The Research Of Energy Discriminant Model Of Mainly Energy Dissipation Equipment For Oilfield Water Injection System

As the oilfield energy-saving and emission-reduction work unceasingly thorough,improve energy management level, and fully tap the potential of fine management has beenimperative. Moreover, the current oilfield management of energy-consuming equipment is asingle non-key management style, we not only ignore deal with each case on its meritsthrough analysis the pros and cons situation, and not have clear control of the powerconsumption issues of potential equipment, therefore, research the energy consumptiondiscriminant model of key energy-consuming equipment is very important, espectially thelarge proportion of oilfield water injection system.In this paper, the foundation of energy consumption model is the main line, research datapreprocessing, energy consumption index prediction and establish discriminant model,ultimately to adequate theoretical basis and conclusive experimental results verify the validityand development of the model, and applied to the practical application, look forward to obtaingood experimental results.The paper mainly studies the content as follows:I. Based on the weight transfer weighted moving method is presented in this paper, todeal with the historical data de-noising, smoothing of the discriminant model. Whichsmoothing algorithm to ensure data and delay the passage of the trend of the experimentaldata at the same time, the method has been improve the hysteresis quality and neglecttransition value, and the feasibility and effectiveness are verified by experiments.II. Aiming at the energy consumption prediction index problem of the model is proposedusing the QPSO algorithm to optimize BP neural network prediction method of mixing theory,the optimazed algorithm has better global optimization, fast convergence speed, andexperiments show that the mixing theory has good effect.III. This paper presents key energy-consuming equipment energy consumptiondistinguish model of water injection system. The model based on equipment energyconsumption index, using supervised learning and intelligent discrimination idea, establishingmathematical model of energy consumption index, and division the region of it, then, inputparameters to the graphics, such discrimination model is formed.