Unmanned metamorphic vehicle parameter optimization method based on multi-configuration grading and integration algorithm

Abstract

The invention discloses an unmanned metamorphic vehicle parameter optimization method based on a multi-configuration grading and integration algorithm. The method comprises the following steps: 1, establishing an unmanned metamorphic vehicle model; 2, establishing evaluation indexes of the unmanned metamorphic vehicle in different states; 3, screening parameters by adopting an SOBOL sensitivity analysis method; 4, formulating a constraint domain of each parameter according to different-state performance and structural limitation; and 5, carrying out grading and integrated optimization on the screened parameters according to a sensitivity analysis result and a constraint domain. According to the method, the parameters which greatly influence the motion performance of the unmanned metamorphic vehicle can be screened and optimized, the optimization complexity is reduced, and the optimization time is shortened, so that the stability of the multi-configuration motion of the unmanned metamorphic vehicle can be improved, and the motion energy consumption is effectively reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned metamorphic vehicle optimization design, and in particular relates to an unmanned metamorphic vehicle parameter optimization method based on multi-configuration classification and integration algorithm. Background Art

[0002] The unmanned metamorphic vehicle is a new type of deformable robot designed by introducing a metamorphic mechanism based on the structure of a car. It can change its configuration according to different road conditions to achieve wheeled driving or legged walking. With the continuous expansion of application scenarios, the use environment of the unmanned metamorphic vehicle is becoming more and more complex, and higher requirements are also placed on the stability and energy consumption of its different configurations.

[0003] In modern scientific research, many research objects are large and complex, and the performance and dynamic changes of the system involve many influencing factors. Specifically in terms of optimization design, the established optimization model contains a large number of design variables. On the one hand, these design variables may have different degrees of influence on the target performance of the system. For example, for the unmanned metamorphic vehicle involved in the present invention, for different motion performance requirements in different configurations, some design variables have a high degree of influence, some design variables have a low degree of influence, and some design variables have zero correlation. On the other hand, these design variables are not only numerous, but also may have different properties. Using different optimization strategies for these design variables will result in completely different optimization results. General multi-design variable optimization strategies all adopt sequential optimization design methods, that is, first optimize the design system structure parameters, and then design the relevant control system and optimize the control parameters. This type of method has some defects:

[0004] 1) The performance requirements of the deformable robot in different configurations are not considered. For example, when a robot has two configurations, the above algorithm only optimizes one configuration first and then optimizes the other configuration.

[0005] 2) The coupling effect of structural parameters and control parameters on system performance is not taken into account.

[0006] At present, there is no optimization program that combines the structural parameter optimization design with the control parameter optimization design process under multiple configurations. In addition, the existing optimization methods optimize all variables and constraints at the same level without considering the influence of the logical order of the design and the coupling relationship between the variables, resulting in low optimization efficiency and the need for multiple iterations of design. Therefore, how to achieve rapid convergence of optimization results and coordinated matching of optimization targets under the mutual coupling between control parameters and structural parameters and the comprehensive influence on multiple optimization targets has become a technical problem in realizing the multi-configuration optimization of unmanned cellular vehicles. Summary of the invention

[0007] The purpose of the present invention is to solve the shortcomings of the above-mentioned prior art and to provide a parameter optimization method for an unmanned metamorphic vehicle based on multi-configuration classification and integrated algorithm, so as to screen and optimize the parameters that have a greater impact on the motion performance of the unmanned metamorphic vehicle, reduce the optimization complexity, and reduce the optimization time, thereby improving the stability of the multi-configuration motion of the unmanned metamorphic vehicle and effectively reducing the energy consumption of motion.

[0008] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0009] The present invention discloses a parameter optimization method for an unmanned metamorphic vehicle based on multi-configuration classification and integrated algorithm, which is applied to the unmanned metamorphic vehicle in the process of driving in a car state, walking in a human-like state, and reconstructing the supporting state between the car state and the human-like state. The unmanned metamorphic vehicle comprises: a front body, a rear body and a sliding mode controller, and two mechanical legs, a center of mass adjustment mechanism, and a lifting mechanism are arranged on the bottom plate of the rear body; wherein each mechanical leg comprises: three leg joint motors, a thigh, a calf and a supporting foot, wherein a knee joint motor is arranged between the thigh and the calf, an ankle joint motor is arranged between the calf and the supporting foot, and a hip joint motor is arranged between the thigh and the bottom plate of the rear body; the characteristic is that the parameter optimization method for the unmanned metamorphic vehicle comprises the following steps:

[0010] Step 1: Establish an unmanned cell-changing vehicle model;

[0011] Step 1.1: Establish a basic coordinate system on the horizontal plane where the support foot of the unmanned vehicle is located. The origin of the basic coordinate system is set at the middle position of the support foot. The positive direction of the axis is the forward direction of the unmanned vehicle, and the basic coordinate system The positive direction of the axis is perpendicular to the horizontal plane and upward. The positive direction of the axis is perpendicular to Axis and The axis is in the plane and points to the outside of the vehicle;

[0012] Step 1.2, establish the kinematic model and Lagrangian dynamics model of the unmanned metamorphic vehicle reconstruction and walking;

[0013] Step 2: Based on the kinematic model and Lagrangian dynamics model, establish the evaluation index of the unmanned metamorphic vehicle in different states;

[0014] Step 2.1: According to formula (1), the evaluation index of the reconstruction stability of the unmanned metamorphic vehicle in the support state is obtained: ;

[0015] (1)

[0016] In formula (1), The zero moment point of the unmanned vehicle in the support state is The value of the axis direction; represents the total time of reconstruction, Represents any moment of the reconstruction process;

[0017] Step 2.2: According to formula (2), the evaluation index of the walking stability of the unmanned metamorphosis vehicle in the humanoid state is obtained: ;

[0018] (2)

[0019] In formula (2), The zero moment point of the unmanned metamorphosis vehicle in the humanoid state is The value of the axis direction; Indicates the support period of a single mechanical leg; Represents a single mechanical leg at any moment in the support process;

[0020] Step 2.3: According to formula (3), the total reconstruction time of the leg joint motor of the unmanned metamorphosis vehicle is obtained: Evaluation index of internal average power ;

