Tham khảo tài liệu 'micro electronic and mechanical systems 2009 part 14', kỹ thuật - công nghệ, cơ khí - chế tạo máy phục vụ nhu cầu học tập, nghiên cứu và làm việc hiệu quả | 446 Micro Electronic and Mechanical Systems property of the given configuration are removed. Then all that parts of the EFSM that cannot be traversed are eliminated. This strategy removes all that constraints that are useless and could also avoid the propagation of observability on the primary outputs. In fact the new generated EFSMs are pruned by all that transitions are not needed to model any possible behaviour that can lead to fault observability. Note that the complexity of the proposed algorithm is linear with the number of transitions and that above all the number of transition is limited and is not affected by the state explosion problem with the EFSM model. Fig. 16. Non-optimized EFSM example. let E be an EFSM and let ob_reg be the observed register reduce_efsm register ob_reg efsm E retrieve the sets of transitions and states of E T transitions_set E S states_set E for each transition t in T delete a transition if the conditions that preclude observability are matched if write_on_register t ob_reg and not read_register t ob_reg delete_transition t _ delete_transition transition t retrieve the in-going state for transition t state s in_going_state t to delete a state it must be different from the state of the current configuration and there are no other transitions except the current one t going into it if not is_configuration_state s and number_of_ingoing_transitions s 1 delete_state s remove transition t from E Fig. 17. Algorithm for removal of EFSM transition precluding observability. Test Generation based on CLP 447 Then another method is proposed in Section to reduce the complexity problem for the solver. The idea is that part of the work performed by the solver to find a sequence could be done earlier and then invoke the solver on the reduced constraints set. In this case no constraints are actually removed but a part of the solution is provided to the solver. In such way different constraints are already satisfied at beginning and .
