| Because of the rapid development of modern science and technology, highly integrated equipments are widely used in both daily life and professional area. Highly integration not only improves performance and reduces cost of equitment, but also makes it harder to maintain. In order to improve the maintainability and reliability of complex system, testability should be considered at early stage of design.Optimal Analysis is based on Huffman coding, it ensures the generated test sequence is optimal. To achieve highest testability index, expensive tests are applied by Optimal Analysis and test cost of system soars. To solve this problem, the proposed Constrained Analysis algorithm can generate optimal test sequence with testability indexes higher than given target, which can efficiently reduce test cost. It is suitable for exorbitant test system to cut down test cost. AO* based Unreliable Test Analysis takes unreliable test into account. It can effectively drop misdiagnosed cost of inaccurate test systems.The paper also proposes the optimization of visual testability model layout.Traditional testability softwares need manual modeling, making system designers burdened and easy to make mistakes. With the help of model information input software,model layout optimization can automatically generate testability model and make the model concise and easy to maintain. At last, the implementation of the introduced optimized testability algorithms are described in detail and the software framework of optimized algorithms integration is introduced.The proposed Constrained Analysis algorithm is experimental verified, proving that it has advantages over traditional algorithms. Also, the proposed testability model layout technology and the introduced algorithms of this paper are integrated in the self-developed testability software, which has been applied in engineering practice. The utilization of the software indicates the practicalbility of the proposed optimized algorithms of this paper. |
PDF Full Text Request