In the existing multi-period robust optimization method of active and reactive power coordination in distribution networks, the reactive power regulation capability of distributed generators (DGs), operation costs of regulating equipment, and the current of shunt capacitance of cables are not taken into account. In this paper, based on branch flow equations, a multi-period two-stage mixed integer second-order cone programming (SOCP) robust optimization model of active and reactive coordination in distribution system with cables is developed considering the reactive power regulation capability of DG, action costs of switched capacitor reactor (SCR), on load tap changer transformer (OLTC), energy storage system (ESS), and current of shunt capacitor of cables. Further, against to the deficiency of low computational rate of column and constraint generation method (CCG), a novel method iteratively solving the first and second stage model on cutting plane is proposed. In the first stage model, the number of variables and constraints keeps constant during the iteration. In the second stage model, it only needs to solve the model of each single time period. Then their results are accumulated. Thus, the computational speed using the proposed method is much higher than CCG. The effectiveness of the proposed method is separately validated on the 4-bus, IEEE 33-bus, and PG69-bus distribution systems.