부산대학교 도서관

Sequential signal processing based on single digital signal processor (DSP) for Coriolis Mass Flowmeter (CMF) can hardly achieve critical real-time performance. A multi-task parallel signal processing scheme based on dual-core DSP for CMF is proposed in this paper. While the sample-based task implements data acquisition and digital filtering, the frame-based task conducts frequency detection and phase difference detection tasks and etc. The two tasks are synchronized via event-based inter-processor communication. Since no real time operating system (RTOS) is needed for task scheduling, the software is of low complexity. Experimental study verifies the higher real-time performance of the proposed scheme over sequential signal processing.