基于最大熵和仿真的导弹装备备件需求预测

Demand Forecasting of Missile Equipment Spare Parts Based on and Maximum Entropy Principle and Simulation

针对小样本条件下确定导弹装备故障数据统计分布的困难,基于最大熵原理与概率加权矩建立了统计分布函数估计的新方法及其实用的数值计算方法。将概率加权矩引入导弹装备寿命分布建模的新方法,克服了基于经典矩在小样本量下统计估计的大偏差问题。其次,针对导弹装备在工作、停机期间及备件在储备期间均可能发生故障或失效的情况,建立了基于保障度的间断工作导弹装备备件需求量预测模型。最后,在建模基础上,分析了仿真流程,规划了仿真步骤。示例仿真结果与解析方法计算的结果吻合较好,表明了仿真模型的适用性和有效性。

Aiming at the difficulty of determining statistic distribution of missile equipment failure date,a new method used to estimate statistic distribution function and its practical date calculation method were presented based on maximum entropy principle and the probability weighted moments.In order to overcome the highly biased problems of statistic and estimation from small sample sizes based on statistical ordinary moment,probability weighted moments was introduced to missile equipment life distribution modeling.Then considering the fault and failure of the missile equipment during working and shutting down and of the spare parts during storage,the demand simulation model of the missile equipment spare parts was established based on the guarantee degree in intermittent work.Finally,based on modeling,simulation flow and programming simulation process were analyzed.The sample simulation result agrees well with the analytic method calculation,which indicates the applicability and effectiveness of the simulation model.