Font Size: a A A

Optimization Optimal Iterative Learning Control Research Based On Niche Shuffled Frog Leaping Algorithm

Posted on:2018-12-15Degree:MasterType:Thesis
Country:ChinaCandidate:D J WangFull Text:PDF
GTID:2348330536980379Subject:Internet of Things works
Abstract/Summary:PDF Full Text Request
Iterative learning control is a branch of control field.It studies a kind of control system which has the repetitive motion properties.By correcting constantly the control input of the system,the actual output is approaching gradually the desired output,and can realize finally the complete tracing for the desired trajectory.Iterative learning control has become an important control method in modern control technology.During recent years,scholars majoring in learning control system carried on the researchinsightfully,including the improvement and the optimization of the algorithm,such as Genetic Algorithm,clone selection algorithm and shuffled leaping frog algorithm.Although these algorithms in solving the general nonlinear system can guarantee the system Convergence,butwhen there is a disturbance to the nonlinear system,the convergence speed and optimization precision are not able to achieve our prediction,in addition,the algorithms need iteration times to completely track the system.Consequently,the traditional algorithms can improve the efficiency in learning system.leading to the low learning efficiency.To solve all the above problems,this dissertation put forward an optimized iterative learning control algorithm which based on the niche shuffled leaping frog algorithm.It has a strong ability and fast convergence which is based on the restrict competitive strategy.The initial population is divided into exclusive mutually and independent subpopulations.In this algorithm,the independent search space can be formed between subpopulations to suppress the convergence among populations and maintain the diversity of understanding.Meantime it can accelerate the elimination of poor fitness of the individual.To optimize PID controller parameters,this algorithm introduces the PID iterative learning controller into the parameters optimization iterative learning control.Because the PID type iterative learning controller can increase the dimension of the solution in the algorithm and the degree of freedom of the parameters,and the parameters of the algorithm are simple and the optimization effect is better.By establishing the parameter optimization iterative learning control,it can optimize the parameters of the system for each iterations.The parameter optimization can reduce the complexity of the algorithms.This dissertation analyzes and proves the convergence of the iterative learning control algorithm based on PID parameter optimization.The niche shuffled leaping frog algorithm is compared with the genetic algorithm and the clonal selection algorithm in the experimental simulation and showing the niche shuffled leaping frog algorithm is faster and the stability is better.Finally,the niche shuffled leaping frog algorithm is applied to the vibration control system,and the simulation results show that the algorithm not only has high search efficiency but also has strong search ability,verifying the feasibility and effectiveness of the algorithm.
Keywords/Search Tags:Learning Control, Iteration, PID, Parameters Optimization, Convergence Speed, Niche Shuffled Frog Leaping Algorithm
PDF Full Text Request
Related items