On the use of Principal Component Analysis for reducing the number of environmental objectives in multi-objective optimization: Application to the design of chemical supply chains

Multi-objective optimization (MOO) has recently attracted an increasing interest in environmental engineering. One major limitation of the existing solution methods for MOO is that their computational burden tends to grow rapidly in size with the number of environmental objectives. In this paper, we study the use of Principal Component Analysis (PCA) to identify redundant environmental metrics in MOO that can be omitted without disturbing the main features of the problem, thereby reducing the associated complexity. We show that, besides its numerical usefulness, the use of PCA coupled with MOO provides valuable insights on the relationships between environmental indicators of concern for decision-makers. The capabilities of the proposed approach are illustrated through its application to the design of environmentally conscious chemical supply chains (SCs).

► We present a novel method for solving MOO with a large number of objectives. ► We apply it to the design of supply chains under economic and environmental criteria. ► Redundant environmental metrics are identified. ► The number of objectives is consequently reduced.