Multi-source data equivalence methods for testability integrated evaluation

In the whole life cycle of equipment, multi-source testability demonstration test data (TDTD) which may be acquired at different stages, in different environment and at different levels (system, subsystem or replaceable unit), can be used for testability integrated evaluation. However, the above multi-source data cannot be used for testability integrated evaluation directly for they share many distinctions according to their physical and mathematical relations. To transform these different types of data into system-level binary data as we need, five equivalence methods are studied based on the prior information of multiple sources. That is, the equivalence methods for expert data based on weight factor, preexposure test data based on approximate processing model, growth test data based on growth factor, replaceable unit data based on structure function and virtual test data based on entropy theory in the field of testability evaluation. Validation of the methods is presented via application examples. Results indicate that our methods can enlarge the amount of data used for testability evaluation indirectly, and also provide readers with a comprehensive reference.