Functional testing of measurement-based control systems: An application to automotive

• We propose a methodology for the functional testing of measurement-based systems.
• Effective statistical techniques for generating plausible test cases are proposed.
• The method takes also into account the time interdependence among input quantities.
• It was applied to test a real automotive system in order to verify its suitability.

The number of measurement-based embedded systems increases continuously in automotive, avionic, defense, and communication industries. In order to ensure the system’s compliance with requirements, adequate solutions for functional testing are needed. The paper describes an original approach for the functional validation of measurement-based complex control systems. The main focus is on the automatic generation of realistic continuous data sets (test stimuli) within the operational ranges of the measurement system under test, by applying enhanced statistical techniques, which are able to reproduce both timing and data dependencies. A thorough case study using an air fuel ratio control system from the automotive domain is performed to assess the approach. Results indicate that the proposed method is suitable for automated functional testing of embedded control systems within a Software-in-the-Loop test environment.