Quantum neural network-based intelligent controller design for CSTR using modified particle swarm optimization algorithm

In this paper, a combination of a multi-layer quantum neural network (QNN) with the particle swarm optimization (PSO) algorithm is used with the aim of controlling a continuous stirred-tank reactor (CSTR) system. The CSTR process is highly non-linear and its dynamics are significantly sensitive to system parameter values. Normally, conventional controllers with fixed coefficients are applied to control this kind of system. In highly non-linear systems, having fixed controller coefficients in different operational conditions may decrease the performance of controllers. In the proposed scheme, by using a multi-layer QNN, an adaptive structure is designed for a PI-D controller. In order to train the QNN, the PSO algorithm is employed. With the aim of improving accuracy and convergence speed of the training process, some modifications have been applied to the movement of each particle towards the optimal point. Furthermore, in order to evaluate the performance of the system, the proposed scheme has been applied in various operational situations in the presence of disturbances and set-point change. The efficiency of the proposed control scheme is compared with PID and a perceptron neural network-based controller, and the simulation results endorse that the proposed scheme shows significantly better performance in different operating conditions.