RBF Neural Network Based-PID Control for Weight on Bit During Drilling Operation
Deep drilling is a costly project and efficiency is of paramount importance. The weight on bit is one of the main operating parameters that influences the drilling efficiency and it was controlled by manual before. But after people saw the giant potential of an auto-drilling system in increasing the drilling efficiency, more and more studies on the feed back control of weight on bit have emerged. This paper mainly studied weight on bit dynamic under the variational formation based on a lumped parameter model and a self-tuning PID controller for weight on bit control. The parameters of the PID controller are tuned by using gradient descent method and RBF neural network identification.