摘要
由于受到温度、磁场等外界因素的干扰,汽车传感器的测量精度降低,致使汽车的整体性能下降。为此,以对温度干扰最敏感的CYJ-101型压力传感器为例,采用18组样本数据对建好的3层前馈BP神经网络进行温度补偿训练。仿真结果表明,温度对压力传感器的干扰波动由补偿前的22%减小到补偿后的2.2%。BP神经网络技术的应用极大地提高了压力传感器的测量精度,并最终改进了汽车的整体性能。作为一种分析、处理温度补偿问题的新技术,它与传统方法相比具有无可比拟的优势。
Owing to the interference by the temperature and magnetic field, the measurable precision of automobile sensor drops greatly, and then results in the decline of the whole function. This paper performs the temperature compensation train to the built 3-floor feed-forward BP neural network based on the 18 groups of sample datum for with the instance of the CYJ-101 pressure sensor which is the most sensible to the temperature interference. The simulated result has shown that as to the interference fluctuation of the pressure sensor, the temperature decreases from 22% to 2.2%. The application of BP neural network technique has greatly increased the measurable precision of the pressure sensor, and ultimately improved the whole function. As a new technique which analyses and deals with the problem of temperature compensation, compared with the traditional methods, it has incomparable advantage.
出处
《农机化研究》
北大核心
2008年第1期71-73,共3页
Journal of Agricultural Mechanization Research
关键词
汽车工程
汽车传感器
理论研究
BP神经网络模型
温度干扰
测量精度
car engineering
automobile sensor
theoretical research
BP neural network model
temperature interference
temperature compensation