基于GA和SVM的水火弯板焰道布置优化方法

水火弯板是目前船舶建造中普遍采用的曲型外板加工方法,由于机理复杂,其加工过程参数确定往往依赖于工人的加工经验,难以实现标准化和规范化。文章提出基于遗传算法(GA)和支持向量机(SVM)的水火弯板焰道布置优化方法,以成形曲面和设计曲面之间的形状误差最小为优化目标,将焰道布置优化问题分解成计算最小形状误差和搜寻最佳匹配位置两个问题,并分别运用嵌套GA和SVM来实现这两个问题的求解。实例验证表明该方法能快速高效确定最优的加热线布置方案。

Optimization method of flame paths layout for line heating based on GA and SVM

Line-Heating is a typical technology for ship-hull curved plate forming process which is generally adopted in shipbuilding presently.Because of the complexity of its mechanism,the parameters determination of Line-Heating always depends on the experience of workers and is difficult to achieve standardization and normalization.The paper provides an optimization method of flame paths layout for line heating based on GA and SVM,in which the optimization goal is the minimum profile error between formed plate and designed plate.The solution of optimal layout of flame paths is obtained by the calculation of minimum profile error and the search for optimal matching position.The two optimization problems are solved by nested Genetic Algorithm(GA) and Support Vector Machine(SVM) respectively.The experimental result shows that the method is effective to get the optimal layout of flame paths.