Abstract: Pointing at the issues of randomness, volatility and intermittence of output brought by high proportion of wind and photovoltaic access, as well as insufficient coordination of source grid load storage and conventional optimization methods in traditional dispatching mode, this paper establishes a robust stochastic multi-objective optimal dispatching model of source grid load storage coordination, uses entropy weight-fuzzy satisfaction method to objectively distribute objective weights, and puts forward an improved chaotic adaptive sparrow search algorithm for solution. According to the IEEE 33-bus system, and combining with real data, simulation outcomes indicate that the strategy we put forward can make the system operation cost decrease by 8.3%, raise the new energy consumption rate by 13.5%, cut the power grid loss to below 4.2%, maintain frequency and voltage within the safe and stable scope, hence the convergence speed and solution precision of the algorithm are obviously superior to the traditional particle swarm optimization algorithm. It can offer theoretical and engineering reference to the efficient, robust and economic dispatching of high proportion new energy electric power systems.