CN113378475A - Vrep-based quadruped robot control method, system and device - Google Patents
Vrep-based quadruped robot control method, system and device Download PDFInfo
- Publication number
- CN113378475A CN113378475A CN202110717553.9A CN202110717553A CN113378475A CN 113378475 A CN113378475 A CN 113378475A CN 202110717553 A CN202110717553 A CN 202110717553A CN 113378475 A CN113378475 A CN 113378475A
- Authority
- CN
- China
- Prior art keywords
- quadruped robot
- entity
- simulation
- robot
- quadruped
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000004088 simulation Methods 0.000 claims abstract description 159
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 75
- 238000011161 development Methods 0.000 claims abstract description 54
- 238000013508 migration Methods 0.000 claims abstract description 14
- 230000005012 migration Effects 0.000 claims abstract description 14
- 238000003062 neural network model Methods 0.000 claims abstract description 7
- 230000001133 acceleration Effects 0.000 claims description 23
- 230000003068 static effect Effects 0.000 claims description 22
- 238000013461 design Methods 0.000 claims description 20
- 238000013528 artificial neural network Methods 0.000 claims description 13
- 230000006870 function Effects 0.000 claims description 11
- 230000033001 locomotion Effects 0.000 claims description 11
- 238000012549 training Methods 0.000 claims description 11
- 239000007787 solid Substances 0.000 claims description 10
- 230000004913 activation Effects 0.000 claims description 7
- 210000001503 joint Anatomy 0.000 claims description 6
- 230000004044 response Effects 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 4
- 238000004891 communication Methods 0.000 claims description 3
- 230000009471 action Effects 0.000 claims description 2
- 230000005540 biological transmission Effects 0.000 claims description 2
- 238000004364 calculation method Methods 0.000 abstract description 4
- 238000010586 diagram Methods 0.000 description 6
- 230000008901 benefit Effects 0.000 description 4
- 238000005094 computer simulation Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000008713 feedback mechanism Effects 0.000 description 1
- 230000009916 joint effect Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Manipulator (AREA)
- Feedback Control In General (AREA)
Abstract
The invention discloses a method, a system and a device for controlling a four-legged robot based on Vrep, wherein the method comprises the following steps: establishing a relationship between simulation and an entity: establishing a simulated quadruped robot of the entity quadruped robot by utilizing the Vrep, and keeping the performances of the quadruped robot and the simulated quadruped robot consistent; algorithm development and migration: the feasibility of algorithm development at the Vrep is checked, and after the algorithm development is completed, a control instruction is sent to the entity quadruped robot through a network serial port to complete the algorithm migration; teleoperation control of the robot: and controlling the entity quadruped robot through a network serial port. According to the invention, by establishing an accurate simulation robot model, the development cost and the development period of the robot can be reduced, and the problems of more parameters, low precision and large calculation amount are solved through the MLP neural network model; through deviation calibration, the program can be actively stopped when the material object and the simulation have larger deviation, so that damage to a robot is prevented; and a quadruped robot teleoperation control platform is established, so that the development of control software is saved.
Description
Technical Field
The invention relates to the field of robots, in particular to a method, a system and a device for controlling a four-legged robot based on Vrep.
Background
At present, the quadruped robot is widely applied to scenes with high danger coefficient and severe working environment, such as rescue and relief work, polar exploration and the like, and becomes a key object of people's research in the field of mobile robots. The application range of the quadruped robot is greatly expanded, and the original wheel type robot is high in speed but cannot adapt to complex terrains such as forests and mountains. Although the track robot has strong off-road capability, the track robot is large in size and poor in flexibility, and the track robot is difficult to pass through in places with dense obstacles. The quadruped robot has the advantages of flexibility, small size and strong off-road capability, and is widely applied to the fields of exploration and detection.
Although the quadruped robot has better cross-country capability, the control algorithm of the quadruped robot is complex, and different control algorithms exist in different environments such as stairs and gravel roads, so that the development period of the quadruped robot is longer, the design difficulty is high, and the damage of the quadruped robot is easily caused by long-time operation of the entity robot.
The prior quadruped robot development and design method is mainly to directly program an entity robot platform, and directly transfer a program to the quadruped robot after software simulation. Compared with a wheeled robot, the quadruped robot has a more complex mechanical structure, the design method is poor in generalization when the physical robot is directly used for development, when the differences of parameters and performances of simulation and the physical robot are large, damage is easily caused to the physical robot, and simulation and the entity are difficult to combine for some complex algorithms.
At present, when a motor of the robot is simulated, an equation is mostly used for establishing a response curve of the motor, and simulation parameters are designed according to a parameter model of the motor. The method has high requirement on the measurement accuracy of the motor parameters, the motor parameters are numerous, and the influence of friction force, current change and the like on the motor parameters is large, so that a stable simulation model is difficult to establish.
The development and design of the conventional quadruped robot lacks an efficient development and design method, and the development and debugging method based on the physical platform has low development efficiency and poor mobility and is extremely easy to damage robots.
