Extension of the Plant Feature Approach Introducing Temporal Relations

Testing of controllers with a large number of inputs and outputs suffers from the curse of dimensionality in terms of the combinatorial explosion of possible input combinations. In particular for conformance testing, this results in unbearable large numbers of test cases. To mitigate this problem, the plant feature approach has been proposed, which guarantees full coverage of the nominal behavior of the controller. This nominal behavior corresponds to the reachable state space when combining in a closed-loop the controller with its fault-free plant. Plant features are identified as intuitive basic knowledge about a system based on its physical behavior. Considering such physical limitations allows to reduce the scope of testing to the actual relevant behavior.