一类非可微广义分式规划的非完全Lagrange函数与鞍点最优性准则

Incomplete Lagrange function and saddle point optimality criteria fora class of nondifferentiable generalized fractional programming

对于一类目标函数中含范数‖Bx‖p的非可微广义分式规划,给出了一个新的非完全Lagrange函数,并利用已有的最优性必要条件,在一类广义(F,α,ρ,d)-凸性的条件下,证明了鞍点最优性准则。

For a class of nondifferentiable generalized fractional programming problems with the norm \$‖Bx‖\-p\$ in the objective function involving, a new incomplete Lagrange function is given, and the saddle point optimal criteria are proven by using the existing necessary optimality conditions, under the assumptions of the class of generalized (\$F,α,ρ,d\$)-convexity.