Yongbiao Yang

Abstract Natural gas is the main energy source and carbon emission source of integrated energy systems (IES). In existing studies, the price of natural gas is generally fixed, and the impact of price fluctuation which may be brought by future liberalization of the terminal side of the natural gas market on the IES is rarely considered. This paper constructs a natural gas price fluctuation model based on particle swarm optimization (PSO) and Dynamic Bayesian networks (DBN) algorithms. It uses the improved epsilon constraint method and fuzzy multi-weight technology to solve the Pareto frontier set considering the system operation cost and carbon emission. The system operation cost is described using Latin Hypercube Sampling (LHS) to predict the stochastic output of the renewable energy source, and a penalty function based on the Predicted Mean Vote (PMV) model to describe the thermal comfort of the user. This is analyzed using the Grey Wolf Optimization (GWO) algorithm. Carbon emissions are calculated using the carbon accounting method, and a ladder penalty mechanism is introduced to define the carbon trading price. Results of the comparison illustrate that the Pareto optimal solution tends to choose less carbon emission, electricity is more economical, and gas is less carbon-intensive in a small IES for end-users when the price of natural gas fluctuates. The impacts of various extents of natural gas price fluctuation for the same load are also discussed.