CN103832504A - Bionic foot-type robot comprehensive simulation strategy - Google Patents

Abstract

The invention discloses a comprehensive simulation strategy of a bionic legged robot, which belongs to the field of robot technology application. The system of the present invention comprises: based on Matlab/Simulink bionic footed robot control model (A), based on Matlab/SimMechanics bionic footed robot kinematics simulation model (B), based on ADAMS bionic footed robot dynamics model (C), bionic foot Experimental prototype of the robot (D). The comprehensive simulation strategy includes the real-time kinematics and dynamics demonstration method of the motion state of the robot experimental prototype, the real-time motion control method of the robot experimental prototype after the robot gait generation simulation verification, the robot virtual co-simulation, and the self-learning adjustment of the hardware-in-the-loop simulation method and method for adaptive multi-coordinated control of robots. The invention has the characteristics of low cost and multiple functions, basically covers the simulation debugging requirements of the traditional bionic legged robot, and has certain universal applicability.

Description

Synthetic Simulation Strategy of Bionic Legged Robot

technical field

The invention belongs to the application field of robot technology, and specifically relates to a comprehensive simulation strategy of a bionic legged robot based on Matlab, ADAMS and semi-physical joint simulation, which is mainly used in motion demonstration, gait generation, self-learning and multi-coordination control of a bionic legged robot .

Background technique

Bionic footed robot is one of the most cutting-edge topics in the field of robot research today. It integrates many disciplines such as machinery, electronics, computers, materials, sensors, control technology and artificial intelligence, reflecting a country's intelligence and automation research. At the same time, as an important symbol of a country's high-tech strength, developed countries have invested heavily in research in this field.

Compared with wheeled and tracked robots, bionic legged robots have superior mobility and can adapt to complex and changeable unstructured natural environments. Because it has more joint degrees of freedom, it is more difficult to debug and control the bionic legged robot. The virtual simulation method is usually used as the pre-verification stage of the robot design, such as the biped robot modeling and simulation based on ADAMS published by Liang Qing et al. ("Computer Simulation" 2010 No. 5), and the semi-physical simulation board can be used for later debugging. Card (such as dSPACE hardware-in-the-loop simulation platform) for hardware-in-the-loop debugging, such as the design based on dSPACE hardware-in-the-loop simulation platform published by Li Xuejun and others ("Journal of Changchun University" 2011 06).

Since each simulation software has its own advantages and disadvantages, two simulation software are also used for joint simulation (for example, Liu Xiaocheng et al. published a joint simulation analysis of imitation crab robot based on MATLAB and ADAMS single foot ("Microcomputer Information" 2010 No. 14) ), but the results of pure virtual co-simulation are often quite different from the actual experimental data, and do not have certain practical guiding significance; a single half-physical simulation based on half-physical simulation boards not only relies on existing high-tech integrated hardware and software ( High cost, poor adaptability to bionic legged robot motion control debugging, etc.), and it is not compatible with other virtual simulation software.

For above situation, the present invention proposes a kind of bionic legged robot comprehensive simulation strategy based on Matlab, ADAMS and half-in-the-loop joint simulation, it is characterized in that utilize the advantage that various virtual simulation software is good at to carry out division of labor joint simulation, borrow bionic legged robot The wireless transmission device of the prototype and the computer control platform carry out wireless data transmission, which really exerts the advantages of Matlab, ADAMS and half-in-the-loop joint simulation, and can realize various functions of bionic legged robot motion debugging, including: bionic legged robot motion demonstration, Gait generation, self-learning and multi-coordination control.

Based on the motion demonstration, gait generation, self-learning and multi-coordination control strategies of bionic legged robots based on Matlab, ADAMS and hardware-in-the-loop co-simulation, a multi-functional debugging strategy is innovatively proposed to provide comprehensive motion control debugging for bionic legged robots It provides new ideas and methods, which have very important theoretical significance and practical value.

Contents of the invention

The purpose of the present invention is to provide a bionic legged robot comprehensive simulation strategy based on Matlab, ADAMS and hardware-in-the-loop joint simulation.

A kind of bionic legged robot comprehensive simulation strategy based on Matlab, ADAMS and hardware-in-the-loop co-simulation, is characterized in that:

The systems used include: the control model of the bionic footed robot based on Matlab/Simulink, the kinematics simulation model of the bionic footed robot based on Matlab/SimMechanics, the dynamics model of the bionic footed robot based on ADAMS, and the experimental prototype of the bionic footed robot.

The strategy includes the real-time kinematics demonstration process of the motion state of the robot experimental prototype. The specific method is: the bionic legged robot experimental prototype wirelessly sends interface robot environment perception and real-time motion feedback data to the bionic legged robot control model based on Matlab/Simulink, based on The Matlab/Simulink bionic footed robot control model transmits the foot end trajectory and gait parameters to the bionic footed robot kinematics simulation model based on Matlab/SimMechanics, and Matlab/SimMechanics has the relevant mechanism kinematics animation demonstration function, real-time display The kinematic state of the robot experimental prototype, the relevant joint motion trajectory and gait data are transmitted to the bionic legged robot control model based on Matlab/Simulink, and the kinematic data is summarized and saved;

The strategy includes the real-time dynamics demonstration process of the motion state of the experimental prototype of the robot. The specific method is: the experimental prototype of the bionic legged robot wirelessly transmits the environment perception and real-time motion feedback data of the interface robot to the control model of the bionic legged robot based on Matlab/Simulink, and based on Matlab /Simulink bionic legged robot control model transfers motion gait and attitude data to the bionic legged robot dynamics model based on ADAMS, and ADAMS has related mechanism dynamics animation demonstration functions, which can display the dynamics of the robot experimental prototype in real time Data, related dynamics co-simulation output data to the control model based on Matlab/Simulink bionic legged robot, and dynamic data collection and storage;

