With the construction of Global Energy Internet, power system has the trend of the scale expansion, the network complexity, the equipment precision and data massivication. Substation automation and power grid equipment level are continuously improved as well, but the traditional power grid dispatcher training simulation system (DTS) has been unable to adapt to the increasingly precise secondary electric equipment simulation requirements. A Big Data analysis based new method for power grid dispatch and control training simulation first uses ETL tool to extract and standardization process the data of secondary equipment and signals they emit. Then do the equipment type association rules mining to equipment and signal data after preprocessing, and correlation matching in accordance with association rules. At last, classified equipment data will be divided and load according to the condition of area, substation, bay and voltage grade, and drive the equipment detailed simulation logic. The method based on the actual operation monitoring data, uses semantic analysis and association rule technology and ETL and ElasticSearch tools to implement the grid primary and secondary equipment signal extraction, parsing, mining and load, so that operation monitoring equipment detailed simulation logic can be driven and the authenticity, accuracy, adaptability and precision of simulation can be improved.