For the problems of low storage efficiency, insufficient writing and query performance of relational databases and time-series databases when processing high-frequency time-series data in the forging industry, an efficient industrial real-time time-series massive measurement point data storage method was proposed. The method innovatively used JSON format for single-column full-storage strategy, effectively circumventing the column number limitations of traditional databases and reducing the storage space occupancy. By using the underlying JSON operator logic of the database for data query and retrieval, the query efficiency of time-series data was improved, especially showing significant advantages in large-scale data scenarios. Experimental results demonstrate that compared with the traditional methods, this method improves the writing efficiency by 5.6 times and saves 170 times of storage space consumption under large data volumes, and has a performance advantage of 3-5 times in data query speed.