Abstract：To facilitate the participation of large-scale, flexible air conditioning loads in demand response programs, enabling "source-load interaction" and ensuring the safe and economical operation of the power grid, various research institutions conducted real-time control simulation and practical studies on building air conditioning demand response. However, accurately estimating and forecasting air conditioning load remains a significant challenge. Current mainstream approaches include model-driven and data-driven methodologies. The model-driven approach relies on air conditioning load modeling, which struggles to capture the complex variations of the load. On the other hand, the data-driven approach depends on extensive data for model training, but often fails to account for the diverse characteristics of air conditioning loads. Therefore, this paper aims to integrate both model-driven and data-driven approaches to intelligently fit air conditioning loads, thereby improving the accuracy and adaptability of air conditioning load forecasting and generation. The paper proposes a load generation method and model AirGAN that combines mechanism models with Generative Adversarial Networks (GANs). This method continuously adjusts the hyperparameters of the physical model to better match the actual air conditioning load characteristics using virtual data generated by the GAN generator. Additionally, the GAN discriminator is employed to evaluate the load predicted by the mechanism model, thereby training the mechanism model to enhance its prediction accuracy.

Abstract：Electrical fires in high-rise buildings are difficult to predict and can cause significant damage. To address this issue, this paper proposes a multi-modal data fusion model for early detection of electrical fires in high-rise buildings. The model integrates data from three different types of sensors: temperature, CO gas concentration, and smoke, leveraging the complementary advantages of each modality. Initially, the gated Multi-Layer Perceptron(gMLP) is used to capture the intrinsic patterns of the three modalities' data, facilitating feature extraction. Subsequently, a fusion method based on multi-head attention is employed to merge the effective information from different modalities, achieving feature fusion and identifying electrical facilities with potential fire hazards. Experiments conducted on a multi-modal dataset under scenarios of no hazard and various electrical facilities with potential fire hazards demonstrate that the multi-modal data fusion model achieves high accuracy in early prediction, highlighting the superiority of the fusion approach.

Abstract：The dynamic aggregation of commercial load clusters is crucial for enhancing the flexibility of power grid dispatch, optimizing demand-side management, and promoting the integration of renewable energy. This paper selects commercial load features using Canonical Correlation Analysis (CCA) and employs DBSCAN and K-means clustering algorithms to classify loads, forming load clusters suitable for different scenarios. Furthermore, three load aggregation criteria are proposed, namely, regulation speed-based, load stability-based, and economic-based standards. The characteristics, applicability, and potential applications of commercial load aggregation under different standards in power dispatch are analyzed.

Abstract：To address the issue of poor recognition performance of existing named entity recognition methods in texts from fields such as electric power safety regulations, this paper introduces a method for named entity recognition in power safety based on machine reading comprehension. Firstly, a pre-trained model is used to encode the text to be recognized to obtain the vector representation of the text. Secondly, a hierarchical attention mechanism is utilized to capture the hierarchical relationships among nested entities and re-allocate the attention weights of the text sequence. On this basis, a classifier is employed to predict the entity scope in the text, and the final entity recognition results are obtained. The method is validated on the ACE 2005 and OntoNotes 4.0 public datasets, achieving optimal recognition performance compared to mainstream approaches. In the context of entity recognition for power safety scenarios, the method attains an accuracy rate of 89.3%, enabling precise identification of named entities in the power safety domain.

Abstract：Aiming at the problem of high retrieval difficulty caused by the complexity and diversity of power system data, this paper studies an adaptive retrieval method for large fluctuation data entries in the power system. Based on the rate of change in power system output, a two-component one-dimensional mixture Gaussian model is selected to construct a probability distribution model for power system fluctuations. Compare the power system fluctuation data simulated by the probability distribution model with the measurement data, identify the large fluctuation data entries in the power system based on the judgment threshold, and construct a data entry retrieval library. Use hash functions to obtain hash features of large fluctuation data entries in the retrieval database and generate binary codes. When retrieving large fluctuation data entries, generate binary codes for user search terms, calculate the Hamming distance between the binary codes of search terms and the binary codes of entries in the search database, and weight them. Use the weighted Hamming distance to sort the data entries and obtain adaptive search results for large fluctuation data entries. The experimental results show that this method can adaptively retrieve large fluctuation data entries in the power system based on user input search terms. The normalized cumulative loss gain of the retrieval results is higher than 0.9, and the retrieval time is less than 500 ms.

