Stochastic optimization of GeantV code by use of genetic algorithms
Abstract
GeantV is a complex system based on the interaction of different modules needed for detector simulation, which include transport of particles in fields, physics models simulating their interactions with matter and a geometrical modeler library for describing the detector and locating the particles and computing the path length to the current volume boundary. The GeantV project is recasting the classical simulation approach to get maximum benefit from SIMD/MIMD computational architectures and highly massive parallel systems. This involves finding the appropriate balance between several aspects influencing computational performance (floatingpoint performance, usage of offchip memory bandwidth, specification of cache hierarchy, etc.) and handling a large number of program parameters that have to be optimized to achieve the best simulation throughput. This optimization task can be treated as a blackbox optimization problem, which requires searching the optimum set of parameters using only pointwise function evaluations. Here, the goal of this study is to provide a mechanism for optimizing complex systems (high energy physics particle transport simulations) with the help of genetic algorithms and evolution strategies as tuning procedures for massive parallel simulations. One of the described approaches is based on introducing a specific multivariate analysis operator that could be used in case ofmore »
 Authors:
 more »
 Sao Paulo State Univ. (UNESP), Sao Paulo (Brazil). Parallel Computing Center
 European Organization for Nuclear Research (CERN), Meyrin (Switzerland)
 Bhabha Atomic Research Centre (BARC), Mumbai (India)
 Fermi National Accelerator Lab. (FNAL), Batavia, IL (United States)
 Intel Corporation, Santa Clara, CA (United States)
 European Organization for Nuclear Research (CERN), Meyrin (Switzerland); Intel Corporation, Santa Clara, CA (United States)
 European Organization for Nuclear Research (CERN), Meyrin (Switzerland); Inst. of Space Sciences, BucharestMagurele (Romania)
 Publication Date:
 Research Org.:
 Fermi National Accelerator Lab. (FNAL), Batavia, IL (United States)
 Sponsoring Org.:
 USDOE Office of Science (SC), High Energy Physics (HEP) (SC25)
 OSTI Identifier:
 1421538
 Report Number(s):
 FERMILABCONF16766CD
Journal ID: ISSN 17426588; 1638148
 Grant/Contract Number:
 AC0207CH11359
 Resource Type:
 Journal Article: Accepted Manuscript
 Journal Name:
 Journal of Physics. Conference Series
 Additional Journal Information:
 Journal Volume: 898; Journal Issue: 4; Journal ID: ISSN 17426588
 Publisher:
 IOP Publishing
 Country of Publication:
 United States
 Language:
 English
 Subject:
 72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS
Citation Formats
Amadio, G., Apostolakis, J., Bandieramonte, M., Behera, S. P., Brun, R., Canal, P., Carminati, F., Cosmo, G., Duhem, L., Elvira, D., Folger, G., Gheata, A., Gheata, M., Goulas, I., Hariri, F., Jun, S. Y., Konstantinov, D., Kumawat, H., Ivantchenko, V., Lima, G., Nikitina, T., Novak, M., Pokorski, W., Ribon, A., Seghal, R., Shadura, O., Vallecorsa, S., and Wenzel, S. Stochastic optimization of GeantV code by use of genetic algorithms. United States: N. p., 2017.
Web. doi:10.1088/17426596/898/4/042026.
Amadio, G., Apostolakis, J., Bandieramonte, M., Behera, S. P., Brun, R., Canal, P., Carminati, F., Cosmo, G., Duhem, L., Elvira, D., Folger, G., Gheata, A., Gheata, M., Goulas, I., Hariri, F., Jun, S. Y., Konstantinov, D., Kumawat, H., Ivantchenko, V., Lima, G., Nikitina, T., Novak, M., Pokorski, W., Ribon, A., Seghal, R., Shadura, O., Vallecorsa, S., & Wenzel, S. Stochastic optimization of GeantV code by use of genetic algorithms. United States. doi:10.1088/17426596/898/4/042026.
Amadio, G., Apostolakis, J., Bandieramonte, M., Behera, S. P., Brun, R., Canal, P., Carminati, F., Cosmo, G., Duhem, L., Elvira, D., Folger, G., Gheata, A., Gheata, M., Goulas, I., Hariri, F., Jun, S. Y., Konstantinov, D., Kumawat, H., Ivantchenko, V., Lima, G., Nikitina, T., Novak, M., Pokorski, W., Ribon, A., Seghal, R., Shadura, O., Vallecorsa, S., and Wenzel, S. 2017.
"Stochastic optimization of GeantV code by use of genetic algorithms". United States.
doi:10.1088/17426596/898/4/042026. https://www.osti.gov/servlets/purl/1421538.