Disclosure of Invention
In order to overcome the defect of low development efficiency of the quadruped robot in the prior art, the invention provides a method, a system and a device for controlling the quadruped robot based on Vrep.
The technical problem of the invention is solved by the following technical scheme:
the invention provides a Vrep-based quadruped robot control method, which is characterized by comprising the following steps of: s1: establishing a relationship between simulation and an entity: receiving an instruction of a user, and establishing a simulated quadruped robot of the entity quadruped robot by utilizing a Vrep simulation environment, so that the performances of the entity quadruped robot and the simulated quadruped robot are kept consistent; s2: algorithm development and migration: receiving a user instruction, checking feasibility of algorithm development in a Vrep simulation environment, and after the algorithm development is completed, sending a control instruction to the entity quadruped robot through a network serial port to complete algorithm migration; s3: teleoperation control of the robot: and receiving the instruction of the user, and controlling the entity quadruped robot through the network serial port.
In some embodiments, the step of S1 includes the steps of: s11: establishing static parameter simulation, namely establishing a simulated quadruped robot according to the entity quadruped robot, measuring static physical parameters of the entity quadruped robot, and establishing a hardware framework of the simulated quadruped robot; s12: establishing dynamic parameter simulation, simulating and matching motor parameters, friction force and quality response parameters between the entity quadruped robot and the simulation quadruped robot, and establishing a motor model of the simulation quadruped robot by acquiring motor data of the entity quadruped robot; s13: and (4) deviation calibration, namely establishing connection between the simulation quadruped robot and the entity quadruped robot through a network serial port, adjusting relevant parameters by controlling the movement of the entity quadruped robot in a virtual environment, and regularly updating information of each joint of the quadruped robot by utilizing the network serial port to reduce accumulated errors. .
In some embodiments, the step of S11 includes the steps of: s111: acquiring parameters of each part of the solid quadruped robot: measuring the quality, length and centroid position of each joint of the entity quadruped robot; s112: and (3) building a simulation quadruped robot according to the parameter information: and (4) building a leg joint model and a body model of the simulation quadruped robot.
In some embodiments, the step of S12 includes the steps of: s121: motor data acquisition: the input quantity is the speed, the acceleration, the position, the moment and the input voltage of the motor at the current moment, wherein the control variable is the input voltage; the output quantity is information of speed, acceleration, position and moment at the next moment; s122: training a neural network: establishing a multilayer neural network which comprises an input layer, a hidden layer and an output layer, wherein all the layers are connected through an activation function; obtaining a neural network model of the motor through training iteration; s123: simulation test: and (3) importing the model into a Vrep simulation platform, simulating the dynamic parameters of the motor of the entity quadruped robot, and controlling the application of the motor of the simulation robot through a voltage signal.
In some embodiments, in step S122: training a network model by acquiring voltage, speed and acceleration parameters of a motor at different moments; and outputting the speed, acceleration and position parameters of the motor at the next moment according to the speed, acceleration and position of the current motor and the input voltage information.
In some embodiments, the step of S13 includes the steps of: s131: running a program, and setting the position of a motor in the simulation program; s132: sending the instruction to the entity quadruped robot through the network serial port, and executing the command by the entity quadruped robot; s133: feeding back position, speed and attitude parameter information by the entity quadruped robot; s134: comparing the simulation data with the entity data, if the deviation is greater than Error1, suspending the program, and restarting after the parameters are modified; and if the deviation is not greater than Error1, updating the simulation parameters and continuing to run the program.
In some embodiments, the step of S2 includes: s21: developing an algorithm based on a simulation platform; s22: migrating an entity robot algorithm; s23: and (6) optimizing an algorithm.
In some embodiments, the step of S3 includes: the simulation quadruped robot is connected with the entity quadruped robot in a network serial port communication mode; the simulation quadruped robot sends a motor control command to the entity quadruped robot; the entity quadruped robot returns the speed, position and posture information of each motor to the simulation quadruped robot, so that the speed and position of the motor of the entity quadruped robot are controlled, the movement of the entity quadruped robot is controlled, and the posture information feedback of the entity quadruped robot is controlled.
The invention also provides a Vrep-based quadruped robot control system, which is characterized by comprising: the system comprises a simulation and entity relationship establishing module, an algorithm development and design module and a robot teleoperation control module; the simulation and entity relationship establishing module is used for establishing a parameter relationship between the entity quadruped robot and the simulation quadruped robot so as to keep the performances of the entity quadruped robot and the simulation quadruped robot consistent; the algorithm development design module is used for simulating a development design algorithm on a Vrep software platform and checking the feasibility of the algorithm; after the algorithm simulation is completed, the actions of all joints during the robot simulation are sent to the entity quadruped robot by using the network serial port, and the algorithm is transferred to the entity robot; the robot teleoperation control module is used for controlling the entity quadruped robot through a network serial port after the simulation quadruped robot is established, and feeding back the state of the entity quadruped robot in real time in software.
