Load Forecasting of a Desert: a Computational Intelligence Approach

This paper presents artificial neural networks and particle swarm optimization (ANN-PSO) based short-term load forecasting model with improved generalization technique for the regional power control center of Saudi Electricity Company, Western Operation Area (SEC-WOA) of Saudi Arabia. Weather, load demand, wind speed, wind direction, heat, sunlight, etc. are quite different in a desert land than other places. Thus this model is different from a typical forecasting model considering inputs and outputs. In this research, two models are implemented - firstly load forecasting model for prediction; however, it is not sufficient for desired level of accurate forecasting, and secondly, optimization to improve the results up to at least better than existing results. This paper includes ANN and PSO models for 24-hours ahead load forecasting. ANN is a mathematical tool for mapping complex relations; it is well proved for the successful use of prediction, function approximation with dynamics, categorization, classification, and so forth. In this research, 24-step ahead calculations are performed in the ANN model and results are moderate. On the other hand, PSO is the most promising optimization tool. It is a swarmed based optimization method; it has better information sharing and conveying mechanism; it has better balance of local and global searching abilities; it can handle huge multi-dimensional optimization problems efficiently with hundreds of thousands of constraints. Thus PSO is chosen as the optimization tool that is applied on the weight matrix of ANN to improve results. In this research, PSO reliably and accurately tracks the continuously changing weights of ANN for uncertain load demand. By analyzing the model of ANN for the load-forecasting problem of SEC-WOA with hundreds of thousands of data and changing-uncertain load demand, the PSO is applied for the ANN weight adjustment and to optimize the uncertain load demand, as the ANN is not an optimization method. Res- lts show that the proposed ANN-PSO performs much better than ANN alone for the load forecasting in a desert like Saudi Arabia.