无约束多目标优化的一种新的拟牛顿法

基于在新拟牛顿方程形式下无约束单目标优化问题改进的拟牛顿法,提出了无约束多目标优化问题的一种新的拟牛顿法,同时在一定的假设条件下,结合Wolfe线性搜索准则,证明了算法具有全局收敛性和超线性收敛性,并进行了数值试验,结果表明,所提的新算法是正确和有效的,并能够迭代得到可使多个目标更优的临界点.

A New Quasi-Newton's Method for Unconstrained Multiobjective Optimization

Based on the improved quasi-Newton's methods for unconstrained optimization problems,we present a new quasi-Newton's method for multiobjective optimization without constraints,and prove its global and superlinear convergence under Wolfe line search.Numerical results also show that the proposed method is efficient,and better critical point can be obtained by iteration.