论文部分内容阅读
目前,钻井地质特征参数的获得主要依赖于地震、测井资料,对待钻井而言,则只有地震信息。而若缺乏详细的地质信息,利用地震信息很难精确地推算各种地质参数。可首先利用已钻井地震信息和测井信息的映射关系,结合待钻井的地震信息,来预测待钻井的测井信息。采用PSO优化的RBF神经网络算法进行地震测井反演,并将该算法应用于准噶尔盆地永字号井。该算法与最小二乘RBF神经网络算法和梯度下降RBF神经网络算法相比,在平均绝对误差、平均相对误差、最大误差、相关系数、数据方差以及收敛速度等方面都是最优的。
At present, the parameters of drilling geological features mainly depend on the seismic and well logging data, but only the seismic information when drilling well. Without detailed geological information, the use of seismic information is difficult to accurately calculate a variety of geological parameters. First, the well logging information to be drilled may be predicted by using the mapping relationship between the well-drilling seismic information and well logging information and the seismic information to be drilled. The PSO-optimized RBF neural network algorithm is used to perform seismic logging inversion. The algorithm is applied to the Yong-zhao well in the Junggar Basin. Compared with the least squares RBF neural network algorithm and the gradient descent RBF neural network algorithm, this algorithm is optimal in terms of average absolute error, average relative error, maximum error, correlation coefficient, data variance and convergence speed.