基于最小二乘估计算法的钻井液流变模式参数优化

OPTIMIZATION MODEL OF DRILLING FLUID RHEOLOGICAL PARAMETER BASED ON LEAST SQUARES ESTIMATION ALGORITHM

借鉴最小二乘估计的优化思想,推导了流变模式(宾汉流变模式、幂律流变模式和赫-巴流变模式)的参数求解方法,并使用了拟牛顿算法求解,该方法避免了因转化成线性模型过程中的统计误差;根据SSE、RMSE和R2,评价了流变参数拟合效果。对实际流变数据进行拟合,结果表明,该方法拟合精度高,效果好,可应用于钻井液流变参数确定、流变模式优选及钻井液性能调整。

Least square estimation algorithm optimization was proposed on the basis of the characteristics of drilling fluid rheological formula,and the solution methods of drilling fluid rheological parameter models( including Bingham rheological model,Power- law model and Herschel- Bulkley rheological model) were deduced and was calculated using Quasi Newton method,which can avoid the statistical errors in the process of change into linear model. Based on the method of SSE、RMSE and R2,the effect of rheological parameter fitting was evaluated.The rheological data fitting results showed that Least squares estimation algorithm has smaller fitting variance. This method can be used to determine the rheological model parameter of drilling fluid,optimize rheological model,adjust and evaluate the performance of drilling fluid.