Search

A fuzzy expert system was designed for modeling the current density in a hard chrome plating plant. First, an analysis of the general theory of hard chrome is carried out and it is compared with the practical execution of the plating process. The most relevant process variables are selected. After that, a model using fuzzy logic is proposed. To do this, membership functions, specific ranges, and a rule base describing the system, are defined. A Mamdani type system is used for modeling current density and its application time in the hard chrome process. The expert system obtained show an average percent error between the actual process and the model calculations lower than 10%. It is concluded that the proposed model is representative of the current chromed system studied. In addition, its implementation would mean a decrease in costs associated with both the saving of overtime of the plant operator as well as with the efficiency in the use of energy. [ABSTRACT FROM AUTHOR]

This paper presents the design of an expert controller applied to voltage level control in a power system. The use of pilot node concept allows to keep stable the list of controllers against topological changes in the network. To achieve the optimal voltage profile, the Alberta Electric System Operator regulations are used. The controller validation is performed by simulating a case library and the solution is compared with results provided by an optimal power flow and a genetic based algorithm. The main advantage of the propose model is the unified treatment of the voltage profile and the reactive dispatch without the need of a two-step algorithm. The proposed controller is suitable for online real-time applications and it is flexible when dealing with problems that present a complex analytical formulation. [ABSTRACT FROM AUTHOR]

This work presents a fault-detection prototype handbook of a pilot line for cheese production, obtained from the knowledge of an expert in the monitoring of the plant. By using the IDEAL methodology (Identification, Development, Execution, Action, Achievement), the acquisition, the conceptualization and the formalization of the knowledge that the expert handles was done. Correct and complete conceptual and formal models of the expert knowledge were established and from them, the first prototype was developed. This prototype had a good answer, with an error of 8% in the fault-detection produced during thirty two cheese elaborations. [ABSTRACT FROM AUTHOR]