The strategy includes the pure virtual joint motion simulation process of the robot. The specific method is: based on the Matlab/Simulink bionic footed robot control model, the foot end motion trajectory and gait parameters are transmitted to the Matlab/SimMechanics bionic footed robot kinematics simulation model, and the robot is passed through Matlab. After /SimMechanics simulation, the joint motion trajectory and gait data are summarized to the control model of the bionic footed robot based on Matlab/Simulink, and the motion gait and posture data are sent to the bionic footed robot based on ADAMS based on the control model of the bionic footed robot based on Matlab/Simulink Dynamics model, after ADAMS dynamics simulation, the output data of dynamics co-simulation is summarized to the control model of bionic legged robot based on Matlab/Simulink, and the joint kinematics simulation of Matlab/SimMechanics and ADAMS is completed;

The strategy includes robot gait generation and the real-time motion control process of the robot experimental prototype. The specific method is: based on the Matlab/Simulink bionic legged robot control model, the foot end trajectory and gait parameters are transmitted to the bionic legged robot based on Matlab/SimMechanics. After Matlab/SimMechanics simulation, the joint motion trajectory and gait data are summarized into the control model of the bionic legged robot based on Matlab/Simulink, and the robot motion gait data package and coordination are based on the Matlab/Simulink bionic legged robot control model. The control command parameters are wirelessly sent to the bionic legged robot experimental prototype, and the data generated by Matlab/SimMechanics gait is completed to control the movement of the robot experimental prototype in real time;

The strategy includes robot gait generation and the real-time motion control process of the robot experimental prototype after simulation verification. The specific method is: based on the Matlab/Simulink bionic legged robot control model, the foot end motion trajectory and gait parameters are transmitted to the Matlab/SimMechanics bionic robot control model. Footed robot kinematics simulation model, after Matlab/SimMechanics simulation, the joint motion trajectory and gait data are summarized into a bionic footed robot control model based on Matlab/Simulink (A), based on Matlab/Simulink bionic footed robot control model The gait and attitude data are sent to the dynamic model of the bionic footed robot based on ADAMS. After the dynamics simulation of ADAMS, the output data of the dynamics co-simulation is summarized to the control model of the bionic footed robot based on Matlab/Simulink, and the Matlab/SimMechanics step is completed. After completing the joint motion simulation verification of Matlab/SimMechanics gait generation and ADAMS The data controls the movement of the robot experimental prototype in real time, and realizes the motion control of the robot experimental prototype more accurately;

The strategy includes the robot virtual co-simulation self-learning adjustment process, the specific way: based on the control model of the Matlab/Simulink bionic legged robot, the trajectory of the foot end and the gait parameters are transmitted to the kinematics simulation model of the bionic legged robot based on Matlab/SimMechanics. After Matlab/SimMechanics simulation, the joint motion trajectory and gait data are summarized to the control model of the bionic footed robot based on Matlab/Simulink, and the motion gait and attitude data are sent to the bionic footed robot based on ADAMS based on the control model of the Matlab/Simulink bionic footed robot The robot dynamics model, after the ADAMS dynamics simulation, aggregates the output data of the dynamics co-simulation to the bionic legged robot control model based on Matlab/Simulink, and completes a joint kinematics simulation of Matlab/SimMechanics and ADAMS. The self-learning module in the bionic legged robot control model adjusts parameters, performs continuous cycle simulation, and finally realizes virtual co-simulation self-learning adjustment, providing a basis for data optimization and adaptive control;

The strategy includes the self-learning adjustment process of robot half-in-the-loop simulation. The specific method is: based on the control model of the Matlab/Simulink bionic legged robot, the trajectory of the foot end and the gait parameters are transmitted to the kinematics simulation model of the bionic legged robot based on Matlab/SimMechanics. After Matlab/SimMechanics simulation, the joint motion trajectory and gait data are summarized to the control model of the bionic legged robot based on Matlab/Simulink, and the robot motion gait data packet and coordination control command parameters are sent to the The experimental prototype of the bionic legged robot, after the motion experiment of the robot experimental prototype, summarized the robot environment perception and real-time motion feedback data into the control model of the bionic legged robot based on Matlab/Simulink, and completed a combination of Matlab/SimMechanics and the semi-physical experimental prototype Motion simulation, by adjusting the parameters of the self-learning module based on the Matlab/Simulink bionic legged robot control model, conduct continuous loop hardware-in-the-loop simulation, and finally realize the self-learning adjustment under the hardware-in-the-loop joint simulation, providing a basis for data optimization and adaptive control;

The strategy includes the robot self-adaptive multi-coordination control process, and the specific method is: based on the Matlab/Simulink bionic footed robot control model, the foot end motion trajectory and gait parameters are transmitted to the bionic footed robot kinematics simulation model based on Matlab/SimMechanics, and the After /SimMechanics simulation, the joint motion trajectory and gait data are summarized to the control model of the bionic footed robot based on Matlab/Simulink, and the motion gait and posture data are sent to the bionic footed robot based on ADAMS based on the control model of the bionic footed robot based on Matlab/Simulink Dynamics model, after ADAMS dynamics simulation, the dynamics co-simulation output data is summarized to the control model of the bionic footed robot based on Matlab/Simulink; at the same time, based on the Matlab/Simulink bionic footed robot control model, the robot motion step State data packets and coordination control instruction parameters are sent to the bionic legged robot experimental prototype. After the robot experimental prototype motion experiment, the robot environment perception and real-time motion feedback data are summarized into the bionic legged robot control model based on Matlab/Simulink; integrated dynamics Co-simulation output data and real-time motion feedback data, the two data are complementary and self-learning control, and the robot adaptive multi-coordination control is realized with the help of simulation data under the condition that the actual robot prototype lacks part of the sensor information.

