In this chapter

Chapter 11 of 13

  1. 11.1Result
  2. 11.2Status of the requirements
  3. 11.3Limits of the statement
  4. 11.4What the test rig achieves nonetheless
All chapters
  1. 01Initial situation
  2. 02Objective and approach
  3. 03State of the art
  4. 04Requirements
  5. 05Solution approaches and concept decision
  6. 06System architecture
  7. 07The test rig
  8. 08The heated-tank model
  9. 09The simulation cycle
  10. 10Five test scenarios
  11. 11Result and limits
  12. 12Outlook
  13. 13Preliminary projects

Chapter 11 Verification

Result and limits

Fulfilled requirements, prototypical parts and limits of the statement

By Chapter 11 of 13 3 min read Updated: 30 September 2026

Summary

The outcome is a test rig at which the same controllers work with an unchanged program against a computational plant model. The functional requirements are met, the plugin interface is considered prototypical, the cycle time is only partially met. The most important limit: without measurement data from a real tank, the model has only been tested for internal consistency.

Result

In the end there is a test rig at which the same controllers work with an unchanged program against a computational plant model. Four things were created:

  • the setup itself with power supply and converter level (Chapter 7),
  • the signal bridge including firmware,
  • the software consisting of gateway, orchestrator and user interface (Chapter 6),
  • the heated-tank model with its derivation (Chapter 8), supplemented by twelve further plant models.

Status of the requirements

The functional requirements are met, with one restriction: the plugin interface does carry 13 models without any change to the orchestrator, but it is agreed upon as a convention and not formally specified; it is therefore considered prototypical. Of the quality requirements, the cycle time (NFA-02) is only partially met: the bound of 100 ms is kept on average but not assured (Chapter 9).

Limits of the statement

The most important limitation

Without measurement data from a real tank, the validation remains limited to internal consistency. What is verified is that the model calculates what the equations demand, not that the equations match a specific real tank.

Further limits in detail:

  • Numerics: Protective limits catch unphysical intermediate values but only apply within the intended operating range. In the pressure build-up (scenario 3), the calculated pressure fluctuates from step to step; the alarm threshold B4 is exceeded by a single peak (Chapter 10).
  • Scope of the tests: A passed run shows that the tested scenario works, not that every conceivable scenario works. The scenarios test the model, not the communication path to the controller.
  • Real time: The cycle measurement comes from the Windows development computer; a general-purpose operating system gives no real-time guarantee.
  • Electrical acceptance: The replacement of the miniature circuit breaker -F2 and the initial inspection by a qualified electrician are still outstanding (Chapter 7).

What the test rig achieves nonetheless

It makes errors visible that previously remained invisible. If the interlock of the heating is missing in the control program, the pressure rises in the model and B4 reports (scenario 3 shows this sequence in the model). That was exactly the starting point of the thesis (Chapter 1). What follows from this next is described in Chapter 12.