George Anescu - Academia.edu (original) (raw)
Uploads
Papers by George Anescu
Since its creation in 2005 by D. Karaboga the ABC algorithm proved to be very effective in approa... more Since its creation in 2005 by D. Karaboga the ABC algorithm proved to be very effective in approaching a wide variety of research optimization problems. However, some drawbacks were also experienced related mainly to a poor exploitation capability (which makes the algorithm relatively slow) and poor success rates when highly non-linear optimization problems with unstructured modes are approached. In order to improve the performance of the ABC algorithm, in both efficiency and success rate, the paper presents a set of proposed enhancements to the original ABC algorithm. The novel proposed ABC variant, Fast ABC (F-ABC), was tested against two known variants of ABC, the original algorithm proposed by D. Karaboga ([1]), and an improved variant, Gbest-guided Artificial Bee Colony (GABC) ([2]). The testing was conducted by employing an original testing methodology over a set of 11 scalable, multimodal, continuous optimization functions (10 unconstrained and 1 constrained) with known globa...
The paper is generalizing and enhancing the modeling principles and concepts introduced by the au... more The paper is generalizing and enhancing the modeling principles and concepts introduced by the authors in previous research in order to surpass known drawbacks affecting the performance of current 3D modeling software applications. The proposed modeling method presents increased flexibility in modifying and reusing existing models and is alleviating the difficulties related to mastering advanced mathematical mechanisms by providing improved visual feedback to the user. An experimental 3D modeling software application was implemented and some of its applications in jewelry design are described.
Review of the Air Force Academy, Oct 20, 2017
The paper is investigating the suitability of the FSA-DE optimization method for solving reliabil... more The paper is investigating the suitability of the FSA-DE optimization method for solving reliability optimization problems by approaching a set of three case studies: a known RAP case study, a FTO case study and a ETO case study. For the RAP case study the numerical results obtained by FSA-DE are compared with the ones obtained by other known optimization methods.
Since its creation in 2005 by D. Karaboga the ABC algorithm proved to be very effective in approa... more Since its creation in 2005 by D. Karaboga the ABC algorithm proved to be very effective in approaching a wide variety of research optimization problems. However, some drawbacks were also experienced related mainly to a poor exploitation capability (which makes the algorithm relatively slow) and poor success rates when highly non-linear optimization problems with unstructured modes are approached. In order to improve the performance of the ABC algorithm, in both efficiency and success rate, the paper presents a set of proposed enhancements to the original ABC algorithm. The novel proposed ABC variant, Fast ABC (F-ABC), was tested against two known variants of ABC, the original algorithm proposed by D. Karaboga ([1]), and an improved variant, Gbest-guided Artificial Bee Colony (GABC) ([2]). The testing was conducted by employing an original testing methodology over a set of 11 scalable, multimodal, continuous optimization functions (10 unconstrained and 1 constrained) with known globa...
2017 19th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC)
Journal of Advances in Mathematics and Computer Science
Journal of Advances in Mathematics and Computer Science
Journal of Advances in Mathematics and Computer Science
Multidimensional scalable test functions are very important in testing the capabilities of new op... more Multidimensional scalable test functions are very important in testing the capabilities of new optimization methods, especially in evaluating their response to the increase of the search space dimension. As a continuation of a previous published paper, new sets of test functions for continuous optimization are proposed, both unconstrained (or only box constrained, 7 new test functions) and constrained (10 new test functions).
Computers & Industrial Engineering
2015 17th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC), 2015
Journal of Advances in Mathematics and Computer Science
Multidimensional scalable test functions are very important in testing the capabilities of new op... more Multidimensional scalable test functions are very important in testing the capabilities of new optimization methods, especially in evaluating their response to the increase of the search space dimension. As a continuation of a previous published paper, new sets of test functions for continuous optimization are proposed, both unconstrained (or only box constrained, 7 new test functions) and constrained (10 new test functions).
