针对塔里木油田产运储销信息化方面的不足以及核心智能应用的缺乏,创新性提出深度融合大数据和人工智能技术的油田智能分析辅助决策子系统开发思路,利用大数据和人工智能技术,通过数据收集、数据清洗和数据变换,全面收集和精准分析油田生产过程中产生的大量数据。在产运储销平衡模块中,拟使用长短期记忆(Long Short-Term Memory,LSTM)神经网络处理和分析动态数据,采用随机森林(Random Forest,RF)构建模拟预测模型,使用遗传算法(Genetic Algorithm,GA)优化目标结果。油田智能分析辅助决策子系统在线诊断模块使用基于阈值的异常检测自编码器模型,实时监测设备运行状态。应急救援在线训练模块采用决策树构建决策支持模型,提供智能决策支持。基于大数据和人工智能的油田智能分析辅助决策子系统打破了传统油田生产管理的局限性,提高了整体运营效率和资源利用效率,为油田行业的智能化发展提供了新的思路和方法。

Aiming at the deficiencies in production, transportation, storage and marketing informatization and the lack of core artificial intelligent applications in Tarim Oilfield, an innovative development idea of oilfield intelligent analysis assisted decision-making sub-system that deeply integrates big data and artificial intelligence technology is proposed, making use of big data technology and comprehensively collects and accurately analyzes a large amount of data generated in the production process of the oilfield by means of data collection, data cleansing, and data transformation. In the production, transportation, storage and marketing balanced module, it is proposed to use Long Short-Term Memory (LSTM) to process and analyze dynamic data, Random Forest (RF)to build a simulation prediction model, and Genetic Algorithm (GA) to optimize the target results. The online diagnosis module of the oilfield intelligent analysis assisted decision-making subsystem uses a threshold-based anomaly detection self-encoder model to monitor the equipment operation status in real time. The online training module for emergency rescue uses a decision tree to construct a decision support model to provide intelligent decision support. The oilfield intelligent analysis assisted decision-making subsystem based on big data and artificial intelligence overcomes the limitations of traditional oilfield production management, improves the overall operational efficiency and resource utilization efficiency, and provides new ideas and methods for the application of artificial intelligence in the development of the entire oilfield industry.