Drilling geologic characteristic parameters estimation and prediction model

In view of the accuracy of the current drilling geologic characteristic parameters estimation and forecast model is not high, this paper presents drilling geologic characteristic parameters estimation and forecast model based on an improved Adaboost-SVM and modal parameters. Firstly, use the ergodicity, randomness and regularity of Logistic map to change SVM parameters, and then propose a measurement function combined with precision and differentiation to evaluate the generalization ability of the integrated classifier, finally, on the basis of the traditional drilling geologic characteristic parameters forecasting techniques, obtain the test results by using the modal parameters of the measured data identification system. Simulation results show that the proposed drilling geologic characteristic parameters estimation and forecast model based on the improved Adaboost-SVM and modal parameters has higher estimation and prediction accuracy. © 2014 CAFET-INNOVA TECHNICAL SOCIETY. All rights reserved.