With the rapid development of cold chain logistics, the carbon emissions from fuel-powered refrigerated vehicles are increasingly exacerbated. However, the poor timeliness of cold chain distribution is resulted from the range limit of electric refrigerated vehicles, thereby enterprise development is impacted. In order to reduce distribution costs for cold chain logistics enterprises while saving energy and reducing emissions, the problem of multi-energy vehicle-based cold chain distribution and optimal charging and changing modes for electric refrigerated vehicles were investigated by constructing a cold chain path optimization model which focused on selecting multi-energy vehicles and optimizing charging and changing methods. To effectively solve the model, a modified adaptive large neighborhood search(MALNS) algorithm was proposed along with relevant damage and repair operators. The experimental results demonstrate that the model not only effectively reduces carbon emissions, but also optimizes distribution costs, thereby facilitating the sustainable development of enterprises, and optimal costs can be searched efficiently by the MALNS algorithm. The research results hold significant guiding implications for cold chain logistics enterprise distribution.