摘要
指出了传统预测方法用于火炮故障预测的不足,提出了基于动态模糊综合评判和“框架+规则”的正反向混合推理故障预测模型。针对火炮故障发生和发展过程的动态变化特性,提出了动态隶属度的概念,并结合变权理论,详细阐述了动态模糊综合评判方法,用于生成候选故障集。基于“框架+规则”的反向推理确认实际发生的故障,输出故障预测报告,为火炮在执行训练和作战任务前实现预知维修提供决策依据。
The deficiencies of traditional forecast methods for gun fault are pointed out. A new fault forecast model based on hybrid inference of dynamic fuzzy comprehensive evaluation and “frame＋rules” is put forward. Considering the dynamic characteristics of fault occurrence and development process, the concept of dynamic membership is presented. Combined with the variable weight theory, the dynamic fuzzy comprehensive evaluation method to produce candidate fault set is discussed in detail. The factual faults can be confirmed by the backward inference process based on frames and rules and the fault forecast report be given, which provides valuable foundations of Decision-making for predictive maintenance before training and battle tasks implementation.
出处
《火炮发射与控制学报》
北大核心
2005年第3期47-52,共6页
Journal of Gun Launch & Control
关键词
数理统计学
故障预测
混合推理
动态隶属度
mathematical statistics
fault forecast
hybrid inference
dynamic membership
