基于人工蜂群算法优化BP神经网络的交通流预测

Traffic Flow Prediction of BP Neural Network Optimized by Artificial Bee Colony Algorithm

建立基于人工蜂群算法(Artificial Bee Colony Algorithm,ABC)优化BP(Back Propagation)神经网络(ABCBP)的分析预测模型,对城市道路短时交通流进行预测。以BP神经网络为基础,通过人工蜂群算法优化神经网络的各个权值和阈值,考虑交通流的时间特性,将历史交通流量作为训练样本,预测某日的交通流量。多种算法的仿真试验对比表明:基于ABC-BP的预测结果比传统BP神经网络、小波预测神经网络以及PSO(Partide Swarm Optimization)-BP神经网络的预测结果更加精确。

The analysis and prediction model of the urban road short-term traffic flow is established based on a neural network optimized by the Artificial Bee Colony Algorithm. The thresholds and weights of the neural network are optimized by the Artificial Bee Colony Algorithm. Taking the time characteristic of the traffic flow into account, the historical traffic flow is used as the training sample to predict the traffic flow of a day. The contrast of stimulation tests of multi-algorithms shows that the prediction results based on ABC-BP are more accurate than those based on the traditional BP neural network, the wavelet neural network and PSO-BP neural network.