On the Application of a Hybrid Genetic-Firework Algorithm for Controllers Structure and Parameters Selection

An approach proposed in this paper uses a new hybrid population-based algorithm. This algorithm is a fusion between genetic algorithm and firework algorithm. Proposed approach aims on solving complex optimization problems in which not only structure parameters of the solution have to be selected, but also the mentioned structure. Proposed approach is based on multiple linear correction terms PID connected using proposed dynamic structure. In simulations a problem of selecting structure and its parameters for automatic control was used. For system evaluation a weighted multi-objective fitness function was used, which can consider elements connected to the simulation problems taken into consideration, such as: RMSE error, oscillations of the controller output signal, controller complexity and overshoot of the control signal.