Parameter Identification of Dynamic Load Model Based on Chaotic Optimization Strategy

Aiming at the parameter identification of power system dynamic load model which is represented as three order induction motor, we propose a high precision chaotic optimization strategy which searches optima by means of ergodicity, regularity and intrinsic stochastic properties and can track any state in a certain scope without repetition according to its own regularity. For optimal problem, it needn't possess continuity and differentiability, so that it can be used for nonlinear and discontinuous problems and overcome the disadvantages of traditional methods which are easily being trapped in local optima and have high request to initial values and bad robustness. The presented algorithm uses two chaotic maps, three search steps and introduces random numbers to speed up its convergence. The results of a practical example prove its validity and accuracy when applied it in parameter identification of dynamic load model.