为了科学合理地预测南疆兵团各师的货运量,以2006~2010年南疆兵团各师货运量数据为基准,基于BP神经网络算法,建立相应的预测模型,并采用Mariab编程求解,得到相应的南疆兵团各师货运量的预测值,通过与实际值比较分析,发现相对误差位于0.8%~7.8%之间,平均相对误差约为4.45%,计算结果符合精度要求,所得结果具有较好的说服力和可信度。
In order to give a scientific prediction of the freight volume of each division in South Xinjiang Corps, tak-ing the freight volume data from 2006 to 2010 of each divi- sion in South Xinjiang Corps as a reference, based on the BP neural network algorithm, the corresponding prediction model was established, and to solve the model by use of Matlab programming. The freight vohune prediction of each division in South Xinjiang Corps was obtained. By compared with the ac- tual values, we found the relative error is between 0.8% to 7.8%, the average relative en'or is about 4.45%, the calcula- tion results meet the accuracy requirements, the results have better persuasion and reliability.