In some embodiments, the simulation and entity relationship establishment module comprises: a static parameter simulation submodule, a dynamic parameter simulation submodule and a deviation calibration submodule; the static parameter simulation submodule is used for building a simulated quadruped robot according to the entity quadruped robot, measuring static physical parameters of the entity quadruped robot and building a hardware framework of the simulated quadruped robot; the dynamic parameter simulation submodule is used for simulating and matching various motor parameters, friction force and quality response parameters between the entity quadruped robot and the simulated quadruped robot; establishing a motor model of the simulated quadruped robot by acquiring motor data of the solid quadruped robot; the deviation calibration submodule is used for establishing the connection between the simulation quadruped robot and the entity quadruped robot through a network serial port and adjusting related parameters by controlling the movement of the entity quadruped robot in a virtual environment; and the information of each joint of the quadruped robot is updated regularly by utilizing a network serial port to reduce the accumulated error.
The invention also provides a four-footed robot control device based on Vrep, which comprises: the system comprises an entity quadruped robot, a router, a computer, an electronic scale, a force sensor and a meter ruler; the entity quadruped robot is used for carrying an embedded computer and sending joint position and acceleration parameter information with an upper computer through a network serial port; the router is used for being responsible for network information transmission; the computer is used for running Vrep simulation software; the electronic scale is used for acquiring the mass of each joint of the entity quadruped robot; the force sensor is used for detecting the motor moment of the solid quadruped robot; the meter ruler is used for measuring the size of the solid quadruped robot.
Compared with the prior art, the invention has the following beneficial effects: according to the invention, by establishing an accurate simulation robot model, the development cost of the robot can be reduced, the development period of the robot is reduced, and the simulation platform can more intuitively reflect the parameter information of the robot, so that the algorithm development is more convenient and faster; the MLP neural network model is used for fitting the input-output relationship of the motor, so that the problems of various parameters, low precision and large calculation amount when a parameter equation is used for fitting the input-output relationship of the motor can be solved.
In some embodiments, the invention has the following advantages compared with the prior art: according to the invention, through a deviation calibration algorithm, the program can be actively stopped when a large deviation occurs between a real object and simulation, so that damage to a robot is prevented;
in some embodiments, the invention has the following advantages compared with the prior art: the invention establishes the quadruped robot teleoperation control platform while completing the quadruped robot modeling, thereby saving the development of robot control software.
Drawings
FIG. 1 is a schematic flow chart of a quadruped robot control method according to an embodiment of the present invention;
FIG. 2 is a diagram of virtual and physical robot relationships according to an embodiment of the present invention;
FIG. 3 is a flowchart of the operation of the static parameter simulation setup submodule according to an embodiment of the present invention;
FIG. 4 is a flow chart of the operation of the dynamic parameter simulation submodule of an embodiment of the present invention;
FIG. 5 is a schematic diagram of establishing an MLP multi-layer neural network according to an embodiment of the present invention;
FIG. 6 is a diagram illustrating a neural network activation function according to an embodiment of the present invention;
FIG. 7 is a flow chart of the offset calibration sub-module of an embodiment of the present invention;
FIG. 8 is a software development sub-module workflow diagram of an embodiment of the present invention.
Detailed Description
The invention will be further described with reference to the accompanying drawings and preferred embodiments. It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict.
It should be noted that the terms of orientation such as left, right, up, down, top and bottom in the present embodiment are only relative concepts to each other or are referred to the normal use state of the product, and should not be considered as limiting.
In order to solve the problems of high algorithm design difficulty and long development period of a quadruped robot, a control method, a system and a device of the quadruped robot based on Vrep are provided. Aiming at a physical platform of a quadruped robot, firstly establishing a 1:1 simulation development model by using Vrep, namely simulating the quadruped robot; and the parameters of the simulation model are adjusted through a feedback mechanism, so that the parameters of the simulation model are consistent with the parameters of the entity quadruped robot. When the algorithm is developed, a robot control algorithm is firstly developed in a virtual scene of simulation software, and after the algorithm is designed, the algorithm is transplanted to an entity quadruped robot.
In order to simulate a dynamic model of a robot motor as much as possible, the invention provides a method for fitting a motor model by using an MLP neural network, wherein the network model is trained by acquiring parameters such as voltage, speed and acceleration of the motor at different moments, and the acquired network model can output parameters such as the speed, the acceleration and the position of the motor at the next moment according to information such as the speed, the acceleration and the position of the current motor and the input voltage; the problems of more parameters and low precision when a parameter equation is used for establishing the motor simulation model are solved.
After the whole development process is completed, a set of control program capable of directly running on the entity quadruped robot can be obtained, a simulation model consistent with physical parameters can be obtained, the feasibility of the robot design is verified in the simulation in the later development, and then the algorithm is directly transferred to the physical model, so that the complexity of the algorithm design is greatly reduced, and the damage of the robot is avoided when the algorithm development is carried out on the entity quadruped robot. In addition, the simulation platform for establishing simulation can be used as a teleoperation control system of the robot to directly control the motion of the entity quadruped robot, and the development time of control software of the quadruped robot is reduced.
Referring to fig. 1, a flow chart of a control method of a quadruped robot is shown.
