基于改进人工蜂群算法优化小波神经网络的短时交通流预测

为了提高城市道路短时交通流量的预测精度，克服小波神经网络预测过程中存在收敛速度较慢、容易陷入局部最优的缺点，提出改进的人工蜂群算法优化小波神经网络预测模型。该算法引入差分进化算法中的自适应变异操作和遗传算法中的选择算子、交叉算子与变异算子来优化传统的人工蜂群算法，改善人工蜂群算法后期收敛速度慢、局部搜索能力弱的缺点。本文使用该算法优化小波神经网络的参数并对短时交通流进行预测，模型的仿真结果表明，改进人工蜂群算法优化小波神经网络预测的结果误差更小，精确度更高，训练次数少，具有较高的实际应用价值。

An optimized wavelet neural network based on improved artificial bee colony algorithm for short-term traffic flow prediction

To improve the forecasting accuracy of short-term traffic flow of urban road, and to overcome the disadvantages of slow convergence and easy to fall into local optimum in the prediction process of wavelet neural network, an improved artificial bee colony algorithm, or ABC for short, was proposed to optimize the wavelet neural network prediction model. The adaptive mutation operation in differential evolution algorithm and the selection operator, crossover operator and mutation operator in genetic algorithm were introduced to optimize the traditional ABC, and to improve such disadvantages as slow convergence and weak local search ability in its later period. In this paper, the algorithm was used to optimize the parameters of the wavelet neural network and predict the short term traffic flow. The simulation results show that compared with the existing model, the improved ABC algorithm has less error, higher accuracy, fewer training times, and has higher practical application value.