In order to quantify the uncertainty of multi-energy flow distribution in the integrated energy system, a probabilistic multi-energy flow calculation method of integrated energy system based on data-driven is proposed. Firstly, a unified multi-energy flow calculation model suitable for different working modes of compressors in integrated energy system is established, and the impact of different operating modes of compressors on the multi-energy flow distribution is also discussed. Secondly, a probabilistic multi-energy flow calculation method based on support vector regression is developed. The method first constructs a data set by calculating deterministic multi-energy flow repeatedly, and then the support vector regression is used to mine the nonlinear mapping relationship between known loads, network node information and unknown node parameters in the integrated energy system. Finally, through case analysis, it is verified that the proposed unified multi-energy flow model can be applied to different compressor working conditions. By comparing with traditional probabilistic multi-energy flow calculation methods, it is shown that the proposed data-driven probabilistic multi-energy flow calculation method has higher computational accuracy and efficiency.

The ever-increasing functionalities and escalating complexity of existing Advanced Driver Assistance Systems inevitably cause the problem of Safety of The Intended Functionality. The identification and generation of trigger conditions play a critical role in SOTIF activities. Most existing trigger condition identification approachesare mainly based on the System-Theoretic Process Analysis method, which however neglect the issues within the system's functional state transitions. This paper adopts a knowledge-driven approach to construct a trigger condition identification mechanism by integrating STPA and Finite State Machine theories to establish an expanded system control structure. Safety analysis is conducted concerning the expanded control architecture and functional state transitions. By considering system limitations and human misuse, trigger conditions are identified, generated, described, classified, and labeled. Finally, the proposed trigger condition generation mechanism is applied to an Integrated Cruise Assistance system, obtaining trigger conditions and their classifications. The generated mechanism is compared with existing trigger condition generation methods, demonstrating its practicality, feasibility, and effectiveness.

The deep learning technology has significantly improved the speed and accuracy of compressed sensing reconstruction. However, the existing deep compressive sensing algorithms usually use the same sampling rate to process different blocks of an image, ignoring the fact that different image blocks have different reconstruction difficulties. In this paper, a compressive sensing algorithm with adaptive sampling and memory enhancement is proposed. Firstly, the reconstruction difficulty of different blocks is estimated based on the reconstruction error of the measurement domain. Then, the rules are designed to adaptively assign the sampling rates, and the sampling matrix is used to sample each image block at a specific sampling rate with the help of a sampling rate mask. Furthermore, the two-branch aggregation module is added to the reconstruction network to enhance the interaction of context memory, and the reconstruction ability of the network is improved by adjusting the channel weight of different memory branches. The experimental results show that the proposed algorithm increases the average SSIM by approximately 0.0269 and the average PSNR by approximately 1.66 dB over other methods on several common datasets.

A lightweight unmanned aerial vehicle (UAV) aerial photography small target detection algorithm GA-YOLO based on YOLOv5 is proposed to address the problem of small target feature size, complex background, and dense distribution in images from the perspective of UAV aerial photography. This algorithm improves the Mosaic data augmentation method and overall network structure, and adds a small object detection head. At the same time, a lightweight global attention module and a parallel spatial channel attention mechanism module are designed to enhance the network's global feature extraction ability and the competition and cooperation between convolutional channels during the training process. Based on the 4.0 version of YOLOv5s, experiments were conducted on the publicly available drone aerial photography dataset VisDrone2019-DET. The results showed that the improved model reduced the number of parameters by 48% and the computational complexity by 26% compared to the original model, and mAP@0.5 improved by 4.9 percentage points, mAP@0.5 0.95 increased by 3.3 percentage points, effectively enhancing the detection capability of unmanned aerial vehicles for dense small targets from an aerial perspective.

This systematic review provides a comprehensive analysis of magnetically actuated capsule robots, focusing on their design, control mechanisms, and diverse applications. Through a systematic literature search and analysis, this review aims to provide valuable insights for researchers, engineers, and practitioners involved in the development and utilization of capsule robots in various domains. The design principles of magnetically actuated capsule robots are summarized in terms of size, shape, material and motion mechanism, the precise control methods and positioning navigation strategies analyzed, and the application of magnetically actuated capsule robots in gastrointestinal disease examination discussed, targeting drug delivery and minimally invasive surgery. According to the scope and objectives of the systematic review, the technical challenges and potential research directions in this field are summarized. Through continuous research and innovation on issues such as positioning accuracy, control strategies, and material properties, the magnetically actuated capsule robot system will bring major breakthroughs and advancements in the fields of medical treatment and biological research.

