Industry News

Your location:Home >> News

Dagang Oilfield has independently developed a large model to build a new intelligent ecosystem

Release time:2025-04-25 13:09:58 Clicks:


As of April 21st, the Port Intelligence application large model of Dagang Oilfield has been online for three months, with an average daily call volume of over 10,000 times, covering more than 80% of the main business units.
The Hongzhi Application large Model is an independently developed project of Dagang Oilfield. It deeply integrates the three major model technologies of Qianwen, DeepSeek and Kunlun. Through local deployment and data security control, it injects new intelligent impetus into the exploration and development, production management and business decision-making of the oilfield. Through in-depth integration with oilfield regulations and rules, meeting materials, leaders' speeches, as well as single-well drilling and logging tests and production dynamic data, the large model has achieved full-chain connection between the oilfield knowledge base and real-time data. Under the premise of ensuring data security, it has launched multiple core functions.
This large model was officially launched at the beginning of this year, supporting precise Q&A on oilfield business knowledge. Employees can quickly obtain production data, system interpretations, policy analyses and other information through natural language interaction. Geological researchers only need to input their requirements to access data from multiple systems with one click, saying goodbye to the inefficient process of repeated cross-platform queries. Meanwhile, the model covers scenarios such as plan formulation, report writing, and code development, and can provide standardized document templates and logical frameworks, helping to increase employees' work efficiency by more than 50%.
In addition, the model integrates the latest industry trends, professional terminology libraries and historical cases, and is equipped with a "knowledge processing plant" function. It supports employees to upload materials independently to build personalized databases, forming iterable smart assets, providing scientific support for the refined management of oil and gas extraction. Chen Zhe, a technical expert from Dagang Oilfield, introduced: "The application of large models by Gangzhi enables data to truly 'speak' and becomes the work think tank for every employee."