The described bionic legged robot comprehensive simulation strategy based on Matlab, ADAMS and half-in-the-loop co-simulation is characterized in that: the described bionic legged robot control model based on Matlab/Simulink is built, in a specific way: first in Matlab/Simulink bionic In the footed robot simulation system, parameters such as the type of simulation algorithm, period, step size, and error range within the system are set, and then a bionic footed robot motion demonstration module based on Matlab/Simulink is selectively established according to the needs of robot debugging. Matlab/Simulink's bionic legged robot motion gait control module, establish a bionic legged robot motion self-learning module based on Matlab/Simulink and establish a bionic legged robot coordination control module based on Matlab/Simulink, and finally connect Matlab/Simulink bionic foot The robot simulation system interface is used to exchange data and coordinate data processing.

The described bionic legged robot comprehensive simulation strategy based on Matlab, ADAMS and half-in-the-loop joint simulation is characterized in that: the described bionic legged robot kinematics simulation model based on Matlab/SimMechanics is built, and the specific method: first create Matlab/SimMechanics SimMechanics bionic footed robot kinematics simulation environment, set the system environment parameters, and then establish the world coordinate system of the bionic footed robot system and the degrees of freedom and directions of the kinematic joints of each leg in Matlab/SimMechanics, setting parameters including mass, inertia, and length Rigid body parameters of each leg member, joint controller type, joint sensor output type, and finally, according to the robot motion control requirements, lead out the input and output interfaces in the Matlab/SimMechanics bionic footed robot kinematics simulation model, ready to accept joint input data and send The joints output data, exchange data, and realize kinematic calculation functions.

The comprehensive simulation strategy of the bionic legged robot based on Matlab, ADAMS and half-in-the-loop co-simulation is characterized in that: the described ADAMS bionic legged robot dynamics model building method is based on the specific method: first create the ADAMS bionic legged robot Dynamics simulation environment, set the system environment parameters, and then import the ground rigid body and the rigid body of each leg of the bionic legged robot in ADAMS, set the rigid body geometric size, material type or density or quality parameters, and then set each in ADAMS Joint degrees of freedom and joint types between rigid bodies, and set the collision type, friction constraints and related parameter settings between each foot end and ground rigid body in ADAMS, and finally set ADAMS bionic legged robot kinematics simulation according to actual requirements The input and output interfaces in the model are ready to receive joint input data, send joint output data and foot collision force data, exchange data, and realize dynamic calculation functions.

The comprehensive simulation strategy of the bionic legged robot based on Matlab, ADAMS and half-in-the-loop joint simulation is characterized in that: the operation process of the bionic legged robot experimental prototype, in a specific way: first the bionic legged robot experimental prototype is initialized, The settings include system initialization parameters, and then the bionic legged robot experimental prototype wirelessly receives the upper-level task control command signal and gait data, and then performs the motion experiment of the bionic legged robot experimental prototype. Self-perceived attitude sensing, joint angles, angular velocities, foot end force information, and infrared distance measurement for environment recognition of the bionic legged robot experimental prototype, machine vision recognition of obstacle distribution, and ground condition information are sent to the computer control platform wirelessly. Working principle of the present invention:

Based on the advantages of Matlab/SimMechanics in the calculation of bionic legged robot kinematics, analyze the robot gait generation, and the obtained data is used as a reference for robot motion; based on the advantages of ADAMS in the calculation of bionic legged robot dynamics, analyze the dynamics of the robot in the motion environment The characteristics and related data are used as the basis for evaluating the dynamic stability of the bionic legged robot; through the establishment of the bionic legged robot software joint simulation environment on the computer control platform, data exchange and complementarity are realized, and the wireless sending and receiving device of the bionic legged robot experimental prototype , carry out semi-physical debugging with the experimental prototype, in which the computer control platform sends the simulated and verified gait data and coordination control instructions, the experimental prototype performs motion experiments, and the sensor information on the relevant prototype is fed back to the computer control platform through the wireless sending device. Various debugging functions of motion gait are realized by corresponding simulation modules (including: motion demonstration, gait generation, self-learning and multi-coordination control, etc.).

Compared with the prior art, the present invention has the following advantages:

1. The present invention can integrate joint simulation applications based on the respective simulation advantages of Matlab/SimMechanics and ADAMS. It has the characteristics of clear division of labor, clear thinking, and practicality. It is helpful for the theoretical research in the early stage and provides a more accurate reference for later prototype experiment debugging. data.

2. The invention has low cost and multiple functions, basically covers the control and debugging requirements of traditional bionic legged robots, and has certain universal applicability.

3. The present invention provides a new solution for the control and debugging of the bionic legged robot, improves the research and development efficiency of the bionic legged robot technology, and provides a broadened range of simulation applications.

Description of drawings

Fig. 1 is a block diagram of a comprehensive simulation strategy of a bionic legged robot based on Matlab, ADAMS and hardware-in-the-loop joint simulation of the present invention.

Fig. 2 is the flow chart of the way of building the control model of the bionic legged robot based on Matlab/Simulink in the present invention.

Fig. 3 is a flow chart of the method for building a kinematics simulation model of a bionic legged robot based on Matlab/SimMechanics in the present invention.

Fig. 4 is a flow chart of the building method of the dynamic model based on ADAMS bionic legged robot in the present invention.

Fig. 5 is a flowchart of the operation of the experimental prototype of the bionic legged robot in the present invention.

