动基座平台式制导工具误差分离的粒子群算法

Particle swarm optimization algorithm for separating the systematic errors of moving-base platform guidance instruments

针对动基座发射飞行器的制导工具误差分离是一个不适定问题,采用经典参数的粒子群算法(PSO)分离制导工具误差和初始误差。目标函数考虑4个因素构建:①逼近遥外差的残差;②地面测试值先验信息;③初始误差的直接或者间接测量信息;④估计结果符合天地一致性的个数。这四个因素充分利用了误差分离前可以利用的有效信息,并利用有效信息提取先验信息。通过构建目标函数,利用先验信息把误差项分为严格有界项和允许越界项。算例表明,所提出的算法优于其他算法,在逼近遥外差的残差足够小的同时,估计结果符合天地一致性的项数为26项,比主成分贝叶斯估计结果多8项。

In view that the separation of the guidance instrumentation systematic errors(GISE) for moving-base spacecraft is an ill-posed problem, an off-the-shelf particle swarm optimization(PSO) is adopted to separate the GISE and the initial error. The objective function is constructed by considering the four factors:(1) the residual approximating the telemetry-tracking deviation;(2) the prior information in ground tests;(3) the direct or indirect measurement information of the initial errors;(4) the number of estimated GISE results consistent with ground tests. The four factors make full use of the available information before the error separation, and extract the prior information by using the effective information. By constructing the objective function and using the prior information, the error terms are divided into strictly bounded terms and permissible transboundary terms for the error separation. The experimental results of error separation show that the proposed algorithm is superior to other methods. In particular, the factor(1) is small enough, and the factor(4) is 26, which is 8 more than the number by the principal Bayes estimation.