Carbon sink fisheries are an important means of producing blue carbon and contributing to carbon neutrality, with significant economic and environmental benefits. However, the implementation of carbon sink fisheries involves multiple parties and is a dynamic game. Existing studies have rarely analyzed carbon sink fisheries as an entry point to include governments, aquaculturists, and enterprises in the system at the same time. Against the background of the success of the pilot projects in Lianjiang and Putian, exploring blue carbon trading based on the perspective of carbon sink fisheries will help to provide scientific recommendations. Therefore, a tripartite evolutionary game model involving local governments, aquaculturists, and enterprises was constructed, and the impact of each subject's strategy choice on the stability of the system was analyzed. The results show that (1) subsidies for carbon sink fisheries have a dual effect on the evolutionary process. Within a certain range, moderate subsidies for carbon sink fisheries can motivate aquaculturists to shift to carbon sink fisheries, but too high subsidies level can cause a heavy financial burden and force the government to give up. (2) The difference between the price of subsidized blue carbon and the carbon tax is a key factor affecting the strategy choice of enterprises. When the price of subsidized blue carbon is lower than the carbon tax, enterprises tend to purchase blue carbon, and the lower the price of subsidized blue carbon, the faster the response of enterprises. (3) Based on the initial scenario, a moderate increase in the subsidy level of the local government can lead the three-party evolutionary game to the ideal outcome. Under this subsidy level, no matter what the initial probability of each party is, it will not affect the realization and continuation of the ideal outcome.

Laser selective quantitative removal of CFRP represents a key technology for structural repair in the aerospace industry. However, the huge difference in properties between carbon fiber and epoxy resin in CFRP makes laser processing very challenging, and thermal damage has always been the main obstacle to the widespread application of CFRP laser processing. In order to study the influence of femtosecond laser processing parameters on the machining quality of CFRP, theoretical analysis and experimental verification were carried out by single factor experimental method. The influence of laser energy density, scanning speed and scanning direction on the ablation rate and the width of heat affected zone were analyzed, and the high-precision selective quantitative removal process of femtosecond laser on CFRP was investigated. The results show that when the process parameters are chosen to be θ = 90°, pulse width of 290 fs, power of 7 W, frequency of 100 kHz, scanning speed of 300 mm/s, and scanning spacing of 60～80 μm, the overall ablation surface quality is superior, and the accuracy (roughness) can be as high as 10 μm, and the heat-affected zone (HAZ) on the surface of the removed area is about 33.9 μm.

The Ag@DMSNs composite was prepared by synthesizing the dendritic mesoporous silica nanoparticles (DMSNs) and subsequently loading the nanostructured Ag in the pore channels of the DMSNs via chemical reduction. Thus-obtained Ag@DMSNs feature an intensive and board absorption for the solar irradiation due to the plasmonic coupling of Ag nanostructures, which are anchored in the pore channels of DMSNs and not prone to aggregation. More importantly, the thermal effect of plasmonic relaxation can efficiently convert solar energy into heat. For example, Ag@DMSNs can increase its surface temperature from 26 ℃ to 70 ℃ within 5 minutes under one sun (1 kW·m ^-2, 420~2500 nm) . When Ag@DMSNs are loaded on the porous polyurethane foam material, the water evaporation rate reaches 1.10 kg·m ^-2·h ^-1 under one sun, and they also exhibit excellent stability in simulated seawater. In addition, the thermal electrons produced during the relaxation of the Ag nanoparticle plasmon in the Ag@DMSNs complex can effectively remove contaminants from water, such as the degradation of methylene blue. These results show that it is an effective way to realize solar-powered clean water production by rational construction of the plasmonic coupling model and utilizing the thermal and hot-electron effect of plasmonic relaxation process, opening new avenues to tackling the deteriorating problem of fresh water scarcity.

The technology of adsorption/absorption atmospheric water harvesting is characterized by ease of operation, low energy consumption, and represents an effective measure to mitigate the scarcity of fresh water resources. A systematic review of recent research progress in this field is presented and an introduction is provided to the basic principles and main material categories of adsorption/absorption air water extraction technology, as well as a summary of the material placement structure, equipment operation mode, and application field of adsorption/absorption equipment. Adsorption/absorption materials include traditional porous structural materials, hygroscopic salts, and new organic air water intake materials. Currently, significant advancements have been made in enhancing the adsorption performance of materials, broadening their application scope, and reducing the energy consumption required for water absorption and release. Nevertheless, there are still challenges related to poor mechanical properties and incapability to be shaped into profiles for adsorption/absorption materials that necessitate further comprehensive investigation. The adsorption/absorption equipment can be utilized for air-water intake applications through various structural configurations, including flat placement, central axis placement, fold placement, etc. This equipment is capable of operating in either a daily cycle or multiple cycle mode. Apart from directly collecting liquid water, the adsorption/absorption atmospheric water harvesting equipment can also serve as an innovative solution for cultivating ecological farms, thereby offering potential strategies to mitigate regional freshwater resource scarcity and food shortages.

Soil cementation and solidification, based on microbially induced calcium carbonate precipitation (MICP) technology, has emerged as research hotspots in geotechnical engineering and geological engineering since the 21st century. In this research, the reinforcement mechanism of MICP technology and the research status of the engineering application and the reinforcement effect and application practice are systematically described and reviewed. The results show that the strength of the site after MICP solidification shows an obvious non-uniformity, and the distribution of calcium carbonate content decreases with depth. In desert environment, the calcium carbonate coating induced by in-situ extraction of bacteria demonstrates superior strength and stability to the traditional Bacillus pasteurelli. The application of new MICP technology, such as microbial cement and microbial brick, shows promising prospects in terms of strength and durability, and new vitality into the realization of China’s double carbon goal. Addressing the factors affecting the precipitation characteristics of calcium carbonate by MICP technology, the uniformity of calcium carbonate distribution at field scale is enhanced, the durability of calcium carbonate skeleton under seasonal changes ensured, and curing efficiency under different environments improved, which should all be prioritized in future research.

The performance of single image super-resolution reconstruction based on standard convolution is limited by the redundancy of the stacked network layers, making it difficult to implement the algorithm on the ground. Moreover, the single residual structure of the feature extraction layer cannot efficiently utilize the feature information obtained from convolution. To address these, this paper proposes a residual distillation reuse module based on the existing residual distillation-based structure to reduce the high-frequency information of the image lost in the residual distillation process. In addition, the base residual block is replaced by a blueprint separable convolution to decouple the spatial correlation of the feature map, such that the weight of highly correlated features can be reduced. As a result, the efficiency of convolution can be improved and the number of parameters can be reduced. We conduct comparative experiments on standard datasets such as Set5 to evaluate the proposed algorithm. The experimental results show that the peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) of the proposed algorithm can be improved by approximately 0.06~0.25 dB and 0.004~0.012, respectively, over the lightweightresidual distillation image super-resolution networks.

Tumor metastasis is an important link in the process of tumor development, and it is also one of the main reasons for cancer deterioration and treatment failure. Taking tumor metastasis as the background, a study is conducted on the generative model of tumor lymphatics based on the interaction between tumor and extracellular matrix (ECM). First, mathematical language is used to sort out the biological principles of tumor lymphangiogenesis, and then assumptions made and mathematical models established and qualitative analysis carried out. The proof of the uniqueness of the existence of local solutions of the model is mainly carried out by means of approximation methods, the qualitative theory of partial differential equations and Banach's immovable point theorem, as well as the uniqueness of the existence of the overall solution of the model with the help of the regularity estimate of the local solution and the embedding inequality. Finally, the difference numerical method is used to carry out numerical simulation to illustrate the reliability and accuracy of the model. This research is of great significance for in-depth understanding the mechanism of tumor metastasis, guiding cancer treatment, and promoting related research.

Identifying potential adverse reactions of drugs can help doctors make clinical medication decisions. In view of the high-dimensional sparse features of previous studies and low prediction accuracy in constructing an independent prediction model for each adverse reaction, a prediction model of adverse reactions based on knowledge graph embedding and deep learning is developed, which can uniformly predict the adverse reactions covered by the experiment. On the one hand, knowledge graph and its embedding technology can fuse the correlation information between drugs and alleviate the deficiency of high-dimensional sparse feature matrix. On the other hand, the efficient training ability of deep learning can improve the prediction accuracy. In the study, drug characteristic data is used to construct a knowledge graph of adverse drug reactions; by analyzing the embedding effect of different embedding strategies, the best embedding strategy is selected to obtain the sample vector. Then a convolutional neural networks model is constructed to predict adverse reactions. The results show that the convolutional neural networks model has the best prediction effect under the DistMult embedding model and the 400-dimensional embedding strategy. The mean values of accuracy, F ₁ score, recall and Area Under Curve were 0.887, 0.890, 0.913 and 0.957, respectively, which are better than those reported in the literature. The prediction model has good prediction accuracy and stability, which can provide an effective reference for safe medication.

Chronic obstructive pulmonary disease (COPD) is a common respiratory disease of the world, and the doctors need a lot of time to read the abdominal CT images for COPD pre-evaluation. To improve the pre-evaluation efficiency, a deep network based on slice correlation information was proposed for COPD auxiliary diagnosis. First, by using a grouping approach, the architecture of the deep network is divided into several network branches, each of which aims to extract the local intra-slice association information of the CT images. Then, the outputs from multiple network branches are integrated via a BiLSTM to extract the global inter-slice association information between the adjacent CT slices. To further improve the ability of local feature extraction for each network branch, the enhanced multi-headed convolutional attention is designed by embedding the ConvNeXt into the existing multi-headed convolutional attention. Experimental results show that the proposed deep network achieves promising effectiveness for CT image classification on auiliarily diagnose of COPD, and the accuracy, sensitivity, specificity and AUC of the proposed network reach to approximately 92.15%, 94.17%, 91.17% and 95.33%, respectively.

The advancement of high-speed and high-precision machining has led to a growing demand for cutting tools. Surface hard coatings substantially enhance the wear resistance and overall lifespan of cutting tools, thereby playing a pivotal role in the development of high-performance tools. This paper first introduces the significance of surface hard coatings in cutting operations, then reviews the research status of common hard coatings such as nitride, boride, and oxide, along with associated physical vapor deposition techniques. Finally, an analysis is conducted to identify the prevailing research and application challenges encountered in physical vapor deposition tool coatings.

Owing to advantages of high theoretical energy density, low cost and environmental friendliness, lithium-sulfur (Li-S) battery is considered as one of the most promising next-generation high-energy-density batteries. The "shuttle effect" of polysulfides is the key issue hindering the commercialization of Li-S batteries. Adoption of "catalytic" strategy to enhance the sulfur redox kinetics has been demonstrated to be an effective way to alleviate the "shuttle effect". Single-atom catalysts (SACs) have received much attention in the field of catalysis due to their uniform metal active centers, unique electronic properties, and theoretically 100% metal atom utilization. In recent years, SACs have been introduced into Li-S systems and studied to achieve fast sulfur conversion kinetics. In this research, the latest progress in the application of SACs in Li-S batteries was reviewed, with special emphasis on the discussion of key factors affecting the catalytic activity of SACs. The prospects of SACs for Li-S batteries were pointed out and highlighted. Important guidance is provided for future design and fabrication of high-performance SACs for Li-S battery application.

Emerging contaminants (ECs) are characterized by stable structures and low concentrations, making them difficult to remove completely using traditional wastewater treatment processes. ECs are posing potential risks to aquatic ecosystems and human health. Advanced oxidation processes (AOPs) can rapidly and effectively degrade persistent pollutants. However, for trace refractory ECs in real water matrices, AOPs require excessive oxidants or consume more energy, resulting in low cost-effectiveness of water treatment and even secondary pollution. Therefore, developing efficient and low-energy selective oxidation processes for treating trace ECs in water has practical significance. Targeted adsorption-transformation technology can effectively enhance the utilization of free radicals and efficiently remove trace ECs. The concept of advanced water purification processes is elaborated based on selective oxidation, with a focus on the technical characteristics and recent development of selective electrochemical adsorption-transformation technology to remove per- and polyfluoroalkyl substances (PFAS) from water. Finally, an outlook is provided on the future research directions and trends.

To address the scheduling problem of multi-department cooperative treatment of patients under the condition of limited medical resources in a hospital, this paper proposes a collaborative treatment scheduling algorithm based on intelligent optimization. The proposed algorithm regards the cooperative treatment scheduling of doctors, nurses and patients in different scenarios as a multi-role cooperative control problem. In order to optimize the role's access behavior, we propose a decision-making model to guide the role to make the optimal access behavior, and introduce an intelligent optimization algorithm to optimize the decision-making model. For the case scenarios of collaborative treatment of patients, doctor ward rounds, and patient physical examinations, we conduct experiment to compare four scheduling strategies, includingthe random, shortest distance, maximum free space, and decision-making models, and comparatively analyze the performance of the genetic algorithms, particle swarm optimization, simulated annealing, and differential evolution in optimizing the decision-making models. The experimental results demonstrates that the decision-making model based on the differential evolution algorithm performs the best, and the optimized decision-making model can find feasible solutions in the case scenarios and also obtain the optimal scheduling results.

Coarctation of aorta (CoA) is a congenital malformation of the aortic arch with a poor natural prognosis, which requires early intervention and even emergency surgery. Meanwhile, postoperative aortic re-coarctation is still a possible problem. At present, the prediction of aortic re-coarctation is mainly carried out based on the risk factor analysis of doctors on the clinical characteristics of patients combining with echocardiography (Ultra Sound Cardiogram) data, which is easy to be misdiagnosed. In this paper, a multimodal data detection framework based on Swin-Unet network is proposed based on the images of the patient's heart from computed tomography (CT) combining with the patient's clinical data. The framework carries out multimodal feature fusion analysis by combining the Swin-Unet network and the machine learning models, aiming to perform early detection of aortic re-coarctation. The experimental results on the clinical dataset show that our proposed methodeffectively improves the prediction effect of aortic re-coarctation when compared with the traditional prediction methods using clinical data. Particularly, we verifie the risk factors related to re-coarctation, the results of which provides a reference for clinical medicine.

