Regional power grid economic dispatching optimization method and system with source load uncertainty

Abstract

The invention discloses a regional power grid dispatching optimization method and system with source load uncertainty, and the method comprises the steps: carrying out the scene generation through employing an improved Gaussian fitting model, constructing a scene with wind power, photovoltaic and load double-side uncertainty, combining a quantile regression method, and carrying out the scene reduction according to the dispatching demands of different regions; and considering system operation cost, constructing an interval objective function and constraint conditions, optimizing the objective function to enable the system operation cost to be minimum, taking unit constraint, start-stop constraint, wind and light output constraint, power flow constraint and other constraints as the constraint conditions, and adopting a stochastic programming method to solve the model to obtain a final scheduling optimization strategy.

Description

technical field [0001] The invention belongs to the field of electric power system dispatching, and in particular relates to an optimization method for regional power grid economic dispatching including source load uncertainty. [0002] technical background [0003] With the continuous large-scale development of wind and solar-based new energy sources, the role of demand-side resources in power grid operation is becoming more and more important. The characteristics of the high proportion of new energy in the power system are becoming more and more prominent. Due to the intermittent, random and anti-peaking characteristics of wind points, as well as the uncertainty of load forecasting, the uncertainty of demand-side resources participating in grid scheduling is increasing, which poses a threat to grid security. Stable operation presents great challenges. [0004] The characteristics of a high proportion of new energy in the power system are becoming more prominent and the rol...

AI Technical Summary

Problems solved by technology

It is essentially a two-stage stochastic method. The objective function has a single-layer structure, and the uncertainty of the variables is accurately described by the scene. The more scenes are included, the more accurate the model is, but how to generate accurate scenes and how to analyze the scenes? The reduction will directly affect the subsequent model accuracy

From the perspective of existing scene reduction technologies, scene reduction is performed through probabilistic methods or clustering algorithms with the concept of "message passing" between data points, but scene reduction cannot be performed according to the scheduling needs of different regions, and there is no general reduction technology To meet the scheduling needs of different regions

Images