解空间定向优化的信号分选算法改进策略  

Improvement Strategy of Signal Sorting Based on Solution Space Orientation Optimization

作　　者：同晓荣 TONG Xiaorong(School of Network security and Informatization,Weinan Normal University,Weinan 714099,China)

机构地区：[1]渭南师范学院计算机学院

出　　处：《探测与控制学报》2019年第4期106-111,117,共7页Journal of Detection & Control

基　　金：教育部产学合作协同育人项目资助(201702030020);陕西省教育厅科研计划项目资助(18JK0286);渭南市科研发展计划项目资助(2019JCYJ-2-6)

摘　　要：为了减少声音信号分选算法应用过程中的计算量与耗时,提出解空间定向优化的改进策略,应用于信号分选以提升运算效率。类比于基于药物辅助的改进型免疫算法思想,对信号分选过程中的时频单源区域检测环节实施定向优化。将声信号分选的目标函数视为在时频二维平面寻找最大值,当算法寻找到某一局部极大值后,缩小原始的搜索区域,使得时频平面上的众多无效区域排除在搜索范围外,有效提高算法的收敛速度,降低了计算量。仿真实验结果表明,在正定与超定的情况下,该策略能够减少50%的系统响应时间;欠定情况下,该策略能够减少约30%的系统响应时间。In order to reduce the calculation amount and time consuming of acoustic signal sorting algorithm,the principle of signal sorting algorithm was analysed in this paper,an improvement strategy based on solution space orientation optimization was proposed,and then,its application in signal sorting was studied for improving operational efficiency.According to the auxiliary medicine based immune algorithm,some orienteering optimizations were used in single source point detection link in the process of signal sorting.The target function of acoustic signal sorting was regarded as maximum search in time-frequency two-dimensional plane,when local maximum was found out,the original search area was reduced for excluding a large number of invalid area,so that,the proposed strategy effectively improved the convergence speed,and reduces the amount of computation.This paper focuses on the analysis of the working principle of improvement strategy,improved ideology of the algorithm and key steps.At last,the simulation experiments proved the feasibility and superiority of the strategy.The simulation results showed that the strategy could reduce the system response time by 50%under positive and overdetermined conditions,and about 30%under undetermined conditions.

关 键 词：定向优化 收敛性 搜索区域 信号分选 时频极值 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]