FPGA based Monitoring Platform for Condition Monitoring in Cylindrical Grinding

Abstract Sensor systems for machine condition monitoring face many challenges in the world of digitized data with an emphasis being placed on the application of high performance feature extraction computing systems. These systems must be robust, reliable and economically viable before being adopted by industry. This paper describes a platform for condition monitoring that has been developed using a field programmable gate array (FPGA) for the real time signal processing in a complex grinding process. An architecture which can sample 16 channels at 12.5 kHz and perform 1024 bin FFTs on a National Instruments CompactRIO with a Xilinx Virtex 5 LX50 is described. FPGA resource utilization figures for multiple configurations of this FFT are reported. The FPGA also performs an exponentially time weighted RMS on 10 acceleration channels and samples four quadrature encoded axis channels in real time. Results are displayed as conditioned data to a machine operator HMI for machine and process evaluation. The research demonstrates how multisensory multiprocessing platform approach can be realized and implemented in industry for future high end condition and process monitoring applications.