Fault estimation for rotary steerable drilling tool systems with strong vibrations and unknown terms

In this paper, the problem of gyro fault estimation is studied for the dynamic point-the-bit rotary steerable drilling tool system (DPRSDTS), and it is essentially the fault estimation problem for nonlinear systems with strong vibrations and unknown terms. First of all, considering the mechanical structure of the engineering system, the DPRSDTS is described by a polynomial nonlinear system, where strong vibrations are modeled as high-frequency noise, and the load torque and the interference of the outer housing are modeled as unknown inputs. High-frequency noise caused by strong vibrations during drilling is suppressed by the low-pass filter. By unknown input observer, unknown inputs generated by engineering modeling are decoupled. Then, a novel polynomial nonlinear fault estimator is constructed, and gain matrices are obtained by solving the parameter-dependent linear matrix inequality through the sum-of-squares decomposition method. Moreover, the sufficient condition is derived to guarantee the H performance of the estimation error system. Finally, an experiment is provided to validate the feasibility and effectiveness of the proposed fault estimation scheme.