Abstract: The welding process which is one of the most important assembly processes is widespread in the modern manufacturing industry, including aerospace, automotive and engineering machinery. The welding shop scheduling greatly impacts the efficiency of whole production system. However, few studies on the welding shop scheduling problem (WSSP) were reported. In this paper, a mathematical model and an improved discrete artificial bee colony algorithm (DABC) are proposed for the WSSP. Firstly, it is defined where multi-machine can process one job at the same time in the WSSP. Secondly, the mathematical models of WSSP have been constructed. Thirdly, an effective DABC is proposed to solve the WSSP, considering job permutation and machine allocation simultaneously. To improve the performance of proposed DABC algorithm, the effective operators have been designed. Three instances with different scales are used to evaluate the effectiveness of proposed algorithm. The comparisons with other two algorithms including genetic algorithm and grey wolf optimizer are also provided. Experimental results show that the proposed model and algorithm achieve good performance. Finally, the proposed model and DABC algorithm are applied in a real-world girder welding shop from a crane company in China. The results show that proposed model and algorithm reduces 55.17% production time comparing with the traditional algorithm and the scheduled machine allocation provides more reasonable arrangements for workers and machine loads. PubDate: 2019-03-18

Abstract: The problem of integrated project portfolio selection and scheduling (PPSS) is among the most important and highly pursed subjects in project management. In this study, a mathematical model and algorithm are designed specifically to assist decision makers decide which projects are to be chosen and when these projects are to be undertaken. More specifically, the PPSS problem is first formulated as a nonlinear multi-objective model with simultaneous consideration of benefit and risk factors. Due to the complexity and uncertainty involved in most real life situations, fuzzy numbers are incorporated into the model, which can provide decision makers with more flexibility. Then, an inverse modeling based multi-objective evolutionary algorithm using a Gaussian Process is presented to obtain the Pareto set. Finally, an illustrative example is used to demonstrate the high efficacy of the foregoing approach, which can provide decision makers with valuable insights into the PPSS process. The proposed algorithm is found to be more effective compared with two other popular algorithms. PubDate: 2019-03-07

Abstract: In the last decades, a plethora of dedicated heuristic and meta-heuristic algorithms have been crafted to solve complex optimization problems. However, it is noted that the majority of these algorithms are restricted to instances of small to medium size only. In today’s world, with the rapid growth in communication and computation technologies, massive volumes of data are generated and stored daily, making it vital to explore learning and optimization techniques that can handle ‘big’ problems. In this paper, we take an important step in the aforementioned direction by proposing a novel, theoretically motivated compressed representation with high-level meme evolution for big optimization. In contrast to existing heuristics and meta-heuristics, which work directly on the solution space, the proposed meme evolution operates on a high-level meme space. In particular, taking knapsack problem as the case study, a meme, in the present case, represents a knowledge-block as an instruction for solving the knapsack problem. Since the size of the meme, as defined in this paper, is not strongly sensitive to the number of items in the underlying knapsack problem, the search in meme space provides a compressed form of optimization. In order to verify the effectiveness of the proposed approach we carry out a variety of numerical experiments with problem sizes ranging from the small (100 items) to the very large (10,000 items). The results provide strong encouragement for further exploration, in order to establish meme evolution as the gold standard in big optimization. PubDate: 2019-03-01

Abstract: The incremental extreme learning machine (I-ELM) was proposed in 2006 as a method to improve the network architecture of extreme learning machines (ELMs). To improve on the I-ELM, bidirectional extreme learning machines (B-ELMs) were developed in 2012. The B-ELM uses the same method as the I-ELM but separates the odd and even learning steps. At the odd learning step, a hidden node is added like I-ELM. At the even learning step, a new hidden node is added via a formula based on the former added node result. However, some of the hidden nodes generated by the I-ELM may play a minor role; thus, the increase in network complexity due to the B-ELM may be unnecessary. To avoid this issue, this paper proposes an enhanced B-ELM method (referred to as EB-ELM). Several hidden nodes are randomly generated at each odd learning step, however, only the nodes with the largest residual error reduction will be added to the existing network. Simulation results show that the EB-ELM can obtain higher accuracy and achieve better performance than the B-ELM under the same network architecture. In addition, the EB-ELM can achieve a faster convergence rate than the B-ELM, which means that the EB-ELM has smaller network complexity and faster learning speed than the B-ELM. PubDate: 2019-03-01

