SOCEMO: Surrogate Optimization of Computationally Expensive Multiobjective Problems

Abstract

We present the algorithm SOCEMO for optimization problems that have multiple conflicting computationally expensive black-box objective functions. The computational expense arising from the objective function evaluations considerably restricts the number of evaluations that can be done to find Pareto-optimal solutions. Frequently used multiobjective optimization methods are based on evolutionary strategies and generally require a prohibitively large number of function evaluations to find a good approximation of the Pareto front. SOCEMO, in contrast, employs surrogate models to approximate the expensive objective functions. These surrogate models are used in the iterative sampling process to decide at which points in the variable domain the next expensive evaluations should be done. Therefore, fewer expensive objective function evaluations are needed, and a good approximation of the Pareto front can be found efficiently. Previous algorithms have generally been tested on problems with few variables (up to 10) and few objective functions (up to 5). In our numerical study, we show that our algorithm performs well for benchmark problems with up to 35 dimensions and up to 10 objective functions, as well as two engineering application problems. We compared the performance of SOCEMO to a variant of NSGA-II and show that SOCEMO’s sophisticated search strategy is more efficient than NSGA-II when the number of allowable function evaluations is low. The online supplement is available at https://doi.org/10.1287/ijoc.2017.0749 .

Analysis of energetics and economics of sub‐ambient hybrid post‐combustion carbon dioxide capture

“…The cycle employed for the VPSA process consists of four steps: (1) Light product pressurization, (2) adsorption, (3) co-current blowdown, and (4) counter current evacuation. To compute Pareto frontiers efficiently, the VPSA-cycle simulation is treated as a black-box and the multiobjective optimization algorithm "Surrogate Optimization of Computationally Expensive Multiobjective Problems" (SOCEMO), developed by Müller, 56 was implemented in Matlab. 57 This optimizer allows computation of Pareto frontiers at a fraction of the computational cost required for other commonly applied genetic algorithms.…”

Evolutionary multiobjective optimization: open research areas and some challenges lying ahead

“…2. Surrogates When using this approach, an empirical model that approximates the real problem is built through the use of information gathered from actual objective function evaluations [43,121,157]. Then, the empirical model (on which evaluating the fitness function is computationally inexpensive) is used to predict promising new solutions [157].…”

Evolutionary Many-Objective Algorithms for Combinatorial Optimization Problems: A Comparative Study

“…Although solving a MOOP using exact algorithms such as cutting-plane [13], polynomial-time approximation scheme [14], branch and bound [15,16], and branch and cut [17] has been attempted, most MOOP methods are heuristics and meta-heuristics. For example, evolutionary computation [18][19][20][21], Pareto ant colony optimization [22], decomposition methods based on Lagrangian relaxation [23], diversity maximization approach [24], and zigzag search [12] have been employed for addressing MOOPs. Some meta-heuristic algorithms are well-suited for solving global optimization problems such as non-convex and discontinuous problems [25][26][27].…”