scholarly journals Quality-Oriented Study on Mapping Island Model Genetic Algorithm onto CUDA GPU

Symmetry ◽  
2019 ◽  
Vol 11 (3) ◽  
pp. 318 ◽  
Author(s):  
Xue Sun ◽  
Ping Chou ◽  
Chao-Chin Wu ◽  
Liang-Rui Chen
Keyword(s):  
Genetic Algorithm ◽  
Execution Time ◽  
Large Scale ◽  
Parallel Architecture ◽  
Computation Time ◽  
Search Space ◽  
Island Model ◽  
Process Unit ◽  
Np Hard ◽  
Solution Quality

Genetic algorithm (GA), a global search method, has widespread applications in various fields. One very promising variant model of GA is the island model GA (IMGA) that introduces the key idea of migration to explore a wider search space. Migration will exchange chromosomes between islands, resulting in better-quality solutions. However, IMGA takes a long time to solve the large-scale NP-hard problems. In order to shorten the computation time, modern graphic process unit (GPU), as highly-parallel architecture, has been widely adopted in order to accelerate the execution of NP-hard algorithms. However, most previous studies on GPUs are focused on performance only, because the found solution qualities of the CPU and the GPU implementation of the same method are exactly the same. Therefore, it is usually previous work that did not report on quality. In this paper, we investigate how to find a better solution within a reasonable time when parallelizing IMGA on GPU, and we take the UA-FLP as a study example. Firstly, we propose an efficient approach of parallel tournament selection operator on GPU to achieve a better solution quality in a shorter amount of time. Secondly, we focus on how to tune three important parameters of IMGA to obtain a better solution efficiently, including the number of islands, the number of generations, and the number of chromosomes. In particular, different parameters have a different impact on solution quality improvement and execution time increment. We address the challenge of how to trade off between solution quality and execution time for these parameters. Finally, experiments and statistics are conducted to help researchers set parameters more efficiently to obtain better solutions when GPUs are used to accelerate IMGA. It has been observed that the order of influence on solution quality is: The number of chromosomes, the number of generations, and the number of islands, which can guide users to obtain better solutions efficiently with moderate increment of execution time. Furthermore, if we give higher priority on reducing execution time on GPU, the quality of the best solution can be improved by about 3%, with an acceleration that is 29 times faster than the CPU counterpart, after applying our suggested parameter settings. However, if we give solution quality a higher priority, i.e., the GPU execution time is close to the CPU’s, the solution quality can be improved up to 8%.

A Fast and Scalable Heuristic for the Solution of Large-Scale Capacitated Vehicle Routing Problems

2021 ◽  
Author(s):  
Luca Accorsi ◽  
Daniele Vigo
Keyword(s):  
Vehicle Routing ◽  
Large Scale ◽  
Computing Time ◽  
Search Space ◽  
Iterated Local Search ◽  
Solution Quality ◽  
Routing Problem ◽  
Acceptance Criterion ◽  
Acceleration Techniques ◽  
Capacitated Vehicle

In this paper, we propose a fast and scalable, yet effective, metaheuristic called FILO to solve large-scale instances of the Capacitated Vehicle Routing Problem. Our approach consists of a main iterative part, based on the Iterated Local Search paradigm, which employs a carefully designed combination of existing acceleration techniques, as well as novel strategies to keep the optimization localized, controlled, and tailored to the current instance and solution. A Simulated Annealing-based neighbor acceptance criterion is used to obtain a continuous diversification, to ensure the exploration of different regions of the search space. Results on extensively studied benchmark instances from the literature, supported by a thorough analysis of the algorithm’s main components, show the effectiveness of the proposed design choices, making FILO highly competitive with existing state-of-the-art algorithms, both in terms of computing time and solution quality. Finally, guidelines for possible efficient implementations, algorithm source code, and a library of reusable components are open-sourced to allow reproduction of our results and promote further investigations.

scholarly journals Space-time clustering-based method to optimize shareability in real-time ride-sharing

PLoS ONE ◽  
2022 ◽  
Vol 17 (1) ◽  
pp. e0262499
Author(s):  
Negin Alisoltani ◽  
Mostafa Ameli ◽  
Mahdi Zargayouna ◽  
Ludovic Leclercq
Keyword(s):  
Real Time ◽  
Large Scale ◽  
Computation Time ◽  
Clustering Method ◽  
Matching Problem ◽  
Solution Quality ◽  
Mobility Service ◽  
Ride Sharing ◽  
Large Scale Problems ◽  
Spatio Temporal

