 description 24
- 238000010586 diagram Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 230000004888 barrier function Effects 0.000 description 2
- 230000008447 perception Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000019771 cognition Effects 0.000 description 1
- 230000001149 cognitive effect Effects 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000004886 process control Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/10—Simultaneous control of position or course in three dimensions
- G05D1/101—Simultaneous control of position or course in three dimensions specially adapted for aircraft
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention provides a method for supervising and controlling obstacle avoidance of an unmanned aerial vehicle by an unmanned aerial vehicle, which relates to the field of unmanned aerial vehicles, and the invention defines values of environmental factors, obstacle avoidance actions and a supervising and controlling mode, wherein the unmanned aerial vehicle carries out reasoning and judgment on selectable obstacle avoidance actions and carries out judgment according to a judgment result so as to avoid obstacles; for unknown obstacles, the unmanned aerial vehicle must complete evasion under the guidance of a human machine, and the variable autonomous supervision control method fully exerts the autonomous execution capability of the unmanned aerial vehicle and the analysis and judgment capability of a human-machine operator; for extreme conditions such as communication interruption, the unmanned aerial vehicle can guarantee the safety of the unmanned aerial vehicle while waiting for communication recovery by adjusting the supervision control mode, and the variable autonomous supervision control mode has better strain capacity for different obstacle types and environmental conditions by combining the characteristics of the unmanned aerial vehicle and the human-machine.
Description
Technical Field
The invention relates to the field of unmanned aerial vehicles, in particular to an avoidance judging method.
Background
When the unmanned aerial vehicle marchs the in-process and faces the barrier, there is the interaction of man-machine operator and unmanned aerial vehicle, guarantees that unmanned aerial vehicle accomplishes the barrier and avoids. Supervisory control means that there is man-machine to interact with unmanned aerial vehicle intermittently, receives feedback and gives the instruction in order to carry out process control to unmanned aerial vehicle in the task environment.
However, in the practical application of obstacle avoidance of the unmanned aerial vehicle system, the traditional method generally adopts manual operation of an operator to carry Out obstacle avoidance or adopts a path planning, rule or automatic machine mode to carry Out obstacle avoidance by calculating a cost function, so that two problems occur, namely an 'overload' phenomenon of the operator caused by the complexity of the environment ①, an 'Out-of-the-Loop' (OOTL) phenomenon that the operator loses the perception capability of the surrounding environment situation caused by the fact that the unmanned aerial vehicle system is in an overhigh autonomous authority, ②, and the two problems show that the supervision and control of the obstacle avoidance of the unmanned aerial vehicle by a human machine need to be dynamically adjusted according to the actual situation.
Therefore, it is very necessary to research an obstacle avoidance supervision control method for an unmanned aerial vehicle, which can sufficiently exert the high-level cognitive decision-making capability of human-computer operators and the efficient task execution capability of the unmanned aerial vehicle, and simultaneously maintain the situation cognition of the human-computer operators on the environment and the proper workload.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention provides an obstacle avoidance supervision control method, which can complement the advantages of human and machine, can keep the perception capability of an operator on the environment situation and proper workload, and avoids two problems in supervision control, thereby effectively achieving the purpose of obstacle avoidance.
The invention provides an obstacle avoidance supervision control method, which is a method for adjusting different supervision control modes by an unmanned aerial vehicle according to environmental changes, and completing corresponding obstacle avoidance actions by the unmanned aerial vehicle so as to achieve the purpose of obstacle avoidance.
The technical scheme adopted by the invention for solving the technical problem comprises the following steps:
step 1: defining environmental factors, obstacle avoidance actions and supervision control modes
When the unmanned aerial vehicle encounters an obstacle, 4 obstacle avoidance actions are provided, including an autonomous obstacle avoidance action A1, a deceleration waiting instruction action A2, a hover waiting instruction action A3 and a return action A4;
wherein the autonomous obstacle avoiding action a1 is defined as a high-level action; the deceleration waiting instruction a2 and the hover waiting instruction A3 are intermediate-level actions; return a4 to a low level action;
the supervision control mode is 3 modes, and the autonomous capacity of the unmanned aerial vehicle is respectively an exception management mode L3, an agreement management mode L2 and a command control mode L1 from high to low; the initial mode of the unmanned aerial vehicle is an exception management mode L3;
the unmanned aerial vehicle supervision control mode is defined as shown in table 1;
TABLE 1 three supervisory control modes and their corresponding actions
The action execution mode refers to a mode that the unmanned aerial vehicle completes corresponding obstacle avoidance actions under the supervision control of the unmanned aerial vehicle, and is shown in table 2:
TABLE 2 three supervisory control modes and their corresponding action execution modes
1) The environmental factors include 3 aspects of the distance between obstacles, whether the obstacle information is known, and whether the current communication condition is good, which are defined as follows:
① definition of obstacle distance, including obstacle near V3 and obstacle far V5;
with respect to the drone, the position of the obstacle may be expressed as
In the formula (1), P (t) is the position of the obstacle i at the time t, t is time, P (0) is the initial position of the obstacle i, viIs the velocity, v, of the obstacle iu(τ) is the velocity of drone u;
in the formula (2)
For the yaw angle of the drone, the distance between the drone and the obstacle is defined by the time T to reach the obstacle position by the following formula:
② whether the obstacle is known to be divided into a known obstacle V1 and an unknown obstacle V4, the known obstacle being a known obstacle in which both image information and position information of the obstacle are known in the working area, and the unknown obstacle being an obstacle in which at least one of the image information and the position information of the obstacle is unknown in the working area;
③, dividing the current communication status into normal communication and interruption V2, wherein the signal intensity of the interruption means that the signal reaches the unmanned aerial vehicle receiver is less than the sensitivity of the unmanned aerial vehicle receiver, which causes the situation that the normal communication between the unmanned aerial vehicle and the manned aerial vehicle can not be realized;
