Modified differential evolution algorithm and its application in thermal process model identification

The mathematical model of the object in power plant is of extremely significance for the design and analysis of the thermal control system. There are many methods to identify the parameters of the desiring object. In this article, we adopt a modified form of a relatively effective yet simple algorithm called differential evolution algorithm (DE) which is a population based stochastic optimization approach. The differential evolution algorithm uses the difference of randomly sampled pairs of vectors in the population for its mutation operators and is applied mainly in real parameter optimization. Based on analysis of DE searching mechanism, the article proposed the improved differential evolution algorithm with self-adaptive parameters to promote its robust, optima searching capability and speed. In order to prove the effectiveness of the improved differential evolution algorithm, we work out relevant model identifying program on MATLAB and identify the mathematical models. Then we analyze the result using the method of comparing.