Research On Storage And Quick Retrival Of Real-Time Data

With the extensive application of real-time simulation system, the real-time simulation supporting system appears which is mainly used to manage data generated by real-time simulation system. The nuclear power real-time simulation system generates massive real-time batch data, the real-time data need to be stored in limited time and the quickly retrieving the records from the massive historical data is very time-consuming. In the dissertation, will research and implement storage and fast retrieval of the real-time data, which is generated by nuclear power real-time simulation system.Based on the research of theory and related technology about real-time database, main memory and historical database, the dissertation proposes real-time data management mechanism which uses the object as the unit of memory allocation and recovery, and the object pool technology as the container to manage memory objects. Through improving the existing mechanisms for real-time database index, a new multi-index mechanism is proposed, which combined the tree index and hashing index. In the dissertation, real-time data retrieval model and the algorithm RTRLD (real-time retrieved algorithm based on Large Data) are proposed. Besides, in order to improve retrieval performance of real-time historical data, many optimization methods are used. After reference the design and implementation of open source memory database and the data processing flow of the existing nuclear real-time simulation platform, the real-time data storage and quickly retrieving system focus on the real-time nuclear power simulation platform is finished.By the relevant experiments, the performance of real-time data storage and retrieving data from massive data records are tested respectively. The experimental results and theoretical analysis prove that the new strategies and algorithms for real-time data storage and retrieval are improved in performance In the last part, the author's research is listed, from the prospect of real-time database applications and technical, the problems and direction to further study are pointed out.