Self-Optimizing Multi-Agent Manufacturing Systems Using Hybrid Optimization and Intelligent Control Techniques

This project framework explores the development of a Self-Optimizing Multi-Agent Manufacturing System (SOMAMS) designed to address the operational complexities of decentralized, dynamic production environments. Utilizing a multi-layered architecture, the proposed system integrates supervisory coordination with a hybrid optimization algorithm combining Genetic Algorithms (GA) and Particle Swarm Optimization (PSO). This hybrid approach balances exploration and exploitation to solve multi-objective optimization problems, specifically targeting the minimization of makespan, energy consumption, tardiness, and operational costs under strict physical constraints. To handle real-time environmental disturbances and dynamic variations, reinforcement learning-based control is incorporated within decentralized agent structures representing products, machines, and resources. The methodology provides a robust simulation-based framework for evaluating agent-based self-organization, resilience,

and resource utilization. Project development support focuses on structuring the mathematical formulation of the multi-objective functions, guiding the implementation of the hybrid GA-PSO algorithm, and reviewing the reinforcement learning agent training protocols. The resulting model serves as an advanced research direction for evaluating decentralized decision-making and adaptive control strategies in complex process and manufacturing systems.

Message: Hi FE, I need help with “Self-Optimizing Multi-Agent Manufacturing Systems Using Hybrid Optimization and Intelligent Control Techniques” in “Chemical Engineering”