Domain adaptive control architecture for advanced manufacturing systems
The main objective of this study is to design, develop, and investigate domain adaptive control architecture for advanced manufacturing systems. The proposed control architecture is a swarm-based, multi-agent system that exhibits adaptive and emergent behavior that has been inspired from social insects, such as wasps.
The proposed control architecture is built on a decentralized model, which does not require any user-defined control parameters but rather employs stigmergic communication that results in emerging behavior with minimal negotiation overhead. Inspired from natural systems, the architecture is based on simple rules that self-adapts automatically based on the real-time status of the manufacturing system. A 3-layer modular software platform that includes MATLAB ® codes, ARENA® simulation models, Visual Basic ® modules, and MS ACCESS® database, is developed to facilitate experimentation and benchmarking.
The proposed control architecture is applied to a variety of manufacturing domains and benchmarked against studies from the literature. A variety of experiments, involving internal and external dynamic disturbances, is performed. The experimental results clearly show that the proposed model is more reliable, adaptable, scalable, survivable, and robust and outperforms all the benchmarked models and heuristics. This research adds to the literature the advantages of applying stigmergy as a means of communication among agents in a multi-agent system. The research clearly shows as how the mathematical model developed for wasp life-cycle can be directly mapped and used to proactively control manufacturing systems.
0800: Artificial intelligence