This paper proposes a dynamic electric vehicle path optimization method for intelligent garbage collection and transportation systems to address the increasing uncertainty of urban household waste and the carbon emissions caused by vehicles in traditional garbage collection and transportation processes. With the goal of minimizing total cost and maximizing garbage collection and transportation, adjust the soft time window in a timely manner based on whether the amount of garbage generated by the intelligent garbage bin reaches the collection and transportation threshold. Taking into account the position of the garbage bin to be cleared, the position of the garbage collection and transportation vehicle, the amount of electricity, and the loading capacity, dynamically adjust the vehicle path. Due to the fact that genetic algorithms require a long time to converge to an acceptable solution, this paper uses greedy genetic algorithms to comprehensively utilize their advantages in local search and global search capabilities to achieve faster convergence solutions. The research results indicate that compared with traditional static garbage collection and transportation modes, dynamic garbage collection and transportation schemes can reduce total costs and increase garbage collection and transportation volume, and reasonable planning of the path of electric collection and transportation vehicles can effectively reduce carbon emissions. Finally, the effectiveness and efficiency of the algorithm were verified through comparative analysis of numerical examples.