Applied Soft Computing最新文献

{"title":"A survey on resilient microgrid system from cybersecurity perspective","authors":"Zhibo Zhang ,&nbsp;Benjamin Turnbull ,&nbsp;Shabnam Kasra Kermanshahi ,&nbsp;Hemanshu Pota ,&nbsp;Ernesto Damiani ,&nbsp;Chan Yeob Yeun ,&nbsp;Jiankun Hu","doi":"10.1016/j.asoc.2025.113088","DOIUrl":"10.1016/j.asoc.2025.113088","url":null,"abstract":"<div><div>Due to increases in communication speed, computation, the liberalization of the electrical service business, and the environmental impact of traditional power generation technologies, Distributed Energy Resources (DERs) power systems such as microgrids are gaining in popularity. It is therefore imperative to develop resilient microgrid systems capable of withstanding cyber physical threats. The capacity to integrate Machine Learning (ML) and Deep Learning (DL) to analyze energy data has created opportunities for businesses and academia to explore the possibilities of enhancing the cybersecurity of microgrid systems. This study surveys and discusses recent developments, challenges, and opportunities in cybersecurity for microgrid systems, from both attack and defense perspectives. In this paper, we address the current state and future directions in cybersecurity in industrial communication networks, and endpoint security in distributed control systems. This paper discusses attack types including Man-In-The-Middle (MITM), False Data Injection (FDI), and Distributed Denial of Service (DDoS) attacks, alongside defensive mechanisms including AI-based detection and multi-layered security frameworks. Furthermore, this survey offers comprehensive insights into benchmark datasets and open-source tools frequently utilized in experimental research and practical applications. It includes an in-depth comparison, discussion, and opportunities for future research to guide the research community’s focus and advancing progress in the field.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"175 ","pages":"Article 113088"},"PeriodicalIF":7.2,"publicationDate":"2025-04-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143816198","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Towards sustainable societies: Convolutional neural networks optimized by modified crayfish optimization algorithm aided by AdaBoost and XGBoost for waste classification tasks","authors":"Ana Tasic ,&nbsp;Luka Jovanovic ,&nbsp;Nebojsa Bacanin ,&nbsp;Miodrag Zivkovic ,&nbsp;Vladimir Simic ,&nbsp;Miroslav Popovic ,&nbsp;Milos Antonijevic","doi":"10.1016/j.asoc.2025.113086","DOIUrl":"10.1016/j.asoc.2025.113086","url":null,"abstract":"<div><div>The categorization of waste is playing a pivotal role in addressing and alleviating the environmental and health repercussions linked to waste. It is essential for safeguarding the environment, as improper handling of hazardous waste may result in soil, water, and air contamination, posing significant threats to ecosystems and human well-being and maintaining a sustainable society. Effective waste classification enhances the efficacy of waste management by organizing waste into distinctive groups based on characteristics that include toxicity, flammability, recyclable potential, and biodegradability. This research introduces a methodology that relies on employing convolutional neural networks and the AdaBoost and XGBoost models for the purpose of waste classification. It emphasizes the necessity of customizing every deep learning method to suit the specific problem that needs to be solved. An altered form of the latterly proposed crayfish optimization algorithm is suggested, explicitly developed to meet the requirements of the particular waste classification task in hand. The assessment of the presented method using real-world datasets consistently demonstrates that models configured by the proposed modified algorithm achieve an accuracy level that exceeds 89.6140%. This pinpoints the considerable potential of this method in effectively addressing pressing problems in waste management within real-world scenarios.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"175 ","pages":"Article 113086"},"PeriodicalIF":7.2,"publicationDate":"2025-04-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143807428","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Quaternary converter based balanced graph contrastive learning for recommendation","authors":"Fan Ye ,&nbsp;Hongwei Li ,&nbsp;Zhangling Duan ,&nbsp;Zhaolong Ling","doi":"10.1016/j.asoc.2025.113096","DOIUrl":"10.1016/j.asoc.2025.113096","url":null,"abstract":"<div><div>Graph contrastive learning (GCL) has emerged as a highly effective collaborative filtering method for recommender systems in recent years. However, existing collaborative filtering methods based on GCL often excessively prioritize user-side information, leading to inadequate exploration of user–item information. Furthermore, these methods generate contrastive views through data augmentation, which is prone to noise interference. To address these issues, we propose a balanced graph contrastive learning framework (BGCL). Specifically, BGCL incorporates a quaternary converter that introduces negative user based on triples (user, positive item, negative item), to provide the GCL module with embeddings that treat users and items balancedly. Subsequently, BGCL includes a noiseless GCL module that conducts contrastive learning on the embeddings after approximately infinite layers of convolution and the embeddings after k-layer graph convolutional networks to mitigate noise interference. We conducted experiments comparing our algorithm with 15 alternative approaches using real-world datasets, and the results demonstrate that our algorithm outperforms state-of-the-art methods in terms of recommendation accuracy and convergence speed.