Hepatobiliary stone disease is a common liver disease and has become the main cause of death from non-neoplastic biliary diseases in China, and it is important to achieve interpolated segmentation reconstruction between slices of hepatobiliary ducts. In this research, an end-to-end framework for super-resolution abdominal CT image processing is proposed based on deep correlation mechanism. The framework cascades an inter-slice interpolation network and a segmentation network, in which the ConvLSTM is introduced to enhance the extraction of high-dimensional feature information of hepatobiliary ducts between slices. A novel loss is designed by combining the loss of the interpolation network and the loss of the segmentation network. Experimental results show that the proposed framework is superior to the existing deep learning methods for the segmentation of hepatobiliary ducts, which is beneficial for the 3D reconstruction of hepatobiliary ducts.

Traffic flow prediction is an important problem in the field of intelligent transportation systems. Most existing traffic-flow prediction methods have made good progress, which however still face the following two key challenges. (1) The underlying pattern of traffic flow depends on not only the historical information along the timeline, but also the information of spatially adjacent areas, making it a challenging problem on how to balance the two temporal-spatial patterns; (2) Due to fact that time information has the characteristic of multiple granularity, such as hour, day and week, how to capture the multi-grained temporal patterns is another challenge problem. In this paper, we design a multi-grained spatio-temporal deep regression model (MGSTDR) to address the above challenges. By extending the typical autoregressive integrated moving average model (ARIMA) on the basis of multi-grained spatio-temporal traffic flow information, the proposed model can effectively use historical information along the timeline as well as the information of adjacent regions, such that the prediction of multi-grained spatio-temporal traffic flow can be performed. Experimental results on several datasets demonstrate that the proposed model outperforms existing benchmark methods on the task of multi-granularity, and particularly obtains an approximately 5.66% improvement in the hourly traffic flow prediction.

As a pivotal part of the task-oriented dialogue system, dialogue policy can be trained by using the discriminative deep Dyna-Q framework. However, the framework uses vanilla deep Q-network method in the direct reinforcement learning phase and adopts MLPs as the basic network of world model, which limits the efficiency and stability of the dialogue policy learning. In this paper, we purpose an improved discriminative deep Dyna-Q method for task-oriented dialogue policy learning. In the improved direct RL phase, we first employ a NoisyNet to improve the exploration method, and then combine the dual-stream architecture of Dueling Network, Double-Q Network and n-step bootstrapping to optimize the calculation of the Q values. Moreover, we design a soft-attention-based model to replace the MLPs in the world model. The experimental results show that our proposed method achieves better results than other baseline models in terms of task success rate, average dialog turns and average reward. We further validate the effectiveness of proposed method by conducting both ablation and robustness analysis.

To address the problems of missing detection and low detection accuracy of the existing helmet wearing detection algorithms for small and crowded targets detection, this paper proposes a helmet wearing detection method based on improved YOLOv5 and transfer learning. First, different from the default priori frame that is not suitable for the task, we use the K-means algorithm to cluster the suitable priori frame size for the detection task. Then, in the back of the feature extraction network, we introduce a spatial channel mixed attention module to strengthen the learning of relevant weights and suppress the weights of irrelevant backgrounds, respectively. Further, we improve the judgment metric of the non-maximum-suppression (NMS) algorithm in the post-processing stage of YOLOv5 to reduce the phenomenon of false deletion and missing of prediction boxes. After that, the proposed network is trained based on the strategy of transfer learning, which can overcome the scarcity of limited existing data sets and improve the generalization ability of the model. Finally, we build a cascade judgment framework for helmet wearing deployed in visual sensor networks. The experimental results show that our proposed method improves the average accuracy (IOU=0.5) to 93.6%, which is 5% higher than the original model in the helmet wearing data set. The proposed model also outperforms other state-of-the-art algorithms by obviously improving the accuracy of helmet wearing detection in the construction scenarios.

In view of the information asymmetry and opacity of the existing system for centralizing the regulation of grain reserves, which reduces the reliability of data, and at the same time, the independence of various business links of grain reserve supervision, which brings information islands, a modular, pluggable and configurable model of grain multi-chain reserve management is proposed based on the characteristics of blockchain technology. The model adds channel technology in Hyperledger Fabric to isolate public traceability data and private data. An authorization mechanism for enterprise users is set up. An intelligent contract for authorization information is designed to check and judge the data format in order to secure the chain and so on. Through the construction of the supervision model and the experimental test results, it is shown that the system can well guarantee the data upload, query and safe storage under the condition of complete functions. It also realizes the supervision of the whole process of the data on the chain, the efficient processing and traceability of the problem events. It can enhance the credibility of grain reserve supervision data and ensure the safety of grain reserves.

In order to meet the high-speed working requirements of the fifth-generation mobile communication, a high gain antenna for 5G millimeter wave based on SIW is proposed. Through the coupling between the rectangular patch and the resonant cavity, and arranging parasitic units on both sides of the main unit, the working bandwidth of the antenna ( S ₁₁≤−10 dB) reaches 10.7% (3.36~3.74 GHz) . The maximum gain in the working range reaches 7.78 dBi. Compared with other similar antenna, the proposed antenna can reduce 2/3 of the phase shifter and it uses Fr-4 material,which is of greater cost advantage.

