Incorporating machine intelligence in a parameter-based control system: a neural-fuzzy approach

Abstract

The capabilities of the two computational intelligence technologies including neural network and fuzzy logic can be synergized through the formation of an integrated and unified model which capitalizes on the benefits and concurrently offsets the flaws of the involved technologies. In this paper, a neural-fuzzy model, which is characterized by its ability to suggest the appropriate change of process parameters in a relatively complex parameter-based control situation involving multiple parameters, is presented. This model is particularly useful in multiple input and multiple output situations where complex mathematical calculations are required if conventional control approach is adopted. In particular, it serves to acquire knowledge from the information base for extracting rules, which are then fuzzified based on fuzzy principle. To validate the feasibility of this approach, a test has been conducted based on the neural-fuzzy model with the objective to achieve heat transfer enhancement in rectangular ducts using transverse ribs. This paper describes the roadmap for the deployment of this hybrid model to enhance machine intelligence of a complex system with the description of a case study to exemplify its underlying principles.