Abstract: With the development of digital multimedia technologies, image matting has become one of the most popular research problem in academic field and been widely applied in industrial communities. The key challenge of image matting is how to extract the foreground region (target region) from a given image accurately. Sampling-based image matting technology implements matting by sampling some foreground pixels and background pixels from known regions and finding the best foreground–background sample pair for every undetermined pixel. The best foreground–background sample pair represents the true foreground and background colors of the corresponding undetermined pixel and they can estimate the region of this undetermined pixel accurately. Therefore, the quality of matting depends on whether the best sample pair can be found. This search process can be regarded as a combinational optimization problem. In this paper, in order to obtain more accurate matting result, we applied a bio-inspired metaheuristic algorithm to solve this problem, which is based on the promising earthworm optimization algorithm (EWA). By analyzing the property of this optimization problem, we upgrade two reproductions and the cauchy mutation of EWA to discrete calculations. The proposed algorithm is called as the discrete earthworm optimization algorithm (D-EWA). By comparing with existing optimization algorithms on a standard benchmark dataset, the experimental results show that the proposed D-EWA can obtain more accurate matting results on both visual effect and quantitative metric. PubDate: 2019-03-01

Abstract: This paper proposes an improved weighted extreme learning machine (IW-ELM) for imbalanced data classification. By incorporating voting method into weighted extreme learning machine (weighted ELM), three major steps are involved in the proposed method: training weighted ELM classifiers, eliminating unusable classifies to determine proper classifiers for voting, and finally determining the classification result based on majority voting. Simulations on many real world imbalanced datasets with various imbalance ratios have demonstrated that the proposed algorithm outperforms weighted ELM and other related classification algorithms. PubDate: 2019-03-01

Abstract: Brain storm optimization (BSO) is a novel evolutionary algorithm which originates from the human brainstorming process. The successful applications of BSO on various problems demonstrate its validity and efficiency. To theoretically analyze the performance of algorithm from the viewpoint of population evolution, the population interaction network (PIN) is used to construct the relationship among individuals in BSO. Four experiments in different dimensions, parameters, combinatorial parameter settings and related algorithms are implemented, respectively. The experimental results indicate the frequency of average degree of BSO meets a power law distribution in the functions with low dimension, which shows the best performance of algorithm among three kinds of dimensions. The parameters of BSO are investigated to find the influence of the population interaction with the power law distribution on the performance of algorithm, and respective parameter can change the relationship among individuals. In addition, the mutual effect among parameters is analyzed to find the best combinatorial result to significantly enhance the performance of BSO. The contrast among BSO, DE and PSO demonstrates a power law distribution is more effective for boosting the population interaction to enhance the performance of algorithm. PubDate: 2019-03-01

Abstract: The aim of this article is to study the two-objective coverage problem of wireless sensor networks (WSNs) by means of differential evolution algorithm. Firstly, in order to reduce the computing redundancy of multi-objective optimization, namely to reduce the number of individuals which participate in non-dominated solution sorting, we introduced a fast two-objective differential evolution algorithm (FTODE). The FTODE contains a fast non-dominated solution sorted and a uniform crowding distance calculation method. The fast sorting method just handles the highest rank individuals but not all individuals in the current population. Meanwhile, during the individuals sorted, it can choose some of individuals into next generation and reduce the time complexity. The uniform crowding distance calculation can enhance the diversity of population due to it will retain the outline of optimal solution set by choosing the individual uniformly. Secondly, we use the FTODE framework to research the two-objective coverage problem of WSNs. The two objectives are formulated as: the minimum number of sensor used and the maximum coverage rate. For this specific problem, decimal integer encoding are used and a recombination operation is introduced into FTODE, which performs after initialization and guarantees at least one critical target’s sensor is divided into different disjoint sets. Finally, the simulation experiment shows that the FTODE provides competitive results in terms of time complexity and performance, and it also obtains better solutions than comparison algorithms on the two-objective coverage problem of WSNs. PubDate: 2019-03-01

Abstract: The area under receiver operating characteristic curve (AUC) is one of the widely used metrics for measuring imbalanced data classification results. Designing multi-objective evolutionary algorithms for AUC maximization problem has attracted much attention of researchers recently. However, most of these methods either search the Pareto front directly, or perform tailored convex hull search for AUC maximization. None of them take the advantage of multi-level knee points found in the process of evolution for AUC maximization. To this end, this paper proposes a multi-level knee point based multi-objective evolutionary algorithm (named MKnEA-AUC) for AUC maximization on the basis of a recently developed knee point driven evolutionary algorithm for multi/many-objective optimization. In MKnEA-AUC, an adaptive clustering strategy is proposed for automatically determining the knee points on the current population. By utilizing the preference of found knee points, the evolution of the population can converge quickly. We verify the effectiveness of the proposed algorithm MKnEA-AUC on 13 widely used benchmark data sets and the experimental results demonstrate that MKnEA-AUC is superior over the state-of-the-art algorithms for AUC maximization. PubDate: 2019-02-09

