Research On The Design And Control Strategy Of A Rope-driven Snake-like Robot

With the continuous improvement of equipment integration in the fields of aerospace,aviation and nuclear power plants,the internal workspace is getting narrower and smaller,the environment is more and more complex,and the requirements of internal operation are higher and higher.Because of its strong flexibility and flexibility,the serpentine robot can better complete the maintenance and other tasks in the narrow space.Considering that the rope driving mode can reduce the volume and weight of the robot,a design scheme of rope driven serpentine robot for working in narrow space is proposed in this paper,and its control strategy is studied.The main work of this paper is as follows:(1)A design scheme of rope-driven serpentine robot is presented in this paper.For the hardware structure of the serpentine robot,such as joint,connecting rod and drive box,the detailed hardware structure and component model diagram are given.According to the structure of the serpentine robot designed in this paper,the mathematical model of kinematics and dynamics of the serpentine robot is established,which lays a foundation for the research of robot control strategy.(2)In order to solve the problems of poor stability and slow response of traditional PID control strategy of snake robot,a BP neural network PID control strategy based on fireworks algorithm optimization is proposed.First of all,the self-learning and adaptive ability of BP neural network is used to improve the response speed and stability of PID control strategy.Secondly,in order to prevent the BP neural network from falling into the local optimal value,the fireworks algorithm is used to optimize the initial weight of the BP neural network in order to accelerate the convergence speed of the BP neural network and prevent the local optimal value.Finally,the proposed control strategy is simulated based on MATLAB,and the simulation results show that the control strategy has better stability and position control accuracy.(3)In order to solve the problem that it is difficult to model a serpentine robot in some complex environments,a data-driven control strategy based on depth deterministic strategy gradient is proposed in this paper.The control strategy uses the learning and decision-making ability of the deep deterministic strategy gradient to train the off-line and on-line input and output data of the serpentine robot control system to design the robot controller.The simulation of the control strategy is based on the data of MATLAB reinforcement learning toolbox and serpentine robot control system.The simulation results show that the control strategy has anti-interference,stability,convergence and high position control accuracy.(4)The prototype is made.According to the design scheme of the serpentine robot proposed in this paper,the hardware components are made by using 3D printing technology and the driving equipment is selected,and the prototype of the serpentine robot is assembled.In this paper,through the experiment of the prototype,the feasibility of the design scheme of the serpentine robot and the effectiveness of the control strategy are verified.