2) the value range of the values of the environmental factors and the obstacle avoidance actions is {0,1}, wherein 1 represents that the state is effective, and 0 represents that the state is invalid; the values of the specific environmental factors and the action states are shown in the following tables 3 and 4:
TABLE 3 values of environmental factors
Wherein, the irrelevant item represents that the value can be '0' or '1', and the value does not affect the corresponding state under the value, but only affects the state under the condition that the value needs to be taken as a specific value;
table 4 value of action state of unmanned aerial vehicle
Step 2: inference judgment of unmanned aerial vehicle selectable obstacle avoidance actions
1) Writing out the state vector from Step1
The state vector is represented by C (k), wherein k represents the number of times, C (0) represents the initial state vector, the state vector comprises all contents of obstacle avoidance actions and environmental factors, and the format is as follows:
in the avoidance maneuver, the default drones A1-A4 all have values of 0, that is, values of
V1-V5 are values obtained at Step1, and are constant values at Step 2;
2) reasoning is performed using the state vector c (k), the reasoning process being as follows:
① multiplying the state vector C (k) by the neighboring weight matrix W to obtain an intermediate vector X (k):
X(k)＝C(k)W (6)
wherein, the adjacent weight matrix W between the nodes is as follows:
② processing each component x of the intermediate vector x (k) with a state transfer function f (x) which is a binary step function:
wherein x is a component of the intermediate vector x (k), the dimension of the vector x (k) is 4, and the value range of the component x is {0,1 };
③ update the state vector C (k) with the intermediate vector X (k), i.e., the
④ repeating steps ①, ② and ③, namely updating the state vector C (k) until the state vector C (k + n +1) is equal to C (k + n), wherein n represents the number of times, i.e. the value of the (k + n +1) th state vector is the same as the value of the (k + n) th state vector;
⑤ the drone provides optional obstacle avoidance maneuver results to the drone;
step 3: final obstacle avoidance action determination
1) When one of the following conditions a), and b) is satisfied, then the drone needs to interact with the human-machine:
a) the environmental factors include an unknown obstacle V4 with a value of "1", i.e., V (, 1,), where x represents an irrelevant term and the irrelevant term represents a value of "0" or "1";
b) the selectable obstacle avoidance actions do not accord with action grades corresponding to the current supervision control mode of the unmanned aerial vehicle;
the action in the current supervisory control mode is performed as follows:
①, when the supervision control mode of the unmanned aerial vehicle is an exception management mode L3 and the unmanned aerial vehicle needs interaction, if the unmanned aerial vehicle cannot provide an obstacle avoidance action reasoning result within 15s, namely an obstacle is unknown, the unmanned aerial vehicle reduces the supervision control mode to an agreement management mode L2;
②, when the supervision control mode of the unmanned aerial vehicle is the agreement management mode L2 and the unmanned aerial vehicle needs to interact, if the unmanned aerial vehicle cannot provide an obstacle avoidance action reasoning result within 15s, namely an obstacle is unknown, the unmanned aerial vehicle reduces the supervision control mode to the instruction control mode L1, if the unmanned aerial vehicle provides the obstacle avoidance action reasoning result to the human machine, the unmanned aerial vehicle actively feeds back in an intelligent combination mode to wait for approval of an operator, if the operator does not deny within 15s, the obstacle avoidance action reasoning result provided by the unmanned aerial vehicle is executed, if the operator does not deny within 15s, the unmanned aerial vehicle does not execute, and reduces the supervision control mode to the instruction control mode L1;
③ when the supervision control mode of the unmanned aerial vehicle is the instruction control mode L1 and the unmanned aerial vehicle needs to interact with each other, the unmanned aerial vehicle provides the image information and the position information of the obstacle to the human machine, provides the state vector C (k) and the obstacle avoiding action reasoning result, and the human machine executes the action in a takeover mode:
if (obstacle type known), then (select autonomous obstacle avoidance decision avoidance)
if (obstacle type is unknown and far away), then (select slow down wait instruction) (9)
if (obstacle type unknown and distance close), then (select hover wait instruction)
Giving an obstacle avoidance action of the unmanned aerial vehicle through a command, waiting for a decision of an operator in 15s of the unmanned aerial vehicle, executing according to a decision result, and returning to the base if no decision is made when the decision is over;
2) when the selectable obstacle avoiding action meets the action grade corresponding to the current supervision control mode of the unmanned aerial vehicle and the environmental factors do not appear under the condition of the judgment condition a) of the step 1) in step2, the unmanned aerial vehicle does not need to interact with the human-computer, and the selectable obstacle avoiding action is executed according to the following priority:
A1>A2>A3>A4 (10)
the action execution mode in the corresponding supervision control mode is as follows:
① when the unmanned aerial vehicle does not need interaction, in the exception management mode L3, the actions allowed to be executed are an autonomous obstacle avoidance action a1, a deceleration waiting instruction action a2, a hover waiting instruction action A3 and a return action a4, and the unmanned aerial vehicle adopts an autonomous execution mode according to the priority of the formula (10);
② when the unmanned aerial vehicle does not need interaction, in the agreement management mode L2, the actions allowed to be executed are a deceleration waiting command action A2, a hover waiting command action A3 and a return action A4, and the unmanned aerial vehicle adopts an autonomous execution mode according to the priority of the formula (10);
③ when the unmanned aerial vehicle does not need interaction, under the instruction control mode L1, the action allowed to be executed is a return action A4, and the unmanned aerial vehicle adopts an autonomous execution mode according to the priority of the formula (10);
step 4: updating the environmental factors and the obstacle avoidance action values of Step1 every 30s, if the unmanned aerial vehicle supervision control mode is lowered, updating the unmanned aerial vehicle supervision control mode to the lowered mode, circularly reciprocating according to Step 1-Step 3, executing the finally obtained obstacle avoidance action by the unmanned aerial vehicle, finishing obstacle avoidance, and after the obstacle avoidance is successful, recovering the unmanned aerial vehicle supervision control mode to be the exception management mode L3.
