Resource aggregation and optimization strategy for virtual power plants based on multi-energy collaboration

At present, the regulation and control of virtual power plants are mostly deterministic optimizations, relying on predictive data and combining internal system constraints and electricity price conditions to obtain operation plans. However, with the large-scale integration of renewable energy into the grid, uncertainty algorithms have gradually shown their advantages. This paper proposes a two-stage uncertainty optimization strategy that takes into account demand response. Firstly, a wind-solar power prediction model is constructed. The correlation degree of environmental factors is calculated through gray-level correlation. Then, the meteorology is classified according to the probability values in the Gaussian distribution. Finally, the prediction model is constructed in combination with the gated recurrent neural network model. In the pre-regulation, the operation plan is obtained based on the predicted values and with cost minimization as the objective function. Subsequently, re-regulation is carried out, and the robust optimization method is employed. By determining the worst-case scenario in the uncertain set, the optimal regulation scheme is obtained. Through case analysis, it is found that demand response has the effect of shaving peaks and filling valleys. When comparing deterministic and uncertainty regulation strategies, the operating cost of uncertainty regulation is relatively high in the pre-regulation stage. However, the uncertainty regulation took into account the uncertainty of wind and solar power. In the re-regulation stage, zero abandoned wind and solar power and load cutting are achieved, and ultimately the operating cost of the virtual power plant was reduced by 11.20%.