Due to factors such as loss of coherence between repeated SAR acquisitions, the interferograms are often very noisy. The quadtree filter can remove such discrete noise in unwrapped interferograms and obtain clean deformation results. A mining area in Yunnan is investigated, discussing how the minimum division window and the gross error threshold affect the performance of quadtree filtering. The surface deformation time series of the mining area is obtained using Sentinel-1 satellite data from 2019 to 2021. The results show that the mining area has significant surface deformation during our monitoring period, focusing mainly on three regions with both uplift and subsidence. The deformation time series present wave-like variations in the time domain. The maximum incremental subsidence and uplift are up to 67.3 and 79.4 mm in 12 days, respectively. The difference is $ \pm 10.9\;\mathrm{m}\mathrm{m} $ between InSAR and 41 repeated observations from 32 ground-based total stations.

In this paper, a circumvention design method which taking patent group as target, function requirement as the basis for circumvention and the extension theory as the theoretical tool was proposed. Firstly, online product reviews were used as requirement acquisition resources, which were transformed into several functional requirements by using natural language processing technology and extension primitive model. Secondly, the target patent group was researched by the keywords from the online reviews and IPC classification code, then the patent group function model was built by analyzing each patent text. Thirdly, the obtained function requirements were matched with the patent group function model to confirm the contents in patent group function model to be circumvented. Fourthly, the function requirements were guided by the extension analysis as the path for circumvention analysis, and the extension transformation was used to express the circumvention solution operation, then the new scheme was determined whether exists the technical features of infringement. Finally, a medicine storage box patent group was taken as case study to confirm the feasibility of the above process of patent group circumvention design. Taking functional requirements as the basis for circumvention can help reduce the influence of subjective judgment in confirming the content to be circumvented in the process of patent group circumvention, and at the same time ensure that the subsequent circumvention solution is carried out in the direction of satisfying users’ function requirements.

Aiming at the problems of small defect targets and complex backgrounds, which leads to low accuracy in insulator burst defect detection, an improved detection algorithm YOLOv4-MP based on YOLOv4 is proposed. First, the Shuffle Attention module is included in the feature extraction network to limit the interference of complicated background, enabling the model to extract more effective feature information. Then, the spatial pyramid pooling is enhanced by the dilated pooling layer, which effectively increases the receptive field and improves the effect of feature fusion. Finally, the Mish function is utilized as the activation function of the path enhancement network to limit the loss of low-level information. The experimental results show that the mean average precision of YOLOv4-MP reaches 93.60%, which is 6.37% higher than that of the YOLOv4 algorithm. Compared with the commonly used detection algorithms, YOLOv4-MP has better detection performance and has high application value for the detection of insulator burst defects.

A new method is proposed for self-supervised monocular depth estimation. Although recent methods have been able to produce high-precision depth maps, previous work has ignored channel-wise information in the image. To solve this problem, Channel Attention is introduced and improvements made in two aspects of the network structure: (a) a Squeeze-and-Excitation (SE) block is injected into the corresponding networks to capture the channel relation in feature map; (b) a Channel Attention Dense Connection (CADC) block is applied to combine multi-scale features and recalibrate channel-wise feature. Experiments on the KITTI dataset show the effectiveness of the proposed approach, which outperforms quantitatively and qualitatively the state-of-the-art Self-Supervised Depth Estimation methods.

DNA-N6 methyladenine (6-mA) methylation modification is one of the most important epigenetic modification markers. The aberrant 6-mA modification can affect gene expression and lead to serious diseases. Therefore, the work of predicting the 6-mA site is of great significance for the understanding of the pathogenesis and treatment of diseases. In this paper, a long short-term memory (LSTM) neural network based on K-mer encoding method and one hot encoding method is proposed to predict methylation sites.Firstly, the information content of gene sequence is increased through K-mer coding method. Secondly, the information content after one hot encoding is converted into a composite encoding matrix. The LSTM model can extract more feature dimensions and types from the encoding matrix, to improve the prediction performance of the LSTM model for gene sequence. The cross validation experiment show that the proposed method can achieve accuracy of 93.7% on benchmark datasets. The sensitivity, specificity and matthews correlation coefficient of the trained model were 93.0%, 94.5% and 0.875, which outperformed existing 6-mA prediction methods. On the other six different species datasets, the proposed method can achieve the area under the curve (AUC) values from 0.9055 to 0.9262,which shows the applicability of the proposed method on animals, plants and microorganisms methylation tasks. The proposed method was applied on rice gene NC_ 029258.1, and the predictions were verified by the recently published online prediction tools. The results show that 86% to 96% of the prediction results are supported by these tools, indicating that the proposed method can be applied to large-scale site prediction and analysis of different species.

