为了保证飞机机翼的最优位姿装配，提出一种基于容差约束的机翼位姿评价及优化算法。该方法以翼身交点的三维相对偏差和水平测量点的相对高差的加权组合作为机翼装配误差的综合评价函数，以该装配误差最小作为目标，翼身交点的位置容差和水平测量点的高度容差作为约束条件，建立机翼位姿评价的非线性优化模型。通过奇异值分解法获得机翼位姿的初值，再采用Lagrange法结合拟牛顿法迭代求解，得到机翼的最优位姿，使得机翼在满足容差约束的同时，综合装配误差最小。最后，通过仿真算例验证了算法的有效性。

To ensure the alignment accuracy of wing in aircraft assembly, an evaluation algorithm is proposed, which is based on tolerance constraints. With relative deviation of wing-fuselage joints and level-measuring points, an evaluation function of assembly errors is constructed. The optimization model of wing digital alignment is established to minimize the synthesis assembly error with tolerance requirements, which are made of 3D tolerance of wing-fuselage joints and relative tolerance between level-measuring points. The non-linear constrained optimization problem is solved by Lagrange multiplier method and quasi-Newton method and its initial value is provided by SVD method. The optimized position and orientation of wing is obtained, which makes the tolerance requirements be met and the synthesis assembly error minimal. Compared with the result of SVD method through simulated data, the algorithm is validated.