Abstract: Adaptive cutting planning of shearer based on the fluctuation of coal seam is one of the key problems to realize intelligent unmanned mining in coal mine. However, the adaptability of existing shearer cutting planning scheme considering the changes of complex geological conditions is relatively weak. Aiming at the problems of incomplete consideration of coal seam distribution information characteristics, lack of appropriate shearer cutting path planning model for different undulating conditions and poor continuous planning accuracy in fully mechanized coal mining faces, an adaptive cutting path optimization model for complex undulating coal seams is proposed. Firstly, the time series prediction model of shearer drum height is established based on particle swarm optimization least squares support vector machine, which can realize accurate and advanced prediction of shearer drum height. Then, the optimization models of adaptive cutting path planning for fully mechanized working face under near horizontal conditions and inclined mining conditions are constructed respectively, with the objective of minimizing the deviation between the cutting lines of shearers and the boundary lines of coal seams. Finally, the optimal path is solved by multi-constraint optimization algorithm to realize the comprehensive mining working face adaptive cutting path planning for various mining conditions. Through data simulation, the accuracy of shearer drum cutting height prediction model is above 84.11%, and the maximum average absolute percentage error between the optimized cutting path and the simulated coal seam boundary is 3.13. The proposed method can achieve high-precision prediction of the cutting trajectory of the shearer drum and achieve continuous adaptive cutting path planning for complex conditions in fully mechanized mining faces, providing a reference for the application of adaptive cutting path planning for shearers in complex undulating working faces.