Virtual power plant energy storage optimization scheduling strategy based on model predictive control

Abstract

The invention discloses a virtual power plant energy storage optimization scheduling strategy based on model prediction control. The strategy comprises the following steps: predicting short-term load and new energy output, which is to realize that employ historical data of a certain region 1 a is used to train a network, wherein the time interval between the data is 5 min, so as to enable the network to learn internal rules of a time sequence, wherein the data is normalized before use, the prediction time step length is 24 h, the time resolution is 15 min, and a random forest algorithm is used to predict the load, wind power and photoelectric output of 1 d in advance; establishing a model. The method is advantageous in that: the random forest algorithm is used for predicting load demands and the new energy output in a period of time in the future, an optimal scheduling model with a minimum operating cost of a virtual power plant as an objective function is established, and an improved particle swarm optimization algorithm of a parameter omega is changed through a linear decline weight strategy to obtain an optimal solution, so that the influence of prediction errors on an optimization result is reduced, and the short-term optimization effect is improved.

Description

technical field [0001] The invention relates to the field of electric power optimal dispatching, in particular to a virtual power plant energy storage optimal dispatching strategy based on model predictive control. Background technique [0002] The proportion of distributed power generation represented by renewable energy in the power grid is increasing year by year, but this form of power generation has inherent randomness and intermittent characteristics, which make the safe operation of power supply and distribution equipment and the reliability of power supply of the power grid new problems arise; [0003] The traditional scheduling algorithm is based on the load and the output of distributed generation, and the optimal scheduling value of each unit at each moment is obtained through offline stochastic optimization. The premise of this is to assume that the predicted value is accurate. Uncertainty, intermittent, and output fluctuations of distributed power sources are d...

Images