Fractional order PID controller design using predator effect inspired artificial bee colony algorithm

This paper presents an artificial bee colony with predator effect algorithm to design the fractional order proportional integral derivative (PID) controller which is a constrained optimization problem with design space of variable size. To address this complex challenge, a better solution approach with a right balance between exploration and exploitation is required. The proposed approach utilizes a dynamic adaptation mechanism, defined by a random predator-based Euclidean distance operator, which markedly improves self-adaptive search abilities within multidimensional hyperspace. The efficacy of the proposed methodology is evaluated using three different test systems, each posing distinct challenges: a second-order process with a right half-plane zero ( G 1 ( s ) ) , an integrating system with significant dead time ( G 2 ( s ) ) , and a jacketed stirred tank reactor ( G 3 ( s ) ) characterized by complex dynamics. The choice of these systems establishes a strong foundation for assessing the adaptability and effectiveness of the suggested algorithm. Comparison of the proposed approach with established algorithms reported in the literature has been carried out. This assessment emphasizes essential performance indicators, such as time domain response criteria, control effort, and computational efficiency. The simulation studies reveal minimum values for rise time ( t r ) and settling time ( t s ) i.e. 10.71 sec and 14.56 sec respectively for system G 1 ( s ) with mean square error performance index. Similarly, for G 2 ( s ) the maximum overshoot 4.63% while t r and t s are 26.23 sec and 71.24 sec respectively for weighted mean of error and control effort performance criteria, which is better than its counterparts. Finally, for G 3 ( s ) the maximum overshoot is greatly reduced to 1.49% compared to others. This study also examines the convergence behavior of the algorithm, providing insights into its computational dynamics under different settings. To verify the robustness of the proposed approach, delay has been introduced in a test system. The trials validate the resilience of the suggested strategy, emphasizing its ability to effectively address real-world uncertainties in control system design. • Basic ABC algorithm lacks population diversity resulting in poor solution quality. • Proposed ABCPE algorithm show better convergence due to predator based Euclidean distance operator. • ABCPE algorithm is successfully used to tune the parameters of FOPID controller. • Tuning problem for an inverse response system and a large dead time process is considered. • Robustness of ABCPE approach is verified by introducing delay in the test systems.