A Parallel Signal Processing Method Of Optical Current Transformer Based On FPGA

With the introduction and development of the smart grid, digital substation requires transformer to have a digital output, and with intelligent features. The optical current transformer based on Faraday magneto-optical effect has excellent insulating properties and linearized dynamic response capability that is independent of the current size and waveform. It was adapted to the requirements of digital substation. However, in the measurement of small current, optical current transformer photoelectric signal is smaller than the internal noise signal and real-time requirements of the optical current transformer is high. In addition, under different working conditions, OCT measurement accuracy affected by temperature is relatively obvious.So the noise and signal in OCT should be seperated fastly to meet the measurement accuracy and improve real-time ability.This article put a parallel signal processing method of optical current transformer based on FPGA as the research object. We conducted a series of work such as mathematical modeling, rapid Kalman filter algorithm programming and FPGA Implementation, temperature-current acquisition system creation and temperature compensation to current online.It introduces the measuring principle of the optical current transformer, and outlines the OCT structure and the noise characteristic of this paper. By spectral characteristics of noise, the OCT output signal to noise characteristics are analyzed.In order to meet the optical current transformer signal to noise characteristics and real-time requirements, this paper presents a fast Kalman filter algorithm adapted to the FPGA implementation. In the currents of AC, DC, AC-DC three cases, We have established an appropriate rapid Kalman filter dynamic spatial models. LabVIEW simulation verifies rapid Kalman filter’s effectiveness in three cases. The results showed that rapid Kalman filter can effectively separate current signal to noise, and improve signal to noise ratio.Based on LabVIEW FPGA hardware, the fast Kalman filter algorithm implemented in parallel on the hardware platform. The key researches include offline calculation algorithm module, data access module, states after-estimates module and filter results processing module. By using IP core optimization techniques, we improved algorithm FPGA implementation structure, so that each module can run in parallel.It improves the speed of while the filtering accuracy is ensured,and it reduces the usage of resources.Finally,I designed the temperature-current collection experimental system. The system uses real-time signal processing platform based on FPGA. It collects and processes current data for each temperature,then we can get the temperature-current curve. Which verifies the validity of algorithm based on FPGA and is laying the foundation for the work of OCT current compensation online.