2016 International Symposium on INnovations in Intelligent SysTems and Applications (INISTA), 2016
Reliability Engineering & System Safety, 1991
2015 17th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC), 2015
2014 16th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing, 2014
Clustering optimization methods for continuous numerical multivariable functions have been used f... more Clustering optimization methods for continuous numerical multivariable functions have been used for increasing the efficiency in the selection of the start points in multi-start global optimization methods. Methods of this kind usually have three steps: (1) sampling points in the search domain, (2) transforming the sampled points in order to obtain points grouped in neighbourhoods of local optima, (3) using a clustering technique to identify the clusters. After the clusters are successfully identified , the set of local optima (and from it the global optimum) can be easily determined by applying a local optimization method for each cluster. The novel Particle Swarm Clustering Optimization (PSCO) method proposed in this paper is concerned with simultaneous integration of steps (1), (2) and (3) from the classical clustering optimization methods by applying Swarm Intelligence (SI) techniques. Two existing SI methods provided inspiration in the design of the PSCO method: Particle Swarm Optimization (PSO) and Firefly Algorithm (FA).
The paper presents the experimental results of some tests conducted with the purpose to gradually... more The paper presents the experimental results of some tests conducted with the purpose to gradually and cumulatively improve the classical DE scheme in both efficiency and success rate. The modifications consisted in the randomization of the scaling factor (a simple jitter scheme), a more efficient Random Greedy Selection scheme, an adaptive scheme for the crossover probability and a resetting mechanism for the agents. After each modification step, experiments have been conducted on a set of 11 scalable, multimodal, continuous optimization functions in order to analyze the improvements and decide the new improvement direction. Finally, only the initial classical scheme and the constructed Fast Self-Adaptive DE (FSA-DE) variant were compared with the purpose of testing their performance degradation with the increase of the search space dimension. The experimental results demonstrated the superiority of the proposed FSA-DE variant.
The problem of economic optimization of Shell-and-Tube Heat Ex-changers (ST HE) is well known in ... more The problem of economic optimization of Shell-and-Tube Heat Ex-changers (ST HE) is well known in the literature. Since traditional design approaches do not guarantee the reach of the optimal solution, some heuris-tic approaches were developed and their results are published in the literature. Here is proposed a new method inspired from the multistart methods and Ant Colony Optimization (ACO) methods for continuous problems, the Distributed MultiStart Ant Colony Optimization (DM SACO) method. As a local optimization method used by DM SACO a novel Swarm Intelligence (SI) method is proposed, the Particle Swarm Local Optimization (P SLO) method. The results for two case studies are finally compared to those obtained by other approaches from literature, proving that the DM SACO global optimization method combined with P SLO local optimization method can be successfully applied to the ST HE economic design problem.
Conference Presentations by George Anescu
Problema optimizării economice a schimbătorului de căldură cu țevi și manta (STHE) este o problem... more Problema optimizării economice a schimbătorului de căldură cu țevi și manta (STHE) este o problemă de optimizare deschisă cunoscută. Deoarece abordările de proiectare tradiționale nu garantează obținerea soluției optime, au fost dezvoltate unele abordări euristice și rezultatele acestora sunt publicate în literatură. În articolul prezent propunem o metodă nouă inspirată din metodele PSO canonică și ABC, metoda PSO fără Viteze și Coeficienți (NSC-PSO). Rezultatele pentru două studii de caz sunt în final comparate cu cele obținute de către alte abordări din literatură, demonstrându-se faptul că metoda de optimizare globală NSC-PSO poate fi aplicată cu succes la problema proiectării economice a STHE. The problem of economic optimization of Shell-and-Tube Heat Exchanger (STHE) is a known open optimization problem. Since traditional design approaches do not guarantee the reach of the optimal solution, some heuristic approaches were developed and their results are published in the literature. In the present paper we propose a new method inspired from the canonical PSO and ABC methods, No Speeds and Coefficients PSO (NSC-PSO) method. The results for two case studies are finally compared to those obtained by other approaches from literature, proving that the NSC-PSO global optimization method can be successfully applied to the STHE economic design problem.
