实现采煤机智能化调高，关键是解决煤岩界面识别问题、截割路径优化问题及采煤机调高控制问题。即使煤岩界面被精确识别，受到实际工作中顶底板的平整性和液压支架的推移滑溜等要求的限制，采煤机滚筒无法完全跟随煤岩界面曲线，因此需要基于煤岩界面识别结果，对起伏变化的煤岩界面曲线进行截割路径优化，得到采煤机调高控制的目标轨迹。滚筒截割路径优化是基于煤岩界面估计曲线，在采煤工艺、煤质要求和设备的适应能力等限制条件的约束下，得到使回采最大化的平滑轨迹。针对上述采煤机截割路径约束优化问题，提出一种基于B样条曲线拟合和蜉蝣算法的采煤机截割路径约束优化方法。为了提高截割路径优化效果和降低计算复杂度，以B样条曲线节点系数作为设计变量，构建一种新型截割路径优化目标函数；考虑采煤机截割工艺、煤质要求等限制，使用多段赋值罚函数法处理约束，根据约束的不满足程度动态改变罚函数系数值，避免优化陷入局部最值和约束不能起到实际作用；为了进一步提高优化效果和收敛速度，使用修正蜉蝣算法寻找最优截割路径。最后，考虑实际煤岩界面中褶皱、陷落柱、断层等典型地质构造，进行仿真研究，结果表明，所提方法能在满足实际约束下快速得到平滑的截割优化路径，实时性好、适用性高。

In order to achieve the intelligent height adjustment control of the shearer, the key techniques are the coal-rock interface recognition, cutting path optimization, and shearer height adjustment control. Although the coal-rock interface is accurately identified, the shearer drum cannot completely follow the estimated coal-rock interface due to the flatness requirement for the roof and floor of the coal seam which guarantees the working of hydraulic supports. Therefore, the cutting trajectory should be optimized based on the coal-rock interface recognition results, which is regarded as the target trajectory of shearer height-adjusting control. To solve this issue, the cutting path optimization is required. Based on the estimated coal-rock interface and considering the limitations in practical application, the cutting path is optimized to maximize the recovery ratio. To improve the optimization results, satisfy the restricting condition, and reduce the computational complexities, this paper proposed a novel constraint optimization method of shearer cutting path based on the mayfly algorithm and B-spline curve fitting. A novel objective function is built, in which the curve node coefficients are chosen as the design variants, and the optimization target is minimizing the difference between the fitness curve and the coal-rock interface, leading to the much less designed variants and the lower computational load. The piecewise penalty function is used to deal with the constraints, which assists the exploration process in escaping from local maxima and make sure the constraints work. And then the modified mayfly algorithm is used to find the optimized cutting path to further improve the optimization effect and the convergence rate. Finally, the simulations of cutting path optimization are conducted under the condition of folds, subsidence and faults, which indicate that the proposed method can obtain the optimized smooth cutting path with the limitation of the constraints quickly, and have high real-time behavior, and good applicability.