Multiobjective optimization including design robustness objectives for the embodiment design of a two-stage flash evaporator

Abstract

During the early phases of design processes, designers have to take the best decisions in order to converge quickly toward the most preferred design solution. More precisely, in a preliminary design context, the phase of embodiment design consists in determining the best values for the main dimensioning parameters of the system to improve its performances. In general, these performances must satisfy functional or technical requirements related to design objectives. The balancing act between the satisfaction of constraints and objectives can be treat as a multiobjective optimization problem. Although designers often face to a qualitative evaluation of design solutions, difficulty arises from his ability to express preferences and expectations throughout the objective function of the multiobjective optimization problem. Illustrated through the preliminary design of a two stage flash evaporator, the purpose of this article aims to develop a methodology to formulate both constraints and preferences into the optimization model. Thus design solutions are evaluated with a single global indicator of confidence. Flash evaporators are thermal systems which are mainly used for flash cooling and juice concentration applications. The design of such a system has to respect specific constraints related to these areas of applications and should meet multiple design objectives. Several goal-oriented design solutions are discussed, in particular when low design sensitivity is expected.

Tracking the consequences of design decisions in mechatronic Systems Engineering

“…For example, a dynamical model of the wheelchair can compute the 'Current consumption' from the values of iDDPs such as 'nominal torque of the engine', 'diameter of the wheel', 'reduction ratio of the gearbox' and others impacting iDDPs. In general however, T is not explicitly known and, in the most favorable case, it can be approximated by a predictive model [33].…”

Robust design of flash evaporators for must concentration applications in wine production

“…Basseur in [32] or Greiner [33], have introduced uncertainties into multiobjective optimization. In a recent study [34], we proposed to formulate the sensitivity of the design as a particular design objective to be traded-off. However, this approach doesn't allow designers to express a compromise between the performance and the sensitivity while, these objectives must be obviously balanced according to the designers' expectations.…”