旋转导向钻井稳定平台的RBF网络滑模变结构控制

Sliding Mode Variable Structure Control of Stabilized Platform in Rotary Steerable Drilling System Based on RBF Neural Network

针对旋转导向钻井稳定平台存在的摩擦问题带来的不确定性,提出一种基于RBF神经网络的自适应滑模变结构控制方法,以提高稳定平台控制的精确性和抗干扰能力。使用RBF神经网络对稳定平台模型中的不确定性进行逼近,通过设计RBF网络节点的唤醒与激活阈值来减少网络规模,同时设计权值调整的自适应律,并结合滑模控制增强系统的鲁棒性。分别采用一般滑模变结构控制方法和RBF神经网络滑模变结构控制方法进行仿真实验,结果表明,RBF神经网络滑模变结构控制方法能够有效地逼近控制对象模型,有较强的鲁棒性。

Aiming at the uncertainty of the stabilized platform in rotary steerable drilling system caused by friction, an adaptive variable structure control method based on RBF neural network is proposed to improve the accuracy and the anti-interference ability in the control of the stabilized platform. RBF neural network is used for approximating the uncertainty of the control model of the stabilized platform,the size of the RBF neural network is reduced by designing the wake-up and activation threshold of RBF network nodes, and the system robustness is increased by the adaptive law with adjustable weights and sliding mode control. The simulated result using general sliding mode variable structure control method is compared with that using the sliding mode variable structure control method based on RBF neural network, and it is shown that, the sliding mode variable structure control method based on RBF neural network can effectively approximate the control object model and has strong robustness.