IPTV System Fault Expert System Design

Automatic diagnosis fault of IPTV system has always been considered a hard nut in expertsystem. This paper reports some pioneering research on a possible reasoning algorithm and itsapplication to the fault diagnosis of IPTV system. In light of the IPTV system fault diagnosis, thispaper designed the knowledge representation of production rule, the log analysis describelanguage to match the log files’ pattern, the way of forward and backward reasoning, the fuzzyreasoning based on deterministic theory, the wizard-style human-machine interface base onspeech recognition and text-to-speech technology, the distributed structure of client/server. Asshown by tests, the system constructed on the basis of the production rule model designed in thispaper is basically capable of diagnosing some simple situation of fault.IPTV fault diagnosis expert system knowledge and inference engine design, uncertaintyhandling several aspects of system design. Initial realization of the choice of operating systems,architecture, development tools, database, and gives three simple examples.IPTV fault diagnosis expert system knowledge representation, we compared the expertsystem knowledge representation: semantic network notation, frame notation, productionrepresentation. IPTV network characteristics, production rules is the best option.In the IPTV fault diagnosis expert system inference engine design, this paper due todeductive reasoning, production systems of reasoning, forward reasoning algorithm, the reverseanalysis of the characteristics of inference algorithms and other methods, based on the thoughtprocesses of human experts, the mechanization of this process, formalized the log analysisdescription Language used to describe the characteristics of the log file, and these strings to theknowledge base in the title table.IPTV fault diagnosis expert system, uncertainty handling, combined with other methods ofprobability theory and fuzzy set theory, using a kind of inexact reasoning method, commonlyreferred to as the CF model. CF-based model system, knowledge is generally expressed as: E.For the initial realization of the difficulty, from a cost, to achieve ease of use, performance ofmultiple considerations, made a choice of operating systems, architecture, development tools,database analysis. And combined with my existing experimental conditions, gives three specificexamples of the Windows+Access+VB6+C/S. The automatically generate staff IPTV Fault diagnosis reasoning for the report.