The unmanned aerial vehicle has the beneficial effects that for known obstacles, the unmanned aerial vehicle can fully exert the autonomous ability thereof and independently complete obstacle avoidance; for unknown obstacles, the unmanned aerial vehicle must complete evasion under the guidance of a human machine, and the variable autonomous supervision control method can fully exert the autonomous execution capability of the unmanned aerial vehicle and the analysis and judgment capability of a human-machine operator; for the extreme conditions such as communication interruption, the unmanned aerial vehicle can guarantee the safety of the unmanned aerial vehicle by adjusting the supervision control mode while waiting for communication recovery. The variable autonomous supervision control mode can combine the characteristics of an unmanned aerial vehicle and a human machine, and has better strain capacity for different obstacle types and environmental conditions.
Drawings
FIG. 1 is a model block diagram of unmanned aerial vehicle obstacle avoidance of the present invention.
Fig. 2 is an obstacle avoidance process diagram of embodiment 1 of the present invention.
Fig. 3 is an obstacle avoidance process diagram of embodiment 2 of the present invention.
Fig. 4 is an obstacle avoidance process diagram according to embodiment 3 of the present invention.
Detailed Description
The invention is further explained below by combining the attached drawings and an embodiment, and fig. 1 is a model block diagram of unmanned aerial vehicle obstacle avoidance of the invention.
Step 1: defining environmental factors, obstacle avoidance actions and supervision control modes
When the unmanned aerial vehicle encounters an obstacle, 4 obstacle avoidance actions are provided, including an autonomous obstacle avoidance action A1, a deceleration waiting instruction action A2, a hover waiting instruction action A3 and a return action A4;
wherein the autonomous obstacle avoiding action a1 is defined as a high-level action; the deceleration waiting instruction a2 and the hover waiting instruction A3 are intermediate-level actions; return a4 to a low level action;
the supervision control mode is 3 modes, and the autonomous capacity of the unmanned aerial vehicle is respectively an exception management mode L3, an agreement management mode L2 and a command control mode L1 from high to low; the initial mode of the unmanned aerial vehicle is an exception management mode L3;
the unmanned aerial vehicle supervision control mode is defined as shown in table 1;
TABLE 1 three supervisory control modes and their corresponding actions
The action execution mode refers to a mode that the unmanned aerial vehicle completes corresponding obstacle avoidance actions under the supervision control of the unmanned aerial vehicle, and is shown in table 2:
TABLE 2 three supervisory control modes and their corresponding action execution modes
1) The environmental factors include 3 aspects of the distance between obstacles, whether the obstacle information is known, and whether the current communication condition is good, which are defined as follows:
① definition of obstacle distance, including obstacle near V3 and obstacle far V5;
with respect to the drone, the position of the obstacle may be expressed as
In the formula (1), P (t) is the position of the obstacle i at the time t, t is time, P (0) is the initial position of the obstacle i, viIs the velocity, v, of the obstacle iu(τ) is the velocity of drone u;
in the formula (2)
For the yaw angle of the drone, the distance between the drone and the obstacle is defined by the time T to reach the obstacle position by the following formula:
② whether the obstacle is known to be divided into a known obstacle V1 and an unknown obstacle V4, the known obstacle being a known obstacle in which both image information and position information of the obstacle are known in the working area, and the unknown obstacle being an obstacle in which at least one of the image information and the position information of the obstacle is unknown in the working area;
③, dividing the current communication status into normal communication and interruption V2, wherein the signal intensity of the interruption means that the signal reaches the unmanned aerial vehicle receiver is less than the sensitivity of the unmanned aerial vehicle receiver, which causes the situation that the normal communication between the unmanned aerial vehicle and the manned aerial vehicle can not be realized;
2) the value range of the values of the environmental factors and the obstacle avoidance actions is {0,1}, wherein 1 represents that the state is effective, and 0 represents that the state is invalid; the values of the specific environmental factors and the action states are shown in the following tables 3 and 4:
TABLE 3 values of environmental factors
Wherein, the irrelevant item represents that the value can be '0' or '1', and the value does not affect the corresponding state under the value, but only affects the state under the condition that the value needs to be taken as a specific value;
table 4 value of action state of unmanned aerial vehicle
Step 2: inference judgment of unmanned aerial vehicle selectable obstacle avoidance actions
1) Writing out the state vector from Step1
The state vector is represented by C (k), wherein k represents the number of times, C (0) represents the initial state vector, the state vector comprises all contents of obstacle avoidance actions and environmental factors, and the format is as follows:
in the avoidance maneuver, the default drones A1-A4 all have values of 0, that is, values of
V1-V5 are values obtained at Step1, and are constant values at Step 2;
