ABSTRACT
In this work we investigated the concept of desirability functions for a many-objective optimization of a hybrid car controller with five objectives from different domains. We study this problem from the perspective of preference expression. Specifically we are looking at the impact of wrongly defined desirabilities and how this can be corrected using a Graphical User Interface (GUI). Overall we find that a desirability-based many-objective optimization approach could be well suited for real-world problems with objectives from many domains as it is becoming more and more common in industrial settings.
- Dimo Brockhoff, Johannes Bader, Lothar Thiele, and Eckart Zitzler. 2013. Directed multiobjective optimization based on the weighted hypervolume indicator. Journal of Multi-Criteria Decision Analysis 20, 5--6 (2013), 291--317.Google ScholarCross Ref
- EC Harrington. 1965. The desirability function. Industrial quality control 21, 10 (1965), 494--498.Google Scholar
- Ben Paechter, RC Rankin, Andrew Cumming, and Terence C Fogarty. 1998. Timetabling the classes of an entire university with an evolutionary algorithm. In International Conference on Parallel Problem Solving from Nature. Springer Berlin Heidelberg, 865--874. Google ScholarDigital Library
- Tobias Rodemann, Kaname Narukawa, Michael Fischer, and Mohammed Awada. 2015. Many-objective optimization of a hybrid car controller. In European Conference on the Applications of Evolutionary Computation. Springer, 593--603.Google ScholarCross Ref
- Tobias Wagner and Heike Trautmann. 2010. Integration of preferences in hypervolume-based multiobjective evolutionary algorithms by means of desirability functions. Evolutionary Computation, IEEE Transactions on 14, 5 (2010), 688--701. Google ScholarDigital Library
- K Wonggasem, T Wagner, H Trautmann, D Biermann, and C Weihs. 2013. Multi-objective Optimization of Hard Turning of AISI 6150 Using PCA-based Desirability Index for Correlated Objectives. Procedia CIRP 12 (2013), 13--18.Google ScholarCross Ref
Recommendations
A set-based genetic algorithm for solving the many-objective optimization problem
Many-objective optimization problems are very common and important in real-world applications, and there exist few methods suitable for them. Therefore, many-objective optimization problems are focused on in this study, and a set-based genetic algorithm ...
Approximate non-dominated sorting for evolutionary many-objective optimization
Non-dominated sorting has widely been adopted and shown to be very effective in dominance based evolutionary multi-objective optimization where the number of objectives is two or three. In dealing with many-objective optimization problems, where the ...
Objective reduction for many-objective optimization problems using objective subspace extraction
Multi-objective evolutionary algorithms (MOEAs) have shown their effectiveness in exploring a well converged and diversified approximation set for multi-objective optimization problems (MOPs) with 2 and 3 objectives. However, most of them perform poorly ...
Comments