Label names in the above figures: A, Matlab/Simulink-based bionic legged robot control model; B, Matlab/SimMechanics-based bionic legged robot kinematics simulation model; C, ADAMS-based bionic legged robot dynamics model; D, bionic legged robot 1. Foot movement trajectory and gait parameters; 2. Joint movement trajectory and gait data; 3. Dynamics co-simulation output data; 4. Motion gait and posture data; Real-time motion feedback data; 6. Robot motion gait data packets and coordination control instruction parameters.

The components A and B in the figure belong to the simulation environment of the bionic footed robot software based on Matlab, and the components A, B and C belong to the joint simulation environment of the bionic footed robot software based on the computer control platform.

Detailed ways

Below in conjunction with accompanying drawing and specific embodiment the present invention is described in further detail:

In conjunction with Fig. 1, 2, 3, the present embodiment is a block diagram of a comprehensive simulation strategy for a bionic footed robot based on Matlab, ADAMS and half-in-the-loop co-simulation, including: based on Matlab/Simulink bionic footed robot control model A, based on Matlab/ SimMechanics bionic footed robot kinematics simulation model B, bionic footed robot dynamics model C based on ADAMS, bionic footed robot experimental prototype D, foot end movement trajectory and gait parameters 1, joint movement trajectory and gait data 2; power Joint simulation output data 3, motion gait and attitude data 4, robot environment perception and real-time motion feedback data 5, robot motion gait data package and coordination control instruction parameters 6.

Among them, the input and output data between the control model of the bionic footed robot based on Matlab/Simulink and the kinematics simulation model of the bionic footed robot based on Matlab/SimMechanics are foot end motion trajectory and gait parameters and joint motion trajectory and gait data; based on Matlab/SimMechanics The input and output data between the control model of the Simulink bionic footed robot and the dynamic model of the bionic footed robot based on ADAMS are dynamics co-simulation output data and motion gait and posture data; the control model and the bionic footed robot based on Matlab/Simulink The input and output data between the robot experimental prototypes are robot environment perception and real-time motion feedback data, robot motion gait data packets and coordination control instruction parameters.

Among them, the bionic footed robot control model based on Matlab/Simulink and the kinematics simulation model based on Matlab/SimMechanics bionic footed robot software simulation environment based on Matlab bionic footed robot; the control model based on Matlab/Simulink bionic footed robot, based on Matlab/ SimMechanics bionic footed robot kinematics simulation model and bionic footed robot dynamics model based on ADAMS constitute the bionic footed robot software co-simulation environment based on computer control platform.

As shown in Figure 1, a bionic legged robot comprehensive simulation strategy block diagram based on Matlab, ADAMS and hardware-in-the-loop co-simulation, in which the bionic legged robot control model A based on Matlab/Simulink and the bionic legged robot kinematics simulation based on Matlab/SimMechanics The input and output data between model B are the foot end trajectory and gait parameters 1 and the joint movement trajectory and gait data 2; the control model A based on Matlab/Simulink bionic footed robot and the dynamic model C based on ADAMS bionic footed robot The input and output data between the dynamic co-simulation output data 3 and the motion gait and attitude data 4; the input and output data between the bionic legged robot control model A and the bionic legged robot experimental prototype D based on Matlab/Simulink are the robot environment perception and Real-time motion feedback data 5 and robot motion gait data packets and coordination control instruction parameters 6.

As shown in Figure 1, a bionic legged robot comprehensive simulation strategy block diagram based on Matlab, ADAMS and hardware-in-the-loop co-simulation, in which the bionic legged robot control model A based on Matlab/Simulink and the bionic legged robot kinematics simulation based on Matlab/SimMechanics Model B constitutes the simulation environment of the bionic footed robot based on Matlab; the control model A of the bionic footed robot based on Matlab/Simulink, the kinematics simulation model B of the bionic footed robot based on Matlab/SimMechanics and the dynamics model C of the bionic footed robot based on ADAMS The software co-simulation environment of the bionic legged robot based on the computer control platform is formed.

As shown in Figure 1, a bionic footed robot comprehensive simulation strategy block diagram based on Matlab, ADAMS and hardware-in-the-loop co-simulation, in which the robot environment perception and real-time motion feedback data 5 and robot motion gait data packets and coordination control instruction parameters 6 are passed Wireless devices (such as wireless serial devices, wireless wifi and other equipment) perform wireless transmission and reception.

As shown in Figure 2, based on the flow chart of the Matlab/Simulink bionic footed robot control model building method, first set the parameters of the Matlab/Simulink bionic footed robot simulation system (such as the type of simulation algorithm inside the system, period, step size, error range, etc. Parameter setting), and then selectively establish the motion demonstration module of the bionic legged robot based on Matlab/Simulink, establish the gait control module of the bionic legged robot based on Matlab/Simulink, and establish the gait control module based on Matlab/Simulink The bionic footed robot motion self-learning module and the establishment of the bionic footed robot coordination control module based on Matlab/Simulink, and finally connect the Matlab/Simulink bionic footed robot simulation system interface (including data streams 1, 2, 3, 4, 5, 6's input and output interfaces are interconnected for data exchange and coordinate data processing).

As shown in Figure 3, the flow chart of the construction method of the kinematics simulation model based on Matlab/SimMechanics bionic footed robot, first create the Matlab/SimMechanics bionic footed robot kinematics simulation environment (including system environment parameter settings), and then create Establish the world coordinate system of the bionic legged robot system (including base 1, base 2, etc.), the kinematic joints of each leg (including joint degrees of freedom, direction, etc.), and the rigid body parameters of each leg member (including mass, inertia, length, etc.) , joint controller type, joint sensor output type, etc. Finally, according to the robot motion control requirements, the input and output interfaces in the Matlab/SimMechanics bionic legged robot kinematics simulation model (ready to receive joint input data, send joint output data, and Data exchange, realizing kinematics calculation function).

