论文部分内容阅读
近年来,由于猪肉安全事故和供求关系等因素的影响,导致猪肉市场价格波动大,养猪业难以持续稳定发展。科学预测猪肉需求量,对科学指导生猪生产和宏观调控猪肉市场价格意义重大。笔者根据历年数据,采用BP神经网络预测方法,实现在MATLAB中运行,通过对模型的多次训练,选择隐层神经元数目为6个,达到了期望效果。预测结果表明,2012~2016年,我国人均猪肉需求量分别为34.10、38.12、38.66、40.28和40.60kg/a。
In recent years, due to the impact of factors such as pork safety accidents and supply and demand, the price of pork market fluctuates greatly and it is difficult for the pig-raising industry to maintain a sustained and stable development. Scientific prediction of pork demand, the scientific guidance of pig production and macro-control of pork market price is significant. Based on historical data, the author uses BP neural network prediction method to run in MATLAB. Through several trainings on the model, the number of hidden neurons is selected to be 6, which achieves the desired effect. The forecast results show that from 2012 to 2016, the per capita pork demand in China was 34.10, 38.12, 38.66, 40.28 and 40.60kg / a respectively.