The work aims to propose an adaptive gray wolf optimization algorithm based on the convergence factor and dynamic changes of weights to solve problems such as low precision, slow convergence rate and poor local search ability of basic wolf algorithm in function optimization. A focusing distance changing rate for dynamically updating the convergence factor was given to maintain a balance between global search and local search of the algorithm. The position updating formula of the algorithm was adjusted by introducing the adaptive weighting factor, to improve the convergence speed and precision of the algorithm. The simulation results showed that the improved algorithm had a significant improvement in convergence accuracy and speed, and overcame the shortcoming of the gray wolf algorithm that it was easy to fall into a local optimum when processing multi-modal functions. For pulp concentration control systems, the control effect was relatively ideal. The PID controller parameters set by the improved gray wolf algorithm can obviously improve the performance indicators such as the control accuracy of the system, and can better meet the requirements of practical application.