基于神经网络的拨禾导向装置可靠性预测研究

The Research about Reliability Prediction of Raking and Conveying Device Based on Neural Network

针对玉米收获过程中拨禾导向装置的高故障率问题,在分析玉米收获机拨禾导向装置工作原理的基础上,运用MATLAB神经网络工具箱建立拨禾导向装置可靠性预测的BP神经网络模型。根据不同型号玉米收获机拨禾导向装置的可靠性数据,进行网络设计和训练。选取传动系统、果穗输送器、左拨禾链、右拨禾链、摘穗辊组、张紧装置、机架的可靠度为输入变量,该装置总体可靠度为输出变量进行预测分析。结果表明,预测误差较小,能够有效预测其可靠性,为玉米收获机拨禾导向装置的优化设计及参数选择提供了依据。

Aimed to the problem of high failure rate appearing on the raking and conveying device during the process of harvesting corn .We analyzed the working principle of it and then established a BP neural network model to predict the reliability of this device on the basis of using the MATLAB neural network toolbox .The neural network was designed and trained according to the device reliability data from different type corn harvest .Selected these reliability data which were from transmission system , scraper conveyer , left raking chain , right raking chain , snapping roll , tightener and frame and used it as the input variables then used the all device ’ s reliability data as the output variable .Predicted results showed that the neural network with small prediction error can forecast the reliability of raking and conveying device effectively . Moreover , it provided a foundation for the optimization design and parameter selection of this device .