Miguel A. Jaramillo-Morán

Monthly Electric Energy Demand Forecasting Based on Trend Extraction

Medium-term electric energy demand forecasting is an essential tool for power system planning and operation, mainly in those countries whose power systems operate in a deregulated environment. This paper proposes a novel approach to monthly electric energy demand time series forecasting, in which it is split into two new series: the trend and the fluctuation around it. Then two neural networks are trained to forecast them separately. These predictions are added up to obtain an overall forecasting. Several methods have been tested to find out which of them provides the best performance in the trend extraction. The proposed technique has been applied to the Spanish peninsular monthly electric consumption. The results obtained are better than those reached when only one neural network was used to forecast the original consumption series and also than those obtained with the ARIMA method

Forecasting of the electric energy demand trend and monthly fluctuation with neural networks

Electric energy demand forecasting is a fundamental tool for production and distribution companies because it provides them with a prediction of the market demand of electric energy. Two kinds of forecasting may be performed: short term and medium to long term. This work is focused on monthly prediction, which is useful for the maintenance planning of grids and as market research for electricity producers and resellers. The timed series of monthly electric energy demands presents a rising tendency due to the influence of economic and technological evolution on the electric market. Embedded in this general trend is a fluctuation caused by the difference in demand from month to month. This paper proposes the extraction of that trend to perform separate predictions of both tendency and fluctuation with neural networks, which will be summed up to obtain the series forecasting. A Mean Absolute Percentage Error (MAPE) of about 2% has been obtained.

Multimicrocomputer implementation of three-dimensional neural networks

Abstract The simulation of neural networks by means of the development of software strategies implemented on a computer always comes up against the limitation of execution time from running the programs onconventional von Neumann type systems and is still not really solved with the use of new parallel architectures of microcomputers. In this paper, we present our work in progress aimed at taking advantage of the facilities of generation of parallel systems of microcomputers with a bus specially prepared for the simultaneous functioning of several CPU boards sharing resources (memory, input/output, etc) and under the overall control of a master. Each board includes a microprocessor, a floating point coprocessor, and a certain amount of private memory shared with the bus, which in this case is the VME Bus of 32 bits.

Sliding Mode Control of a Wastewater Plant with Neural Networks and Genetic Algorithms

In this work a simulated wastewater treatment plant is controlled with a sliding mode control carried out with softcomputing techniques. The controller has two modules: the first one performs the plant control when its dynamics lies inside an optimal working region and is carried out by a neural network trained to reproduce the behavior of the technician who controls an actual plant, while the second one drives the system dynamics towards that region when it works outside it and is carried out by a corrective function whose parameters have been adjusted with a genetic algorithm. The controller so defined performs satisfactory even when extreme inputs are presented to the model.

Three-Dimensional Neural Network Synthesis

Abstract A method of synthesizing neural networks is presented in which use is made of the VME buss multiprocessing capabilities to configure the network as a spatial grid in which the concept of neural plane is eliminated to be replaced by one of 3-dimensional neighbourhoods. This arrangement is accompanied by substantial modification of the function implemented by each neuron, and of the concept of the networks learning.