模糊广义学习向量量化算法在交通事故预测方面的应用

Traffic Safety Forecast with GLVQ Based on Fuzzy Generalized Neural Network

模糊逻辑系统易于理解,而神经网络则有极强的自适应能力。本文将模糊数学方法和神经网络结合起来建立组合模型,用广义模糊神经网络(FGNN)预测道路交通事故。运用MATLAB语言编程,利用模糊广义学习向量量化算法(FGLVQ)建立模糊神经网络模型,并应用于交通事故预测中,改进了交通事故预测的计算方法。理论分析和实例表明,设计的模糊神经网络模型具有良好的非线性映射功能和泛化功能,对预测交通事故有较好的适应性。

Fuzzy logic system is easy to understand,and the neural network has very strong adaptive ability.A hybrid model combined GLVQ(generalized learning vetor quantization) and GFNN for forecasting traffic safety had been put forward in this paper,which used MATLABL programming and fuzzy revised generalized learning vector quantization to cope with the subjective factors in traffic accident prediction.The model improved the computational methods for traffic safety.The theoretical analysis and experimental results indicated that the model was effective in nonlinear mapping and generalization.