存在不确定性系统的半定量故障预报方法

给出了半定量故障预报的问题描述,在推导得到一步和多步模糊卡尔曼预报器以及给出故障预报准则的基础上,提出了一种存在不确定性系统的半定量故障预报方法.该方法能够对一类存在不确定性系统的缓变故障进行实时预报,克服了传统方法需要已知系统精确解析模型和噪声统计特性的不足.另外,由于将噪声描述成模糊定性区间的形式,因此该方法具有一定的鲁棒性.文章通过仿真实验验证了该方法的有效性,并给出了未来可能的研究方向.

Research on a Semi-Quantitative Fault Prediction Method of Systems with Uncertainty

The problem formulation of semi-quantitative fault prediction is given firstly in this paper. Based on one step and multi-step Fuzzy Kalman predictors and the rule of fault prediction, a semi-quantitative fault prediction method is proposed to predict the soft faults of a class of systems with uncertainty. The accurate analytic models and the statistical characteristics of noises, which are necessary in the traditional methods, are not used in this method. In addition, the noises are depicted as the fuzzy qualitative interval, so this method has some robustness. Finally, the simulation results demonstrate the effectiveness of the proposed approach, and the possible research direction in the future is given.