[0021] (3)

[0022] In formula (3), is the torque of the i-th leg joint motor in the two mechanical legs, express The angular velocity of the i-th joint of the two robotic legs at the moment; is the total number of leg joint motors, and =6;

[0023] Step 2.4: According to formula (4), the leg joint motor of the unmanned metamorphosis vehicle is obtained in a gait cycle during human-like walking. Evaluation index of internal average power ;

[0024] (4)

[0025] In formula (4), Represents a gait cycle At any time while walking inside, express The angular velocity of the i-th joint of the two robotic legs at the moment;

[0026] Step 2.5: According to formula (5), the evaluation index of the reconstruction control quality of the unmanned metamorphic vehicle in the support state is obtained: ;

[0027] (5)

[0028] In formula (5), They are Constantly adjust the tracking error and error derivative of the slider in the center of mass mechanism, is the sliding surface parameter of the sliding mode controller;

[0029] Step 3: Use SOBOL sensitivity analysis method to select parameters;

[0030] Step 3.1: Define the variable structural parameters of the unmanned metamorphosis vehicle, including: the installation height of the hip joint motor , Installation height of lifting mechanism , Adjust the installation height of the center of mass mechanism , longitudinal installation position of hip joint motor , Longitudinal installation position of lifting mechanism , adjust the longitudinal installation position of the center of mass mechanism , the length of the calf And the length of the thigh ;

[0031] Define the variable posture parameters of the unmanned vehicle, including: the landing position of the support foot ;

[0032] The unmanned metamorphic vehicle is composed of variable structural parameters and variable posture parameters. parameters, and any parameter is recorded as , is the number of parameters; z=7;

[0033] Step 3.2, calculate the bth parameter The first-order influence index of the yth evaluation index of the unmanned vehicle ;

[0034] Step 3.3: Calculate the bth parameter according to formula (7): The total effect index of the yth evaluation index of the unmanned metamorphosis vehicle , thus the first-order impact index Total effect index Composition of the bth parameter Sensitivity analysis results of

[0035] (7)

[0036] In formula (7), Indicates that only the bth parameter is included The expectation of the corresponding y-th evaluation index value; Except for the bth parameter The variance calculated for the yth evaluation index value corresponding to all parameters except ; Indicates the bth parameter The corresponding y-th evaluation index value;

[0037] Step 4: According to different state performance and structural constraints, formulate the constraint domain of each parameter;

[0038] Step 4.1: According to formula (8), construct the unmanned metamorphosis vehicle in the humanoid state The constraint domain of

[0039] (8)

[0040] In formula (8), Indicates the minimum value of leg length ratio, Indicates the maximum value of leg length ratio;

[0041] Step 4.2: According to equations (9) to (11), construct the unmanned metamorphic vehicle in the car state The constraint domain of

[0042] (9)

[0043] (10)

[0044] (11)

[0045] In formula (9) to formula (11), Indicates that the hip joint motor is in the longitudinal installation position The minimum value of Indicates that the hip joint motor is installed in the longitudinal position The maximum value of Indicates the hip joint motor at the installation height The minimum value of Indicates the hip joint motor at the installation height The maximum value of Indicates the installation height of the lifting mechanism The minimum value of Indicates the installation height of the lifting mechanism The maximum value of

[0046] Step 4.3: According to equations (12) to (15), construct the unmanned metamorphosis vehicle in the support state and reconstruction process , The constraint domain of

[0047] (12)

[0048] (13)

[0049] (14)

[0050] (15)

[0051] In formula (12) to formula (15), is the distance from the center of the left ankle joint motor to the supporting foot; represents the sliding surface parameters of the sliding mode controller The minimum value of represents the first parameter of the reaching law of the sliding mode controller The minimum value of The second parameter of the reaching law of the sliding mode controller is The minimum value of The second parameter of the reaching law of the sliding mode controller is The maximum value of

[0052] Step 5: According to The sensitivity analysis results of each parameter are used to formulate a hierarchical optimization strategy;

[0053] Step 5.1: Evaluation index of walking stability Evaluation index of average walking power As two optimization objectives, They are respectively used as the first-level optimization variables, and the artificial hummingbird algorithm is used to perform the first-level optimization on the first-level optimization variables to obtain the optimal thigh and calf lengths of each mechanical leg of the unmanned metamorphic vehicle;

[0054] Step 5.2: Based on the results of the first-level optimization strategy, reconstruct the stability evaluation index , Evaluation index of reconstruction average power and reconstruction control quality evaluation index To optimize the goal, the unmanned vehicle , And the sliding surface parameters of the sliding mode controller , the first parameter of the reaching law and the second parameter of the reaching law As the second-level optimization variables, the NSGA-Ⅲ algorithm is used to perform the second-level multi-objective integrated optimization on the second-level optimization variables, so as to obtain the optimal variable structure parameter values ​​and variable posture parameter values ​​of the unmanned vehicle, as well as the optimal sliding surface parameter values ​​of the sliding mode controller, the optimal first parameter value of the reaching law and the optimal second parameter value of the reaching law.

[0055] The unmanned metamorphic vehicle parameter optimization method based on multi-configuration classification and integrated algorithm described in the present invention is also characterized in that step 3.2 includes:

[0056] Step 3.2.1, define variable j and initialize j=1;

[0057] Step 3.2.2: Substitute any j parameters from the parameters into the kinematic model and the Lagrangian dynamics model, and use equations (1) to (4) to calculate the four evaluation index values ​​corresponding to the j parameters, so as to obtain the four evaluation index values ​​corresponding to each j parameter, and calculate their variances;

[0058] Step 3.2.3, after assigning j+1 to j, return to step 3.2.2 until j>z; thus, the total variance is obtained by summing up the variance calculated each time. ;

[0059] Step 3.2.4: Calculate the bth parameter using formula (6): The first-order influence index of the yth evaluation index of the unmanned vehicle ;

[0060] (6)

[0061] In formula (6), Indicates the bth parameter The variance calculated for the corresponding y-th evaluation index value, Indicates the bth parameter The corresponding y-th evaluation index value, y=1,2,3,4; Indicates that the bth parameter is The expectation of the y-th evaluation index value corresponding to all parameters except .