Real-time ride-sharing has become popular in recent years. However, the underlying optimization problem for this service is highly complex. One of the most critical challenges when solving the problem is solution quality and computation time, especially in large-scale problems where the number of received requests is huge. In this paper, we rely on an exact solving method to ensure the quality of the solution, while using AI-based techniques to limit the number of requests that we feed to the solver. More precisely, we propose a clustering method based on a new shareability function to put the most shareable trips inside separate clusters. Previous studies only consider Spatio-temporal dependencies to do clustering on the mobility service requests, which is not efficient in finding the shareable trips. Here, we define the shareability function to consider all the different sharing states for each pair of trips. Each cluster is then managed with a proposed heuristic framework in order to solve the matching problem inside each cluster. As the method favors sharing, we present the number of sharing constraints to allow the service to choose the number of shared trips. To validate our proposal, we employ the proposed method on the network of Lyon city in France, with half-million requests in the morning peak from 6 to 10 AM. The results demonstrate that the algorithm can provide high-quality solutions in a short time for large-scale problems. The proposed clustering method can also be used for different mobility service problems such as car-sharing, bike-sharing, etc.

Optimized Crossover JumpX in Genetic Algorithm for General Routing Problems

2018 ◽  
pp. 205-230
Author(s):  
Hicham El Hassani ◽  
Said Benkachcha ◽  
Jamal Benhra
Keyword(s):  
Genetic Algorithm ◽  
Optimization Methods ◽  
Search Space ◽  
Optimal Solutions ◽  
Solution Quality ◽  
Path Representation ◽  
Representation Technique ◽  
Hard Problems ◽  
The Traveling Salesman Problem

Inspired by nature, genetic algorithms (GA) are among the greatest meta-heuristics optimization methods that have proved their effectiveness to conventional NP-hard problems, especially the traveling salesman problem (TSP) which is one of the most studied Supply chain management problems. This paper proposes a new crossover operator called Jump Crossover (JMPX) for solving the travelling salesmen problem using a genetic algorithm (GA) for near-optimal solutions, to conclude on its efficiency compared to solutions quality given by other conventional operators to the same problem, namely, Partially matched crossover (PMX), Edge recombination Crossover (ERX) and r-opt heuristic with consideration of computational overload. We adopt the path representation technique for our chromosome which is the most direct representation and a low mutation rate to isolate the search space exploration ability of each crossover. The experimental results show that in most cases JMPX can remarkably improve the solution quality of the GA compared to the two existing classic crossover approaches and the r-opt heuristic.

scholarly journals A GPU-Enabled Compact Genetic Algorithm for Very Large-Scale Optimization Problems

Mathematics ◽  
2020 ◽  
Vol 8 (5) ◽  
pp. 758
Author(s):  
Andrea Ferigo ◽  
Giovanni Iacca
Keyword(s):  
Genetic Algorithm ◽  
Large Scale ◽  
Optimization Problems ◽  
Search Space ◽  
Large Scale Problems ◽  
Processing Power ◽  
Compact Genetic Algorithm ◽  
Estimation Of Distribution ◽  
Feasible Solutions ◽  
Compact Optimization

The ever-increasing complexity of industrial and engineering problems poses nowadays a number of optimization problems characterized by thousands, if not millions, of variables. For instance, very large-scale problems can be found in chemical and material engineering, networked systems, logistics and scheduling. Recently, Deb and Myburgh proposed an evolutionary algorithm capable of handling a scheduling optimization problem with a staggering number of variables: one billion. However, one important limitation of this algorithm is its memory consumption, which is in the order of 120 GB. Here, we follow up on this research by applying to the same problem a GPU-enabled “compact” Genetic Algorithm, i.e., an Estimation of Distribution Algorithm that instead of using an actual population of candidate solutions only requires and adapts a probabilistic model of their distribution in the search space. We also introduce a smart initialization technique and custom operators to guide the search towards feasible solutions. Leveraging the compact optimization concept, we show how such an algorithm can optimize efficiently very large-scale problems with millions of variables, with limited memory and processing power. To complete our analysis, we report the results of the algorithm on very large-scale instances of the OneMax problem.

scholarly journals Genetic Algorithm-Based Variable Value Control Method for Solving the Ground Target Attacking Weapon-Target Allocation Problem

2019 ◽  
Vol 2019 ◽  
pp. 1-9 ◽  
Author(s):  
Chao Wang ◽  
Guangyuan Fu ◽  
Daqiao Zhang ◽  
Hongqiao Wang ◽  
Jiufen Zhao
Keyword(s):  
Genetic Algorithm ◽  
Large Scale ◽  
Control Method ◽  
Search Space ◽  
Initial Population ◽  
Improved Genetic Algorithm ◽  
Mutation Strategy ◽  
Ground Target ◽  
Value Range ◽  
Target Allocation

Key ground targets and ground target attacking weapon types are complex and diverse; thus, the weapon-target allocation (WTA) problem has long been a great challenge but has not yet been adequately addressed. A timely and reasonable WTA scheme not only helps to seize a fleeting combat opportunity but also optimizes the use of weaponry resources to achieve maximum battlefield benefits at the lowest cost. In this study, we constructed a ground target attacking WTA (GTA-WTA) model and designed a genetic algorithm-based variable value control method to address the issue that some intelligent algorithms are too slow in resolving the problem of GTA-WTA due to the large scale of the problem or are unable to obtain a feasible solution. The proposed method narrows the search space and improves the search efficiency by constraining and controlling the variable value range of the individuals in the initial population and ensures the quality of the solution by improving the mutation strategy to expand the range of variables. The simulation results show that the improved genetic algorithm (GA) can effectively solve the large-scale GTA-WTA problem with good performance.

scholarly journals Improving Genetic Algorithm with Fine-Tuned Crossover and Scaled Architecture

2016 ◽  
Vol 2016 ◽  
pp. 1-10 ◽  
Author(s):  
Ajay Shrestha ◽  
Ausif Mahmood
Keyword(s):  
Genetic Algorithm ◽  
Evolutionary Biology ◽  
Optimization Problems ◽  
Search Space ◽  
Island Model ◽  
Fine Tuning ◽  
Small Subset ◽  
Combinatorial Optimization Problems ◽  
Mutation Operators ◽  
New Concepts

Genetic Algorithm (GA) is a metaheuristic used in solving combinatorial optimization problems. Inspired by evolutionary biology, GA uses selection, crossover, and mutation operators to efficiently traverse the solution search space. This paper proposes nature inspired fine-tuning to the crossover operator using the untapped idea of Mitochondrial DNA (mtDNA). mtDNA is a small subset of the overall DNA. It differentiates itself by inheriting entirely from the female, while the rest of the DNA is inherited equally from both parents. This unique characteristic of mtDNA can be an effective mechanism to identify members with similar genes and restrict crossover between them. It can reduce the rate of dilution of diversity and result in delayed convergence. In addition, we scale the well-known Island Model, where instances of GA are run independently and population members exchanged periodically, to a Continental Model. In this model, multiple web services are executed with each web service running an island model. We applied the concept of mtDNA in solving Traveling Salesman Problem and to train Neural Network for function approximation. Our implementation tests show that leveraging these new concepts of mtDNA and Continental Model results in relative improvement of the optimization quality of GA.

scholarly journals Design and performance analysis of global path planning techniques for autonomous mobile robots in grid environments

2017 ◽  
Vol 14 (2) ◽  
pp. 172988141666366 ◽  
Author(s):  
Imen Chaari ◽  
Anis Koubaa ◽  
Hachemi Bennaceur ◽  
Adel Ammar ◽  
Maram Alajlan ◽  
...  
Keyword(s):  
Path Planning ◽  
Execution Time ◽  
Large Scale ◽  
Optimal Path ◽  
Heuristic Methods ◽  
Solution Quality ◽  
Good Trade ◽  
And Performance ◽  
Path Planner ◽  
The Cost

This article presents the results of the 2-year iroboapp research project that aims at devising path planning algorithms for large grid maps with much faster execution times while tolerating very small slacks with respect to the optimal path. We investigated both exact and heuristic methods. We contributed with the design, analysis, evaluation, implementation and experimentation of several algorithms for grid map path planning for both exact and heuristic methods. We also designed an innovative algorithm called relaxed A-star that has linear complexity with relaxed constraints, which provides near-optimal solutions with an extremely reduced execution time as compared to A-star. We evaluated the performance of the different algorithms and concluded that relaxed A-star is the best path planner as it provides a good trade-off among all the metrics, but we noticed that heuristic methods have good features that can be exploited to improve the solution of the relaxed exact method. This led us to design new hybrid algorithms that combine our relaxed A-star with heuristic methods which improve the solution quality of relaxed A-star at the cost of slightly higher execution time, while remaining much faster than A* for large-scale problems. Finally, we demonstrate how to integrate the relaxed A-star algorithm in the robot operating system as a global path planner and show that it outperforms its default path planner with an execution time 38% faster on average.

scholarly journals Parallel Implementation of Non-slicing Floorplans with MPI and OpenMP

2021 ◽  
Author(s):  
Oluvaseun Owojaiye
Keyword(s):  
Message Passing ◽  
Large Scale ◽  
Message Passing Interface ◽  
Parallel Implementation ◽  
Computation Time ◽  
Sequential Algorithm ◽  
Design Stage ◽  
Single Chip ◽  
Solution Quality ◽  
Early Design Stage

Advancement in technology has brought considerable improvement to processor design and now manufacturers design multiple processors on a single chip. Supercomputers today consists of cluster of interconnected nodes that collaborate together to solve complex and advanced computation problems. Message Passing Interface and Open Multiprocessing are the popularly used programming models to optimize sequential codes by parallelizing them on the different multiprocessor architecture that exist today. In this thesis, we parallelize the non-slicing floorplan algorithm based on Multilevel Floorplanning/placement of large scale modules using B*tree (MB*tree) with MPI and OpenMP on distributed and shared memory architectures respectively. In VLSI (Very Large Scale Integration) design automation, floorplanning is an initial and vital task performed in the early design stage. Experimental results using MCNC benchmark circuits show that our parallel algorithm produced better results than the corresponding sequential algorithm; we were able to speed up the algorithm up to 4 times, hence reducing computation time and maintaining floorplan solution quality. On the other hand, we compared both parallel versions; and the OpenMP results gave slightly better than the corresponding MPI results.

scholarly journals Evolutionary Models applied to Multiprocessor TaskScheduling: Serial and Multipopulation Genetic Algorithm

2019 ◽  
Vol 26 (1) ◽  
pp. 11
Author(s):  
Bruno Well Dantas Morais ◽  
Gina Maira Barbosa de Oliveira ◽  
Tiago Ismailer De Carvalho
Keyword(s):  
Genetic Algorithm ◽  
Execution Time ◽  
Evolutionary Models ◽  
Solution Quality ◽  
Communication Costs

This work presents the development of a multipopulation genetic algorithm for the task schedulingproblem with communication costs, aiming to compare its performance with the serial genetic algorithm. For thispurpose, a set of instances was developed and different approaches for genetic operations were compared.Experiments were conducted varying the number of populations and the number of processors available forscheduling. Solution quality and execution time were analyzed, and results show that the AGMP with adjustedparameters generally produces better solutions while requiring less execution time.

Toll Pricing and Heterogeneous Users

2005 ◽  
Vol 1923 (1) ◽  
pp. 28-36 ◽  
Author(s):  
Hani S. Mahmassani ◽  
Xuesong Zhou ◽  
Chung-Cheng Lu
Keyword(s):  
Shortest Path ◽  
Large Scale ◽  
Approximation Error ◽  
Computation Time ◽  
Time Dependent ◽  
Approximation Schemes ◽  
Value Of Time ◽  
Solution Quality ◽  
Shortest Path Problems ◽  
Toll Pricing

This paper presents both exact and approximation algorithms for finding extreme efficient time-dependent shortest paths for use with dynamic traffic assignment applications to networks with variable toll pricing and heterogeneous users (with different value of time preferences). A parametric least-generalized cost path algorithm is presented to determine a complete set of extreme efficient time-dependent paths that simultaneously consider travel time and cost criteria. However, exact procedures may not be practical for large networks. For this reason, approximation schemes are devised and tested. Based on the concept of ε-efficiency in multiobjective shortest path problems, a binary search framework is developed to find a set of extreme efficient paths that minimize expected approximation error, with the use of the underlying value of time distribution. Both exact and approximation schemes (along with variants) are tested on three actual traffic networks. The experimental results indicate that the computation time and the size of the solution set are jointly determined by several key parameters such as the number of time intervals and the number of nodes in the network. The results also suggest that the proposed approximation scheme is computationally efficient for large-scale bi-objective time-dependent shortest path applications while maintaining satisfactory solution quality.

Sign in / Sign up

Export Citation Format

Share Document