基于DBN-PID的木材干燥窑参数检测系统  

Detecting System of Wood Dry Kilns' Parameters Based on DBN-PID

作　　者：夏春江 王培良[1,2] 

机构地区：[1]杭州电子科技大学新型电子器件研究所,杭州310018 [2]湖州师范学院信息与控制技术研究所,浙江湖州313000

出　　处：《计算机测量与控制》2015年第1期99-101,105,共4页Computer Measurement &Control

基　　金：浙江省自然科学基金资助项目(LY12F03008)

摘　　要：针对目前在木材干燥过程中检测参数精度低的问题,设计了一种木材干燥窑参数检测系统;该系统对干燥窑有关参数进行实时采集、调理,经ARM处理器进行数据处理后,将采集的数据通过WiFi无线网络上传至PC机;为改进系统的控制算法,引入深度学习方法,提出了一种基于DBN-PID的控制算法,并与BP-PID算法进行了实验比较;实验结果表明,DBN-PID控制算法应用在木材干燥窑参数检测系统中具有更高的检测精度;为进一步说明DBN-PID算法的性能,还与BP-PID算法进行了仿真比较;仿真表明,DBN-PID算法能够很好的近似非线性对象,具有较强的自适应能力。For the problem of the low measured precision of the parameters in the process of the wood drying,this paper design a detecting system of wood dry kilns＇ parameters.The system will capture the parameters of wood dry kilns real time,regulate and process the data with ARM microprocessor,and then send the data to the PC with the WiFi.To improve the control algorithm of system,this paper introduces the deep learning method,and put forward a control algorithm of DBN-PID.In the same hardware platform,compared with the algorithm of BP-PID.The experimental results show that the detecting system of wood dry kilns with DBN-PID has the more higher precision.For further clarification about the performance of DBN-PID,compared with the algorithm of BP-PID in the simulation.The results of simulation show that DBN-PID can approach the nonlinear object more better,and has the more stronger adaptive ability.

关 键 词：深度学习 深度信念网络 木材干燥窑参数 嵌入式系统 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]