Multi-objective optimization of the design of two-stage flash evaporators: Part 2. Multi-objective optimization

摘要

Flash evaporation process is currently developing in the wine industry where it is used for flash-cooling or concentration. The design of flash evaporators is faced with specific constraints and must take into account multiple design objectives. In this paper, the development of a multi-objective optimization method is investigated for the joint optimization of design objectives such as process transportability, environmental efficiency, operative cost or cooling power. The optimization method is based on the aggregation of design objectives through desirability functions and indexes. Desirability functions are suitable for formulating design constraints more precisely than inequality relations and, moreover, the global design model results in an unconstrained optimization problem. However, aggregation methods do make it difficult to compute the global optimum of the design problem. This difficulty has been addressed by developing a distributed genetic algorithm which is not so sensitive to this type of numerical solving difficulty. Another difficulty arises from the weighting method for the aggregation of desirability functions since weight parameters have no physical meaning. This weighting problem is approached through a sensitivity analysis of the weight parameters and by observing their relative influence.