2) reasoning is performed using the state vector c (k), the reasoning process being as follows:
① multiplying the state vector C (k) by the neighboring weight matrix W to obtain an intermediate vector X (k):
X(k)＝C(k)W (6)
wherein, the adjacent weight matrix W between the nodes is as follows:
② processing each component x of the intermediate vector x (k) with a state transfer function f (x) which is a binary step function:
wherein x is a component of the intermediate vector x (k), the dimension of the vector x (k) is 4, and the value range of the component x is {0,1 };
③ update the state vector C (k) with the intermediate vector X (k), i.e., the
④ repeating steps ①, ② and ③, namely updating the state vector C (k) until the state vector C (k + n +1) is equal to C (k + n), wherein n represents the number of times, i.e. the value of the (k + n +1) th state vector is the same as the value of the (k + n) th state vector;
⑤ the drone provides optional obstacle avoidance maneuver results to the drone;
step 3: final obstacle avoidance action determination
1) When one of the following conditions a), and b) is satisfied, then the drone needs to interact with the human-machine:
a) the environmental factors include an unknown obstacle V4 with a value of "1", i.e., V (, 1,), where x represents an irrelevant term and the irrelevant term represents a value of "0" or "1";
b) the selectable obstacle avoidance actions do not accord with action grades corresponding to the current supervision control mode of the unmanned aerial vehicle;
the action in the current supervisory control mode is performed as follows:
①, when the supervision control mode of the unmanned aerial vehicle is an exception management mode L3 and the unmanned aerial vehicle needs interaction, if the unmanned aerial vehicle cannot provide an obstacle avoidance action reasoning result within 15s, namely an obstacle is unknown, the unmanned aerial vehicle reduces the supervision control mode to an agreement management mode L2;
②, when the supervision control mode of the unmanned aerial vehicle is the agreement management mode L2 and the unmanned aerial vehicle needs to interact, if the unmanned aerial vehicle can not provide the obstacle avoidance action reasoning result within 15s, namely the obstacle is unknown, the unmanned aerial vehicle reduces the supervision control mode to the instruction control mode L1, if the unmanned aerial vehicle provides the obstacle avoidance action reasoning result to the human machine, the unmanned aerial vehicle actively feeds back in an intelligent combination mode to wait for the approval of the operator, if the operator does not deny within 15s, the obstacle avoidance action reasoning result provided by the unmanned aerial vehicle is executed, if the operator negates within 15s, the unmanned aerial vehicle does not execute, and reduces the supervision control mode to the operator decision mode L1;
③ when the supervision control mode of the unmanned aerial vehicle is the operator decision mode L1 and the unmanned aerial vehicle needs to interact with each other, the unmanned aerial vehicle provides the image information and the position information of the obstacle for the human, provides the state vector C (k) and the obstacle avoiding action reasoning result, and the human adopts the takeover mode to execute the action:
giving an obstacle avoidance action of the unmanned aerial vehicle through a command, waiting for a decision of an operator in 15s of the unmanned aerial vehicle, executing according to a decision result, and returning to the base if no decision is made when the decision is over;
2) when the selectable obstacle avoiding action meets the action grade corresponding to the current supervision control mode of the unmanned aerial vehicle and the environmental factors do not appear under the condition of the judgment condition a) of the step 1) in step2, the unmanned aerial vehicle does not need to interact with the human-computer, and the selectable obstacle avoiding action executes according to the following priority
A1>A2>A3>A4 (10)
The action execution mode in the corresponding supervision control mode is as follows:
① when the unmanned aerial vehicle does not need interaction, in the exception management mode L3, the actions allowed to be executed are an autonomous obstacle avoidance action a1, a deceleration waiting instruction action a2, a hover waiting instruction action A3 and a return action a4, and the unmanned aerial vehicle adopts an autonomous execution mode according to the priority of the formula (10);
② when the unmanned aerial vehicle does not need interaction, in the agreement management mode L2, the actions allowed to be executed are a deceleration waiting command action A2, a hover waiting command action A3 and a return action A4, and the unmanned aerial vehicle adopts an autonomous execution mode according to the priority of the formula (10);
③ when the unmanned aerial vehicle does not need interaction, under the instruction control mode L1, the action allowed to be executed is a return action A4, and the unmanned aerial vehicle adopts an autonomous execution mode according to the priority of the formula (10);
step 4: updating the environmental factors and the obstacle avoidance action values of Step1 every 30s, if the unmanned aerial vehicle supervision control mode is lowered, updating the unmanned aerial vehicle supervision control mode to the lowered mode, circularly reciprocating according to Step 1-Step 3, executing the finally obtained obstacle avoidance action by the unmanned aerial vehicle, finishing obstacle avoidance, and after the obstacle avoidance is successful, recovering the unmanned aerial vehicle supervision control mode to be the exception management mode L3.
