Study On Immune Evolutionary Algorithm And Its Application To Water Problem

As key technologies to solve complex optimization problems, evolutionary algorithms are studied continuously both at home and abroad, and great achievement has been made. In the paper, based on the existed evolutionary algorithms, further study has been done, and main works are illustrated as follows:1. This paper presents a comparatively systematic review about progress, basic principle, processing technologies and application of the existed evolutionary algorithms, and stress of this part is mainly put on related problems of genetic algorithm. Besides that, disadvantages of genetic algorithm are also analyzed, and possible improved approaches are pointed out.2. Based on study of the existed evolutionary algorithms' property and enlightened by the immune principle of creature, a novel evolutionary algorithm-IEA(Immune Evolutionary Algorithm) is creatively proposed. As to this algorithm, our research consists of the next three parts. First, we introduce the design idea, flow chart and characteristics of SIEA(Simple Immune Evolutionary Algorithm), put forward general convergence theory about it, and prove it to befeasible and correct from the aspect of application through the testing results on functions with multi-model. Second, based on SIEA, we propose two improved algorithms respectively, which are the IEA-BIT(Immune Evolutionary Algorithm Based on Interval Transition) and IEA-BNR(Immune Evolutionary Algorithm Based on Network Regulation). Compared with SIEA, IEA-BIT not only achieves better results in application to functions with multi-model and GA deception problem, but also further improves computation efficiency and enhances generalization of the model. IEA-BNR is an improved algorithm that bases itself on the above two algorithms, and through primary application to complex constrained optimization problem, it is proved to have the ability to solve more complex problems. Finally, we make detailed discussion about the processing technologies of IEA, and then summarize the characteristics of it.3. IEA is applied to solve some optimization problems in water science, and the results show that IEA is simple, efficient and robust, and can preferably overcome the disadvantages of traditional optimization methods and the existed evolutionary algorithms.Above all, IEA is an algorithm with good properties, and it can be widely applied to solve complex optimization problems in water science.