交互式多智能体进化算法及其应用

将多智能体技术应用于交互式进化计算领域，提出一种新的交互式多智能体进化算法。该方法让固定在网格上的相邻智能体之间展开竞争和熏优智能体本身的自学习，来提高智能体的能量，从而使得算法获得较强的全局收敛能力和局部搜索能力。算法以较快的进化速度收敛，并且用户每次只需要选择2个左右最感兴趣的个体，使得评价具有易操作性和轻松性，从而有效缓解用户的疲劳。服装设计的仿真实验证了算法的可行性和有效性。

Interactive Multi-Agent evolutionary Algorithm and Its Application

By introducing multi-agent technique into the domain of interactive evolutionary computation, a new interactive multi-agent evolutionary algorithm (IMAGA) was developed. Every agent fixed on a lattice-point competes with its neighbors and the optimal one also can do some self-learning in order to increase energy, which improves the algorithm globally and locally searching ability. The proposed algorithm obtains quick speed of convergence, and the user only needs selecting about two mostly interesting individuals in every evolutionary procedure, which leads to the work of evaluating easy and tractable and contributes to overcome human fatigue. The simulation of fashion design shows the developed algorithm is feasible and effective.