Abstract: In order to compute traffic capacity of hybrid channels which changes from one-way to two way navigation,a quantitative model is developed based on clustering of ship types.According to the regulations of ship navigation in seaports,different ship types are classified.Characteristics of ship types and their probability distribution are determined by using a cluster algorithm.Compared with traditional methods,this method is more accurate to quantify traffic capacity of channels by assigning weights of different types of ships according to their occupied spaces.Moreover,the influences of different navigation modes on traffic capacity are also considered,which makes the model more accurate.The Dagusha Shipping Channel in Tianjin Port is taken as a case study to verify the proposed model.Current capacity of this channel is around 1 905 ships per month,while computed capacity is 2 350 ships per month.Moreover,while using the model for hybrid channels,the result is 4 455 ships per month,which is significantly larger than actual capacity of the channel.The results show that the method based on clustering of ship types can reflect the actual situation of the channel.