神经网络自适应模糊IDSS系统研究

提出了采用神经网线络与模糊控制相结合的方法来设计自适应ＩＤＳＳ。阐述了神经网络控制器的设计原则及其算法，给出了神经网络自适应ＩＤＳＳ的总体设计和系统结构。

On Neural Network Self- Adaptive Fuzzy IDSS

Sine 1990, the study of neural network self- adaptive fuzzy IDSS(Intelligent DecisionSupport SyStem)has been one of the important tasks of information processing syStem. Wecombine neural network with fUZZy control system to form neurl network self adaptive fuzzyIDSS. The main results of our research are: (1) the conversion of numerical computation ofn6ural network into its fUZZy computationl (2) establishment of rules for neural networkfUel control through combining control rules with control tactics ; (3) use of neural networkto set up a neural network self-adaptive syStem in the tegular parts of fUzzy syStem (Fig. 1);and (4) definition of neural network fuzzy control function and that of fuzzy controlfunction. Our main results provide the design rules that can integrate neural networkself-adaptive fuzzy control with fuzzy database (Fis. 2). We give a new algorithm, calledNSAFC by us, for use in our controller. Compared with ordinary IDSS controller, our controller possesses stronger capacity for solving unstructured decision problems and enjoysgreater fault tolerance.