An Evolved Prediction Method of Gearbox Life Based on Ant Colony Neural Network
In order to predict the gearbox life quickly and exactly, the artificial neural network with ant colony was introduced. Since there were some disadvantages in Back Propagation (BP) algorithms, such as low convergence speed, easily falling into local minimum point and weak global search capability. For the input of the prediction model, the vibration and temperature information were fuses to rationally utilize all the feature information of gearbox and improve accuracy. At the same time, an ant colony neural network learning algorithm was developed which used to train BP neural network. The simulation model and its results showed that the evolutionary neural network based on genetic algorithm could effectively overcome the shortcoming of falling into local minimum point. This new method could obtain higher prediction accuracy. As a result, it improves the efficient working hours of gearbox and saves production cost in the manufacturing execution system.
information fusion ant colony neural network product life prediction
Xiaohui Chen Junxing Li
The State Key Laboratory of Mechanical Transmission,Chongqing University, China Industrail Engineering Department of School of Mechanical Engineering, Chongqing University, China
国际会议
The Institute Industrial Engineera Asian Conference 2011(2011年国际工业工程师协会亚洲会议)
上海
英文
534-540
2011-06-10（万方平台首次上网日期，不代表论文的发表时间）