Procedia Computer Science最新文献_第5页

Accurate Generative Model of Machine Translation System Based on Reinforcement Learning Algorithm 基于强化学习算法的机器翻译系统精确生成模型

Procedia Computer Science Pub Date : 2026-01-01 Epub Date: 2026-06-02 DOI: 10.1016/j.procs.2026.04.247

Ruomu Wang

引用次数: 0

Construction and Empirical Analysis of Multi-source Data Fusion Intelligent Algorithm Model for Dual Control of Carbon Emissions 碳排放双重控制的多源数据融合智能算法模型构建与实证分析

Procedia Computer Science Pub Date : 2026-01-01 Epub Date: 2026-05-30 DOI: 10.1016/j.procs.2026.03.226

Meng Zhang, Qian Zhang

{"title":"Construction and Empirical Analysis of Multi-source Data Fusion Intelligent Algorithm Model for Dual Control of Carbon Emissions","authors":"Meng Zhang, Qian Zhang","doi":"10.1016/j.procs.2026.03.226","DOIUrl":"10.1016/j.procs.2026.03.226","url":null,"abstract":"<div><div>In recent years, Hainan has actively promoted the construction of zero-carbon demonstration parks, aiming to achieve dual control of regional carbon emissions total and intensity through clean energy utilization, low-carbon industrial layout, and smart energy management. However, the parks still face many bottlenecks in energy monitoring, carbon emission assessment, and dynamic regulation, such as high data heterogeneity, complex spatiotemporal coupling, and insufficient real-time feedback mechanisms, resulting in limited accuracy in carbon emission prediction and making it difficult to provide reliable support for park management and policy optimization. In response to this, this paper built a multi-source data fusion intelligent algorithm model to control dual carbon emission. This model automatically derives time-series features, industry-related features using multi-layer nonlinear mapping, and introduces two constraint terms of the total carbon emissions and intensity in the objective to optimize dynamically the emission prediction and control performance. The experimental findings indicate that the model is optimally performed with a deep autoencoding fusion approach, and R2 reaches 0.96, which means that this approach is effective in capturing the nonlinear relationships between time, space, and industry, which improve the capabilities of the model to represent features.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"279 ","pages":"Pages 191-200"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148182502","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Design and Accuracy Improvement of Distributed Data Fusion Algorithms for UAV Swarms 无人机群分布式数据融合算法设计与精度提高

Procedia Computer Science Pub Date : 2026-01-01 Epub Date: 2026-05-30 DOI: 10.1016/j.procs.2026.03.249

Zuhui Shan

引用次数: 0

Research On a Framework for Generating and Maintaining Automated Drawline Syntax Trees for Complex Programming Languages 复杂编程语言自动拖线语法树生成与维护框架研究

Procedia Computer Science Pub Date : 2026-01-01 Epub Date: 2026-05-30 DOI: 10.1016/j.procs.2026.03.251

Boyang Liu

{"title":"Research On a Framework for Generating and Maintaining Automated Drawline Syntax Trees for Complex Programming Languages","authors":"Boyang Liu","doi":"10.1016/j.procs.2026.03.251","DOIUrl":"10.1016/j.procs.2026.03.251","url":null,"abstract":"<div><div>In the development of modern programming languages, how to efficiently generate and maintain abstract syntax trees (AST) is the key to improve the ability of program design and optimization. Abstract syntax trees not only play a core role in compilers, but also are widely used in code refactoring, performance analysis and other automation fields. With the popularity of multi-core processors, concurrent programming has become an important means to improve performance, and the choice of threading mechanism is very important to the efficiency and maintainability of programs. Therefore, it is of practical significance to design a framework for automatically generating and maintaining abstract syntax trees to help code optimization. This paper proposes an automatic abstract syntax tree generation and maintenance framework for multithreading in Java language. It uses static code analysis to transform and optimize the thread model, especially in the process of migrating traditional thread model to efficient thread pool framework. The framework automatically completes code refactoring and performance improvement by parsing program structure layer by layer. Through the experimental evaluation in Eclipse environment, the results show that the framework can quickly complete the transformation of thread mechanism, significantly improve the efficiency of concurrent program development, optimize the performance and maintainability of programs, and provide a new idea of parallel programming in multi-core processor environment.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"279 ","pages":"Pages 420-428"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148182747","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Automatic Generation of Labor Cost Analysis Report Combining GAN and NLP Technology 结合GAN和NLP技术的人工成本分析报表自动生成

