Automobile systems monitoring on the basis of diagnostic signal analysis

Abstract

The specificity of various mechatronic systems functioning in a modern car gives opportuinity to analyze diagnostic signals (oscillograms) of different nature, which are read from the diagnostics characteristic points or can be obtained from the on-board diagnostics regular system (OBD). Such analysis allows achieving a higher level of automation in searching forСерія: Технічні науки298malfunctions causes in case of certain number of diagnostic parameters received presence. Therefore, it is important to improve the system for collecting, storing and processing such diagnostic information.The paper describes the principle of diagnostic signals (oscillograms) database forming. Reference oscillograms and oscillograms of malfunctions are saved. The basic approach is that the diagnostic information is stored as separate ranges (fragments) of oscillograms in the form of fourth-order polynomials coefficients. In addition, the database contains objects of typical diagnostic features, diagnostic parameters and structural parameters, as well as possible changes of structural parameters i.e. malfunctions. Between all these objects there are «many to many» connections. As you know, when there are typical malfunctions in diagnostic signals, certain anomalies appear. Interrelations between database tables characterize the relationship between typical malfunctions and anomalies of diagnostic signals fragments. The degree of influence of the malfunctions causes on the behavior of the diagnostic signal is determined by the weighting coefficients. The determination and correction of these coefficients is the basis of the diagnostic model developed on the basis of a neuro-fuzzy network.Also the general architecture of the monitoring system is described, which is based on the implementation of diagnostic automation approaches with the implementation of intelligent-expatcher monitoring and modern IT technologies. Architecture is based on the principle of three levels. Each level is encapsulated and has an accessible communication interface. The first level contains the models of the database entities and the access functions to the database, the second level contains the system's logic and diagnostic models, the third level, that is the level of visualization and communication with the user.