Abstract
The paper deals with the efficiency study of the classifier developed based on the probabilistic neural network for multi-class diagnostics of a complex spatial object in the presence of multi-site damage. For recognition, the multidimensional diagnostic feature vector is used, the values of the features may have a deviation of ±5 % for the defect-free condition of an object, and exceed the permissible deviation in case of occurrence and development of damage. For the vector containing 5 diagnostic features, 6 classes of technical condition of an object are substantiated. Formation of sets of training and test input vectors, used for the classifier training and testing is performed. In order to evaluate the multi-class recognition efficiency, the coefficient, which is a percentage of the probability of correct classification of test vectors, is used. The analysis of the dependence of the efficiency coefficient on the characteristics of the classifier and the set of training vectors is carried out. It is found that error-free multi-class recognition of the object condition over the entire set of input vectors with different values of deviation of feature elements is provided in the range of values of the classifier parameter spread of [0,02; 0.07]. It is revealed that the greater the diagnostic feature deviation in test vectors, the greater the influence of the dimension of the set of training vectors on the multi-class recognition efficiency. The minimum size of the set of training vectors (68 vectors) and the limit value of diagnostic feature deviation in test vectors (17 %), which provide error-free multi-class recognition by the developed classifier are determined.
Highlights
Ensuring the reliability and efficiency of operation of complex spatial objects is a topical issue in the aviation, power, oil and gas indries, as well as for special-purpose engineering structures
Design of structural elements of such objects is based on the principle of safe damage, which allows for a microdefect, but such that does not lead to efficiency loss and object destruction [1,2,3]
When developing neural network classifiers, it is important to analyze the effect of the dimension of the set of training diagnostic feature vectors on the classification accuracy in order to determine the possibility of error-free multi-class recognition at a certain minimum number of training vectors
Summary
Ensuring the reliability and efficiency of operation of complex spatial objects is a topical issue in the aviation, power, oil and gas indries, as well as for special-purpose engineering structures. The presence of welded or rivet joints of structural elements of complex spatial objects poses a threat of the emergence and development of multi-site damages This may lead to destruction characterized by a sudden and rapid propagation due to combining among themselves and absorbing small-size cracks. In order to ensure safe and effective operation of such objects, it is necessary to provide multi-class diagnostics for timely detection of damage, assessment of its extent, monitoring of its development and interaction on large-sized surfaces of complex spatial objects. This will contribute to ensuring the reliability and efficiency of operation, preventing the destruction of complex spatial objects and averting catastrophic consequences
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
