The Vague optimization method sorted out is a special case of the pattern recognition method of the Vague set. Its specific application steps are as follows: 1 the set-up of characters to determine the evaluation; 2 the establishment of a collection of new varieties of wheat to be optimized; 3 the extract of the ideal set of new wheat varieties ; 4 the construction of Vague environment to collect all varieties of Vague sets; 5 the Vague optimization: to calculate similarity measures between Vague sets to obtain the new wheat varieties based on numerical similarity measures and to propose the similarity measures formula between the Vague sets. This optimization formula is supported by Vague class techniques. Breeding new varieties of wheat with comprehensive quality traits is one of the directions in wheat breeding. It is a new attempt to study the wheat assessment of the new wheat varieties with the application of optimization methods. The wheat assessment case of the new wheat varieties shows that both the formula and the method are practical.