论文部分内容阅读
本文提出一种聂曼 皮尔逊准则下分布式多传感器检测系统优化设计的混合优化算法 ,该算法以进化策略方法进行全局搜索 ,辅以Powell方向加速法进行局部搜索 ,无须使用目标函数的导数信息 .对分布式恒虚警检测系统的优化设计结果表明 ,混合优化算法在收敛性和精度上都优于梯度算法 .
In this paper, a hybrid optimization algorithm based on Niemann-Pearson criterion for optimal design of distributed multisensor detection system is proposed. This algorithm uses global search by evolution strategy and complements local search by Powell direction-accelerating method without using derivative information of objective function The results of optimization design of the distributed CFAR detection system show that the hybrid optimization algorithm is superior to the gradient algorithm in convergence and accuracy.