As shown in Figure 4, the flow chart of building a dynamic model based on the ADAMS bionic legged robot is first created. The dynamics simulation environment of the ADAMS bionic legged robot (including system environment Rigid bodies of each leg of the robot (set parameters such as rigid body geometry, material type or density or quality), and then set the degree of freedom and joint type of joints between each rigid body in ADAMS, and set each foot end in ADAMS Collision type with ground rigid body, friction constraints and related parameter settings, and finally according to actual requirements, set the input and output interfaces in the ADAMS bionic legged robot kinematics simulation model (ready to accept joint input data, send joint output data and foot impact force and other data, exchange data, and realize dynamic calculation function).

The operation flow chart of the bionic legged robot experimental prototype is shown in Figure 5. First, the bionic legged robot experimental prototype is initialized (including system parameter initialization settings, etc.), and then the bionic legged robot experimental prototype wirelessly receives the upper control command signal and gait Data (including gait data, task instructions, etc.), and then conduct the motion experiment of the bionic legged robot experimental prototype, and at the same time wirelessly send the sensory information of the bionic legged robot experimental prototype in real time during the experiment (such as gyroscope, accelerometer, geomagnetic Attitude sensors such as gauges, joint angle and angular velocity sensors, foot force sensors, etc.) and wirelessly send environmental recognition information (ultrasonic or infrared ranging, machine vision recognition of obstacle distribution, ground conditions, etc.) to computer control platform.

As shown in Figure 1, a bionic legged robot comprehensive simulation strategy block diagram based on Matlab, ADAMS and hardware-in-the-loop co-simulation, such as following the flow sequence (D→5→A→1→B→2→A), the robot experiment can be realized Real-time kinematics demonstration of the motion state of the prototype. Specific method: The experimental prototype D of the bionic footed robot wirelessly sends the interface robot environment perception and real-time motion feedback data 5 to the control model A of the bionic footed robot based on Matlab/Simulink, and the control model A of the bionic footed robot based on Matlab/Simulink transmits the foot end The motion trajectory and gait parameters 1 are transferred to the bionic legged robot kinematics simulation model B based on Matlab/SimMechanics, and Matlab/SimMechanics has the relevant mechanism kinematics animation demonstration function, which can display the kinematics state of the robot experimental prototype in real time. The joint trajectory and gait data 2 of the joints are transferred to the control model A of the bionic legged robot based on Matlab/Simulink, and the kinematics data is summarized and saved.

As shown in Figure 1, a bionic footed robot comprehensive simulation strategy block diagram based on Matlab, ADAMS and hardware-in-the-loop co-simulation, such as following the flow sequence (D→5→A→4→C→3→A), the robot experiment can be realized Real-time dynamics demonstration of the motion state of the prototype. Specific method: The experimental prototype D of the bionic legged robot wirelessly sends the interface robot environment perception and real-time motion feedback data 5 to the control model A of the bionic legged robot based on Matlab/Simulink, and the control model A of the bionic legged robot based on Matlab/Simulink transmits the motion steps state and attitude data 4 into the dynamic model C based on ADAMS bionic legged robot, and ADAMS has the relevant mechanism dynamics animation demonstration function, which can display the dynamics data of the robot experimental prototype in real time, and the relevant dynamics co-simulation output data 3 Transfer to the bionic legged robot control model A based on Matlab/Simulink to summarize and save the dynamic data.

As shown in Figure 1, a comprehensive simulation strategy block diagram of a bionic legged robot based on Matlab, ADAMS and hardware-in-the-loop co-simulation, such as in accordance with the process sequence (A→1→B→2→A→4→C→3→A), It can realize the pure virtual joint motion simulation of the robot. Specific method: Based on the Matlab/Simulink bionic legged robot control model A, transfer the foot end trajectory and gait parameters 1 to the bionic legged robot kinematics simulation model B based on Matlab/SimMechanics, after Matlab/SimMechanics simulation, the joint motion trajectory and The gait data 2 is summarized to the control model A of the bionic footed robot based on Matlab/Simulink, and the motion gait and posture data 4 is sent to the dynamic model C of the bionic footed robot based on ADAMS, based on the Matlab/Simulink bionic footed robot control model A, After the ADAMS dynamics simulation, the dynamics co-simulation output data 3 is summarized to the bionic legged robot control model A based on Matlab/Simulink, and the joint motion simulation of Matlab/SimMechanics and ADAMS is completed.

As shown in Figure 1, a bionic footed robot comprehensive simulation strategy block diagram based on Matlab, ADAMS and hardware-in-the-loop co-simulation, such as following the flow sequence (A→1→B→2→A→6→D), the robot step can be realized State generation, real-time motion control of robot experimental prototype. Specific method: Based on the Matlab/Simulink bionic legged robot control model A, transfer the foot end trajectory and gait parameters 1 to the bionic legged robot kinematics simulation model B based on Matlab/SimMechanics, after Matlab/SimMechanics simulation, the joint motion trajectory and The gait data 2 is summarized to the bionic footed robot control model A based on Matlab/Simulink, and the robot motion gait data packet and coordination control command parameter 6 are wirelessly sent to the bionic footed robot experiment based on the Matlab/Simulink bionic footed robot control model Prototype D, the data generated by Matlab/SimMechanics gait is used to control the movement of the robot experimental prototype in real time.

