Multi-Agent Credit Assignment and Bankruptcy Game for Improving Resource Allocation in Smart Cities.

Area of Science:

• Computer Science
• Artificial Intelligence
• Urban Planning

Background:

• Smart city development necessitates efficient resource allocation.
• Multi-agent systems (MASs) are suitable for modeling smart city environments.
• Existing resource allocation methods face challenges in complex urban systems.

Purpose of the Study:

• To propose an efficient resource allocation approach for smart cities.
• To integrate the multi-agent credit assignment problem (MCA) with bankruptcy games.
• To evaluate the performance of the proposed methods in a multi-score problem (MSP) setting.

Main Methods:

• Mapping the smart city resource allocation problem to MCA and bankruptcy games.
• Introducing a task start threshold (TST) constraint to convert MCA into a bankruptcy problem.
• Presenting three solution methods: TS-Only, TS + MAS, and TS + ExAg.
• Utilizing a multi-score problem (MSP) for experimental evaluation.

Main Results:

• The proposed approach demonstrates superior performance compared to existing methods.
• The evaluation considered parameters such as learning rate, confidence, expertness, efficiency, certainty, and correctness.
• The new methods showed significant improvements in five key performance metrics.

Conclusions:

• The proposed method offers an effective solution for resource allocation in smart cities.
• The integration of MCA and bankruptcy games provides a robust framework for urban resource management.
• This research contributes to the advancement of intelligent resource management in smart urban environments.