The invention provides a four-legged robot control method based on a Vrep simulation platform, which comprises three parts, namely simulation and entity relation establishment, algorithm development design and a robot teleoperation control module; the method comprises the following specific steps: s1: establishing a relationship between simulation and an entity; establishing a simulation quadruped robot of the entity quadruped robot by utilizing a Vrep simulation environment, so that the performances of the entity quadruped robot and the simulation quadruped robot are kept consistent; s2: algorithm development and migration; carrying out algorithm development in a Vrep simulation environment; after algorithm development is completed, a control instruction is sent to the entity quadruped robot through a network serial port, and algorithm migration is completed; s3: teleoperation control of the robot; and controlling the entity quadruped robot through a network serial port.
The establishment of the simulation and entity relationship mainly establishes a simulation system, namely, a parameter relationship between the simulation quadruped robot and the entity quadruped robot, so that the performances of the simulation quadruped robot and the entity quadruped robot are kept consistent in all aspects. The algorithm development design mainly carries out rapid algorithm migration after the simulation relation is established. The robot teleoperation control module mainly realizes teleoperation control of the robot on a Vrep software control platform.
The step S1 of simulating the establishment of the entity relationship includes the following steps: s11: establishing static parameter simulation; s12: establishing dynamic parameter simulation; s13: and (5) calibrating deviation.
The step S11 of the static parameter simulation setup includes the following steps: s111: acquiring parameters of each part of the entity quadruped robot; measuring the quality, length and centroid position of each joint of the entity quadruped robot; s112: constructing a simulation quadruped robot according to the parameter information; and (4) building a leg joint model and a body model of the simulation quadruped robot.
The step S12 of dynamic parameter simulation establishment includes the following steps: s121: collecting motor data; the input quantity is the speed, the acceleration, the position, the moment and the input voltage of the motor at the current moment, wherein the control variable is the input voltage; the output quantity is information of speed, acceleration, position and moment at the next moment; s122: training a neural network; establishing a multilayer neural network which comprises an input layer, a hidden layer and an output layer, wherein all the layers are connected through an activation function; obtaining a neural network model of the motor through training iteration; s123: carrying out simulation test; and (3) importing the model into a Vrep simulation platform, simulating the dynamic parameters of the motor of the entity quadruped robot, and controlling the application of the motor of the simulation robot through a voltage signal.
In the step S122 of neural network training, a network model is trained by acquiring voltage, speed and acceleration parameters of the motor at different moments; and outputting the speed, acceleration and position parameters of the motor at the next moment according to the speed, acceleration and position of the current motor and the input voltage information.
The step S13 of offset calibration includes the following steps: s131: running a program, and setting the position of a motor in the simulation program; s132: sending the instruction to the entity quadruped robot through the network serial port, and executing the command by the entity quadruped robot; s133: feeding back position, speed and attitude parameter information by the entity quadruped robot; s134: comparing the simulation data with the entity data; if the deviation is larger than Error1, suspending the program, and restarting after the parameters are modified; and if the deviation is not greater than Error1, updating the simulation parameters and continuing to run the program.
The step S2 of algorithm development migration includes: s21: developing an algorithm based on a simulation platform; s22: migrating an entity robot algorithm; s23: and (6) optimizing an algorithm.
Referring to fig. 2, a diagram of virtual and physical robot relationships; the simulation quadruped robot 1 in the Vrep simulation environment is connected with the entity quadruped robot 2 in a network serial port communication mode; the simulation quadruped robot 1 in the Vrep simulation environment sends a motor control command to the entity quadruped robot 2, the entity quadruped robot 2 returns speed, position and posture information of each motor to the simulation quadruped robot 1 in the Vrep simulation environment, the speed and the position of the motor of the entity quadruped robot 2 are controlled, the movement of the entity quadruped robot 2 is controlled, and the posture information feedback of the entity quadruped robot 2 is controlled.
The invention provides a Vrep-based quadruped robot efficient control system, which comprises: the system comprises a simulation and entity relationship establishing module, an algorithm development and design module and a robot teleoperation control module.
The simulation and entity relationship establishing module comprises: the device comprises a static parameter simulation submodule, a dynamic parameter simulation submodule and a deviation calibration submodule. The simulation and entity relationship building module is mainly used for building 1:1, keeping the performance of the entity quadruped robot and the simulation quadruped robot consistent by using a parameter relation; the accuracy of the related parameters directly influences the errors of the later algorithm development design module.
The static parameter simulation submodule is mainly responsible for simulating the static parameter according to the solid quadruped robot 1:1, building a corresponding virtual simulation model, and mainly finishing a hardware simulation part of the virtual robot. The static parameter simulation building submodule is mainly used for building a hardware frame of the robot and measuring static physical parameters of the robot through an electronic scale, a meter ruler and the like. The static parameter simulation establishes the work flow of the submodule: first, a simulation model is built in Vrep according to an entity quadruped robot. Referring to fig. 3, the static parameter simulation building submodule is used for acquiring parameters of each part of the physical robot, measuring the quality of each joint of the robot, measuring the length of each joint of the robot, and measuring the centroid position of the robot; and then the method is used for building a virtual robot, and a leg joint model and a body model of the simulation robot are built. Firstly, parameters such as mass, length, mass center and the like of each joint of the entity quadruped robot are obtained by using tools such as an electronic scale, a meter ruler and the like, and then a corresponding simulation robot is established in simulation software according to the parameter information.
And the dynamic parameter simulation submodule is mainly responsible for simulating and matching response parameters such as motor parameters, friction force, quality and the like between the entity quadruped robot and the simulation robot. And the dynamic parameter simulation submodule is used for establishing a motor model in simulation by collecting motor data. The work flow of the dynamic parameter simulation submodule refers to fig. 4, and includes motor data acquisition, neural network training and simulation testing. Referring to fig. 5, an MLP multi-layer neural network is established, which includes an input layer, a hidden layer, and an output layer, and the respective layers are connected by an activation function. The input quantity is the speed, the acceleration, the position, the moment and the input voltage of the motor at the current moment, wherein the control variable is the input voltage, and the output quantity is the speed, the acceleration, the position and the moment information at the next moment. Referring to fig. 6, a diagram of an activation function of a neural network is shown, and the activation function is a sigmoid function. A neural network model of the motor can be obtained through training for multiple iterations, dynamic parameters of the motor of the solid robot can be simulated by introducing the model into a Vrep simulation platform, and the application of the motor of the simulation robot is controlled through voltage signals.
After the dynamic parameters and the static parameters of the quadruped robot are established, some errors exist inevitably; and the deviation calibration submodule is used for regularly updating the information of each joint of the quadruped robot by utilizing a network serial port to reduce the accumulated error. And the deviation calibration submodule establishes connection between the virtual robot and the entity robot through a network serial port and adjusts related parameters by controlling the movement of the entity robot in a virtual environment. Referring to fig. 7, firstly, a program is run, a motor position is set in the simulation program, an instruction is sent to the physical robot through a network serial port, the physical robot executes a command, the physical robot feeds back parameter information such as a position and a speed parameter, and simulation data are compared with physical data; if the deviation is greater than Error1, the program is halted; if the deviation is not larger than Error1, the simulation parameters are updated, and the simulation program is continuously run. The module is used for controlling the entity robot and the simulation robot to move simultaneously through a network serial port in Vrep simulation software after establishing a static and dynamic model of the quadruped robot, detecting deviation existing in the moving process of the quadruped robot and the simulation robot, establishing a negative feedback control method, feeding back joint information of the entity robot by using the network serial port at intervals, comparing the joint information with simulation information, and having a deviation alarm mechanism.
The algorithm development design module mainly realizes the development of a high-order algorithm of the quadruped robot, completes the program design under complex road conditions such as stairs and gravel roads, and completes the algorithm migration from a simulation algorithm to an entity robot platform. And simulating the designed algorithm on a Vrep software platform, and checking the feasibility of the algorithm. After algorithm simulation is completed by using Vrep software, joint actions during robot simulation are sent to the quadruped robot by using a network serial port, and the algorithm is transferred to the entity robot. Referring to fig. 8, the software development submodule includes: algorithm development based on a simulation platform, entity robot algorithm migration and algorithm optimization. After the simulation model of the entity quadruped robot is established, the algorithm of the quadruped robot can be developed and designed in a simulation environment under the condition that the entity robot does not need to be connected. Because the parameters of the quadruped robot in simulation and the parameters of the entity quadruped robot are kept consistent through simulation environment modeling, the development and the design of a calculation method in the simulation environment are high in speed and low in cost, related parameters can be better visualized, and the development time is greatly shortened. After algorithm development is completed in a simulation environment, the control instruction can be directly sent to the entity robot through the network serial port, and algorithm migration is completed.
The robot teleoperation control module mainly realizes the movement of the robot under the control of a software platform. After the simulation model of the quadruped robot is established, the robot can be directly controlled through a network serial port, and the state of the entity robot can be fed back in real time in software. After the robot simulation modeling is completed, the entity robot can be controlled through a network serial port, and the robot simulation modeling method has the following functions: 1. controlling the speed and position of a robot motor; 2. movement of the quadruped robot; 3. and (5) feeding back the posture information of the quadruped robot.
Referring to table 1, the efficient control device for the Vrep-based quadruped robot in the embodiment of the invention comprises a quadruped robot, a router, a computer, an electronic scale, a force sensor and a meter ruler. The quadruped robot is used for carrying an embedded computer and can send parameter information such as joint position, acceleration and the like with an upper computer through a network serial port; the router is used for transmitting network information; the computer is used for running Vrep simulation software; the electronic scale is used for acquiring the weight of each joint of the quadruped robot; the force sensor is used for detecting the motor moment of the quadruped robot; the meter ruler is used for measuring the size of the solid quadruped robot.
TABLE 1 device composition table
The method has the advantages that the MLP neural network model is used for fitting the input-output relationship of the motor, so that the problems of various parameters, low precision and large calculation amount when a parameter equation is used for fitting the input-output relationship of the motor can be solved; by establishing an accurate simulation robot model, the development cost of the robot can be reduced, the development period of the robot can be reduced, and the simulation platform can more intuitively reflect the parameter information of the robot, so that the algorithm development is more convenient and faster; by means of a deviation calibration algorithm, the program can be actively stopped when a large deviation occurs between a real object and simulation, and damage to a robot is prevented; the quadruped robot remote operation control platform is established while the quadruped robot modeling is completed, and the development of robot control software is saved.
The invention can be applied to the fields of simulation and modeling of the quadruped robot, teleoperation control of the quadruped robot, algorithm development of the quadruped robot and the like. The functions of the device are as follows: the capability of establishing a high-precision simulation model of the quadruped robot is realized; the capability of simulating the rapid algorithm migration to a real object is realized; the capability of accurately simulating the input and output relations of the robot motor is realized; the automatic power-off protection function is realized when the difference between simulation parameters of the simulation and the physical robot is large; the method has the capability of performing algorithm simulation on a software platform; has the teleoperation control capability of the quadruped robot.
The foregoing is a more detailed description of the invention in connection with specific preferred embodiments and it is not intended that the invention be limited to these specific details. For those skilled in the art to which the invention pertains, several equivalent substitutions or obvious modifications can be made without departing from the spirit of the invention, and all the properties or uses are considered to be within the scope of the invention.
Claims (11)
1. A four-foot robot control method based on Vrep is characterized by comprising the following steps:
s1: establishing a relationship between simulation and an entity: receiving an instruction of a user, and establishing a simulated quadruped robot of the entity quadruped robot by utilizing a Vrep simulation environment, so that the performances of the entity quadruped robot and the simulated quadruped robot are kept consistent;
s2: algorithm development and migration: receiving a user instruction, and developing an algorithm in a Vrep simulation environment; after algorithm development is completed, a control instruction is sent to the entity quadruped robot through a network serial port, and algorithm migration is completed;
s3: teleoperation control of the robot: and receiving the instruction of the user, and controlling the entity quadruped robot through the network serial port.
2. The Vrep-based quadruped robot control method of claim 1, characterized by:
the step of S1 includes the steps of:
s11: establishing static parameter simulation, namely establishing a simulated quadruped robot according to the entity quadruped robot, measuring static physical parameters of the entity quadruped robot, and establishing a hardware framework of the simulated quadruped robot;
s12: establishing dynamic parameter simulation, simulating and matching motor parameters, friction force and quality response parameters between the entity quadruped robot and the simulation quadruped robot, and establishing a motor model of the simulation quadruped robot by acquiring motor data of the entity quadruped robot;
s13: and (4) deviation calibration, namely establishing connection between the simulation quadruped robot and the entity quadruped robot through a network serial port, adjusting relevant parameters by controlling the movement of the entity quadruped robot in a virtual environment, and regularly updating information of each joint of the quadruped robot by utilizing the network serial port to reduce accumulated errors.
3. The Vrep-based quadruped robot control method of claim 2, characterized by:
the step of S11 includes the steps of:
s111: acquiring parameters of each part of the entity quadruped robot; measuring the quality, length and centroid position of each joint of the entity quadruped robot;
s112: constructing a simulation quadruped robot according to the parameter information; and (4) building a leg joint model and a body model of the simulation quadruped robot.
4. The Vrep-based quadruped robot control method of claim 2, characterized by:
the step of S12 includes the steps of:
s121: motor data acquisition: the input quantity is the speed, the acceleration, the position, the moment and the input voltage of the motor at the current moment, wherein the control variable is the input voltage; the output quantity is information of speed, acceleration, position and moment at the next moment;
s122: training a neural network: establishing a multilayer neural network which comprises an input layer, a hidden layer and an output layer, wherein all the layers are connected through an activation function; obtaining a neural network model of the motor through training iteration;
s123: simulation test: and (3) importing the model into a Vrep simulation platform, simulating the dynamic parameters of the motor of the entity quadruped robot, and controlling the application of the motor of the simulation robot through a voltage signal.
5. The Vrep-based quadruped robot control method according to claim 4, wherein in step S122: training a network model by acquiring voltage, speed and acceleration parameters of a motor at different moments; and outputting the speed, acceleration and position parameters of the motor at the next moment according to the speed, acceleration and position of the current motor and the input voltage information.
6. The Vrep-based quadruped robot control method of claim 2, characterized by:
the step of S13 includes the steps of:
s131: running a program, and setting the position of a motor in the simulation program;
s132: sending the instruction to the entity quadruped robot through the network serial port, and executing the command by the entity quadruped robot;
s133: feeding back position, speed and attitude parameter information by the entity quadruped robot;
s134: comparing the simulation data with the entity data, if the deviation is greater than Error1, suspending the program, and restarting after the parameters are modified; and if the deviation is not greater than Error1, updating the simulation parameters and continuing to run the program.
7. The Vrep-based quadruped robot control method of claim 1, characterized by:
the step of S2 includes: s21: developing an algorithm based on a simulation platform; s22: migrating an entity robot algorithm; s23: and (6) optimizing an algorithm.
