A two-tier optimization framework for urban integrated energy systems incorporating PSO-LSTM data-driven prediction and low-carbon demand response

The combination of data-driven renewable energy power forecasting and low-carbon dispatch optimization will become a new model for optimizing the operation of urban integrated energy system (IES). In addition, the optimized operation taking into account economic and environmental benefits needs to consider load demand response (DR) and appropriate carbon trading mechanism (CTM). To address this issue, this paper proposes a two-tier low-carbon optimization framework for synergizing carbon trading and DR in an urban IES based on data-driven prediction. First, PSO-LSTM is used in the first data-driven layer to predict the active power of renewable energy. Second, a comprehensive carbon cost model is introduced in the second layer, combining the cost of carbon trade (COCT) and the cost of carbon storage (COCS). This model, along with the costs of energy purchase (COEP), the cost of operation and maintenance (COOM), and the cost of time-of-use comfort (COTUC) serves as the objective function in the proposed IES optimization. Then, a novel IES optimization model is developed that incorporates both DR and frequency security control. The proposed model is validated using a real-world IES located in eastern China. The results show that the proposed model outperforms the benchmarks from three perspectives: 1) a 2.02 % reduction in total operation costs, 2) a 5.77 % reduction in carbon emissions of IES, and 3) a 34 % reduction in steady-state frequency deviation within the IES.