CN113091999A - Foot type robot leg one-dimensional force sensor calibration method and system - Google Patents
Foot type robot leg one-dimensional force sensor calibration method and system Download PDFInfo
- Publication number
- CN113091999A CN113091999A CN202110354203.0A CN202110354203A CN113091999A CN 113091999 A CN113091999 A CN 113091999A CN 202110354203 A CN202110354203 A CN 202110354203A CN 113091999 A CN113091999 A CN 113091999A
- Authority
- CN
- China
- Prior art keywords
- force sensor
- detection value
- dimensional force
- joint
- foot end
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01L—MEASURING FORCE, STRESS, TORQUE, WORK, MECHANICAL POWER, MECHANICAL EFFICIENCY, OR FLUID PRESSURE
- G01L25/00—Testing or calibrating of apparatus for measuring force, torque, work, mechanical power, or mechanical efficiency
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Manipulator (AREA)
Abstract
The invention relates to a method and a system for calibrating a leg one-dimensional force sensor of a foot robot, which control the position of a leg driving system by an upper computer, load the force of a foot end by a two-dimensional force sensor to obtain the actual detection value of each joint one-dimensional force sensor, construct a leg driving system virtual model by computer software, input the actual detection values of the two-dimensional force sensor and each joint driving unit displacement sensor into the virtual model for simulation to obtain the theoretical detection value of each joint one-dimensional force sensor in the virtual model, finally obtain a calibration curve between the actual detection value and the theoretical detection value of each joint one-dimensional force sensor by a least square method, obtain the correction calibration coefficient of each joint according to the calibration curve, and calibrate the one-dimensional force sensor by the coefficient, the recalibration of the one-dimensional force sensor is realized.
Description
Technical Field
The invention relates to the field of sensor detection equipment of a foot type robot, in particular to a method and a system suitable for calibrating a one-dimensional force sensor of a leg of the foot type robot.
Background
The robot is more and more widely applied in various industries, and can replace human beings to carry out work with high risk and strong repeatability. The legged robot simulates the physiological characteristics of legged animals to carry out structural design, so that the legged robot has the characteristic of obvious discontinuous support, and has stronger adaptability to complex environments.
During walking, the foot end of the foot robot sometimes generates a large contact force with the ground. If the contact force of the foot end cannot be well controlled, the electronic equipment carried by the robot body can bear great impact to cause damage, so that the robot is required to detect the contact force between the foot end and the ground in real time.
There are generally two methods to detect the contact force between the foot end and the ground: the first method is to install a multidimensional force sensor at the foot end to directly detect the contact force of the foot end; the second method is to install a one-dimensional force sensor at the top of each joint driving unit of the leg of the robot and combine statics to indirectly obtain the contact force of the foot end. The multidimensional force sensor is expensive and easy to damage, and the one-dimensional force sensor has high data stability and long service life, so that in the foot type robot, the contact force between the foot end of the robot and the ground is solved indirectly by adopting a second method.
The one-dimensional force sensor is calibrated when leaving a factory, and a user can directly determine a calibration coefficient according to the linear relation between the measuring range and the output voltage in the sample. However, after long-term use, due to the change of related factors such as friction force and pretightening force, a certain error will be generated between the detection value and the actual value of the one-dimensional force sensor, and at this time, the one-dimensional force sensor needs to be recalibrated. In the traditional calibration method, the one-dimensional force sensor needs to be detached from the robot leg, but most parts of the robot leg are in interference fit, and if the robot leg is frequently detached, the assembly relation of the robot leg is damaged, so that the motion control performance of the robot is influenced, and an unreasonable installation process influences a calibration coefficient and even damages the one-dimensional force sensor.
In summary, in the technical field of sensor detection equipment for foot robots, a method capable of completing recalibration of a sensor without disassembling a one-dimensional force sensor of a leg joint is urgently needed.
Disclosure of Invention
The invention aims to provide a method and a system for calibrating a leg one-dimensional force sensor of a foot type robot, which can finish the recalibration of the one-dimensional force sensor under the condition of not disassembling the force sensor.
In order to achieve the purpose, the invention provides the following scheme:
a method for calibrating a leg one-dimensional force sensor of a legged robot, comprising the following steps:
in the process of loading and unloading the acting force of the foot end, recording a two-dimensional force sensor detection value for loading, a displacement sensor detection value of each joint driving unit and a one-dimensional force sensor detection value by using an upper computer;
inputting the detection value of the displacement sensor and the detection value of the two-dimensional force sensor into a leg driving system virtual model built based on a leg mechanical structure for simulation, and recording the simulation detection value of the one-dimensional force sensor of each joint driving unit in the virtual model;
for each joint, subtracting the detection value of the one-dimensional force sensor when the foot end is at the initial position from the detection value of the one-dimensional force sensor and the simulation detection value of the one-dimensional force sensor to obtain the actual detection value of each joint and the simulation detection value of each joint, wherein the detection value of the one-dimensional force sensor at the initial position is the detection value of the one-dimensional force sensor when acting force is not loaded on the foot end;
solving a one-dimensional force sensor calibration curve of each driving unit by using a least square method according to a linear relation between each actual detection value and each simulated detection value;
obtaining a correction calibration coefficient of each driving unit according to the calibration curve of each driving unit and an original calibration coefficient of each driving unit, wherein the original calibration coefficient is a calibration coefficient of the one-dimensional force sensor when the sensor leaves a factory;
and calibrating the one-dimensional force sensor according to the coefficient corrected and calibrated by each driving unit.
