Translate

Παρασκευή 30 Αυγούστου 2019

Special issue on Intelligence Computation Evolutionary Computation: ICEV2018

A gas source localization algorithm based on NLS initial optimization of particle filtering

Abstract

A gas source localization algorithm based on the initial NLS value optimization of particle filter (PF-NLS) is proposed to solve the problem that the positioning accuracy of NLS algorithm is reduced due to rough initial value estimation and improper weight allocation of nodes. Firstly, the state space model of the system is established by using the gas turbulence diffusion model, the particle weight is updated by constructing the likelihood function, and then the initial position and initial source strength of the gas source are obtained by using the NLS algorithm. Finally, the true information of the gas source is accurately estimated by using the PF-NLS algorithm. The simulation results show that compared with NLS, the algorithm has higher positioning accuracy, further improves the convergence speed of the algorithm by optimizing the initial value, and can be well applied to practical scenarios.

Imbalanced data classification algorithm with support vector machine kernel extensions

Abstract

Learning from the imbalanced data samples so as to achieve accurate classification is an important research content in data mining field. It is very difficult for classification algorithm to achieve a higher accuracy because the uneven distribution of data samples makes some categories have few samples. A imbalanced data classification algorithm of support vector machines (KE-SVM) is proposed in this article, this algorithm achieve the initial classification of data samples by training the maximum margin classification SVM model, and then obtaining a new kernel extension function. based on Chi square test and weight coefficient calculation, through training the samples again by the new vector machine with kernel function to improve the classification accuracy. Through the simulation experiments of real data sets of artificial data set, it shows that the proposed method has higher classification accuracy and faster convergence for the uneven distribution data.

Research on cloud computing in the resource sharing system of university library services

Abstract

As a combination of technology and application, cloud technology brings new opportunities for library data storage. Cloud computing in the resource sharing system of library service area has first proposed the cloud library platform to achieve resource sharing among digital libraries. First, this paper needs to build cloud library platform architecture, and then we introduce the cloud computing related technology, such as Hadoop. Finally, the system is applied it to this map. The building of the library constitutes a local library cloud platform uses Web technology to achieve the scheduling of local libraries, and then implements the most common resource retrieval service in the library on the cloud platform framework, and finally discusses issues that need attention in the follow-up development of the platform. Finally, a multi-agent sharing model for educational information resources under the cloud-computing environment has established to serve the construction of library information resources better. In addition, this article innovatively puts forward the function of library construction of employment information services for college students, which is better to serve college students.

A staged approach to evolving real-world UAV controllers

Abstract

A testbed has recently been introduced that evolves controllers for arbitrary hover-capable UAVs, with evaluations occurring directly on the robot. To prepare the testbed for real-world deployment, we investigate the effects of state-space limitations brought about by physical tethering (which prevents damage to the UAV during stochastic tuning), on the generality of the evolved controllers. We identify generalisation issues in some controllers, and propose an improved method that comprises two stages: in the first stage, controllers are evolved as normal using standard tethers, but experiments are terminated when the population displays basic flight competency. Optimisation then continues on a much less restrictive tether, effectively free-flying, and is allowed to explore a larger state-space envelope. We compare the two methods on a hover task using a real UAV, and show that more general solutions are generated in fewer generations using the two-stage approach. A secondary experiment undertakes a sensitivity analysis of the evolved controllers.

Oppositional elephant herding optimization with dynamic Cauchy mutation for multilevel image thresholding

Abstract

This paper presents an improved elephant herding optimization (IEHO) to solve the multilevel image thresholding problem for image segmentation by introducing oppositional-based learning (OBL) and dynamic cauchy mutation (DCM). OBL accelerates the convergence rate and enhances the performance of standard EHO whereas DCM mitigates the premature convergence. The suggested optimization approach maximizes two popular objective functions: ‘Kapur’s entropy’ and ‘between-class variance’ to estimate optimized threshold values for segmentation of the image. The performance of the proposed technique is verified on a set of test images taken from the benchmark Berkeley segmentation dataset. The results are analyzed and compared with conventional EHO and other four popular recent metaheuristic algorithms namely cuckoo search, artificial bee colony, bat algorithm, particle swarm optimization and one classical method named dynamic programming found from the literature. Experimental results show that the proposed IEHO provides promising performance compared to other methods in view of optimized fitness value, peak signal-to-noise ratio, structure similarity index and feature similarity index. The suggested algorithm also has better convergence than the other methods taken into consideration.

