## Simulated Annealing Algorithms For Continuous Global Optimization (2000)

Citations: | 30 - 1 self |

### BibTeX

@MISC{Locatelli00simulatedannealing,

author = {M. Locatelli},

title = {Simulated Annealing Algorithms For Continuous Global Optimization},

year = {2000}

}

### Years of Citing Articles

### OpenURL

### Abstract

INTRODUCTION In this paper we consider Simulated Annealing algorithms (SA in what follows) applied to continuous global optimization problems, i.e. problems with the following form f = min x2X f(x); (1.1) where X ` ! n is a continuous domain, often assumed to be compact, which, combined with the continuity or lower semicontinuity of f , guarantees the existence of the minimum value f . SA algorithms are based on an analogy with a physical phenomenon: while at high temperatures the molecules in a liquid move freely, if the temperature is slowly decreased the thermal mobility of the molecules is lost and they form a pure crystal which also corresponds to a state of minimum energy. If the temperature is decreased too quickly (the so called quenching) a liquid metal rather ends up in a polycrystalline or amorphous state with

### Citations

2243 |
Equation of state calculations by fast computing machines
- Metropolis, AW, et al.
(Show Context)
Citation Context ...ollowing. Step 1. Let Y0∈X be a given starting point, let z0G{(Y0, f(Y0))}, f *0 Gf(Y0), and kG0. Step 2. Sample a point XkC1 from a distribution D(·; zk). Step 3. Sample a uniform random number p in =-=[0, 1]-=- and set YkC1G� XkC1, if pYA(Yk,XkC1,tk), Yk, otherwise, where A is a function with values in [0, 1] and tk is a parameter called the temperature at iteration k. Step 4. Set zkC1Gzk∪{(XkC1, f(XkC1))};... |

313 |
Thermodynamical approach to the traveling salesman problem: An efficient simulation algorithm
- Černý
- 1985
(Show Context)
Citation Context ...here D(·, ·) is a Markov kernel, absolutely continuous with respect to the Lebesgue measure m and with density β bounded away from zero, i.e., D(x, B)G� β(x, y) dy and inf β(x, y)GρH0, B x,y∈X ∀B∈B , =-=(2)-=- where B is a σ-algebra over X. (A2) For any G⊆X open, D(x, G) is continuous in x. (A3) For any initial state Y0 and any initial temperature t0, (A4) tkGU(zk)→0, (3) with probability 1. X is a compact... |

248 | Global Optimization: Deterministic Approaches (3rd edition - Horst, Tuy - 2003 |

232 | Efficient global optimization of expensive black-box functions - Jones, Schonlau, et al. - 1998 |

219 |
Cooling Schedules for Optimal Annealing
- Hajek
- 1988
(Show Context)
Citation Context ...ys implied by Assumption 3.8. Indeed, for any x∈B22¯ \B2/2, the probability in (12) can be split in the following sum P[ f(YkC1)Xf(Yk)Cuk�YkGx, RkGrk, zk] CP[ f(Yk)Ff(YkC1)Ff(Yk)Cuk�YkGx, RkGrk, zk]. =-=(13)-=-130 JOTA: VOL. 104, NO. 1, JANUARY 2000 By observing the Metropolis acceptance function (1), it follows that an upper bound for the first probability in (13) is the value exp{−(uk�ck)}, while from As... |

204 | Lipschitzian Optimization without the Lipschitz Constant - Jones, Perttunen, et al. - 1993 |

188 | The parallel genetic algorithm as function optimizer - Mühlenbein, Schomisch, et al. - 1991 |

181 | Very fast simulated re-annealing - Ingber - 1989 |

164 |
Minirnizing multimodal functions of continuous variables with the 'Simulated Annealing' algorithm
- Corana, Marchesi, et al.
- 1987
(Show Context)
Citation Context ...point XkC1 is equivalent to the uniform distribution over S(Yk, Rk)∩X; i.e., there exist constants g, GH0, gYG such that g[m(B)�m(S(Yk, Rk)∩X)]YD(B; zk)YG[m(B)�m(S(Yk, Rk)∩X)], ∀B⊆S(Yk, Rk)∩X, ∀Yk∈X. =-=(7)-=- The second assumption introduces restrictions on the objective function f and the feasible set X. Assumption 3.2. The feasible set X is compact, convex, and fulldimensional, and the objective functio... |