@article{osti_1421538,
title = {Stochastic optimization of GeantV code by use of genetic algorithms},
author = {Amadio, G. and Apostolakis, J. and Bandieramonte, M. and Behera, S. P. and Brun, R. and Canal, P. and Carminati, F. and Cosmo, G. and Duhem, L. and Elvira, D. and Folger, G. and Gheata, A. and Gheata, M. and Goulas, I. and Hariri, F. and Jun, S. Y. and Konstantinov, D. and Kumawat, H. and Ivantchenko, V. and Lima, G. and Nikitina, T. and Novak, M. and Pokorski, W. and Ribon, A. and Seghal, R. and Shadura, O. and Vallecorsa, S. and Wenzel, S.},
abstractNote = {GeantV is a complex system based on the interaction of different modules needed for detector simulation, which include transport of particles in fields, physics models simulating their interactions with matter and a geometrical modeler library for describing the detector and locating the particles and computing the path length to the current volume boundary. The GeantV project is recasting the classical simulation approach to get maximum benefit from SIMD/MIMD computational architectures and highly massive parallel systems. This involves finding the appropriate balance between several aspects influencing computational performance (floatingpoint performance, usage of offchip memory bandwidth, specification of cache hierarchy, etc.) and handling a large number of program parameters that have to be optimized to achieve the best simulation throughput. This optimization task can be treated as a blackbox optimization problem, which requires searching the optimum set of parameters using only pointwise function evaluations. Here, the goal of this study is to provide a mechanism for optimizing complex systems (high energy physics particle transport simulations) with the help of genetic algorithms and evolution strategies as tuning procedures for massive parallel simulations. One of the described approaches is based on introducing a specific multivariate analysis operator that could be used in case of resource expensive or time consuming evaluations of fitness functions, in order to speedup the convergence of the blackbox optimization problem.},
doi = {10.1088/17426596/898/4/042026},
journal = {Journal of Physics. Conference Series},
number = 4,
volume = 898,
place = {United States},
year = 2017,
month =
}

Fuel management optimization is a hard problem for traditional optimization techniques. Loading pattern optimization is a large combinatorial problem without analytical derivative information. Therefore, methods designed for continuous functions, such as linear programming, do not always work well. Genetic algorithms (GAs) address these problems and, therefore, appear ideal for fuel management optimization. They do not require derivative information and work well with combinatorial. functions. The GAs are a stochastic method based on concepts from biological genetics. They take a group of candidate solutions, called the population, and use selection, crossover, and mutation operators to create the next generation of bettermore »

Stochastic reservoir modeling using simulated annealing and genetic algorithms
This paper discusses and compares three different algorithms based on combinatorial optimization schemes for generating stochastic permeability fields. The algorithms are not restricted to generating Gaussian random fields and have the potential to accomplish geologic realism by combining data from many different sources. The authors have introduced a ``heatbath`` algorithm for simulated annealing (SA) as an alternative to the commonly used ``Metropolis`` algorithm and a new stochastic modeling technique based on the ``genetic`` algorithm. The authors applied these algorithms to a set of outcrop and tracer flow data and examined the associated uncertainties in predictions. All three algorithms reproduce themore » 
Nuclear fuel management optimization using genetic algorithms
The code independent genetic algorithm reactor optimization (CIGARO) system has been developed to optimize nuclear reactor loading patterns. It uses genetic algorithms (GAs) and a codeindependent interface, so any reactor physics code (e.g., CASMO3/SIMULATE3) can be used to evaluate the loading patterns. The system is compared to other GAbased loading pattern optimizers. Tests were carried out to maximize the beginning of cycle k{sub eff} for a pressurized water reactor core loading with a penalty function to limit power peaking. The CIGARO system performed well, increasing the k{sub eff} after lowering the peak power. Tests of a prototype parallel evaluation methodmore » 
Global optimization of a dryer by using neural networks and genetic algorithms
For many optimum design problems, the objective function is the result of a complex numerical code and may not be differentiable and explicit. The first aim is to propose a way of solving such complexity on an example problem. A novel and global strategy involving artificial neural networks and a genetic algorithm is presented and validated for an industrial convective dryer. To begin with, a method to represent a drying model using artificial neural networks is defined. This method is tested and the results are compared with those obtained with classic numerical methods. This approach drastically reduces simulation times andmore » 
Fuel management optimization using genetic algorithms and expert knowledge
The CIGARO fuel management optimization code based on genetic algorithms is described and tested. The test problem optimized the core lifetime for a pressurized water reactor with a penalty function constraint on the peak normalized power. A bitstring genotype encoded the loading patterns, and genotype bias was reduced with additional bits. Expert knowledge about fuel management was incorporated into the genetic algorithm. Regional crossover exchanged physically adjacent fuel assemblies and improved the optimization slightly. Biasing the initial population toward a known priority table significantly improved the optimization.