Abstract

Under the inspiration of particle swarm optimization (PSO), this paper proposed an efficient artificial immune network (EaiNet) for function optimization. This algorithm absorbs the learning mechanism of PSO, i.e., the elite learning that each individual is capable of learning from the best in the social population. The insert of the elite learning mechanism quickens the convergence of EaiNet. Besides, EaiNet has self-learning capability, especially when it is fell in the local optima, which will result in finer global optima. Compared to the cononical artificial immune network (aiNet), EaiNet proposed in this paper has better solution quality and faster convergence speed, which indicates that EaiNet is an effective optimization method. The results of an industrial application in PID controller also demonstrate the effectiveness and robustness of the proposed EaiNet algorithm.

Paper Details