Preferences in Evolutionary Multi-Objective Optimisation with Noisy Fitness Functions: Hardware in the Loop Study
|
|
- Marian Cross
- 5 years ago
- Views:
Transcription
1 Proceedings of the International Multiconference on ISSN Computer Science and Information Technology, pp PIPS Preferences in Evolutionary Multi-Objective Optimisation with Noisy Fitness Functions: Hardware in the Loop Study Piotr Woźniak Technical University of Lodz, Control Theory Group, 18/22 Stefanowskiego St., Lodz, Poland Abstract. Multi-objective optimisation (MOO) is an important class of problem in engineering. The conflict of objectives in MOO places the issue of compromise in a central position. Since no single solution optimises all objectives, decision-making based on human preference is a part in solving MOO problems. In this paper application of the evolutionary MOO to the dynamic system controller design by use of the hardware in the loop is presented. Thanks to this approach problems of un-modelled plant dynamics and uncertainty of parameters are alleviated because no mathematical model is needed. The a-priori search of one solution does not require knowledge of a whole Pareto front. 1 Introduction Real-world optimisation problems are characterised by the existence of multiple, often conflicting, objectives. In the absence of information about the relative importance of the various criteria, such problems typically admit multiple optimal solutions where any improvement in one objective can only be obtained at the expense of degradation in other objectives. Edgeworth and Pareto captured this notion mathematically in the criterion widely known as Pareto optimality. The general MOO problem may be stated as follows. Without loss of generality, the minimisation of objectives is considered in the discussion throughout this paper. Definition 1 Find the vector of decision variables X = {x 1, x 2,, x n} which satisfies the m inequality constraints: p equality constraints: g i X 0, i=1,2,., m h i X 0, i=1,2,., p and minimises the k objective functions: f X =[ f 1 X f 2 X f k X ]. In general the objectives are non-commensurable and in conflict with each other. 337
2 338 Piotr Woźniak Definition 2. A vector of decision variables X j dominates X i in the Pareto-optimality sense if and only if X j performs better in at least one objective and at least as good as X i in the rest X i X j f n j f n i for 1 n k and f m j f m i for m [1, k ], s.t. m n. If neither of the two candidate solutions dominates the other, they are nondominated. The Pareto-optimal set of a MOO problem consists of solutions which dominate other solutions in the feasible region. The image of this set in the objective space is the Pareto-front. The immediate effect, when applying the criterion of Pareto optimality to highdimensional problems, is that a large number of solutions within any generation are non-dominated and, therefore, qualitatively indistinguishable from each other. It is empirically evidenced in [7, pp ]. Therefore, when the number of objectives is large, Pareto dominance-based ranking procedures become ineffective in sorting out the quality of solutions and selection operators play a minor role to compensate for this effect, which translates into a lack of selective pressure. The dimensionality reduction techniques may be applied to alleviate this issues. This approach takes advantage of possibility that some objectives may behave in a non-conflicting manner near the Pareto-optimal region although a conflict exists elsewhere. In such cases, the Pareto-optimal front will be of dimension lower than the number of objectives. This technique, has been recently successfully applied to the EMOO electric synchronous machine design [1, 2]. Another possibility, typical for engineering approach to the MOO, is by search for set of solutions in a region of interest to the DM without approximating points on the complete high-dimensional Pareto-optimal front. Thus, the final solution of the Evolutionary MMO (EMOO) problem results, not only from the search process, but also from the decision process. As such, this paper provides a basis for online solutions selection with hardware in the loop (HiL) setup with Permanent Magnet Synchronous Machine (PMSM). The DM preferences in the EMOO are discussed in the next Section. In Section 3 the robustness issues of the EMOO of the HiL setup. Design problem is introduced in Section 4 along the hardware overview. Results of the research are presented in Section 5. Section 6 concludes the presentation. 2 Issues in the EMOO algorithms Evolutionary algorithms are global parallel search and optimisation methods based around Darwinian principles, working on a population of potential solutions to a problem. Every individual in the population represents a particular solution to the EMOO problem. The population is evolved over a series of generations (iterations of the algorithm) to produce better solutions. The quality of the solution is evaluated by the fitness assignment i.e. functions which characterises the individual s performance in the problem domain. Most of published results on EMOO algorithms base on explicit objective functions. In this paper we consider the approach when each fitness evaluation is run on the hardware (as introduced in Section 4).
3 Preferences in Evolutionary Multi-Objective Optimisation 339 The relative degree of the fitness value determines the level of propagation of the individual s genes to the next iteration. In the EMOO algorithm NSGA-II [7] in use here, the rank of a certain individual corresponds to the number of individuals in the current population by which it is dominated. All non-dominated individuals are assigned rank zero, while dominated ones are penalised according to the population density of the corresponding region of the trade-off surface. 2.1 Preferences in the EMOO Solutions on the Pareto-optimal set perform better in some objectives and worse in other. Since no single solution optimises all objectives, Decision-Making (DM) based on subjective human preference is a natural aspect in solving EMOO problems. In real problems only a set of solutions in a region of interest to the DM is required. One may consider introducing some component to overall fitness measure for each solution. This preference related fitness component zeroes on a subset of the Paretooptimal solutions considered of high relevance to the DM for the particular EMOO problem. The vagueness and context-dependence of the DM s preference structure result, unsurprisingly, in various mathematical models as well as techniques proposed for the incorporation of preferences into EMOO [3-5, 8, 10]. The integration of the DM preferences in the EMOO process must allow the selection. Two different classes of methods can be defined, taking into account the way how a search and decision processes are integrated with EMOO [3]: i) A-priori methods: The DM preferences are defined in terms of an aggregating function which combines objectives values into a single utility value. ii) A-posteriori methods: The approximation of the Pareto optimal set is supplied to the DM, which selects the most preferred among the alternatives using e.g. Weighted Metrics approach [6]. In this paper we consider the a-priori approach based on Reference Point Method [8]. We consider it as most suitable in the problem under investigation although for this methodology it is necessary to know beforehand the criteria space, and also the DM needs to have some knowledge about the location of regions of the best solutions. No need for the whole Pareto front approximation is the main advantage over all a-posteriori approaches. 2.2 Reference Point Method In the Reference Point Method the aim is to reach a Pareto front region located near a specific pre-defined reference point or points. This method can be applied using or not a weight vector. The main ideas of this method are the following: i) Solutions closer to the reference points, in the objective space, are to be more preferred, ii) Solutions within a ε-neighbourhood to a near-reference point solution are to be preferred, in order to maintain a diverse set of solutions near each reference point.
