Multi objective optimization in the level controller project in a pilot plant

Abstract

Controlling the process variables at the desired point represents from economic gain, quality in the final product and control of industrial safety. The optimization aims to search for the best solution, so that this occurs, algorithms such as Firefly Colony and Differential Evolution are used, which minimize the relative error of the controlled variable output signal. In order to compare the differential evolution algorithms and the bio-inspired: firefly colony and the classic Ziegler-Nichols method, the PI control and PID control project applied to a pilot plant was carried out, considering two final elements control systems (valve and pump) for level control, establishing the minimum effort of the manipulated variable as a performance criterion. The comparison of the results obtained between the designed controllers, using the three methods for the two elements under study, demonstrated that the optimization algorithms showed an excellent control of the process, without any overshoot (maximum deviation from the value of the controlled variable, with the setpoint value). The results obtained demonstrated that both optimization algorithms are good methods for the design of controllers when worked with the minimum effort of the manipulated variable.