RBF-Sliding Mode Variable Structure Control For Steering Drilling Stable Platform

Steering drilling stable platform is the key of the rotary steering drilling tool,the accurate and fast control can directly affect the efficiency and reliability of drilling.Because of the environment and various friction affected by the complex changeable,it is difficult to establish accurate mathematical model of the stable platform.So the traditional control method cannot control the stable platform.In this paper,combine the RBF neural network with sliding mode variable structure control theory and realized the control of the stable platform.Aiming at the uncertainty about friction for the stabilized platform of rotary steerable drilling system,RBF neural network adaptive sliding mode control method(RBFASMC)is proposed to improve the accuracy and anti-jamming capability in stabilized platform control.RBF neural network is used in stabilized platform control approach to the uncertainty in the model,meanwhile the size of the network be reduced through design the wake and the activation threshold of RBF network nodes.At last the weight adjustment adaptive law be devised and RBF network is combined with the sliding mode control to enhancement system robustness.Simulation experiments show that under the control method for stable platform system can achieve high precision tracking on the tool face angle,and the experimental robustness analysis,the system has strong robustness and can effectively eliminate the parameters and disturbance.In order to improve the speed of platform tracking tool angle,proposed RBF neural network direct adaptive sliding mode control method(RBFDASMC).This method is through design the controller parameters to improve the system convergence and approaching speed switching.Meanwhile,RBF neural network is used to approximate the unknown model of the system,and considering the unknown system model will have a negative impact on quality control,designed the new adaptive law,to prove the stability of the system.Simulation results show that the method is good for tracking tool face angle in parameter perturbation.The system has strong robustness,uncertain model and effective approximation system.