关键词：传感器；传感器故障检测和隔离；统计变化检测/隔离算法
摘 要：This chapter addresses the early detection and isolation of sensor faults in a systematic and unified way and illustrates the approach on wind turbine simulation data. Three problems are successively considered: individual signal monitoring, fault detection and isolation (FDI) in redundant sensors, and FDI based on analytical redundancy. In all three cases, a specific approach to generate fault indicators, also called residuals, is presented and combined with an online statistical change detection/isolation algorithm. The considered case studies consist of wind turbine generator speed monitoring, as well as FDI in the stator current and voltages of a wind-driven doubly fed induction generator. For the latter problem, the fact that the three-phase signals are balanced allows one to determine a simple signal model from which a multiobserver scheme is designed for residual generation.