With the large-scale access of intermittent renewable energy, the active distribution networks (ADN) is facing great challenges on achieving full consumption and efficient use of intermittent renewable energy on the premise of ensuring safe operation. To resolve these issues, a novel multi-period robust dispatch strategy for ADN based on scene analysis method is proposed in this paper. Then, the second-order cone relaxation (SOCR) is used to transform the original model into a mixed integer second-order cone programming (SOCP), which gains as much improvement as possible in computational performance at the expense of as few losses as possible in accuracy. Finally, numerical test on the modified IEEE 33-bus test system demonstrates that on the premise of realizing the maximum utilization of DG, the proposed strategy can also achieve the purposes of energy-saving, loss reduction and improvement in voltage level.