The case of the unmanned aerial vehicle when the obstacle is avoided can be divided into two categories according to normal communication and communication interruption, meanwhile, the case of the unmanned aerial vehicle when the obstacle is avoided can be divided into two cases facing known obstacles and unknown obstacles, and the following processes of the three cases are simulated respectively, as shown in table 5:
TABLE 5 unmanned aerial vehicle obstacle avoidance
The embodiments shown in table 5 are classified into two general cases of normal communication and interrupted communication, and the normal communication case is classified into two types of known obstacles and unknown obstacles. The following describes the simulation process of task execution of the unmanned aerial vehicle under three embodiments:
example 1
(1) Step 1: representation of environmental factors and obstacle avoidance actions
As can be seen from table 5, if the type of the obstacle is a known obstacle, V1 is 1, and V4 is 0; when the obstacle distance is far, V3 is 0, and V5 is 1; when the communication status is normal, V2 is 0, the drone actions a1 to a4 are all initialized to 0, and the initial supervisory control mode is the exception management mode L3;
(2) step 2: inference of selectable obstacle avoidance maneuvers
The state vector is:
iterating the state vector, wherein the state vector of the output node is as follows:
the output result shows that the selectable action of the unmanned aerial vehicle in the embodiment 1 is autonomous obstacle avoidance A1, so that the unmanned aerial vehicle can fully exert the autonomous capability of the known obstacle and independently complete obstacle avoidance.
(3) Step 3: final obstacle avoidance action determination
For embodiment 1, the obstacle type is a known obstacle, the unmanned aerial vehicle has an ability to avoid the autonomous obstacle, and in the exception management mode L3, the unmanned aerial vehicle has a right to execute an autonomous obstacle avoiding action, so that the unmanned aerial vehicle can autonomously complete an obstacle avoiding process without interacting with a human machine, and the process is as shown in fig. 2.
Case 2
(1) Step 1: representation of environmental factors and obstacle avoidance actions
As can be seen from table 5, if the type of the obstacle is unknown, V1 is 0, and V4 is 1; when the obstacle distance is short, V3 is 1, and V5 is 0; if the communication status is normal, V2 is 0, the drone actions a1 to a4 are all initialized to 0, and the initial supervisory control mode is the exception management mode.
(2) Step 2: inference of selectable obstacle avoidance maneuvers
The state vector is:
iterating the state vector, wherein the state vector of the output node is as follows:
the output result shows that in the embodiment 2, the unmanned aerial vehicle cannot avoid the obstacle autonomously, and then the hovering waiting instruction action a3 is executed.
(3) Step 3: final obstacle avoidance action determination
Unmanned aerial vehicle can not accomplish the obstacle and avoid by oneself, need interact with someone, returns the image information and the positional information of obstacle simultaneously, and someone's machine operator is according to the information of obstacle, updates the obstacle data, is about to unknown obstacle information update for known obstacle information.
Updating the representation of the environmental factors and the obstacle avoidance action of Step1, wherein if the obstacle type is a known obstacle, V1 is 1, and V4 is 0; when the obstacle distance is short, V3 is 1, and V5 is 0; if the communication status is normal, V2 is 0, the drone actions a1 to a4 are all initialized to 0, and the supervisory control mode is the exception management mode L3.
Update inference judgment of optional obstacle avoidance action of Step2
The state vector is:
iterate it, the output state vector is:
the output result indicates that the action selected by the drone in the case of scenario 2 is the autonomous obstacle avoidance a 1.
Update Step 3: final obstacle avoidance action determination
For embodiment 2, the obstacle type is unknown, and the unmanned aerial vehicle has no capability of avoiding autonomous obstacles, so that interaction with the unmanned aerial vehicle is required; the man-machine receives obstacle image information and position information returned by the unmanned aerial vehicle, and the obstacle information is updated to change the type into a known obstacle; the unmanned aerial vehicle further performs inference and judgment of selectable obstacle avoiding actions according to the environment state under the current condition, at the moment, the obstacle type is a known obstacle, the unmanned aerial vehicle has the capability of autonomous obstacle avoiding, and in the exception management mode L3, the unmanned aerial vehicle has the authority of executing the autonomous obstacle avoiding actions, so that the unmanned aerial vehicle can autonomously complete an obstacle avoiding process without interacting with a human machine, and the process is shown in FIG. 3.
Example 3
(1) Step 1: representation of environmental factors and obstacle avoidance actions
From example 3, if the type of the obstacle is unknown, then V1 is 0, and V4 is 1; when the obstacle distance is far, V3 is 0, and V5 is 1; when the communication status is communication interruption, V2 is equal to 1, the drone operations a1 to a4 are all initialized to 0, and the supervisory control mode is the exception management mode L3.
