Design and Hardware Implementation of a Filtering Algorithm for a Fast-Response MEMS Thermal Gas Flowmeter

Thermal gas flowmeters are widely employed in energy metering, industrial control systems, and other fields. However, the gas flow process is characterized by turbulent pulsation, pipe vibration, and other nonstationary noise interference problems. This article presents an integrated filtering algorithm and hardware implementation scheme for thermal gas flow sensors. A miniaturized platinum (Pt)-resistance thermal sensor was fabricated using microelectromechanical system (MEMS) micromachining technology and integrated into a flow measurement device combining tube-bundle structures and bypass channels to enhance the interference resistance. Meanwhile, a robust strong-tracking Kalman filter (KF) algorithm with rapid convergence was developed by systematically analyzing innovation vectors and noise characteristics. The hardware implementation efficiency was improved by optimizing the algorithm computation flow and reusing the computation modules. Finally, the hardware implementation was accomplished by Xilinx ARTIX-7 series XC7A35T field-programmable gate array (FPGA). In comparison with the conventional KF, the enhanced algorithm exhibits superior theoretical performance in terms of filtering accuracy and convergence speed, and the dynamic response time was reduced by 83.3%. Under 50-MHz operation, each iteration requires 61 clock cycles, with a computational latency of $1.22 \; \mu $ s and an energy efficiency of 63.07 nJ/op.