利用神经网络的学习功能对控制器的隶属度函数及推理规则进行修正，以提高其自适应能力。
The membership functions and the inference rules in the controller are modified using the learning functions of neural network so that the adaptability of the controller is further enhanced.
然后结合隶属度变量构建优化模型，利用具有动态惩罚函数的遗传算法求解，计算得到各方案的所属类别。
Then optimization model with the membership degree is constructed. It is solved by using genetic algorithms with dynamical castigatory function.
分析过程中用隶属度函数衡量各影响因子代表的发电商串通报价可能性的程度，并利用变权法改变权重，突出薄弱环节。
The membership grade function is used to analyze the collusion bidding possibility of the contributed factors and the variable weight method is used to show the weak point.
应用推荐