The invention also provides a leg one-dimensional force sensor calibration system of the foot robot, which comprises the following components:
the sensor detection value acquisition module is used for recording a two-dimensional force sensor detection value for loading, a displacement sensor detection value of each joint driving unit and a one-dimensional force sensor detection value by using an upper computer in the process of loading and unloading acting force on the foot end;
the simulation module is used for inputting the detection value of the displacement sensor and the detection value of the two-dimensional force sensor into a leg driving system virtual model built based on a leg mechanical structure for simulation, and recording the simulation detection value of the one-dimensional force sensor of each joint driving unit in the virtual model;
the variable value and sampling value acquisition module is used for subtracting the detection value of the one-dimensional force sensor when the foot end is at the initial position from the detection value of the one-dimensional force sensor and the simulation detection value of the one-dimensional force sensor for each joint to obtain the actual detection value of each joint and the simulation detection value of each joint, wherein the detection value of the one-dimensional force sensor at the initial position is the detection value of the one-dimensional force sensor when the acting force is not loaded on the foot end;
the calibration curve acquisition module is used for solving a one-dimensional force sensor calibration curve of each driving unit by using a least square method according to the linear relation between each actual detection value and each simulated detection value;
and the calibration module is used for obtaining a correction calibration coefficient of each driving unit according to the calibration curve of each driving unit and an original calibration coefficient of each driving unit, wherein the original calibration coefficient is a calibration coefficient of the one-dimensional force sensor when the one-dimensional force sensor leaves a factory, and the one-dimensional force sensor is calibrated according to the correction calibration coefficient of each driving unit.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the method comprises the steps of carrying out position control on a leg driving system through an upper computer, carrying out force loading on a foot end by utilizing two-dimensional force sensors to obtain actual detection values of all joint one-dimensional force sensors, then building a leg driving system virtual model by utilizing computer software, inputting the actual detection values of the two-dimensional force sensors and all joint driving unit displacement sensors into the virtual model for simulation to obtain simulated detection values of all joint one-dimensional force sensors in the virtual model, finally obtaining a calibration curve between the actual detection values and the simulated detection values of all joint one-dimensional force sensors by utilizing a least square method, obtaining a calibration coefficient of each joint correction calibration according to the calibration curve, and calibrating the one-dimensional force sensors by the coefficient compared with the conventional method, wherein the calibration can be finished without taking down the one-dimensional force sensors from the legs of the robot, so that the damage to the assembly relation of the legs of the robot in the disassembly and assembly process and the calibration coefficient of the force sensors in The influence of (c).
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
Fig. 1 is a flowchart of a method for calibrating a leg one-dimensional force sensor of a legged robot according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a loading force loading experimental scheme according to a first embodiment of the present invention;
fig. 3 is a schematic diagram of a kinematic forward model of a leg driving system of a legged robot according to an embodiment of the present invention;
fig. 4 is a schematic view of a kinematic inverse model of a leg driving system of a legged robot according to an embodiment of the present invention;
FIG. 5 is a diagram illustrating a relationship between a joint angle and an extension length of a joint driving unit of the legged robot according to an embodiment of the present invention;
fig. 6 is a schematic diagram of a leg one-dimensional force sensor calibration method of a legged robot according to a first embodiment of the present invention;
fig. 7 is a calibration curve of a one-dimensional force sensor of each joint driving unit of the leg driving system of the legged robot according to the first embodiment of the present invention;
FIG. 8 is a schematic view of a static model of a leg driving system of a legged robot according to an embodiment of the present invention;
fig. 9 is a diagram illustrating a relationship between a moment applied to a joint of the legged robot and a stress applied to a joint driving unit according to the first embodiment of the present invention;
FIG. 10 is a foot end contact force curve in the loading direction for scenario 1 according to a first embodiment of the present invention;
FIG. 11 is a graph of loading direction foot end contact force for scenario 2 in accordance with a first embodiment of the present invention;
FIG. 12 is a graph of loading direction foot end contact force for scenario 3 according to a first embodiment of the present invention;
FIG. 13 is a graph of loading direction foot end contact force for scenario 4 according to a first embodiment of the present invention;
FIG. 14 is a graph of loading direction foot end contact force for scenario 5 in accordance with a first embodiment of the present invention;
FIG. 15 is a foot end contact force curve for a loading direction of scenario 6 according to a first embodiment of the present invention;
FIG. 16 is a graph of the loading direction foot end contact force of scenario 7 in accordance with a first embodiment of the present invention;
FIG. 17 is a graph of loading direction foot end contact force for scenario 8 according to a first embodiment of the present invention;
fig. 18 is a schematic structural diagram of a leg one-dimensional force sensor calibration system of a legged robot according to a second embodiment of the present invention.
