根据免疫学研究中抗原与抗体同时进化的特性,通过对动目标预测系统特点的分析,提出一种改进的免疫算法.算法以目标当前运动时刻的拟合信息为初始抗原,下一时刻信息的改变为抗原进化,进行参数优化搜索.并进一步使用所得最佳参数对预测模型进行拟合得出结果.仿真结果表明,该算法可以满足动目标预测系统的高实时性要求.
Based on the property that antigen and antibody are evolving simultaneously in immunology, an improved immune algorithm is presented through analysis of moving object forecast system. The algorithm employs the current information as the initial antigen and next one as the antigen evolution in the optimization search algorithm. The algorithm satisfies the high real time demands of moving object forecast system, reducing effectively the system' s estimation error in highly non-liner system.