Research On RBC Test Sequence Generation And Optimization Method Of Train Control System Based On Timed Automata

In recent years,with the rapid development of social economy,high-speed railroads between Chinese cities have been opened one after another.In order to improve the train running speed and ensure its safety,experts have designed the Chinese Train Control System Level 3(CTCS-3),which is suitable for Chinese railroads,through the scientific decision of introduction,digestion,absorption and reinvention.CTCS-3 level column control system consists of two parts: ground equipment and on-board equipment.Among them,the Radio Block Center(RBC)is the core subsystem of the ground equipment that controls the safe operation of trains by generating travel permits from train information provided by other equipment.In the actual field,in order to ensure that the RBC system meets the requirements of the technical specifications,the testers will conduct comprehensive simulation tests on the RBC to ensure its functional integrity before the trains are officially opened.At present,most of the RBC simulation tests are conducted using manually generated test sequences,which not only has a long test cycle,but also makes it difficult to guarantee the completeness and validity of the tests.Therefore,in order to improve test efficiency,reduce test cost and improve the scientific and validity of test sequences,it is of strong practical significance to conduct research on test sequence generation and optimization for RBC systems.The thesis adopts a model-based testing method to formally model typical functional scenarios of RBC systems.After verifying that the model meets the requirements of the wireless occlusion center specifications,the model file is parsed,traversed and concatenated to generate test sequences that meet the testing requirements,and finally the model-generated test sequences are optimized with an improved firefly algorithm.The main research content of this thesis is as follows:Firstly,the functional characteristics and technical requirements of the RBC system are analyzed according to the technical specifications and the requirement specifications,and a formal modeling of the travel permit generation and RBC switching scenarios is performed using UPPAAL,a modeling tool based on the theory of timed automaton.The dynamic behavior of the model is then analyzed by simulation timing diagrams generated by the simulator,while the security and restricted activity of the model are verified using the validator.Through analysis and verification,both the travel permit generation model and the RBC switching model meet the functional attributes required by the above two scenarios in the wireless occlusion center technical specification.Secondly,the file format of the time automaton model is analyzed,and the model file is firstly parsed,followed by traversal of the model file using the full variation coverage criterion and depth-first search algorithm to obtain the test case set.Finally,test cases are concatenated to generate test sequences according to the test case concatenation process and test sequence generation principles.In order to achieve automatic test sequence generation and improve test automation,we use the QT Creator tool to design an application for automatic test sequence generation.When the model file is loaded in the application,the model state,migration and other information can be automatically extracted,and the required test sequence can be automatically generated by selecting the start and end states of the test sequence and the critical state.Finally,for the problem of high redundancy of the test sequences generated by the model,the improved Firefly algorithm is selected to optimize the generated test sequences.By comparing with the test sequences generated by the model,it is found that the optimized test sequences effectively reduce the repetition rate of the sequences while ensuring the coverage of the original test sequences and reducing repeated tests,and the improved firefly algorithm outperforms the traditional firefly algorithm in terms of convergence speed,which can effectively improve the testing efficiency.