| The beam pumping unit is the most widely used oil pumping equipment in oil fields,but it has the defects of low efficiency and high energy consumption in the process of oil and gas production,especially in the low-permeability oil field.The intelligent control system of the beam pumping unit developed in this paper is based on the actual production data from a large crude oil extraction enterprise in northeast China to study the quantification method of oil well fluid supply capacity based on deep learning,and combine the edge computing technology to make up for the shortcomings of the traditional beam pumping unit control system in terms of real-time,reliability and intelligence,thereby promoting energy saving and efficiency improvement in oilfield production and push the intelligent manufacturing construction of the crude oil extraction industry in China.In the first chapter,the background and significance of this subject,the oil production technology of the beam pumping unit,the intelligent control technology of the pumping unit and the research status of the quantification technology of the oil supply capacity are summarized,and then the challenges of this study are analyzed.The main structure and content arrangement of this paper are given in the end.In the second chapter,the overall scheme of the intelligent control system of the beam pumping unit is designed.Aiming at the shortcomings of the traditional intelligent control system of beam pumping unit in real-time,reliability and intelligence level,an intelligent control system based on deep learning and edge computing technology is proposed.The design goals of the system are analyzed,and the overall architecture and key technical routes of the system are planned.In the third chapter,the edge layer intelligent monitoring gateway based on NVIDIA JETSON NANO module is designed and developed.Combined with the functional requirements of the intelligent monitoring gateway,the overall hardware architecture of the gateway is designed and each functional module is described in detail.At the same time,the software of the core functions of the gateway is designed and developed.The gateway realizes the collection,processing,analysis and transmission of oil well working condition data on the edge side of the system,as well as the frequency conversion and start-stop control of the pumping unit.It improves the realtime and reliability of the system,and provides a hardware platform for the deployment of subsequent artificial intelligence algorithms.In the fourth chapter,the quantification algorithm of oil well fluid supply capacity based on deep learning is studied.Firstly,a quantification scheme for oil well fluid supply capacity was designed,and then the principle of the Gaussian heatmap matching model based on deep learning was analyzed,and the model is applied to locate the valve opening and closing points of the dynamometer card.At the same time,some improvements and optimizations are proposed for the model considering the particularity of the location task of the valve opening and closing points.Finally,through the comparison and ablation experiments,the superiority of the improved model in the positioning accuracy of the valve opening and closing points and the effectiveness of each improvement measure are verified.The algorithm realizes the accurate quantification of the liquid supply capacity of the oil well,and provides the most direct basis for the control of the beam pumping unit.In the fifth chapter,the oil well information cloud platform is designed and implemented.Firstly,the overall architecture of the platform is designed according to the specific needs of the cloud platform.Then,starting from the table structure,the pumping unit operating condition database is designed.Finally,the functional module of oilfield data management and intelligent algorithm verification is designed and developed,which further improves the efficiency of oilfield production management.In the sixth chapter,the test and joint debugging of each functional module of the intelligent monitoring gateway and the oil well information cloud platform are completed.At the same time,the industrial application value of the intelligent control system for the beam pumping unit proposed in this paper is confirmed by the actual application test of the oil well.Finally,the problems encountered in the system development process and the corresponding solutions are expounded.The intelligent control system of the beam pumping unit developed in this paper can effectively improve the system efficiency,reduce the system energy consumption in the process of oil production,and realize efficient,real-time,accurate and reliable control of the pumping unit.This study has totally achieved the expected design objectives in the testing environment of the crude oil extraction enterprise and prepared to be deployed and implemented.It surely has promising industrial application value in the future. |