[0062] Further, the step 5.1 includes:

[0063] Step 5.1.1, define the population size as M, the dimension as d, and the maximum number of iterations , the current number of iterations is N, and N=1 is initialized;

[0064] Let the mth hummingbird food source The position is the length of the thigh and calf of any mechanical leg of the mth group of unmanned metamorphic vehicles in the humanoid state; define and initialize the optimal thigh and calf length as ;

[0065] Step 5.1.2, initialize m=1;

[0066] Step 5.1.3: Calculate the length of the mth group of thighs and calves in the Nth iteration using equations (2) and (4): Evaluation index of walking stability Evaluation index of average walking power , thus obtaining The fitness value of = + ; and with optimal thigh and calf length The fitness values ​​of the two groups are compared, and the set of thigh and calf lengths corresponding to the smaller value is taken as ;

[0067] Step 5.1.4, after assigning m+1 to m, return to step 5.1.3 and execute sequentially until m > M;

[0068] Step 5.1.5: After assigning N+1 to N, return to step 5.1.2 and execute sequentially until N > So far, the optimal thigh and calf length of each mechanical leg of the unmanned cell car is output .

[0069] Further, the step 5.2 includes:

[0070] Step 5.2.1. Define the maximum number of iterations , the current number of iterations is V, and V=1 is initialized to determine the selection, crossover operator and mutation operator;

[0071] Randomly generate a V-th generation parent population containing A individuals ,and Each individual in contains 7-dimensional genes, i.e., the second-level optimization variables;

[0072] Step 5.2.2: For the Vth generation parent population After selection, crossover, and mutation, a V-generation offspring population containing A individuals is generated. ; V generation parent population and the V generation progeny population After merging, we get the Vth generation merged population containing 2A individuals. ;

[0073] Step 5.2.3: Define variable a as the Vth generation merged population The ath individual in , and initialize a=1;

[0074] Step 5.2.4: Use equations (1), (3) and (5) to calculate the Vth generation combined population Evaluation index of the reconstruction stability of the ath individual , Evaluation index of reconstruction average power and reconstruction control quality evaluation index , thus obtaining the Vth generation merged population The fitness value of the ath individual in ;

[0075] Step 5.2.5, after assigning a+1 to a, return to step 5.2.4 and execute sequentially until a > 2A;

[0076] Step 5.2.6, according to The fitness value of each individual in the Vth generation merged population The individuals in the non-dominated sort are sorted, and then according to the non-dominated relationship, a number of individuals A are selected to form the V+1 generation parent population. ;

[0077] Step 5.2.7, after assigning V+1 to V, return to step 5.2.2 and execute sequentially until V > So far, output The Pareto solution set in the iterative parent population is used as the optimal parameter value for the second-level multi-objective integrated optimization of the unmanned metamorphic vehicle.

[0078] An electronic device of the present invention includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the unmanned metamorphic vehicle parameter optimization method, and the processor is configured to execute the program stored in the memory.

[0079] The present invention provides a computer-readable storage medium, and the computer-readable storage medium stores a computer program, which is characterized in that when the computer program is run by a processor, the steps of the unmanned metamorphic vehicle parameter optimization method are executed.

[0080] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0081] 1. The present invention first optimizes the design variables that affect the humanoid walking motion performance, and then optimizes the design variables that affect the automobile driving motion and the reconstruction motion performance between different states. This hierarchical optimization method optimizes the design variables with greater influence in a targeted manner according to different optimization goals, reduces the optimization design time by reducing the number of optimization design variables, and significantly improves the optimization time efficiency. In addition, this hierarchical optimization method can shield the optimization interference of design variables with lower correlation, thereby further improving the accuracy of the optimization design.

[0082] 2. In the second level optimization stage of hierarchical optimization, the present invention innovatively combines variable structural parameters, variable posture parameters and sliding mode controller parameter optimization design, comprehensively considers the coupling effects of different types of parameters on the driving performance of the unmanned vehicle state and the multi-objective optimization of the reconstructed motion performance between different states, and performs multi-parameter integrated optimization design. The integrated optimization design can effectively avoid the optimized parameter value that can only meet the motion performance requirements of one of the configurations, while increasing the optimization calculation time a little, thereby obtaining the optimal parameter value that can meet the motion performance requirements of different configurations. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] Figure 1 This is a simplified diagram of the unmanned cell-changing vehicle mechanism movement;

[0084] Figure 2 This is the quality principle diagram of sliding mode control;

[0085] Figure 3 This is a schematic diagram of the parameters of the unmanned cell-changing vehicle;

[0086] Figure 4 This is the sensitivity analysis result diagram;

[0087] Figure 5 This is the flow chart of the artificial hummingbird algorithm;

[0088] Figure 6 This is the flow chart of NSGA-Ⅲ. DETAILED DESCRIPTION

[0089] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments:

[0090] The unmanned metamorphic vehicle involved in the present invention comprises: a front and rear body and a sliding mode controller, and two mechanical legs, a mass center adjustment mechanism, and a lifting mechanism are arranged on the bottom plate of the rear body; wherein each mechanical leg comprises: three leg joint motors, a thigh, a calf, and a supporting foot, wherein a knee joint motor is arranged between the thigh and the calf, an ankle joint motor is arranged between the calf and the supporting foot, and a hip joint motor is arranged between the thigh and the bottom plate of the rear body. The unmanned metamorphic vehicle involves problems such as motion stability and energy consumption during driving in a car state, walking in a humanoid state, and reconstructing motion, so it is necessary to control and optimize the multi-objective performance of the motion process. In an embodiment, the present invention first establishes a kinematic model and a Lagrangian dynamics model for reconstruction and walking of the unmanned metamorphic vehicle, and then uses the SOBOL sensitivity analysis method to screen parameters that have a greater impact on the motion performance of different states, and establishes a hierarchical optimization method for different states, thereby significantly reducing the computational complexity of subsequent optimization. In the first-level optimization stage, the artificial hummingbird algorithm is used to optimize the leg length structural parameters to improve the stability of humanoid walking and reduce energy consumption. On this basis, the second level optimization is carried out to optimize the stability performance of steady-state steering driving in the vehicle state by setting the constraint domain of the parameters; and in the reconstruction process from the vehicle state to the humanoid state, a sliding mode stability controller is designed for stability control. Considering the coupling influence of variable structural parameters, variable posture parameters and sliding mode controller parameters on the reconstructed motion performance of the system, the NSGA-Ⅲ algorithm is used to integrate and optimize the variable structural parameters, variable posture parameters and sliding mode controller parameters. This parameter optimization method can meet the stability and energy consumption requirements of each configuration state of the unmanned metamorphic vehicle. Specifically, a parameter optimization method for an unmanned metamorphic vehicle based on multi-configuration classification and integrated algorithm includes the following steps:

