With the development of many new technologies in aircraft manufacturing area and the increasing competition of the global market, aircraft manufacturing enterprises have to reduce their production time and increase the cost-efficiency, with the consideration of high speed response to the changes inside enterprises or in the environment. Production scheduling is a significant process in manufacturing, especially for complicated part or component processing. This paper proposes an agent based multi-objective optimization approach for production scheduling based on Genetic Algorithms. It aims to minimize the total production cost and simultaneously reducing the emission released during production, and the delivery time and equipment constraints are satisfied as well. The new approach is tested in a virtual plant for turbine blade manufacturing. Experimental results show that a group of Pareto optimal solutions are obtained, which can be provided to the decision maker of the manufacturer to select according to different actual conditions.