Certain practical considerations in the model-following method of decentralised control

In the paper a recent decentralised algorithm is tested on a practical nonstandard problem, i.e. the decentralised control of part of the energy cycle of a container ship. The problem is a realistic one and it is numerically rather ill-conditioned. The algorithm is modified in order to fit such nonstandard problems. The ship problem is used to clarify a number of issues which were left somewhat vague in the original algorithm.