Research On Emergency Load Shedding Of Multi-infeed HVDC Receiving Power System

The long-distance transmission of massive electrical power is an important feature of China’s power grid and even the global energy interconnection in the future.The continuous construction of power system improves the complexity of grid structure.Large scale wind power,photovoltaic and other clean energy units integrated with the grid reduce the inertia level of the system.The occurrence probability of hidden failure and cascading failure in power grid increases.The disconnection of UHV tie-line will lead to a large amount of power shortage in the receiving power grid,which will lead to the deterioration of frequency,voltage and other transient indicators.It is necessary to configure emergency control scheme for faults to prevent the triggering of the last defense line of the system,so as to ensure the safe and stable operation of the power grid.This paper mainly studies the emergency load control of multi-infeed HVDC receiving power system.The specific researches are as follows:A grid model of multi-infeed HVDC receiving power system is built.The feasibility of the grid model as the research object is verified.An emergency load shedding optimization model considering various transient index constraints is proposed.The emergency control scheme is evaluated by the degree of constraint violation.An emergency load shedding optimization method based on particle swarm optimization is proposed.The optimization ability of the algorithm is improved through the improvement of the initial population generation method and the adaptive adjustment of the motion parameters.The iterative formula is improved to make the optimization scheme suitable for project,and the time-consuming of the algorithm is reduced by parallel computing technology.The improvement effect of the algorithm and the optimization effect of the emergency load shedding scheme are verified.A fast optimization method of load shedding based on parameter fitting model is proposed.Radial basis function neural network is used to fit the functional relationship between the transient parameters and the control vector.After the implementation of the emergency control scheme,the transient index of the system is predicted to evaluate the control scheme.Considering the expensive computation of time-domain simulation,an emergency load shedding optimization method based on surrogate-assisted differential evolution is proposed.The characteristics of different optimization methods for load shedding are studied.The load shedding distribution of each load bus under different transient constraints and the possibility of obtaining the global optimal solution are analyzed.Clustering analysis is carried out on the parameters which are highly correlated with the load shedding scheme to extract the typical operation characteristics of the system.The data dimension of high-dimensional operation characteristics is reduced,and the coordinate system of key parameters used to represent the operation state of power grid is constructed.The typical operation scenarios of power grid are extracted.An emergency load shedding optimization method considering multi-operation modes of the system is proposed,and an emergency load shedding scheme suitable for a certain time scale is obtained.The validity of the proposed method is verified.