基于模糊感知器的路面性能预测方法

利用模糊感知器人工神经元网络尝试建立一个智能型路面性能预测模型，对神经网络原理和模型结构进行了系统的理论分析，给出了详细的算法，并以平整度历年调查资料为数据基础，对该模型进行了实例分析，证明了模型的理论与实际运用的可行性。该方法利用神经网络与模糊方法各自的优点，针对不同性能指标，建立对应的神经网络预测模型，来对今后的指标值进行预测，较好地解决了路面性能预测中的模糊性和不确定性。

A Pavement Performance Predicting Method on the Fuzzy Perceptron

By using fuzzy perceptron artificial neural network, an intelligent predicting model of the pavement performance is attempt to establish. The principle of artificial neural network and model structure are analyzed by the numbers and the detailed arithmetic is made out. On the basis of surveying data, the predicting model is validated and proved to be used in practice. And according to different performance indices, the corresponding predicting model of the neural network can be established. It is used to predict the indices in next year. The virtues of the neural network and fuzzy theory are combined, so the fuzzy and uncertain matter is solved preferably in the prediction of the pavement performance.