Theoretical Study On Stable Platform Control Of Full Rotary Closed-loop Directional Drilling Tools

The different guiding control methods of the rotary steerable drilling tool guiding actuator directly determine the difference of the guiding ability of the drilling tool.Among them,the stable platform is the core of the rotary steerable drilling tool.Only the precise closed-loop control of the stable platform can provide the system with stable and controllable azimuth and guiding force under dynamic working conditions.This paper studies and analyzes the stable platform of the full-rotation closed-loop directional drilling tool.The mathematical model of the generalized control object of the stable platform is established,and its state space expression and transfer function are obtained,and its controllability and observability analysis are carried out.According to the working principle of the dynamic directional guiding drilling tool,the control method of the double eccentric ring biasing mechanism of the full rotation closed-loop directional drilling tool is analyzed.The RBF neural network adaptive control is used to approximate the nonlinear model of the drilling system,and the overall approach and block approximation are used to design the control mode.Finally,the simulation analysis is carried out to verify the effectiveness of the control method.Aiming at the problems of adaptive control of RBF neural network,a two-loop integral sliding mode controller is designed by using the combination of speed loop and position loop.The stability is verified by Lyapunov function.The trajectory tracking of the inner and outer eccentric rings under four sets of different position commands proves that the control method has better robustness and anti-interference.For the problem of chattering phenomenon and noise of control system,a double-loop integral sliding mode control algorithm based on adaptive strong tracking Kalman filter is proposed.After adding adaptive Kalman filter,the controller can effectively estimate according to noise statistics.The ground filter removes white noise and accurately predicts the current state,and the interference against external noise achieves an ideal control effect.