Research On Load Modeling And Non-intrusive Monitoring Of Several Typical Residents Loads

With the development of science and technology and the improvement of people's living standards,the number and types of household appliances and the proportion of residents' electricity consumption to the total electricity consumption of the society are also gradually increasing,so energy conservation has also become an important part of energy conservation.In order to describe the characteristics of various loads better,we need to model the common typical loads mathematically for further study.In addition,study shows that the load side electricity informa tion fed back to the user,the residents will also have a greater energy-saving response on the load side.Household load modeling and identification technology can not only greatly help power companies know the load composition of residents,electricity ha bits,then timely adjust the power supply plan,but also help the power supplier understand household electricity information,from the user side to implement rational use of electricity,promote energy conservation and emission reduction.Firstly,in the aspect of load modeling,this paper summarizes the current situation of traditional load modeling,introduces the ways and methods of load modeling,describes various static and dynamic load model structures according to load characteristics,and summarizes the significance of load modeling and model selection principles;On the basis of parameter identification algorithm,the basic principle of least square method and improved algorithm,steady-state constrained weighted least square method,are expounded,and the mathematical model of common typical load obtained by using this method is given.Different types of loads usually exhibit unique characteristic information due to their different characteristics during transient response such as switching.In ord er to monitor the load transient events,this paper applies a new algorithm based on the sliding bilateral window accumulation sum(CUSUM),which can automatically,accurately and efficiently monitor the occurrence of transient events such as the opening and closing of the load in the household according to the active power consumed during the operation of the load.Secondly,based on the research foundation of load modeling and transient event monitoring,non-invasive load identification is studied,and a non-invasive load identification algorithm based on improved chicken group algorithm is proposed.the normal distribution measurement function considering the steady harmonic current and power characteristics is designed as the fitness function.The experi mental results show that the algorithm can effectively improve the accuracy of load identification and has wide applicability.Finally,the mathematical model of common typical load is established by using experimental data,and it is applied in MATALB simulation software to simulate the household electrical system.the application effect of CUSUM algorithm transient event monitoring and non-invasive load identification algorithm based on improved chicken group algorithm is verified.The results show that the non-invasive load monitoring system can reduce the electricity expenditure,and achieve energy saving and consumption reduction.