Research On The Measurement Of Power Harmonic Based On DSP

As the development of power electronic technology and the massive application of non-linear components, the harmonic pollution of power system has become a serious problem, which causes great damages to electric equipments. So the correct measurement of harmonics is the first step towards the inhibition of damaging effect. It provides a scientific evidence for power department to the management of electric network and is of far-reaching significance to the solution of harmonic pollution. The main work of this project includes researching an effective algorithm of harmonic measurement, designing the prototype based on DSP and measuring the magnitude and phase of 1-50 harmonics.The dissertation adopts windows and interpolation algorithm based on FFT after analysing the current situation and developing trend of harmonic measurement at home and abroad. Then some improvements are made on this algorithm and a tri-spectrum interpolation FFT algorithm based on nuttall window is put forward. The corresponding program of this algorithm is written and the hardware platform is designed based on TMS320F2812. Then this algorithm is implemented on DSP and some experiments are done to validate it.First, after researching on the theory of Fast Fourier Transform, the spectral leakage during FFT is the key point of causing measurement errors. The usual way to solve this problem is to use the window and interpolation algorithm. The simulation experiment using MATLAB shows that the revised algorithm can greatly improve the accuracy.Secondly, the system software is designed. The whole process can be summarized as follows: first, the harmonic is transformed into square wave through voltage comparator so that the frequency can be measured by the capture module of 2812. Then the sampling frequency can be determined according to the fundamental frequency. When the sampling point reaches 1024, each harmonic parameter can be obtained by the algorithm.Finally, the fixed-point DSP chip TMS320F2812 are chosen as the core processor and the minimum system and the front signal processing circuit are designed. Some experiments are done and the possible solutions are proposed after analysing the error sources. The experiment shows that this algorithm is very effective, the softwarel and hardware function normally and this module can meet the design requirements.