Maintenance Robots In Unstructured And Complex Environments Control Strategy And Stability Studies

In recent years,intelligent machinery has been favored,and the mobile robot industry is also developing rapidly.In order to improve maintenance efficiency and safe operation management,maintenance machinery is also in urgent need of intelligent transformation.With the development of science and technology and robot industry,mobile robots began to move from structured environments such as factories and laboratories to unstructured environments such as tunnels,mining areas,and water,and entered the team of road surface maintenance and maintenance machinery,and their task requirements gradually developed from welding,handling,installation and other process-oriented work to irregular and changing task scenarios such as inspection,maintenance,and replacement.When performing tasks in an unstructured and complex environment,the maintenance robot system has inherent nonlinear characteristics and the model is inaccurate or unknown due to the disturbance of the external environment,and cannot be tracked and controlled under the accurate model.In addition,the response speed and continuous stability of the manipulator work will be affected when the maintenance robot is physically constrained and interfered with by the outside world,so this paper comprehensively uses the intelligent control method and the environmental information fusion control method,based on the tunnel lighting maintenance robot prototype developed by this topic,and conducts in-depth research on the control system of the manipulator arm and its end in terms of external interference,shaking at the end of the manipulator,response speed,etc.,and the main work and conclusions are as follows:Firstly,aiming at the uncertainty problem of the system,the dynamic modeling is carried out based on the PIEPER criterion and the Lagrange equation,and a strategy based on high-order sliding mode variable structure control(HO-SMC)is proposed.Secondly,considering the problem of system finite time tracking,on the basis of high-order sliding mode variable structure feedback control,the homogeneous continuous control item is adopted,and the adaptive compensation item is added to ensure the stability of the system in a limited time.Furthermore,in order to reduce the jitter at the end of the robotic arm,the superspiral adaptive law(STA)is designed to increase the robustness of the system.Finally,in order to reduce the approximation error of the model,the radial basis function neural network is used for approximation,and the design of the high-order sliding mode variable structure adaptive neural network composite controller(HO-SMC-NNA)is completed.In order to further improve the finite time convergence speed and end tracking accuracy of the three-joint rigid robotic arm under the unstructured complex environmental environment,the improved particle swarm algorithm(IPSO)based on dynamically updated parameters is studied on the basis of sliding mode neural network control,and the neural network(NNA)control is optimized.When optimizing the neural network by particle swarm algorithm,in order to improve the global search ability of particles,chaos theory is introduced,and the inertia weight and acceleration constant in the particle position are improved by evaluating the particle moderate value in the entropy value of the particle discreteness,so as to realize the update of the weight of the middle layer of the radial basis function neural network,so as to avoid the "precocious" phenomenon that the traditional particle swarm algorithm is easy to jump out of the local optimal solution,and improve the approximation accuracy of the radial basis function neural network for modeling non-deterministic information in the system.Through simulation experiments,it is proved that the high-order sliding mode composite control strategy(IPSO-NNA-HO-SMC)based on improved particle swarm optimization neural network has improved tracking accuracy and convergence speed.Considering that the actual manipulator drive should have the flexibility characteristics of smoother motion and higher error tolerance,the adaptive control algorithm CSMC-STA coupled with the graded sliding mode variable structure control and superspiral algorithm is obtained by using the adaptive law combining the graded sliding mode variable structure control(CSMC)and the gain of the superspiral algorithm,so that the system can be stable in a limited time while continuously controlling,and the sliding mode jitter is weakened,which improves the smoothness of the motion.Finally,the Matlab/Simulink simulation proves the scientificity and effectiveness of the method.Then,aiming at the problems in the information fusion and feedback of the end effector in the complex environment,the consistency algorithm based on the distributed network is studied to model the system sensing information and the multi-target tracking problem under the finite set,and the adaptive thrust reverse algorithm controller of the end device is designed to improve the reliability of the end device,and the effectiveness of the control method is proved by Matlab/Simulink simulation.Through research,this project designed and manufactured a prototype tunnel lighting maintenance robot,and used it as an experimental platform,and the composite controller studied in this paper was used to verify the end position accuracy and maintenance time of the robot manipulator,and the results proved the correctness of the theoretical research.By installing and maintaining tunnel lighting fixtures with robots,the effectiveness of the control strategy designed in this paper is verified.