As shown in Figure 1, a bionic legged robot comprehensive simulation strategy block diagram based on Matlab, ADAMS and hardware-in-the-loop co-simulation, such as in accordance with the flow sequence (A→1→B→2→A→4→C→3→A→6 →D), which can realize the real-time motion control of the robot experimental prototype after robot gait generation and simulation verification. Specific method: Based on the Matlab/Simulink bionic legged robot control model A, transfer the foot end trajectory and gait parameters 1 to the bionic legged robot kinematics simulation model B based on Matlab/SimMechanics, after Matlab/SimMechanics simulation, the joint motion trajectory and The gait data 2 is summarized to the control model A of the bionic footed robot based on Matlab/Simulink, and the motion gait and posture data 4 is sent to the dynamic model C of the bionic footed robot based on ADAMS, based on the Matlab/Simulink bionic footed robot control model A, After the ADAMS dynamics simulation, the dynamics co-simulation output data 3 is summarized to the Matlab/Simulink bionic legged robot control model A, and the Matlab/SimMechanics gait generation and ADAMS joint kinematics simulation verification are completed, and then the robot motion step The state data packet and the coordination control instruction parameter 6 are wirelessly sent to the bionic legged robot experimental prototype D, and the joint motion simulation verification by Matlab/SimMechanics gait generation and ADAMS is completed. Motion control of a robotic experimental prototype.

As shown in Figure 1, a bionic footed robot comprehensive simulation strategy block diagram based on Matlab, ADAMS and hardware-in-the-loop co-simulation, such as in accordance with the process sequence (A→1→B→2→A→4→C→3→A∙∙ ∙→A→1→B→2→A→4→C→3→A∙∙∙), which can realize the self-learning adjustment of robot virtual co-simulation. Specific method: Based on the Matlab/Simulink bionic legged robot control model A, transfer the foot end trajectory and gait parameters 1 to the bionic legged robot kinematics simulation model B based on Matlab/SimMechanics, after Matlab/SimMechanics simulation, the joint motion trajectory and The gait data 2 is summarized to the control model A of the bionic footed robot based on Matlab/Simulink, and the motion gait and posture data 4 is sent to the dynamic model C of the bionic footed robot based on ADAMS, based on the Matlab/Simulink bionic footed robot control model A, After the ADAMS dynamics simulation, the dynamics co-simulation output data 3 was summarized to the control model A of the bionic legged robot based on Matlab/Simulink, and a joint motion simulation of Matlab/SimMechanics and ADAMS was completed, and the bionic legged robot based on Matlab/Simulink The self-learning module in the robot control model A adjusts the parameters, performs continuous cycle simulation, and finally realizes the virtual co-simulation self-learning adjustment, which provides a basis for data optimization and adaptive control.

As shown in Figure 1, a bionic footed robot comprehensive simulation strategy block diagram based on Matlab, ADAMS and hardware-in-the-loop co-simulation, such as in accordance with the process sequence (A→1→B→2→A→6→D→5→A∙∙ ∙→A→1→B→2→A→6→D→5→A∙∙∙), which can realize the self-learning adjustment of the robot half-in-the-loop simulation. Specific method: Based on the Matlab/Simulink bionic legged robot control model A, transfer the foot end trajectory and gait parameters 1 to the bionic legged robot kinematics simulation model B based on Matlab/SimMechanics, after Matlab/SimMechanics simulation, the joint motion trajectory and The gait data 2 is summarized to the bionic legged robot control model A based on Matlab/Simulink, and the robot motion gait data package and coordination control instruction parameter 6 are sent to the bionic legged robot experimental prototype based on the Matlab/Simulink bionic legged robot control model A D. After the motion experiment of the robot experimental prototype, the robot environment perception and real-time motion feedback data 5 were summarized into the Matlab/Simulink-based bionic legged robot control model A, and a joint motion simulation of Matlab/SimMechanics and the semi-physical experimental prototype was completed. By adjusting the parameters of the self-learning module based on the Matlab/Simulink bionic legged robot control model A, the continuous loop hardware-in-the-loop simulation is carried out, and finally the self-learning adjustment under the hardware-in-the-loop joint simulation is realized, which provides a basis for data optimization and adaptive control.

As shown in Figure 1, a comprehensive simulation strategy block diagram of a bionic legged robot based on Matlab, ADAMS and hardware-in-the-loop co-simulation, for example, according to the process sequence ((A→1→B→2→A→4→C→3)+( A→6→D→5)→A ∙∙∙→(A→1→B→2→A→4→C→3)+(A→6→D→5)→A ∙∙∙), achievable Adaptive multi-coordination control for robots. Specific method: Based on the Matlab/Simulink bionic legged robot control model A, transfer the foot end trajectory and gait parameters 1 to the bionic legged robot kinematics simulation model B based on Matlab/SimMechanics, after Matlab/SimMechanics simulation, the joint motion trajectory and The gait data 2 is summarized to the control model A of the bionic footed robot based on Matlab/Simulink, and the motion gait and posture data 4 is sent to the dynamic model C of the bionic footed robot based on ADAMS, based on the Matlab/Simulink bionic footed robot control model A, After the ADAMS dynamics simulation, the dynamics co-simulation output data 3 is summarized to the control model A of the bionic legged robot based on Matlab/Simulink. At the same time, based on the Matlab/Simulink bionic legged robot control model A, the robot motion gait data packet and the coordination control instruction parameter 6 are sent to the bionic legged robot experimental prototype D. After the robot experimental prototype motion experiment, the robot environment The perception and real-time motion feedback data 5 are summarized into the control model A of the bionic legged robot based on Matlab/Simulink. Integrating dynamics co-simulation output data 3 and real-time motion feedback data 5, the two data are complemented and self-learning control, and the robot adaptive multi-coordination control is realized with the help of simulation data under the condition that the actual robot prototype lacks part of the sensor information .

