1.04k likes | 1.16k Views
Evolution Strategies for Industrial Applications. Thomas Bäck November 26, 2004. Natural Computing Leiden Institute for Advanced Computer Science (LIACS) Niels Bohrweg 1 NL-2333 CA Leiden baeck@liacs.nl Tel.: +31 (0) 71 527 7108 Fax: +31 (0) 71 527 6985. Managing Director / CTO
E N D
Evolution Strategiesfor Industrial Applications Thomas Bäck November 26, 2004 Natural Computing Leiden Institute for Advanced Computer Science (LIACS) Niels Bohrweg 1 NL-2333 CA Leiden baeck@liacs.nl Tel.: +31 (0) 71 527 7108 Fax: +31 (0) 71 527 6985 Managing Director / CTO NuTech Solutions GmbH / Inc. Martin-Schmeißer-Weg 15 D – 44227 Dortmund baeck@nutechsolutions.de Tel.: +49 (0) 231 / 72 54 63-10 Fax: +49 (0) 231 / 72 54 63-29
Background I Daniel Dannett: Biology = Engineering
Background II Realistic Scenario ....
Background III Phenotypic evolution ... and genotypic
Optimization • E.g. costs (min), quality (max), error (min), stability (max), profit (max), ... • Difficulties: • High-dimensional • Nonlinear, non-quadratic: Multimodal • Noisy, dynamic, discontinuous • Evolutionary landscapes are like that !
Overview • Evolutionary Algorithm Applications: Examples • Evolutionary Algorithms: Some Algorithmic Details • Genetic Algorithms • Evolution Strategies • Some Theory of EAs • Convergence Velocity Issues • Other Examples • Drug Design • Inverse Design of Cas • Summary
Modeling – Simulation - Optimization Modeling / Data Mining ??? !!! !!! Simulation !!! ??? !!! Optimization ??? !!! !!!
Model from Data Simulation Function Function(s) Subjective Experiment Evaluation Business Process Model EA-Optimizer General Aspects
Examples I: Inflatable Knee Bolster Optimization MAZDA Picture Low Cost ES: 0.677 GA (Ford): 0.72 Hooke Jeeves DoE: 0.88 Initial position of knee bag model deployed knee bag (unit only) Tether FEM #5 Tether Volume of 14L Load distribution plate FEM #3 Support plate FEM #4 Support plate Knee bag FEM #2 Load distribution plate Straps are defined in knee bag(FEM #2) Vent hole
IKB: Problem Statement Objective: Min Ptotal Subject to: Left Femur load <= 7000 Right Femur load <= 7000
IKB Results I: Hooke-Jeeves Quality: 8.888 Simulations: 160
IKB Results II: (1+1)-ES Quality: 7.142 Simulations: 122
Optical Coatings: Design Optimization • Nonlinear mixed-integer problem, variable dimensionality. • Minimize deviation from desired reflection behaviour. • Excellent synthesis method; robust and reliable results.
Dielectric Filter Design Problem • Dielectric filter design. • n=40 layers assumed. • Layer thicknesses xi in [0.01, 10.0]. • Quality function: Sum of quadratic penalty terms. • Penalty terms = 0 iff constraints satisfied. Client: Corning, Inc., Corning, NY
Results: Overview of Runs • Factor 2 in quality. • Factor 10 in effort. • Reliable, repeatable results. Benchmark
Problem Topology Analysis: An Attempt • Grid evaluation for 2 variables. • Close to the optimum (from vector of quality 0.0199). • Global view (left), vs. Local view (right).
Examples II: Bridgman Casting Process 18 Speed Variables (continuous) for Casting Schedule Turbine Blade after Casting · FE mesh of 1/3 geometry: 98.610 nodes, 357.300 tetrahedrons, 92.830 radiation surfaces large problem: · run time varies: 16 h 30 min to 32 h (SGI, Origin, R12000, 400 MHz) · at each run: 38,3 GB of view factors (49 positions) are treated!
Examples II: Bridgman Casting Process Global Quality Turbine Blade after Casting GCM(Commercial Gradient Based Method) Initial (DoE) Evolution Strategy Quality Comparison of the Initial and Optimized Configurations
Examples IV: Traffic Light Control Client: Dutch Ministry of Traffic Rotterdam, NL • Generates green times for next switching schedule. • Minimization of total delay / number of stops. • Better results (3 – 5%) / higher flexibility than with traditional controllers. • Dynamic optimization, depending on actual traffic (measured by control loops).
Examples V: Elevator Control Client: Fujitec Co. Ltd., Osaka, Japan • Minimization of passenger waiting times. • Better results (3 – 5%) / higher flexibility than with traditional controllers. • Dynamic optimization, depending on actual traffic.
Examples VI: Metal Stamping Process Client: AutoForm Engineering GmbH, Dortmund • Minimization of defects in the produced parts. • Optimization on geometric parameters and forces. • Fast algorithm; finds very good results.
Examples VII: Network Routing Client: SIEMENS AG, München • Minimization of end-to-end-blockings under service constraints. • Optimization of routing tables for existing, hard-wired networks. • 10%-1000% improvement.
Examples VIII: Nuclear Reactor Refueling Client: SIEMENS AG, München • Minimization of total costs. • Creates new fuel assembly reload patterns. • Clear improvements (1%-5%) of existing expert solutions. • Huge cost saving.
Two-Phase Nozzle Design (Experimental) Experimental design optimisation: Optimise efficieny. ... evolves... Initial design Final design: 32% improvement in efficieny.