4 340 Piotr Woźniak To allow the identification of the solutions located near the reference point the normalised Euclidian distance (d ij) between each solution of the best nondominance level and the reference point is calculated 2 k f i (X) z i ij = (1) i max min i= 1 f i f i d w max min where : f i and f i are the maximum and the minimum value of the objective function for criterion i, and zi and w i are the i-th component of the reference point and weight vector, respectively. The distance (1) takes into account simultaneously the relative importance of the different criteria, quantified through the weight vector, and the distance between the solution and the reference point. The solutions with lower distance to the reference point and, simultaneously, with high weight vector, will be selected preferentially. This procedure was been implemented [9] through the modification of the NSGA-II algorithm [5] niching strategy: Step 1: For each reference point, the distance (1) of each solution of the front is calculated and the solutions are sorted in ascending order of distance. This way, the solution closest to the reference point is assigned a rank of one, Step 2: After computations in Step 1 are performed for all reference points, the solutions with a smaller preference distance are preferred in the tournament selection, Step 3: To control the extent of obtained solutions, an ε - neighbourhood idea is used in the niching operator. First, a random solution is picked from the nondominated set. Thereafter, all solutions having a sum of normalised difference in objective values of ε or less from the chosen solution are assigned an artificial large preference distance. This way, only one solution within an ε-neighbourhood is emphasised. Then, another solution from the nondominated set is picked and the above procedure is performed. The use of the ε-based selection strategy ensures a spread of solutions near the preferred Pareto-optimal regions. The main change in NSGA-II was made on the crowding operator, since in this case the aim is not to obtain a great diversity of solutions along the Pareto frontier but only a subset of these solutions that minimise the distance to the reference point. 3 Robustness issues in the EMOO based control system design The performance of a particular control design is fundamentally tied to the accuracy of the model upon which it is based. This is especially true for iterative control design and optimisation procedures. The substitution of Hhardware in the loop (HiL) for the software model in the EMOO opens up new possibilities for design based on real world performance indices. In real optimisation problems, a wide range of uncertainties have to be taken into account. Generally, uncertainties in EMOO can be categorised into four classes. Each
5 Preferences in Evolutionary Multi-Objective Optimisation 341 of them is presented with additional comments concerning the HiL setup to be presented in Section 4.2. i. Noise; in the HiL it comes from measurement errors from sensors (e.g. the Sensors Block from the system presented on Fig. 2), ii. Errors in fitness approximations based on mathematical modelling; introducing the HiL approach we eliminate this type of uncertainty, iii. Time-varying parameters of fitness; in the considered HiL based control system design changes in environment may be neglected because in this study evaluations are available in less than 1 [s] that is in time shorter than all time constants in the system [11], iv. Perturbations of parameters and environmental parameters after the optimal solution has been determined; this type is not considered in this paper, however it may be easily included in the HiL by the appropriate design of experiments sampling approaches. In general noisy or uncertain scenario the proximity of other fitness values, even if only close on one objective, can influence how the rank is assigned. Limits on objectives, constraints on the chromosomes, and sharing can all be implemented easily within this ranking framework, allowing interactive DM with any noisy systems possible. At the current stage of research no disturbance was introduced. 4 Hardware in the Loop in the EMOO of the Permanent Magnet Synchronous Motor speed control design The structure of a PI speed controller considered it this paper is given by 1 t u(t)= k pe( t ) e( )d (2) T i 0 where u(t) is the output of the controller at time t, e(t) is the error between the actual and the desired speed of the motor shaft, k p, T i are constants of the controller, and decision parameters of the EMOO problem. Depending on the choice of k p, T i the response of a controller to a given error signal will vary significantly. Various methods exist to tune the gains of a PI controller off-line to attain the prescribed transient response and steady-state error criteria [12] because there is no analytical way of finding the optimal set of constants (k p, T i). Empirical methods such as Ziegler-Nichols tuning can be used. Up-to-date search methods generally involve some form of iterative approach to achieve performance criteria such as risetime, overshoot and settling-time [12]. For that reason it was decided to use a EMOO approach to find several possible parameter settings that would result in good performance.
6 342 Piotr Woźniak 4.1 Objective functions When a controller is designed for a specific system it is often desired that multiple design requirements are met. The drive system is no exception. When evaluating the performance of the PI controller parameters for the speed control loop a step response is considered and different performance measures are investigated [11]. The fitness functions to be minimised are of the form: i) the speed tracking error denoted as ITAE [12] ITAE e (t) tdt (3) p ii) composite objective based on the time period when output current of the converter is on constraints. This objective (denoted hereafter as I constr) is related to the power consumption of the system [11]. Int(i) i(t) dt (4) In this way, it is required the search be performed basing on two noncommensurable objectives. They may be visualised in form of fitness landscapes over constrained parameters space as presented on Fig.1. The values for the sample visualisation are obtained from a sequence of step response tests for 0 t 1 [s]. Fig. 1. Fitness landscape for objective functions (3) and (4). 4.2 Hardware overview In the application under consideration here, the software implemented PI controller to be optimised (Fig. 2) performs tracking of a speed demand for the PMSM. It is achieved by calcultion of a current demand for Power Module. Its value depends on the shaft speed (from the Speed Sensor), and the Load (from the Manometer).
7 Preferences in Evolutionary Multi-Objective Optimisation 343 Software Hardware-in-the-Loop EMOO Algorithm dspace Control System Power Module PI MLIB NO STOP Conditions YES Position and Speed Sensors Results Decision Maker MATLAB Kollmorgen PMSM Momentometer DutymAx DS Load Fig. 2. On-line PI speed controller EMOO design with the hardware in the loop The actual view of the experimental installation is presented on Fig. 3. Fig. 3. Actual view of the experimental installation. The measurement of fitness functions were realised in each iteration of the EMOO as a set of step response tests. There is one test for each member of population under consideration. Thanks to the symmetry in the PMSM operation it is possible to alter the sign of successive tests. It enables cutting down the time of experiments. Sample sequence of six, 0,8 [s] long tests is presented on Fig. 4.