Claims (5)

1. a kind of bionic legged robot comprehensive simulation strategy based on Matlab, ADAMS and hardware-in-the-loop co-simulation, is characterized in that: The system used includes: Matlab/Simulink-based bionic footed robot control model (A), based on Matlab/SimMechanics bionic footed robot kinematics simulation model (B), ADAMS-based bionic footed robot dynamics model (C), bionic footed robot Robot experimental prototype (D); The strategy includes the real-time kinematics demonstration process of the motion state of the robot experimental prototype, in a specific way: the bionic legged robot experimental prototype (D) wirelessly sends interface robot environment perception and real-time motion feedback data (5) to the bionic legged model based on Matlab/Simulink In the robot control model (A), the bionic legged robot control model (A) based on Matlab/Simulink transfers the foot end trajectory and gait parameters (1) to the bionic legged robot kinematics simulation model (B) based on Matlab/SimMechanics , while Matlab/SimMechanics has the relevant mechanism kinematics animation demonstration function, real-time display of the kinematics state of the robot experimental prototype, the relevant joint motion trajectory and gait data (2) transmitted to the bionic legged robot control model based on Matlab/Simulink In (A), the kinematics data is summarized and saved; The strategy includes the real-time dynamics demonstration process of the motion state of the experimental prototype of the robot. The specific method is: the experimental prototype of the bionic legged robot (D) wirelessly sends the interface robot environment perception and real-time motion feedback data (5) to the bionic legged robot based on Matlab/Simulink In the control model (A), the control model (A) based on Matlab/Simulink bionic legged robot transfers motion gait and attitude data (4) to the bionic legged robot dynamic model (C) based on ADAMS, and ADAMS has The relevant mechanism dynamics animation demonstration function can display the dynamics data of the robot experimental prototype in real time, and the relevant dynamics co-simulation output data (3) is transferred to the bionic footed robot control model (A) based on Matlab/Simulink for dynamics analysis. data aggregation and storage; The strategy includes the pure virtual joint motion simulation process of the robot. The specific method is: based on the Matlab/Simulink bionic footed robot control model (A), the foot end motion trajectory and gait parameters (1) are transferred to the bionic footed robot based on Matlab/SimMechanics. Learning simulation model (B), after Matlab/SimMechanics simulation, the joint trajectory and gait data (2) are summarized into the control model of bionic footed robot based on Matlab/Simulink (A), and the control model of bionic footed robot based on Matlab/Simulink (A) Send the motion gait and attitude data (4) to the ADAMS-based bionic legged robot dynamics model (C), after the ADAMS dynamics simulation, the dynamics co-simulation output data (3) is summarized to the Matlab/ Simulink bionic legged robot control model (A), completed the joint motion simulation of Matlab/SimMechanics and ADAMS; The strategy includes robot gait generation, the real-time motion control process of the robot experimental prototype, in a specific way: based on Matlab/Simulink bionic legged robot control model (A) transmission foot end motion trajectory and gait parameters (1) to the Matlab/Simulink based SimMechanics bionic legged robot kinematics simulation model (B), after Matlab/SimMechanics simulation, the joint trajectory and gait data (2) are summarized into the bionic legged robot control model based on Matlab/Simulink (A), based on Matlab/Simulink The bionic legged robot control model (A) wirelessly sends the robot motion gait data packet and the coordination control instruction parameters (6) to the bionic legged robot experimental prototype (D), and completes the real-time control of the data generated by Matlab/SimMechanics gait Robot experiment prototype movement; The strategy includes robot gait generation, real-time motion control process of the robot experimental prototype after simulation verification, specific method: based on Matlab/Simulink bionic footed robot control model (A) transmission of foot end motion trajectory and gait parameters (1) To the bionic legged robot kinematics simulation model (B) based on Matlab/SimMechanics, after Matlab/SimMechanics simulation, the joint trajectory and gait data (2) are summarized to the bionic legged robot control model (A) based on Matlab/Simulink, Based on the Matlab/Simulink bionic footed robot control model (A), the motion gait and attitude data (4) are sent to the ADAMS-based bionic footed robot dynamic model (C). After the ADAMS dynamics simulation, the dynamics joint simulation The output data (3) is summarized into the control model (A) of the bionic footed robot based on Matlab/Simulink, and the joint motion simulation verification of Matlab/SimMechanics gait generation and ADAMS is completed, and then the robot motion gait data package and coordination control command parameters (6) Send wirelessly to the bionic legged robot experimental prototype (D), complete the joint motion simulation verification by Matlab/SimMechanics gait generation and ADAMS, and control the movement of the robot experimental prototype in real time, more accurately realizing the robot experimental prototype sport control; The strategy includes the robot virtual co-simulation self-learning adjustment process, the specific way: based on the Matlab/Simulink bionic legged robot control model (A) transfer foot end motion trajectory and gait parameters (1) to the Matlab/SimMechanics bionic legged robot Kinematics simulation model (B), after Matlab/SimMechanics simulation, the joint trajectory and gait data (2) are summarized into the control model of the bionic footed robot based on Matlab/Simulink (A), based on the control of the bionic footed robot based on Matlab/Simulink The model (A) sends the motion gait and attitude data (4) to the ADAMS-based bionic legged robot dynamics model (C). After the ADAMS dynamics simulation, the dynamics co-simulation output data (3) is summarized to the /Simulink bionic footed robot control model (A), completed a joint motion simulation of Matlab/SimMechanics and ADAMS, and performed continuous cycle simulation by adjusting parameters based on the self-learning module in the Matlab/Simulink bionic footed robot control model (A) , and finally realize the self-learning adjustment of virtual co-simulation, providing a basis for data optimization and adaptive control; The strategy includes the robot hardware-in-the-loop simulation self-learning adjustment process, the specific way: based on Matlab/Simulink bionic legged robot control model (A) transfer foot end motion trajectory