Incentive-based demand response strategies for natural gas considering carbon emissions and load volatility
Huibin Zeng,Bilin Shao,Hongbin Dai,Ning Tian,Wei Zhao
DOI: https://doi.org/10.1016/j.apenergy.2023.121541
IF: 11.2
2023-10-01
Applied Energy
Abstract:With the continuous development of the global economy, the demand for natural gas is increasing year by year. At the same time, the widespread adoption of emission reduction measures around the world prompts many countries to switch to cleaner energy sources, which is driving the growth in demand for natural gas. In order to regulate the balance of natural gas supply and demand, maintain the stable operation of natural gas systems and reduce environmental pollution, this paper proposes an incentive-based demand respond (DR) model for natural gas, which takes carbon emissions and load volatility into account, and in the end, it concludes an effective response strategy. In order to get the multi-objective incentive-based DR, first of all, the reward function is constructed by adopting the TOPSIS method, and then it is solved by using the DQN algorithm, in the end, a multi-objective reinforcement learning method based on DQN is proposed. The DR model is evaluated by adopting a real load data from Xi'an city and DR strategies are proposed for different gas users in two periods: the heating and non-heating periods. The DR strategy proposed in this paper can effectively increase gas users motivation, reduce peak loads, maximize the revenue of gas suppliers and gas users, mitigate load volatility and reduce carbon emissions.
energy & fuels,engineering, chemical