The traditional frost detection method based on image processing technology is difficult to flexibly and accurately detect the cold chain refrigeration unit in complex production environment. It is also easy to misjudge due to the influence of environmental factors such as illumination and fog. Therefore, a cold chain refrigerating machine frosting detection method based on spiking neural network is designed. Taking the refrigerator image as the input, the dynamic change of the frosting area of the refrigerator evaporator is automatically detected, the abnormalities caused by interference factors such as illumination and fog are corrected, and the double threshold divided by the cumulative value of pulse emission rate is used as the judgment basis of frosting degree. Experiments are carried out on several cold chain refrigerators put into the production environment. The results show that the designed spiking neural network can adaptively detect the dynamic region of the frosting area of the refrigerator evaporator under the actual production environment and the double threshold which is divided accurately can judge the frosting degree of the evaporator. The detection effect is good and the stability is strong, which can provide a reliable basis for the defrosting time of the defrosting strategy of the cold chain refrigeration unit.

Brain tumors are a major problem affecting the health status of the nation. To determine the extent of brain tumor in order to determine the next step in treatment, a puncture biopsy procedure of brain tumor tissue is often required. Magnetic resonance imaging (MRI) is commonly used to detect brain tumors due to its better soft-tissue resolution. Therefore, research related to MRI-compatible robots is necessary. Based on an MRI-compatible hydraulically driven puncture surgery robot, the kinematic model of the robot is derived based on the principle of hydraulic linker, and the dynamic model of the robot is obtained based on the relevant theory of fluid dynamics. In order to realize the accurate control of the designed robot system, the state feedback H _∞ control rate of this hydraulically driven system is designed according to the H _∞ control theory, which enables the robot to track the target signal quickly and stably. Finally, the average positioning accuracy of the designed robot system in x-axis, y-axis, z-axis, pitch-axis and roll-axis are obtained through experimental studies, and they are, respectively, 0.41 mm, 0.6 mm, 0.67 mm, 0.886° and 1.17°. Experimental results verify the performance of robot-assisted positioning puncture needles, and the dynamic model and control method of the robot have been given, which provide certain reference value for the research of the control algorithm of puncture robots.

Identifying the sentiment polarity of movie customer group reviews may inspire the platform to optimize movie recommendation algorithms and improve services and provide suggestions for consumers' movie choices. A MABM sentiment polarity recognition model is proposed based on the multi-head self-attention bidirectional long-short term mechanism. Using the online review data of the well-known movie review website Douban Review as the corpus, text mining tools are used to pre-process the data. Then dividing the data into training and test sets according to 8:2, and using 10 machine learning models and 4 deep learning models as the control group, the effectiveness and robustness of the MABM model are evaluated by cross-validation and test set validation comparison. In the two sets of comparison experiments, it is found that the deep neural network model predicts overall better than the machine learning model, and the MABM model used as the control set has the best classification results. The MABM model can effectively identify the sentiment polarity of consumer reviews, providing management insights and algorithm improvement suggestions for movie recommendation platforms.

To improve the tracking accuracy of the gait trajectory of the lower limb exoskeleton robot (LLER) , an experimental method of parameter identification is proposed for the LLER’s two-link dynamic model, including static and dynamic experiment. Combined with the wearer’s human parameters, the accurate dynamic model of the LLER human-machine collaborative system is deduced. The sliding mode control (SMC) with the upper bound of the model and a low-pass filter are adopted to track the gait trajectory precisely by MATLAB. The simulation shows that the waveforms of the hip and knee torques measured by experiment are basically consistent with the theoretical values by calculation, and the dynamic parameter identification results are correct. The human-machine collaborative system based on SMC can realize the accurate tracking of the reference gait trajectories of the hip and knee joints, and the low-pass filter can effectively reduce the high-frequency chattering caused by SMC. This research provides a solution for the identification of the dynamic parameters of LLER, a reference model for the model-based control method, and a reference method for precisely tracking the gait trajectory of LLER human-machine collaborative system

The transaction volume of cosmetics is surging, as e-commerce advances rapidly in the context of the Internet and big data, resulting in a large amount of packaging waste that is difficult to recycle. To address the contradictory issues between the high cost of recycling and processing and the low recycling value in the plastic recycling and processing system for cosmetic bottles, reducing the cost of recycling and processing to within target cost range, and improving recycling efficiency to promote the recycling of cosmetics bottles, this paper applies the first creating method to the design idea generation of a plastic recycling and processing system for cosmetics bottles. We first analyzed the unmet needs of the recycling party, established an Extenics model of the contradictory problem, and determined the functional affair-element, the property characteristic-element and the substantial characteristic-element of the recycling system. We then performed extensible analysis and extension transformation on the conjugate part, obtained multiple schemes, judged whether the contradictory problem was solved or not, and used the dependent function of superiority evaluation method to derive the optimal creativity and deepen it. Using the superiority evaluation method, it was calculated that creativity S ₃ has the highest degree of superiority, the actual cost of recycling and processing was lower than the target cost after the enterprise assessment, and the contradiction problem was solvedand the creativity has been adopted by the enterprise. Finally, we verify the validity of the application of the first creating method in the design of plastic recycling and processing system for cosmetic bottles.