8. The Vrep-based quadruped robot control method of claim 1, characterized by:
the step of S3 includes:
the simulation quadruped robot is connected with the entity quadruped robot in a network serial port communication mode;
the simulation quadruped robot sends a motor control command to the entity quadruped robot;
the entity quadruped robot returns the speed, position and posture information of each motor to the simulation quadruped robot, so that the speed and position of the motor of the entity quadruped robot are controlled, the movement of the entity quadruped robot is controlled, and the posture information feedback of the entity quadruped robot is controlled.
9. A Vrep-based quadruped robot control system comprising:
the system comprises a simulation and entity relationship establishing module, an algorithm development and design module and a robot teleoperation control module;
the simulation and entity relationship establishing module is used for establishing a parameter relationship between the entity quadruped robot and the simulation quadruped robot so as to keep the performances of the entity quadruped robot and the simulation quadruped robot consistent;
the algorithm development design module is used for simulating a development design algorithm on a Vrep software platform and checking the feasibility of the algorithm; after the algorithm simulation is completed, the actions of all joints during the robot simulation are sent to the entity quadruped robot by using the network serial port, and the algorithm is transferred to the entity robot;
the robot teleoperation control module is used for controlling the entity quadruped robot through a network serial port after the simulation quadruped robot is established, and feeding back the state of the entity quadruped robot in real time in software.
10. The Vrep based quadruped robot control system of claim 9, wherein:
the simulation and entity relationship establishing module comprises: a static parameter simulation submodule, a dynamic parameter simulation submodule and a deviation calibration submodule;
the static parameter simulation submodule is used for building a simulated quadruped robot according to the entity quadruped robot, measuring static physical parameters of the entity quadruped robot and building a hardware framework of the simulated quadruped robot;
the dynamic parameter simulation submodule is used for simulating and matching various motor parameters, friction force and quality response parameters between the entity quadruped robot and the simulated quadruped robot; establishing a motor model of the simulated quadruped robot by acquiring motor data of the solid quadruped robot;
the deviation calibration submodule is used for establishing the connection between the simulation quadruped robot and the entity quadruped robot through a network serial port and adjusting related parameters by controlling the movement of the entity quadruped robot in a virtual environment; and the information of each joint of the quadruped robot is updated regularly by utilizing a network serial port to reduce the accumulated error.
11. A Vrep-based quadruped robot control device, comprising:
the system comprises an entity quadruped robot, a router, a computer, an electronic scale, a force sensor and a meter ruler;
the entity quadruped robot is used for carrying an embedded computer and sending joint position and acceleration parameter information with an upper computer through a network serial port;
the router is used for being responsible for network information transmission;
the computer is used for running Vrep simulation software;
the electronic scale is used for acquiring the mass of each joint of the entity quadruped robot;
the force sensor is used for detecting the motor moment of the solid quadruped robot;
the meter ruler is used for measuring the size of the solid quadruped robot.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110717553.9A CN113378475B (en) | 2021-06-28 | 2021-06-28 | Four-foot robot control method, system and device based on Vrep |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110717553.9A CN113378475B (en) | 2021-06-28 | 2021-06-28 | Four-foot robot control method, system and device based on Vrep |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113378475A true CN113378475A (en) | 2021-09-10 |
| CN113378475B CN113378475B (en) | 2023-06-16 |
Family
ID=77579520
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110717553.9A Active CN113378475B (en) | 2021-06-28 | 2021-06-28 | Four-foot robot control method, system and device based on Vrep |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113378475B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114839880A (en) * | 2022-06-02 | 2022-08-02 | 淮阴工学院 | Self-adaptive control method based on flexible joint mechanical arm |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070073442A1 (en) * | 2005-09-28 | 2007-03-29 | Canadian Space Agency | Robust impedance-matching of manipulators interacting with unknown environments |
| CN105446821A (en) * | 2015-11-11 | 2016-03-30 | 哈尔滨工程大学 | Improved neural network based fault diagnosis method for intelligent underwater robot propeller |
| CN111208822A (en) * | 2020-02-17 | 2020-05-29 | 清华大学深圳国际研究生院 | Quadruped robot gait control method based on reinforcement learning and CPG controller |
| CN112440281A (en) * | 2020-11-16 | 2021-03-05 | 浙江大学 | Robot trajectory planning method based on digital twins |