[0091] Step 1: Establish an unmanned cell-changing vehicle model;

[0092] Step 1.1: Establish a basic coordinate system on the horizontal plane where the support foot of the unmanned vehicle is located. The origin of the basic coordinate system is set at the middle position of the support foot. The positive direction of the axis is the forward direction of the unmanned vehicle, and the basic coordinate system The positive direction of the axis is perpendicular to the horizontal plane and upward. The positive direction of the axis is perpendicular to Axis and The axis is in the plane and points to the outside of the vehicle;

[0093] The unmanned metamorphic vehicle can be simplified into a multi-link system. The reconstruction and walking of the unmanned metamorphic vehicle are essentially the movement of each rod around the joint. Specifically, Figure 1 As shown. Figure 1 middle, They represent the supporting foot, calf, thigh and equivalent rods of the rear body respectively. They respectively represent the lower rod of the V-shaped lifting rod in the lifting mechanism, the equivalent rod of the front body, the electric push rod in the lifting mechanism and the upper rod of the V-shaped lifting. and They are used to adjust the horizontal and vertical displacements of the slider of the mass center mechanism. Each rod has an attached coordinate system fixed to it. Counterclockwise is positive, where Rotation around the X axis is called roll motion. Rotation around the Y axis is called pitch motion. Rotation around the Z axis is called yaw motion. The rotational freedom is indicated by an arrow. The right-hand rule is used to determine , and positive and negative.

[0094] Step 1.2, establish the kinematic model and Lagrangian dynamics model of the unmanned metamorphic vehicle reconstruction and walking;

[0095] First, the homogeneous transformation matrix of the positional relationship between adjacent coordinate systems is calculated;

[0096] The motor part of the left ankle joint has two rotational degrees of freedom. Describe the coordinate system in The homogeneous transformation matrix for roll rotation around the X-axis is:

[0097] (1)

[0098] In the base coordinate system The attached coordinate system is described in The homogeneous transformation matrix for pitch rotation around the Y axis is:

[0099] (2)

[0100] Attached coordinate system The origin of the coordinate system and the basic coordinate system The distance between the origin of the coordinate system can be described as;

[0101] (3)

[0102] In formula (3), is the distance from the center of the left ankle joint motor to the supporting foot. Then, in the basic coordinate system The attached coordinate system is described in The homogeneous transformation matrix of is:

[0103] (4)

[0104] Similarly, the homogeneous transformation matrix between two adjacent attached coordinate systems can be calculated , through the homogeneous transformation matrix and the rod In the attached coordinate system The centroid position We can calculate the center of mass in the basic coordinate system The coordinates below ;

[0105] (5)

[0106] The ZMP calculation model can be obtained based on the centroid coordinates;

[0107] (6)

[0108] Then Lagrangian dynamics is used to model the dynamics of the unmanned metamorphic vehicle. Define generalized coordinates above;

[0109] (7)

[0110] In formula (7), Contains the generalized coordinates of all rotational and translational positions, the mass matrix Contains the mass properties of all components converted to the joint axis;

[0111] (8)

[0112] In formula (8), represents the linear velocity Jacobian matrix of the i-th component, represents the angular velocity Jacobian matrix of the i-th component, Represents the inertia tensor of the i-th component, then the kinetic energy of the unmanned cell car during reconstruction.

[0113] (9)

[0114] Assuming the support foot landing position surface as the zero potential energy surface, the potential energy of the i-th component for:

[0115] (10)

[0116] In formula (10), is the z coordinate of the centroid of the i-th component;

[0117] Then the Lagrangian function of the unmanned metamorphic vehicle is for:

[0118] (11)

[0119] The torque of each rotation joint is:

[0120] (12)

[0121] In formula (12), yes Vector, representing torque or force. Then the instantaneous power of each joint angle is for:

[0122] (13)

[0123] In formula (13), For the The torque or force of each leg joint motor.

[0124] Step 2: Based on the kinematic model and Lagrangian dynamics model, establish the evaluation index of the unmanned metamorphic vehicle in different states;

[0125] Step 2.1: When the unmanned metamorphic vehicle is reconstructing, the closer the zero moment point is to the center of the support domain, the more stable the motion process is. Therefore, the absolute value of the deviation between the zero moment point and the center of the support domain is selected in the entire reconstruction time. The integral of is used as an evaluation index of reconstruction stability . The smaller the value, the more stable the reconstruction process is. There is no Y-axis centroid offset during the reconstruction process, so it does not need to be considered. Finally, according to formula (14), the evaluation index of the reconstruction stability of the unmanned metamorphic vehicle in the support state is obtained: ;

[0126] (14)

[0127] In formula (14), The zero moment point of the unmanned vehicle in the support state is The value of the axis direction; represents the total time of reconstruction, Represents any moment in the reconstruction process.

[0128] Step 2.2: When walking in a human-like manner, Xiang and The displacement towards the center of mass, therefore, needs to be considered simultaneously and At the same time, in a gait cycle, the support area of ​​the single-leg support period is significantly smaller than that of the double-leg support period, and it is more likely to fall. In the example, the step stability when the slider of the mass center adjustment mechanism is moved to the lateral limit position is taken as the optimization target. According to formula (15), the evaluation index of the walking stability of the unmanned metamorphic vehicle in the humanoid state is obtained: ;

[0129] (15)

[0130] In formula (15), The zero moment point of the unmanned metamorphosis vehicle in the humanoid state is The value of the axis direction; Indicates the support period of a single mechanical leg; Represents a single robotic leg at any moment in the support process.