158 | Simulated annealing: Practice versus theory - Ingber - 1993 |

126 | Rogers(1994),“Global Optimization of Statistical Functions with Simulated Annealing - Goffe, Ferrier, et al. |

99 | B.: Genetic algorithms and very fast simulated reannealing: a comparison - Ingber, Rosen - 1992 |

95 |
Optimization by Simulated
- Kirkpatrick, Gellat, et al.
- 1983
(Show Context)
Citation Context ...nd inf β(x, y)GρH0, B x,y∈X ∀B∈B , (2) where B is a σ-algebra over X. (A2) For any G⊆X open, D(x, G) is continuous in x. (A3) For any initial state Y0 and any initial temperature t0, (A4) tkGU(zk)→0, =-=(3)-=- with probability 1. X is a compact set, f is a continuous function, and m(B2)H0, ∀2H0. Under the above assumptions, it has been proved that lim P[Yk∈B2]G1, k→S ∀2H0. Note that (3) is much less restri... |

76 | Fast simulated annealing - Szu, Hartley - 1987 |

73 | Global optimization by multilevel coordinate search - Huyer, Neumaier - 1999 |

71 | Adaptative simulated annealing (ASA): Lessons learned - Ingber - 1996 |

66 | Generalized simulated annealing for function optimization - Bohachevsky, Johnson, et al. - 1986 |

49 | A radial basis function method for global optimization - Gutmann |

46 |
A Monte Carlo simulated annealing approach to optimization over continuous variables
- Vanderbilt, Louie
- 1984
(Show Context)
Citation Context ...following result can be proved.128 JOTA: VOL. 104, NO. 1, JANUARY 2000 Theorem 3.1. let Assumptions 3.1–3.3 hold. Then, Let tk and Rk be given by (4) and (6), respectively, and P[∃k: Yk∈B2]G1, ∀2H0. =-=(10)-=- Proof. See Theorem 2 in Ref. 11. � Basically, Theorem 3.1 states that, under the given assumptions, the record value f *k converges to the global optimum value f * with probability 1. However, Assump... |

39 |
Convergence theorem for a class of simulated annealing algorithms on R d
- Belisle
- 1992
(Show Context)
Citation Context ...of accepting an ascent step decreases to 0 as the iteration counter k increases. Assumption 3.7. There exists a sequence bkGbk(2)→0 such that P[ f(YkC1)Hf(Yk)�YkGx, RkGrk, zk]Ybk, ∀x∈B22¯ \B2/2, ∀zk. =-=(12)-=- While Assumption 3.7 is what we need for the proof of (11), we would like to introduce an alternative assumption, which is less general than Assumption 3.7, but gives a better indication about how {c... |

33 | Efficient simulated annealing on fractal energy landscapes. Algorithmica 6, 367–418. ECCC ISSN 1433-8092 http://www.eccc.uni-trier.de/eccc ftp://ftp.eccc.uni-trier.de/pub/eccc ftpmail@ftp.eccc.uni-trier.de, subject ’help eccc - Sorkin - 1991 |

29 |
Simulated annealing for constrained global optimization
- Romeijn, Smith
- 1994
(Show Context)
Citation Context ...the assumption is the following. Assumption 3.3. The sequences {ck} and {rk} decrease to 0 slowly enough; specifically, the following condition must be satisfied: ∑ S ψ sq j exp{A[sqj ∆Fqj�cqjC1]}GS, =-=(9)-=- jG1 where ψ∈(0, 1) is a given constant. While condition (9) is quite general, it is quite difficult to appreciate its dependency on the two sequences {ck} and {rk}. In particular, while the sequence ... |

29 | Enhanced simulated annealing for globally minimizing functions of many continuous variables - Siarry, Berthiau, et al. - 1997 |

28 | Global optimization and simulated annealing - Dekkers, Aarts - 1991 |

27 | Asymptotic global behavior for stochastic approximation and diffusions with slowly decreasing noise effects: global minimization via Monte Carlo - Kushner - 1987 |

26 | Recursive stochastic algorithms for global optimization in R d - Gelfand, Mitter - 1991 |