Since its creation in 2005 by D. Karaboga the ABC algorithm proved to be very effective in approa... more Since its creation in 2005 by D. Karaboga the ABC algorithm proved to be very effective in approaching a wide variety of research optimization problems. However, some drawbacks were also experienced related mainly to a poor exploitation capability (which makes the algorithm relatively slow) and poor success rates when highly non-linear optimization problems with unstructured modes are approached. In order to improve the performance of the ABC algorithm, in both efficiency and success rate, the paper presents a set of proposed enhancements to the original ABC algorithm. The novel proposed ABC variant, Fast ABC (F-ABC), was tested against two known variants of ABC, the original algorithm proposed by D. Karaboga ([1]), and an improved variant, Gbest-guided Artificial Bee Colony (GABC) ([2]). The testing was conducted by employing an original testing methodology over a set of 11 scalable, multimodal, continuous optimization functions (10 unconstrained and 1 constrained) with known globa...
The paper is generalizing and enhancing the modeling principles and concepts introduced by the au... more The paper is generalizing and enhancing the modeling principles and concepts introduced by the authors in previous research in order to surpass known drawbacks affecting the performance of current 3D modeling software applications. The proposed modeling method presents increased flexibility in modifying and reusing existing models and is alleviating the difficulties related to mastering advanced mathematical mechanisms by providing improved visual feedback to the user. An experimental 3D modeling software application was implemented and some of its applications in jewelry design are described.
Review of the Air Force Academy, Oct 20, 2017
The paper is investigating the suitability of the FSA-DE optimization method for solving reliabil... more The paper is investigating the suitability of the FSA-DE optimization method for solving reliability optimization problems by approaching a set of three case studies: a known RAP case study, a FTO case study and a ETO case study. For the RAP case study the numerical results obtained by FSA-DE are compared with the ones obtained by other known optimization methods.
Since its creation in 2005 by D. Karaboga the ABC algorithm proved to be very effective in approa... more Since its creation in 2005 by D. Karaboga the ABC algorithm proved to be very effective in approaching a wide variety of research optimization problems. However, some drawbacks were also experienced related mainly to a poor exploitation capability (which makes the algorithm relatively slow) and poor success rates when highly non-linear optimization problems with unstructured modes are approached. In order to improve the performance of the ABC algorithm, in both efficiency and success rate, the paper presents a set of proposed enhancements to the original ABC algorithm. The novel proposed ABC variant, Fast ABC (F-ABC), was tested against two known variants of ABC, the original algorithm proposed by D. Karaboga ([1]), and an improved variant, Gbest-guided Artificial Bee Colony (GABC) ([2]). The testing was conducted by employing an original testing methodology over a set of 11 scalable, multimodal, continuous optimization functions (10 unconstrained and 1 constrained) with known globa...
2017 19th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC)
Journal of Advances in Mathematics and Computer Science
Journal of Advances in Mathematics and Computer Science
Journal of Advances in Mathematics and Computer Science
Multidimensional scalable test functions are very important in testing the capabilities of new op... more Multidimensional scalable test functions are very important in testing the capabilities of new optimization methods, especially in evaluating their response to the increase of the search space dimension. As a continuation of a previous published paper, new sets of test functions for continuous optimization are proposed, both unconstrained (or only box constrained, 7 new test functions) and constrained (10 new test functions).
Computers & Industrial Engineering
2015 17th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC), 2015
Journal of Advances in Mathematics and Computer Science
Multidimensional scalable test functions are very important in testing the capabilities of new op... more Multidimensional scalable test functions are very important in testing the capabilities of new optimization methods, especially in evaluating their response to the increase of the search space dimension. As a continuation of a previous published paper, new sets of test functions for continuous optimization are proposed, both unconstrained (or only box constrained, 7 new test functions) and constrained (10 new test functions).
