Explore the vast range of titles available.

Forest fires have frequently occurred and caused great harm to people’s lives. Many researchers use machine learning techniques to predict forest fires by considering spatio-temporal data features. However, it is difficult to efficiently obtain the features from large-scale, multi-source, heterogeneous data. There is a lack of a method that can effectively extract features required by machine learning-based forest fire predictions from multi-source spatio-temporal data. This paper proposes a forest fire prediction method that integrates spatio-temporal knowledge graphs and machine learning models. This method can fuse multi-source heterogeneous spatio-temporal forest fire data by constructing a forest fire semantic ontology and a knowledge graph-based spatio-temporal framework. This paper defines the domain expertise of forest fire analysis as the semantic rules of the knowledge graph. This paper proposes a rule-based reasoning method to obtain the corresponding data for the specific machine learning-based forest fire prediction methods, which are dedicated to tackling the problem with real-time prediction scenarios. This paper performs experiments regarding forest fire predictions based on real-world data in the experimental areas Xichang and Yanyuan in Sichuan province. The results show that the proposed method is beneficial for the fusion of multi-source spatio-temporal data and highly improves the prediction performance in real forest fire prediction scenarios.

Effectively managing multi-source heterogeneous data remains a critical challenge in distributed cyber-physical systems (CPS). To address this, we present a novel and edge-centric computing framework integrating four key technological innovations. Firstly, a hybrid OPC communication stack seamlessly combines Client/Server, Publish/Subscribe, and P2P paradigms, enabling scalable interoperability across devices, edge nodes, and the cloud. Secondly, an event-triggered adaptive Kalman filter is introduced; it incorporates online noise-covariance estimation and multi-threshold triggering mechanisms. This approach significantly reduces state-estimation error by 46.7% and computational load by 41% compared to conventional fixed-rate sampling. Thirdly, temporal asynchrony among edge sensors is resolved by a Dynamic Time Warping (DTW)-based data-fusion module, which employs optimization constrained by Mahalanobis distance. Ultimately, a content-aware deterministic message queue data distribution mechanism is designed to ensure an end-to-end latency of less than 10 ms for critical control commands. This mechanism, which utilizes a \"rules first\" scheduling strategy and a dynamic resource allocation mechanism, guarantees low latency for key instructions even under the response loads of multiple data messages. The core contribution of this study is the proposal and empirical validation of an architecture co-design methodology aimed at ultra-high-performance industrial systems. This approach moves beyond the conventional paradigm of independently optimizing individual components, and instead prioritizes system-level synergy as the foundation for performance enhancement. Experimental evaluations were conducted under industrial-grade workloads, which involve over 100 heterogeneous data sources. These evaluations reveal that systems designed with this methodology can simultaneously achieve millimeter-level accuracy in field data acquisition and millisecond-level latency in the execution of critical control commands. These results highlight a promising pathway toward the development of real-time intelligent systems capable of meeting the stringent demands of next-generation industrial applications, and demonstrate immediate applicability in smart manufacturing domains.

To enhance the safety of microgrid switching and the identification of misoperations, we propose Time-Synchronized Misoperation Recognition (TS-MR), a method tailored to switching operations. The approach performs rule-based pre-screening grounded in operating procedures and anti-misoperation interlocking, achieves millisecond-level time synchronization of multi-source heterogeneous data via a two-stage scheme that combines variational Bayesian inference with a UKF, and employs a fusion of a Transformer, a TCN, and a GNN for cross-modal representation with interpretable discrimination. Laboratory records constitute the training, validation, and test sets; HIL data are used solely for independent cross-validation; and public datasets are used only for cross-domain robustness calibration, and none contributes to training, validation, or threshold tuning. Under a unified evaluation protocol, TS-MR attains 94.69% accuracy and an AUC of 0.977 in typical switching scenarios; end-to-end latency is about 80 ms; the core forward-pass latency is about 42 ms; the per-inference latency is 55 ms; and computational complexity is about 3.4 GFLOPs. Compared with CNN-BiLSTM, ConvLSTM, and GAT under identical preprocessing, time synchronization, and fixed random seeds, TS-MR improves accuracy by 0.9 to 3.7% points and AUC by 0.024 to 0.057. These results indicate that TS-MR provides high-confidence misoperation recognition and interpretable assessment for microgrid switching while satisfying engineering-grade real-time constraints.

Incorporating renewable energy solutions into the power grid is essential for facilitating a shift in energy sources. Ensuring the effectiveness and dependability of communication technologies within smart distribution grids is vital, as these grids serve as the link between end users and the central station, thereby contributing to the grid’s stable functioning and the effective dissemination of energy. This study introduces a framework for communication that tackles the challenges associated with the integration of multi-source, heterogeneous data in smart distribution grids. The framework leverages multi-source heterogeneous data fusion and capitalizes on edge computing to efficiently process and consolidate data from diverse sensors and devices. An optimal weight allocation algorithm is proposed to mitigate data redundancy and to improve the energy efficiency of communication. The paper demonstrates, through theoretical examination and empirical tests, the proposed system’s enhanced communication capabilities, especially in terms of data transmission effectiveness and energy usage optimization. The discussion highlights the advantages of smart distribution networks over conventional communication methods and offers insights for further advancements in communication technologies within this sector.

