Stability-Preserving Automatic Tuning of PID Control with Reinforcement Learning

TL;DR

This paper introduces a reinforcement learning-based framework for automatic PID tuning that guarantees closed-loop stability during the tuning process, using a novel episodic approach and stability supervision.

Contribution

It presents a stability-preserving RL tuning method with an episodic framework, baseline initialization, and a supervisor to prevent instability during PID parameter optimization.

Findings

Maintains closed-loop stability during RL-based PID tuning.

Converges quickly to optimal PID parameters.

Validated on a second-order plus dead-time system.

Abstract

PID control has been the dominant control strategy in the process industry due to its simplicity in design and effectiveness in controlling a wide range of processes. However, traditional methods on PID tuning often require extensive domain knowledge and field experience. To address the issue, this work proposes an automatic PID tuning framework based on reinforcement learning (RL), particularly the deterministic policy gradient (DPG) method. Different from existing studies on using RL for PID tuning, in this work, we consider the closed-loop stability throughout the RL-based tuning process. In particular, we propose a novel episodic tuning framework that allows for an episodic closed-loop operation under selected PID parameters where the actor and critic networks are updated once at the end of each episode. To ensure the closed-loop stability during the tuning, we initialize the training with a conservative but stable baseline PID controller and the resultant reward is used as a benchmark score. A supervisor mechanism is used to monitor the running reward (e.g., tracking error) at each step in the episode. As soon as the running reward exceeds the benchmark score, the underlying controller is replaced by the baseline controller as an early correction to prevent instability. Moreover, we use layer normalization to standardize the input to each layer in actor and critic networks to overcome the issue of policy saturation at action bounds, to ensure the convergence to the optimum. The developed methods are validated through setpoint tracking experiments on a second-order plus dead-time system. Simulation results show that with our scheme, the closed-loop stability can be maintained throughout RL explorations and the explored PID parameters by the RL agent converge quickly to the optimum.