Abstract: Abstract： The frequent of the monitoring signal impact working efficiency of the dispatchers，which causes carelessness and endanger the security of power grids . This research proposed scheme of controlling methods. By using Dynamic decision algorithm，alarm threshold is planned flexibly; By using Q-learning AI algorithm, the fundamental reason of the frequent monitoring signals is analyzed, and the solution is automatically given. Case shows the research is of great practical value, which will provide instructive reference for power grid dispatch. 0 Introduction Some scholars classify the causes of frequent signals to substation primary and secondary equipment. Some scholars have proposed solutions from the substation level，including strengthen the planning, procurement and other standards of equipment and so on; Some scholars have proposed solutions from the grid monitoring system level, including a set of delay shielding strategy is developed for Different signal types. The above research can reduce the number of frequent monitoring information from many aspects, but it is lack of intelligence and flexibility. This study introduces dynamic decision AI algorithm (DD) and Q-learning AI algorithm into the controlling methods of frequent monitoring signal. The expert opinion database can be established to identify the frequent information in real time, and to put forward the solutions intelligently and flexibility. Research findings will achieve the purpose of managing the frequent monitoring signal from the source automatically. 1. Model for monitoring information Model for monitoring information includes m type signal, as shown in the Formula 1-2 and Figure 1. 2. Dynamic decision AI algorithm the electric network supervisory signal has the characteristic of dimensional and multi-phase, in order to assess the alarm threshold of the frequent monitoring signal in real-time, the Dynamic decision is used. As shown by the formula 3-5, if the formula 6 are met, the system gives the alarm. 3. Q-learning AI algorithm In order to analysis of the frequent causes of monitoring information intelligently, Q-learning AI algorithm is used to analyze historical monitoring signal data and give the fundamental reason of signal frequency in real time. As shown by the formula 6-7. 4. The frequent monitoring signal control system Based on data of OMS system, a database is set up. The system includes Frequency information precontrol module and Expert experience analysis optimization module. System framework as shown on pic 2. 5. Case analysis In this paper, monitoring of a certain regional distribution is taken as a case, the frequent monitoring signal includes AVC latch, Relay communication fault, Fault of microcomputer device. Analysis results proved that by using the method proposed in the paper, the frequent signals in 2017 was significantly reduced, The Comparison of above three kinds of frequent signals between 2016 and 2017 is shown in Figure 3. The signal number and threshold curve of the three kinds of signals at a certain substation in a certain month in 2017 are shown in Figure 4. This research introduces Dynamic decision AI algorithm (DD) and Q-learning AI algorithm into the management of frequent monitoring signal in regional distribution, which has the characteristic of intelligence and flexibility. 1) flexibility: To attract the attention of the dispatcher in early stage, day threshold is analyzed according to the actual situation of the frequent information daily. 2) Intelligence: Based on the historical database of the frequent monitoring signal, Q-learning is used to judge the fundamental reason intelligently. The research can effectively reduce the number of frequent signals intelligently and flexibly, which will provide guidance for power grid dispatch automatically.