Research On Fault Diagnosis Technology Of Sensors For Rotary Steerable Drilling Tool

Rotary Steerable Drilling Tool(RSDT)has become an indispensable equipment in oil and gas exploration and development.The sensor fault diagnosis technology of RSDT,which was closely related to its reliability and safety,has received extensive attention.In this thesis,the RSDT sensor fault diagnosis technology was studied based on a model-based fault diagnosis method for a self-developed dynamic point-the-bit RSDT prototype.1.The working principle,measurement and control technology of RSDT were introduced,and the mathematical model of the RSDT system was established.The influence of modeling errors,unknown inputs and sensor faults on the system model were determined.The common faults of RSDT sensors were modeled.2.According to the RSDT measurement and control system’s mathematic model,a robust fault detection observer was designed.Firstly,the system observer was developed based on the principle of unknown input observer(UIO)to make sure the residual can be decoupled from the unknown inputs entirely.Secondly,the robustness of the modeling error and the sensibility of sensors fault were obtained by combining H_∞norm with H_-index,a finite frequency H_-/H_∞UIO was designed,and the optimal solution algorithm of observer matrix was proposed.Thirdly,the transmission characteristics of sensor faults in the residual system were analyzed,which laid the foundation for sensor fault isolation.3.The Weighted Sum-Squared Residual(WSSR)method was used to evaluate the residual error and the calculation method of fault detection threshold was proposed,which realized the accurate fault detection of RSDT sensors.The fault detection performance of the observer was verified by MATLAB simulation and prototype experimental test.4.An RSDT sensor fault isolation scheme was proposed based on the observer fault detection results,the accelerometers’spatial configuration was improved,and the odd-even equation method was used to isolate the accelerometer fault.The coordinate transformation method was used to solve the current sensors’fault isolation problem.The experimental results showed that the proposed scheme could accurately locate the fault sensor.