Description of the symbols: o-hip joint; e-knee joint; g-ankle joint; a-the connecting position of the hip joint driving unit and the base; b-the position of the connection between the hip joint driving unit and the thigh; c-the connecting position of the knee joint driving unit and the thigh; d, connecting the knee joint driving unit with the crus; f- (E) -B-the position of the ankle joint drive unit connected to the lower leg; h, connecting positions of the ankle joint driving units and the foot ends; i-the apex of the semi-cylindrical foot end; OE-thigh length; EG-shank length; GI-foot length; theta1-hip joint rotation angle: thigh and x0The positive direction of the axis forms an included angle; theta2-knee joint rotation angle: the included angle formed by the shank EG and the extension line of the thigh OE; theta3-ankle joint rotation angle: the extension line direction of the shank EG forms an included angle with the leg component GI; alpha is the included angle formed by EG and ED; beta-OA and x0The positive direction of the axis forms an included angle; AB-total length of extension of hip drive unit; CD-Total extension of the knee drive unit; FH — total extension length of the ankle drive unit; Δ Fs1-the hip joint drive unit force sensor detects a signal; Δ Fs2-the knee joint drive unit force sensor detects a signal; Δ Fs3-the ankle joint drive unit force sensor detects a signal;
-the leg foot end is stressed in the x-axis direction;
-the leg foot end is stressed in the y-axis direction; tau is1The moment to which the hip joint is subjected; tau is2-the moment on the knee joint; tau is3-the moment on the ankle joint;
-the leg drive system tests the extension of the hip drive unit of the experimental platform;
-the leg drive system tests the extension of the knee joint drive unit of the experimental platform;
-the leg drive system tests the extension length of the experimental platform ankle drive unit;
-a leg drive system test experiment platform hip joint drive unit one-dimensional force sensor detection value;
-the leg drive system tests the test values of the one-dimensional force sensors of the experimental platform knee joint drive unit;
-the leg drive system tests the test platform ankle joint drive unit one-dimensional force sensor detection value;
-simulating a detection value by a virtual model hip joint drive unit force sensor;
-the virtual model knee joint drive unit force sensor simulates a detection value;
-simulating a detection value by a virtual model ankle joint drive unit force sensor;
-a two-dimensional force sensor x-axis direction detection value;
-a two-dimensional force sensor y-axis direction detection value;
the two-dimensional force sensor directly detects the actual contact force of the foot end and the ground in the x-axis direction;
the two-dimensional force sensor directly detects the actual contact force of the foot end and the ground in the y-axis direction;
the contact force of the foot end and the ground in the x-axis direction is obtained by resolving an uncorrected calibration curve of the joint driving unit force sensor;
the contact force of the foot end and the ground in the y-axis direction is obtained by resolving an uncorrected calibration curve of the joint driving unit force sensor;
calculating the contact force of the foot end and the ground in the x-axis direction by using the corrected force sensor calibration curve of the joint driving unit;
and resolving the corrected force sensor calibration curve of the joint driving unit to obtain the contact force between the foot end and the ground in the y-axis direction.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention aims to provide a method and a system for calibrating a leg one-dimensional force sensor of a foot type robot, which can finish the recalibration of the one-dimensional force sensor under the condition of not disassembling the force sensor.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
The hydraulic drive based foot robot has the advantages of large power-weight ratio, high response speed and the like, so the foot robot mentioned in the application can be a hydraulic drive type foot robot, but is not limited to the hydraulic drive type foot robot.
Example one
The embodiment provides a method for calibrating a leg one-dimensional force sensor of a legged robot, referring to fig. 1, the method includes:
s1, in the process of loading and unloading the acting force on the foot end, recording a two-dimensional force sensor detection value for loading, a displacement sensor detection value of each joint driving unit and a one-dimensional force sensor detection value by using a DSpace controller;
specifically, the method for loading the force on the foot end may be that a wooden cuboid fixed with a two-dimensional force sensor is slowly pulled by hand to load the force on the foot end, the force is slowly unloaded after a preset pulling force is reached, and a detection value of the two-dimensional force sensor during the loading and unloading process is recorded (step (c) ((ii) is a detection value of the two-dimensional force sensor during the loading and unloading process
And
) And displacement sensor detection values of the respective joint drive units (c)
And
) And a one-dimensional force sensor detection value (
And
)。
fig. 2 is a schematic diagram of a loading force loading experimental scheme, and as can be seen from fig. 2, the leg driving system of the legged robot consists of leg mechanical components and three joint driving units of a hip, a knee and an ankle. The one-dimensional force sensor is arranged at the tail end of a piston rod of the joint driving unit and used for detecting the stress of each joint driving unit.
As an optional implementation manner, before step S1, the method further includes: determining the mapping relation between the extension length of each joint driving unit and each joint corner according to the size and the geometric relation of the leg mechanical structure; and controlling the leg driving system to be in an initial position to be immobile by utilizing an upper computer based on the mapping relation and a kinematic model obtained by the leg mechanical structure.
The method for acquiring the kinematic model comprises the following steps: simplifying the leg mechanical structure of the legged robot to obtain parameters of a single-leg mechanical structure; determining a positive solution part of the mechanical structure of the leg according to the parameters of the mechanical structure of the single leg, wherein the positive solution part is determined according to the rotation angle theta of the hip joint O1Rotation angle theta of knee joint E2Rotation angle theta of ankle joint G3Thigh length OE, shank length EG and foot end length GI, and solving a function of the position of the foot end relative to hip joint O;
determining an inverse solution part of the leg mechanical structure according to parameters of the single-leg mechanical structure, wherein the inverse solution part is used for solving the hip joint rotation angle theta according to the position of the foot end relative to the hip joint O, the thigh length OE, the shank length EG and the foot end GI1Angle of rotation of knee joint theta2And ankle joint rotation angle theta3A function of (a); and obtaining the kinematic model according to the forward solution part and the backward solution part.
The calculation method for obtaining the kinematic model is described in detail below.
Fig. 3 is a simplified schematic diagram of a leg kinematics forward model of a leg driving system of the legged robot.
The D-H parameters of the resulting mechanical structure are shown in Table 1 with reference to FIG. 3.
TABLE 1D-H PARAMETER TABLE FOR MECHANICAL STRUCTURE OF FOOT DRIVING SYSTEM
In Table 1，ai-1Is denoted as zi-1To ziAlong xi-1The measured distance. Alpha is alphai-1Is denoted as zi-1To ziAround xi-1The angle of rotation. diIs denoted by xi-1To xiAlong ziThe measured distance. ThetaiIs denoted by xi-1To xiAround ziThe angle of rotation. z is a radical ofiDenotes xi、yiIn the coordinate system with xiAxis, yiThe axes are all perpendicular to the z-axis, i 1, 2.
According to the leg D-H coordinate system shown in figure 3, the pose relationship can be changed by connecting rods
And (4) showing.