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"175 ","pages":"Article 113096"},"PeriodicalIF":7.2,"publicationDate":"2025-04-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143785592","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A radial visualization method based on knee point information for many-objective optimization","authors":"Gui Li ,&nbsp;Renbin Xiao","doi":"10.1016/j.asoc.2025.113064","DOIUrl":"10.1016/j.asoc.2025.113064","url":null,"abstract":"<div><div>Numerous high-dimensional solutions for many-objective optimization problems (MaOPs) usually impose a high cognitive burden on decision makers (DMs). Pareto front (PF) of MaOPs can express the problem characteristics, and then provide prior knowledge for solving the MaOPs. However, the existing high-dimensional visualization methods usually do not establish the relationship between PF information and decision making. Therefore, a novel radial visualization (RadViz) method called KRadViz that incorporates knee point information is proposed to visualize the information of PF shape and aid decision making. The relationship between the optimized performance information and PF shape is established, and the PF shape identification method is constructed. KRadViz is constructed by combining the optimization performance and PF shape. Three preferred solution selection methods are proposed to quickly screen out a few preferred solutions in different scenarios. The proposed KRadViz is compared with three high-dimensional visualization methods. The experimental results show that KRadViz can effectively display the high-dimensional PF shape, and give the optimization performance information of different solutions. The selection preferences of the three methods are also analyzed, and the effectiveness of the assisted decision process is verified. For the DTLZ2 and real-world MaOPs, the individual hypervolume (HV) contribution of preferred solutions increased by 9.98 % and 10.95 %, respectively.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"175 ","pages":"Article 113064"},"PeriodicalIF":7.2,"publicationDate":"2025-04-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143816050","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Twin-population multiple knowledge-guided transfer prediction framework for evolutionary dynamic multi-objective optimization","authors":"Shijie Zhao ,&nbsp;Tianran Zhang ,&nbsp;Miao Chen ,&nbsp;Lei Zhang","doi":"10.1016/j.asoc.2025.113113","DOIUrl":"10.1016/j.asoc.2025.113113","url":null,"abstract":"<div><div>Dynamic multi-objective evolutionary algorithms (DMOEAs) have been widely studied, and one of the main tasks is the need for algorithms to track Pareto optimal front (POF) under dynamic environmental changes. Existing methods integrate transfer learning (TL) techniques to predict the initial population for the new environment. However, the lack of transferred individual diversity and inaccurate moving directions lead to poor performance of DMOEAs. Therefore, this work proposes a Twin-population Multiple Knowledge-guided Transfer prediction (TMKT) framework to form an initial population for the new environment. Three strategies, i.e., Twin Populations Guided prediction (TPG), SVM-based Multi-knowledge prediction (SVM-M) and Kernel Subspace Alignment for Transfer prediction (KSA-T), are designed to mine and transfer positive historical knowledge for accurately predicting changing POFs. First, TPG is used to obtain new approximate individuals and provide potential directions of subsequent transfer, which splits the population into two twin populations based on upper and lower quartile points of the first objective and their angles. Subpopulations transmit information by different similarity methods to obtain their new positions. Secondly, to obtain solutions with better diversity and convergence, SVM-M trains a certain classifier that can discriminate between positive and negative solutions and predicts labels of noisy solutions based on useful knowledge from the first two environments. Third, KSA-T is proposed to further enhance the accuracy of the new population prediction. The kernel trick and second-order feature alignment are introduced in subspace alignment to develop a new TL technique called Kernel Subspace Alignment (KSA) for adaptively achieving homotypic distributions of the source domain and target domain. Solutions predicted by TPG as the target domain are employed to guide the evolution, and obtained-SVM-M positive solutions are transferred to the new environment via KSA. TMKT is integrated with two baseline algorithms MOEA/D and NSGA-II to construct DMOEAs. Numerical results on 14 functions of different variation types and a real parameter optimization problem of control system validate the superior dynamic optimization performance of TMKT compared with five state-of-the-art algorithms.