Comparison of Neural and Evolutionary Approaches to Peak Load Estimation in Distribution Systems

In distribution systems the knowledge of load characteristics at system buses is one of the top requirements for developing a precise analysis and taking good decisions with respect to the optimal operation and planning of the system. This paper presents a comparative study of the peak load estimation in power systems using two approaches based on neural networks and genetic programming. The first approach used the resilient propagation algorithm applied to multi-layer perceptron neural networks, while the second one was based on genetic programming and symbolic regression. A series of case studies were performed to compare the two approaches and to discover the best solution for each method and to evaluate their performances