用模糊神经网络预测蘑菇生长过程

Predicting Mushroom Growth Process with Fuzzy Neural Network

文章探讨了模糊神经网络的基本构造和原理,结合蘑菇生长过程预测系统重点分析了FNNC摸型的推理和学习方法。并在此基础上提出了TPH学习方法。该方法吸收了梯度下降算法和随机搜索算法的优点,能够使生长过程预测系统的学习以很大概率快速收敛在系统误差的最优点附近。最后文章指出模糊神经网络以及TPH学习算法在农业生产过程的应用。

Fuzzy Neural Network is widely used to deal with fuzzy knowledge.This paper designs a mushroom growth process prediction system based on fuzzy neural network controller principles and presents a new training algorithm named two phases hybrid algorithm, which combines the genetic algorithm and the back-propagation algorithm together in order to quickly complete the training near the point of the minimum system errors in a soundly probability. At last the paper believes that the fuzzy neural network and the TPH algorithm could have more applications in agriculture production.