No abstract

No abstract

No abstract

No abstract

In this paper, a performance comparison between the Marine Predators Algorithm (MPA), a metaheuristic paradigm, and two other designed variants for the tuning of a fractional proportional-integrative-derivative (PID) controller in a multiple-input multiple-output (MIMO) application is presented. The practical system plant corresponds to a ball mill pulverizing system, whose structure presents two inputs and two outputs. To encounter the optimal response on the MIMO control of this system a MPA approach applied to PID tuning is suitable, as it presents both the capability to diversify the search space (exploration) and to improve the quality of current solutions (exploitation) in search space. The MPA is a metaheuristic inspired by the extensive hunting strategy of ocean predators called Lévy and Brownian movements, it focuses on an optimal confront rate procedure in natural interaction between predator and prey in the marine ecosystem. The original MPA itself presents a satisfactory performance, in terms of statistical metrics. Nevertheless, it can be improved through the modification and addition of distinct techniques. In order to achieve those modifications, three variants are implemented exploring different procedures namely the oppositional-based learning and application of quantum mechanics. The optimal parameter values for the PID controller are analyzed by minimizing the integral time squared error (ITSE) index of the system’s response. The simulations are performed using the SIMULINK/MATLAB computational environment. Statistical measures including best, mean, median and standard deviation of the system response error for the tuned controllers are evaluated and compared over fifty runs. The obtained results suggest that the use of the mentioned proposals has an advantage in enhancing the tuning efficiency of the MPA in this application.