Abstract： In this paper, a composite antenna working in L and C bands is designed. The antenna adopts the composite design of discone antenna and microstrip array gold layer. By loading metal short circuit pull, the discone antenna is miniaturized and broadband, it can transmit horizontally polarized omnidirectional electromagnetic waves at 0.95 to 1.6 GHz. The 2×2 microstirp array transmits high-gain directional electromagnetic waves at 4.25 to 4.35 GHz. The composite antenna is embedded in a metal cavity with a diameter of 200 mm and a height of 26.5 mm, which is easy to conform with the aircraft skin. It can meet the requirements of L-band database function and C-band radio altimetry function. The antenna can be widely used in aviation and aerospace fields.

Abstract： In order to solve the problem that the receiver of millimeter-wave high-power radar seeker is easy to burn, this paper designs a millimeter-wave microstrip one-divided-eight-sector fan-shaped power divider working in Ka band, which is used in millimeter wave receiving front-end to achieve the function of anti-burning and low noise. The power divider has a compact structure, and has the advantages of miniaturization, low insertion loss and good phase consistency. The power divider, ultra-low noise and low noise amplifier, limiter and its corresponding eight-channel synthesizer are applied to the receiving front-end as a whole, which breaks the bottleneck of the development of millimeter wave high-power seeker and has broad application prospects.

Abstract：In this paper, a V-band waveguide slots array antenna suitable for FOD detection system is presented. The operation frequency range of the proposed array antenna is 74~79 GHz and the array contains 2 048 elements. The proposed antenna has excellent characteristics such as high gain and low profile by using the form of a multi-mode waveguide slot array antenna and a waveguide power division network. The thickness of the antenna is within 10 mm and the radiation efficiency is above 50%. The experimental results show that the measured and simulated results are agreed well with each other, so it is of good practical engineering application.

Abstract：A 5.5 GHz band wideband directional pattern diversity antenna based on metasurface is designed, which can be used in single band on-/off-body dual communication mode. By analyzing the characteristic mode of the metasurface antenna, the appropriate directional and omnidirectional characteristic mode is selected, the metasurface structure is modified,the higher-order mode is suppressed, and a suitable feed structure is designed to excite it. Finally, the simulation and test results show that port 1 has omnidirectional radiation characteristics and can be used in on-body communication mode, and port 2 has directional radiation characteristics and can be used in off-body communication mode. The working bandwidth of the two modes of the antenna is 23.6% (4.98 GHz~6.25 GHz), and the in-band isolation is higher than 30 dB.

Abstract： A frequency and pattern reconfigurable microstrip patch antenna is designed, which consists of a metal floor with an inverted "Ω" slot and a radiation patch, printed on both sides of the substrate. By changing the operating state of the switching diode located at the center of the floor slot, the local structure and the surface current distribution of the antenna will change, and the antenna's working frequency and direction pattern will be reconstructed. The simulation and measurement results show that the antenna can work in both single-frequency and dual-frequency modes. When the diode is in the on-off state, it first works in single-frequency mode, the resonant frequency of the antenna is at 3 GHz, and its working bandwidth covers the main frequency band of world interoperability for microwave access (WiMAX). While the diode is in the cut-off state, the resonant point of the antenna is at 3 GHz and 5.8 GHz respectively, realizing the dual-frequency working mode. And the operating band covers WIMAX and part of the wireless local area networks (WLAN) band, and the antenna far-field pattern changes with different working modes.

Abstract：According to the classification of the standard architecture given in the new generation AI standard system construction guide, the paper studied systematically on the results of standardization research in AI field in our country, which involves the basic conceptions and the key general technologies. From the statistical data of the effective standards and the projects under study, the conclusion can be obtained that the standards system has been formed of the key general technologies, in addition, the study results have leaded the world on the subdivided technologies and industry applications. However, there are still some outstanding problems, such as unbalanced development of standard research, weak theoretical basis and imperfect ecosphere. Finally, the author put forward some suggestions on the development of relevant standardization system.

Abstract： Convolutional neural networks have established themselves as a prominent technique in computer vision and related fields. By leveraging features such as local connections, weight sharing, and pooling operations, these networks are able to effectively extract local features from images, reducing network complexity, and exhibiting fewer parameters and greater robustness. As a result, they have garnered significant attention from researchers and have led to the rapid development of classification models that are lighter, faster, and more efficient. This article presents a timeline of typical network models used in convolutional neural network development, analyzes their innovative points and advantages and disadvantages, and offers insights into their future development directions.

Abstract： Neural machine translation for the legal domain is of great value for application scenarios such as contract text translation. Due to the scarcity of bilingual corpora in the legal domain, the machine translation performance is still not satisfactory. A practical method to address this problem is to integrate prior knowledge such as translation memory(TM) or templates. However, texts in the legal domain mostly have fixed expression structures and precise wording specifications. The performance of translation in the legal field can be further improved by using both sentence structure information and semantic information in the translation memory. Based on this, this paper proposes a new framework that uses monolingual TM and performs learnable memory retrieval in a cross-language manner. Firstly, this monolingual translation memories contain translation memory and translation template, which can provide richer external knowledge to the model. Secondly, the retrieval model and the translation model can be jointly optimized. Experiments on the MHLAW dataset show that this model surpasses baseline models up to 1.28 BLEU points.

