5. Conclusion

Many malfunctions are the result of a poor estimate of the complexity of the problem to be solved: PID loops set independently when they are coupled, or conversely, an adaptive controller with overly sensitive criteria in an installation where a "robust" PID setting would be sufficient to combat frequent but small fluctuations. It's better to spend a little time on preliminary studies, but to be sure of the suitability of the equipment and settings to the problem at hand.

Knowledge of the physical system, whatever its origin, and even if it is limited to a qualitative aspect, is a valuable asset, as it facilitates validation of the model and enables us to assess its limits of use. It must be stressed that many loops only operate correctly around a nominal reference regime, which is why 2 criteria are sometimes used (e.g. PID and fuzzy) if wide variations are possible....