2016 International Symposium on INnovations in Intelligent SysTems and Applications (INISTA), 2016
Reliability Engineering & System Safety, 1991
2015 17th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC), 2015
2014 16th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing, 2014
Clustering optimization methods for continuous numerical multivariable functions have been used f... more Clustering optimization methods for continuous numerical multivariable functions have been used for increasing the efficiency in the selection of the start points in multi-start global optimization methods. Methods of this kind usually have three steps: (1) sampling points in the search domain, (2) transforming the sampled points in order to obtain points grouped in neighbourhoods of local optima, (3) using a clustering technique to identify the clusters. After the clusters are successfully identified , the set of local optima (and from it the global optimum) can be easily determined by applying a local optimization method for each cluster. The novel Particle Swarm Clustering Optimization (PSCO) method proposed in this paper is concerned with simultaneous integration of steps (1), (2) and (3) from the classical clustering optimization methods by applying Swarm Intelligence (SI) techniques. Two existing SI methods provided inspiration in the design of the PSCO method: Particle Swarm Optimization (PSO) and Firefly Algorithm (FA).
The paper presents the experimental results of some tests conducted with the purpose to gradually... more The paper presents the experimental results of some tests conducted with the purpose to gradually and cumulatively improve the classical DE scheme in both efficiency and success rate. The modifications consisted in the randomization of the scaling factor (a simple jitter scheme), a more efficient Random Greedy Selection scheme, an adaptive scheme for the crossover probability and a resetting mechanism for the agents. After each modification step, experiments have been conducted on a set of 11 scalable, multimodal, continuous optimization functions in order to analyze the improvements and decide the new improvement direction. Finally, only the initial classical scheme and the constructed Fast Self-Adaptive DE (FSA-DE) variant were compared with the purpose of testing their performance degradation with the increase of the search space dimension. The experimental results demonstrated the superiority of the proposed FSA-DE variant.
The problem of economic optimization of Shell-and-Tube Heat Ex-changers (ST HE) is well known in ... more The problem of economic optimization of Shell-and-Tube Heat Ex-changers (ST HE) is well known in the literature. Since traditional design approaches do not guarantee the reach of the optimal solution, some heuris-tic approaches were developed and their results are published in the literature. Here is proposed a new method inspired from the multistart methods and Ant Colony Optimization (ACO) methods for continuous problems, the Distributed MultiStart Ant Colony Optimization (DM SACO) method. As a local optimization method used by DM SACO a novel Swarm Intelligence (SI) method is proposed, the Particle Swarm Local Optimization (P SLO) method. The results for two case studies are finally compared to those obtained by other approaches from literature, proving that the DM SACO global optimization method combined with P SLO local optimization method can be successfully applied to the ST HE economic design problem.
Problema optimizării economice a schimbătorului de căldură cu țevi și manta (STHE) este o problem... more Problema optimizării economice a schimbătorului de căldură cu țevi și manta (STHE) este o problemă de optimizare deschisă cunoscută. Deoarece abordările de proiectare tradiționale nu garantează obținerea soluției optime, au fost dezvoltate unele abordări euristice și rezultatele acestora sunt publicate în literatură. În articolul prezent propunem o metodă nouă inspirată din metodele PSO canonică și ABC, metoda PSO fără Viteze și Coeficienți (NSC-PSO). Rezultatele pentru două studii de caz sunt în final comparate cu cele obținute de către alte abordări din literatură, demonstrându-se faptul că metoda de optimizare globală NSC-PSO poate fi aplicată cu succes la problema proiectării economice a STHE. The problem of economic optimization of Shell-and-Tube Heat Exchanger (STHE) is a known open optimization problem. Since traditional design approaches do not guarantee the reach of the optimal solution, some heuristic approaches were developed and their results are published in the literature. In the present paper we propose a new method inspired from the canonical PSO and ABC methods, No Speeds and Coefficients PSO (NSC-PSO) method. The results for two case studies are finally compared to those obtained by other approaches from literature, proving that the NSC-PSO global optimization method can be successfully applied to the STHE economic design problem.
