Towards Analysing and Controlling Self-Organising Systems with Socio-economic Principles

摘要

The increasing complexity of large-scale distributed applications motivates the study and development of self-organising systems. However, engineering self-organising systems is still a challenge. Despite their benefits, self-organising systems suffer from a lack of control and stability, so that methods and tools are needed for improving their understanding and control. In this paper, we consider socio-economic concepts as a tool for analysing and controlling self-organising systems. We propose an approach to enhance existing self-organising mechanisms with high-level representation and reasoning abilities based on socio-economic models. This approach has the advantage of preserving the typical performance, flexibility and robustness of a self-organising system, while improving its stability and making it more understandable to its users. We illustrate the approach with the example of a flexible manufacturing system we have evaluated, and we explore further directions based on this example.