Process planning is an essential part of the manufacturing system linking the designing and practical manufacturing. However, the reported process planning models are too simple to describe all characteristics because of the complexity of process planning. Therefore, a new mixed-integer linear programming (MILP) mathematical model is established based on OR-node of the network graph. In the model, the linear expression of the OR-node controlling function as well as three types of changing costs are first established. Beside, considering the OR-node selection state in the encoding and decoding method, a hybrid evolutionary algorithm (HEA) is designed to combine a genetic algorithm with a simulated annealing algorithm. The tournament selection method is adopted in the proposed HEA, and the discussion on the tournament size is conducted on the open problems to make the algorithm designing more reasonable and scientific. The HEA and the new MILP model are both tested on series of numerical experiments which are carried on the existing benchmarks as well as some randomly generated cases. The behavior of both two methods can verify their effectiveness and superiority successfully.

工艺计划是连接设计和实际制造的制造系统的重要组成部分。但是，由于过程计划的复杂性，所报告的过程计划模型过于简单，无法描述所有特征。因此，基于网络图的“或”节点建立了一个新的混合整数线性规划（MILP）数学模型。在该模型中，首先建立“或”节点控制函数的线性表达式以及三种变化的成本。此外，考虑到编码和解码方法中的“或”节点选择状态，设计了一种混合进化算法（HEA），将遗传算法与模拟退火算法相结合。建议的HEA采用了比赛选择方法，通过对公开赛问题的讨论，使算法设计更加合理，科学。HEA和新的MILP模型都在一系列数值实验中进行了测试，这些实验是在现有基准以及一些随机生成的情况下进行的。两种方法的行为均可成功验证其有效性和优越性。