To address indeterminism in the bilevel knapsack problem,an uncertain bilevel knapsack problem(UBKP)model is proposed.Then,an uncertain solution for UBKP is proposed by defining thePE Nash equilibrium andPE Stackelber...To address indeterminism in the bilevel knapsack problem,an uncertain bilevel knapsack problem(UBKP)model is proposed.Then,an uncertain solution for UBKP is proposed by defining thePE Nash equilibrium andPE Stackelberg-Nash equilibrium.To improve the computational efficiency of the uncertain solution,an evolutionary algorithm,the improved binary wolf pack algorithm,is constructed with one rule(wolf leader regulation),two operators(invert operator and move operator),and three intelligent behaviors(scouting behavior,intelligent hunting behavior,and upgrading).The UBKP model and thePE uncertain solution are applied to an armament transportation problem as a case study.展开更多
针对模糊C-均值聚类算法用于点云分割时对初始值敏感且易于陷入局部最优,导致点云分割效果不理想,不稳定的问题。提出了一种基于曲率约束的改进狼群算法优化模糊C-均值聚类的混合算法(IWPAFCM)。该算法首先在狼群算法中引入佳点集初始...针对模糊C-均值聚类算法用于点云分割时对初始值敏感且易于陷入局部最优,导致点云分割效果不理想,不稳定的问题。提出了一种基于曲率约束的改进狼群算法优化模糊C-均值聚类的混合算法(IWPAFCM)。该算法首先在狼群算法中引入佳点集初始化种群分布;然后利用自适应步长简化参数设定、平衡寻优与收敛时间;进一步应用交互策略增强狼群的内部交流,提升狼群全局寻优的能力;最后对头狼加入高斯扰动机制使其具有跳出局部最优的能力,将改进狼群算法(improved wolf pack algorithm,IWPA)得到的聚类中心作为模糊聚类的初始值进行迭代,由此得到准确的聚类中心。在此基础上,基于点云的法矢量和曲率对点云之间的距离进行定义并替换传统欧式距离,实现了理想的点云分割效果。以ModelNet40公开数据集中Chair和Stool点云模型和实测点云机械零件和汽车覆盖件点云模型为例对算法可行性进行验证,并与FCM算法、FAFCM算法、WPAFCM算法和MACWPAFCM算法进行对比。结果表明,对于4种点云模型,本文算法相比4种对比算法在以数值高为优的V_(PC)聚类性能指标上平均提高0.4%~11.95%,在以数值低为优的适应度函数值J_(m)、V_(PE)和V_(XB)聚类指标上分别平均减少0.2%~11.97%、0.65%~7.35%、0.3%~19.47%,在两种ModelNet40点云模型上平均迭代次数减少8~21次,在两种实测点云模型上平均迭代次数减少39~57次,表明本文算法收敛速度快,迭代次数少,聚类效果佳,具有更高的聚类准确性和更好的综合性能。展开更多
基金Project supported by the National Science and Technology Innovation 2030 Major Project of the Ministry of Science and Technology of China(No.2018AAA0101200)the National Natural Science Foundation of China(No.61502534)+1 种基金the Natural Science Foundation of Shaanxi Province,China(No.2020JQ-493)and the Domain Foundation of China(No.61400010304)。
文摘To address indeterminism in the bilevel knapsack problem,an uncertain bilevel knapsack problem(UBKP)model is proposed.Then,an uncertain solution for UBKP is proposed by defining thePE Nash equilibrium andPE Stackelberg-Nash equilibrium.To improve the computational efficiency of the uncertain solution,an evolutionary algorithm,the improved binary wolf pack algorithm,is constructed with one rule(wolf leader regulation),two operators(invert operator and move operator),and three intelligent behaviors(scouting behavior,intelligent hunting behavior,and upgrading).The UBKP model and thePE uncertain solution are applied to an armament transportation problem as a case study.
文摘针对模糊C-均值聚类算法用于点云分割时对初始值敏感且易于陷入局部最优,导致点云分割效果不理想,不稳定的问题。提出了一种基于曲率约束的改进狼群算法优化模糊C-均值聚类的混合算法(IWPAFCM)。该算法首先在狼群算法中引入佳点集初始化种群分布;然后利用自适应步长简化参数设定、平衡寻优与收敛时间;进一步应用交互策略增强狼群的内部交流,提升狼群全局寻优的能力;最后对头狼加入高斯扰动机制使其具有跳出局部最优的能力,将改进狼群算法(improved wolf pack algorithm,IWPA)得到的聚类中心作为模糊聚类的初始值进行迭代,由此得到准确的聚类中心。在此基础上,基于点云的法矢量和曲率对点云之间的距离进行定义并替换传统欧式距离,实现了理想的点云分割效果。以ModelNet40公开数据集中Chair和Stool点云模型和实测点云机械零件和汽车覆盖件点云模型为例对算法可行性进行验证,并与FCM算法、FAFCM算法、WPAFCM算法和MACWPAFCM算法进行对比。结果表明,对于4种点云模型,本文算法相比4种对比算法在以数值高为优的V_(PC)聚类性能指标上平均提高0.4%~11.95%,在以数值低为优的适应度函数值J_(m)、V_(PE)和V_(XB)聚类指标上分别平均减少0.2%~11.97%、0.65%~7.35%、0.3%~19.47%,在两种ModelNet40点云模型上平均迭代次数减少8~21次,在两种实测点云模型上平均迭代次数减少39~57次,表明本文算法收敛速度快,迭代次数少,聚类效果佳,具有更高的聚类准确性和更好的综合性能。