8 344 Piotr Woźniak Fig. 4. Registered sequence of six 0,8 [s] long, speed step-response tests with altering sign for v desired = 25 [rad/s] t 4.3 EMOO setup The parameters of the optimisation algorithm NSGA-II implemented during EMOO of the problem defined by (2)-(4) under constraints: i) in the parameters space defined by inequalities 0,25 < k p < 20 ; 25< 1/T i < 200, ii) in the H-i-L installation (not presented; for details see [11]), are presented in Table 1. Table 1. Basic parameters of the NSGA-II algorithm Algorithm parameter Value Population N 60 Generations G 50 Pool size N/2 30 Tour size 2 Crossover 0.9 probability Mutation probability Results The result of measurement s noise influence on the EMOO speed control design is presented after normalisation to [0,1] on Fig. 5 in form of a Pareto set. This approximation obtained using NSGA-II [4] without DM preferences modifications (as discussed in Section 2.2) presents significant influence of noise which prevents even spread of non-dominated solutions along the front. To alleviate this effect further research on relation between the bounded noise and quality of the whole Pareto set approximation is needed.
9 Preferences in Evolutionary Multi-Objective Optimisation 345 Fig. 4. Single solution selected from the Pareto front approximation (stars) with all elite solutions from 100 iterations. There are several techniques for selection of one exact, or with some acceptable tolerance, Pareto-optimal solution. For this study the Reference Point Method modified NSGA-II algorithm is used to select single Pareto-optimal solution k p=2.01, 1/T i=29.3 [1/s]. The step response for final values gives the speed step response presented in Fig.5. Fig. 5. Comparison of speed step response for the selected single Pareto-optimal solution (bold line) and four extreme decision pairs. Further research is needed to evaluate influence of the measurement noise on the radius of solution set under DM preferences.
10 346 Piotr Woźniak 6 Conclusions This paper investigates the design of a dynamic controller for the drive system directly onto hardware. A multi-objective evolutionary algorithm is applied to the task of controller development, while an objective function ranks the system response to find the Pareto-optimal set of controllers. The measurement noise present during each evaluation at run-time deteriorate the quality of Pareto set approximation. To prevent decline of results the Decision Maker preferences are included in the EMOO process in form of the a-priori type Reference Point Method. This approach demonstrates very good properties providing the designer the set of optimisation parameters in the area surrounding the reference point. References 1. Woźniak P., Witczak P.: Dimensionality Reduction in Evolutionary Multiobjective Design of Permanent Magnet Generator, In: Proc.of Fourth Int. Conference on Evolutionary Multi- Criterion Optimization EMO 2007 (2007) Woźniak P.: Dimensionality Reduction in Evolutionary Multiobjective Design: Case Study, In: Proc.of ACM-SIGEVO on Genetic and Evolutionary Computation GECCO 07 (2007) Coello C.A.: Handling Preferences in Evolutionary Multiobjective Optimization: A Survey, in Proc. Congr. Evolutionary Computation (2000) Cvetkovic D., Parmee I. C.: Preferences and Their Application in Evolutionary Multiobjective Optimization, IEEE Trans. on Evol. Comput., 6, 1 (2002) Branke J., Deb K.: Integrating User Preferences into Evolutionary Multi-Objective Optimization, in Yaochu Jin (editor), Knowledge Incorporation in Evolutionary Computation, Springer (2005) Miettinen K. M.: Nonlinear Multiobjective Optimization, Kluwer, Boston (1999). 7 Deb K.: Multi-Objective Optimization using Evolutionary Algorithms, Wiley, Chichester (2001). 8 Wierzbicki A. P.: The use of reference objectives in multiobjective optimization, In Fandel, G., Gal, T. Eds, Multiple Criteria Decision Making Theory and Applications. Springer, Berlin (1980) Deb K., Sundar J., Bhaskara U., Chaudhuri S.: Reference Point Based Multi-Objective Optimization Using Evolutionary Algorithms, ISSN International Journal of Computational Intelligence Research, 2, 3 (2006) Ferreira J. C., Fonseca C. M., Gaspar-Cunha A.: Methodology to Select Solutions from the Pareto-Optimal Set: A Comparative Study, In: Proc.of ACM-SIGEVO on Genetic and Evolutionary Computation GECCO 07 (2007) Woźniak P., Sobieraj T.: Multi-objective control design by experiment-based evolutionary methods : case study, approved for presentation at the Conference on Control for Power Electronics and Electrical Drives, SENE 07 (2007). 12 Ogata, K.: Modern Control Engineering, 3 rd Ed., Prentice Hall, 1997.
Finding a preferred diverse set of Pareto-optimal solutions for a limited number of function calls
Finding a preferred diverse set of Pareto-optimal solutions for a limited number of function calls Florian Siegmund, Amos H.C. Ng Virtual Systems Research Center University of Skövde P.O. 408, 541 48 Skövde,
More informationA Distance Metric for Evolutionary Many-Objective Optimization Algorithms Using User-Preferences
A Distance Metric for Evolutionary Many-Objective Optimization Algorithms Using User-Preferences Upali K. Wickramasinghe and Xiaodong Li School of Computer Science and Information Technology, RMIT University,
More informationRecombination of Similar Parents in EMO Algorithms
H. Ishibuchi and K. Narukawa, Recombination of parents in EMO algorithms, Lecture Notes in Computer Science 341: Evolutionary Multi-Criterion Optimization, pp. 265-279, Springer, Berlin, March 25. (Proc.
More informationMultiobjective Optimisation. Why? Panorama. General Formulation. Decision Space and Objective Space. 1 of 7 02/03/15 09:49.
