Information Features Research Articles

ERT-GFAN: A multimodal drug–target interaction prediction model based on molecular biology and knowledge-enhanced attention mechanism

In drug discovery, precisely identifying drug–target interactions is crucial for finding new drugs and understanding drug mechanisms. Evolving drug/target heterogeneous data presents challenges in obtaining multimodal representation in drug–target prediction(DTI). To deal with this, we propose ‘ERT-GFAN’, a multimodal drug–target interaction prediction model inspired by molecular biology. Firstly, it integrates bio-inspired principles to obtain structure feature of drugs and targets using Extended Connectivity Fingerprints(ECFP). Simultaneously, the knowledge graph embedding model RotatE is employed to discover the interaction feature of drug–target pairs. Subsequently, Transformer is utilized to refine the contextual neighborhood features from the obtained structure feature and interaction features, and multi-modal high-dimensional fusion features of the three-modal information constructed. Finally, the final DTI prediction results are outputted by integrating the multimodal fusion features into a graphical high-dimensional fusion feature attention network (GFAN) using our innovative multimodal high-dimensional fusion feature attention. This multimodal approach offers a comprehensive understanding of drug–target interactions, addressing challenges in complex knowledge graphs. By combining structure feature, interaction feature, and contextual neighborhood features, ‘ERT-GFAN’ excels in predicting DTI. Empirical evaluations on three datasets demonstrate our method’s superior performance, with AUC of 0.9739, 0.9862, and 0.9667, AUPR of 0.9598, 0.9789, and 0.9750, and Mean Reciprocal Rank(MRR) of 0.7386, 0.7035, and 0.7133. Ablation studies show over a 5% improvement in predictive performance compared to baseline unimodal and bimodal models. These results, along with detailed case studies, highlight the efficacy and robustness of our approach.

Heart Diseases Recognition Model Based on HRV Feature Extraction over 12-Lead ECG Signals.

Heart Rate Variability (HRV) refers to the capability of the heart rhythm to vary at different times, typically reflecting the regulation of the heart by the autonomic nervous system. In recent years, with advancements in Electrocardiogram (ECG) signal processing technology, HRV features reflect various aspects of cardiac activity, such as variability in heart rate, cardiac health status, and responses. We extracted key features of HRV and used them to develop and evaluate an automatic recognition model for cardiac diseases. Consequently, we proposed the HRV Heart Disease Recognition (HHDR) method, employing the Spectral Magnitude Quantification (SMQ) technique for feature extraction. Firstly, the HRV signals are extracted through electrocardiogram signal processing. Then, by analyzing parts of the HRV signal within various frequency ranges, the SMQ method extracts rich features of partial information. Finally, the Random Forest (RF) classification computational method is employed to classify the extracted information, achieving efficient and accurate cardiac disease recognition. Experimental results indicate that this method surpasses current technologies in recognizing cardiac diseases, with an average accuracy rate of 95.1% for normal/diseased classification, and an average accuracy of 84.8% in classifying five different disease categories. Thus, the proposed HHDR method effectively utilizes the local information of HRV signals for efficient and accurate cardiac disease recognition, providing strong support for cardiac disease research in the medical field.

Digital toy design: A doll head modeling method based on model matching and texture generation

With the development of technology and the improvement of living standards, consumer demand for toy products has also changed. More people are showing strong interest and demand for digital toy products. However, in the current digital toy design process, extracting the hairstyle features of the doll’s head is still a challenge. Therefore, this study extracts the two-dimensional contour of the target hairstyle and matches it with the template hairstyle in the database. Then, combining hair texture information and hairstyle structural features, fine geometric texture details are generated on the hairstyle mesh surface. Finally, a doll head modeling method based on model matching and texture generation is proposed. The results showed that the hairstyle of the Generative model was almost the same as the real hairstyle. Compared with the modeling methods based on interactive genetic algorithm and digital image, the average F1 value of this method was 0.95, and the mean absolute error was the smallest. The accuracy of the target model modeling was 95.4%, and the area enclosed by the receiver operating characteristic curve and coordinate axis was 0.965. In summary, the doll head modeling method based on model matching and texture generation proposed in this study can generate high-precision and realistic hairstyle models corresponding to the hairstyle. The overall shape and local geometric details of the hairstyle can meet the needs of 3D printing, providing certain reference significance for hairstyle reconstruction.