(2) Step 2: inference of selectable obstacle avoidance maneuvers
The state vector is:
iterating the state vector, wherein the state vector of the output node is as follows:
the output result shows that the unmanned aerial vehicle has no autonomous obstacle avoidance capability under the condition of the embodiment 3, and then decelerates to wait for the command a2, and the unmanned aerial vehicle cannot receive the command of the human machine due to communication interruption, and executes a return operation a4 to ensure the safety of the unmanned aerial vehicle.
(3) Step 3: final obstacle avoidance action determination
The unmanned aerial vehicle can not finish obstacle avoidance decision autonomously, meanwhile, due to communication interruption, the unmanned aerial vehicle can not interact with the unmanned aerial vehicle, at the moment, the unmanned aerial vehicle selects and adjusts a supervisory control mode, the supervisory control mode is lowered to an operator decision mode L2 after 15s, communication interruption still exists after 15s, interaction cannot be achieved, the unmanned aerial vehicle lowers the supervisory control mode to an instruction control mode L1, communication interruption still exists after 15s, interaction cannot be achieved, the unmanned aerial vehicle selects self-protection behavior according to the instruction control mode L1, and returns to the base, namely return-to-base action A4 is executed, and the process is shown in FIG. 4.
Claims (1)
1. A supervision control method for avoiding obstacles of unmanned aerial vehicles by people is characterized by comprising the following steps:
step 1: defining environmental factors, obstacle avoidance actions and supervision control modes
When the unmanned aerial vehicle encounters an obstacle, 4 obstacle avoidance actions are provided, including an autonomous obstacle avoidance action A1, a deceleration waiting instruction action A2, a hover waiting instruction action A3 and a return action A4;
wherein the autonomous obstacle avoiding action a1 is defined as a high-level action; the deceleration waiting instruction a2 and the hover waiting instruction A3 are intermediate-level actions; return a4 to a low level action;
the supervision control mode is 3 modes, and the autonomous capacity of the unmanned aerial vehicle is respectively an exception management mode L3, an agreement management mode L2 and a command control mode L1 from high to low; the initial mode of the unmanned aerial vehicle is an exception management mode L3;
the unmanned aerial vehicle supervision control mode is defined as shown in table 1;
TABLE 1 three supervisory control modes and their corresponding actions
The action execution mode refers to a mode that the unmanned aerial vehicle completes corresponding obstacle avoidance actions under the supervision control of the unmanned aerial vehicle, and is shown in table 2:
TABLE 2 three supervisory control modes and their corresponding action execution modes
1) The environmental factors include 3 aspects of the distance between obstacles, whether the obstacle information is known, and whether the current communication condition is good, which are defined as follows:
① definition of obstacle distance, including obstacle near V3 and obstacle far V5;
with respect to the drone, the position of the obstacle may be expressed as
In the formula (1), P (t) is the position of the obstacle i at the time t, t is time, P (0) is the initial position of the obstacle i, viIs the velocity, v, of the obstacle iu(τ) is the velocity of drone u;
in the formula (2)
For the yaw angle of the drone, the distance between the drone and the obstacle is defined by the time T to reach the obstacle position by the following formula:
② whether the obstacle is known to be divided into a known obstacle V1 and an unknown obstacle V4, the known obstacle being a known obstacle in which both image information and position information of the obstacle are known in the working area, and the unknown obstacle being an obstacle in which at least one of the image information and the position information of the obstacle is unknown in the working area;
③, dividing the current communication status into normal communication and interruption V2, wherein the signal intensity of the interruption means that the signal reaches the unmanned aerial vehicle receiver is less than the sensitivity of the unmanned aerial vehicle receiver, which causes the situation that the normal communication between the unmanned aerial vehicle and the manned aerial vehicle can not be realized;
2) the value range of the values of the environmental factors and the obstacle avoidance actions is {0,1}, wherein 1 represents that the state is effective, and 0 represents that the state is invalid; the values of the specific environmental factors and the action states are shown in the following tables 3 and 4:
TABLE 3 values of environmental factors
Wherein, the irrelevant item represents that the value can be '0' or '1', and the value does not affect the corresponding state under the value, but only affects the state under the condition that the value needs to be taken as a specific value;
table 4 value of action state of unmanned aerial vehicle
Step 2: inference judgment of unmanned aerial vehicle selectable obstacle avoidance actions
1) Writing out the state vector from Step1
The state vector is represented by C (k), wherein k represents the number of times, C (0) represents the initial state vector, the state vector comprises all contents of obstacle avoidance actions and environmental factors, and the format is as follows:
in the avoidance maneuver, the default drones A1-A4 all have values of 0, that is, values of
V1-V5 are values obtained at Step1, and are constant values at Step 2;
2) reasoning is performed using the state vector c (k), the reasoning process being as follows:
① multiplying the state vector C (k) by the neighboring weight matrix W to obtain an intermediate vector X (k):
X(k)＝C(k)W (6)
wherein, the adjacent weight matrix W between the nodes is as follows:
