Gain tuning PI controllers for boiler turbine unit using a new hybrid jump PSO

Abstract

In this paper, a new hybrid jump PSO (HJPSO) is proposed for tuning the gains of PI controllers to the boiler turbine unit. HJPSO based Gaussian and Cauchy mutation is proposed to improve the standard PSO performance. The new strategy is based on observing the local and global best particles which are not improved in a predefined number of iterations and moving these particles to a new best position. Besides, forming a new particle that handles the minimum error of each controller to replace the global best particle if it has best fitness. The simulation results show that the proposed algorithm has its ability in optimizing the control parameters and effectively achieved better performance when compared with other PSO algorithms.

Keywords

• Boiler-turbine unit
• Particle swarm optimization (PSO)
• Gaussian and Cauchy mutation and PI controller