Unsuitable wind farm layout design influences wind energy conversion efficiency as well as wind farm noise radiation. In the past, more attentions were paid on high energy efficiency wind farm design, but little progress was achieved on noise constrained wind farm optimization. The present work aims at developing a numerical optimization framework that combines the modelling of wind farm wake and wind farm noise with the annual energy production in an optimization algorithm. Firstly, a wind turbine wake model is implemented in the optimization framework to obtain the detailed flow field of a wind farm. Secondly, the noise radiation around a wind farm is simulated with an advanced method based on solving the parabolic wave equation. Thirdly, the annual energy production of a wind farm is calculated with respect to a prescribed wind speed probability distribution. Finally, wind farm layout optimization is carried out with a cost model built on the levelised cost of energy and an efficient iterative approach. According to the results, the developed optimization framework is capable of seeking a trade-off between maximum energy yield and minimum negative environmental effect.