LAUSR

Abstract As the renewable energy generation keeps growing, the capacity of wind-solar power generation penetrated into the grid system increases year by year, and the influence of renewable generation power fluctuations on the stability of grid after integrated also increases, and even affects the normal operation of grid system in serious cases. In this case, we consider to use a dynamic load to maintain the stability of the grid. Since datacenters show great potential as dynamic loads to guarantee the grid stability, this paper focused on studying the impact of dynamic load regulation of datacenters on the stability of integrated power grid systems. Accordingly, we proposed a dynamic load regulation method based on a stochastic search algorithm, collecting historical data to carry out a series of experiments. Experimental results show that the peak off-limit rate could be reduced up to 72% under the strategy we proposed, which means that the strategy could help the grid to keep the frequency stability of the grid as well as avoiding bus voltage and power flow violations. Furthermore, practical factors including the flexibility and the action delay of the datacenter load adjustment were also incorporated and examined. Through a series of comparative simulation experiments, results also show that the total power loss could be reduced on the basis of keeping the stability, which could help the grid to enhance its sustainability. Finally, test and analysis about efficiency and scalability illustrate that the proposed method is both feasible and effective.