This paper firstly analyzes the small and medium-sized source-grid-load-storage system and mathematically models the system structure based on the power supply model, the grid model, the load model and the energy storage model. Secondly, based on blockchain technology, the multivariate coordinated scheduling objectives and constraints of small and mediumsized source-grid load-storage systems are modeled through the Cosmos platform, and the model is optimized and solved by the Gray Wolf optimization algorithm. Finally, the impact of multivariate coordinated scheduling on small and medium-sized source-network load storage systems is examined using the parameters of the example and the micro-network design. The results show that compared to the total operating cost before optimization, the total operating cost decreased by 1163.7 yuan. In terms of the peak-valley difference of load electricity consumption, the peak-valley difference before the implementation of the demand response is 0.48, the peak-valley difference of load after the implementation of the demand response is 0.35, and the real-time residual capacity of the energy storage device is always maintained between 60 and 250 kW. The multivariate dispatch model based on this study has a positive impact on small and medium-sized distribution grids with source-grid load-storage systems.