Optimization Of Biomimetic Robotic Fish Motion Control Based On CPG

Biomimetic robotic fish has great underwater application prospects,especially multi-mode motion control based on the central pattern generator(CPG)mechanism brings new opportunities and challenges to the research of biomimetic robotic fish.This paper takes biomimetic robotic fish as the research object.From the angle of CPG control and optimization,we built a biomimetic robotic fish monitoring system and a biomimetic robotic fish dynamic model.Then,we analyzed the CPG mathematical model.Based on the optimized parameters of the CPG mathematical model,we optimized the swimming speed of the biomimetic robotic fish and the propulsion efficiency.Inspired by biological movements,we studied stratified motion control based on Spiking neural network and CPG.The main tasks are as follows:First,the research background of the swimming control of biomimetic robotic fish based on CPG and the current research situation at home and abroad are expounded.On the basis of the previous research work in the laboratory,the experimental platform of the biomimetic robotic fish monitoring system is built.In this process,the design and development of biomimetic robotic fish,hardware development of monitoring system experimental platform,communication protocol and protocol formulation,and monitoring software development were carried out.Secondly,the parameters of the CPG model of the biomimetic robotic fish motion control are analyzed and optimized,and the problem of setting the parameters of the CPG model is solved.There are many parameters such as vibration frequency,vibration amplitude,excitation,down-coupling parameters,and up-coupling parameters in the CPG model of biomimetic robotic fish swimming control.The traditional test-and-conquer method has low efficiency and poor stability.For this reason,the particle swarm optimization(PSO)method is used in this paper.The parameters of the CPG model were optimized,and simulation and experimental platform tests were performed.The results obtained verified the feasibility of using the PSO algorithm to set the parameters of the CPG model.Thirdly,in order to improve the speed of swimming and the efficiency of propulsion,the problem of the average swimming speed of the biomimetic robotic fish and the optimization control of the propulsive efficiency are proposed.By using the additional mass method to analyze the force of the head and tail of the biomimetic machine fish,the hydrodynamic modeling of the biomimetic robotic fish is carried out.Then,the PSO optimization was performed on the important parameters such as the frequency and amplitude related to the propulsion speed and the propulsion efficiency in the dynamic model.Last,the average speed of the biomimetic robotic fish swimming was maximized and the propulsion efficiency was maximized.The efficiency of the strategy is maximized.The experimental results show that the optimization method is effective.Finally,inspired by biological movement,a hierarchical control method of biomimetic robot fish based on CPG and Spiking neural network is proposed.The Spiking neural network imitates the information of the human brain neuron to receive the external environment information and produces an exciting signal,which is input into the CPG model to produce rhythmic motion signals,thus controlling the motion of the various joints of the biomimetic robotic fish.In order to prove the feasibility of the method,the simulation was performed on MATLAB.Finally,the CPG output signals such as direct travel and turn control were obtained,and the hierarchical control of biomimetic robotic fish based on CPG and Spiking neural network was realized.