Procedia Computer Science Pub Date : 2026-01-01 Epub Date: 2026-05-30 DOI: 10.1016/j.procs.2026.03.253

Enlai Zhao, Heyi Zhang, Ming Li, Xueliang Zhang, Shuo Wang, Anran Wang

{"title":"Automatic Generation of Labor Cost Analysis Report Combining GAN and NLP Technology","authors":"Enlai Zhao, Heyi Zhang, Ming Li, Xueliang Zhang, Shuo Wang, Anran Wang","doi":"10.1016/j.procs.2026.03.253","DOIUrl":"10.1016/j.procs.2026.03.253","url":null,"abstract":"<div><div>In recent years, companies have faced a large number of data processing and report generation needs in the field of labor cost analysis. Traditional labor cost analysis report generation methods have the disadvantages of long generation time and error-proneness, so it is necessary to automatically and intelligently generate accurate labor cost analysis reports. To solve this problem, this study combines the Generative Adversarial Network (GAN) and Natural Language Processing (NLP) technology to automatically generate a labor cost analysis report that meets the requirements through the collaborative work of deep learning models. First, this paper uses GAN to generate simulated labor cost data to ensure the diversity and realism of training data; then, the text generation model in NLP technology is used to generate a structured analysis report based on this data to ensure the professionalism and accuracy of the report content; finally, by evaluating and optimizing the generated results, the quality and generation efficiency of the report are further improved. In the experimental results, when the data volume increased to 1000, the BLEU (Bilingual Evaluation) score increased to 0.80 and the ROUGE (Recall-Oriented Summary Evaluation) score increased to 0.85. In addition, the Flesch-Kincaid readability index decreased from 65 to 51, and the manual score increased from 4.2 to 5.0. The method proposed in this paper that combines GAN and NLP technology can effectively improve the performance of labor cost analysis reports and has broad application prospects.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"279 ","pages":"Pages 437-444"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148182750","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Construction of Regional Digital Economy Vitality Evaluation Model Driven by Federated Learning 基于联邦学习的区域数字经济活力评价模型构建

Procedia Computer Science Pub Date : 2026-01-01 Epub Date: 2026-05-30 DOI: 10.1016/j.procs.2026.03.269

Chenyu Liu

{"title":"Construction of Regional Digital Economy Vitality Evaluation Model Driven by Federated Learning","authors":"Chenyu Liu","doi":"10.1016/j.procs.2026.03.269","DOIUrl":"10.1016/j.procs.2026.03.269","url":null,"abstract":"<div><div>In the context of regional digital economy vitality assessment facing the traditional model ignoring cross-regional collaboration, this study aims to build an evaluation model based on federated learning to balance data privacy and evaluation accuracy. First, an evaluation index system including eight dimensions such as digital infrastructure investment, industrial digitalization scale, and innovation and R&D intensity is designed; secondly, a federated learning framework is constructed: provincial nodes use CNN-LSTM hybrid networks to train regional economic time series data locally, and only upload encrypted model parameters to the central server; the central server aggregates parameters through the federated averaging algorithm and sends updates, iterating 50 rounds; finally, the global model is integrated to predict the vitality index. The experiment selects data from 2018 to 2022 from six provinces in eastern, central and western China for testing: the evaluation accuracy of the federated model on the test set reaches 95%, and the ablation experiment confirmed that the federated framework contributes 68% of the error reduction. The model significantly improves the evaluation accuracy while protecting data privacy, effectively solves the problem of cross-domain data collaboration, and provides a reliable quantitative basis for the formulation of differentiated regional digital policies.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"279 ","pages":"Pages 577-585"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148182698","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Multi-sensor Fusion Localization and Navigation Algorithm for Hotel Service Robots in Complex and Dynamic Environments 复杂动态环境下酒店服务机器人多传感器融合定位与导航算法

