GPU implementation of discrete particle swarm optimization algorithm for endmember extraction from hyperspectral image

Endmember extraction (EE) is one of the most important issues in hyperspectral mixture analysis, and it is also one of the most challenging tasks due to the intrinsic complexity of remote sensing images and the lack of priori knowledge. In recent years, a number of EE methods have been developed, and several different optimization objectives have been proposed from different perspectives. In all of these methods, there is only one objective function to be optimized, which represents a specific characteristic of the remote sensing image. However, one single-objective function may not provide satisfactory results because of the complexity of remote sensing images. In this paper, a multiobjective discrete particle swarm optimization algorithm (MODPSO) is utilized to tackle the problem of EE, where two objective functions, namely, volume maximization and root-mean-square error (RMSE) minimization are simultaneously optimized. The result set of Pareto-optimal solutions contains a number of non-dominated solutions, from which the user can judge relatively and pick up the most promising one according to the problem requirements. Experiments on two real hyperspectral images were conducted to evaluate the proposed MODPSO algorithm, which confirmed the effectiveness of the proposed algorithm.

The booming cloud computing, together with edge computing, lays the computing resource foundation for larger and more granular simulation. To achieve better execution of the simulation in cloud and edge environments, it is necessary to assign simulation components to cloud computing centers and edge computing nodes appropriately. This is an optimization problem and a particle swarm optimization(PSO) algorithm is proposed for this purpose. However, deployment problems are a kind of discrete optimization problem and PSO algorithm was originally designed to solve continuous optimization problems, thus, various discretization methods had been proposed to modify original PSO algorithm to handle discrete problems. Unsatisfactorily, many discrete PSO algorithms suffer from random movements of particles due to improper use of the guidance information obtained from the global best particle and personal best particle. To counter this challenge, we proposed a selection-based discrete PSO algorithm(SPSO) and conducted simulation experiments on MATLAB platform. The results showed that SPSO got better performance in simulation components allocation problems compared with NSGA-II and other pure discrete PSO algorithms.

Endmember extraction is a key step in hyperspectral unmixing. This paper proposes a new endmember extraction framework based on the swarm intelligence algorithm. We adopt a discrete structure because pixels exist within a discrete frame. Traditional swarm intelligence algorithms produce stacked solutions based on similar endmembers in the same class. We introduce a “distance” factor into the objective function to limit the number of endmembers per class. We then propose three endmember extraction methods based on the artificial bee colony (ABC), ant colony optimization (ACO) and particle swarm optimization (PSO) algorithms. Experiments with both simulated and actual hyperspectral image data indicate that the proposed framework can significantly improve the accuracy of endmember extraction.

With the rapid development of renewable energy, the proportion of small hydropower as a clean energy in the distribution network (DN) is increasing. However, the randomness and intermittence of small hydropower has brought new challenges to the operation of DN; especially, the problems of increasing network loss and reactive voltage exceeding the limit have become increasingly prominent. Aiming at the above problems, this paper proposes a reactive power optimization control method for DN with hydropower based on an improved discrete particle swarm optimization (PSO) algorithm. Firstly, this paper analyzes the specific characteristics of small hydropower and establishes its mathematical model. Secondly, considering the constraints of bus voltage and generator RP output, an extended minimum objective function for system power loss is established, with bus voltage violation serving as the penalty function. Then, in order to solve the following problems: that the traditional discrete PSO algorithm is easy to fall into local optimization and slow convergence, this paper proposes an improved discrete PSO algorithm, which improves the global search ability and convergence speed by introducing adaptive inertia weight. Finally, based on the IEEE-33 buses distribution system as an example, the simulation analysis shows that compared with GA optimization, the line loss can be reduced by 3.4% in the wet season and 13.6% in the dry season. Therefore, the proposed method can effectively reduce the network loss and improve the voltage quality, which verifies the effectiveness and superiority of the proposed method.