ITNPD8/CSCU9YO Multiobjective Optimisation An Overview Nadarajen Veerapen (nve@cs.stir.ac.uk) University of Stirling Why? Classic optimisation: 1 objective Example: Minimise cost Reality is often more
More informationMulti-objective Optimization
Jugal K. Kalita Single vs. Single vs. Single Objective Optimization: When an optimization problem involves only one objective function, the task of finding the optimal solution is called single-objective
More informationMulti-objective Optimization
Some introductory figures from : Deb Kalyanmoy, Multi-Objective Optimization using Evolutionary Algorithms, Wiley 2001 Multi-objective Optimization Implementation of Constrained GA Based on NSGA-II Optimization
More informationComparison of Evolutionary Multiobjective Optimization with Reference Solution-Based Single-Objective Approach
Comparison of Evolutionary Multiobjective Optimization with Reference Solution-Based Single-Objective Approach Hisao Ishibuchi Graduate School of Engineering Osaka Prefecture University Sakai, Osaka 599-853,
More informationFinding Knees in Multi-objective Optimization
Finding Knees in Multi-objective Optimization Jürgen Branke 1, Kalyanmoy Deb 2, Henning Dierolf 1, and Matthias Osswald 1 1 Institute AIFB, University of Karlsruhe, Germany branke@aifb.uni-karlsruhe.de
More informationReference Point-Based Particle Swarm Optimization Using a Steady-State Approach
Reference Point-Based Particle Swarm Optimization Using a Steady-State Approach Richard Allmendinger,XiaodongLi 2,andJürgen Branke University of Karlsruhe, Institute AIFB, Karlsruhe, Germany 2 RMIT University,
More informationEvolutionary Algorithms: Lecture 4. Department of Cybernetics, CTU Prague.
Evolutionary Algorithms: Lecture 4 Jiří Kubaĺık Department of Cybernetics, CTU Prague http://labe.felk.cvut.cz/~posik/xe33scp/ pmulti-objective Optimization :: Many real-world problems involve multiple
More informationI-MODE: An Interactive Multi-Objective Optimization and Decision-Making using Evolutionary Methods
I-MODE: An Interactive Multi-Objective Optimization and Decision-Making using Evolutionary Methods Kalyanmoy Deb and Shamik Chaudhuri Kanpur Genetic Algorithms Laboratory (KanGAL) Indian Institute of Technology,
More informationFinding Sets of Non-Dominated Solutions with High Spread and Well-Balanced Distribution using Generalized Strength Pareto Evolutionary Algorithm
16th World Congress of the International Fuzzy Systems Association (IFSA) 9th Conference of the European Society for Fuzzy Logic and Technology (EUSFLAT) Finding Sets of Non-Dominated Solutions with High
More informationLamarckian Repair and Darwinian Repair in EMO Algorithms for Multiobjective 0/1 Knapsack Problems
Repair and Repair in EMO Algorithms for Multiobjective 0/ Knapsack Problems Shiori Kaige, Kaname Narukawa, and Hisao Ishibuchi Department of Industrial Engineering, Osaka Prefecture University, - Gakuen-cho,
More informationMulti-Objective Memetic Algorithm using Pattern Search Filter Methods
Multi-Objective Memetic Algorithm using Pattern Search Filter Methods F. Mendes V. Sousa M.F.P. Costa A. Gaspar-Cunha IPC/I3N - Institute of Polymers and Composites, University of Minho Guimarães, Portugal
More informationBi-Objective Optimization for Scheduling in Heterogeneous Computing Systems
Bi-Objective Optimization for Scheduling in Heterogeneous Computing Systems Tony Maciejewski, Kyle Tarplee, Ryan Friese, and Howard Jay Siegel Department of Electrical and Computer Engineering Colorado
More informationR2-IBEA: R2 Indicator Based Evolutionary Algorithm for Multiobjective Optimization
R2-IBEA: R2 Indicator Based Evolutionary Algorithm for Multiobjective Optimization Dũng H. Phan Department of Computer Science University of Massachusetts, Boston Boston, MA 02125, USA Email: phdung@cs.umb.edu
More informationUsing ɛ-dominance for Hidden and Degenerated Pareto-Fronts
IEEE Symposium Series on Computational Intelligence Using ɛ-dominance for Hidden and Degenerated Pareto-Fronts Heiner Zille Institute of Knowledge and Language Engineering University of Magdeburg, Germany
More informationCHAPTER 2 MULTI-OBJECTIVE REACTIVE POWER OPTIMIZATION
19 CHAPTER 2 MULTI-OBJECTIE REACTIE POWER OPTIMIZATION 2.1 INTRODUCTION In this chapter, a fundamental knowledge of the Multi-Objective Optimization (MOO) problem and the methods to solve are presented.
More informationAn Evolutionary Algorithm for the Multi-objective Shortest Path Problem
An Evolutionary Algorithm for the Multi-objective Shortest Path Problem Fangguo He Huan Qi Qiong Fan Institute of Systems Engineering, Huazhong University of Science & Technology, Wuhan 430074, P. R. China
More informationA Similarity-Based Mating Scheme for Evolutionary Multiobjective Optimization
A Similarity-Based Mating Scheme for Evolutionary Multiobjective Optimization Hisao Ishibuchi and Youhei Shibata Department of Industrial Engineering, Osaka Prefecture University, - Gakuen-cho, Sakai,
More informationEvolutionary Algorithm for Embedded System Topology Optimization. Supervisor: Prof. Dr. Martin Radetzki Author: Haowei Wang
Evolutionary Algorithm for Embedded System Topology Optimization Supervisor: Prof. Dr. Martin Radetzki Author: Haowei Wang Agenda Introduction to the problem Principle of evolutionary algorithm Model specification
More informationNEW DECISION MAKER MODEL FOR MULTIOBJECTIVE OPTIMIZATION INTERACTIVE METHODS
NEW DECISION MAKER MODEL FOR MULTIOBJECTIVE OPTIMIZATION INTERACTIVE METHODS Andrejs Zujevs 1, Janis Eiduks 2 1 Latvia University of Agriculture, Department of Computer Systems, Liela street 2, Jelgava,
More informationMechanical Component Design for Multiple Objectives Using Elitist Non-Dominated Sorting GA
Mechanical Component Design for Multiple Objectives Using Elitist Non-Dominated Sorting GA Kalyanmoy Deb, Amrit Pratap, and Subrajyoti Moitra Kanpur Genetic Algorithms Laboratory (KanGAL) Indian Institute
More informationUnsupervised Feature Selection Using Multi-Objective Genetic Algorithms for Handwritten Word Recognition
Unsupervised Feature Selection Using Multi-Objective Genetic Algorithms for Handwritten Word Recognition M. Morita,2, R. Sabourin 3, F. Bortolozzi 3 and C. Y. Suen 2 École de Technologie Supérieure, Montreal,
More informationMULTI-OBJECTIVE GENETIC LOCAL SEARCH ALGORITHM FOR SUPPLY CHAIN SIMULATION OPTIMISATION
MULTI-OBJECTIVE GENETIC LOCAL SEARCH ALGORITHM FOR SUPPLY CHAIN SIMULATION OPTIMISATION Galina Merkuryeva (a), Liana Napalkova (b) (a) (b) Department of Modelling and Simulation, Riga Technical University,
More informationIncorporation of Scalarizing Fitness Functions into Evolutionary Multiobjective Optimization Algorithms
H. Ishibuchi, T. Doi, and Y. Nojima, Incorporation of scalarizing fitness functions into evolutionary multiobjective optimization algorithms, Lecture Notes in Computer Science 4193: Parallel Problem Solving
More informationLight Beam Search Based Multi-objective Optimization using Evolutionary Algorithms
Light Beam Search Based Multi-objective Optimization using Evolutionary Algorithms Kalyanmoy Deb and Abhay Kumar KanGAL Report Number 275 Abstract For the past decade or so, evolutionary multiobjective
More informationAn Improved Progressively Interactive Evolutionary Multi-objective Optimization Algorithm with a Fixed Budget of Decision Maker Calls
An Improved Progressively Interactive Evolutionary Multi-objective Optimization Algorithm with a Fixed Budget of Decision Maker Calls Ankur Sinha, Pekka Korhonen, Jyrki Wallenius Firstname.Secondname@aalto.fi,
More informationINTERACTIVE MULTI-OBJECTIVE GENETIC ALGORITHMS FOR THE BUS DRIVER SCHEDULING PROBLEM
Advanced OR and AI Methods in Transportation INTERACTIVE MULTI-OBJECTIVE GENETIC ALGORITHMS FOR THE BUS DRIVER SCHEDULING PROBLEM Jorge PINHO DE SOUSA 1, Teresa GALVÃO DIAS 1, João FALCÃO E CUNHA 1 Abstract.
More informationDevelopment of Evolutionary Multi-Objective Optimization
A. Mießen Page 1 of 13 Development of Evolutionary Multi-Objective Optimization Andreas Mießen RWTH Aachen University AVT - Aachener Verfahrenstechnik Process Systems Engineering Turmstrasse 46 D - 52056
More informationDETERMINING PARETO OPTIMAL CONTROLLER PARAMETER SETS OF AIRCRAFT CONTROL SYSTEMS USING GENETIC ALGORITHM
DETERMINING PARETO OPTIMAL CONTROLLER PARAMETER SETS OF AIRCRAFT CONTROL SYSTEMS USING GENETIC ALGORITHM Can ÖZDEMİR and Ayşe KAHVECİOĞLU School of Civil Aviation Anadolu University 2647 Eskişehir TURKEY
More informationPerformance Assessment of DMOEA-DD with CEC 2009 MOEA Competition Test Instances
Performance Assessment of DMOEA-DD with CEC 2009 MOEA Competition Test Instances Minzhong Liu, Xiufen Zou, Yu Chen, Zhijian Wu Abstract In this paper, the DMOEA-DD, which is an improvement of DMOEA[1,
More informationMULTI-OBJECTIVE OPTIMIZATION
MULTI-OBJECTIVE OPTIMIZATION Introduction Many real-world problems require the simultaneous optimization of a number of objective functions. Some of these objectives may be in conflict. Example 1:optimal
More informationLate Parallelization and Feedback Approaches for Distributed Computation of Evolutionary Multiobjective Optimization Algorithms
Late arallelization and Feedback Approaches for Distributed Computation of Evolutionary Multiobjective Optimization Algorithms O. Tolga Altinoz Department of Electrical and Electronics Engineering Ankara
More informationAttribute Selection with a Multiobjective Genetic Algorithm
Attribute Selection with a Multiobjective Genetic Algorithm Gisele L. Pappa, Alex A. Freitas, Celso A.A. Kaestner Pontifícia Universidade Catolica do Parana (PUCPR), Postgraduated Program in Applied Computer
More informationProceedings of the 2012 Winter Simulation Conference C. Laroque, J. Himmelspach, R. Pasupathy, O. Rose, and A.M. Uhrmacher, eds
Proceedings of the 2012 Winter Simulation Conference C. Laroque, J. Himmelspach, R. Pasupathy, O. Rose, and A.M. Uhrmacher, eds REFERENCE POINT-BASED EVOLUTIONARY MULTI-OBJECTIVE OPTIMIZATION FOR INDUSTRIAL
More informationTowards Understanding Evolutionary Bilevel Multi-Objective Optimization Algorithm
Towards Understanding Evolutionary Bilevel Multi-Objective Optimization Algorithm Ankur Sinha and Kalyanmoy Deb Helsinki School of Economics, PO Box, FIN-, Helsinki, Finland (e-mail: ankur.sinha@hse.fi,
More informationAutomated Test Case Generation as a Many-Objective Optimisation Problem with Dynamic Selection of the Targets
1 Automated Test Case Generation as a Many-Objective Optimisation Problem with Dynamic Selection of the Targets Annibale Panichella, Fitsum Meshesha Kifetew, Paolo Tonella SnT - University of Luxembourg,
More informationGraphical User Interface For Multi-Objective Decision Support
Master Thesis Graphical User Interface For Multi-Objective Decision Support Author: Benjamin Keller Supervisor: Dr. Thomas Hanne A master thesis presented to the School of Business of the University of
More informationMulti-objective Optimization Algorithm based on Magnetotactic Bacterium
Vol.78 (MulGrab 24), pp.6-64 http://dx.doi.org/.4257/astl.24.78. Multi-obective Optimization Algorithm based on Magnetotactic Bacterium Zhidan Xu Institute of Basic Science, Harbin University of Commerce,
More informationStandard Error Dynamic Resampling for Preference-based Evolutionary Multi-objective Optimization
Standard Error Dynamic Resampling for Preference-based Evolutionary Multi-objective Optimization Florian Siegmund a, Amos H. C. Ng a, and Kalyanmoy Deb b a School of Engineering, University of Skövde,
More informationEvolutionary multi-objective algorithm design issues
Evolutionary multi-objective algorithm design issues Karthik Sindhya, PhD Postdoctoral Researcher Industrial Optimization Group Department of Mathematical Information Technology Karthik.sindhya@jyu.fi
More informationApproximation-Guided Evolutionary Multi-Objective Optimization
Approximation-Guided Evolutionary Multi-Objective Optimization Karl Bringmann 1, Tobias Friedrich 1, Frank Neumann 2, Markus Wagner 2 1 Max-Planck-Institut für Informatik, Campus E1.4, 66123 Saarbrücken,
More informationNeural Network Regularization and Ensembling Using Multi-objective Evolutionary Algorithms
Neural Network Regularization and Ensembling Using Multi-objective Evolutionary Algorithms Yaochu Jin Honda Research Institute Europe Carl-Legien-Str 7 Offenbach, GERMANY Email: yaochujin@honda-ride Tatsuya
More informationMultiobjective Formulations of Fuzzy Rule-Based Classification System Design
Multiobjective Formulations of Fuzzy Rule-Based Classification System Design Hisao Ishibuchi and Yusuke Nojima Graduate School of Engineering, Osaka Prefecture University, - Gakuen-cho, Sakai, Osaka 599-853,
More informationMulticriterial Optimization Using Genetic Algorithm
Multicriterial Optimization Using Genetic Algorithm 180 175 170 165 Fitness 160 155 150 145 140 Best Fitness Mean Fitness 135 130 0 Page 1 100 200 300 Generations 400 500 600 Contents Optimization, Local
More informationAn Interactive Evolutionary Multi-Objective Optimization Method Based on Progressively Approximated Value Functions
An Interactive Evolutionary Multi-Objective Optimization Method Based on Progressively Approximated Value Functions Kalyanmoy Deb, Ankur Sinha, Pekka Korhonen, and Jyrki Wallenius KanGAL Report Number
More informationMechanical Component Design for Multiple Objectives Using Elitist Non-Dominated Sorting GA
Mechanical Component Design for Multiple Objectives Using Elitist Non-Dominated Sorting GA Kalyanmoy Deb, Amrit Pratap, and Subrajyoti Moitra Kanpur Genetic Algorithms Laboratory (KanGAL) Indian Institute
More informationDynamic Resampling for Preference-based Evolutionary Multi-Objective Optimization of Stochastic Systems
Dynamic Resampling for Preference-based Evolutionary Multi-Objective Optimization of Stochastic Systems Florian Siegmund a, Amos H. C. Ng a, and Kalyanmoy Deb b a School of Engineering, University of Skövde,
More informationA Comparative Study of Dynamic Resampling Strategies for Guided Evolutionary Multi-Objective Optimization
A Comparative Study of Dynamic Resampling Strategies for Guided Evolutionary Multi-Objective Optimization KanGAL Report Number 203008 Florian Siegmund, Amos H.C. Ng Virtual Systems Research Center University
More informationApproximation Model Guided Selection for Evolutionary Multiobjective Optimization
Approximation Model Guided Selection for Evolutionary Multiobjective Optimization Aimin Zhou 1, Qingfu Zhang 2, and Guixu Zhang 1 1 Each China Normal University, Shanghai, China 2 University of Essex,
More informationDynamic Uniform Scaling for Multiobjective Genetic Algorithms
Dynamic Uniform Scaling for Multiobjective Genetic Algorithms Gerulf K. M. Pedersen 1 and David E. Goldberg 2 1 Aalborg University, Department of Control Engineering, Fredrik Bajers Vej 7, DK-922 Aalborg
More informationHYBRID GENETIC ALGORITHM WITH GREAT DELUGE TO SOLVE CONSTRAINED OPTIMIZATION PROBLEMS
HYBRID GENETIC ALGORITHM WITH GREAT DELUGE TO SOLVE CONSTRAINED OPTIMIZATION PROBLEMS NABEEL AL-MILLI Financial and Business Administration and Computer Science Department Zarqa University College Al-Balqa'
More informationA New Ranking Scheme for Multi and Single Objective Problems
ISSN (Print) : 2347-671 (An ISO 3297: 27 Certified Organization) Vol. 4, Issue 3, March 215 A New Ranking Scheme for Multi and Single Objective Problems A. R. Khaparde 1, V. M Athawale 2 Assistant Professor,
More informationEVOLUTIONARY algorithms (EAs) are a class of
An Investigation on Evolutionary Gradient Search for Multi-objective Optimization C. K. Goh, Y. S. Ong and K. C. Tan Abstract Evolutionary gradient search is a hybrid algorithm that exploits the complementary
More informationA genetic algorithms approach to optimization parameter space of Geant-V prototype
A genetic algorithms approach to optimization parameter space of Geant-V prototype Oksana Shadura CERN, PH-SFT & National Technical Univ. of Ukraine Kyiv Polytechnic Institute Geant-V parameter space [1/2]
More informationDimensionality Reduction in Multiobjective Optimization: The Minimum Objective Subset Problem
Eckart Zitzler ETH Zürich Dimo Brockhoff ETH Zurich Gene Expression Data Analysis 1 Computer Engineering and Networks Laboratory Dimensionality Reduction in Multiobjective Optimization: The Minimum Objective
More informationCHAPTER 5 STRUCTURAL OPTIMIZATION OF SWITCHED RELUCTANCE MACHINE
89 CHAPTER 5 STRUCTURAL OPTIMIZATION OF SWITCHED RELUCTANCE MACHINE 5.1 INTRODUCTION Nowadays a great attention has been devoted in the literature towards the main components of electric and hybrid electric
More informationDEMO: Differential Evolution for Multiobjective Optimization
DEMO: Differential Evolution for Multiobjective Optimization Tea Robič and Bogdan Filipič Department of Intelligent Systems, Jožef Stefan Institute, Jamova 39, SI-1000 Ljubljana, Slovenia tea.robic@ijs.si
More informationDesign of Curves and Surfaces Using Multi-Objective Optimization
Design of Curves and Surfaces Using Multi-Objective Optimization Rony Goldenthal and Michel Bercovier Abstract. Design by optimization of curves and surfaces is a powerful design technique. The mathematical
More informationUsing an outward selective pressure for improving the search quality of the MOEA/D algorithm
Comput Optim Appl (25) 6:57 67 DOI.7/s589-5-9733-9 Using an outward selective pressure for improving the search quality of the MOEA/D algorithm Krzysztof Michalak Received: 2 January 24 / Published online:
More informationTowards an Estimation of Nadir Objective Vector Using Hybrid Evolutionary and Local Search Approaches
Towards an Estimation of Nadir Objective Vector Using Hybrid Evolutionary and Local Search Approaches Kalyanmoy Deb, Kaisa Miettinen, and Shamik Chaudhuri KanGAL Report Number 279 Abstract Nadir objective
More informationEvolving Human Competitive Research Spectra-Based Note Fault Localisation Techniques
UCL DEPARTMENT OF COMPUTER SCIENCE Research Note RN/12/03 Evolving Human Competitive Research Spectra-Based Note Fault Localisation Techniques RN/17/07 Deep Parameter Optimisation for Face Detection Using
More informationOptimizing Delivery Time in Multi-Objective Vehicle Routing Problems with Time Windows
Optimizing Delivery Time in Multi-Objective Vehicle Routing Problems with Time Windows Abel Garcia-Najera and John A. Bullinaria School of Computer Science, University of Birmingham Edgbaston, Birmingham
More informationInvestigating the Effect of Parallelism in Decomposition Based Evolutionary Many-Objective Optimization Algorithms
Investigating the Effect of Parallelism in Decomposition Based Evolutionary Many-Objective Optimization Algorithms Lei Chen 1,2, Kalyanmoy Deb 2, and Hai-Lin Liu 1 1 Guangdong University of Technology,
More informationReference Point Based Evolutionary Approach for Workflow Grid Scheduling
Reference Point Based Evolutionary Approach for Workflow Grid Scheduling R. Garg and A. K. Singh Abstract Grid computing facilitates the users to consume the services over the network. In order to optimize
More informationMultiobjective Job-Shop Scheduling With Genetic Algorithms Using a New Representation and Standard Uniform Crossover
Multiobjective Job-Shop Scheduling With Genetic Algorithms Using a New Representation and Standard Uniform Crossover J. Garen 1 1. Department of Economics, University of Osnabrück, Katharinenstraße 3,
More informationPart II. Computational Intelligence Algorithms
Part II Computational Intelligence Algorithms 126 Chapter 5 Population-based Single-objective Algorithms One bee makes no swarm. French proverb This chapter provides an overview of two CI algorithms that
More informationHierarchical Clustering of Evolutionary Multiobjective Programming Results to Inform Land Use Planning
Hierarchical Clustering of Evolutionary Multiobjective Programming Results to Inform Land Use Planning by Christina Moulton A thesis presented to the University of Waterloo in fulfillment of the thesis
More informationNCGA : Neighborhood Cultivation Genetic Algorithm for Multi-Objective Optimization Problems
: Neighborhood Cultivation Genetic Algorithm for Multi-Objective Optimization Problems Shinya Watanabe Graduate School of Engineering, Doshisha University 1-3 Tatara Miyakodani,Kyo-tanabe, Kyoto, 10-031,
More informationHeuristic Optimisation
Heuristic Optimisation Part 10: Genetic Algorithm Basics Sándor Zoltán Németh http://web.mat.bham.ac.uk/s.z.nemeth s.nemeth@bham.ac.uk University of Birmingham S Z Németh (s.nemeth@bham.ac.uk) Heuristic
More informationImproved S-CDAS using Crossover Controlling the Number of Crossed Genes for Many-objective Optimization
Improved S-CDAS using Crossover Controlling the Number of Crossed Genes for Many-objective Optimization Hiroyuki Sato Faculty of Informatics and Engineering, The University of Electro-Communications -5-
More informationAdjusting Parallel Coordinates for Investigating Multi-Objective Search
Adjusting Parallel Coordinates for Investigating Multi-Objective Search Liangli Zhen,, Miqing Li, Ran Cheng, Dezhong Peng and Xin Yao 3, Machine Intelligence Laboratory, College of Computer Science, Sichuan
More informationAPPLICATION OF SELF-ORGANIZING MAPS IN VISUALIZATION OF MULTI- DIMENSIONAL PARETO FRONTS
Zeszyty Naukowe WSInf Vol 15, Nr 1, 2016 Tomasz Schlieter Institute of Computational Mechanics and Engineering, Silesian University of Technology, ul. Konarskiego 18A, 44-100 Gliwice email: tomasz.schlieter@polsl.pl
More informationSolving Multi-objective Optimisation Problems Using the Potential Pareto Regions Evolutionary Algorithm
Solving Multi-objective Optimisation Problems Using the Potential Pareto Regions Evolutionary Algorithm Nasreddine Hallam, Graham Kendall, and Peter Blanchfield School of Computer Science and IT, The Univeristy
More informationSolving Bilevel Multi-Objective Optimization Problems Using Evolutionary Algorithms
Solving Bilevel Multi-Objective Optimization Problems Using Evolutionary Algorithms Kalyanmoy Deb and Ankur Sinha Department of Mechanical Engineering Indian Institute of Technology Kanpur PIN 2816, India
More informationGenetic Algorithm Performance with Different Selection Methods in Solving Multi-Objective Network Design Problem
etic Algorithm Performance with Different Selection Methods in Solving Multi-Objective Network Design Problem R. O. Oladele Department of Computer Science University of Ilorin P.M.B. 1515, Ilorin, NIGERIA
More informationEngineering design using genetic algorithms
Retrospective Theses and Dissertations 2007 Engineering design using genetic algorithms Xiaopeng Fang Iowa State University Follow this and additional works at: http://lib.dr.iastate.edu/rtd Part of the
More informationGenerating Uniformly Distributed Pareto Optimal Points for Constrained and Unconstrained Multicriteria Optimization
Generating Uniformly Distributed Pareto Optimal Points for Constrained and Unconstrained Multicriteria Optimization Crina Grosan Department of Computer Science Babes-Bolyai University Cluj-Napoca, Romania
More informationCompromise Based Evolutionary Multiobjective Optimization Algorithm for Multidisciplinary Optimization
Compromise Based Evolutionary Multiobjective Optimization Algorithm for Multidisciplinary Optimization Benoît Guédas, Xavier Gandibleux, Philippe Dépincé To cite this version: Benoît Guédas, Xavier Gandibleux,
More informationAn Optimality Theory Based Proximity Measure for Set Based Multi-Objective Optimization
An Optimality Theory Based Proximity Measure for Set Based Multi-Objective Optimization Kalyanmoy Deb, Fellow, IEEE and Mohamed Abouhawwash Department of Electrical and Computer Engineering Computational
More informationGECCO 2007 Tutorial / Evolutionary Multiobjective Optimization. Eckart Zitzler ETH Zürich. weight = 750g profit = 5.
Tutorial / Evolutionary Multiobjective Optimization Tutorial on Evolutionary Multiobjective Optimization Introductory Example: The Knapsack Problem weight = 75g profit = 5 weight = 5g profit = 8 weight
More informationCOMPENDIOUS LEXICOGRAPHIC METHOD FOR MULTI-OBJECTIVE OPTIMIZATION. Ivan P. Stanimirović. 1. Introduction
FACTA UNIVERSITATIS (NIŠ) Ser. Math. Inform. Vol. 27, No 1 (2012), 55 66 COMPENDIOUS LEXICOGRAPHIC METHOD FOR MULTI-OBJECTIVE OPTIMIZATION Ivan P. Stanimirović Abstract. A modification of the standard
More informationMulti-Objective Optimization Techniques for VLSI Circuits
Multi-Objective Techniques for VLSI Circuits Fatemeh Kashfi, Safar Hatami, Massoud Pedram University of Southern California 374 McClintock Ave, Los Angeles CA 989 E-mail: fkashfi@usc.edu Abstract The EDA
More informationEvolutionary Multi-Objective Optimization Without Additional Parameters
Evolutionary Multi-Objective Optimization Without Additional Parameters Kalyanmoy Deb Department of Mechanical Engineering Indian Institute of Technology Kanpur Kanpur, PIN 8, India Email: deb@iitk.ac.in
More informationVisualization of Pareto Front Points when Solving Multi-objective Optimization Problems
ISSN 9 4X, ISSN 884X (online) INFORMATION TECHNOLOGY AND CONTROL,, Vol.4, No.4 Visualization of Pareto Front Points when Solving Multi-objective Optimization Problems Olga Kurasova,, Tomas Petkus, Ernestas
More informationHybrid Genetic Algorithms for Multi-objective Optimisation of Water Distribution Networks
Hybrid Genetic Algorithms for Multi-objective Optimisation of Water Distribution Networks Edward Keedwell and Soon-Thiam Khu Centre for Water Systems, School of Engineering and Computer Science and Mathematics,
More informationMulti-objective Evolutionary Design of Fuzzy Autopilot Controller
Multi-objective Evolutionary Design of Fuzzy Autopilot Controller Anna L. Blumel, Evan J. Hughes, and Brian A. White Department of Aerospace, Power, and Sensors, Cranfield University, Royal Military College
More informationAn Evolutionary Algorithm Approach to Generate Distinct Sets of Non-Dominated Solutions for Wicked Problems
An Evolutionary Algorithm Approach to Generate Distinct Sets of Non-Dominated Solutions for Wicked Problems Marcio H. Giacomoni Assistant Professor Civil and Environmental Engineering February 6 th 7 Zechman,
More informationExperimental Study on Bound Handling Techniques for Multi-Objective Particle Swarm Optimization
Experimental Study on Bound Handling Techniques for Multi-Objective Particle Swarm Optimization adfa, p. 1, 2011. Springer-Verlag Berlin Heidelberg 2011 Devang Agarwal and Deepak Sharma Department of Mechanical
More informationSimulation of Robot Manipulator Trajectory Optimization Design
International Journal of Research in Engineering and Science (IJRES) ISSN (Online): -96, ISSN (Print): -956 Volume 5 Issue ǁ Feb. 7 ǁ PP.7-5 Simulation of Robot Manipulator Trajectory Optimization Design
More informationBI-OBJECTIVE EVOLUTIONARY ALGORITHM FOR FLEXIBLE JOB-SHOP SCHEDULING PROBLEM. Minimizing Make Span and the Total Workload of Machines
International Journal of Mathematics and Computer Applications Research (IJMCAR) ISSN 2249-6955 Vol. 2 Issue 4 Dec - 2012 25-32 TJPRC Pvt. Ltd., BI-OBJECTIVE EVOLUTIONARY ALGORITHM FOR FLEXIBLE JOB-SHOP
More informationImproving interpretability in approximative fuzzy models via multi-objective evolutionary algorithms.
Improving interpretability in approximative fuzzy models via multi-objective evolutionary algorithms. Gómez-Skarmeta, A.F. University of Murcia skarmeta@dif.um.es Jiménez, F. University of Murcia fernan@dif.um.es
More informationA gradient-based multiobjective optimization technique using an adaptive weighting method
10 th World Congress on Structural and Multidisciplinary Optimization May 19-24, 2013, Orlando, Florida, USA A gradient-based multiobjective optimization technique using an adaptive weighting method Kazuhiro
More informationA GENETIC ALGORITHM APPROACH FOR TECHNOLOGY CHARACTERIZATION. A Thesis EDGAR GALVAN
A GENETIC ALGORITHM APPROACH FOR TECHNOLOGY CHARACTERIZATION A Thesis by EDGAR GALVAN Submitted to the Office of Graduate Studies of Texas A&M University in partial fulfillment of the requirements for
More informationComputational Intelligence
Computational Intelligence Winter Term 2016/17 Prof. Dr. Günter Rudolph Lehrstuhl für Algorithm Engineering (LS 11) Fakultät für Informatik TU Dortmund Slides prepared by Dr. Nicola Beume (2012) Multiobjective
More informationScenario-based Refactoring Selection
BABEŞ-BOLYAI University of Cluj-Napoca Faculty of Mathematics and Computer Science Proceedings of the National Symposium ZAC2014 (Zilele Academice Clujene, 2014), p. 32-41 Scenario-based Refactoring Selection
More informationSTUDY OF MULTI-OBJECTIVE OPTIMIZATION AND ITS IMPLEMENTATION USING NSGA-II
STUDY OF MULTI-OBJECTIVE OPTIMIZATION AND ITS IMPLEMENTATION USING NSGA-II A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF Bachelor of Technology in Electrical Engineering.
More informationAssessing the Convergence Properties of NSGA-II for Direct Crashworthiness Optimization
10 th International LS-DYNA Users Conference Opitmization (1) Assessing the Convergence Properties of NSGA-II for Direct Crashworthiness Optimization Guangye Li 1, Tushar Goel 2, Nielen Stander 2 1 IBM
More informationIncrementally Maximising Hypervolume for Selection in Multi-objective Evolutionary Algorithms
Incrementally Maximising Hypervolume for Selection in Multi-objective Evolutionary Algorithms Lucas Bradstreet, Student Member, IEEE, Lyndon While, Senior Member, IEEE, and Luigi Barone, Member, IEEE Abstract
More information