In order to improve the forecasting and warning ability of sluice pumping station monitoring, considering the multi-factor affecting the safety of sluice pumping station and the huge amount of monitoring data, a method for mining association rules of sluice pumping station monitoring data was proposed, which combines the cloud model with the improved Apriori algorithm. Firstly, the attribute space of monitoring data was established, and the soft partition of attribute space was realized by using the reverse cloud model, and the monitoring data were discretized. In order to avoid the disadvantage of long time taken for multi-scanning database with the traditional Apriori algorithm, the binary was used to storage data and frequent itemsets were got through ‘and operation’. Considering the dynamic increase of data, an incremental update method was proposed to update rules. Finally, the monitoring data analysis of a sluice pumping station in a city was taken as an example to verify the effectiveness of the proposed method.