Ingenieria Electrica

Fuzzy-split range control system is able to regulate a fermentation process. Reducing control effort is possible for similar performance index. Reducing utilities demands is possible for similar performance index. Fuzzy-split range control system provides smooth control output.

The Platanares geothermal area, Departamento de Cop~in, Honduras, is located within a graben that is complexly faulted. The graben is bounded on the north by a highland composed of Paleozoic (?) metamorphic rocks in contact with Cretaceous-Tertiary redbeds of unknown thickness. These are unconformably overlain by Tertiary andesitic lavas, rhyolitic ignimbrites, and associated sedimentary rocks. The volcanic rocks are mostly older than 14 Ma, and thus are too old to represent the surface expression of an active crustal magma body. Thermal fluids that discharge in the area are heated during deep circulation of meteoric water along faults in a region of somewhat elevated heat flow. Geothermometry based upon the chemical composition of thermal fluids from hot springs and from geothermal gradient coreholes suggests that the reservoir equilibrated at temperatures as high as 225 to 240°C, within the Cretaceous redbed sequence. Three c, atinuously cored geothermal gradient holes have been drilled; fluids of about 165°C have been produced from two drilled along a NW-trending fault zone, from depths of 250 to 680 m. A conductive thermal gradient of 139°C/km, at a depth of 400 m, was determined from the third well, drilled 0.6 km west of that fault zone. These data indicate that the Platanares geothermal area holds considerable promise for electrical generation by moderate-to hightemperature geothermal fluids.

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsi

Mathematically, determining whether a non linear system is controllable or not can be a very difficult task, even an impossible one. In this paper, an approximate method is used to check the controllability of a system, with a specific reference input. The method is based on neuro-fuzzy techniques (ANFIS, more specifically) and it proves to provide very good approximate results, useful to determine whether the system is controllable or not. Although the paper refers to a single case, concerning a discrete time system, the method explained can be applied to any other system whose dynamics can be described by means of a difference equation or whose input – output data have been previously collected. Additionally, when the checking procedure is completed, a fuzzy logic controller has also been developed.

From a theorical point of view, underactuated mechanical systems can be controlled by means of the so called IDA-PBC method. In this method, in order to achieve the control objective, the stabilization mechanism follows two basic stages: (1) energy holding stage, which consists on shaping the total energy function of the system in order to assign the desired equilibrium state, and (2) damping introduction stage, necessary to achieve asymptotic stability. In order to mantain an energy approach of the stabilization process, it is necessary to obtain a port controlled Hamiltonian form for the closed loop system. However, the theoretical development faces practical dificulties when applied to real systems, due to the difficult solution to the partial differential equations envolved. The Acrobot system is a prototype of a underactuated mechanical system widely studied by the non linear control community. In this paper, a controller is designed taking into account the port controlled Hami...

The purpose of this investigation was to develop a simple methodology to implement GRAFCET diagrams on programmable logic controllers using ladder language (in any of its versions). In this way, even small scale PLC users can benefit from the capability of GRAFCET diagrams or SFC (sequential function chart) to divide complex processes into simple stages, therefore making it easier to plan and implement its control. The methodology included allows the ladder language programs obtained to be easily read and maintained, while the implementation time is reduced.

This paper introduces a dynamic model of a system of proton exchange membrane fuel cells (PEMFC) using neuro-fuzzy identification (ANFIS). The static and dynamic behavior of the neuro-fuzzy model is validated by means of a mathematical model of a PEMFC stack of fuel cells, obtained from its electrochemical equations. This mathematical model is also used to obtain the training data for the neuro-fuzzy model. The static behavior was tested using the polarization curve of the cell. The dynamic behavior was simulated using input current pulses. The neuro-fuzzy model shows a behavior very similar to that of the original mathematical model, proving the usefulness of the ANFIS technique in this particular area, involving a highly non-linear system.