【摘要】 为将试井分析算法从传统单一计算节点转换为并行式、分布式计算模式,在OpenStack云计算架构基础上,通过Nova以及Swift项目实现计算节点分配和面向对象存储,使用改进后调度算法降低了系统计算资源负载。在保证计算精度的前提下,极大缩短了计算时间,实现了对计算效率要求很高的实时拟合过程。利用云平台的并行及分布式计算特点,提出试井分析中井底压力计算方法的并行新策略,实现了并行新算法,研发了基于云平台的试井分析软件。经100余口井例的计算对比,结果表明,基于OpenStack的新架构可以实现试井分析中偏微分方程复杂解析解的实时计算,是云平台试井分析的重要基础。更多还原

【Abstract】 On the OpenStack architecture, Nova and Swift projects were used to implement computing node allocation and object-oriented storage to convert well test algorithm from a traditional single node to a parallel, distributed computing model, and consequently the computing resource load was reduced by using an improved scheduling algorithm. Under the premise of ensuring the calculation accuracy, the calculation time was greatly shortened, and the real-time fitting process that requires highly efficient calculation was realized. Taking the advantage of parallel and distributed computation of the cloud platform, a new parallel strategy was proposed for calculating the bottom hole pressure in well test analysis, a new parallel algorithm was established, and finally well test analysis software based on the cloud platform was developed. Applications in more than 100 wells show that the new architecture based on the OpenStack architecture can real-time calculate the complex analytical solutions of partial differential equations in well test analysis. It provides an important foundation for well test analysis on the cloud platform.更多还原