基于神经网络的多功能收割机发动机性能仿真及优化

摘要

With the rapid development of the global economy and industry, energy crisis and environmental protection has become more and more prominent, the traditional internal combustion engine has been a tremendous impact.Therefore, the research of high efficiency and energy saving of the engine is very important.This paper presents a compression ratio of 10.6 multifunctional harvester, designs a GT-Power neural network model based on MATLAB platform, automatic simulation and storage by using the neural network training and testing data, using the Latin hypercube sampling algorithm design, simplify the operation process and improve the searching efficiency.The experimental results show that the neural network model, the torque is small, fuel consumption ratio parameters and temperature model predictive error precision is high, which can be used to predict the performance of multifunctional harvester engine, the parameters optimization.%随着全球经济和工业的快速发展,能源危机和环境保护问题越来越突出,传统内燃式发动机受到了巨大冲击,因此研究高效、节能的发动机显得尤为重要.为此,研究一种压缩比为10.6多功能收割机,并设计了基于MatLab仿真平台的GT-Power神经网络模型,利用神经网络训练和测试的数据自动进行仿真和储存,采用拉丁超立方采样算法设计试验,简化运算过程提高寻优效率.实验结果表明:神经网络模型转矩、比油耗和温度等参数模型预测误差很小、精度很高,可用于多功能收割机发动机的性能预测,使其各项指标参数最优化.