基于GA和ARMA模型的制导弹药技术状态预测

针对定期检测判断制导弹药技术状态影响弹药寿命和实时性较差的问题,提出了基于遗传算法(GA)优化的自回归与移动平均(ARMA)模型制导弹药技术状态预测方法。在获取到制导弹药状态特征量历史检测数据的基础上,采用遗传算法对ARMA模型的阶数进行优化,通过实例将遗传算法优化后的ARMA模型与经典ARMA模型的预测结果进行比较分析,结果表明,经过优化后的ARMA模型具有更好的预测效果。该方法提高了制导弹药技术状态确定的实时性,对降低、消除制导弹药故障危害具有实际意义。

Technical Conditions Prognostics for Guided Munition Based on Genetic Algorithm and ARMA Model

Aiming at the problem of periodic inspections which would affect the life of the ammunition when judging thetechnical conditions of the guided munition and have poor real-time performance, prediction method of guided munitiontechnical conditions based on autoregressive and moving average(ARMA)model optimized by genetic algorithm (GA) wasproposed. Based on the historical data of the state characteristic quantity of the guided munition, the genetic algorithm wasused to optimize the order number of the ARMA model. The prediction results of ARMA model optimized by genetic algo?rithm and classic ARMA model were comparatively analyzed by an example. The results show that the optimized ARMAmodel has a better prediction effect. The method improves the real-time determination of the technical conditions of theguided munition, and has practical significance for reducing and eliminating the damage of the munition.