Research And Implementation To Knowledge Mining Model Of Oil Well Measure Adjustment Schemes

Oil well measure adjustment schemes are closely related to oil field production, thequality of the adjustment schemes have a direct impact on oil production. Traditional designfor adjustment schemes rely on the experience of technical personnel, through inquiring largeamountsofdata, comprehensive analysis, and ultimately formatting a practical implementationscheme. The whole process is cumbersome and inefficient, but also due to the differencesbetween the experiences of relevant technical staff lead to deviation of effect of theadjustment measures scheme. In view of the actual situation above, to ensure to offer a timelyand effective measure adjustment scheme when an oil field exception occurs, this text mainlyuses data mining technology to find the general law and implicit knowledge about measuresadjustment scheme from large amount of historical data of the oil well measure adjustmentschemes, and applying the knowledge and mode to actual production to guide the activities ofoil field production.The knowledge and rules generated by decision tree model is easy to understand andimplement. And combining with the actual situation of the oil field production, the authorsdecided to use the decision tree model of data dining, after researching various algorithms ofthe decision tree model, including: ID3algorithm, CLS algorithm, CART algorithm and C4.5algorithm. By comparisonjkls finally decide to adopt the relatively stable and effective ID3algorithm in decision tree model, and make improvements for the ID3algorithm, the contentof the major improvements consists as follows:First, because the ID3algorithm adopts log for calculation, so the operation is not simple.In order to solve the shortcomings of the ID3algorithm computational complexity, citedMaclaurin formula, raised the ID3simplify algorithm basing on the ID3algorithm, so thatsimply the computing.Secondly, since the ID3algorithm has the property of biased in selecting more values,selecting more values are not necessarily optimal. By using the binary tree data structure tostore the decision tree, proposed the ID3simplified algorithm of binary tree storagealgorithms that combination of ID3simplified algorithm and binary ordinary tree algorithmbasing on ID3algorithm.To provide decision support in order to make better use of mining knowledge andconclusions for the oil field development, the article also states the management framework ofwell measures adjustment scheme, the framework is divided into three levels: data layer, knowledge mining layer and application layer. The use of the three-tier structure frameworkimproves the efficiency of data storage, allowing the rapid data mining, using the results tothe application layer, and feedback the practical effect of the knowledge application to theknowledge mining layer, then amending the data mining model. Hereby to give a reasonableand efficient measures adjustment scheme with well exceptions, and to improve the efficiencyof oil field production.