Optimal Design Of Intelligent Controller For Shaking Table Based On Hybrid Intelligent Algorithm

Electromagnetic seismic simulation shaker represents an essential experimental equipment for vibration pickup calibration and model structure earthquake resistance.The waveform reproduction accuracy of the shaker is highly relevant to the vibration pickup calibration and model structure seismic test results.In recent years,iterative learning control and fuzzy control technology have been widely applied in this field,yet some problems must be solved urgently.This paper takes the parameter optimization algorithm of the small electromagnetic seismic simulation shaking table controller as the research object,improves the algorithm convergence accuracy,the convergence speed and the optimal solution ability to escape from the local by upgrading intelligent optimization algorithm,realizing the adjusting of the control system parameters and enhancing the accuracy of the seismic waveform reproduction by the shaking table of electromagnetic.(1)For the insufficiency of the traditional vibration table control system parameter tuning effect that is affected by human factors,particle swarm optimization algorithm is adopted to achieve automatic optimization and tuning of controller parameters.However,the algorithm is deficient in three aspects: convergence speed,convergence accuracy,and escape from the local optimal solution.Hence,this paper presents a hybrid optimization algorithm based on particle swarm and gravity search algorithm and the convergence accuracy is improved.Simplify the velocity iteration formula in the particle swarm iterative formula,change the iterative formula into a first-order equation,decrease the amount of calculation,elevate the convergence speed,choose the best in uniformly distributed Zaslavskii chaotic map and shrinkage factor in the 0-1 interval,increasing the population of the algorithm diversity and helping the algorithm to escape from the local optimal solution faster.Eight different test functions with multiple local extreme points,grand search space,multiple peaks,and difficulty in convergence are selected to verify the effectiveness of the algorithm.The result indicates that after the expression of the velocity is simplified,the time-consuming process of the particle swarm iteration process is significantly less than the standard particle swarm.When chaos mapping is added to the particle swarm algorithm,the descending speed is fast,and the staying time at the local optimal solution is short.The Zaslavskii chaotic map is more convergent than the Logicstic chaotic map.Compared with eight commonly used intelligent optimization algorithms,in multiple measurements,the hybrid algorithm in this paper shows good performance in four aspects: the best result,the worst result,the average result and the running time.(2)To improve the accuracy of the waveform reproduction of the seismic simulation shaking table,the open loop model,velocity feedback model and displacement feedback model of the electromagnetic seismic simulation shaking table have been constructed.Based on the measured frequency response data of the vibration table,the mean square error is adopted as the objective function in optimizing the identification model data,which improves the identification accuracy and reduces the difficulty of the algorithm.The results indicates that the hybrid intelligent optimization algorithm is applied to model parameter identification and is superior in high identification accuracy,fast speed,and simple application.(3)For the problem that traditional iterative learning algorithm has the problems of slow convergence speed and quantities of iterations in the vibration table control.Add the feedforward controller before the iteration to broaden the system frequency band.The learning factor is introduced into the iterative learning law to improve the convergence accuracy and stability of the iterative learning law.Introduce the derivative of the previous error and the current error signal to the feedback control law as correction items to increase the convergence speed and solve the non-causal problem caused by the derivation of the current error signal.Reduce the number of iterations,and quickly improve the accuracy of the vibration table waveform reproduction.Calculate the iterative learning law parameters offline with the hybrid optimization algorithm to reduce the number of repeated iterations of the vibration table in actual control.Compared with the traditional iterative learning algorithm,the results indicates that after three iterations,the feedback-assisted PD-type iterative learning algorithm with forgetting factor is superior to the traditional iterative learning algorithm in number of iterations and convergence accuracy.(4)To improve the real-time output waveform reproduction accuracy of the shaking table,the control rules of the electromagnetic seismic simulation shaking table fuzzy controller are constructed,and combined with the characteristics of high sensitivity of the triangular membership function and good stability of the Gaussian membership function,the membership function of the combination of the two is constructed.The parameter adjusting in the fuzzy controller is greatly influenced by human factors.Hence,the intelligent optimization algorithm is applied to optimize the parameters of the fuzzy PID controller.The results indicates that the fuzzy PID controller optimized by the intelligent optimization algorithm is improved in the tracking effect and distortion of the waveform under the effects of the sine wave signal and the seismic wave signal.