Abstract: A new differential evolution algorithm with two dynamic speciation-based mutation strategies (DSM-DE) is proposed to solve single-objective optimization problems. An explorative mutation “DE/seeds-to-seeds” and an exploitative mutation “DE/seeds-to-rand” are employed simultaneously in DSM-DE in the evolutionary process. A Dynamic Speciation Technique is designed to assist the two mutations in order to utilize the potential of selective portioning of critical individuals in the population. It dynamically divides the population into numbers of species whilst taking species seeds as centers. The best individuals for each species are used as base vectors in each species in the proposed mutation strategies. “DE/seeds-to-seeds” selects individuals from species seeds and current species to constitute difference vectors whereas “DE/seeds-to-rand” selects from the whole population. Thus the two mutation strategies can accelerate the convergence process without decreasing diversity of the population. Comparison results with four classic DE variants, one state-of-art DE variant and two improved non-DE variants on CEC2014, CEC2015 benchmark, and Lennard-Jones potential problem reveal that the overall performance of DSM-DE is better than that of the other seven DE algorithms. In addition, experiments also substantiate the effectiveness and superiority of two seeds-guided mutation strategies in DSM-DE. PubDate: 2019-01-11

Abstract: The single-factory manufacturing is gradually transiting to the multi-factory collaborative production with the globalization. The decentralization of resources and the heterogeneity of the production modes make it difficult to solve this kind of problem. Therefore, the distributed assembly flexible job shop scheduling problem (DAFJSP) is studied. DAFJSP can be decomposed into several flexible job shop scheduling problems and several single machine factory scheduling problems. To begin with, a mixed integer linear programming model for the DAFJSP is formulated to minimize the earliness/tardiness and the total cost simultaneously. Then, an improved differential evolution simulated annealing algorithm (IDESAA) is proposed. The balanced scheduling algorithm is designed to trade off the two objectives. Two crossover and mutation operators are designed. Due to its strong robustness, simulated annealing is integrated to local search the best Pareto solutions. The greedy idea combined with the Non-Dominated Sorted selection is employed to select the offspring. Finally, comprehensive experiments are conducted and the results show that the proposed algorithm can solve DAFJSP effectively and efficiently. PubDate: 2018-12-20

Abstract: With the increasing installed power of the wind turbines, the necessity of condition monitoring for wind turbine drivetrain cannot be neglected any longer. A reliable and rapid response fault diagnosis is vital for the wind turbine drivetrain system. The existing manual inspection-based methods are difficult to accomplish the real-time compound-fault monitoring task. To solve this problem, this paper proposes a novel dual extreme learning machines (Dual-ELMs) based fault diagnostic framework for feature extraction and fault pattern recognition. At the stage of feature learning, this paper applies the local mean decomposition (LMD) method to extract the production functions from the raw vibration signals. Compared with the traditional empirical mode decomposition method, the LMD method has a stronger ability to restrain the mode mixing and endpoints effect. At the stage of compound-fault classification, unlike the other widely-used classifiers, the proposed Dual-ELM networks inherit the advantages of the original extreme learning machines (ELMs), that employs two basic ELM networks for the compound-fault classification, and it does not need iterative fine-tuning of parameters. Thus the learning speed is faster than the other combinations of classifiers. The experimental validity of the proposed algorithm was conducted on a test rig for vibration analysis, which demonstrated that the proposed Dual-ELMs based fault diagnostic method provides an effective measure for the observed machinery than the other available fault diagnostic methods in aspects of feature extraction and compound-fault recognition. PubDate: 2018-12-12

Abstract: Brain storm optimization (BSO) is a newly proposed optimization algorithm inspired by human being brainstorming process. After its appearance, much attention has been paid on and many attempts to improve its performance have been made. The search ability of BSO has been enhanced, but it still suffers from sticking into stagnation during exploitation phase. This paper proposes a novel method which incorporates BSO with chaotic local search (CLS) with the purpose of alleviating this situation. Chaos has properties of randomicity and ergodicity. These properties ensure CLS can explore every state of the search space if the search time duration is long enough. The incorporation of CLS can make BSO break the stagnation and keep the population’s diversity simultaneously, thus realizing a better balance between exploration and exploitation. Twelve chaotic maps are randomly selected for increasing the diversity of the search mechanism. Experimental and statistical results based on 25 benchmark functions demonstrate the superiority of the proposed method. PubDate: 2018-12-01

