基于FPGA的多组分气体监测装置设计

Design of multi-component gas monitoring device based on FPGA

气体传感器往往存在交叉敏感性,传统复合气体检测的精度低且耗时长。设计了一种基于现场可编程门阵列(FPGA)的多组分气体监测装置,以气体传感器阵列为基础,在FPGA上设计硬件神经网络控制器,对多传感器数据进行融合。通过分段拟合及流水线的设计思路,在FPGA上实现了硬件Sigmoid函数。仿真及应用表明:设计的装置具有检测精度高、响应快的优势及良好的扩展性。

Gas sensors always have cross sensitivity,and the precision of traditional compound gas detection is low and time consuming is long. A multi component gas monitoring device based on freld programmable gate array（ FPGA） is designed,which is based on the gas sensor array. A hardware neural network controller is designed on the FPGA to fuse the multi-sensor data. Among them,the Sigmoid function is an important part of the hardware to realize the neural network,and the hardware Sigmoid function is implemented on the FPGA through the method of subsection fitting and pipeline design. Simulation and application show that the device has the advantages of high detection precision and fast response,and has good scalability.