Study On The Assessment Technologies Of Collective Harmonic Impacts Of The Residential Customers

Power quality is one of the key issues of the safe operation of the active distribution network(ADN),and harmonic is the most critical problem.The harmonic generated by an individual residential load is usually ignored because of its small amount.But with widespread applications of the power electronic based residential appliances,the electric vehicles(EVs)and the photo-voltaic(PV)generators,the harmonics generated by those distributed residential loads inject into the power distribution networks.The harmonic emission level of the residential customers is increasing.The impacts of their collective harmonics become significant and cannot be negelected.With the electricity market reform and the active distribution network development,it is urgent to study the assessment techniques of the collective harmonic impacts of the distributed residential loads.However,the study on the theory and techniques in this topic is far from enough.In this thesis,it studies the harmonic emission characteristics,the collective harmonic modeling and assessment techniques,which has significance in theory and industrial applications.For the distributed harmonic sources with mass quantity and individual small amount,it studies the distribution of the harmonic sources based on the analysis of the grid topology structure.It studies the emission characteristics of distributed harmonic sources,and conducts experimental study on the typical residential appliances,PVs and EVs.The results show that,those residential harmonics have the emission characteristics of distribution,stochasticity and nonindependence,which are different from traditional industrial harmonics.For the low accuracy of modeling of distributed harmonic sources with the traditional constant current source,it studies the harmonic modeling methods based on the crossed admittance matrix and Norton equivalent circuit,respectively.Both harmonic models of each typical residential appliance are built through experimental tests and simulation.Comparison results show that,the Norton equivalent circuit model has better accuracy than the constant current model and lower complexity than the crossed admittance matrix model,with high practicability.In order to describe the impacts of the customer habits on distributed harmonics,it studies the behavior modeling of typical residential loads based on the Markov Monte Carlo method.With consideration of several impact factors on the use of residential appliances,it builds two probability fuctions of the household population and daily time of use profile,and obtains the probability of appliance switch-on.The behavior modeling of a typical PV system is built up with the irradiation intensity.In addition,the behavior modeling of a typical EV is built up with consideration of the start time and duration of charging.Simulation results approve the effectiveness of those methods.For the problem that the collective impacts of distributed harmonic sources cannot be analyzed with deterministic analysis methods,it proposes a bottom-up based assessment system for the collective harmonic impacts.An assessment index system is presented that consists of the harmonic current contribution rate,harmonic maximum probability value,and probability density distribution of harmonic current amplitude.The impedance gathering trend method is adopted to study the harmonic contribution of the system side and customer side.Finally,it uses the cloud model to describe the probability distribution of the collective residential harmonics,and adopts the cloud transform to fit the probability distribution curves.The case study proves the effectiveness of the proposed techniques,and the study results can provide reference to identify the harmonic sources and to determine the supervision policy.