26 |
S.K.,Metropolis-Type Annealing Algorithms for Global Optimization in R d
- GELFAND, MITTER
- 1993
(Show Context)
Citation Context ...rges with probability 1 to the global optimum. Theorem 4.2. Let 2˜ be the same as in Theorem 4.1. If Assumptions 3.1–3.7 hold, and if in (4) and (6) 2¯ ∈(0, 2˜ ] is chosen, then lim P[Xk∈B2]G1, ∀2H0. =-=(14)-=- k→S Proof. See Theorem 4 in Ref. 11. � As already remarked in the previous section, a drawback of the above result is that generally it is not possible to know in advance a suitable value for 2¯. How... |

25 | Global optimization and stochastic differential equations - Aluffi-Pentini, Parisi, et al. - 1985 |

22 | Diffusion for global optimization in R - Chiang, Hwang, et al. - 1987 |

19 | Isotropic effective energy simulated annealing searches for low energy molecular cluster states - Coleman, Shalloway, et al. - 1993 |

19 | A Parallel Buildup Algorithm for Global Energy Minimizations of Molecular Clusters Using Effective Energy VOL
- COLEMAN, SHALLOWAY, et al.
- 2000
(Show Context)
Citation Context ...phere for the distribution D(·; zk); i.e., we should have RkXδ 2H0, if f(Yk)Af *k H2¯, (6a) RkGrk, otherwise, (6b) where {rk} is a deterministic nonincreasing sequence converging to 0. The meaning of =-=(6)-=- is that, if we are in a point whose function value is poor with respect to the current record, then we guarantee the possibility of performing steps XkC1AYk whose length is bounded away from 0, while... |

19 | Towards global optimization 2 - Dixon, Szegö - 1978 |

17 | Simulated annealing process in general state space - Haario, Saksman - 1991 |

15 | D.J.; Evolution of the Potential Energy Surface with Size for Lennard-Jones Clusters - Doye, Miller, et al. - 1999 |

15 |
G.W.,An Adaptive Simulated Annealing Algorithm for Global Optimization over Continuous Variables
- JONES, FORBES
- 1995
(Show Context)
Citation Context ...the case of finitely many global optima is trivial). Note that the continuity of f and the compactness of X imply that f is uniformly continuous over X; i.e., ∀δH0, ∃γGγ (δ)H0:d(x,y)Yγ⇒�f(x)Af(x)�Yδ. =-=(8)-=-JOTA: VOL. 104, NO. 1, JANUARY 2000 127 Next, we introduce an assumption about the speed of convergence to 0 for the sequences {ck} and {rk}. First, we need to introduce some notation. Let skGmin� s:... |

14 | Optimization of functions with many minima - Bilbro, Snyder - 1991 |

14 | Molecular conformation on the CM5 by Parallel Two-Level Simulated annealing - Xue - 1994 |

13 | Rapid determination of the critical temperature in simulated annealing inversion - Basu, Frazer - 1990 |

12 | Simulated Annealing Type Algorithms for Multivariate Optimization,” Algorithmica - Gelfand, Mitter - 1991 |

12 | Minimum-Energy Configurations of Atomic Clusters: New Results Obtained by Simulated Annealing - Wille - 1987 |

12 | Applications of simulated annealing to peptides - Wilson - 1990 |

11 | Convergence of the simulated annealing algorithm for continuous global optimization - Yang |

10 | Random tunneling by means of acceptance-rejection sampling for global optimization - Lucidl, Piccioni - 1989 |

10 | Weak convergence rates for stochastic approximation with application to multiple targets and simulated annealing - Pelletier - 1998 |

9 | The double-funnel energy landscape of the 38-atom Lennard-Jones cluster - Doye, Miller, et al. - 1999 |

8 | Aspiration based simulated annealing algorithm - Ali, Storey - 1997 |

8 | Combining simulated annealing and local optimization for efficient global optimization - Desai, Pateil - 1996 |

8 | Large-time behavior of perturbed diffusion Markov processes with applications to the second eigenvalue problem for Fokker–Planck operators and simulated annealing - Hwang, Sheu - 1990 |

8 | AARTS, E.H.L.: Simulated Annealing: Theory and Applications - LAARHOVEN - 1987 |

7 | Simulated annealing and adaptive search in global optimization - Romeijn, Smith - 1994 |