② processing each component x of the intermediate vector x (k) with a state transfer function f (x) which is a binary step function:
wherein x is a component of the intermediate vector x (k), the dimension of the vector x (k) is 4, and the value range of the component x is {0,1 };
③ update the state vector C (k) with the intermediate vector X (k), i.e., the
④ repeating steps ①, ② and ③, namely updating the state vector C (k) until the state vector C (k + n +1) is equal to C (k + n), wherein n represents the number of times, i.e. the value of the (k + n +1) th state vector is the same as the value of the (k + n) th state vector;
⑤ the drone provides optional obstacle avoidance maneuver results to the drone;
step 3: final obstacle avoidance action determination
1) When one of the following conditions a), and b) is satisfied, then the drone needs to interact with the human-machine:
a) the environmental factors include an unknown obstacle V4 with a value of "1", i.e., V (, 1,), where x represents an irrelevant term and the irrelevant term represents a value of "0" or "1";
b) the selectable obstacle avoidance actions do not accord with action grades corresponding to the current supervision control mode of the unmanned aerial vehicle;
the action in the current supervisory control mode is performed as follows:
①, when the supervision control mode of the unmanned aerial vehicle is an exception management mode L3 and the unmanned aerial vehicle needs interaction, if the unmanned aerial vehicle cannot provide an obstacle avoidance action reasoning result within 15s, namely an obstacle is unknown, the unmanned aerial vehicle reduces the supervision control mode to an agreement management mode L2;
②, when the supervision control mode of the unmanned aerial vehicle is the agreement management mode L2 and the unmanned aerial vehicle needs to interact, if the unmanned aerial vehicle cannot provide an obstacle avoidance action reasoning result within 15s, namely an obstacle is unknown, the unmanned aerial vehicle reduces the supervision control mode to the instruction control mode L1, if the unmanned aerial vehicle provides the obstacle avoidance action reasoning result to the human machine, the unmanned aerial vehicle actively feeds back in an intelligent combination mode to wait for approval of an operator, if the operator does not deny within 15s, the obstacle avoidance action reasoning result provided by the unmanned aerial vehicle is executed, if the operator negates within 15s, the unmanned aerial vehicle does not execute, and reduces the supervision control mode to the instruction control mode L1;
③ when the supervision control mode of the unmanned aerial vehicle is the instruction control mode L1 and the unmanned aerial vehicle needs to interact with each other, the unmanned aerial vehicle provides the image information and the position information of the obstacle to the human machine, provides the state vector C (k) and the obstacle avoiding action reasoning result, and the human machine executes the action in a takeover mode:
if (obstacle type known), then (select autonomous obstacle avoidance decision avoidance)
if (obstacle type is unknown and far away), then (select slow down wait instruction) (9)
if (obstacle type unknown and distance close), then (select hover wait instruction)
Giving an obstacle avoidance action of the unmanned aerial vehicle through a command, waiting for a decision of an operator in 15s of the unmanned aerial vehicle, executing according to a decision result, and returning to the base if no decision is made when the decision is over;
2) when the selectable obstacle avoiding action meets the action grade corresponding to the current supervision control mode of the unmanned aerial vehicle and the environmental factors do not appear under the condition of the judgment condition a) of the step 1) in step2, the unmanned aerial vehicle does not need to interact with the human-computer, and the selectable obstacle avoiding action is executed according to the following priority:
A1>A2>A3>A4 (10)
the action execution mode in the corresponding supervision control mode is as follows:
① when the unmanned aerial vehicle does not need interaction, in the exception management mode L3, the actions allowed to be executed are an autonomous obstacle avoidance action a1, a deceleration waiting instruction action a2, a hover waiting instruction action A3 and a return action a4, and the unmanned aerial vehicle adopts an autonomous execution mode according to the priority of the formula (10);
② when the unmanned aerial vehicle does not need interaction, in the agreement management mode L2, the actions allowed to be executed are a deceleration waiting command action A2, a hover waiting command action A3 and a return action A4, and the unmanned aerial vehicle adopts an autonomous execution mode according to the priority of the formula (10);
③ when the unmanned aerial vehicle does not need interaction, under the instruction control mode L1, the action allowed to be executed is a return action A4, and the unmanned aerial vehicle adopts an autonomous execution mode according to the priority of the formula (10);
step 4: updating the environmental factors and the obstacle avoidance action values of Step1 every 30s, if the unmanned aerial vehicle supervision control mode is lowered, updating the unmanned aerial vehicle supervision control mode to the lowered mode, circularly reciprocating according to Step 1-Step 3, executing the finally obtained obstacle avoidance action by the unmanned aerial vehicle, finishing obstacle avoidance, and after the obstacle avoidance is successful, recovering the unmanned aerial vehicle supervision control mode to be the exception management mode L3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710553624.XA CN107943066B (en) | 2017-07-08 | 2017-07-08 | Method for supervising and controlling obstacle avoidance of unmanned aerial vehicle by using human |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710553624.XA CN107943066B (en) | 2017-07-08 | 2017-07-08 | Method for supervising and controlling obstacle avoidance of unmanned aerial vehicle by using human |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107943066A CN107943066A (en) | 2018-04-20 |
