基于C# .NET与Matlab接口和BP网络的汽车产量预测  

Vehicle Production Forecasting Based on BPNN and Interface Between C# .NET and Matlab

作　　者：黄一丹[1] 严洪森[1] 冯丽娟[1] 吴奇[1] 

机构地区：[1]东南大学复杂工程系统测量与控制教育部重点实验室,江苏南京210096

出　　处：《计算机技术与发展》2008年第11期36-40,共5页Computer Technology and Development

基　　金：国家863计划资助项目(2007AA04Z112);国家自然科学基金资助项目(60574062)

摘　　要：为了提高软件预测系统的开发效率和精度,研究了误差反向传播神经网络(Error Back Propagation Neural Network,BPNN)在Matlab 2006a环境下的实现方法,采用迭代法选择神经网络训练参数。探讨了Matlab与C# .NET开发平台的接口技术,能成功实现C# .NET对Matlab神经网络工具箱的调用。以汽车产量预测为应用案例,开发了基于C# .NET和Matlab接口的神经网络预测系统。实验结果表明,该预测系统对汽车产量预测有着良好性能,接口技术的应用提高了软件开发效率。In order to improve the efficiency and accuracy of software forecasting system, researched into the implementing method of BPNN （Error Back Propagation Neural Network） in Matlab 2006a. The training parameters of BPNN were chosen by iterative method. The interface technology between C#. NET and Matlab was discussed and then Matlab NN toolbox could be called by C#. NET successfully. A BPNN forecasting system based on the interface between C#. NET and Matlab was developed for vehicle production forecasting. Experiment results indicate that this system can forecast the vehicle production accurately and the application of interface technology can improve the software development efficiently,

关 键 词：人工神经网络 接口技术 预测 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]