进化算法在求解连续空间函数优化问题上的比较

Evolutional algorithms in optimizing continuous function

优化算法就是一种搜索过程和规则,它是基于某种思想和机制,通过一定的途径和规则来得到满足用户要求问题的解.优化问题也是人们在工程技术、科学研究和经济管理等诸多领域中经常遇到的问题.鉴于工程优化问题的复杂性、约束性、非线性、多极小等特点,分析各种算法,寻求一种适合要求的算法很重要.目前,进化算法主要包括遗传算法、模拟退火法、混沌算法、蚂蚁算法、粒子群算法等.分别介绍了5种算法的基本思想以及它们的特点,并且通过对几个连续函数的优化比较,阐述了算法的各自不足.

Optimum algorithm is a searching process based on a certain idea and mechanism to solve problem. The optimization design is also the usual problem in engineering, science research and economy management fields. Finding a proper algorithm is important due to the complexity, restriction, nonlinearity and multi-minimum of engineering problems, The evolutional algorithm includes genetic algorithm, chaos algorithm, ant algorithm, particle swarm algorithm, simulated annealing and so on. The basic concepts and characteristics of five optimization algorithms are introduced, and compared by optimizing several continuous functions to point out the disadvantages of these algorithms.