| CN112560263A (en) * | 2020-12-11 | 2021-03-26 | 太原理工大学 | Mobile robot state monitoring and maintenance system based on digital twins |
| CN112631131A (en) * | 2020-12-19 | 2021-04-09 | 北京化工大学 | Motion control self-generation and physical migration method for quadruped robot |
| US20210138651A1 (en) * | 2019-11-11 | 2021-05-13 | Rockwell Automation Technologies, Inc. | Robotic digital twin control with industrial context simulation |
-
2021
- 2021-06-28 CN CN202110717553.9A patent/CN113378475B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070073442A1 (en) * | 2005-09-28 | 2007-03-29 | Canadian Space Agency | Robust impedance-matching of manipulators interacting with unknown environments |
| CN105446821A (en) * | 2015-11-11 | 2016-03-30 | 哈尔滨工程大学 | Improved neural network based fault diagnosis method for intelligent underwater robot propeller |
| US20210138651A1 (en) * | 2019-11-11 | 2021-05-13 | Rockwell Automation Technologies, Inc. | Robotic digital twin control with industrial context simulation |
| CN111208822A (en) * | 2020-02-17 | 2020-05-29 | 清华大学深圳国际研究生院 | Quadruped robot gait control method based on reinforcement learning and CPG controller |
| CN112440281A (en) * | 2020-11-16 | 2021-03-05 | 浙江大学 | Robot trajectory planning method based on digital twins |
| CN112560263A (en) * | 2020-12-11 | 2021-03-26 | 太原理工大学 | Mobile robot state monitoring and maintenance system based on digital twins |
| CN112631131A (en) * | 2020-12-19 | 2021-04-09 | 北京化工大学 | Motion control self-generation and physical migration method for quadruped robot |
Non-Patent Citations (6)
| Title |
|---|
| LINQI YE ET AL: "Multi-task Control for a Quadruped Robot with Changeable Leg Configuration", 《2020 IEEE/RSJ INTERNATIONAL CONFERENCE ON ROBOTICS AND BIOMIMETICS》 * |
| LINQI YE ET AL: "Multi-task Control for a Quadruped Robot with Changeable Leg Configuration", 《2020 IEEE/RSJ INTERNATIONAL CONFERENCE ON ROBOTICS AND BIOMIMETICS》, 10 February 2021 (2021-02-10), pages 3944 - 3949 * |
| 乐斌等: "V-REP机器人仿真远程控制方法研究", 《工业控制计算机》 * |
| 乐斌等: "V-REP机器人仿真远程控制方法研究", 《工业控制计算机》, vol. 31, no. 09, 25 September 2018 (2018-09-25), pages 41 - 43 * |
| 钟云胜 等: "基于 MLP 和 SPSO 的机器人行为选择与运动控制方法", 《计算机应用研究》 * |
| 钟云胜 等: "基于 MLP 和 SPSO 的机器人行为选择与运动控制方法", 《计算机应用研究》, 31 August 2018 (2018-08-31), pages 2379 - 2382 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114839880A (en) * | 2022-06-02 | 2022-08-02 | 淮阴工学院 | Self-adaptive control method based on flexible joint mechanical arm |
| CN114839880B (en) * | 2022-06-02 | 2024-04-19 | 淮阴工学院 | Self-adaptive control method based on flexible joint mechanical arm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113378475B (en) | 2023-06-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA3080184C (en) | Deployment of trained neural network based rss fingerprint dataset | |
| CA3049070C (en) | Neural network training for mobile device rss fingerprint-based indoor navigation | |
| CN106030430A (en) | Multi-sensor fusion for robust autonomous filght in indoor and outdoor environments with a rotorcraft micro-aerial vehicle (MAV) | |
| WO2017031718A1 (en) | Modeling method of deformation motions of elastic object | |
| Lin et al. | A leg configuration measurement system for full-body pose estimates in a hexapod robot | |
| CN111156998A (en) | Mobile robot positioning method based on RGB-D camera and IMU information fusion | |
| Gechter et al. | Virtual intelligent vehicle urban simulator: Application to vehicle platoon evaluation | |
| CA3077683C (en) | Crowd-sourced training of a neural network for rss fingerprinting | |
| CN106607910B (en) | A kind of robot imitates method in real time | |
| WO2018056391A1 (en) | Method for creating positioning geomagnetism map, position measurement method, noise measurement method, and system for creating positioning geomagnetism map | |
| CN113378475B (en) | Four-foot robot control method, system and device based on Vrep | |
| CN115019825B (en) | Acoustic channel digital twin method, device and system for underground tunnel environment | |
| Aslinezhad et al. | ANN-assisted robust GPS/INS information fusion to bridge GPS outage | |
| US20230278214A1 (en) | Robot localization using variance sampling | |
| CN105824292A (en) | Robot distributed controller and control method | |
| CA3078072A1 (en) | Maintaining a trained neural network for mobile device rss fingerprint based indoor navigation | |
| CN107504979A (en) | Move distance computational methods and device and wearable device | |
| CN114047697A (en) | Four-footed robot balance inverted pendulum control method based on deep reinforcement learning | |
| Bailly et al. | Estimating the center of mass and the angular momentum derivative for legged locomotion—a recursive approach | |
| CN109857127B (en) | Method and device for calculating training neural network model and aircraft attitude | |
| WO2022105023A1 (en) | Robot control method and apparatus, and computer readable storage medium, and robot | |
| CN116225029B (en) | Robot path planning method | |
| CN110007617A (en) | A kind of uncertainty transmission analysis method of hardware-in-the-loop system | |
| CN105818145A (en) | Distributed control system and method for humanoid robot | |
| WO2024021744A1 (en) | Method and apparatus for controlling legged robot, electronic device, computer-readable storage medium, computer program product and legged robot |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |