Parametric Identification with Performance Assessment of Wiener Systems Using Brain Storm Optimization Algorithm

This paper proposes a performance assessment-based system identification of different practically useful open-loop and closed-loop Wiener systems using an evolutionary computational algorithm named as brain storm optimization (BSO) algorithm. Different performance measures of the estimation process in practical scenario, i.e., accuracy; precision; consistency; and computational time, are measured with a properly selected fitness function which is the output mean square error (MSE) between the desired and the estimated outputs. Bias and variance of the output MSE have been found negligible for each plant model to show the accuracy and consistency limits, and the corresponding statistical test results have been shown to establish the consistency of the BSO-based identification approach. Efficient identification of each plant under a noisy environment ensures the robustness and the stability of the proposed evolutionary-based identification approach with BSO. The BSO-based optimum MSE values, corresponding estimated parameter values, computational times and the other statistical information are all compared and are found to be superior to those of the other approaches reported earlier.