In view of the suitability assessment of forest land resources, a consistent fuzzy assessment method with heterogeneous information is proposed. Firstly, some formulas for transforming large-scale real data and interval data into fuzzy numbers are provided. To derive the unified representation of multi-granularity linguistic assessment information, a fuzzy quantitative transformation for multi-granularity uncertain linguistic information is proposed. The proofs of the desirable properties and some normalized formulas for the trapezoidal fuzzy numbers are presented simultaneously. Next, the objective weight of each assessment indicator is further determined by calculating the Jaccard–Cosine similarity between the trapezoidal fuzzy numbers. Moreover, the trapezoidal fuzzy numbers corresponding to the comprehensive assessment values of each alternative are obtained. The alternatives are effectively ranked according to the distance from the centroid of the trapezoidal fuzzy number to the origin. Finally, based on the proposed consistent fuzzy assessment method, the suitability assessment of forest land resources is achieved under a multi-source heterogeneous data setting.

With the development of big data and blockchain technology, a large amount of multi-source heterogeneous data has been accumulated in the agricultural field by before, during and after production. Agricultural information service systems are often targeted at specific regions, specific applications and specific data resources. Due to the lack of effective analysis and refining, the conversion efficiency of data resources into useful information is too low, resulting in contradiction between the continuous enrichment of agricultural data resources and the relative lack of agricultural information services. Therefore, in view of the multi-source heterogeneous characteristics of agricultural data and the specific business needs of different agricultural scenarios, the intelligent processing method of agricultural data is analysed, and a heuristic algorithm based on K-Means limited clustering number is proposed to judge the accuracy of abnormal data processing. By inputting test sample data for testing, the algorithm has improved accuracy by nearly 30% compared to traditional K-Means.

Surface residual stresses ( Rs ) have a significant impact on the performance of machined parts, including fatigue life and corrosion resistance. To enable online monitoring of Rs , many studies have focused on obtaining real-time Rs . However, direct measurement methods, including destructive and non-destructive techniques, will consume too much time or even damage the machined surface. Meanwhile, prediction methods rarely consider dynamic factors as identifying key features from dynamic data is challenging for humans. Therefore, this paper proposes an intelligent prediction method of Rs based on multi-source heterogeneous data, which contain cutting force, cutting temperature, power consumption, and cutting noise. Firstly, an Improved Convolutional Neural Network is established to identify features from the dynamic heterogeneous data. The mean training identification accuracy reaches 99.6%, which is significantly better than that (71%) obtained by the original convolutional neural network. Then, the Principal Component Analysis is built to automatically determine the key features, which benefit the subsequent Rs prediction. Finally, based on the key features, the Gaussian Process Regression is proposed to predict Rs in two directions. From the various experiments, the performance of the intelligent prediction method is validated, and the prediction accuracy rates for two directions reach 99.10% and 99.13%, respectively. Based on the proposed method, the real-time Rs can be predicted with the key features, which are automatically extracted from the multi-source heterogeneous data. This provides the basis for surface quality monitoring based on online data and greatly improves the level of intelligent manufacturing.

Industrial process data have the characteristics of heterogeneity, dimensional inconsistency and multi time scales, which increase the difficulty of condition evaluation in industrial process using multi-source data. To address these problems, a multi-source heterogeneous data fusion model is proposed for the condition evaluation of aluminum electrolysis cell. Firstly, the deep residual network (ResNet) is used to extract the superheat degree features of the fire hole video, the ResNet and wavelet packet are used to extract the cell voltage features, thereby achieving the isomorphism of heterogeneous data. An anode current features extraction method based on dynamic pivotal sequence is used to reduce the dimension of anode current data and extract features. Then, a fusion model of feature layer and data layer based on CatBoost is proposed, which comprehensively considers the material-energy balance mechanism knowledge and current efficiency to describe the dynamic coupling relationship between various data sources. The experimental evaluation results on the actual industrial aluminum electrolysis dataset show that our method improves the performance by 2.3% compared with existing multi-source data fusion method.

The prediction of fuel consumption and Carbon Intensity Index (CII) of ships is crucial for optimizing decarbonization strategies in the maritime industry. This study proposes a ship fuel consumption prediction model based on the Long Short-Term Memory with Self-Attention Mechanism (SA-LSTM). The model is applied to a container ship of 2400 TEU to predict its hourly fuel consumption, hourly CII, and annual CII rating. Four different feature sets are selected from these data sources and are used as inputs for SA-LSTM and another ten models. The results demonstrate that the SA-LSTM model outperforms the other models in prediction accuracy. Specifically, the Mean Absolute Percentage Error (MAPE) for fuel consumption predictions using the SA-LSTM model is reduced by up to 20% compared to the XGBoost and by up to 12% compared to the LSTM model. Additionally, the SA-LSTM model achieves the highest accuracy in annual CII predictions.

In this paper, the digital twin platform for the indoor environment is constructed by combining digital twin technology and modern indoor environment elements to innovate the modern indoor environment design method. The digital twin platform is designed for modern indoor environments in two aspects: data acquisition and 3D model visualization for indoor environments. The indoor environment data are collected, cleaned and quasi-exchanged using sensors, the collected multi-source heterogeneous data of the indoor environment are fused by the time alignment method, and the 3D model of the indoor environment is driven by the design of the 3D model’s operations of translation, rotation and scaling. On this basis, the performance of the indoor environment digital twin platform is analyzed, and the data acquisition method and driving effects of the 3D model are explored. The results show that the data transmission measurement delay is within 20ms, the display delay is within 70ms, the transmission frames per second are basically stabilized at about 200FPS, 100FPS, 60FPS, the accuracy reaches 0.9 in the case of multiple data acquisition, and the fusion speed is about 3.4m/s, and the success rate of the driving operation of the overall three-dimensional model of the indoor environment is all greater than 0.96.