Affiliations: [a]
School of Electrical Engineering, Shenyang University of Technology, Shenyang, China
| [b]
College of Information, Shenyang Institute of Engineering, Shenyang, China
| [c]
School of Computer Science and Informatics, De Montfort University, Leicester LE1 9BH, UK
| [d]
College of Energy and Power, Shenyang Institute of Engineering, Shenyang, China
Correspondence:
[*]
Corresponding author. L. Wang, E-mail: [email protected].
Abstract: A low-carbon economic scheduling method based on a Q-learning-based multiobjective memetic algorithm (Q-MOMA) is proposed to improve the economy of cogeneration system scheduling and reduce carbon emission. First, the model incorporates a carbon capture device, a heat storage device, and a demand response mechanism to enhance the system’s flexibility and wind power consumption. In addition, the Q-MOMA algorithm combines global and local search and uses a Q-learning algorithm to dynamically adjust the crossover and mutation probabilities to improve the algorithm’s searchability. Finally, the fuzzy membership function method is used to make a multiobjective decision, which balances the economy and low carbon of the system, and a compromise scheduling scheme is given. The effectiveness of the proposed model and solution method is verified through the simulation calculation of the improved system and compared with the simulation results of various optimization algorithms. The simulation results show that the proposed model can improve the wind power consumption space and the system’s economy and reduce carbon emissions. The Q-MOMA algorithm has a relatively better optimization ability in the low-carbon economic scheduling of the cogeneration system.