Abstract: The cumulative capacitated vehicle routing problem (CCVRP) is a combinatorial optimization problem which aims to minimize the sum of arrival times at customers. This paper presents a brain storm optimization algorithm to solve the CCVRP. Based on the characteristics of the CCVRP, we design new convergent and divergent operations. The convergent operation picks up and perturbs the best-so-far solution. It decomposes the resulting solution into a set of independent partial solutions and then determines a set of subproblems which are smaller CCVRPs. Instead of directly generating solutions for the original problem, the divergent operation selects one of three operators to generate new solutions for subproblems and then assembles a solution to the original problem by using those new solutions to the subproblems. The proposed algorithm was tested on benchmark instances, some of which have more than 560 nodes. The results show that our algorithm is very effective in contrast to the existing algorithms. Most notably, the proposed algorithm can find new best solutions for 8 medium instances and 7 large instances within short time. PubDate: 2018-12-01

Authors:Jingyu Wang; Junfeng Chen; Xingsi Xue; Jingjing Du Abstract: The canonical brain storm optimization (BSO) employs clustering, creating and selecting operators, which are all connected and have great impacts on the optimization performance. In this paper, the state-of-the-art search strategies are introduced and the potential strengths and weaknesses of these strategies in the BSO algorithm are analyzed and compared from the perspective of the analytical model rather than any metaphors. The numerical experiments are carried out to artificially amplify and highlight the performance of various strategies. Finally, the progressive directions of the BSO algorithm are discussed for further research. PubDate: 2018-06-01 DOI: 10.1007/s12293-018-0261-x

Authors:Satyabrata Dash; Deepak Joshi; Gaurav Trivedi Abstract: This paper presents a multiobjective analog/RF circuit sizing tool using an improved brain storm optimization (IMBSO) algorithm with the purpose of analyzing the tradeoffs between competing performance specifications of analog/RF circuit block. A number of improvements are incorporated into IMBSO algorithm at different steps. At first, the clustering step of IMBSO algorithm is augmented with k-means \(++\) seeding technique to select the initial cluster centroids while clustering using k-means clustering technique. As a second improvement, the proposed IMBSO algorithm makes use of random probabilistic decision-making of river formation dynamics scheme to select optimal cluster centroids during population generation step. As a third improvement, an adaptive mutation operator is incorporated inside the IMBSO algorithm to generate new population. Finally, two separate constraint handling techniques are employed to handle both boundary and functional constraints during analog/RF circuit optimization. The performance of the proposed IMBSO algorithm is demonstrated in finding optimal Pareto fronts among different performance specifications of a two-stage operational amplifier circuit, a folded cascode amplifier circuit and a low noise amplifier circuit. PubDate: 2018-05-31 DOI: 10.1007/s12293-018-0262-9

Authors:Michał Bereta Abstract: The study is concerned with the Baldwin effect and Lamarckian evolution in a memetic algorithm for Euclidean Steiner tree problem (ESTP). The main aim is to examine the importance of the proposed local search procedures for the ability of the algorithm to find solutions to ESTP. Two local search procedures are proposed as a part of an evolutionary algorithm for ESTP, and the effect of their use is carefully analyzed. The correction procedure modifies the given chromosome so that it satisfies the degree and angle conditions in ESTP. The improvement procedure actively adds new Steiner points to the evaluated solution. The modified solutions are accepted with varying probability, leading to a spectrum of algorithms with a Baldwin effect on one end and Lamarckian evolution on the other. The results are carefully analyzed using proper statistical procedures (Friedman test and post-hoc procedures). To further check the ability of the proposed algorithm to find the optimal or near optimal solutions, results for problems from OR-Lib are provided. PubDate: 2018-03-28 DOI: 10.1007/s12293-018-0256-7

Authors:Haibin Duan; Daifeng Zhang; Yuhui Shi; Yimin Deng Abstract: Close formation flight of swarm unmanned aerial vehicles (UAVs) has drawn much attention from scholars due to its significant importance in many aspects. In this paper, we focus on an advanced controller design for swarm UAV close formation based on a novel bio-inspired algorithm, i.e., metric-distance brain storm optimization (MDBSO). The proposed method utilizes the brain storm optimization (BSO) which has been extensively adopted in complicated systems with great performances and modifies its basic operators to formulate the formation flight controller design. The original clustering operator in BSO is replaced by a fresh clustering method based on metric distances, while the individual updating operator utilizes Lévy distribution to extend search steps to fit into the metric searching regions. Then the proposed algorithm is applied to optimize the benchmark controller in swarm UAV close formation to enhance the tracking performances under complicated circumstances. Simulation results demonstrate that our approach is more superior in stable configuration of swarm UAV close formations by comparing with several generic methods. PubDate: 2018-03-06 DOI: 10.1007/s12293-018-0251-z