Intelligent tuning of PID controllers: Comprehensive approach based on modified Particle Swarm Optimization (PSO) algorithm

This paper introduces a novel approach for tuning PID controllers by framing the tuning problem as an optimization task, solved using a modified version of Particle Swarm Optimization (PSO). The modified PSO algorithm achieves high precision without requiring manual adjustment of hyper-parameters. Unlike conventional optimization-based tuning methods that focus solely on minimizing a specific type of error, this approach employs a multi-objective formulation to simultaneously minimize Integral Absolute Error (IAE), control action aggressiveness, and settling time. The optimization problem is treated as a single-objective problem with weighted indices for each minimization goal. To enhance adaptability in controller tuning, the optimization problem incorporates constraints to precisely manage specific parameters of both the system response and the control action. Users can adjust the constraints and weights to achieve precise and adaptable control over the system’s overshoot and settling time. Additionally, the amplitude and maximum variation of the control signal can be regulated, resulting in a versatile solution adaptable to various design requirements. This approach was evaluated on a thermal process model based on a Peltier cell. The results underscore the effectiveness of this method, offering a comprehensive and versatile tool for the precise tuning of PID controllers.