Study On Probabiistic Assessment Methods Of Harmonic Resonance In Power Systems

With the rapid development of(power)electronic and information technology in recent years,the massive access of new energy systems and(power)electronic equipment has significantly changed the nature of harmonic problems in power systems.Their randomn ess and distributed characteristics and the continuous expansion of the scale of power systems have brought new challenges to the study of harmonic propagation and harmonic resonance.On the basis of the existing harmonic propagation and resonance analysis methods,new methods are urgently needed to solve the problems that the existing analysis methods are complicated in calculation,can only obtain qualitative analysis results,and do not consider system uncertainties.To this end,this thesis conducts theoretical research on the harmonic resonance problem of power systems,focusing on solving a series of theoretical problems such as rapid calculation of harmonic resonance modal analysis,quantitative assessment of impact area and severity,and probability assessment.The main work of the thesis can be summarized as follows.(1)By comparing the existing harmonic resonance analysis methods,the characteristics and limitations of each method are summarized,and problems that need to be further solved in harmonic resonance analysis are pointed out.(2)A fast resonance mode calculation method based on improved power iteration is studied.As the traditional modal analysis method based on eigenvalue decomposition calculation has high computational space and time complexity,this thesis uses the power iteration method to calculate the eigenvalues of the network impedance matrix and improve it.According to the characteristics of the frequency scanning process of the resonance mo de analysis method,a new power iteration initial vector selection method is proposed.According to the alternating characteristics of “peak” and “valley” of the key mode resonance curve of the power system,a new iteration termination criterion is proposed.According to the test,it is found that when a larger frequency interval is used as the calculation step size,the resonance frequency can still be determined more accurately.Based on this,a "two calculation steps" analysis method is proposed.A larger step size is firstly used to find the approximate resonance frequency,and then a smaller step is used for more detailed analysis near these approximate resonance frequencies.This can further increase the calculation speed of harmonic resonance mode analysis.This method is tested on IEEE14 bus system and IEEE 30 bus system,respectively.The results show that the method can improve the calculation speed by more than 90%,and the improvement is more obvious for the system with larger node size.(3)A quantitative assessment method of harmonic resonance is studied.Aiming at the limitation of the resonance mode analysis method,the resonances at the harmonic source bus and the non-harmonic source bus are distinguished,that is,local resonance and nonlocal resonance,and specific definitions are given.Based on the harmonic electrical distance,a set of indices is derived for the quantitative analysis of harmonic propagation and resonance amplification.In addition to a local sensitivity analysis,a global sensitivity method is used to further analyze the uncertain parameters of the system,and three types of sensitivity parameters are distinguished and defined.Compared with the existing harmonic resonance analysis methods,the main advantage of this set of indices is that it can distinguish between local resonance condition and non-local resonance condition,quantify the severity of resonance amplification,and perform more types of sensitivity analysis.(4)A probabilistic harmonic resonance mode assessment method with considering the power system uncertainties is studied.The uncertainties existing in the power system are classified and modeled,and a probabilistic harmonic resonance mode assessment method is established based on the resonance mode analysis method and Monte Carlo simulation method.The system-level rmodal frequency band,the probabilistic expression of the modal impedance and the sensitivity information of the bus level and the component level,provide a theoretical basis for the in-depth study of the random behaviors of the harmonic resonance in uncertain power systems.The potential application of this method is discussed and an application example of the probabilistic resonance frequency shift technique is given.In addition,the improved power iteration method is embedded in the Monte Carlo simulation process,which effectively reduces the computational space complexity and time complexity.(5)A probabilistic assessment method of harmonic propagation and resonance amplification severity considering uncertainties is further studied to reveal the randomness of harmonic propagation and resonance amplification in power systems.Based on the proposed quantitative evaluation method of harmonic resonance,the uncertainties of power system are further divided into two types,including instantaneous uncertainty and epistemic uncertainty.The instantaneous randomness and cognitive randomness of harmonic propagation and resonance amplification in power systems are studied respectively.The randomness of harmonic propagation and amplification under different load models is compared and analyzed,and the high sensitivity of the harmonic response to the load modeling method is revealed.Extensive analysis and simulation studies have been performed in this thesis to illustrate and verify the above methods.The results confirm that,compared with the existing methods,these proposed methods have obvious advantages in analyzing harmonic propagation and resonance amplification problems in power systems.The research in this thesis has effectively promoted the rapid calculation of harmonic resonance analysis,progress from qualitative analysis to quantitative analysis,and deterministic analysis to probability analysis.