Abstract: Fully exploiting the load-side controllable resources through the load aggregators（LAs） is an important way to realize the congestion management in the distribution network. Firstly, in response to the problem of huge differences of LAs in several aspects, such as scale, reliability and congestion improvement effects, this paper proposes a multiple congestion management method for distribution networks based on optimization classification for LAs. Then, three indicators are defined to measure the reliability level of LA congestion management, such as congestion management efficiency, contribution and default rate, which are prioritized by the grey fixed weight clustering. In order to take into account the uncertainty of distributed power supply, an opportunity constraint planning model of congestion management in the distribution network is established. The model is solved by a stochastic simulation particle swarm algorithm, taking into account the reliability and economy of congestion management.Finally, the effectiveness of the proposed method is verified by case simulation.