Research On Intelligent Optimization Method Of Logging Interpretation Model Based On Knowledge Graph

With the increasingly complex exploration and development of oilfields,abundant logging knowledge has been accumulated in the field of well logging,and the requirements for the processing and interpretation of logging data are also increasing.In the process of data processing and interpretation,it is required to configure different parameters and select different models for different regions or even different layers in the same region to improve the conformity of logging interpretation.Therefore,in the evaluation of complex reservoirs,how to automatically select the best parameters and interpretation models for various target intervals with different properties and characteristics to intelligently process logging curves has become a very important issue.In response to the above problems,this study proposes to introduce knowledge graphs to manage the logging processing and interpretation software platform itself,related knowledge in the field of logging interpretation,and the knowledge of people with rich experience.Regional interpretation features,narrow the search range as much as possible and quickly give users an optimal strategy,provide users with the most suitable interpretation parameters and models,reduce the workload of staff and improve the overall performance of the logging processing interpretation software platform.The main research contents of this paper are as follows:Firstly,the knowledge graph of logging interpretation field is constructed.According to the knowledge characteristics in the field of logging interpretation and the actual business needs of logging interpretation,combined with expert experience,following the logic of the knowledge graph system architecture,and taking the business model layer and the original entity data layer in the field of logging interpretation as the benchmark,the entity recognition based on Bi LSTM+CRF is used.Technology and relation extraction technology based on domain knowledge rule template realize knowledge acquisition,complete knowledge fusion through ontology matching and entity alignment,and use Neo4 j graph database to realize knowledge storage.The knowledge is displayed,and the construction of the knowledge map in the field of logging interpretation is completed.Secondly,a method for optimizing the parameter model of logging interpretation based on knowledge graph is proposed.In the process of data processing and interpretation,there are many logging interpretation models and the configuration of parameters in different intervals in different regions is complicated.Typical reservoir parameters and logging interpretation models are sorted out,and correlation analysis techniques are used to complete the results that can indicate the oil content of the reservoir.The key parameters of the function mapping relationship between gas volume and logging parameters are determined,and by formulating the interpretation model to identify standard expression specifications and interpretation rules,rule inference is carried out,and the optimization method of logging interpretation parameter model based on knowledge graph is realized,and the Example application,the effect is good.