Attitude Control of Rotary Steering Drilling Stabilized Platform Based on Improved Deep Deterministic Policy Gradient

Abstract

Summary The rotary steerable drilling system is an advanced drilling technology, with stabilized platform toolface attitude control being a critical component. Due to a multitude of downhole interference factors, coupled with nonlinearities and uncertainties, challenges arise in model establishment and attitude control. Furthermore, considering that stabilized platform toolface attitude determines the drilling direction of the entire drill bit, the effectiveness of toolface attitude control will directly impact the precision and success of drilling tool guidance. In this paper, a mathematical model and a friction model of the stabilized platform are established, and an improved deep deterministic policy gradient (I_DDPG) attitude control method is proposed to address the friction nonlinearity problem existing in the rotary steering drilling stabilized platform. A prioritized experience replay based on temporal difference (TD) error and policy gradient is introduced to improve sample usage, and high similarity samples are pruned to prevent overfitting. Furthermore, SumTree structure is adopted to sort samples for reducing computational effort, and a double critic network is used to alleviate the overestimated value. Numerical simulation results illustrate that the stabilized platform attitude control system based on I_DDPG can achieve high control accuracy with both strong anti-interference capability and good robustness.