Litcius/Paper detail

Accelerometer Fault Detection for Rotary Steerable Drilling Tool Systems Under Strong Noises

Yichun Niu, Li Sheng, Ming Gao, Donghua Zhou

2022IEEE Transactions on Instrumentation and Measurement22 citationsDOI

Abstract

This paper is concerned with the problem of accelerometer fault detection for dynamic point-the-bit rotary steerable drilling tool systems (DPRSDTSs). The essential difficulty of this problem lies primarily in the fact that the accelerometer fault is usually submerged in strong noises induced by the mechanical vibration during the drilling process. Firstly, the DPRSDTS is modeled by a time-varying system with multiplicative noises. Subsequently, the fault detection filter and residual generator are designed by means of the Kalman filtering algorithm and the weighted moving average method. By choosing suitable weight matrices and window length, the detectability of accelerometer fault can be ensured in a probabilistic sense. In such a case, the real false alarm rate and missed detection rate are reduced below the allowable level. Finally, the simulations and experiments performed on the DPRSDTS prototype are given to show the feasibility and effectiveness of the developed method.

Topics & Concepts

AccelerometerFault detection and isolationFault (geology)Kalman filterControl theory (sociology)Constant false alarm rateResidualComputer scienceVibrationNoise (video)EngineeringArtificial intelligenceAlgorithmAcousticsSeismologyControl (management)PhysicsActuatorImage (mathematics)GeologyOperating systemAdvanced machining processes and optimizationFault Detection and Control SystemsDrilling and Well Engineering