New PID Parameter Autotuning for Nonlinear Systems Based on a Modified Monkey–Multiagent DRL Algorithm

• Main Authors: Hongming Zhang, Wudhichai Assawinchaichote, Yan Shi
• Format: Article
• Language: English
• Published: IEEE 2021-01-01
• Series: IEEE Access
• Subjects: PID controller; monkey algorithm (MA); multiagent deep deterministic policy gradient (MADDPG); modified monkey–multiagent DRL (MM-MADRL) algorithm; optimization
• Online Access: https://ieeexplore.ieee.org/document/9440392/

Proportional–integral–derivative (PID) control is the most widely used control law in industrial processes. Although various new controllers continue to emerge, PID controllers are still in a dominant position due to their simple structure, easy implementation, and good robustness. In the design and application of PID controllers, one of the core issues is parameter tuning. Accurately and effectively selecting the best tuning parameters of the PID is the key to achieving an effective PID controller. Therefore, this paper proposes a novel modified monkey-multiagent DRL (MM-MADRL) algorithm and uses it to tune PID parameters to improve the stability and performance of automatic parameter optimization. The MM-MADRL algorithm is a new version of the basic monkey group algorithm (MA) and the multiagent reinforcement learning algorithm known as the multiagent deep deterministic policy gradient (MADDPG). This paper selects a typical nonlinear quadcopter system for simulation; its principle and data are given below. MM-MADRL, the genetic algorithm (GA), particle swarm optimization (PSO), the sparse search algorithm (SSA) and differential evolution (DE) are used to adjust the parameters. The simulation results show that the overall performance of the MM-MADRL algorithm is better than that of the other algorithms.