Online load modeling parallel computing method based on electric energy quality monitoring system

Abstract

The invention relates to a power system load modeling parallel computing method based on an electric energy quality monitoring system. According to the method, the electric energy quality monitoring system acquires the disturbance data of a power grid, and an improved clone selection parallel computing algorithm is used for performing parallel computing identification processing on a model; and the fitting condition of the output power of the model with actually-measured load power is checked under different failure conditions, and a Mifare (MI) card and a fitting curve are output according to a Bonneville power administration (BPA) motor model. The electric energy quality monitoring system is used for acquiring the data, and the improved clone selection parallel computing algorithm is used for identifying load model parameters, so that the method has the characteristics of high identification accuracy, global convergence, high running speed and fitting real-time performance, and is applied to practical engineering application; high scalability and a high speed-up ratio are realized according to actual needs; the timeliness and usability of a load modeling process are improved; and the running efficiency of a software platform for load modeling operation under complex conditions is improved to a great extent.

Description

technical field [0001] The invention relates to a calculation method, in particular to a parallel calculation method for online load modeling of a power system based on a power quality monitoring system. Background technique [0002] The basis of digital simulation is to establish a mathematical model of each component of the power system - modeling for short, and its accuracy directly affects the error of the simulation result. At present, the models of generating units and transmission network are relatively mature. In contrast, the development of load modeling is relatively slow, and the load model has become one of the key factors restricting the calculation accuracy. Load modeling is a very complicated problem, because the load of the power system is composed of many different electrical equipment, and there are many types; and the load composition and load change randomly with time; and Lack of precise data on load composition; and load uncertainty and nonlinearity, w...

AI Technical Summary

Problems solved by technology

However, the application of this model requires many parameters to be identified, with multiple peaks, slight differences in extreme points, and complex solution space.

The most commonly used methods for solving parameter identification include traditional optimization algorithms such as the least square method and artificial intelligence algorithms represented by genetic algorithms. Traditional optimization algorithms are greatly affected by the initial value and have difficulty in converging

The genetic algorithm has poor global convergence, is prone to problems such as individual degradation, and takes a long time to calculate

Images