[0131] Step 2.3: The energy consumption during the reconstruction process mainly comes from the mechanical legs, the center of mass adjustment mechanism, and the lifting mechanism. Compared with the mechanical legs, the other mechanisms consume very little energy during the reconstruction process, so only the energy consumption optimization of the mechanical legs is considered. The average power within is used as the evaluation index , the smaller the value, the less energy consumption of reconstruction. According to formula (16), the leg joint motor of the unmanned metamorphic vehicle is Evaluation index of internal average power ;

[0132] (16)

[0133] In formula (16), is the torque of the e-th leg joint motor in the two mechanical legs, express The angular velocity of the e-th joint of the two robotic legs at the moment; is the total number of leg joint motors, and =6.

[0134] Step 2.4: When walking in a humanoid posture, only the ankle, knee, and hip joint motors move. Therefore, according to formula (17), the leg joint motors of the unmanned metamorphosis vehicle move in a gait cycle during humanoid walking. Evaluation index of internal average power ;

[0135] (17)

[0136] In formula (17), Represents a gait cycle At any time while walking inside, express The angular velocity of the e-th joint of the two robotic legs at time .

[0137] Step 2.5: In order to improve the quality of sliding mode control, it is necessary to make the system state reach the set sliding mode surface as quickly as possible and suppress the chattering caused by the sign function. With time The product of the total reconstruction time The integral of , can comprehensively evaluate the approach speed of the sliding surface and the chattering amount in the sliding mode area. The smaller the value, the better the control quality. Specifically, Figure 2 According to formula (18), the evaluation index of the reconstruction control quality of the unmanned metamorphic vehicle in the support state is obtained: ;

[0138] (18)

[0139] In formula (18), They are Constantly adjust the tracking error and error derivative of the slider in the center of mass mechanism, is the sliding surface parameter of the sliding mode controller.

[0140] Step 3: Use SOBOL sensitivity analysis method to select parameters;

[0141] Step 3.1: Define the variable structural parameters of the unmanned metamorphosis vehicle, including: the installation height of the hip joint motor , Installation height of lifting mechanism , Adjust the installation height of the center of mass mechanism , longitudinal installation position of hip joint motor , Longitudinal installation position of lifting mechanism , adjust the longitudinal installation position of the center of mass mechanism , the length of the calf And the length of the thigh ;

[0142] Define the variable posture parameters of the unmanned vehicle, including: the landing position of the support foot ;

[0143] The variable structural parameters and variable posture parameters of the unmanned metamorphic vehicle are defined as follows: Figure 3 As shown;

[0144] The unmanned metamorphic vehicle is composed of variable structural parameters and variable posture parameters. parameters, and any parameter is recorded as , is the number of parameters; =7;

[0145] Step 3.2, calculate the bth parameter The first-order impact index of each evaluation index of the unmanned vehicle ;

[0146] Step 3.2.1, define variable j and initialize j=1;

[0147] Step 3.2.2: Substitute any j parameters from the parameters into the kinematic model and the Lagrangian dynamics model, and use equations (14) to (17) to calculate the four evaluation index values ​​corresponding to the j parameters, so as to obtain the four evaluation index values ​​corresponding to each j parameter, and calculate their variances;

[0148] Step 3.2.3, after assigning j+1 to j, return to step 3.2.2 until j>z; thus, the total variance is obtained by summing up the variance calculated each time. ;

[0149] Step 3.2.4: Calculate the bth parameter using formula (19): The first-order influence index of the yth evaluation index of the unmanned vehicle ;

[0150] (19)

[0151] In formula (19), Indicates the bth parameter The variance calculated for the corresponding y-th evaluation index value, Indicates the bth parameter The corresponding y-th evaluation index value, y=1,2,3,4; Indicates that the bth parameter is The expectation of the y-th evaluation index value corresponding to all parameters except .

[0152] Step 3.3: Calculate the bth parameter according to formula (20): The total effect index of the yth evaluation index of the unmanned metamorphosis vehicle , thus the first-order impact index Total effect index Composition of the bth parameter The sensitivity analysis results are as follows: Figure 4 shown. Figure 4 (a) indicates The sensitivity analysis results of the parameters for the first evaluation index of the unmanned metamorphic vehicle are shown in (b). The sensitivity analysis results of the parameters for the second evaluation index of the unmanned metamorphic vehicle are shown in (c). The sensitivity analysis results of the parameters for the third evaluation index of the unmanned metamorphic vehicle are shown in (d). The sensitivity analysis results of the bth parameter for the fourth evaluation index of the unmanned metamorphic vehicle. First-order impact index Total effect index The larger the parameter, the The greater the impact on the calculation of the yth evaluation index of the unmanned vehicle, the greater the impact on the calculation of the yth evaluation index of the unmanned vehicle. First-order impact index Total effect index is zero, indicating that the parameter It has no impact on the calculation of evaluation indicators.

[0153] (20)

[0154] In formula (20), Indicates that only the bth parameter is included The expectation of the corresponding y-th evaluation index value; Except for the bth parameter The variance calculated for the y-th evaluation index value corresponding to all parameters except .

[0155] Step 4: According to different state performance and structural constraints, formulate the constraint domain of each parameter;

[0156] Step 4.1: According to formula (21), construct the unmanned metamorphosis vehicle in the humanoid state The constraint domain of

[0157] When optimizing the structure of the unmanned metamorphosis vehicle, Optimize, the optimization index is the evaluation index of walking stability Evaluation index of average walking power When using the inverted pendulum model for gait planning, the total length of the mechanical leg cannot exceed the center of mass height of 0.9 m planned by the inverted pendulum model. In order to make the humanoid leg length of the unmanned vehicle closer to the human leg length ratio, ;

[0158] (twenty one)

[0159] Step 4.2: According to equations (22) to (24), construct the unmanned metamorphic vehicle in the car state constraint domain.

