Study On Key Issues Of System Level Testability Design

With the rapid development of equipment,practical usage requirement in real battle and the reform of logistic support mode,equipment test and diagnosis have to face higher requirement and larger challenge.Thus,the reform of test mode is imperative and equipment testability engineering is right the product of these changes.At present,the testability design has become perfect in the general sense of engineering,but for specific complicated systems,it still faces unsettled problems,such as non-consideration of integrated faults and system-level test,and difficulty in system-level fault diagnosis isolation.Therefore,supported by the ministry-level pre-research projects,this paper focuses on the system-level testability design problems and difficulties in complicated systems.Firstly,the form mechanism and influences of integrated fault are analyzed.Based on the mechanism analysis,the key techniques of system-level testability design are systematically studied,which includes the testability allocation,testability modeling,test selection and diagnosis strategy at the system-level.The main research contents and innovations are as follows:1.Considering the integrated faults in the testability analysis of complicated system,we proposed the concept and definition of the integrated fault.The integrated fault is classified according to its producing level and causes.The influence of integrated fault on the system testability is analyzed with the fault mechanism.Focusing on the fault mechanism of connection-type integrated fault,we analyze the influence of integration fault on testability allocation and fault diagnosis.Further,the connotation,classification of integration test is studied as well as the influence mechanism of integration test on testability allocation and diagnosis,which lays the theoretical foundation for the key technologies of system-level testability such as testability allocation,test selection and diagnosis of complex equipment.2.The current one-time index allocation method cannot provide reasonable system-level allocation result for complex equipment.To solve this problem,a two-time iteration index allocation method is proposed.For the initial allocation,an integrated fault rate calculation method is proposed,which takes the integrated fault as the allocation index of the equivalent unit.Meanwhile,an improved fault rate allocation method based on the arc tangent function is proposed.For the second allocation,a fault rate to quantitatively describe the degree of mutual detection between units is defined.Used as variable of the allocation function,an allocation function with arc tangent form are theoretically derived and the index allocation is realized by considering unit mutual measurement.This allocation method integrates system-level scheme design and index allocation,and effectively solves the unreasonable allocation problem of testability index in complex equipment system caused by integration failure and unit test.3.Since the existing testability models and test selection methods are not suitable for the system-level test selection of complex equipment,a mixed relevance matrix model is established by using the initial allocated unit-level fault detection rate replacing the "1" data in the conventional relevance matrix,which solves the lack of the unit information when built the system testability model;based on this,this research proposes the prediction method of the system-level fault detection rate and system-level fault isolation rate based on mixed relevance matrix,and puts forward the prediction method of integrated fault detection rate,establishes the mathematical description model of system-level test selection based on the mixed relevance matrix and testability“three rates”,adopts Binary Genetic Algorithm to get the solution and realizes the system-level test selection better.4.For the diagnosis strategy design problem under the situation of insufficient and unreliable system test of complicated equipment,this paper proposes one diagnosis technology integrating three kinds of diagnosis inference methods,which adopts the merits of different inference methods to realize complementation and enhance the fault diagnosis ability.Specifically,a verification method of test effectiveness the uncertain situations of missing and wrong reports is provided to solve the reference problem in face of uncertain test.For parallel diagnosis and multi-fault diagnosis inference method based on fault directional graph the ambiguity group problem is made which uses hidden information in the directional graph to complete diagnosis solves the diagnosis inference problem in face of insufficient test.5.Taking an Aircraft Automatic Flight Control system as the object,the testability design and diagnostic strategy design were carried out by using the new technology route and method proposed in this paper.The verification results show that the system-level testability analysis,modeling,BIT design and diagnostic strategy design proposed in this paper can greatly improve the testability level and ability of complex systems,and solve the problem that the current testability design and diagnostic reasoning cannot meet the engineering practice requirements.In summary,through the analysis of the influence mechanism of integrated fault to integrated test,this paper focuses on the studies of the key issues of system-level index allocation,testability modeling and analysis,test selection and diagnosis strategy design,which effectively solves the difficult testability design analysis and diagnosis problems in complicated equipment.The techniques proposed here can reach a better testability design for huge complicated system,which has important academic and engineering practical value.