Modeling of a multi-circuit system for automated control of the well drilling process based on subcontrollers

Abstract

Decentralized automated control systems (ACS) based on proportional-integral-differential (PID) controllers remain the most common solution for controlling industrial processes with multiple inputs and outputs (MIMO). This is due to their flexibility, simplicity, and inherent fault tolerance compared to centralized control structures. Even when more sophisticated strategies, such as model predictive control (MPC), are applied, PID controllers are usually used at lower control levels of the respective systems. The paper considers an automated control system for the well drilling process implemented on the basis of an electric drive. In the studied hierarchical ACS, a structure is used at its lower level that includes three circuits based on proportional-integral-differential (PID) controllers that provide control of the bit rotation speed RPM, the bit load WOB, and the parameters of the well cleaning system (primarily its performance). Drilling process modeling shows a close relationship between these parameters, which requires appropriate adjustment of each control loop. The optimal tuning of decentralized PID controllers in multi-loop ACS remains a difficult and complex task. Most of the known methods of tuning multi-loop PID controllers are similar in that they use the tuning rules of one loop to obtain initial values for individual controllers, and then detune the individual loops to maintain the stability of the entire system. In the conditions of a real production process, this approach does not always allow achieving the desired performance and control reliability. An algorithm for tuning multicircuit SACs based on PID controllers, which implements the generalized internal model method (IMC), is investigated. A criterion based on determining the parameters of the closed-loop frequency response is used to achieve the desired performance and control reliability. The frequency response of the closed-loop ACS is obtained by calculating the system output signal in response to a sinusoidal input. The results obtained indicate that the tuning of a three-circuit ACS based on PID controllers using the investigated method allows achieving the most accurate compliance with the desired performance and control reliability. The investigated method can be applied to dynamic models of any order. In this case, the feedback characteristics of the closed-loop control are set in advance. In addition, as a result of the corresponding algorithm, information is provided on the stability margins and sensitivity characteristics of the ACS.