Procedia Computer Science Pub Date : 2026-01-01 Epub Date: 2026-05-30 DOI: 10.1016/j.procs.2026.03.264

Weiwei Li, Jun Zhang, Wei Xiao

{"title":"Multi-sensor Fusion Localization and Navigation Algorithm for Hotel Service Robots in Complex and Dynamic Environments","authors":"Weiwei Li, Jun Zhang, Wei Xiao","doi":"10.1016/j.procs.2026.03.264","DOIUrl":"10.1016/j.procs.2026.03.264","url":null,"abstract":"<div><div>With the rapid development of intelligent hotel industry, service robot has become the core tool to improve service efficiency and quality. However, in a highly traffic-friendly and dynamic hotel environment, they often face issues such as positioning and navigation errors. Traditional single-sensor localization solutions are susceptible to light, enclosure, and electromagnetic interference and cannot meet high accuracy and durability requirements in complex scenes. This paper first developed a multi-sensor fusion system integrating laser radar, visual camera, inertial measurement unit (IMU) and ultra-wide band (UWB). Secondly, an improved adaptive Kalman filter based sensor data fusion algorithm is proposed to solve the data synchronization problem under dynamic noise. Finally, a dynamic route planning strategy was developed to integrate semantic information about the surrounding environment. The proposed method has a completion time of about 42s, an average speed of 1.90m/s, a completion time of about 55s and a speed of 1.45m/s for LiDAR-IMU, a completion time of about 72s and a speed of 1.14m/s for LiDAR-only. The navigation efficiency of the proposed method is better.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"279 ","pages":"Pages 534-542"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148182769","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Exploration of Urban Ecological Planning and Urbanization Balance Model Driven by Simulated Annealing Algorithm 模拟退火算法驱动的城市生态规划与城市化平衡模型探索

Procedia Computer Science Pub Date : 2026-01-01 Epub Date: 2026-05-30 DOI: 10.1016/j.procs.2026.03.210

Xiaotian Jiang, You Li

{"title":"Exploration of Urban Ecological Planning and Urbanization Balance Model Driven by Simulated Annealing Algorithm","authors":"Xiaotian Jiang, You Li","doi":"10.1016/j.procs.2026.03.210","DOIUrl":"10.1016/j.procs.2026.03.210","url":null,"abstract":"<div><div>Maintaining ecological balance and development has been one of the biggest challenges in the field of urban planning. This paper seeks to achieve the optimal balance between urban ecological planning and urbanization through SA algorithm. The study first constructs a comprehensive evaluation model for the balance between the urban ecological planning and urbanization, which includes multiple dimensions such as economic development, environment protection, resource utilization and social welfare. There after SA algorithm is introduced as an optimization tool for finding the optimal solution of the model by simulating the energy changes and the state transitions during the solid annealing. This paper specifies the initial temperature of the algorithm, and generates an initial urban planning scheme at random as starting point. Subsequently, the neighborhood solutions are produced by changing the parameters in the scheme. When evaluating these neighborhood solutions, these are compared by the value of the objective function and then a decision of accepting a new solution which is inferior to the current solution is made by the Metropolis criterion. The temperature is slowly decreased in the iteration process so that the algorithm slowly converges to an approximate optimal solution. The SA algorithm succeeds in detecting the best scheme of urban planning with the average green coverage rate of 38.81%, building density values range primarily in [0.2, 0.4] and the traffic flow becomes higher as compared to PSO algorithm. These optimization results not only greatly improve the comprehensive benefits of the city but also provide a scientific basis for the sustainable development of the city.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"279 ","pages":"Pages 46-54"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148182967","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Carbon Emission Prediction System for Comprehensive Utilization of Straw Based on Multi-Source Data 基于多源数据的秸秆综合利用碳排放预测系统

