Parameter Estimation of Uncertain Fractional-Order Chaotic Power System Using an Improved State Transition Algorithm

Fractional-order chaotic power system parameter estimation is a multidimensional function optimization problem, and it is a prerequisite for controlling such systems. Aiming at the problem that the parameter estimation of a fractional-order chaotic power system is easily affected by external interference and estimation is difficult, a state transition algorithm based on the reverse learning strategy of lens imaging is proposed, taking a fractional six-order chaotic power system as the study model. First, the tent chaotic mapping to initialize the population, thus increasing population diversity. Second, we optimize parameters and of the basic state transition algorithm to improve its global and single-dimensional search abilities. We use the lens imaging learning strategies to avoid the local optima. The proposed improved state transition algorithm shows high estimation accuracy and convergence speed, and is superior to the traditional state transition, particle swarm optimization, genetic and gray wolf algorithms. The simulation results show that the parameters of the fractional sixth-order chaotic power system are identified accurately. even in the presence of white noise, demonstrating the strong robustness and versatility of the proposed algorithm