A multi-objective discrete particle swarm optimization (MODPSO) algorithm is useful in accurately identifying communities in a network by avoiding the pitfalls of modularity optimized discrete PSO algorithms. Inertia weights in a PSO can be used to guide the flight of particles in PSO by modifying step size of the particles. In this paper, we present a new adaptive inertia weight based MODPSO and compare it with other good inertia weight approaches by applying them to three real-world datasets. Our algorithm demonstrates consistently the best results among various inertia weight strategies in three real-world datasets with maximum Q (modularity score) values of 0.457, 0.527728 and 0.60457 for Zachary’s Karate Club, Bottlenose Dolphins and American College Football datasets, respectively. Adaptive inertia weight strategy is able to perform consistently by adaptively determining the step size of the velocity update equation. To the best of our knowledge, this is the first such attempt to explore the adaptive inertia weight technique with MODPSO in the field of community detection in complex networks.

A novel discrete particle swarm optimization algorithm for meta-task assignment in heterogeneous computing systems

A kind of discrete Particle Swarm Optimization (PSO) algorithm named NDPSO was proposed for extending the classic PSO model to solve the discrete optimization problems with high effectiveness and stability. The concept of comparison with probability was also introduced based on NDPSO, and then the stochastic repairing operator was constructed for heuristic search to solve the Winner Determination Problem (WDP) in combinatory auction. The experimental results show that NDPSO has great advantages in both success rate and convergence speed compared with other discrete PSO algorithms and genetic algorithm.

In assembly optimisation, assembly sequence planning and assembly line balancing have been extensively studied because both activities are directly linked with assembly efficiency that influences the final assembly costs. Both activities are categorised as NP-hard and usually performed separately. Assembly sequence planning and assembly line balancing optimisation presents a good opportunity to be integrated, considering the benefits such as larger search space that leads to better solution quality, reduces error rate in planning and speeds up time-to-market for a product. In order to optimise an integrated assembly sequence planning and assembly line balancing, this work proposes a multi-objective discrete particle swarm optimisation algorithm that used discrete procedures to update its position and velocity in finding Pareto optimal solution. A computational experiment with 51 test problems at different difficulty levels was used to test the multi-objective discrete particle swarm optimisation performance compared with the existing algorithms. A statistical test of the algorithm performance indicates that the proposed multi-objective discrete particle swarm optimisation algorithm presents significant improvement in terms of the quality of the solution set towards the Pareto optimal set.

A high call blocking rate is a consequence of an inefficient utilization of system resources, which is often caused by a load imbalance in the network. Load imbalances are common in wireless networks with a large number of cellular users. This paper investigates a load-balancing scheme for mobile networks that optimizes cellular performance with constraints of physical resource limits and users quality of service demands. In order to efficiently utilize the system resources, an intelligent distributed antenna system (IDAS) fed by a multi base transceiver station (BTS) has the ability to distribute the system resources over a given geographic area. To enable load balancing among distributed antenna modules we dynamically allocate the remote antenna modules to the BTSs using an intelligent algorithm. A self-optimizing network for an IDAS is formulated as an integer based linear constrained optimization problem, which tries to balance the load among the BTSs. A discrete particle swarm optimization (DPSO) algorithm as an evolutionary algorithm is proposed to solve the optimization problem. The computational results of the DPSO algorithm demonstrate optimum performance for small-scale networks and near-optimum performance for large-scale networks. The DPSO algorithm is faster with marginally less complexity than an exhaustive search algorithm.

Spectral unmixing is an important problem for remotely sensed hyperspectral data exploitation. Automatic spectral unmixing can be viewed as a three-stage problem, where the first stage is subspace identification, the next one is endmember extraction, and the final one is abundance estimation. In this sequence, endmember extraction is the most challenging problem. Many researchers have attempted to extract endmembers from hyperspectral images using spectral information only. However, it is well known that the inclusion of spatial information can improve the endmember extraction task. In this article, we introduce a new endmember extraction algorithm that exploits both spectral and spatial information. A main innovation of the proposed algorithm is that spatial information is exploited using entropy, while spectral information is exploited using convex set optimization. In the literature, none of the spatial–spectral algorithms has used entropy as spatial information. The inclusion of this entropy-based spatial information improves the accuracy of the endmember extraction process. The results obtained by the proposed algorithm are compared (using a variety of metrics) with those obtained by other state-of-the-art methods, using both synthetic and real datasets. Our experimental results demonstrate that the proposed algorithm outperforms many available algorithms.

