Spatial electric distribution load forecasting

Electric distribution systems planning needs know the magnitude of the power that must be delivered, in order to plan the investment expansion and the efficient operation of its equipment. However, beyond these estimates, knowing the spatial distribution of the electric load increases is fundamental. A simulation model of market electric expansion is proposed, to take into account time and space. This model is based on spatial temporal disorder ideas. The spatial load evolution, is simulated using a dynamic system, whose main components are the iteration among the consumers and the consumption migration of dense regions. The model presents various possible configurations of the distribution market. For this, the model works with a parameter set, in order to model the load evolution type. The spatial disorder is generated by an algorithm. To adjust the simulation results to a real data set of load evolution, a simulation set is used for each test parameter set. So, the result of each simulation set is converted in a number. This number is calculated using a first and second order moments arrangement. The best parameter set is obtained using simulated annealing. The evaluation function of the simulated annealing results of the simulation and real data comparison, is shown by the number. The ideal parameter search processes in hyperspace, in a nonlinear topology. The results of several tests, show a parameter set able to generate a similar spatial disorder to the real market, by means of a predefined statistic criterion.