METHOD OF SOFTWARE VERIFICATION OF AIR OBJECTS CLASSIFICATION FUZZY LOGICAL SYSTEM
Article Sidebar
Main Article Content
Abstract
Objects of different classes are detected in the process of monitoring airspace. The classification of an air object is the process of establishing its belonging to a preassigned class. Classes are automatically determined or set automated. The unambiguous assignment of air objects to a particular class is an actual scientific task. The purpose of the article is to develop a method for verifying software for a fuzzy logical system for classifying air objects. This problem is solved in a fuzzy setting. To solve this problem, an appropriate software verification method has been developed. The method is based on fuzzy colored Petri nets and uses a base of fuzzy productional rules. The structure of the fuzzy network verification model has been developed. The basis of the model is a fuzzy colored Petri net for representing the base of fuzzy production rules for classifying air objects. For the convenience of visualization of the fuzzy network verification model, the interpretation of the elements of the fuzzy colored Petri net is introduced. The analysis of the state space of a fuzzy network verification model reflects all possible markings. The state space allows to obtain the values of the indicators of all the basic properties of the Petri net. The CPN Tools modeling system is used to build and analyze the state space. The full standard report for the fuzzy logical classification system of air objects was obtained from the simulation results. The report fragment with conclusions about the correctness of the model is given. The report contains sections of state space statistics - the number of nodes, arcs and status, indicators of the properties of reversibility, limitation, survivability and fairness of transitions. The method includes five steps. 1. A base of fuzzy production rules is being developed. 2. The set of interpretation rules transforms the base of fuzzy production rules into the form of fuzzy colored Petri nets. 3. The model is examined for proper functioning. 4. When an error is detected, its type is analyzed. After its correction, the program repeats, starting with any of stages 1, 2 or 3. 5. Reports on the total space of states with various combinations of source data are issued. A final report is issued after analyzing the correctness of the set of reports and correcting errors that have occurred.
Article Details
References
Timochko, A.A. (2018), “Representation of dynamic interacting processes in the state space of a fuzzy colored Petri net for verification of a software product of a fuzzy logical classification system”, Science and Technology of the Air Force of Ukraine, No. 4(33), pp. 82-89.
Timochko, A.A. (2018), “Development of fuzzy productive rules of a fuzzy logical classification system of air objects in the process air space control and a generalized algorithm of their realization”, Systems of Control, Navigation and Communication, PNTU, Poltava, No. 5(51), pp. 33-37, available at: https://doi.org/10.26906/SUNZ.2018.5.033.
Hasegama, Т., Furuhashi, Т., Uchikama, Y. (1996), “Stability analysis of fuzzy control systems based on Petri nets“, Proc. Int. Discourse on Fuzzy Logic and the Management of Comflexity, FLAMOC'96, pp. 191-195.
Nasareth, D.L. (1993), “Investigating the applicability of Petri nets for rule-based system verification“, IEEE Trans. Software Eng., 4, pp. 402-415.
Polat, F., Guvenir, H. (1993), “UVT: A unification-based tool for knowledge bast verification“, IEEE Expert, 8, pp. 69-75.
Zaitsev, D.A., Shmeleva, T.R. (2006), Simulating Telecommunication Systems with CPN Tools, ONAT, Odessa, 60 p.