Finding the parameters of the controller of a nonlinear twin-motor multiple-input multiple-output system using metaheuristic optimization methods

The twin-rotor multi-input and multi-output system (TRMS), a helicopter simulator, is a nonlinear system. The TRMS system used in this study consists of main and tail propellers. In order to keep the system in balance, the speeds of the DC motors that move the propellers must be controlled. For this control, a discrete-time Proportional-Integral-Derivative (Discrete-PID) controller is used. Finding the controller parameters of nonlinear systems is very difficult with classical approaches. Metaheuristic algorithms are basically algorithms inspired by the behaviour of animals moving in herds. In this study, the discrete time PID parameters to keep the system in equilibrium are determined using metaheuristic algorithms such as particle swarm optimization (PSO), butterfly based particle swarm optimization (BFPSO), dragonfly algorithm (Dragon), genetic algorithm (GA), grey wolf optimisation algorithm and cuckoo. The results of simulation and experimental studies using the parameters obtained from the algorithms are compared. The comparison was made in terms of error and performance. In terms of error, integral absolute error (IAE), integral squared error (ISE), integral time squared error (ITSE) and integral time absolute (ITAE) values were used, and in terms of performance, rise time to reference value, settlement time, percentage overshoot and peak values were used. If a quick response is needed and it is believed that the error will not be a significant issue, then the parameters obtained from the GA should be used. However, if the error will be a problem and time is not an issue, the grey wolf or KTPSO algorithms can be utilized. In this context, the GA, grey wolf, and KTPSO algorithms have become prominent in determining PID controller coefficients for a multi-input, multi-output system (TRMS) with twin motors.