阻尼器测试试验台的新变步长LMS自适应滤波器设计  

Design of a New Variable Step Size LMS Adaptive Filter for Damper Test Bench

作　　者：丁兆轩 刘凯磊 强红宾 康绍鹏 Ding Zhaoxuan;Liu Kailei;Qiang Hongbin;Kang Shaopeng(School of Mechanical Engineering,Jiangsu University of Technology,Jiangsu Changzhou 213001;Sinomach Changlin Co.,Ltd.,Jiangsu Changzhou 213136;Research Center of Fluid Machinery Engineering and Technology,Jiangsu University,Jiangsu Zhenjiang 212013)

机构地区：[1]江苏理工学院机械工程学院,江苏常州213001 [2]国机重工集团常林有限公司,江苏常州213136 [3]江苏大学流体机械工程技术研究中心,江苏镇江212013

出　　处：《南方农机》2023年第21期31-34,共4页

基　　金：江苏省高等学校自然科学基金项目(22KJB460021);常州市科技支撑计划(社会发展)(CE20209002);常州市领军型创新人才引进培育项目资助(CQ20210093)。

摘　　要：【目的】提升阻尼器测试试验台的控制品质,进而得到更好的被测阻尼器参数曲线,需要对试验台反馈数据进行滤波提取。【方法】课题组通过对传统LMS自适应滤波器的研究,针对其步长选值问题,提出了一种基于麻雀搜索算法(Sparrow Search Algorithm,SSA)的变步长LMS算法,进而进行了新变步长LMS自适应滤波器设计。并通过Tent混沌映射优化了SSA算法在LMS自适应滤波器中的使用,加强了该算法的跟踪和收敛能力。【结果】仿真结果表明,优化后的算法较优化前误差总值下降约75.72%,优化效果较为显著。同时,相较于其他文献中的LMS自适应滤波算法,本设计中的算法能在快速收敛的同时保持较低的稳态误差,误差总值下降约35.17%。【结论】经过Tent混沌映射所得出的SSA算法初始种群能够提高LMS算法的跟踪能力、收敛速度和稳态精度。课题组设计的滤波器表现良好,具有较强的跟踪能力和稳态精度在低范围的保持能力,为阻尼器测试试验台性能优化和反馈数据的滤波提取提供了新的思路和方案。[Objective]Improve the control quality of the damper test bench and obtain better parameter curves of the tested damper,it is necessary to filter the feedback data of the test bench.[Method]Through research on the traditional LMS adaptive filter,the research team proposed a variable step size LMS algorithm based on Sparrow Search Algorithm(SSA)to address the problem of step size selection,and further designed a new variable step size LMS adaptive filter.And the use of SSA algorithm in LMS adaptive filters was optimized by Tent chaotic mapping,enhancing the tracking and convergence capabilities of the algorithm.[Result]The simulation results show that the optimized algorithm reduces the total error by about 75.72%compared to the pre optimized algorithm,and the optimization effect is relatively significant.Meanwhile,compared to the LMS adaptive filtering algorithm in other literature,the algorithm in this design can maintain a low steady-state error while rapidly converging,with a total error reduction of approximately 35.17%.[Conclusion]The initial population of SSA algorithm obtained by Tent chaotic mapping can improve the tracking ability,convergence speed,and steady-state accuracy of LMS algorithm.The filter designed by the research group performs well,with strong tracking ability and ability to maintain steady-state accuracy in a low range,providing new ideas and solutions for optimizing the performance of the damper test bench and filtering and extracting feedback data.

关 键 词：阻尼器测试试验台 LMS自适应滤波器 麻雀搜索算法 Tent混沌映射 

分 类 号：TN713[电子电信—电路与系统]