Modeling and control of hierarchical systems with fuzzy systems (Q1362424)

The paper proposes to use a fuzzy methodology to model the supervision and planning levels of a three level hierarchical control system. Only the first level of the process and the controller itself remain crisp. The main advantage of this system is that all levels of this hierarchical system can be formulated in a same mathematical framework. In this way, a mathematically rigorous model can be developed which is stable and for which the tracking error converges to zero. This fuzzy approach also avoids to have to combine differential equations with discrete event systems. This results from the possibility that fuzzy systems have to transform rule based instructions in continuous mathematical formulas using fuzzy logic. The paper shows how this principle can be applied to a robot planning and control problem and also to an intelligent vehicle highway system. Even if the author asserts that the approach, as described in the paper, is still preliminary, it looks very interesting and is certainly very well formulated and documented. Although some more tests are probably required, the work already done looks promising and worthwhile to study.