基于后随机非线性动态粒子群算法的地震模拟振动台参数优化设计

Parameter Optimization of Shaking Table Based on Later Random and Nonlinear Dynamic Particle Swarm Optimization

地震模拟振动台系统控制参数的质量影响着台面波形复现精度,进而影响试验效果。为实现对系统控制参数的高效校调与智能化寻优,提高系统对台面的控制效果,在标准粒子群算法的基础上提出了一种基于后随机非线性动态粒子群算法的三参量控制参数寻优方法,用于地震模拟振动台系统控制参数的优化整定。仿真结果表明,改进算法寻优参数控制下复现波形的相关系数比理论参数控制时提高了25%以上;与试验反馈数据波形仿真结果相比,寻优参数控制下的系统仿真波形相关度均提升至0.95以上,绝对峰值误差降至10%以内;对比结果证明了改进算法可行有效,优化参数控制下系统对不同输入波形具有更好的适应性,系统受控效果良好。

The quality of control parameters of shaking table affects the precision of table waveform reproduction and the test results.In order to realize the efficient adjustment and intelligent optimization of the system control parameters,and to improve the control effect of the system,a three-parameter control parameter optimization method based on Later Random and Nonlinear Dynamic Particle Swarm Optimization(LRNDPSO)is proposed to optimize the control parameters of shaking table system.The results show that the correlation coefficient of the regenerated waveform under the optimization parameter control of the improved algorithm is more than 25%higher than that under the theoretical parameter control.Compared with the simulation results of experimental feedback data,the correlation degree of simulation waveform under the control of optimization parameters increases to more than 0.95,and the absolute peak error decreases to less than 10%.The comparison results show that the algorithm is feasible and effective,the system has better adaptability to different input waveforms under the control of optimization parameters,the system is better controlled.