论文部分内容阅读
针对limit-checking故障检测方法的缺点,提出了一种基于模糊集的故障检测方法。该方法基于传感器测量值的不确定性和门限值具有模糊的客观属性,在检测值所有可能值的论域上划分模糊子集,将检测值偏离门限值的程度用隶属函数与模糊子集报警权重表示。考虑到时间因素对检测结果的影响,提出采用时间窗函数的方法计算时间窗内的报警权值之和作为模糊检测决策依据。仿真结果表明,基于模糊集的故障检测方法相比于持续性计数(Persistence-Counter)方法具有更好的检测性能。
Aiming at the shortcoming of limit-checking fault detection method, a fault detection method based on fuzzy set is proposed. The method is based on the uncertainty of the measured value of the sensor and the objective value of the fuzzy threshold. The fuzzy subset is divided on the universe of possible values of the detected value. The deviation of the detected value from the threshold value is determined by membership function and fuzzy sub Set alarm weight said. Considering the influence of time factor on the test results, a method of time window function is proposed to calculate the sum of the alarm weights in the time window as the basis of fuzzy detection decision. The simulation results show that the fault detection method based on fuzzy sets has better detection performance than the Persistence-Counter method.