Table 1 the D-H parameters of the main mechanical structure of the leg drive system are substituted into formula (1) and the transformation matrix between adjacent links can be obtained as follows:
the product can be obtained by multiplying the formula (2), the formula (3), the formula (4) and the formula (5) in this order
The following were used:
then the foot end position can be obtained from equation (6)
Relative position relation with hip joint O, wherein40 means no θ4The kinematics is positive solution
In the formula I1Length of thigh member,/2Length of the lower leg member, /)3Is the length of the foot member, θ1Angle of rotation of hip joint O, theta2Angle of rotation, θ, of the knee joint E3The rotation angle of the ankle joint G;
fig. 4 is a simplified inverse model diagram of leg kinematics of the leg driving system of the legged robot. In order to avoid the problem of uncertain solution when inverse kinematics is solved, the invention keeps the hip joint O, the ankle joint G and the foot end I of the leg driving system collinear.
In FIG. 4, the line on which OGI is located is parallel to x0Angle theta formed by positive direction of axis0Comprises the following steps:
in Δ OEG, angle EOG is:
the hip joint rotation angle theta can be obtained from fig. 21Comprises the following steps:
in Δ OEG, angle OEG is:
the knee joint rotation angle theta can be obtained from fig. 22Comprises the following steps:
in Δ OEG, angle OGE is:
the rotation angle theta of the ankle joint can be obtained from the attached figure 43Comprises the following steps:
the following describes a method of calculating a mapping relationship between the extension length of each joint drive unit and each joint rotation angle:
in fig. 5, the parts a) to c) are respectively a graph of the angle of the hip joint O, the knee joint E and the ankle joint G as a function of the extension of the drive unit.
Wherein, hip joint drive unit stretches out total length AB, knee joint drive unit stretches out total length CD, ankle joint drive unit stretches out total length FH and is:
in the formula I01Is the initial length of the hip drive unit, /)02Is the initial length of the knee joint drive unit,/03For the initiation of the ankle joint drive unitLength.
In fig. 5a) Δ AOB, < AOB is obtained by the cosine theorem as follows:
from the geometrical relationship in fig. 5a), the hip joint rotation angle θ can be obtained1Comprises the following steps:
the extension length Deltax of the hip joint driving unit can be obtained by the above formulap1Comprises the following steps:
in fig. 5b) Δ CDE, the cosine theorem is used to obtain ≈ CED as:
from the geometrical relationship in fig. 5b), the hip joint rotation angle θ can be obtained2Comprises the following steps:
the extension length delta x of the knee joint driving unit can be obtained by the above formulap2Comprises the following steps:
in fig. 5c) Δ FGH, the cosine theorem can be used to find that ≈ FGH is:
from the geometrical relationship in fig. 5c), the hip joint rotation angle θ can be obtained3Comprises the following steps:
the extension length delta x of the ankle joint driving unit can be obtained by the above formulap3Comprises the following steps:
after obtaining the leg mechanical structure kinematic model and the mapping relation between the extension length of each joint driving unit and each joint corner, writing a program of the leg mechanical structure kinematic model obtained in the above steps, the mapping relation between each joint corner and each joint driving unit extension length in the DSpace controller, and controlling the extension length of each driving unit of the leg by the program to enable the foot end to be positioned at the initial position
The method comprises the steps of (1) fixedly, horizontally and leftwards slowly pulling a wooden cuboid fixed with a two-dimensional force sensor by a hand to load force in the horizontal direction on a foot end, slowly unloading after the loading force reaches a preset tension, repeating for multiple times, and recording detection values of displacement sensors of all joint driving units in the loading and unloading processes (a)
And
) And the detection value of the one-dimensional force sensor of each joint driving unit (C:)
And
) And two-dimensional force sensor detection value (
And
)。
it should be noted that the initial position mentioned here is a position where the foot end is located when no force is applied to the foot end, and may be any position as long as the position corresponds to the leg drive system virtual model. Meanwhile, the loading of the horizontal force to the foot end mentioned above is only one case in the embodiment, and the application is not limited to loading the horizontal force to the foot end, and the direction of the force may be any.
S2, inputting the detection value of the displacement sensor and the detection value of the two-dimensional force sensor of each joint driving unit into a leg driving system virtual model built based on a leg mechanical structure for simulation, and recording the simulation detection value of the one-dimensional force sensor of each joint driving unit in the virtual model;
the leg driving system virtual model is a leg driving system virtual model built in Simulink/SimMechanics based on a leg mechanical structure of the legged robot.
In this embodiment, a SimMechanics connected Generation is selected by a SimMechanics Link tool in the solid works software, an assembly file of a leg mechanical structure is exported into an XML file and a plurality of STL files, then an smimmoport command is run in an MATLAB command window, and the MATLAB automatically reads the corresponding XML file and STL file and automatically constructs a virtual model of a leg driving system of a legged robot.
Prism module in virtual model: the moving pair module is used for connecting the piston rod and the driving unit cylinder body to enable the two rigid bodies to translate relatively, the moving position of the moving pair can be set through the establishment attribute, the stress of the moving pair can be measured through the sensing attribute, and the stress of the moving pair can be measured through the establishment attributeThe extension length of each joint driving unit is arranged (
And
) Detecting the simulation detection value of each joint driving unit force sensor in the virtual model by sensing attribute (
And
) (ii) a External Force and Torque module: the External Force and Torque module can directly apply interference Force to the foot end through the Force attribute, namely, a two-dimensional Force sensor detection value (through an External Force and Torque module)
And
) Inputting a leg driving system virtual model for simulation.
