Large Relational Database Query Optimization Technique Research Based On Genetic Algorithm

The rapid development of the database techniques and the wider usage of information systems not only improve management efficiency but also result in the accumulation of massive amounts of data, the data in some large-scale systems, such as securities system, the bank system and the population data system has exceeded TB level. From most of the current systems applications, query operation takes the largest proportion of operations in databases; it will take a long time to query,which means the query speed has limited the system performance. However, the ordinary algorithms have intrinsic limitation when dealing with the query optimization problems. So the paper uses genetic algorithm in the large relational database to solve the query optimization problems.Main work is as follows:①Analyze the current research results in this field both in domestic and oversea industry, and give the necessity of the query optimization.②Emphasize on the structure of the query optimizer, the function of each model and analyze functions of the overwriting and planning optimization.③Analyze the principle of multi-join query and the theory of execution engine, research ordinary research algorithms and summarize their features.④Analyze the theory and key techniques of genetic algorithm and modify the varietal operator by using two varietal operators to solve the problem that crossover operator generats low new offsprings; then compare the genetic algorithm with other algorithms.⑤Design a model of multi-join query optimization based on the genetic algorithm according to the feature of genetic algorithm and test the model.The result proves that the query efficiency is highly improved by using genetic algorithm compared with other algorithms.