Procedia Computer Science Pub Date : 2026-01-01 Epub Date: 2026-05-30 DOI: 10.1016/j.procs.2026.03.207

Xinyuan Huang, Tingting Li, Ping Lu, Jiaheng Lin

{"title":"Carbon Emission Prediction System for Comprehensive Utilization of Straw Based on Multi-Source Data","authors":"Xinyuan Huang, Tingting Li, Ping Lu, Jiaheng Lin","doi":"10.1016/j.procs.2026.03.207","DOIUrl":"10.1016/j.procs.2026.03.207","url":null,"abstract":"<div><div>In order to overcome the drawbacks of the conventional methods of carbon emission prediction, which include low precision and the inability to adapt effectively since they rely on single sources of data and are unable to generalize predictions to different items, this paper develops a prediction system that yields on multi-source data. To identify the spatiotemporally aligned feature engineering framework, a composite of heterogeneous data, which are remote sensing, meteorology, and statistically based, is established. Then a hybrid prediction model that integrates mechanistic constraints and adaptive graph neural networks is suggested, which directs the learning of the data through the laws of physics and describes the spatial relationship of the carbon emissions dynamically. Lastly, a prediction engine that is based on multi-scenario simulation is created. It has been demonstrated that the system minimizes RMSE by 22.5 percent compared to the XGBoost (eXtreme Gradient Boosting) model, increases R² to 0.923, and the generalization stability in untrained regions improves a lot. This system can efficiently address the issue of precise prediction and cross-region adaption, which presents a stable analytical instrument of managing the carbon emission of agriculture.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"279 ","pages":"Pages 19-27"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148182972","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Full-Link Non-Intrusive Distributed Call Chain Tracing and Continuous Function Profiling Based on eBPF 基于eBPF的全链路非侵入式分布式调用链跟踪和连续功能分析

Procedia Computer Science Pub Date : 2026-01-01 Epub Date: 2026-05-30 DOI: 10.1016/j.procs.2026.03.292

Libin Li, Wei Sun, Jie Cui, Yunzhi Liu, Yan Li

{"title":"Full-Link Non-Intrusive Distributed Call Chain Tracing and Continuous Function Profiling Based on eBPF","authors":"Libin Li, Wei Sun, Jie Cui, Yunzhi Liu, Yan Li","doi":"10.1016/j.procs.2026.03.292","DOIUrl":"10.1016/j.procs.2026.03.292","url":null,"abstract":"<div><div>With the increasing complexity of distributed systems, traditional distributed call chain tracking methods are invasive, performance is significantly reduced, and the granularity of tracking is insufficient to meet the demands of high reliability and high-performance surveillance of large-scale distributed systems. First, the limits of existing distributed tracking technology were analyzed to propose a distributed call chain tracking solution based on ebpf, non-invasive and continuous. In this study, the researchers used ebpf’s dynamic detection technology to collect non-intrusive data about key nodes, such as system calls and function execution. They designed a lightweight tracking data protocol and efficient transmission mechanism, built a complete link call topology card, and developed a functional-level continuous contour module to identify performance bottlenecks in real time. Experimental results have shown that this solution reduces the average execution time of the inventory request function from 65 to 6 milliseconds, the average execution time of the payment processing function from 42 to 4 milliseconds, and the average execution time of the registration function from 120 to 120 milliseconds, while ensuring the accuracy and integrity of tracking. Once optimized, the payment interface response time was reduced by approximately 90%, demonstrating the practical importance of this article in terms of performance optimization.</div></div>","PeriodicalId":20465,"journal":{"name":"Procedia Computer Science","volume":"279 ","pages":"Pages 784-791"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148182831","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0