神经网络结构优化在肉鸡生长性能预测中的应用

Application of the Neural Network structural optimization in broiler growth performance prediction

动物生长性能预测是养殖业信息技术的核心应用之一。分析季节性因素对肉鸡生长状况的影响,进而建模和预测肉鸡生长性能,可以为大规模肉鸡养殖提供良好的决策支持。基于神经网络模型及其结构优化理论,采用国内著名肉鸡养殖公司提供的案例数据,进行系统的神经网络结构优选试验,得到了较为稳定和有效的预测模型。

Animal growth performance prediction was one of core applications of the information technology in breeding industry. By analyzing the seasonal influence on broiler breeding and making further modeling and prediction, modern poultry breeding companies could get better decision support for large scale poultry raising. Based on the Neural Network and its structural optimization theory and the broiler growth dataset provided by the most famous poultry raising company in China, this paper deployed systematic experiments on the structural optimization of Neural Network, and achieved a relatively stable and effective prediction model.