[0160] Due to different longitudinal installation positions of hip joint motors The overall position of the mechanical leg will change, which will affect the distance from the center of mass to the front and rear axes, and ultimately lead to different steering characteristics. In order to meet the understeering characteristics of the car, the constraint domain of the longitudinal installation position of the hip joint motor is selected: ;

[0161] In order to meet the passing performance of the vehicle in the state, considering the minimum ground clearance of the vehicle, the lowest position of the hip joint motor installation height is 0.25 m. The lower the center of mass position, the better the driving stability of the vehicle. Therefore, the installation height of the hip joint motor should not be too high. The constraint domain of the hip joint motor installation height is selected as follows: The installation height of the lifting mechanism The constraint domain is: ;

[0162] (twenty two)

[0163] (twenty three)

[0164] (twenty four).

[0165] Step 4.3: According to equations (25) to (28), construct the unmanned metamorphosis vehicle in the support state and reconstruction process , The constraint domain of

[0166] Support foot landing position The initial posture of the unmanned vehicle can be changed, so it is called a variable posture parameter. When the thigh is horizontal, the supporting foot is at the left extreme position. When the knee joint motor is about to touch the ground, the supporting foot is at the right extreme position, so that the supporting foot landing position The constraint domain is:

[0167] (25)

[0168] In order to ensure the stability of reconstruction during the reconstruction process, a sliding mode controller based on the power reaching law is designed to control the reconstruction stability of the slider on the mass center adjustment mechanism. By controlling the movement of the slider in the lateral and longitudinal positions, the zero moment point can be made to approach the center of the support foot to ensure stable reconstruction. The sliding mode control method based on the power reaching law has the sliding surface parameter , and the first parameter of the reaching law The second parameter of the reaching law , according to the design requirements, the constraint domain of the sliding mode controller parameters is selected as:

[0169] (26)

[0170] (27)

[0171] (28).

[0172] Step 5: According to The sensitivity analysis results of the parameters are used to formulate a hierarchical optimization strategy, including: evaluation indicators for walking stability Evaluation index of average walking power , using the artificial hummingbird algorithm to analyze the unmanned vehicle The strategy for the first-level optimization and the evaluation index for the reconstruction stability based on the results of the first-level optimization strategy , Evaluation index of reconstruction average power and reconstruction control quality evaluation index , using NSGA-Ⅲ algorithm to analyze the unmanned vehicle , And the sliding surface parameters of the sliding mode controller , the first parameter of the reaching law and the second parameter of the reaching law Strategies for performing second level optimization.

[0173] Step 5.1: and As two optimization objectives, As the first-level optimization variables, the artificial hummingbird algorithm is used to optimize the first-level optimization variables to obtain the optimal thigh and calf lengths of each mechanical leg of the unmanned metamorphic vehicle. Specifically, Figure 5 As shown;

[0174] Step 5.1.1, define the population size as 30, the dimension as 2, and the maximum number of iterations , the current number of iterations is N, and N=1 is initialized;

[0175] Let the mth hummingbird food source The position is the length of the thigh and calf of any mechanical leg of the mth group of unmanned metamorphic vehicles in the humanoid state; define and initialize the optimal thigh and calf length as ;

[0176] Step 5.1.2, initialize m=1;

[0177] Step 5.1.3: Calculate the length of the mth group of thighs and calves in the Nth iteration using equations (15) and (17): Evaluation index of walking stability Evaluation index of average walking power , thus obtaining The fitness value of = + ; and with optimal thigh and calf length The fitness values ​​of the two groups are compared, and the set of thigh and calf lengths corresponding to the smaller value is taken as ;

[0178] Step 5.1.4, after assigning m+1 to m, return to step 5.1.3 and execute sequentially until m > 30;

[0179] Step 5.1.5: After assigning N+1 to N, return to step 5.1.2 and execute sequentially until N > 50, thereby outputting the optimal thigh and calf lengths for each mechanical leg of the unmanned metamorphosis vehicle. .

[0180] Step 5.2: To optimize the goal, , , As the second-level optimization variable, the NSGA-Ⅲ algorithm is used to perform the second-level multi-objective integrated optimization on the second-level optimization variable, so as to obtain the optimal variable structure parameter value and variable posture parameter value of the unmanned vehicle, as well as the optimal sliding surface parameter value of the sliding mode controller, the optimal first parameter value of the reaching law and the optimal second parameter value of the reaching law. Specifically, Figure 6 shown.

[0181] Step 5.2.1, define the maximum number of iterations as 50, the current number of iterations as V, and initialize V=1, determine the selection, crossover and mutation operators;

[0182] Randomly generate a V-th generation parent population containing 100 individuals ,and Each individual in contains 7-dimensional genes, i.e., the second-level optimization variables;

[0183] Step 5.2.2: For the Vth generation parent population After selection, crossover, and mutation, a V-generation offspring population containing 100 individuals is generated. ; V generation parent population and the V generation progeny population After merging, we get the Vth generation merged population containing 200 individuals. ;

[0184] Step 5.2.3: Define variable a as the Vth generation merged population The ath individual in , and initialize a=1;

[0185] Step 5.2.4: Use equations (14), (16) and (18) to calculate the Vth generation combined population respectively: Evaluation index of the reconstruction stability of the ath individual , Evaluation index of reconstruction average power and reconstruction control quality evaluation index , thus obtaining the Vth generation merged population The fitness value of the ath individual in ;

[0186] Step 5.2.5, after assigning a+1 to a, return to step 5.2.4 and execute sequentially until a > 200;

[0187] Step 5.2.6, according to The fitness value of each individual in the Vth generation merged population The individuals in are sorted by non-dominance, and then 100 individuals are selected according to the non-dominance relationship to form the V+1 generation parent population. ;

[0188] Step 5.2.7, after assigning V+1 to V, return to step 5.2.2 and execute sequentially until V > So far, the Pareto solution set in the 50th iteration parent population is output as the optimal parameter value for the second-level multi-objective integrated optimization of the unmanned metamorphic vehicle.

[0189] Finally, the parameter optimization results of the embodiment are obtained, specifically, as shown in Table 1:

[0190] Table 1 Optimization results of the embodiment

[0191]

[0192] Comparing the performance of the unmanned cell car before and after optimization in Table 1, the evaluation index of reconstruction stability The optimized unmanned vehicle has reduced its average power by 69.81% compared with that before optimization. After optimization, it was reduced by 11.81% compared with before optimization; in the evaluation index of reconstruction control quality After optimization, the value was reduced by 69.42% compared with that before optimization. After optimization, it was reduced by 9.84% compared with before optimization; in the evaluation index of average walking power After optimization, the walking speed is reduced by 8.48% compared with that before optimization. Since the walking process is optimized only in a single-leg support period, and the walking process is composed of multiple gait cycles, the walking stability and energy consumption over long distances are better than before optimization. The optimization results of this embodiment show that the parameter optimization method of the unmanned metamorphosis vehicle is feasible.

[0193] In this embodiment, an electronic device includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the above method, and the processor is configured to execute the program stored in the memory.

[0194] In this embodiment, a computer-readable storage medium stores a computer program on the computer-readable storage medium, and the computer program executes the steps of the above method when executed by a processor.

Claims

1. A parameter optimization method for an unmanned metamorphic vehicle based on multi-configuration classification and integrated algorithm is applied to the unmanned metamorphic vehicle in the process of driving in a car state, walking in a human-like state, and reconstructing the supporting state between the car state and the human-like state. The unmanned metamorphic vehicle includes: The front and rear bodies and the sliding mode controller are provided with two mechanical legs, a mass center adjustment mechanism and a lifting mechanism on the bottom plate of the rear body; wherein each mechanical leg comprises: three leg joint motors, a thigh, a calf and a supporting foot, wherein a knee joint motor is provided between the thigh and the calf, an ankle joint motor is provided between the calf and the supporting foot, and a hip joint motor is provided between the thigh and the bottom plate of the rear body; characterized in that the parameter optimization method of the unmanned cell vehicle comprises the following steps: Step 1: Establish an unmanned cell-changing vehicle model; Step 1.1: Establish a basic coordinate system on the horizontal plane where the support foot of the unmanned vehicle is located. The origin of the basic coordinate system is set at the middle position of the support foot. The positive direction of the axis is the forward direction of the unmanned vehicle, and the basic coordinate system The positive direction of the axis is perpendicular to the horizontal plane and upward. The positive direction of the axis is perpendicular to Axis and The axis is in the plane and points to the outside of the vehicle; Step 1.2, establish the kinematic model and Lagrangian dynamics model of the unmanned metamorphic vehicle reconstruction and walking; Step 2: Based on the kinematic model and Lagrangian dynamics model, establish the evaluation index of the unmanned metamorphic vehicle in different states; Step 2.1: According to formula (1), the evaluation index of the reconstruction stability of the unmanned metamorphic vehicle in the support state is obtained: ; (1) In formula (1), The zero moment point of the unmanned vehicle in the support state is The value of the axis direction; represents the total time of reconstruction, Represents any moment of the reconstruction process; Step 2.2: According to formula (2), the evaluation index of the walking stability of the unmanned metamorphosis vehicle in the humanoid state is obtained: ; (2) In formula (2), The zero moment point of the unmanned metamorphosis vehicle in the humanoid state is The value of the axis direction; Indicates the support period of a single mechanical leg; Represents a single mechanical leg at any moment in the support process; Step 2.3: According to formula (3), the total reconstruction time of the leg joint motor of the unmanned metamorphosis vehicle is obtained: Evaluation index of internal average power ; (3) In formula (3), is the torque of the i-th leg joint motor in the two mechanical legs, express The angular velocity of the i-th joint of the two robotic legs at the moment; is the total number of leg joint motors, and =6; Step 2.4: According to formula (4), the leg joint motor of the unmanned metamorphosis vehicle is obtained in a gait cycle during human-like walking. Evaluation index of internal average power ; (4) In formula (4), Represents a gait cycle At any time while walking inside, express The angular velocity of the i-th joint of the two robotic legs at the moment; Step 2.5: According to formula (5), the evaluation index of the reconstruction control quality of the unmanned metamorphic vehicle in the support state is obtained: ; (5) In formula (5), They are Constantly adjust the tracking error and error derivative of the slider in the center of mass mechanism, is the sliding surface parameter of the sliding mode controller; Step 3: Use SOBOL sensitivity analysis method to select parameters; Step 3.1: Define the variable structural parameters of the unmanned metamorphosis vehicle, including: the installation height of the hip joint motor , Installation height of lifting mechanism , Adjust the installation height of the center of mass mechanism , longitudinal installation position of hip joint motor , Longitudinal installation position of lifting mechanism , adjust the longitudinal installation position of the center of mass mechanism , the length of the calf And the length of the thigh ; Define the variable posture parameters of the unmanned vehicle, including: the landing position of the support foot ; The unmanned metamorphic vehicle is composed of variable structural parameters and variable posture parameters. parameters, and any parameter is recorded as , is the number of parameters; z=7; Step 3.2, calculate the bth parameter The first-order influence index of the yth evaluation index of the unmanned vehicle ; Step 3.3: Calculate the bth parameter according to formula (7): The total effect index of the yth evaluation index of the unmanned metamorphosis vehicle , thus the first-order impact index and total effect index Composition of the bth parameter Sensitivity analysis results of (7) In formula (7), Indicates that only the bth parameter is included The expectation of the corresponding y-th evaluation index value; Except for the bth parameter The variance calculated for the yth evaluation index value corresponding to all parameters except ; Indicates the bth parameter The corresponding y-th evaluation index value; Step 4: According to different state performance and structural constraints, formulate the constraint domain of each parameter; Step 4.1: According to formula (8), construct the unmanned metamorphosis vehicle in the humanoid state The constraint domain of (8) In formula (8), Indicates the minimum value of leg length ratio, Indicates the maximum value of leg length ratio; Step 4.2: According to equations (9) to (11), construct the unmanned metamorphosis vehicle in the car state The constraint domain of (9) (10) (11) In formula (9) to formula (11), Indicates that the hip joint motor is in the longitudinal installation position The minimum value of Indicates that the hip joint motor is in the longitudinal installation position The maximum value of Indicates the hip joint motor at the installation height The minimum value of Indicates the hip joint motor at the installation height The maximum value of Indicates the installation height of the lifting mechanism The minimum value of Indicates the installation height of the lifting mechanism The maximum value of Step 4.3: According to equations (12) to (15), construct the unmanned metamorphosis vehicle in the support state and reconstruction process , The constraint domain of (12) (13) (14) (15) In formula (12) to formula (15), is the distance from the center of the left ankle joint motor to the supporting foot; represents the sliding surface parameters of the sliding mode controller The minimum value of The first parameter of the reaching law of the sliding mode controller is The minimum value of The second parameter of the reaching law of the sliding mode controller is The minimum value of The second parameter of the reaching law of the sliding mode controller is The maximum value of Step 5: According to The sensitivity analysis results of each parameter are used to formulate a hierarchical optimization strategy; Step 5.1: Evaluation index of walking stability Evaluation index of average walking power As two optimization objectives, They are respectively used as the first-level optimization variables, and the artificial hummingbird algorithm is used to perform the first-level optimization on the first-level optimization variables to obtain the optimal thigh and calf lengths of each mechanical leg of the unmanned metamorphic vehicle; Step 5.2: Based on the results of the first-level optimization strategy, reconstruct the stability evaluation index , Evaluation index of reconstruction average power and reconstruction control quality evaluation index To optimize the goal, the unmanned vehicle , And the sliding surface parameters of the sliding mode controller , the first parameter of the reaching law and the second parameter of the reaching law As the second-level optimization variables, the NSGA-Ⅲ algorithm is used to perform the second-level multi-objective integrated optimization on the second-level optimization variables, so as to obtain the optimal variable structure parameter values ​​and variable posture parameter values ​​of the unmanned vehicle, as well as the optimal sliding surface parameter values ​​of the sliding mode controller, the optimal first parameter value of the reaching law and the optimal second parameter value of the reaching law.