Deep learning has achieved great success with big data and powerful computing, but its performance is poor under sample constraint, mainly due to the construction of function space (clusters) and the design of algorithms under dataset constraint. Accordingly, a categorical review of deep learning under restricted samples is presented. In addition, according to the current research on the brain, the cognitive process of humankind is categorized in the brain with different regions, and the cognitive functions of each region are also different. Therefore, the training function of each region should also be different. At this point, an idea of deep learning method using functional evolution is proposed, trying to create a network structure composed of multiple functional modules, and the training procedure of the functional module used in this method is studied, aiming to explore the new area of "humanoid learning".

With the continuous development of manipulator technology, the traditional rigid manipulator, such as space manipulator, surgical manipulator and man-machine interactive manipulator, etc. gradually has difficulties in meeting the needs of some new manipulator systems in terms of lightweight, motion flexibility, large space range, etc. More and more researchers pay attention to the research of flexible manipulator and its high-precision motion control. At present, some effective motion control methods have been proposed for flexible manipulator. However, when an actuator of the manipulator fails, it is difficult for conventional control methods to ensure the original control performances and may even lead to system instability. The research on the flexible manipulator motion control under the effect of actuator faults has important theoretical significance and application prospect. Firstly, the existing motion control methods of flexible manipulator are summarized. Then, according to the types of actuator faults, the effect of actuator performance fault, actuator completely damaged fault and sudden actuator fault on the system are analyzed, and the state of art methods that are used to overcome these actuator faults are summarized. Finally, the key problems to be further solved in the motion control of flexible manipulator under the effect of actuator faults are discussed, and three prospect directions summarized, which has reference value for the further research of flexible manipulator motion control.

Adaptive Dynamic Programming (ADP) algorithm, as a research focus in the field of optimal control, has been widely applied in the field of Energy Management System (EMS). ADP is an effective tool for solving optimal control problems in complex nonlinear systems, which can adaptively adjust the control strategy through the input and output data of the system. The research progress of ADP algorithm and its application in EMS are introduced. Then the research status and algorithm implementation of ADP algorithm in discrete EMS and continuous EMS are analyzed. And at last the Real-time Adaptive Dynamic Programming (RT-ADP) algorithm and its feasibility are introduced.

In the energy harvesting based mobile edge computing (MEC) system, the energy arrivals and wireless channels for computing offloading are both dynamically changing in time and space, which results in dynamic adaptation between communication/computational resource management and task execution. To address such problems, based on the criterion of maximizing the system’s computing throughput, the predication models for renewable energy random arrival and wireless channel are established, and a novel online design framework is proposed for dynamically managing communication/computation resources over time. This solution solves the convex optimization problem time slot by time slot, and based on the optimal structure of offline resource dynamic management and control, real-time resource management strategies are formulated, and it has low computational complexity. Numerical results show that the proposed online sliding window design scheme is superior to the existing benchmark schemes in terms of system computational throughput performance, and has better robust performance against channel/energy state information prediction errors.

High-speed rail needs 6 hours of repair time in the early morning every day, when it cannot be used. However, with the development of society, it is necessary to reduce the window time to improve the operation efficiency of high-speed rail. The traditional deformation monitoring system of high-speed railway track subgrade based on BDS (BeiDou Navigation Satellite System) needs one day observation time to accurately monitor the subgrade, but such observation time during the empty window period of ultra-high railway cannot play a role in improving the operation efficiency of high-speed railway. Particle filter algorithm is added to the original deformation monitoring algorithm. While trying to reduce the observation time to the high-speed railway empty window period, the particle filter algorithm is used to ensure that the coordinate values of the solved monitoring points meet the positioning accuracy requirements of the high railway base under the condition that the sampling data decreases substantially. An experimental simulation is carried out using the measured sampling data of BDS deformation monitoring system of Guangzhou-Shantou high-speed railway, and the effectiveness of particle filter algorithm is verified. The experimental results show that, with the help of particle filter algorithm, the observation time reduced to 15 min can ensure that the a, B and H coordinates of the monitoring points meet the accuracy requirement of ±5 mm for high railway base positioning, which provides an effective method and idea for reducing the time of high-speed railway gap period and improving the operation efficiency of high-speed railway.