Abstract: Through applying an improved BP(back propagation)algorithm to city consume，a BP model is built．At the same time，a assessment system of city consume based on ANN is designed，and the results of application show that the algorithm is of high speed for convergence and of high accuracy for prediction and is an efective predictive method．

Key words: BP neural network, BP algorithm, city consume, assessment

摘要： 针对城市消费预测问题，应用一种改进的BP算法，建立了相应的BP网络模型，并设计了基于BP神经网络的 城市消费预测系统．应用实例证明：该算法可以提高当前的预测精度，是一种较为有效的预测方法．

关键词: 人工神经网络, BP算法, 城市消费, 评价