多属性决策神经网络模型与应用

Neural network model for multi-attribute decision making

针对多属性决策问题,提出了基于理想解的竞争神经网络模型方法.采用竞争神经网络对备选方案进行学习、分类,并结合多属性决策的理想解概念,根据方案间欧式距离远近,对备选方案进行排序.最后使用Matlab软件进行仿真计算,得到码头较优的结构方案,表明该方法是可行的.

Considering multi-attribute decision making problem,a competing neural network model was established based on ideal solution.The network was adapted to learn and classify alternative solutions.Then,the concept of ideal solution was introduced.The problem was conducted by the Euclidean distances between the alternative solutions and the ideal solution.Simulations show that the neural network can achieve an advisable decision making for a certain wharf structure and the developed method is feasible.