用于复杂机械系统状态监测的综合决策模型研究

针对复杂机械系统中体现机器运行的状态信息所具有的的多元化、强耦合性等特点，提出了一种对多源信息进行优化综合、分类决策的监控模型，通过建立多元信息矩阵并向关联矩阵、特征矩阵以及故障状态矩阵的依次转化，最终通过建立集成神经网络实现了对系统运行状态的有效监测。仿真结果证明了决策模型的可靠性。

Research on Synthetically Decision Model Used in Complex Mechanical System Monitoring

According to the particulars of state information of complex mechanical system such as multiples and coupling, a state-monitoring model is proposed, which can realize the optimal synthesis and classify decision of multiple information. In this model, the multiple information matrixes can be transformed into feature matrix and diagnosis state matrix in sequence, and sent into integrated neural network,and finally to realize the monitoring of the mechanical system. The reliability of the decision model is demonstrated by simulation results.