This paper presents a technique that incorporates preference information within the framework of multi-objective evolutionary algorithms for the solution of many-objective optimization problems. The proposed approach employs a single reference point to express the preferences of a decision maker, and adaptively biases the search procedure toward the region of the Pareto-optimal front that best matches its expectations. Experimental results suggest that incorporating preferences within these algorithms leads to improvements in several quality criteria, and that the proposed approach is capable of yielding competitive results when compared against existing algorithms.