UAV Visual Tracking with Enhanced Feature Information

Unmanned aerial vehicles (UAVs) visual tracking is an important research direction. The tracking object is lost due to the problems of target occlusion, illumination variation, flight vibration and so on. Therefore, based on a Siamese network, this study proposes a UAVs visual tracker named SiamDFT++ to enhance the correlation of depth features. First, the network width of the three-layer convolution after the full convolution neural network is doubled, and the appearance information of the target is fully utilized to complete the feature extraction of the template frame and the detection frame. Then, the attention information fusion module and feature deep convolution module are proposed in the template branch and the detection branch, respectively. The feature correlation calculation methods of the two depths can effectively suppress the background information, enhance the correlation between pixel pairs, and efficiently complete the tasks of classification and regression. Furthermore, this study makes full use of shallow features to enhance the extraction of object features. Finally, this study uses the methods of deep cross-correlation operation and complete intersection over union to complete the matching and location tasks. The experimental results show that the tracker has strong robustness in UAVs short-term tracking scenes and long-term tracking scenes.

Intelligent fault diagnosis of multi-source cross-machine bearings based on center-weighted optimal transport and class-level alignment domain adaptation

Abstract Most of the current domain adaptation research primarily focuses on the single-source or multi-source domain transfer constructed under different working conditions of the same machine. However, when faced with cross-machine tasks with significant domain discrepancies, forcing the direct feature alignment between source and target domain samples may lead to negative transfer, thereby reducing the model’s diagnostic performance. To overcome the above limitations, this paper proposes a multi-source deep transfer model based on center-weighted optimal transport (CWOT) and class-level alignment domain adaptation. Firstly, to enhance the representation capability of deep features, a multi-structure feature representation network is constructed to enrich the information capacity embedded within the deep features, thereby achieving better domain adaptation capabilities. Then, the local maximum mean discrepancy is introduced to fully exploit fine-grained information and discriminative features among different source domains, minimizing the distribution differences among the source domains to the greatest extent, thus capturing reliable and highly generalized multi-source domain invariant features. On this basis, a CWOT strategy is designed, which comprehensively considers the transport cost of intra-domain uncertainty and inter-domain correlation among samples, establishing a more effective transport between source and target domains, alleviating the problem of sample negative transfer, and improving the model’s cross-machine diagnostic performance. Finally, instance studies are conducted through multiple cross-machine transfer diagnostic tasks, demonstrating that the proposed method outperforms existing domain adaptation methods in terms of diagnostic accuracy and fault transfer capability. This research provides a reliable fault diagnosis method for detecting the health status of rotating machinery equipment, promoting the application of domain adaptation technology in practical industry.

DERE-Net: A dual-encoder residual enhanced U-Net for muscle fiber segmentation of H&E images

Accurate segmentation of hematoxylin-eosin (H&E) muscle fiber images is crucial for the diagnosis of weightless muscle atrophy. However, uneven contrast, blurred fiber boundaries and adhesions in H&E images are still challenges for accurate segment complete muscle fiber. Although the existing single-encoder based U-shaped networks can extract local information about the target in H&E images, ignores remote dependencies between muscle fibers and low-level detail features, resulting in poor segmentation accuracy when facing H&E images with complex contrast. Therefore, we propose a dual-encoder residual enhanced U-Net segmentation model (DERE-Net) for effective segment muscle fibers in H&E images. DERE-Net uses Transformer and residual CNN to extract global context information and local features of muscle fibers in parallel, and fuses these two features to enable the model to accurately identify the muscle fibers. The multi-scale semantic information fusion (MSIF) module also utilizes the differences between multi-scale features to recover undetected muscle fiber, thus improving the network’s ability to recognize the object regions. In addition, the attention module is added to the skip connection, making the muscle fiber regions information more prominent and improving the robustness of the model. To demonstrate the DERE-Net effect, the comparative experiments are conducted on our own muscle fiber H&E image dataset, GLAS and MoNuSeg. DERE-Net achieved excellent performance on the muscle fiber H&E dataset (DSC 0.919), GLAS dataset (DSC 0.912), and MoNuSeg dataset (DSC 0.821). These results indicate that DERE-Net can accurately predict muscle fiber regions, providing a new method for accurate diagnosis of muscle atrophy in the future.