The spatial resolution of hyperspectral data is low, so there are a large number of mixed pixels, which is also one of the main reasons that reduce the accuracy of hyperspectral image target classification. Hyperspectral unmixing is an important subject in the field of remote sensing. Hyperspectral unmixing generally consists of three steps: reduction, endmember extraction and inversion. As one of the key steps of hyperspectral unmixing, efficient and rapid endmember extraction is an important object in hyperspectral remote sensing. In this paper, the endmember extraction of hyperspectral data is implemented based on PCA and a new SGA algorithm, which solves the dimension limitation of traditional SGA algorithm and the new SGA algorithm without data redundancy caused by data dimensionality reduction.

The signed network has received widespread attention because it can well reflect the cooperation and conflict relationship. Structural balance is an important global feature in signed networks, which can well reflect the structural characteristics of the network. Existing structural balance calculation algorithms define the global and local balance computation problems as an optimization problem, and then optimize their respective objective functions through optimization algorithms, but these algorithms ignore the correlation between the two problems. In this paper, we combine the multifactorial evolutionary algorithm and the discrete particle swarm optimization algorithm, and further propose the multifactorial discrete particle swarm optimization algorithm (MFDPSO). This algorithm designs the knowledge transfer function and optimization algorithm based on the correlation of the strong and weak structure balance and optimizes the two problems at the same time. The experimental results on 8 real networks demonstrate the effectiveness of the MFDPSO.

The current methods that use hyperspectral remote sensing imagery to extract and monitor marine oil spills are quite popular. However, the automatic extraction of endmembers from hyperspectral imagery remains a challenge. This paper proposes a data field-spectral preprocessing (DSPP) algorithm for endmember extraction. The method first derives a set of extreme points from the data field of an image. At the same time, it identifies a set of spectrally pure points in the spectral space. Finally, the preprocessing algorithm fuses the data field with the spectral calculation to generate a new subset of endmember candidates for the following endmember extraction. The processing time is greatly shortened by directly using endmember extraction algorithms. The proposed algorithm provides accurate endmember detection, including the detection of anomalous endmembers. Therefore, it has a greater accuracy, stronger noise resistance, and is less time-consuming. Using both synthetic hyperspectral images and real airborne hyperspectral images, we utilized the proposed preprocessing algorithm in combination with several endmember extraction algorithms to compare the proposed algorithm with the existing endmember extraction preprocessing algorithms. The experimental results show that the proposed method can effectively extract marine oil spill data.

Process mining, which aims to mine a high-quality process model from event log, provides a powerful tool to support the design, enactment, management, and analysis of operational business processes. However, the task is not easy because the algorithm needs to discover various complex process structures, handle noisy and incomplete event logs and balance multiple performance indicators. In this paper, a novel algorithm (called PSOMiner) for process mining is proposed, which consists of a discrete particle swarm optimization algorithm and guided local mutation. The former is in charge of searching the solution space of causal matrix and the latter is used to help the algorithm skip out the local optimum when it suffers from premature. A fine-grained scoring strategy which used to assign a score to each position of a particle (i.e. causal matrix) is presented to direct the mutation. The experiments were performed on 28 synthetic event logs with/without noise and 4 real-life event logs, and three classical algorithms of process mining (ETM, Hybrid ILP Miner, HM) were chosen for comparison. The results show that (1) PSOMiner achieved the best f-score on 25 synthetic event logs; (2) The average f-score of PSOMiner is 0.825 on 4 real-life event logs, which is superior to ETM whose average f-score is 0.703.

Parameter Selection of Discrete Particle Swarm Optimization Algorithm for the Quadratic Assignment Problems