Removal of rain video based on temporal intensity and chromatic constraint of raindrops

Abstract

An improved algorithm of frame time difference is proposed and applied to raindrops removal of video image.This paper analyzes the temporal intensity waveform and chromatic constraint properties of raindrops, and the method is optimized by these two properties. We make use of the difference between rain and non-rain moving objects in the pixels’ intensity changes, which realized a broad classification between the rain and non-rain moving object pixel. The candidate raindrops pixels are optimized in combination with the chromatic constraint property. The experimental results show that the proposed algorithm has a better effect of rainy day in video image restoration than Garg’s, and it is simple and effective. The algorithm has a strong applicability, and it can be further used for many applications, such as air pollution control, management, outdoor surveillance, remote sensing and intelligent vehicles.

Automatic segmentation of sub-acute ischemic stroke lesion by using DTCWT and DBN with parameter fine tuning

Abstract

In image processing the ischemic stroke lesion segmentation is a major procedure used to extricate suspicious regions from the given MRI brain image. For classification and segmentation of MRI in this paper, we proposed a three-step framework. To remove noise the initial step utilizes a de-noising technique based on dual tree complex wavelet transform (DTCWT) test without affective the essential image features and content. In the second step, an un-supervised deep belief network (DBN) is intended for learning the unlabelled features. Here, the noise in MRI can cause a significant corruption of data that impedes the execution of DBNs. The DTCWT in the initial step enhances execution of DBNs. Additionally, we manage the issue of DBNs parameters fine-tuning by means of a quick meta-heuristic approach named salp swarm algorithm. Based on the simulation behaviour of salps this new meta-heuristic algorithm is planned to solve optimisation issues. It is validated against different benchmark test functions and afterward contrasted with well known state-of-the-art optimisation algorithms like genetic algorithm, particle swarm optimisation, bat algorithm, artificial bee colony algorithm and cuckoo search algorithm for performance efficiency.

Implementation of self adaptive mutation factor and cross-over probability based differential evolution algorithm for node localization in wireless sensor networks

Abstract

Node localization or positioning is essential for many position aware protocols in a wireless sensor network. The classical global poisoning system used for node localization is limited because of its high cost and its unavailability in the indoor environments. So, several localization algorithms have been proposed in the recent past to improve localization accuracy and to reduce implementation cost. One of the popular approaches of localization is to define localization as a least square localization (LSL) problem. During optimization of LSL problem, the performance of the classical Gauss–Newton method is limited because it can be trapped by local minima. By contrast, differential evolution (DE) algorithm has high localization accuracy because it has an ability to determine global optimal solution to the LSL problem. However, the convergence speed of the conventional DE algorithm is low as it uses fixed values of mutation factor and cross-over probability. Thus, in this paper, a self-adaptive mutation factor cross-over probability based differential evolution (SA-MCDE) algorithm is proposed for LSL problem to improve convergence speed. The SA-MCDE algorithm adaptively adjusts the mutation factor and cross-over probability in each generation to better explore and exploit the global optimal solution. Thus, improved localization accuracy with high convergence speed is expected from the SA-MCDE algorithm. The rigorous simulation results conducted for several localization algorithms declare that the propose SA-MCDE based localization has about (40–90) % more localization accuracy over the classical techniques.

A polychromatic sets theory based algorithm for the input/output scheduling problem in AS/RSs

Abstract

This paper proposes a hybrid polychromatic sets theory based genetic algorithm (GA) for the input/output scheduling problem of automated storage and retrieval systems (AS/RSs). In order to overcome the drawbacks of precocious phenomena and lack of stability of the GA, a novel hybrid model through integration of genetic algorithm (GA) and polychromatic sets theory (PS) is proposed. During the solution process, polychromatic sets contour matric is used to assign input/output goods location reasonably to improve the quality of initial population. During the iterative process, simulate anneal algorithm (SA) is invoked to jump out of the local optimum to obtain a satisfactory solution. The experiment results showed that the hybrid PS-SA-GA genetic algorithm has obvious advantages in solving the input/output scheduling problems: the introduction of contour matrix constraint module can effectively improve the quality of initial population; the simulated annealing algorithm is invoked to reduce the possibility of premature convergence, and the global optimal solution can be obtained efficiently.

Δεν υπάρχουν σχόλια:

Δημοσίευση σχολίου

Αρχειοθήκη ιστολογίου

Translate