Abstract：Sulfur Hexafluoride(SF6) circuit breaker is an important equipment to ensure the stable operation of the power grid, but it is prone to air leakage problems in long-term use, which not only affects the performance of the equipment, but also threatens the safety of the power grid. In order to accurately diagnose the air leakage failure of SF6 circuit breaker, this paper proposes a CNN-BiLSTM-Attention combination model based on Gini index feature selection and Bayesian optimization (BO). First, feature mapping and importance analysis are conducted for internal and external factors affecting SF6 circuit breaker air leakage, and KMeans-SMOTE technology is used to solve the problem of uneven data distribution. Second, the Gini index-based method is used to screen the key features and optimize the hyperparameters of the fine-tuning CNN-BiLSTM-Attention model by Bayesian to improve the classification performance. The experimental results show that the equipment defects, operation life, operation and maintenance level, weather and temperature are the main factors leading to air leakage. Compared with other models, the proposed method shows higher classification accuracy and robustness in the 0/1 classification task of air leakage failure. The study not only verifies the effectiveness of the method, but also reveals the key factors that cause the leakage of SF6 circuit breaker, which provides scientific support for equipment inspection and operation and maintenance management, and further improves the security and reliability of power grid operation.

Abstract：Complex landscape images contain various objects with different characteristics, and traditional style transfer methods are unable to perform local style transfer on different objects within the same image. CycleGAN can achieve style transfer without paired samples through a pseudo-supervised strategy. However, CycleGAN fails to transfer styles between different categories of objects in complex landscape images; moreover, it lacks generalization ability in complex scenes and has high complexity. Therefore, this paper proposes a method for generating complex landscape multi-state images based on semantic categories, namely Semantic Category Style Transfer (SCST), which effectively combines local features for the generation of complex landscape images. Additionally, this paper introduces a context-aware style transfer model called GCycleGAN. Experimental results show that the performance of the proposed GCycleGAN is superior to that of deep learning-based image generation models such as CycleGAN, DualGAN, and Munit.

Abstract：In order to solve the problem of data redundancy conflict and lack of association, a knowledge graph visualization fusion method for multi-source heterogeneous data is studied to improve the reliability of data fusion. The domain ontology database and global ontology database corresponding to multi-source heterogeneous data are established by using Web Ontdogy Languge(OWL), so that knowledge entity extraction and knowledge fusion are carried out under the same framework. Based on the Long Short-Term Memory network(LSTM) and Conditional Random Field(CRF) model, knowledge entities conforming to domain definition are extracted from heterogeneous data from multiple sources under the constraint of ontology library. The knowledge fusion model based on hierarchical filtering is used to visualize the extracted knowledge entities, solve the redundant information and inconsistency problems in multi-source heterogeneous data, and form an accurate, complete and reliable multi-source heterogeneous data visualization fusion knowledge graph, which helps to find potential data associations and complete the missing data associations. The experimental results show that with the increase of the proportion of missing data, the scaling coefficient and attribute coverage begin to decrease, and the lowest scaling coefficient and attribute coverage are 0.86 and 0.87, which are significantly higher than the corresponding thresholds. When dealing with four data sources, the visual clarity of the proposed method is 93%~97%, and the information fusion is 92%~96%, which are better than the comparison methods. It shows that the method can effectively extract the knowledge entities of multi-source heterogeneous data, establish the knowledge graph, and realize the visualization fusion of multi-source

Abstract：The Generalized Matern Hard-Core Point Process (GMHCPP) is a repulsive point process that extends the hard core point process by introducing competition and thinning functions to constrain the retention probability. In this paper, we conduct an in-depth analysis of the performance of macro base stations and micro base stations within a two-layer heterogeneous cellular network model. We derive approximate expressions for the coverage probability and mean achievable throughput of this model. First, based on the performance differences and repulsive characteristics of the macro base stations, a retention probability function is derived. Subsequently, the deployment of macro base stations is modeled using GMHCPP. Then, Monte Carlo simulations are employed to simulate the base station distribution, and the relationship between the average interference-to-signal ratio and the path loss factor is plotted. From this function, the gain factor expression of GMHCPP is derived by fitting with the Poisson point process(PPP) approximation. Finally, the gain factor is incorporated into the coverage probability and mean achievable throughput formulas to simulate and analyze the performance metrics of the two-layer heterogeneous cellular network model.

