Abstract: Training by tank driving simulator is an important way of improving the operating skill of equipment. The statistical analysis method is difficultly used to find the knowledge and rules from the complex training data in simulation training. A mining method of association rules based on bidirectional concept lattice is proposed to analyze the simulation training results of tank driving. The original data table is transformed into single value formal background by using Boolean matrix operation to complete the multiple value data processing of tank driving simulation training. Connotative rank and denotative rank are defined to search the concept nodes from the top and bottom of concept lattice, and the redundant nodes are reduced by combining the support thresholds. Post rule is designed as constraint condition to filter the irrelevant frequent item set, and extract the association rules that meet the mining target of users. The experimental results show that the association rules mining method based on bidirectional concept lattice has obvious advantages in run time and generating frequent item set. It is applied to analyze the training results of a tank driving simulator to extract the valuable reduction association rules, which further verifies the feasibility and effectiveness of the method.

Key words: tank, drivingsimulator, bidirectionalconceptlattice, associationrule, simulationtraining