对变分不等式问题,定义信赖域子问题的最优值函数,研究最优值函数的性质,并在单调的条件下,利用最优值函数为可行解进行一个局部误差估计。
For the variational inequality problems, we define an optimal value function of trust region subproblems and study its properties. And then under monotonic conditions,we use the optimal value function as the feasible solution to provide a local error estimate.