摘要
为了提高恒温箱温度控制的精度,采用STM 32F 103系列芯片作为主控制器,结合恒温箱箱内情况和高精度的控制要求,设计实现了温度测量和温度控制等环节,控制系统软件部分采用嵌入式操作系统μC/OSII基于C语言在Keil平台实现。将由外界环境温度经BP神经网络预测,并计算得到的温度变化率作为模糊控制的其中一维输入,使控制器具有超前控制的作用。经计算可知,预测温度的绝对误差的最大值为0.23720℃,平均值为0.10271℃,均方差为0.06396℃。从预测结果来看,箱内温度的预测可以通过BP神经网络预测模型来实现,有效解决了恒温箱温度控制的滞后性问题,对电气设备温控设计领域起到一定的理论支撑作用。
In order to improve thermostat temperature control accuracy,temperature measurement,temperature control and other elements were designed and realized by using STM32F103 series chips as main controller under consideration of the conditions within the thermostat as well as the requirement for high-precision control.The control system software was implemented on the Keil platform with embedded operating systemμC/OSII based on C language.Temperature change rate predicted and calculated by BP neural network from ambient temperature was taken as one of the inputs for fuzzy control,so that the controller could acquire the function of advance control.As seen from estimation results,the maximum absolute error of the predicted temperature was 0.2372℃,with an average of 0.10271℃and mean square error of 0.06396℃.As seen from the prediction result,the temperature within the thermostat could be predicted through the BP neural network prediction model.The proposed approach effectively solved the hysteresis of thermostat temperature control and could provide a certain theoretical support in the field of temperature control design for electrical equipment.
作者
苏力
雷瑛
Su Li;Lei Ying(College of Electronic Engineering, Xi’an University of Aeronautics, Xi’an Shaanxi 710077, China;College of Mechanical Engineering, Xi’an University of Transportation Engineering, Xi’an Shaanxi 710300, China)
出处
《电气自动化》
2021年第2期35-37,共3页
Electrical Automation
基金
航空科学基金资助项目(2019 ZH0T 7001)。
关键词
温度测控
系统设计
BP神经网络
预测分析
模型
恒温箱
控制器
误差
temperature measurement and control
system design
BP neural network
predictive analysis
model
thermostat
controller
error