and gait parameters (1) to the bionic legged robot based on Matlab/SimMechanics Kinematics simulation model (B), after Matlab/SimMechanics simulation, the joint trajectory and gait data (2) are summarized into the control model of the bionic footed robot based on Matlab/Simulink (A), based on the control of the bionic footed robot based on Matlab/Simulink Model (A) sends the robot motion gait data packet and coordination control instruction parameters (6) to the bionic legged robot experimental prototype (D), after the robot experimental prototype motion experiment, the robot environment perception and real-time motion feedback data (5 ) to the control model (A) of the bionic footed robot based on Matlab/Simulink, and completed a joint motion simulation of Matlab/SimMechanics and the half-in-the-loop experimental prototype. The module adjusts the parameters, performs continuous loop hardware-in-the-loop simulation, and finally realizes the self-learning adjustment under the hardware-in-the-loop joint simulation, providing a basis for data optimization and adaptive control; The strategy includes the robot self-adaptive multi-coordination control process, the specific way: based on the Matlab/Simulink bionic legged robot control model (A) transfer foot end motion track and gait parameters (1) to the bionic legged robot motion based on Matlab/SimMechanics Learning simulation model (B), after Matlab/SimMechanics simulation, the joint trajectory and gait data (2) are summarized into the control model of bionic footed robot based on Matlab/Simulink (A), and the control model of bionic footed robot based on Matlab/Simulink (A) Send the motion gait and attitude data (4) to the ADAMS-based bionic legged robot dynamics model (C), after the ADAMS dynamics simulation, the dynamics co-simulation output data (3) is summarized to the Matlab/ Simulink bionic legged robot control model (A); at the same time, based on Matlab/Simulink bionic legged robot control model (A), the robot motion gait data packet and coordination control command parameters (6) are sent to the bionic legged robot Experimental prototype (D), after the motion experiment of the robot experimental prototype, the robot environment perception and real-time motion feedback data (5) are summarized into the Matlab/Simulink-based bionic legged robot control model (A); the output data of the integrated dynamics co-simulation ( 3) and real-time motion feedback data, the two data are complementary and self-learning regulation, and the robot adaptive multi-coordination control is realized with the help of simulation data under the condition that the actual robot prototype lacks part of the sensing information. 2. the comprehensive simulation strategy of the bionic legged robot based on Matlab, ADAMS and half-in-the-loop co-simulation according to claim 1, is characterized in that: described based on Matlab/Simulink bionic legged robot control model (A) construction mode, Specific method: first set parameters such as system internal simulation algorithm type, period, step size, and error range in the Matlab/Simulink bionic legged robot simulation system, and then selectively establish a Matlab/Simulink-based bionic simulation system according to the needs of robot debugging. Footed robot motion demonstration module, establishment of bionic footed robot motion gait control module based on Matlab/Simulink, establishment of bionic footed robot motion self-learning module based on Matlab/Simulink, and establishment of bionic footed robot coordination control based on Matlab/Simulink The module is finally connected to the Matlab/Simulink bionic legged robot simulation system interface for data exchange and coordination of data processing. 3. the bionic legged robot comprehensive simulation strategy based on Matlab, ADAMS and half-in-the-loop co-simulation according to claim 1, is characterized in that: described based on Matlab/SimMechanics bionic legged robot kinematics simulation model (B) builds Method, specific method: first create the Matlab/SimMechanics bionic footed robot kinematics simulation environment, set the system environment parameters, and then establish the bionic footed robot system world coordinate system and the degrees of freedom and directions of the kinematic joints of each leg in Matlab/SimMechanics , set the rigid body parameters of each leg member including mass, inertia, length, joint controller type, joint sensor output type, and finally, according to the robot motion control requirements, lead to the input and output of the Matlab/SimMechanics bionic legged robot kinematics simulation model The interface is ready to accept joint input data and send joint output data for data exchange to realize kinematic calculation functions. 4. the comprehensive simulation strategy of the bionic footed robot based on Matlab, ADAMS and half-in-the-loop co-simulation according to claim 1, is characterized in that: described based on ADAMS bionic footed robot dynamics model (C) construction mode, specifically Method: First create the dynamics simulation environment of the ADAMS bionic footed robot, set the system environment parameters, then import the ground rigid body and the rigid body of each leg of the bionic footed robot into ADAMS, and set the geometric size, material type or density or quality of the rigid body parameters, and then set the joint degrees of freedom and joint types between each rigid body in ADAMS, and set the collision type between each foot end and the ground rigid body, friction constraints and related parameter settings in ADAMS, and finally according to actual requirements, set Define the input and output interfaces in the ADAMS bionic legged robot kinematics simulation model, prepare to receive joint input data, send joint output data and foot end collision force data, exchange data, and realize dynamic calculation functions. 5. the comprehensive simulation strategy of the bionic legged robot based on Matlab, ADAMS and half-in-the-loop co-simulation according to claim 1, is characterized in that: described bionic legged robot experimental prototype (D) operation process, concrete mode: first The bionic legged robot experimental prototype (D) is initialized, setting includes system initialization parameters, then the bionic legged robot experimental prototype (D) wirelessly receives the upper task control command signal and gait data, and then performs the bionic legged robot experimental prototype ( D) Motion experiment, at the same time, the attitude sensing, joint angle, angular velocity, foot end force information and the environment recognition of the bionic footed robot experimental prototype (D) are collected in real time during the experiment. Infrared ranging, machine vision recognition of obstacle distribution, and ground condition information are sent wirelessly to the computer control platform.

Images