摘要
介绍了一种具有DeviceNet现场总线通信功能的智能差压传感器的设计,探讨了传感器输入输出特性曲线的非线性校正方法。通过内嵌的总线控制器(SJA1000)、报文收发器(82C251)和P89C668单片机等,该传感器可直接作为一个DeviceNet从节点工作;对传感器的输入输出特性曲线进行了建模,以软件手段实现高精度的非线性自校正功能。测试结果表明:该智能差压传感器不但具有DeviceNet现场总线输出功能,且经过多项式或神经网络建模的传感器非线性误差分别可达0.04%FS和0.02%FS。这为高精度传感器的制作提供了一种可行的方法。
Design of a smart differential pressure sensor system with communication function, of DeviceNet on-site bus is presented,and the non-linearity rectification methods of the sensor＇ s input-output characteristic curve are discussed. It can work as a DeviceNet slave node with embedded bus controller（ SJA1000）, message transceiver （82C251） and MCU（P89C668）. The input-output characteristic curve is modeled to realize high precision nonlinearity self-rectification function by software method. The test results show that this smart differential pressure sensor not only has communication function of DeviceNet on-site bus, but also can reach non-linearity error of 0.04 % FS and 0.02 % FS through polynomial and neural network curve fitting. So a feasible method to produce high precision sensors is gained,
出处
《传感器技术》
CSCD
北大核心
2005年第12期74-76,88,共4页
Journal of Transducer Technology
基金
上海市科委计划资助项目(03dz1101
04515120)
