Considering the problem of the imprecise matching of the parallel simulation entity model under the condition of incomplete and imprecise information for the battlefield situation, an intelligence driven dynamic generation method for parallel simulation entity is proposed. Firstly, a componentized model system for a parallel simulation entity is established,and a model description method based on metadata is proposed. Then, the running status feature data from realtime intelligence is extracted, the rules of battlefield entity behavior are established, and the dynamic matching between the simulation model and intelligence data is realized based on rules. Then on the basis of these, the information interaction relationship table between parallel simulation entity model components is set up. This table is updated through the simulation middleware in real time, and the description information and the executive body of components are dynamically loaded, realizing the purpose of the parallel simulation entity dynamically generated. Finally, take the joint operation of a region as an example, use the proposed method to dynamically match the simulation model and combine parallel simulation entity model components, in order to verify the effectiveness of the proposed method.