| CN107943066B true CN107943066B (en) | 2020-01-14 |
Family
ID=61928567
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710553624.XA Active CN107943066B (en) | 2017-07-08 | 2017-07-08 | Method for supervising and controlling obstacle avoidance of unmanned aerial vehicle by using human |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107943066B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109101039A (en) * | 2018-06-29 | 2018-12-28 | 太原理工大学 | Vertical detection method and system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5581250A (en) * | 1995-02-24 | 1996-12-03 | Khvilivitzky; Alexander | Visual collision avoidance system for unmanned aerial vehicles |
| EP2674723A2 (en) * | 2012-06-11 | 2013-12-18 | Honeywell International Inc. | Systems and methods for unmanned aircraft system collision avoidance |
| CN103592948A (en) * | 2013-12-04 | 2014-02-19 | 成都纵横自动化技术有限公司 | Unmanned aerial vehicle flying anti-collision method |
| CN103699132A (en) * | 2013-12-05 | 2014-04-02 | 中国航空无线电电子研究所 | Device and method for assisting visual disc to precess and approach |
| CN106094569A (en) * | 2016-07-06 | 2016-11-09 | 西北工业大学 | Multi-sensor Fusion unmanned plane perception with evade analogue system and emulation mode thereof |
| CN106647810A (en) * | 2017-01-10 | 2017-05-10 | 山东科技大学 | UAV automatic collision avoidance method based on negative-proportion guiding |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9562773B2 (en) * | 2014-03-15 | 2017-02-07 | Aurora Flight Sciences Corporation | Autonomous vehicle navigation system and method |
-
2017
- 2017-07-08 CN CN201710553624.XA patent/CN107943066B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5581250A (en) * | 1995-02-24 | 1996-12-03 | Khvilivitzky; Alexander | Visual collision avoidance system for unmanned aerial vehicles |
| EP2674723A2 (en) * | 2012-06-11 | 2013-12-18 | Honeywell International Inc. | Systems and methods for unmanned aircraft system collision avoidance |
| CN103592948A (en) * | 2013-12-04 | 2014-02-19 | 成都纵横自动化技术有限公司 | Unmanned aerial vehicle flying anti-collision method |
| CN103699132A (en) * | 2013-12-05 | 2014-04-02 | 中国航空无线电电子研究所 | Device and method for assisting visual disc to precess and approach |
| CN106094569A (en) * | 2016-07-06 | 2016-11-09 | 西北工业大学 | Multi-sensor Fusion unmanned plane perception with evade analogue system and emulation mode thereof |
| CN106647810A (en) * | 2017-01-10 | 2017-05-10 | 山东科技大学 | UAV automatic collision avoidance method based on negative-proportion guiding |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107943066A (en) | 2018-04-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107491085B (en) | Supervision control method for preventing threat of human to unmanned aerial vehicle | |
| Burger et al. | Cooperative multiple vehicle trajectory planning using miqp | |
| Sheridan | Adaptive automation, level of automation, allocation authority, supervisory control, and adaptive control: Distinctions and modes of adaptation | |
| Meng et al. | Dynamic decision making in lane change: Game theory with receding horizon | |
| CN107491086B (en) | Unmanned aerial vehicle formation obstacle avoidance method and system under time-varying network topology | |
| CN110470306B (en) | Multi-robot formation navigation method capable of guaranteeing connectivity constraint and based on deep reinforcement learning | |
| Shi et al. | Offline reinforcement learning for autonomous driving with safety and exploration enhancement | |
| US11579574B2 (en) | Control customization system, control customization method, and control customization program | |
| Gharib et al. | Multi-objective optimization of a path-following MPC for vehicle guidance: A Bayesian optimization approach | |
| Wang et al. | Interaction-aware model predictive control for autonomous driving | |
| Sun et al. | Cooperative strategy for pursuit-evasion problem in the presence of static and dynamic obstacles | |
| CN107943066B (en) | Method for supervising and controlling obstacle avoidance of unmanned aerial vehicle by using human | |
| CN115993845B (en) | Coordinated motion planning and formation control method for cluster intelligent system | |
| Ding et al. | Executive decision support | |
| CN110806758B (en) | Unmanned aerial vehicle cluster autonomous level self-adaptive adjustment method based on scene fuzzy cognitive map | |
| Asl et al. | A new approach on multi-agent Multi-Objective Reinforcement Learning based on agents' preferences | |
| Chen et al. | Learning Agile Quadrotor Flight in Restricted Environments with Safety Guarantees | |
| Fridovich-Keil et al. | Approximate solutions to a class of reachability games | |
| Johnson et al. | A framework for understanding automation in terms of levels of human control abstraction | |
| Honda et al. | Mpc builder for autonomous drive: Automatic generation of mpcs for motion planning and control | |
| Chu et al. | Distributed Composite Learning Adaptive Fault-Tolerant Control for Multiple Marine Vehicles With Event-Triggered Communication | |
| Németh et al. | Hierarchical control design of automated vehicles for multi-vehicle scenarios in roundabouts | |
| Huang et al. | A behavior control scheme for multi-robot systems under human intervention | |
| CN114995486A (en) | Method, device and equipment for controlling unmanned aerial vehicle to carry out avoidance operation and storage medium | |
| Zhang et al. | Velocity control in a right-turn across traffic scenario for autonomous vehicles using kernel-based reinforcement learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20200812 Address after: 100190, No. 9 West Fourth Ring Road, Beijing, Haidian District Patentee after: Beijing Yuanchen Huasheng Technology Co., Ltd Address before: 710072 Xi'an friendship West Road, Shaanxi, No. 127 Patentee before: Northwestern Polytechnical University |