Abstract：The loop lock time of the digital receiver with feedback structure timing recovery is uncertain, which cannot meet the demand for fast synchronization in burst communication. In order to obtain timing error rapidly, taking into account the complexity and resources of implementation, a timing recovery method based on 2 times sampling feedforward structure is proposed. Firstly, according to the correlation peak detected by the synchronization segment, the timing error is obtained by using parabolic estimation. Then, the sampled data is reconstructed through four points piecewise parabolic interpolator to recover the data of timing synchronization. Finally, the method is implemented and verified by field programmable gate array(FPGA) hardware platform. The experimental results show that the proposed method has less computation and low implementation complexity. When the bit error rate is 10-5, the simulation performance loss is 0.05 dB and the realization performance loss is less than 0.2dB compared with ideal timing.

Abstract：In view of the current situation of weak rural tourism infrastructure, limited publicity means and single function of existing rural roaming systems, this paper takes Qingshandian Village in Lin'an District as a demonstration, uses Blender 3D modeling technology, and combines Unity3D platform to create immersive scene interaction design. Through LOD optimization, partition loading and frustum culling strategy, a highly immersive and interactive virtual rural roaming system is developed. The system is not only compatible with the web platform, but also supports VR bicycle roaming, making the user experience more real and convenient, aiming to promote the dissemination of rural culture and promote the sustainable development and utilization of tourism resources.

Abstract：In order to improve the accuracy and efficiency of invoice detection, a CenterNet based invoice detection algorithm is proposed. Firstly, the algorithm model adopts a backbone network similar to CSPDarkNet, introducing Triplet Attention into the CSP structure to form a TA-CSP structure, and introducing ASPP at the end of the backbone network to improve the receptive field range of the network, enabling the model to better understand the contextual information of the image; Secondly, in the Neck part of the network, CBAM is used to guide the fusion of high-level and low-level features, and the semantic information in high-level feature maps is used to supervise low-level feature maps to suppress background noise in low-level feature maps; Thirdly, in the Head section of the network, the algorithm adds four channels of feature map outputs based on the CenterNet network, achieving invoice orientation prediction while detecting invoices; Finally, an orientation loss term is added to the loss function to optimize the orientation of invoices. The experimental results on the test dataset show that the mAP of the proposed algorithm in this paper is superior to CenterNet and YOLOv5s algorithms reaching 84.3%, effectively improving the accuracy and robustness of invoice detection.

Abstract：This paper proposes a two-layer optimization algorithm based on cloud-edge collaborative energy management, aiming to optimize the scheduling of power grid operators and user aggregators. The algorithm enhances the computational capabilities of user aggregators through edge computing, reducing the operational costs and computational complexity of the energy system. A heuristic algorithm based on the modified optimal value function is used to solve the bilinear optimization problem. Experiments demonstrate that the proposed two-layer optimization scheduling framework and its closed-loop control model effectively improve the collaborative efficiency between power grid operators and user aggregators, providing technical support and a theoretical basis for the efficient operation of energy management systems.

Abstract：According to the communication requirements of multi-vehicle cooperative system, a design method of transceiver-integrated RF device with high isolation, wide dynamic and fast response is provided. This paper discusses the architectural of the transceiver-integrated RF device, and the key indexes as AGC circuit, high isolation, and linearity are designed from the engineering application. A high isolation transceiver-integrated 4×2 RF switch matrix is designed. Through the field test of the system, the test results show that the transceiver isolation index is greater than 99 dB, and the receiving dynamic range is greater than 65 dB. The RF indexes are stable and reliable to meet the needs of multi-vehicle system requirements, which provides a reference method for the further study of the multi-vehicle system data exchange equipment.