A multi-channel spatial information feature based human pose estimation algorithm

Human pose estimation is an important task in computer vision, which can provide key point detection of human body and obtain bone information. At present, human pose estimation is mainly utilized for detection of large targets, and there is no solution for detection of small targets. This paper proposes a multi-channel spatial information feature based human pose (MCSF-Pose) estimation algorithm to address the issue of medium and small targets inaccurate detection of human key points in scenarios involving occlusion and multiple poses. The MCSF-Pose network is a bottom-up regression network. Firstly, an UP-Focus module is designed to expand the feature information while reducing parameter computation during the up-sampling process. Then, the channel segmentation strategy is adopted to cut the features, and the feature information of multiple dimensions is retained through different convolutional groups, which reduces the parameter lightweight network model and makes up for the loss of the feature information associated with the depth of the network. Finally, the three-layer PANet structure is designed to reduce the complexity of the model. With the aid of the structure, it also to improve the detection accuracy and anti-interference ability of human key points. The experimental results indicate that the proposed algorithm outperforms YOLO-Pose and other human pose estimation algorithms on COCO2017 and MPII human pose datasets.

Geometric-aware RGB-D representation learning for hand–object reconstruction

Reconstructing hand–object interaction from single images, crucial for interactive applications, is challenging due to the diversity of hand–object poses and shapes. Depth maps effectively complement RGB data in understanding these interactions geometrically within challenging scenes. However, most existing methods do not fully utilize the potential benefits of RGB-D information fused at different feature levels for reconstruction, limiting their capability to capture the geometric details of hand–object interaction. To address this, we propose an implicit geometric-aware RGB-D representation learning approach using adaptive bidirectional RGB-D feature fusion (ABF) and geometric Fourier feature encoding (GFFE). This method innovatively leverages the hierarchical and complementary high-level features of RGB-D information to enhance the neural implicit representations of hand–object interaction during the feature extraction and encoding stages of RGB-D data. Initially, a two-stream RGB-point cloud encoder, combining CNN and Transformer architectures, extracts appearance and geometric information from RGB and point cloud data. Subsequently, ABF fuses this information, generating dense RGB-D features by leveraging the complementary of appearance and geometric features along with the varying sensitivities of different layers. Finally, GFFE introduces frequency domain information to capture both high-frequency and low-frequency features of objects equally to precisely capture the geometric details of hand–object interaction. Experiments conducted on the real-world DexYCB and the synthetic ObMan benchmarks demonstrate that our approach significantly outperforms existing approaches.

Design of Intelligent Financial System Based on Adaptive Learning Algorithm

The high-frequency trading system in the financial domain has long been a focal point of investigation. This study posits an intelligent financial system design framework predicated on a cross-adaptive self-entropy projection clustering model, aimed at enhancing the efficacy of high-frequency trading systems. A composite distribution model of financial data is formulated to derive sequences of financial data activities. And cross-adaptive learning algorithm is employed to ascertain the interrelated attributes of financial data. Following this, the support vector machine algorithm is applied for the classification processing of these interrelated features, yielding a set of financial data feature vectors, which are then fed into the gray correlation-based information feature extraction model. Through extensive empirical evaluations with authentic trading data, the proposed intelligent financial system design framework exhibits commendable performance, furnishing a viable solution for the intelligent optimization of high-frequency trading systems.

A Novel Multi-modal Sentiment Analysis Based on Multiple Kernel Learning with Margin-Dimension Constraint

Most of the sentiment analysis studies focus on the sentimental classification of pictures in the video, ignoring the spatio-temporal information of the sequence of picture frames as well as text and audio information. The multiple kernel learning is a new hotspot in the field of nuclear machine learning, capable of handling multiple modalities. For multiple kernel learning, it is easy to ignore the basic features that are not discriminative, and cannot make full use of the base features of different modes. This paper puts forward a novel multi-modal fusion model for sentiment analysis, in which a multiple kernel learning algorithm based on convolution margin-dimension constraint is proposed for feature fusion. Moreover, the 3D convolutional neural network is used to extract the features of visual information, and the multiple kernel learning algorithm based on margin-dimension constraint is used to fuse visual, text and audio sentiment features. Experiments conducted on the MOUD and IEMOCAP sentiment databases show that the proposed model outperforms existing models in the field of multi-modal sentiment analysis research.

HRDLNet: a semantic segmentation network with high resolution representation for urban street view images

Semantic segmentation of urban street scenes has attracted much attention in the field of autonomous driving, which not only helps vehicles perceive the environment in real time, but also significantly improves the decision-making ability of autonomous driving systems. However, most of the current methods based on Convolutional Neural Network (CNN) mainly use coding the input image to a low resolution and then try to recover the high resolution, which leads to problems such as loss of spatial information, accumulation of errors, and difficulty in dealing with large-scale changes. To address these problems, in this paper, we propose a new semantic segmentation network (HRDLNet) for urban street scene images with high-resolution representation, which improves the accuracy of segmentation by always maintaining a high-resolution representation of the image. Specifically, we propose a feature extraction module (FHR) with high-resolution representation, which efficiently handles multi-scale targets and high-resolution image information by efficiently fusing high-resolution information and multi-scale features. Secondly, we design a multi-scale feature extraction enhancement (MFE) module, which significantly expands the sensory field of the network, thus enhancing the ability to capture correlations between image details and global contextual information. In addition, we introduce a dual-attention mechanism module (CSD), which dynamically adjusts the network to more accurately capture subtle features and rich semantic information in images. We trained and evaluated HRDLNet on the Cityscapes Dataset and the PASCAL VOC 2012 Augmented Dataset, and verified the model’s excellent performance in the field of urban streetscape image segmentation. The unique advantages of our proposed HRDLNet in the field of semantic segmentation of urban streetscapes are also verified by comparing it with the state-of-the-art methods.

Spatial Information Entropy-Assisted Integrated Sensing and Communication for Integrated Satellite-Terrestrial Networks

To better meet communication needs, 6G proposes Integrated Satellite-Terrestrial Networks. Integrated Sensing and Communication (ISAC) is one of the key technologies of Integrated Satellite-Terrestrial Networks, which can reduce the energy consumption of the system, improve communication efficiency, and increase the utilization rate of spectrum resources. In the existing technology, the Modulated Wideband Converter (MWC) system can provide support for the miniaturization and intelligence of wireless device sensing and communication systems. Therefore, the MWC system can be used as a preliminary application of ISAC technology. However, the reconstruction effect of the conventional MWC system under the influence of noise is not stable. Therefore, we propose a signal processing optimization scheme for the MWC system based on spatial information entropy. First, the subsequent reconstruction algorithm is considered to require the dynamic and flexible processing of the sampled signals to reduce the influence of noise. Second, for the shortcomings of the original Orthogonal Matching Pursuit (OMP) algorithm, the concept of the genetic algorithm is used to optimize the algorithm by constructing the feature factor through spatial information gain and spatial information features. According to the simulation results, compared with the traditional MWC system, the scheme proposed in this paper is improved in all indicators.

Research on wireless monitoring system and algorithms for preload force utilizing machine learning and electromechanical impedance

The root mean square deviation is a common way to measure the electromechanical impedance of piezoceramic transducers that are used for traditional preload monitoring. However, due to the impedance signal’s high sensitivity to temperature, most current research is carried out under the same temperature conditions to avoid its effects. Even so, it is impossible to ignore the temperature factor in practical engineering, which hinders the application of impedance technology. Therefore, this paper proposes a novel approach for preload monitoring in actual engineering: using a wireless impedance signal acquisition platform to collect long-distance impedance signals and then establishing a predicted model based on machine learning (ML) considering the temperature. The wireless impedance signal acquisition platform includes a smart washer, a piezoelectric impedance wireless sensing device, and the host computer software. This test established a dataset under various operating conditions and developed a machine-learning-based predictive model. After comparing five typical ML algorithms, it was discovered that XGBoost performed better in prediction accuracy and generalization ability. Moreover, the Shapley additive explanation method is utilized to further analyze and interpret the XGBoost model. It indicates that ML primarily relies on numerical features (such as the area of each subinterval) to identify the impedance signal and predict the prestress, whereas information features (such as temperature value, peak, etc.) have little influence on the model’s output. Finally, the results above demonstrate that the ML-based models can predict the preload at different temperatures, effectively reducing temperature interference.

Attack detection model for BCoT based on contrastive variational autoencoder and metric learning

With development of blockchain technology, clouding computing and Internet of Things (IoT), blockchain and cloud of things (BCoT) has become development tendency. But the security has become the most development hinder of BCoT. Attack detection model is a crucial part of attack revelation mechanism for BCoT. As a consequence, attack detection model has received more concerned. Due to the great diversity and variation of network attacks aiming to BCoT, tradition attack detection models are not suitable for BCoT. In this paper, we propose a novel attack detection model for BCoT, denoted as cVAE-DML. The novel model is based on contrastive variational autoencoder (cVAE) and deep metric learning (DML). By training the cVAE, the proposed model generates private features for attack traffic information as well as shared features between attack traffic information and normal traffic information. Based on those generated features, the proposed model can generate representative new samples to balance the training dataset. At last, the decoder of cVAE is connected to the deep metric learning network to detect attack aiming to BCoT. The efficiency of cVAE-DML is verified using the CIC-IDS 2017 dataset and CSE-CIC-IDS 2018 dataset. The results show that cVAE-DML can improve attack detection efficiency even under the condition of unbalanced samples.

A General Image Super-Resolution Reconstruction Technique for Walnut Object Detection Model

Object detection models are commonly used in yield estimation processes in intelligent walnut production. The accuracy of these models in capturing walnut features largely depends on the quality of the input images. Without changing the existing image acquisition devices, this study proposes a super-resolution reconstruction module for drone-acquired walnut images, named Walnut-SR, to enhance the detailed features of walnut fruits in images, thereby improving the detection accuracy of the object detection model. In Walnut-SR, a deep feature extraction backbone network called MDAARB (multilevel depth adaptive attention residual block) is designed to capture multiscale information through multilevel channel connections. Additionally, Walnut-SR incorporates an RRDB (residual-in-residual dense block) branch, enabling the module to focus on important feature information and reconstruct images with rich details. Finally, the CBAM (convolutional block attention module) attention mechanism is integrated into the shallow feature extraction residual branch to mitigate noise in shallow features. In 2× and 4× reconstruction experiments, objective evaluation results show that the PSNR and SSIM for 2× and 4× reconstruction reached 24.66 dB and 0.8031, and 19.26 dB and 0.4991, respectively. Subjective evaluation results indicate that Walnut-SR can reconstruct images with richer detail information and clearer texture features. Comparative experimental results of the integrated Walnut-SR module show significant improvements in mAP50 and mAP50:95 for object detection models compared to detection results using the original low-resolution images.

Research on the Application of Deep Learning in Human Spinal Image Segmentation

Abstract Traditional segmentation methods can only segment grayscale images, which limits their application; The segmentation process often depends on the doctor’s experience, which can lead to subjective factors affecting the results; Therefore, the accuracy and efficiency of segmentation are difficult to achieve practical application results. The deep learning model is a structural model that mimics the neural connections within the human brain. The deep learning model can accurately extract multi-level features of key information in images from low-level to high-level, and provide feedback on data interpretation, thereby achieving accurate and efficient image segmentation results. Introducing deep learning algorithms into medical image segmentation can accurately express the key information at a deeper level in spinal images, achieving better image segmentation results.

The rapid improvement method of fatigue life reliability for single-crystal aeroengine turbine blades: Casting orientation design technology

The uncertainty in the casting orientation of single-crystal aeroengine turbine blades is an intractable engineering problem, and breaking through the crystal orientation design technology is vitally important to enhance the fatigue lifetime reliability of the blade. In this paper, a three-dimensional decoupling model for casting orientation is established according to the crystallographic characteristics of single-crystal blade, and the distribution information and statistical features of the orientation deviation angle of single-crystal turbine blades are obtained for the first time according to this decoupling model. Subsequently, the orientation-dependent models and generalized regression neural networks (GRNNs) analysis system of reliability are established, and the analysis of fatigue life response for single-crystal blades in four quadrants is carried out based on the neural network analysis system. Finally, three orientation design schemes are developed, and the improvement effect of each scheme on lifetime reliability are investigated. For the secondary orientation design scheme, the lifetime reliability is improved from 62 % to more than 95 % at the design life of 15,000 cycles; For the integrated orientation design scheme, the life reliability is improved from 50 % to more than 95 % at the design life of 20,000 cycles. The results show that the orientation design can effectively improve the fatigue life reliability of the product, especially in the long-life design interval of 10,000 to 20,000 cycles. Therefore, the orientation design should be highly emphasized in the design of single-crystal turbine blades for aeroengine.

Construction and validation of a prediction model for inguinal lymph node metastasis of penile malignancy.

Penile squamous cell carcinoma is a relatively rare malignancy among male malignancies, there are more than 30,000 new cases and more than 10,000 deaths of penile cancer annually. In patients with penile malignancy, inguinal lymph node metastasis (ILNM) significantly reduces patient survival. Thus, we identified the risk factors for ILNM in penile malignancies, aiming to develop a precise prediction model. We retrospectively analyzed 112 male patients with penile cancer. All subjects underwent penile surgery and inguinal lymphadenectomy at the same time, and postoperative pathology confirmed ILNM. Fisher's exact test, t-test, and Wilcoxon rank sum test were used to assess differences in demographic information and clinical features between the two groups, followed by logical least absolute shrinkage and selection operator (LASSO) regression analysis to determine risk factors of ILNM. The prediction model was constructed using nomogram. LASSO regression revealed that age [β=-0.005, odds ratio (OR) =0.995], smoking history (β=-0.006, OR =0.994) and interleukin 2 (IL-2) level (β=-0.0112, OR =0.989) were protective against ILNM. However, lymph node diameter (β=0.3117, OR =1.366), T-stage (β=0.1254, OR =1.134), fibrinogen (β=0.0377, OR =1.038), IL-4 level (β=0.004, OR =1.001), and neutrophil-to-lymphocyte ratio (β=0.0355, OR =1.034) were risk factors for developing ILNM. When assessing the risk of metastasis, it is crucial to balance these factors. The aforementioned characteristics were utilized to establish the predictive model, which demonstrated a good predictive ability with an area under the curve (AUC) value of 0.81. Moreover, internal leave-one-way cross-validation was used to construct a nomogram showing consistency, with an AUC of 0.75. The diagnosis of ILNM in penile malignant tumors can be predicted through clinicopathological features, biochemical tests, and prediction models based on tumor markers.

Deep-learning radiomics based on ultrasound can objectively evaluate thyroid nodules and assist in improving the diagnostic level of ultrasound physicians.

The incidence rate of thyroid nodules has reached 65%, but only 5-15% of these modules are malignant. Therefore, accurately determining the benign and malignant nature of thyroid nodules can prevent unnecessary treatment. We aimed to develop a deep-learning (DL) radiomics model based on ultrasound (US), explore its diagnostic efficacy for benign and malignant thyroid nodules, and verify whether it improved the diagnostic level of physicians. We retrospectively included 1,076 thyroid nodules from 817 patients at three institutions. The radiomics and DL features of the US images were extracted and used to construct radiomics signature (Rad_sig) and deep-learning signature (DL_sig). A Pearson correlation analysis and least absolute shrinkage and selection operator (LASSO) regression analysis were used for feature selection. Clinical US semantic signature (C_US_sig) was constructed based on clinical information and US semantic features. Next, a combined model was constructed based on the above three signatures in the form of a nomogram. The model was constructed using a development set (institution 1: 719 nodules), and the model was evaluated using two external validation sets (institution 2: 74 nodules, and institution 3: 283 nodules). The performance of the model was assessed using decision curve analysis (DCA) and calibration curves. Furthermore, the C_US_sigs of junior physicians, senior physicians, and expers were constructed. The DL radiomics model was used to assist the physicians with different levels of experience in the interpretation of thyroid nodules. In the development and validation sets, the combined model showed the highest performance, with areas under the curve (AUCs) of 0.947, 0.917, and 0.929, respectively. The DCA results showed that the comprehensive nomogram had the best clinical utility. The calibration curves indicated good calibration for all models. The AUCs for distinguishing between benign and malignant thyroid nodules by junior physicians, senior physicians, and experts were 0.714-0.752, 0.740-0.824, and 0.891-0.908, respectively; however, with the assistance of DL radiomics, the AUCs reached 0.858-0.923, 0.888-0.944, and 0.912-0.919, respectively. The nomogram based on DL radiomics had high diagnostic efficacy for thyroid nodules, and DL radiomics could assist physicians with different levels of experience to improve their diagnostic level.

An intelligent analysis method for power flow of bulk power grid based on fusion of indicator system feature and image feature

Abstract At present, the adjustment of power flow non-convergence in large power grid is still mostly based on manual trial and error, which relies heavily on expert experience and is inefficient. Therefore, an intelligent power flow analysis method based on index system and image feature fusion is proposed. Firstly, the power flow index system is constructed according to the actual demand, which can provide data support for the following power flow convergence judgment problems. Secondly, taking into account the constraints of similarity and ternary loss, the image feature information is used to guide the output of the information feature representation and to realize the organic interaction of the two modal features. On this basis, the calculation result of the index is correlated with the actual state of power flow to judge the convergence of power flow. Finally, the proposed method is verified on the CEPRI36 nodes system and a large power grid in China. The simulation results show that the accuracy of power flow convergence judgment is 100% for CEPRI36 nodes system and 99.83% for some domestic large power grid. It can effectively predict the convergence of the power flow before the calculation of the power flow required by the project and can provide the basis and theoretical support for the power flow optimization adjustment based on the indicator data.