基于免疫算法的微硬盘双级伺服系统最优控制研究  

Study on optimal control of microdisk dual-stage servo systems based on immune algorithm

作　　者：涂歆[1] 张著洪[1] 徐志丹[1] 

机构地区：[1]贵州大学系统科学及信息技术研究所,贵州贵阳550025

出　　处：《哈尔滨工程大学学报》2006年第B07期233-236,共4页Journal of Harbin Engineering University

基　　金：国家自然科学基金资助项目（60565002）;贵州省优秀人才省长基金资助项目（20040706）;贵州省科技基金资助项目（20052108）.

摘　　要：微硬盘伺服系统的双级控制一直是微硬盘控制技术研究的热点．在此将寻求双级控制器的问题转化为高维最优化问题，并基于变量分解和免疫网络优化思想，设计改进的并行免疫网络优化算法寻找有限步的最优控制量．算法设计中，引入并行合作进化、抗体分级进化、抗体群通讯的思想提高网络进化能力和执行效率．数值实验、比较表明，系统在所获最优控制量的作用下，磁头输出能快速跟踪参考信号、磁头定位精度高，噪声衰减能力强．Optimal control of microdisk servo,dual-stage control systems is an important and meaningful topic of microdisk control techniques. In this work, an improved parellel network optimal algorithm is suggested to search for the discrete values of the optimal controller, in which the optimal control problem is transformed into a high dimension optimization problem through variable partition and the idea of immune network optimization. To strengthen the evolutionary ability and speed up the search rate of the algorithm, the ideas of parellel coevolution, classification of antbodies, and communication between antibody populations are introduced. The experiment shows that, when employing the controller, the output of magnetic head can track the reference signal with high speed, orientational precision, and strong noise reduction.

关 键 词：微硬盘伺服系统 最优控制 免疫算法 并行合作进化 变量分解 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]