摘要
复杂机械产品运行状态预测分析往往准确率较低、推理时间较长、难于推理,难以获得有效结果。为此,给出了基于改进可拓神经网络的复杂机械产品运行状态预测分析模型。提出了改进的可拓距,基于该可拓距构建可拓神经元,建立复杂机械产品运行状态分析经典域模型,并对运行数据进行训练,形成复杂机械产品运行状态可拓神经网络预测分析模型。通过具体案例对算法和模型进行验证,并对比BP神经网络,结果表明了模型与算法的有效性和可行性。
The prediction and analysis of the operation status of complex mechanical products often have problems such as low accuracy,long reasoning time,and difficulty in reasoning,which make it difficult to obtain effective results for the prediction of the operation status of complex mechanical products.Therefore,the method for predicting and analyzing the operating state of complex mechanical products was discussed,and the model for predicting and analyzing the operating state of complex mechanical products based on improved extension neural network was proposed.First,an improved extension distance was proposed,which is used to construct extension neurons.Subsequently,the classic domain model for the analysis of the operating state of complex mechanical products was established,the operating data was trained,and the extensional neural network predictive analysis model for the operating state of complex mechanical products was proposed.Finally,a specific case was used to verify the algorithm and model,and the BP neural network was used to compare.The results show that the model and algorithm are effective and feasible.
作者
王体春
方磊磊
童昌圣
WANG Tichun;FANG Leilei;TONG Changsheng(College of Mechanical and Electrical Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)
出处
《重庆理工大学学报（自然科学）》
CAS
北大核心
2020年第12期96-103,共8页
Journal of Chongqing University of Technology：Natural Science
基金
国家自然科学基金项目(51775272,51005114)
中国国家留学基金项目(201906835046)。
关键词
预测分析模型
可拓神经网络
运行状态
复杂机械产品
predictive analysis model
extension neural network
operating state
complex mechanical products
