In order to solve the low-carbon vehicle routing problem under time-dependent network, the relationship between discrete vehicle travel speed and continuous vehicle travel time is analyzed. According to the principle of ``first in first out'', the calculation method of road travel time across time periods based on time division is designed. The calculation function of carbon emissions is employed by considering vehicle speed, real-time load of vehicle, vehicle travel distance and road slope. The mathematical model of the low-carbon time-dependent vehicle routing problem(LCTDVRP) is established with the goal of minimizing the total carbon emissions. The traffic congestion index is employed to design the traffic congestion avoidance approach. An improved ant colony algorithm is designed according to the characteristics of the LCTDVRP model. The experimental results show that the proposed approaches can effectively avoid traffic congestion, shorten vehicle travel time, reduce vehicle carbon emissions, and promote the harmonious development of logistics distribution and ecological environment.