S3, for each joint, subtracting the detection value of the one-dimensional force sensor when the foot end is at the initial position from the detection value of the one-dimensional force sensor and the simulation detection value of the one-dimensional force sensor to obtain the actual detection value of each joint and the simulation detection value of each joint, wherein the detection value of the one-dimensional force sensor at the initial position is the detection value of the one-dimensional force sensor when the acting force is not loaded on the foot end;
s4, solving a one-dimensional force sensor calibration curve of each driving unit by using a least square method according to the linear relation between each actual detection value and each simulated detection value;
the least square method is a classic and most basic parameter identification method, and is also the most widely applied identification method, and the basic formula is as follows:
Zm＝Hmλ+Vm (25)
wherein Z ismRepresents the sampled value, HmRepresenting variable values, λ representing parameter values, VmRepresenting the sampling noise.
The idea of the least squares method is to find an estimate of λ
So that the measured values Z of each timei(i 1, …, m) and the estimated value
Determined measurement value estimation
The sum of the squares of the differences is minimal, i.e.:
according to the theorem of extreme values, we can obtain:
equation (27) can be simplified as follows:
when the number of samples m is not less than the number n of identification parameters,
full rank, then
Are present. At this time, the least squares estimate of λ is:
the estimated parameters at this time minimize the sum of squares of the deviations of the calculation formulas, and the error of the overall parameter estimation value is minimized, which is beneficial to reducing the influence caused by actual noise interference and measurement error. After a batch of samples are obtained, parameter identification is carried out according to the obtained samples, so that the batch processing method is very suitable for off-line identification, and when the sample size is sufficient and relatively accurate, a corresponding parameter estimation value can be estimated.
FIG. 6 is a schematic diagram of a leg one-dimensional force sensor calibration method of the legged robot.
In order to avoid the influence of the gravity term on the calibration curve in the leg dynamics, the detection values of the one-dimensional force sensors of the joint driving units obtained by the robot leg driving system test experiment platform in the step S1 are measured (
And
) And (S) simulated detection values of one-dimensional force sensors of the joint drive units obtained from the leg drive system virtual model in step S2
And
) And subtracting the detection value of the one-dimensional force sensor of each joint driving unit when the foot end is positioned at the initial position and is not loaded. The detection value of the force sensor of each joint driving unit of the robot leg driving system test experiment platform after the initial value is subtracted (
And
) As HmWill subtract the initial valueAnalog detection value of force sensor of each joint driving unit of the posterior leg driving system virtual model (
And
) As ZmAnd fitting calibration curves of the hip joint O, the knee joint E and the ankle joint G by using a formula (29).
S5, obtaining a correction calibration coefficient of each drive unit according to the calibration curve of each drive unit and the original calibration coefficient of each drive unit, wherein the original calibration coefficient is the calibration coefficient of the one-dimensional force sensor when the sensor leaves the factory; and calibrating the one-dimensional force sensor according to the corrected and calibrated coefficient of each driving unit.
Specifically, the original calibration coefficient of the one-dimensional force sensor of each joint driving unit is generally multiplied by the slope value of the corresponding calibration curve, and then the intercept of the corresponding calibration curve is added, so that the calibration of the one-dimensional force sensor of each joint driving unit is completed.
As an optional implementation manner, the calibration method of the present invention further includes verifying the effectiveness of the calibration method after calibrating the one-dimensional force sensor.
Fig. 7a) -c) are calibration curves of one-dimensional force sensors of the leg driving system hip joint O, knee joint E and ankle joint G of the legged robot. In the figure, a black solid line is a curve formed by sequentially connecting points which are formed by taking a detection value of a force sensor of a leg driving system performance test experiment platform as an abscissa and taking a detection curve of a joint driving unit force sensor of a virtual model as an ordinate, and a dotted line is a fitting curve obtained by processing a black curve by a least square method and is defined as a force sensor calibration curve.
As can be seen in fig. 7: the slope and intercept of the hip joint calibration curve are 2.463-7.452 respectively; the slope and intercept of the knee joint calibration curve are 2.147 and 2.703 respectively; the slope and the intercept of the ankle joint calibration curve are 2.338-0.4568 respectively; the linearity of the calibration curve fitting of each joint force sensor is good, which indicates that the original calibration coefficient of the force sensor has certain deviation.
As an optional implementation, the verifying the validity of the calibration method includes:
and when the foot end is at the initial position, respectively verifying the effectiveness of the calibration method in the x-axis direction and the y-axis direction of the foot end in the processes of quick loading/unloading and slow loading/unloading of the loading speed of the foot end.
And when the foot end is at the non-initial position, respectively verifying the effectiveness of the calibration method in the x-axis direction and the y-axis direction of the foot end in the processes of quick loading/unloading and slow loading/unloading of the loading speed of the foot end.
Specifically, in order to test whether the force sensor compensation strategy proposed by the present invention is correct, an experimental scheme is formulated as shown in table 2.
TABLE 2 test protocol
In table 2, the scheme 1, the scheme 2, the scheme 5 and the scheme 6 are to test the correction effect of the correction strategy provided by the present invention when the foot end of the leg driving system is at the initial position and the x/y axis direction of the foot end is in the processes of fast loading/unloading and slow loading/unloading, and the scheme 3, the scheme 4, the scheme 7 and the scheme 8 are to test whether the calibration method is accurate when the foot end of the leg driving system is at other positions and the x/y axis direction of the foot end is in the processes of fast loading/unloading and slow loading/unloading.
Since the static model of the leg mechanical structure and the mapping relationship between the stress of each joint driving unit and the angular torque of each joint are needed when calculating the contact force between the foot end and the ground, the method for calculating the static model of the leg mechanical structure and the mapping relationship between the stress of each joint driving unit and the angular torque of each joint will be described in detail below.
Since only the forward solution part of the statics model is used in the present application, only the calculation method of the statics forward solution part is described here.
Fig. 8 is a simplified leg statics model diagram of a leg driving system of the legged robot.
According to the principle of work deficiency, the following can be obtained:
τ＝JT(q)F (30)
wherein F represents the leg foot end stress (N), JT(q) represents the force Jacobian of the leg and τ represents the moment applied to the joint.
