Dynamic analysis of energy sold through the pulse function to model external variables (ARMAX)

Abstract

The summary of research can be illustrated by the work of sales department in the General Company for Electricity Distribution Baghdad as vital and very effective, so it contains three stages of (energy sold, the value of energy sold, cash received) and has been building a model of energy sold using the methodology of Box - Jenkins and by compared it with model artificial networks neural network and through the neural network was identified on the dynamics of energy sold and its impact on the months, and also to identify the loss of the energy sold and processed.Theory / Introduction: The problem of designing control system well is basically access to features dynamic of the process to be analyzed or control, or other words knowledge of effects and properties of the process dynamics, where the approach to the cycle of performance that can be designed such a system (mathematical model) reflects the dynamic process , the system (model) the designer must represent in the form of mathematical relationships statistically; such as time series (Time Series) and differential equations with a time constant (Continuous Time Differential Equations) and difference equations (Difference Equations) and the equations of transfer of persistent or intermittent (Continuous Or Discrete Transfer Functions ) (2) and others, and difficult developments in the systems make them the inputs and outputs (Input & Output) any black box (Black-Box) where you can not understand the relationships only through the representation of data samples and simulation to gain access to the system operations control dynamic.