2. The method for optimizing parameters of an unmanned metamorphic vehicle based on multi-configuration classification and integrated algorithm according to claim 1 is characterized in that: The step 3.2 comprises: Step 3.2.1, define variable j and initialize j=1; Step 3.2.2: Substitute any j parameters from the parameters into the kinematic model and the Lagrangian dynamics model, and use equations (1) to (4) to calculate the four evaluation index values ​​corresponding to the j parameters, so as to obtain the four evaluation index values ​​corresponding to each j parameter, and calculate their variances; Step 3.2.3, after assigning j+1 to j, return to step 3.2.2 until j>z; thus, the total variance is obtained by summing up the variance calculated each time. ; Step 3.2.4: Calculate the bth parameter using formula (6): The first-order influence index on the yth evaluation index of the unmanned vehicle ; (6) In formula (6), Indicates the bth parameter The variance calculated for the corresponding y-th evaluation index value, Indicates the bth parameter The corresponding y-th evaluation index value, y=1,2,3,4; Indicates that the bth parameter is The expectation of the y-th evaluation index value corresponding to all parameters except .

3. The method for optimizing parameters of an unmanned metamorphic vehicle based on multi-configuration classification and integrated algorithm according to claim 2 is characterized in that: The step 5.1 comprises: Step 5.1.1, define the population size as M, the dimension as d, and the maximum number of iterations , the current number of iterations is N, and N=1 is initialized; Let the mth hummingbird food source The position is the length of the thigh and calf of any mechanical leg of the mth group of unmanned metamorphic vehicles in the humanoid state; define and initialize the optimal thigh and calf length as ; Step 5.1.2, initialize m=1; Step 5.1.3: Calculate the length of the mth group of thighs and calves in the Nth iteration using equations (2) and (4): Evaluation index of walking stability Evaluation index of average walking power , thus obtaining The fitness value of = + ; and with optimal thigh and calf length The fitness values ​​of the two groups are compared, and the set of thigh and calf lengths corresponding to the smaller value is taken as ; Step 5.1.4, after assigning m+1 to m, return to step 5.1.3 and execute sequentially until m > M; Step 5.1.5: After assigning N+1 to N, return to step 5.1.2 and execute sequentially until N > So far, the optimal thigh and calf length of each mechanical leg of the unmanned cell car is output .

4. The method for optimizing parameters of an unmanned metamorphic vehicle based on multi-configuration classification and integrated algorithm according to claim 3 is characterized in that: The step 5.2 comprises: Step 5.2.

1. Define the maximum number of iterations , the current number of iterations is V, and V=1 is initialized to determine the selection, crossover operator and mutation operator; Randomly generate a V-th generation parent population containing A individuals ,and Each individual in contains 7-dimensional genes, i.e., the second-level optimization variables; Step 5.2.2: For the Vth generation parent population After selection, crossover, and mutation, a V-generation offspring population containing A individuals is generated. ; V generation parent population and the V generation progeny population After merging, we get the Vth generation merged population containing 2A individuals. ; Step 5.2.3: Define variable a as the Vth generation merged population The ath individual in , and initialize a=1; Step 5.2.4: Use equations (1), (3) and (5) to calculate the Vth generation combined population Evaluation index of the reconstruction stability of the ath individual , Evaluation index of reconstruction average power and reconstruction control quality evaluation index , thus obtaining the Vth generation merged population The fitness value of the ath individual in ; Step 5.2.5, after assigning a+1 to a, return to step 5.2.4 and execute sequentially until a > 2A; Step 5.2.6, according to The fitness value of each individual in the Vth generation merged population The individuals in the non-dominated sort are sorted, and then according to the non-dominated relationship, a number of individuals A are selected to form the V+1 generation parent population. ; Step 5.2.7, after assigning V+1 to V, return to step 5.2.2 and execute sequentially until V > So far, output The Pareto solution set in the iterative parent population is used as the optimal parameter value for the second-level multi-objective integrated optimization of the unmanned metamorphic vehicle.

5. An electronic device, comprising a memory and a processor, characterized in that: The memory is used to store a program that supports the processor to execute the unmanned metamorphic vehicle parameter optimization method described in any one of claims 1 to 4, and the processor is configured to execute the program stored in the memory.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the unmanned metamorphic vehicle parameter optimization method described in any one of claims 1 to 4 are executed.

Images