In the formula (30), the force of the leg is Jacobian JT(q) is as follows:
the formula (31) can be substituted for the formula (30):
through the moment that knee joint E and ankle joint G receive, ask the foot end atress, can obtain statics and just solve:
the following describes a method for calculating a mapping relationship between the stress of each joint driving unit and the rotational angle torque of each joint specifically:
in the attached figure 9, the parts a) to c) are graphs of the relation between the moment borne by the hip joint O, the knee joint E and the ankle joint G and the force borne by the driving unit.
In fig. 9a) Δ AOB, using the cosine theorem coupled equation (15), we can find the angle ABO as:
the detection signal of the force sensor of the hip joint driving unit can be obtained from the formula (34)ΔFs1Moment tau borne by hip joint1The mapping relation of (1) is as follows:
in fig. 9b) Δ CDE, the cosine theorem can be used to obtain ≈ DCE as:
the knee joint drive unit force sensor detection signal DeltaF can be obtained from the equation (36)s2Moment tau borne by knee joint2The mapping relation of (1) is as follows:
in fig. 9c) Δ FGH, the cosine theorem can be used to find out that ═ GFH is:
as can be seen from the equation (38), the ankle joint drive unit force sensor detection signal DeltaFs3Moment tau borne by ankle joint G3The mapping relation of (1) is as follows:
fig. 10 to 17 show the loading direction contact force curves in the embodiments 1 to 8, respectively. The two-dimensional force sensor directly detects the actual contact force between the foot end and the ground
Is a broken line, a calibration curve and a statics model of a leg mechanical structure are calibrated according to the force sensor of the joint driving unit after correction, the stress of each joint driving unit and each jointThe mapping relation between the angular torques, the contact force between the foot end and the ground obtained by calculation
Is a black solid line, and the contact force between the foot end and the ground is obtained by resolving the mapping relation between the force applied to each joint driving unit and the rotation angle torque of each joint
The initial value of the contact force calculated by the method is not zero in order to avoid the dynamics of the leg mechanical structure and the zero drift of the force sensor, and the effect of the compensation strategy provided by the text is convenient to observe, the contact force detected and calculated by the method is subjected to zero adjustment. It can be seen from fig. 10 to 17 that, after the calibration method provided by the present invention is added, the deviation between the actual foot end contact force and the corrected foot end contact force in the loading direction is almost zero, which indicates that the calibration method of the one-dimensional force sensor provided by the present invention is applicable to different positions.
Example two
The present embodiment provides a system for calibrating a leg one-dimensional force sensor of a legged robot, please refer to fig. 18, which includes:
the sensor detection value acquisition module is used for recording a two-dimensional force sensor detection value for loading, a displacement sensor detection value of each joint driving unit and a one-dimensional force sensor detection value by using an upper computer in the process of loading and unloading acting force on the foot end;
the simulation module is used for inputting the detection value of the displacement sensor and the detection value of the two-dimensional force sensor into a leg driving system virtual model built based on a leg mechanical structure for simulation, and recording the simulation detection value of the one-dimensional force sensor of each joint driving unit in the virtual model;
the variable value and sampling value acquisition module is used for subtracting the one-dimensional force sensor detection value when the foot end is at the initial position from the one-dimensional force sensor detection value and the one-dimensional force sensor simulation detection value of each joint to obtain the actual detection value of each joint and the one-dimensional force sensor simulation detection value when the foot end is not loaded with acting force;
the calibration curve acquisition module is used for solving a one-dimensional force sensor calibration curve of each driving unit by using a least square method according to the linear relation between each actual detection value and each simulated detection value;
and the calibration module is used for obtaining a correction calibration coefficient of each driving unit according to the calibration curve of each driving unit and the original calibration coefficient of each driving unit, wherein the original calibration coefficient is the calibration coefficient of the one-dimensional force sensor when leaving the factory, and calibrating the one-dimensional force sensor according to the correction calibration coefficient of each driving unit.
As an optional implementation, the system further comprises: and the verification module is used for verifying the effectiveness of the calibration method after the coefficient calibrated by each driving unit is obtained.
The above-mentioned embodiments are merely illustrative of the preferred embodiments of the present invention, and do not limit the scope of the present invention, and various modifications and improvements of the technical solution of the present invention by those skilled in the art should fall within the protection scope defined by the claims of the present invention without departing from the spirit of the present invention.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.
Claims (10)
1. A method for calibrating a leg one-dimensional force sensor of a foot robot is characterized by comprising the following steps:
in the process of loading and unloading the acting force of the foot end, recording a two-dimensional force sensor detection value for loading, a displacement sensor detection value of each joint driving unit and a one-dimensional force sensor detection value by using an upper computer;
inputting the detection value of the displacement sensor and the detection value of the two-dimensional force sensor into a leg driving system virtual model built based on a leg mechanical structure for simulation, and recording the simulation detection value of the one-dimensional force sensor of each joint driving unit in the virtual model;
for each joint, subtracting the detection value of the one-dimensional force sensor when the foot end is at the initial position from the detection value of the one-dimensional force sensor and the simulation detection value of the one-dimensional force sensor to obtain the actual detection value of each joint and the simulation detection value of each joint, wherein the detection value of the one-dimensional force sensor at the initial position is the detection value of the one-dimensional force sensor when acting force is not loaded on the foot end;
solving a one-dimensional force sensor calibration curve of each driving unit by using a least square method according to a linear relation between each actual detection value and each simulated detection value;
obtaining a correction calibration coefficient of each driving unit according to the calibration curve of each driving unit and an original calibration coefficient of each driving unit, wherein the original calibration coefficient is a calibration coefficient of the one-dimensional force sensor when the sensor leaves a factory;
and calibrating the one-dimensional force sensor according to the coefficient corrected and calibrated by each driving unit.
