Abstract: In view of the great challenges brought by the uncertainty of renewable energy (RE) output such as wind and solar to the planning and operation of distribution networks, a multi-objective bi-level planning model of distribution network based on improved generative adversarial network and carbon footprint is proposed. First, Wasserstein generative adversarial network with gradient penalty (WGAN-GP) model is used to generate a large number of scenes of wind and solar outputs, and scenes reduction is carried out by K-medoids clustering algorithm. Next, the life cycle approach is used to determine the carbon footprint of all kinds of power generation technologies. Then, this paper constructs a bi-level planning model of the distribution network considering the carbon footprint. In the upper layer, the planning scheme of distributed generation (DG), energy storage system (ESS) and capacitor banks (CB) are considered to minimize the annual comprehensive cost. In the lower layer, operation strategy of on-load tap changer (OLTC), reducible load, CB, ESS and DG are considered to minimize operation costs, voltage offset and carbon emissions. Then, the bi-level model is converted to a single layer model by associating and unifying the upper and lower layers of the model, and normalized normal constraint (NNC) method is used to solve the multi-objective model. Finally, the effectiveness of the model is verified by the IEEE 33 node system.