Multipoint Airfoil Optimization (1) Client: Cruise Low Drag! High Lift! Start 22 design parameters.
Find pressure profiles that are a compromise between two given target pressure distributions under two given flow conditions! Multipoint Airfoil Optimization (2) Three compromise wing designs Pareto set after 1000 Simulations
Evolutionary Algorithms: Some Algorithmic Details
Unifying Evolutionary Algorithm t := 0; initialize(P(t)); evaluate(P(t)); while notterminatedo P‘(t) := mating_selection(P(t)); P‘‘(t) := variation(P‘(t)); evaluate(P‘‘(t)); P(t+1) := environmental_selection(P‘‘(t) u Q); t := t+1; od
Evolutionary Algorithm Taxonomy Evolution Strategies Genetic Algorithms Classifier Systems Evolutionary Programming Genetic Programming Many mixed forms; agent-based systems, swarm systems, A-life systems, ...
Genetic Algorithms vs. Evolution Strategies Genetic Algorithm Evolution Strategies • Binary representation • Fixed mutation rate pm (= 1/n) • Fixed crossover rate pc • Probabilistic selection • Identical population size • No self-adaptation • Real-valued representation • Normally distributed mutations • Fixed recombination rate (= 1) • Deterministic selection • Creation of offspring surplus • Self-adaptation of strategy parameters: Variance(s), Covariances
Genetic Algorithms • Often binary representation. • Mutation by bit inversion with probability pm. • Various types of crossover, with probability pc. • k-point crossover. • Uniform crossover. • Probabilistic selection operators. • Proportional selection. • Tournament selection. • Parent and offspring population size identical. • Constant strategy parameters.
0 1 1 1 0 1 0 1 0 0 0 0 0 1 0 Mutation 0 1 1 1 0 0 0 1 0 1 0 0 0 1 0 • Mutation by bit inversion with probability pm. • pm identical for all bits. • pm small (e.g., pm = 1/l).
Crossover • Crossover applied with probability pc. • pc identical for all individuals. • k-point crossover: k points chosen randomly. • Example: 2-point crossover.
Selection • Fitness proportional: • f fitness • l population size • Tournament selection: • Randomly select q << l individuals. • Copy best of these q into next generation. • Repeat l times. • q is the tournament size (often: q = 2).
Evolution Strategies • Real-valued representation. • Normally distributed mutations. • Various types of recombination. • Discrete (exchange of variables). • Intermediate (averaging). • Involving two or more parents. • Deterministic selection, offspring surplus l >> m. • Elitist: (m+l) • Non-elitist: (m,l) • Self-Adaptation of strategy parameters.
Mutation Creation of a new solution: s-adaptation by means of • 1/5-success rule. • Self-adaptation. More complex / powerful strategies: • Individual step sizes si. • Covariances. • Convergence speed: • Ca. 10 n down to 5 n is possible.
Self-Adaptation • Learning while searching: Intelligent Method. • Different algorithmic approaches, e.g: • Pure self-adaptation: • Mutational step size control MSC: • Derandomized step size adaptation • Covariance adaptation
Self-Adaptive Mutation n = 2, ns = 1, na = 0 n = 2, ns = 2, na = 0 n = 2, ns = 2, na = 1
Self-Adaptation: • Motivation: General search algorithm • Geometric convergence: Arbitrarily slow, if s wrongly controlled ! • No deterministic / adaptive scheme for arbitrary functions exists. • Self-adaptation: On-line evolution of strategy parameters. • Various schemes: • Schwefel one s, n s, covariances; Rechenberg MSA. • Ostermeier, Hansen: Derandomized, Covariance Matrix Adaptation. • EP variants (meta EP, Rmeta EP). • Bäck: Application to p in GAs. Step size Direction
Self-Adaptation: Dynamic Sphere • Optimum s: • Transition time proportionate to n. • Optimum s learned by self-adaptation.
Selection (m,l) (m+l)
Possible Selection Operators • (1+1)-strategy: one parent, one offspring. • (1,l)-strategies: one parent, l offspring. • Example: (1,10)-strategy. • Derandomized / self-adaptive / mutative step size control. • (m,l)-strategies:m>1 parents, l>m offspring • Example: (2,15)-strategy. • Includes recombination. • Can overcome local optima. • (m+l)-strategies: elitist strategies.
Advantages of Evolution Strategies • Self-Adaptation of strategy parameters. • Direct, global optimizers ! • Faster than GAs ! • Extremely good in solution quality. • Very small number of function evaluations. • Dynamical optimization problems. • Design optimization problems. • Discrete or mixed-integer problems. • Experimental design optimisation. • Combination with Meta-Modeling techniques.
Robust vs. Fast: • Global convergence with probability one: General, but for practical purposes useless. • Convergence velocity: Local analysis only, specific functions only.
Convergence Velocity Analysis, ES: • A convex function („sphere model“). • Simplest case: (1,l)-ES • Illustration: (1,4)-ES
Convergence Velocity Analysis, ES: • Order statistics: • pn:l(z) denotes the p.d.f. of Zn:l • Idea: Best offspring has smallest r / largest z‘. • The following holds from geometric considerations: • One gets:
Convergence Velocity Analysis, ES: • Using one finally gets • One gets: (dimensionless) l
Convergence Velocity Analysis, ES: • Convergence velocity, (1,l)-ES and (1+l)-ES: