In current decades, various evolutionary algorithms(EAs) raise as well as many kinds of benchmarks are popular in evaluations of EAs' performances. Since there exists randomness in EAs' performances, the evalu...
详细信息
In current decades, various evolutionary algorithms(EAs) raise as well as many kinds of benchmarks are popular in evaluations of EAs' performances. Since there exists randomness in EAs' performances, the evaluations are made by a large number of runs in simulations or experiments in order to present a relatively fair comparison. However, there still exit several problems that have not been well explained. Does it make sense to deem two algorithms have equal ability if they have same final results? Is it convinced to decide winners or losers in comparisons just by tiny difference in performances? Besides the final results, how to compare algorithms' performances during the optimization iterations? In this paper, a neural network classifier based on extreme learning machine (ELM) is proposed to solve these problems. A novel role of classifier is first proposed to convince the differences between algorithms. If the classifier succeeds to classify algorithms based on their performances recorded in all generations, we deem the two algorithms have so convinced difference that comparisons of two algorithms can reflect algorithms' disparity. Therefore, the conclusions to judge the two algorithms are feasible and acceptable. Otherwise, if classifiers cannot distinguish two algorithms, we deem the two have similar performances so that it is meaningless to differ two algorithms just by tiny differences. By employing a set of classical benchmarks and six EAs, the simulations and computations are conducted. According to the analysis results, the proposed classifier can provide more information to reflect true abilities of algorithms, which is a novel view to compare EAs. (C) 2015 Elsevier B.V. All rights reserved.
Distribution system problems, such as planning, loss minimization, and energy restoration, usually involve network reconfiguration procedures. The determination of an optimal network configuration is, in general, a co...
详细信息
Distribution system problems, such as planning, loss minimization, and energy restoration, usually involve network reconfiguration procedures. The determination of an optimal network configuration is, in general, a combinatorial optimization problem. Several evolutionary algorithms (EAs) have been proposed to deal with this complex problem. Encouraging results have been achieved by using such approaches. However, the running time may be very high or even prohibitive in applications of EAs to large-scale networks. This limitation may be critical for problems requiring online solutions. The performance obtained by EAs for network reconfiguration is drastically affected by the adopted computational tree representation. Inadequate representations may drastically reduce the algorithm performance. Thus, the employed representation for chromosome encoding and the corresponding operators are very important for the performance achieved. An efficient data structure for tree representation may significantly increase the performance of evolutionary-based approaches for network reconfiguration problems. The present paper proposes a tree encoding and two genetic operators to improve the EA performance for network reconfiguration problems. The corresponding EA approach was applied to reconfigure large-scale systems. The performance achieved suggests that the proposed methodology can provide an efficient-alternative for reconfiguration problems.
The deployment of an unmanned aerial network (UAV-network) for the optimal coverage of ground nodes is an NP-hard problem. This work focuses on the application of a multi-layout multi-subpopulation genetic algorithm (...
详细信息
The deployment of an unmanned aerial network (UAV-network) for the optimal coverage of ground nodes is an NP-hard problem. This work focuses on the application of a multi-layout multi-subpopulation genetic algorithm (MLMPGA) to solve multi-objective coverage problems of UAV-networks. The multi objective deployment is based on a weighted fitness function that takes into account coverage, fault tolerance, and redundancy as relevant factors to optimally place the UAVs. The proposed approach takes advantage of different subpopulations evolving with different layouts. This feature is aimed at reflecting the evolutionary concept of different species adapting to the search space conditions of the multi-objective coverage problem better than single-population genetic algorithms. The proposed multi-subpopulation genetic algorithm is evaluated and compared against single-population genetic algorithm configurations and other well-known meta-heuristic optimization algorithms, such as particle swarm optimization and hill climbing algorithm, under different numbers of ground nodes. The proposed MLMPGA achieves significantly better performance results than the other meta-heuristic algorithms, such as classical genetic algorithms, hill climbing algorithm, and particle swarm optimization, in the vast majority of our simulation scenarios. (C) 2017 Published by Elsevier B.V.
This study aims to investigate the deployment of a proposed search field division method (SFDM) within evolutionary algorithms (EAs) to enhance the capability of searching for the global optima in nonlinear problems. ...
详细信息
This study aims to investigate the deployment of a proposed search field division method (SFDM) within evolutionary algorithms (EAs) to enhance the capability of searching for the global optima in nonlinear problems. The proposed technique is benchmarked against the following eight widely-used single-modal, multi-modal, and unimodal benchmark functions: Sphere, Rosenbrock, Rastringin, Griewank, Ackley, Fletcher, Quartic, and Schwefel functions, and the outcome is compared to their standard EAs counterparts to validate the effectiveness of the deployed approach in EAs. In the proposed method, we apply three low, medium, high field divisions (1, 2, and 5) dimensions on nine different EAs simultaneously with two different scenarios, 10 and 100 variables, to reach the optimal solution. Then for the validity of our proposed SFDM technique, we examined the exploration-exploitation search space rates and diversity behavior. The results of the implementation of SFDM on eight benchmark test functions show that the consideration of dimensions using SFDM for EAs improves the outcomes of all nine tested EAs. In our proposed method, we find better compatibility with the integration of SFDM in the Particle Swarm Optimization Algorithm concerning searching for the optimum solution relative to the other EAs.
Previous investigations indicated that a flat-walled, multi-layered anechoic lining system, with an overall thickness slightly less than a quarter of a wavelength, could be used to achieve a required cut-off frequency...
详细信息
Previous investigations indicated that a flat-walled, multi-layered anechoic lining system, with an overall thickness slightly less than a quarter of a wavelength, could be used to achieve a required cut-off frequency. However, the work proved to be tedious and time consuming because of the numerous trial-and-error measurements involved. On the other hand, the successful application of a method of calculating the overall acoustic impedance of multi-layered absorbing systems has indicated that the design of multi-layered absorbing systems can be carried out on a desktop computer. In the present work, a MATLAB genetic and evolutionary algorithm toolbox is implemented as the optimiser to aid and speed up the design process. The optimisation results indicate that a three-layered lining system can achieve results comparable with quality wedge-type anechoic linings with overall thickness slightly less than a sixth of a wavelength at the 100 Hz cut-off frequency. (C) 2004 Elsevier Ltd. All rights reserved.
evolutionary algorithms (EAs) are general-purpose optimisers that come with several parameters like the sizes of parent and offspring populations or the mutation rate. It is well known that the performance of EAs may ...
详细信息
evolutionary algorithms (EAs) are general-purpose optimisers that come with several parameters like the sizes of parent and offspring populations or the mutation rate. It is well known that the performance of EAs may depend drastically on these parameters. Recent theoretical studies have shown that self-adjusting parameter control mechanisms that tune parameters during the algorithm run can provably outperform the best static parameters in EAs on discrete problems. However, the majority of these studies concerned elitist EAs and we do not have a clear answer on whether the same mechanisms can be applied for non-elitist EAs. We study one of the best-known parameter control mechanisms, the one-fifth success rule, to control the offspring population size lambda in the non-elitist (1,lambda) EA. It is known that the (1,lambda) EA has a sharp threshold with respect to the choice of lambda where the expected runtime on the benchmark function OneMax changes from polynomial to exponential time. Hence, it is not clear whether parameter control mechanisms are able to find and maintain suitable values of lambda. For OneMax we show that the answer crucially depends on the success rate s (i. e. a one-(s + 1)-th success rule). We prove that, if the success rate is appropriately small, the self-adjusting (1, lambda) EA optimises OneMax in O(n) expected generations and O(n log n) expected evaluations, the best possible runtime for any unary unbiased black-box algorithm. A small success rate is crucial: we also show that if the success rate is too large, the algorithm has an exponential runtime on OneMax and other functions with similar characteristics.
The past five years have seen rapid development of plans and test pilots aimed at introducing connected and autonomous vehicles (CAVs) in public transport systems around the world. While self-driving technology is sti...
详细信息
The past five years have seen rapid development of plans and test pilots aimed at introducing connected and autonomous vehicles (CAVs) in public transport systems around the world. While self-driving technology is still being perfected, public transport authorities are increasingly interested in the ability to model and optimize the benefits of adding CAVs to existing multi-modal transport systems. Using a real-world scenario from the Leeds Metropolitan Area as a case study, we demonstrate an effective way of combining macro-level mobility simulations based on open data with global optimisation techniques to discover realistic optimal deployment strategies for CAVs. The macro-level mobility simulations are used to assess the quality of a potential multi-route CAV service by quantifying geographic accessibility improvements using an extended version of Dijkstra's algorithm on an abstract multi-modal transport network. The optimisations were carried out using several popular population-based optimisation algorithms that were combined with several routing strategies aimed at constructing the best routes by ordering stops in a realistic sequence.
In this paper, we consider system-level synthesis;Is the problem of optimally mapping a task-level specification onto a heterogeneous hardware/software architecture. This problem requires (1) the selection of the arch...
详细信息
In this paper, we consider system-level synthesis;Is the problem of optimally mapping a task-level specification onto a heterogeneous hardware/software architecture. This problem requires (1) the selection of the architecture (allocation) including general purpose and dedicated processors, ASICs, busses and memories, (2) the mapping of the specification onto the selected architecture in space (binding) and time (scheduling), and (3) the design space exploration with the goal to find a set of implementations that satisfy a number of constraints on cost and performance. Existing methodologies often consider a fixed architecture, perform the binding only, do not reflect the tight interdependency between binding and scheduling, do not consider communication (tasks and resources), or require long run-times preventing design space exploration, or yield only one implementation with optimal cost. Here, a model is introduced that handles all mentioned requirements and allows the task of system-synthesis to be specified as an optimization problem. The application and adaptation of an-evolutionary Algorithm to solve the tasks of optimization and design space exploration is described.
evolutionary algorithms (EAs) are general-purpose optimization algorithms, inspired by natural evolution. Recent theoretical studies have shown that EAs can achieve good approximation guarantees for solving the proble...
详细信息
evolutionary algorithms (EAs) are general-purpose optimization algorithms, inspired by natural evolution. Recent theoretical studies have shown that EAs can achieve good approximation guarantees for solving the problem classes of submodular optimization, which have a wide range of applications, such as maximum coverage, sparse regression, influence maximization, document summarization and sensor placement, just to name a few. Though they have provided some theoretical explanation for the general-purpose nature of EAs, the considered submodular objective functions are defined only over sets or multisets. To complement this line of research, this paper studies the problem class of maximizing monotone submodular functions over sequences, where the objective function depends on the order of items. We prove that for each kind of previously studied monotone submodular objective functions over sequences, i.e., prefix monotone submodular functions, weakly monotone and strongly submodular functions, and DAG monotone submodular functions, a simple multi-objective EA, i.e., GSEMO, can always reach or improve the best known approximation guarantee after running polynomial time in expectation. Note that these best-known approximation guarantees can be obtained only by different greedy-style algorithms before. Empirical studies on various applications, e.g., accomplishing tasks, maximizing information gain, search-and-tracking and recommender systems, show the excellent performance of the GSEMO.(c) 2022 Elsevier B.V. All rights reserved.
Memetic algorithms integrate local search into an evolutionary algorithm to combine the advantages of rapid exploitation and global optimisation. We provide a rigorous runtime analysis of memetic algorithms on the Hur...
详细信息
Memetic algorithms integrate local search into an evolutionary algorithm to combine the advantages of rapid exploitation and global optimisation. We provide a rigorous runtime analysis of memetic algorithms on the Hurdle problem, a landscape class of tunable difficulty with a "big valley structure", a characteristic feature of many hard combinatorial optimisation problems. A parameter called hurdle width describes the length of fitness valleys that need to be overcome. We show that the expected runtime of plain evolutionary algorithms like the (1+1) EA increases steeply with the hurdle width, yielding superpolynomial times to find the optimum, whereas a simple memetic algorithm, (1+1) MA, only needs polynomial expected time. Surprisingly, while increasing the hurdle width makes the problem harder for evolutionary algorithms, it becomes easier for memetic algorithms. We further give the first rigorous proof that crossover can decrease the expected runtime in memetic algorithms. A (2+1) MA using mutation, crossover and local search outperforms any other combination of these operators. Our results demonstrate the power of memetic algorithms for problems with big valley structures and the benefits of hybridising multiple search operators. (C) 2020 Elsevier B.V. All rights reserved.
暂无评论