Abstract：A dual-band dual-polarized antenna with 2.4 GHz and 5.8 GHz bands is designed, which can be applied to synchronous dual-channel wireless communication. Firstly, the resonance characteristics of the radiating structure without U-shaped slots are analyzed using the characteristic mode theory, revealing the existence of two distinct resonance modes. Then, U-shaped slots are introduced into the radiating structure, resulting in two resonance modes in the high-frequency band with equal amplitudes and a 90° phase difference, thereby achieving circular polarization. Finally, coupling feeding is introduced at the locations of maximum modal currents to realize the design of the dual-band dual-polarized antenna, which is then experimentally verified. The measurement results demonstrate that the antenna achieves omnidirectional linear polarization in the low-frequency band with an impedance bandwidth of 2.1% (2.38 GHz~2.43 GHz), and directional circular polarization in the high-frequency band with an impedance bandwidth of 22.8% (4.90 GHz~6.16 GHz) and an axial ratio bandwidth of 2.4% (5.75 GHz~5.89 GHz). The measured results are in good agreement with the simulation results, validating the effectiveness of the design.

Abstract：With the popularity of digital audio sources, Digital to Analog Converter (DAC) has become an indispensable component in audio devices, and its accuracy often determines the signal fidelity of the entire system. For this reason, a sigma-delta modulator for high-precision audio DACs is designed using noise shaping technology and implemented by Field-Programmable Gate Array (FPGA) in this paper. By building a testing system, the test results show that the output Signal to Noise Ratio(SNR)of the designed sigma-delta modulator can achieve 107.4 dB when a 1 kHz, 0 dBFS (Full Scale) sinusoidal signal with 1 411.2 kHz sampling frequency is used. When the input signal frequency is within the audio band (with an input signal amplitude of 0 dBFS), the output SNR remains above 104 dB. Additionally, the designed sigma-delta modulator can be utilized in WAV music players.

Abstract：Due to a variety of imperceptible design defects，high-power RF amplifiers are prone to transient pulse, which has a significant impact on the reliability and service life of the hardware，and seriously interfere with the quality of work. Aiming at the transient pulse damage problems of high-power RF amplifiers，this paper focuses on transceiver timing control，frequency control, ALC response. The influence of transceiver timing control，frequency control and ALC response on the reliability and service life of high-power RF amplifiers is discussed. The design principles of high-power RF amplifiers are put forward. Typical applications and verifications show that the correct and reliable transceiver timing control，frequency control and ALC response can effectively solve the problem of transient pulse damage and improve the reliability and service life of the high-power RF amplifiers.

Abstract：Acoustic excitation antenna technology enables a significant reduction in antenna size, facilitating miniaturization of ground penetrating radar systems in the very high-frequency (VHF) band. However, due to the narrow bandwidth of current acoustic excitation antennas, the range resolution is low. To address this issue, this paper proposes a three-dimensional underground target localization method based on circular array direction-of-arrival (DOA) estimation combined with spatial continuous sampling, which does not rely on range resolution. By leveraging the small size advantage of acoustic excitation antennas, a uniform circular array is constructed. The target's azimuth and elevation angles are obtained using two-dimensional DOA estimation. Based on the variation pattern of two-dimensional angles in spatial continuous sampling, the genetic algorithm is employed to estimate the three-dimensional coordinates of underground targets. Numerical simulation results show that the root mean square error of the three-dimensional localization coordinates is 0.039 m, outperforming traditional broadband antenna systems using hyperbolic vertex detection, with good robustness. The electromagnetic simulation results show that this method has the capability of 3D location of underground target

Abstract：To address the difficulties in fault analysis and localization caused by the strong coupling of fault phenomena in complex fault scenarios of airborne radar, this paper proposes a composite fault log design method based on Non-Volatile Random Access Memory (NVRAM) fault logs. This method integrates various data, such as Built-in-Test（BIT）information and status flags, to optimize the fault localization process. Taking a typical fault of an airborne radar as an example, after adopting the composite fault log designed with this method, the fault can be quickly located, and the root cause of the fault can be revealed, along with the identification of the fault mechanism. The effectiveness of the method has been validated. Through fault analysis and localization using the composite fault log, this method effectively solves the uncertainty problem in fault localization of airborne radar products and handles the multiple fault phenomena that occur during fault events, thereby significantly improving the efficiency of fault localization.

Abstract：In the background of artificial intelligence, integration technology of radar, communication and electronic countermeasures was taken researched. The concept of integrated technology was first combined and summarized, then implementation meaning was discussed in the artificial intelligence background, and the research status was summarized and analyzed from two aspects of domestic and abroad countries. The key technologies such as system mechanism, signal waveform and intelligent decision in the implementation of integrated technology were expounded. The main points of integration technology were summarized and discussed. Finally, the foreground of integration technology was prospected.