Research On Harmonic Estimation Algorithms Under Steady State And Faulty Conditions In Power Systems

With the increasing use of power electronic devices and nonlinear loads, harmonic pollution in modern power systems is getting more and more serious, which not only does harm to the primary equipment and secondary equipment, but also leads to negative influences on people’s life and production, causing a great deal of economic loss and deteriorating people’s life standard. Therefore, it is necessary to deal with the harmonics in power systems in order to relieve or even eliminate the damage that power systems and people’s life and production suffer. However, prior to dealing with harmonics, a key step is to accurately estimate the harmonics contained in current or voltage signals in power systems beforehand. After that the harmonics can be dealt with accordingly.In terms of harmonic estimation, researchers at home and abroad primarily conduct related research under the assumption that power systems are in the steady state. However, under faulty conditions, the harmonic contents behave quite differently and accuracy and time consumption for the estimation algorithm are also required differently. Thus, it is recommended that suitable algorithms should be designed to solve the harmonic estimation according to different working states of power systems.In general, this paper considers three scenarios, i.e., online harmonic estimation under steady-state condition, offline harmonic estimation under steady-state condition, online harmonic estimation under faulty condition. As for the first scenario, this paper proposes a Park Transform based harmonic estimation algorithm. By establishing simulation cases in MATLAB and PSCAD/EMTDC, it is validated that this algorithm is more accurate than FFT and IpFFTHW in general. Moreover, the performance of this algorithm is immune to white Gaussian Noise and inter-harmonics and is able to detect deviated frequency. As for the second scenario, this paper improves the conventional Particle Swarm Optimization(PSO) and proposes a harmonic estimation algorithm based on improved PSO and Least Square Method. By validating using the simulation cases established in MATLAB and real-life data, it has been proven that this algorithm is more accurate than some methods in other literatures. Moreover, it is immune to white Gaussian Noise and is able to deal with the situation with frequency deviation, the situation with inter-harmonic and the situation with both. As for the third scenario, this paper proposes an exponentially decaying DC-offset-removal harmonic estimation method based on the second-order differential. Through the validation using the simulation cases established in MATLAB and PSCAD/EMTDC, it has been proven that this algorithm is more accurate and more robust than the conventional Full-cycle Fourier Algorithm, regardless of fault conditions(fault types, fault resistances and fault locations) and the time constant of the DC offset.