Research On Harmonic Detection Instrument Based On TMS320F2812

With the continuous development of modern industry, the share of non-linear load in the grid is increasing constantly, the harmonic problem have become severer and severer and have turned into a common concern of electric power enterprises and users. The real-time online detection of harmonic is significant to improve power quality and reduce the harm of harmonic. A large amount of electric railway has been put into use in China. At the meantime, it brought new challenges to the harmonic detection and control due to its offensive load and wide range.This paper first applies the FFT, wavelet transform and wavelet package transform to detect the stationary harmonic signal, the non-stationary harmonic signal, the transient harmonic signal and the harmonic signal containing noises through MATLAB simulation. The result shows that wavelet package transform is the core method in harmonic detectors for electric railway.Monitoring network constituted by several harmonic detectors is the precondition to evaluate and control the harmonic of electric railway effectively. Considering the features of electric railway control system, the harmonic detector is designed based on a hardware structure of 3-CPU. This harmonic detector is able to complete synchronous sampling of the three-phase voltage and three-phase current. Meanwhile, it shows great computing power and has the function of real-time data processing and measurement results transmission. Among the three CPUs, DSP TMS320F2812 is used as master CPU to capture and handle power signals, calculate the harmonic parameter. LPC2478 is used as slave CPU to implement the human machine interface. Neuron chip 3150 is responsible for network communications. TMS320f2812 and LPC2478 exchange the data via DARAM, TMS320f2812 and 3150 communicate via RS-232.The software is mainly programmed with C language. It achieves functions of DB24 wavelet package method, capture data, handle data, human machine interface and network communication based on DSP platform. Handling data includes calculation of harmonic components and basic power parameters.By experimental test, we inspect the accuracy of basic electrical parameters of the harmonic detector, test the analysis precision of the core method——the wavelet packet transform, and test the capacity of network communication.