An Unconstrained Approach for Instrumentation Network Design and Upgrade

This paper proposes an unconstrained optimization approach to the design and upgrade of sensor networks. The proposition is motivated by the fact that current methods require threshold values for different instrumentation network attributes, which are in turn very difficult to determine. In fact, these attributes are, after all, related to the process economics and therefore should not be arbitrarily constrained. The methodology replaces a recent trend in viewing the problem as a multi-objective one. In this sense, the unconstrained objective can be considered as the proper utility function of the multiobjective problem.