稀土永磁无刷直流电机健康预测管理

Rare earth permanent magnet (REPM) brushless DC motor is the key executive part of aircraft flight control, brake, fuel and other systems. Prediction of the remaining life of the motor for prompt actions against its potential failure is of great significance to the security of the whole aircraft system. The goal of this research is to establish a prognostics and health management (PHM) system for the REPM brushless DC motor, which mainly focuses on the feature extraction method based on the integration of deep learning and traditional signal analysis, the remaining life prediction based on artificial intelligence and statistical process models, and the modeling and optimization of the real-time joint maintenance. It is expected that the system can dig out the complete features which represent the performance degradation state of the REPM motor, reveal the correlation between the motor degradation and the multidimensional features through unsupervised clustering and measurement learning extraction method, realize the accurate remaining life prediction under the stochastic conditions based on Wiener process modeling method, and finally achieve the real-time health management of the REPM brushless DC motor by constructing the real-time joint decision model and minimizing the multi-objective cost functions via evolutionary algorithms.

稀土永磁无刷直流电机是飞机飞控、刹车、燃油等系统中的关键执行部件，预测其剩余寿命，在电机失效前采取有效管理措施，对飞机总体系统的安全保障具有重要意义。本项目的研究目标是建立一套稀土永磁无刷直流电机健康预测管理系统，其核心包括深度学习与传统信号分析相结合的特征提取、人工智能方法与统计过程模型相结合的剩余寿命预测和实时联合维修决策优化建模。预期该系统可利用深度学习网络和传统信号分析方法，从多源监测信号中挖掘出表征电机性能退化的完备特征变量；通过无监督聚类和度量学习相结合的设备健康指标提取方法，揭示电机剩余寿命随多维特征变量的变化规律；基于维纳过程的建模方法，实现不确定条件下的电机剩余寿命的准确预测；通过构建实时联合决策模型，设计多目标优化方法以及进化算法求解最优维修决策变量，最终实现对稀土永磁无刷直流电机的实时健康管理。

稀土永磁无刷直流电机是飞机飞控、刹车、燃油等系统中的关键执行部件，预测其剩余寿.命，在电机失效前采取有效管理措施，对飞机总体系统的安全保障具有重要意义。本项目研究.目标是建立一套稀土永磁无刷直流电机健康预测管理系统，其核心包括深度学习与传统信号分.析相结合的特征提取、人工智能方法与统计过程模型相结合的剩余寿命预测和实时联合维修决.策建模。预期该系统可利用深度学习网络和传统信号分析方法，从多源监测信号中挖掘出表征.电机性能退化的完备特征变量；通过无监督聚类和度量学习相结合的设备健康指标提取方法，.揭示电机剩余寿命随多维特征变量的变化规律；基于维纳过程的建模方法，实现不确定条件下.的电机剩余寿命的准确预测；通过构建实时联合决策模型，设计多目标优化方法以及进化算法.求解最优维修决策变量，最终实现对稀土永磁无刷直流电机的实时健康管理。