Cylindricity evaluation plays an important role in processing quality evaluation. Accordingly, a new algorithm for evaluating cylindricity based on the minimum condition is presented. First, a theoretical “minimax” optimization model based on the minimum condition is linearized into a linear programming model. Then, the solution of the linear model and the soft alignment are performed alternately to approximate the local-minimum solution of the theoretical model. After the initial posture of the measured cylinder reset by the Monte Carlo method, the linear model is sequentially reconstructed and solved. The global-minimum solution of the theoretical model is determined by comparing the objective functions. Finally, computer arbitration for the minimum condition is conducted to determine whether the obtained solution of the linear programming model reaches the global minimum. The digital verification and application analysis proved that the algorithm is reliable, accurate, and applicable to computer-aided in situ cylindricity measurement of industrial context.

圆柱度评估在加工质量评估中起着重要作用。因此，提出了一种基于最小条件的圆柱度评估算法。首先，将基于最小条件的理论“最小极大”优化模型线性化为线性规划模型。然后，交替执行线性模型的求解和软对齐，以逼近理论模型的局部最小解。在通过蒙特卡洛方法重置被测气缸的初始姿态之后，将线性模型依次重建和求解。通过比较目标函数来确定理论模型的整体最小解。最后，进行最小条件的计算机仲裁，以确定所获得的线性规划模型的解是否达到全局最小值。数字验证和应用分析表明，该算法可靠，准确，适用于工业现场计算机圆柱度测量。