多输入模糊神经网络结构优化的快速算法

采用规则前件提取，以获得较少的高效规则，对模糊神经网络(Fuzzy Neural Network)进行结构优化，解决了在多输入模糊系统中因规则数多导致的结构庞大问题，使之适用于多输入模糊系统．结构学习中采用竞争算法优化隶属函数，保证规则前件提取的高效；参数学习中采用梯度下降法调整网络参数。

Fast Learning Algorithm of Small Multi-input Fuzzy Neural Network Structure

The algorithm of rule extraction is applied to accelerate learning process and extend the application of FNN to overcome the dimensionality problem of fuzzy neural network(FNN) in multi-input fuzzy system through decreasing the number of rules and establishing small FNN structure. Competitive algorithm is used to optimize the parameters of membership function in structure learning before rule extraction. In parameter learning,the weights of FNN are adjusted by gradient algorithm. The simulation results show that the proposed method works effectively.