2. The method according to claim 1, wherein the recording of the two-dimensional force sensor detection value, the displacement sensor detection value of each joint driving unit and the one-dimensional force sensor detection value by using an upper computer during the process of loading and unloading the acting force to the foot end specifically comprises:
and slowly pulling the wooden cuboid fixed with the two-dimensional force sensor by hand to load force on the foot end, slowly unloading after preset tension is reached, and recording the detection value of the two-dimensional force sensor, the detection value of the displacement sensor of each joint driving unit and the detection value of the one-dimensional force sensor in the loading and unloading processes.
3. The method according to claim 1 or 2, characterized in that the method further comprises: before loading acting force on the foot end and unloading, determining a mapping relation between the extension length of each joint driving unit and each joint corner according to the size and the geometric relation of a leg mechanical structure;
and controlling the leg driving system to be in an initial position to be immobile by utilizing an upper computer based on the mapping relation and a kinematic model obtained by the leg mechanical structure.
4. The method according to claim 3, characterized in that the method of obtaining the kinematic model comprises in particular:
simplifying the leg mechanical structure of the legged robot to obtain parameters of a single-leg mechanical structure;
determining a positive solution part of the leg mechanical structure according to the parameters of the single-leg mechanical structure; the positive solution part is a function for solving the position of the foot end relative to the hip joint according to the rotation angle of the hip joint, the rotation angle of the knee joint, the rotation angle of the ankle joint, the length of the thigh, the length of the shank and the length of the foot end;
determining an inverse solution part of the leg mechanical structure according to the parameters of the single-leg mechanical structure; the inverse solution part is used for solving functions of hip joint rotation angle, knee joint rotation angle and ankle joint rotation angle according to the position of the foot end relative to the hip joint, thigh length, shank length and foot end length;
and obtaining the kinematic model according to the forward solution part and the backward solution part.
5. The method of claim 1, wherein the leg drive system virtual model is built by a SimMechanics Link tool in Solideworks software.
6. The method of claim 1, wherein the calibrating further comprises verifying the validity of the calibrating method after the calibrating the one-dimensional force sensor.
7. The method according to claim 6, wherein the verifying the validity of the calibration method specifically comprises:
and when the foot end is at the initial position, respectively verifying the effectiveness of the calibration method in the x-axis direction and the y-axis direction of the foot end in the processes of quick loading/unloading and slow loading/unloading of the loading speed of the foot end.
8. The method of claim 6, wherein the verifying the validity of the calibration method further comprises:
and when the foot end is at the non-initial position, respectively verifying the effectiveness of the calibration method in the x-axis direction and the y-axis direction of the foot end in the processes of quick loading/unloading and slow loading/unloading of the loading speed of the foot end.
9. A leg one-dimensional force sensor calibration system of a legged robot, the system comprising:
the sensor detection value acquisition module is used for recording a two-dimensional force sensor detection value for loading, a displacement sensor detection value of each joint driving unit and a one-dimensional force sensor detection value by using an upper computer in the process of loading and unloading acting force on the foot end;
the simulation module is used for inputting the detection value of the displacement sensor and the detection value of the two-dimensional force sensor into a leg driving system virtual model built based on a leg mechanical structure for simulation, and recording the simulation detection value of the one-dimensional force sensor of each joint driving unit in the virtual model;
the variable value and sampling value acquisition module is used for subtracting the detection value of the one-dimensional force sensor when the foot end is at the initial position from the detection value of the one-dimensional force sensor and the simulation detection value of the one-dimensional force sensor for each joint to obtain the actual detection value of each joint and the simulation detection value of each joint, wherein the detection value of the one-dimensional force sensor at the initial position is the detection value of the one-dimensional force sensor when the acting force is not loaded on the foot end;
the calibration curve acquisition module is used for solving a one-dimensional force sensor calibration curve of each driving unit by using a least square method according to the linear relation between each actual detection value and each simulated detection value;
and the calibration module is used for obtaining a correction calibration coefficient of each driving unit according to the calibration curve of each driving unit and an original calibration coefficient of each driving unit, wherein the original calibration coefficient is a calibration coefficient of the one-dimensional force sensor when the one-dimensional force sensor leaves a factory, and the one-dimensional force sensor is calibrated according to the correction calibration coefficient of each driving unit.
10. The system of claim 9, further comprising:
and the verification module is used for verifying the effectiveness of the calibration method after the coefficient calibrated by each driving unit is obtained.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110354203.0A CN113091999A (en) | 2021-04-01 | 2021-04-01 | Foot type robot leg one-dimensional force sensor calibration method and system |
| NL2031262A NL2031262B1 (en) | 2021-04-01 | 2022-03-11 | Method and system for calibrating one-dimensional force sensor for legged robot's leg |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110354203.0A CN113091999A (en) | 2021-04-01 | 2021-04-01 | Foot type robot leg one-dimensional force sensor calibration method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113091999A true CN113091999A (en) | 2021-07-09 |
Family
ID=76672353
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110354203.0A Pending CN113091999A (en) | 2021-04-01 | 2021-04-01 | Foot type robot leg one-dimensional force sensor calibration method and system |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN113091999A (en) |
| NL (1) | NL2031262B1 (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115562076A (en) * | 2022-11-29 | 2023-01-03 | 北京路凯智行科技有限公司 | Simulation system, method and storage medium for unmanned mine car |
| CN117804666A (en) * | 2024-01-09 | 2024-04-02 | 浙江大学 | Leg displacement/force sensor calibration device and method for hydraulic foot robot |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050156551A1 (en) * | 2002-04-26 | 2005-07-21 | Honda Giken Kogyo Kabushiki Kaisha | System for estimating attitude of leg type moving robot itself |
| JP2006239848A (en) * | 2005-03-07 | 2006-09-14 | Toyota Motor Corp | Offset adjusting system of walking robot having supporting leg and offset adjusting method |
| US20130131865A1 (en) * | 2011-11-22 | 2013-05-23 | Disney Enterprises, Inc. | Kinematic and Dynamic Calibration Methods for Legged Robots with Force-Controlled Joints |
| CN107214698A (en) * | 2017-05-05 | 2017-09-29 | 上海交通大学 | Hexapod Robot joint angles scaling method based on body nodal point displacement correction |
| CN108621161A (en) * | 2018-05-08 | 2018-10-09 | 中国人民解放军国防科技大学 | Method for estimating body state of foot type robot based on multi-sensor information fusion |
| CN110202584A (en) * | 2019-07-11 | 2019-09-06 | 燕山大学 | Legged type robot leg exercise modification method and system with semi-cylindrical foot end |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6996177B2 (en) * | 2017-09-11 | 2022-01-17 | セイコーエプソン株式会社 | Robot, offset correction device for force sensor, and robot control method |
| CN111189577B (en) * | 2020-01-16 | 2022-01-07 | 腾讯科技（深圳）有限公司 | Sensor calibration and data measurement method, device, equipment and storage medium |
-
2021
- 2021-04-01 CN CN202110354203.0A patent/CN113091999A/en active Pending
-
2022
- 2022-03-11 NL NL2031262A patent/NL2031262B1/en active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050156551A1 (en) * | 2002-04-26 | 2005-07-21 | Honda Giken Kogyo Kabushiki Kaisha | System for estimating attitude of leg type moving robot itself |
| JP2006239848A (en) * | 2005-03-07 | 2006-09-14 | Toyota Motor Corp | Offset adjusting system of walking robot having supporting leg and offset adjusting method |
| US20130131865A1 (en) * | 2011-11-22 | 2013-05-23 | Disney Enterprises, Inc. | Kinematic and Dynamic Calibration Methods for Legged Robots with Force-Controlled Joints |
| CN107214698A (en) * | 2017-05-05 | 2017-09-29 | 上海交通大学 | Hexapod Robot joint angles scaling method based on body nodal point displacement correction |
| CN108621161A (en) * | 2018-05-08 | 2018-10-09 | 中国人民解放军国防科技大学 | Method for estimating body state of foot type robot based on multi-sensor information fusion |
| CN110202584A (en) * | 2019-07-11 | 2019-09-06 | 燕山大学 | Legged type robot leg exercise modification method and system with semi-cylindrical foot end |
Non-Patent Citations (1)
| Title |
|---|
| 马国梁: "液压驱动型足式机器人腿部模型修正与阻抗新构型研究", 《中国优秀博硕士学位论文全文数据库（硕士） 信息科技辑》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115562076A (en) * | 2022-11-29 | 2023-01-03 | 北京路凯智行科技有限公司 | Simulation system, method and storage medium for unmanned mine car |
| CN115562076B (en) * | 2022-11-29 | 2023-03-14 | 北京路凯智行科技有限公司 | Simulation system, method and storage medium for unmanned mine car |
| CN117804666A (en) * | 2024-01-09 | 2024-04-02 | 浙江大学 | Leg displacement/force sensor calibration device and method for hydraulic foot robot |
Also Published As
| Publication number | Publication date |
|---|---|
| NL2031262A (en) | 2022-10-06 |
| NL2031262B1 (en) | 2023-10-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TWI642523B (en) | Gravity compensation method for load estimation of mechanical arm and gravity compensation system for load estimation | |
| CN105682865B (en) | Determine the method and system of at least one characteristic of mechanical arm | |
| JP5550468B2 (en) | Force sensor calibration method | |
| CN113091999A (en) | Foot type robot leg one-dimensional force sensor calibration method and system | |
| Zhou et al. | Method of designing a six-axis force sensor for stiffness decoupling based on Stewart platform | |
| Ba et al. | A novel one-dimensional force sensor calibration method to improve the contact force solution accuracy for legged robot | |
| JP5242664B2 (en) | In vivo tension calibration in tendon driven manipulators | |
| CN112767493B (en) | Machine vision calibration method for kinematic parameters of Stewart platform | |
| CN109483528A (en) | The offset correction device of robot and force snesor | |
| CN115307816A (en) | Calibration test system and calibration method for harmonic flexible gear torque sensor | |
| JP6485620B2 (en) | Robot control system and robot control method | |
| Phan et al. | A novel 6-DOF force/torque sensor for COBOTs and its calibration method | |
| CN111693040A (en) | Mechanical arm collision detection method based on series elastic driver | |
| Li et al. | A novel piezoelectric 6-component heavy force/moment sensor for huge heavy-load manipulator's gripper | |
| CN109773827B (en) | Method and system for load estimation gravity compensation of mechanical arm | |
| JPH07205075A (en) | Weight compensation method of end effector at force control robot | |
| Kim | Development of 6-axis force/moment sensor for a humanoid robot's foot | |
| Bartsch et al. | Small insect measurements using a custom MEMS force sensor | |
| CN115107074A (en) | Clamping device, robot and force information sensing method | |
| WO2020170770A1 (en) | Detection device and sensor calibration method | |
| Le et al. | In-Situ Calibration for Load Cells in 3d Printed Bipedal Robot Using 3D Modeling in Computer-Aided Design Environment | |
| Lee et al. | Torque sensor calibration using virtual load for a manipulator | |
| Wang et al. | Investigations on nonlinear deformation phenomena in a force calibration system | |
| JPS63109344A (en) | Calibration system of force detector | |
| Huo et al. | Humanoid arm with the integrated serial elastic actuator |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210709 |
|
| RJ01 | Rejection of invention patent application after publication |