Abstract: Thin-walled rotational shells are characterized by low rigidity and susceptibility to deformation, where reasonable fixture layouts can effectively enhance machining accuracy. However, the multi-variable and multi-objective optimization of fixture layouts requires extensive iterative computations. To address this challenge, a multi-point internal support fixture layout optimization method for thin-walled rotational shells is proposed, black-winged kite algorithm support vector regression (BKA-SVR) model and a multi-objective particle swarm optimization (MOPSO) algorithm. Firstly, an optimization model is established with support positions and clamping forces as variables, considering clamping deformation, machining deformation, and vibration. Secondly, optimal Latin hypercube sampling (optimal Latin hypercube sampling, OLHS) is employed to generate a sample dataset to construct a BKA-SVR surrogate prediction model. Thirdly, the Pareto front solutions are obtained using MOPSO, and the optimal fixture layout is determined via the CRITIC weighting method. Finally, taking a sealed cabin as the research object, the optimal layout of a multi-point internal support fixture is derived for weld seam machining. Milling tests demonstrate that the optimized multi-point support configuration satisfies wall thickness accuracy requirements.