预弯底部钻具组合 (bottom hole assembly, BHA) 钻头合导向力对于井眼轨迹控制十分重要, 但是传统的计算方法耗时耗力. 提出了一种基于粒子群算法优化支持向量回归的快速预测预弯 BHA 钻头合导向力的方法. 首先使用加权余量法求解 BHA 的三维小挠度力学模型, 得到钻头合导向力, 形成支持向量回归的样本空间; 然后, 使用粒子群算法优化支持向量回归参数, 得到惩罚因子、核函数参数和不敏感系数的最优值; 最后, 结合实例, 使用粒子群算法调参支持向量回归预测钻头合导向力, 并对预测结果精度进行分析评价. 研究结果表明, 预弯 BHA 钻头合导向力预测精度较高, 决定系数 R 2 达 0.956 8.

The combined guiding force of pre-bent bottom hole assembly (BHA) bit is of paramount significance for well trajectory control, but the traditional calculation method is time-consuming and labor-intensive. A method based on particle swarm optimization and support vector regression is proposed to quickly predict the combined guiding force of prebent BHA bit. Firstly, the weighted margin method is used to solve the three-dimensional small deflection mechanical model of the BHA, and the bit combined guiding force is obtained, and the sample space of support vector regression is formed. Secondly, particle swarm optimization is used to optimize the support vector regression parameters, and the optimal values of penalty factor, kernel function parameters and insensitivity coefficient are obtained. Finally, combined with an example, particle swarm optimization is used to predict the bit combined guiding force, and the accuracy of the prediction results is analyzed and evaluated. The results show that the prediction accuracy of the combined guiding force of pre-bent BHA bit is high, and the coefficient of determination R 2 is 0.956 8.