The Research And Implementation Of Data Mining Method For Exploration Resources Based On Fuzzy Neural Network

At present,the discovery of knowledge hidden in petroleum exploration geological data(hereinafter referred to as exploration data)and the design of automatic mining software based on data mining technology have become a hot research topic in petroleum and petrochemical industry.However,because of the type diversity and structural complexity of petroleum exploration geological data,the relationships among the attributes are not clear and the noise is large,so there is no research on the process of data mining modeling for petroleum exploration geological data.In addition,a large number of petroleum exploration databases have been built at home and abroad to form huge data sets.However,there are few domestic knowledge mining software for exploration data.How to excavate knowledge and establish mining software for large and diverse exploration data has become an urgent need for the current petroleum and petrochemical industry to seek long-term development.In view of the actual situation and the lack of current research,this paper proposes a method of knowledge discovery based on neural network for exploration data and implements it by software.This study is carried out from the following three aspects:(1)Taking the open basin data(one of the exploration data)as the experimental sample,the data samples suitable for mining are prepared through data reprocessing.The steps include data cleaning,data analysis and construction of data.Taking basin data as an example,this paper explains in detail the systematic data reprocessing process proposed in this paper.This part of research is the foundation of establishing the knowledge mining model,and is essential for the establishment of high quality knowledge mining model.(2)According to the model theory of neural network algorithm,it is based on the experiment to explore whether the algorithm is suitable for the basin data in this paper.At the same time,it tests whether the data reprocessing process proposed in this paper is suitable for neural network algorithm.The evaluation criteria for finding the results of knowledge based on the established model are put forward and evaluated and analyzed.(3)Based on the adaptive fuzzy neural network(Adaptive fuzzy neural network)model,the core algorithm of knowledge discovery software based on user needs is proposed.The software together with the knowledge discovery software based on the SVM model of a Research Institute of CNPC has formed a knowledge discovery software system for petroleum exploration resources.The knowledge discovery software based on adaptive fuzzy neural network model based on adaptive fuzzy neural network has achieved good results in practical engineering application.The results of software operation show that the theory of modeling process of data mining proposed in this paper can be realized.