Research On Control Method Of Cricket System Based On Hierarchical Online Optimization

Ball and plate system is an experimental platform for teaching and research development.It is a teaching platform for verification and analysis for classical control theory,modern control theory,motion control theory,computer image processing,machine vision and other courses.Compared with the two-dimensional club control system,the control process of the ball and plate system is more complicated,and the characteristics of nonlinearity and instability are more prominent.The ease of use and typicality of the ball and plate system have attracted extensive attention from experts and scholars.In addition,the ball and plate system is also involved in the nonlinear field,such as system analysis and modeling,mechanical structure design,path planning and other fields.The combination of easy integration makes the ball and plate system have a certain degree in engineering applications and theoretical research significance.When controlling the ball and plate system,the factors affecting the control performance of ball and plate system are as follows: First,the external environment such as illumination and jitter has a great influence on the system control accuracy when the system is running;Second,the ball is on the experimental platform.The scrolling speed,when the experimental target requires the curvature of the small ball running path to be large or the ball rolling speed is required to be fast,puts higher requirements on system stability and control precision.The main work has been done on the above issues:(1)Aiming at the problem that the single control precision of the system parameters is too low,a strategy combining the Shuffled Frog Leaping Algorithm(SFLA)and PID is proposed.The SFLA is used to tune and optimize the parameters of the controller in the simulation experiment.However,this process is mostly an off-line optimization process.The off-line optimization and singleness constraints of parameters are widely used.In this paper,the in-depth analysis and research on the motion control of the ball and plate system is carried out,and the self-learning ability of the frog leaping algorithm is combined with the traditional PID controller.A number of optimizations and improvements are made to the local update strategy of traditional leapfrog algorithm.The aggregation factor and evolution factor are proposed.The subgroup size of the leapfrog algorithm is dynamically adjusted duringthe optimization process,and the online self-learning and online automatic tuning and correction of control parameters are realized.The SFLA-PID self-tuning control algorithm improves the control accuracy of the experimental platform.(2)Design a hierarchical online optimization controller for the problem that the ball movement rate is too fast and the curvature is large.The bottom controller adopts the SFLA-PID self-tuning control algorithm to complete the system parameter self-learning and online self-correction function,and improve the control precision of the experimental platform.The upper controller adopts the auxiliary function controller to supervise the running speed of the small ball.On the GPB2001 ball and plate system platform of the solid high-tech company,the stratified controller is used to carry out the fixed point control test and the tracking trajectory control test respectively.The test results show that the stratified online optimization control method has feasibility,high precision and good robustness.