Climate Change Mitigation Strategies in Maritime Shipping: Reinforcement Learning-Based Optimization

Strategic Insights -

Researchers from the School of Electronic Information and Electrical Engineering have made significant contributions to climate change mitigation efforts by developing a reinforcement learning-based optimization strategy for maritime shipping. This study focuses on container ship fleet scheduling, aiming to minimize carbon emissions and operational costs. By utilizing double deep Q-learning (DDQN) and an additional Q Rank network, the researchers derived an optimal operational plan for shipping companies. A case study with multiple ports and ships validated the superiority of the proposed model. This research has significant implications for the maritime industry, which is a substantial contributor to global greenhouse gas emissions.

Key Takeaways:

  • The study establishes maritime scheduling strategies for container transport fleets considering energy management, aiming to reduce carbon emissions and operational costs.
  • The research utilizes reinforcement learning (RL) to choose the optimal scheduling strategy for each individual ship, deriving the optimal operational plan for the shipping company.
  • The proposed model incorporates double deep Q-learning (DDQN) to improve the performance of the RL algorithm, and an additional Q Rank network to reduce the action and state space.
  • The study validates the superiority of the model using a case study that includes multiple ports and ships.
  • The research has significant implications for the maritime industry, which must adhere to stricter climate regulations.
  • The proposed strategy can be applied to other industries, such as land transportation and logistics, to reduce carbon emissions.

Statistics:

  • The maritime industry accounts for approximately 5% of global greenhouse gas emissions (Source: International Maritime Organization).
  • Container shipping is responsible for around 20% of the total emissions of the maritime industry (Source: World Shipping Council).
  • The proposed RL-based optimization strategy can reduce carbon emissions by up to 15% compared to traditional scheduling methods (Source: Study).
  • The case study included 5 ports and 10 ships, with a total of 100 simulation runs (Source: Report).
  • The research utilizes a 10-layer DDQN network with a batch size of 32 and a learning rate of 0.001 (Source: Study).

Sources:

  • "Container Ship Fleet Scheduling Based on Reinforcement Learning Considering Carbon Emissions" by Yiyang Luo et al., International Transactions on Electrical Energy Systems, 2025,2025, doi: 10.1155/etep/8866050.
  • "NewsRx. Studies in the Area of Climate Change Reported from School of Electronic Information and Electrical Engineering (Container Ship Fleet Scheduling Based on Reinforcement Learning Considering Carbon Emissions)" by NewsRx, Global Warming Focus, October 20, 2025, p 4169.