FPGA-based neural network harmonic estimation for continuous monitoring of the power line in industrial applications

Manufacturing cells are present in almost all the industrial sector. Unfortunately, all machine tools into the manufacturing cell are connected to the same power line, implying that their operation adds nonlinear loads such as harmonics and interharmonics that affect the general machine-tool condition. A novel NN-based methodology for harmonic monitoring through time in transient and stationary signals that satisfies the IEC61000-4-7 standard is presented, as well as its implementation into a field programmable gate array (FPGA) for continuous and online monitoring. The proposed method and the developed instrument have been tested in a real manufacturing cell.

► NN-based methodology which satisfies the IEC61000-4-7 standard is presented.
► Signal analysis for transient and steady state is done by an effective way.
► Time resolution is improved without losing the standard requirements.
► FPGA implementation allows the continuous and online monitoring.
► Developed instrument performs in situ analysis as required by new generation machines.