机构地区: 中国地质大学北京能源学院
出 处: 《石油天然气学报》 2011年第2期142-145,169,共4页
摘 要: 基于最小二乘法,结合参数识别优化技术与多元回归技术,针对Gauss-Newton方法的缺点,提出改进的Gauss-Newton方法,并对脉冲试井参数进行解释,解释最大误差为6.54%。结果表明,改进方法具有全局搜索能力,并具有超线性收敛速度,对勘探开发初期,生产资料较少情况下快速准确求取地层参数提供了一种可靠的方法。 Based on the least squares method,parameter identification optimization techniques were integrated with multiple element regression techniques,in consideration of the shortcomings of Gauss-Newton method,an improved Gauss-Newton method was proposed to explain pulse well testing parameters with the maximum error of 6.54%.The results show that the improved method has a global searching capacity and super-linear convergence rate,which provides a reliable method for rapid and accurate collection of formation parameters for oilfield exploration and development at the early stage.
关 键 词: 脉冲试井 参数识别 自动识别 最小二乘法 优化技术 多元回归
领 域: [石油与天然气工程]