加工参数优化, 数控车削, 分布估计算法, 多道车削

In NC turning operations, it is of important and also practical sense to decrease the unit production cost. Unfortunately it is a non-linear optimization problem with high complexity and many constraints, that is to decide how to select a set of turning parameters to save the unit production cost. To deal with it, UMDArp and UMDAp algorithms based on marginal estimation of distribution were proposed. And under a set of practical machining constraints, the reasonable machining parameters (speed, feed rate and depth of cut for both rough and finish machining, and number of rough machining) were obtained by proposed algorithms along with optimization of the rough and finish machining parameters at the same time. Furthermore, the performance of the proposed algorithms was improved through the constraint dealing method combined gene repair strategy and penalty function. Computer simulation shows the UMDArp can achieve more optimal solution than other algorithms proposed previous to reduce the production cost.