基于模糊状态观测器的燃气轮机转速传感器故障检测研究

Study on Speed Sensor Fault Diagnosis for gas turbine Based on Fuzzy State Observer

为了更好的跟踪状态量可测的燃气轮机转速曲线,将模型参考自适应与模糊控制相结合,构造了燃气轮机转速模糊状态观测器,不断逼近系统实时转速输出,解决了传统状态观测器在模型不精确时易产生的误差累计和观测偏移的问题,并通过估计相对差和自适应阀值的比较实现传感器故障检测。利用Lyapunov稳定性理论证明了观测器跟踪误差在有限时间区间内一致收敛到零。仿真结果验证了算法的可行性和有效性。

In order to better track the measurable speed curves of gas turbine,a fuzzy observer is designed by combining the model reference adaptive control with fuzzy theory,which can constantly approach to the real speed.Comparing to traditional model observer,the method overcomes the difficulty of computing inaccuracy that arise from error accumulation and observer offset.The method of fault detection by comparing the relatively estimating error and adaptive threshold is advanced.The stability of observer is proved by Lyapunov theory,and simulation results shows that the proposed method is feasible and effective.