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"175 ","pages":"Article 113113"},"PeriodicalIF":7.2,"publicationDate":"2025-04-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143807430","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Dynamic task offloading in edge computing for computer access point selection based on adaptive deep reinforcement learning with meta-heuristic optimization","authors":"S. Vidya,&nbsp;R. Gopalakrishnan","doi":"10.1016/j.asoc.2025.113105","DOIUrl":"10.1016/j.asoc.2025.113105","url":null,"abstract":"<div><div>The computationally intensive tasks are processed by mobile devices which include data processing, virtual reality, and artificial intelligence. The computational resources of the mobile devices are very low so they are suited to perform all tasks with low latency. Mobile Edge Computing (MEC) is a cutting-edge computing model that offloads computation-intensive tasks to MEC servers to increase the capability of computing in Mobile Devices (MDs). Due to the extensive use of Wireless Local Area Networks (WLAN), each MD can use numerous Wireless Access Points (WAPs) to offload tasks to a server. In this research work, the task offloading problem is determined by considering the delay-sensitive task along with edge load dynamics to reduce the long-term cost. The distributed algorithm based on Adaptive Deep Reinforcement Learning (ADRL) is introduced, where every device is analyzed for offloading decisions without knowing the task model of other devices. The parameters in the model are optimized using the Fitness-based Piranha Foraging Optimization Algorithm (F-PFOA) to enhance the performance of the model. Finally, the evaluation is done by using the various metrics to showcase the effectiveness of the proposed model, and it gives the throughput is 93.5, which is enhanced than other existing models. Thus, the simulation outcome with a greater number of mobile devices and corresponding edge nodes showed that the developed optimization minimizes the dropped task’s ratio and average task delay respectively. The result of the designed model outperformed better than other available models.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"176 ","pages":"Article 113105"},"PeriodicalIF":7.2,"publicationDate":"2025-04-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143829447","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep neural aggregation for recommending items to group of users","authors":"Jorge Dueñas-Lerín ,&nbsp;Raúl Lara-Cabrera ,&nbsp;Fernando Ortega ,&nbsp;Jesús Bobadilla","doi":"10.1016/j.asoc.2025.113059","DOIUrl":"10.1016/j.asoc.2025.113059","url":null,"abstract":"<div><div>Modern society dedicates a significant amount of time to digital interaction, as social life is more and more related to digital life, the information of groups’ interaction with the elements of the system is increasing. One key tool for the digital society is Recommender Systems, intelligent systems that learn from our past actions to propose new ones that align with our interests. Some of these systems have specialized in learning from the behavior of user groups to make recommendations to a group of individuals who want to perform a joint task. This research presents an innovative approach to representing group user preferences using deep learning techniques, enhancing recommendations for joint tasks. The proposed aggregation model has been evaluated using two different foundational models, GMF and MLP, four different datasets, and nine group sizes. The experimental results demonstrate the improvement achieved by employing the proposed aggregation model compared to the state-of-the-art, and this aggregation strategy can be applied to upcoming models and architectures.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"175 ","pages":"Article 113059"},"PeriodicalIF":7.2,"publicationDate":"2025-04-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143807431","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Design of deep learning networks for nonlinear delay differential system for Stuxnet virus spread in an air gapped critical environment","authors":"Muhammad Junaid Ali Asif Raja ,&nbsp;Zaheer Masood ,&nbsp;Ijaz Hussain ,&nbsp;Aneela Zameer ,&nbsp;Muhammad Asif Zahoor Raja","doi":"10.1016/j.asoc.2025.113091","DOIUrl":"10.1016/j.asoc.2025.113091","url":null,"abstract":"<div><div>Within the tranquil confines of air-gapped environment, the custodians of digital fortitude must recognize the limitations of a singular defense mechanism, the cornerstone of this defensive architecture lies in proactive threat detection and rapid response capabilities. In the presented study, a deep-learning based bidirectional LSTM architecture is designed to accurately capture the time-delay differential propagation dynamics of the Stuxnet virus in an air gapped environment intricately linked with a network of critical control infrastructure. To address the challenges encountered in compromising the air gapped environment, the mathematical model introduces time delay factors <span><math><msub><mrow><mi>τ</mi></mrow><mrow><mn>1</mn></mrow></msub></math></span>, <span><math><msub><mrow><mi>τ</mi></mrow><mrow><mn>2</mn></mrow></msub></math></span> and <span><math><msub><mrow><mi>τ</mi></mrow><mrow><mn>3</mn></mrow></msub></math></span>, necessary for exploiting the susceptible USB media, susceptible air gapped computers utilizing infected USB media and connected susceptible computers using infected computers respectively. Removable storage media serves as a pivotal link in bridging the air gapped environment and controlling the industrial controllers connected to critical systems thereby posing a significant threat to the integrity of the entire system. Synthetic temporal simulations serve as the ground truth for dual-layer bidirectional LSTM networks exactment on various scenarios involving the infiltration of the air-gapped environment by the Stuxnet virus in a time delay differential system. A detailed comparative analysis with numerical outcomes showed minimal disparity between the predictions generated by LSTM networks, with mean squared error (MSE) values falling within the range of <span><math><mrow><mn>1</mn><msup><mrow><mn>0</mn></mrow><mrow><mo>−</mo><mn>7</mn></mrow></msup></mrow></math></span> underscoring the effectiveness, robustness, and stability of the proposed neural networks in predicting the complex dynamics of virus in air gapped situation.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"175 ","pages":"Article 113091"},"PeriodicalIF":7.2,"publicationDate":"2025-04-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143791983","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A biased randomised GRASP for the electric vehicle routing problem with heterogeneous supplemental infrastructures","authors":"Rui Xu ,&nbsp;Bowen Song ,&nbsp;Wei Xiao ,&nbsp;Xing Fan","doi":"10.1016/j.asoc.2025.113109","DOIUrl":"10.1016/j.asoc.2025.113109","url":null,"abstract":"<div><div>Green logistics policies have positioned electric vehicles (EVs) as the preferred choice for logistics. Prompted by technological advancements, more companies are now adopting electric logistics vehicles equipped with both charging and battery swapping capabilities. This study addresses the electric vehicle routing problem (EVRP) by integrating various charging technologies, partial charging strategies, and different battery swapping specifications. A mixed-integer programming (MIP) model is developed to minimise total logistics costs, including vehicle operating costs, energy replenishment costs, and variable mileage costs. To solve this problem, we design a biased randomised-greedy randomised adaptive search procedure (BR-GRASP) algorithm incorporating geometric distribution. This algorithm is complemented by local search operators and energy management strategies designed for heterogeneous supplemental infrastructures (HSI). For efficient iterative optimisation, we employ a variable neighbourhood descent (VND) mechanism. Computational experiments validate the effectiveness of HSI and the proposed algorithm from multiple perspectives. Additionally, a real-world case study demonstrates the significant benefits of applying our methods to a logistics company. The research findings offer decision-making recommendations and managerial insights for logistics companies adopting EVs, as well as for relevant government agencies.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"175 ","pages":"Article 113109"},"PeriodicalIF":7.2,"publicationDate":"2025-04-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143821257","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Dynamic interrelationships among crude oil, green bond, and carbon markets: Evidence from fuzzy logic autoencoders","authors":"Nini Johana Marín-Rodríguez ,&nbsp;Elie Bouri ,&nbsp;Juan David González-Ruiz ,&nbsp;Sergio Botero ,&nbsp;Alejandro Peña","doi":"10.1016/j.asoc.2025.113112","DOIUrl":"10.1016/j.asoc.2025.113112","url":null,"abstract":"<div><div>This paper investigates the dynamic interrelationships among various markets covering crude oil, green bonds, and carbon emissions from January 2014 to October 2022, using a Fuzzy Logistic Autoencoder (FLAE) model, which elevates methodological sophistication and helps capturing intricate and complex relationships across the three markets. Different features of FLAE, such as identifying crossed lags and introducing a novel sigmoid-type activation function, enhance structural stability and establish the model as a reference for studying cross-temporal effects across markets. The key findings indicate that green bond returns negatively impact the returns of carbon emission allowances and Brent oil in the short and medium term. The impact of carbon emission allowance returns and oil returns on the forecast of green bond returns is comparatively trivial. Forecasting green bond returns is primarily driven by its short-term lags. These findings should be useful for portfolio managers in energy markets, environmentally conscious investors, and policy-makers concerned with financial sustainability amid the energy transition.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"175 ","pages":"Article 113112"},"PeriodicalIF":7.2,"publicationDate":"2025-04-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143791932","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}