I'm working on a flight data analysis project. the flight data is represented in a tabular format. Each quarter of a second, we have the status of different parameters including turobreactor parameters and avionic parameters. I intend to use an expert system to analyse the flight data in order to detect anomalies during the flight. for example T4 (temperature) shouldn't surpass 750 °C over 30 seconds. Is the expert system architecture appropriate to such task?
Asked
Active
Viewed 148 times
1 Answers
1
Every expert system consists of the knowledge base and the inference engine.
If you are going to to use the expert system architecture:
you have to make sure that you have this knowledge gathered from factual and heuristic knowledge. Those are the rules, mostly consisting of an
IFpart and aTHENpart.how you will apply this rules, is defined by the inference engine - the problem-solving model, where the common paradigm is chaining of
IF-THENrules (e.g. forward chaining and backward chaining).
Now answering your question, to me your example looks like a specification of a discrete cyber-physical system (Depending on other specifications can be considered hybrid too). A cyber-physical system can also be considered as a state machine which is a system that exists in a limited number of conditions and has forbidden states and progresses from one state to the next according to a fixed set of rules. In addition, if you had possible input and output events in your example, you could design Moore, Mealy machines, Petri Nets, Statecharts of your state machine, given the specifications and then use formal verification techniques to verify it.