Abstract：In recent years, with the development of deep learning technology, the research and application of image segmentation methods based on coding and decoding in the automatic analysis of pathological images have gradually become widespread. However, due to the complexity and variability of gastric cancer lesions, large scale changes, and the blurring of boundaries caused by digital staining images, segmentation algorithms designed solely from a single scale often cannot obtain more accurate lesion boundaries. To optimize the accuracy of gastric cancer lesion image segmentation, this paper proposes a gastric cancer image segmentation algorithm based on multi-scale attention fusion network using the coding and decoding network structure. The coding structure uses EfficientNet as the feature extractor. In the decoder, the deep supervision of the network is realized by extracting and fusing the features of different levels of multi-path. When outputting, the spatial and channel attention is used to screen the multi-scale feature map for attention. At the same time, the integrated loss function is used in the training process to optimize the model.The experimental results show that the Dice coefficient score of this method on the SEED data set is 0.806 9, which to some extent achieves more refined gastric cancer lesion segmentation compared to FCN and UNet series networks.

Abstract：This paper introduces a CLASS-AB OPAMP-based Low Drop Regulator (LDO) with no off-chip capacitor. Based on the high swing rate Error Amplifier (EA), a dynamic bias circuit has been constructed to feedback voltage to the internal dynamic bias transistors of EA, which greatly improves the LDO’s transient response capability. Moreover, the left half plane zero-pole introduced by the dynamic bias circuit ensures the LDO’s loop stability. At the same time, the EA adopts overshoot detection circuit to reduce the output overshoot and shorten the loop stability time. The circuit is designed and simulated based on 65 nm CMOS process. The simulation result shows that the LDO output is stable with no oscillation under the condition of 10 μA~50 mA load-current and 0~50 pF output capacitor. Under the condition of 2.5 V voltage input, 1.2 V voltage output, and no off-chip capacitor, if the load jumps between 10 μA and 50mA, the recovery time of LDO output is 0.7 μs and 0.8 μs, and the undershoot and overshoot voltages are 58 mV and 15 mV.

Abstract：The design process and measurement results of an internal matching power amplifier are detailed in this paper. This amplifier with GaN HEMT process operates at 5~5.8 GHz. It achieves output power of 48 dBm and power added efficiency of 55% with supply voltage of 28 V. Meanwhile, the bias circuits are integrated in the chip instead of extra off-chip bias circuits, hence it combines the performance and miniaturization. It fully reflects the advantages of high power, high efficiency and miniaturization of GaN internal matching circuit.

Abstract：System in a Package (SiP) has become an important technical route to reduce the volume of electronic devices and improve the integration degree in the post Moore era. It is an important support for the multi-function and miniaturization of electronic equipment in the future. Aiming at the requirements of miniaturization and high performance of the signal processing system, this paper uses SiP technology to design a high-performance signal processing circuit. The SiP circuit integrates a variety of bare cores, including DSP, NOR Flash, DDR3 and gigabit network PHY chips, realizing the miniaturization of a general signal processing core module. Compared with the traditional board level circuit, the SiP circuit is smaller in size and weight under the same conditions of function and performance.

Abstract： Hexavalent chromium is an important detection indicator in water quality monitoring and is harmful to the environment. In order to detect the concentration of hexavalent chromium in water, this paper designs a portable detector for hexavalent chromium in water based on Lambert-Beer law. The working principle and hardware system structure of the detector are introduced in detail, and the detection experiment of hexavalent chromium concentration in water is carried out by using the detector. The experimental results show that the hexavalent chromium portable detector can quickly, reliably and accurately detect the concentration of hexavalent chromium in water.

Abstract：With the development of digital twin technology, the power grid industry in China is also gradually developing from the original physical power grid to the digital power grid. Power transmission and transformation equipment, as the pivot equipment of electric energy transmission and transmission in the power grid, the reliability of its operation is directly related to the safe and stable operation of the power grid. Therefore, it is of great significance to grasp the current operating status of the power transmission and transformation equipment and the operating trend of the future period of time, and to achieve an accurate evaluation of the equipment operating status, to ensure the safe and reliable operation of the equipment. However, in the current practical application process, limited by the poor stability of the sensing device, the harsh on-site operating environment, and the complex electromagnetic environment, the state quantity data of the power transmission and transformation equipment will be missing data, resulting in poor data quality. This directly affects the accuracy of the equipment condition assessment model. In this paper, a genetic algorithm-based method for missing data completion of power transmission and transformation equipment is proposed. The method first randomly assigns values to the transform domain, and then achieves the effect of missing point recovery by minimizing the coefficient vector in the sparse domain. Experiments show that the algorithm can recover missing data accurately.

Abstract：In this paper, the fast location problem of intermediate frequency forwarding interference in satellite communication system is studied. Firstly, the causes of intermediate frequency forwarding interference are analyzed, and the steps and methods of locating intermediate frequency forwarding interference are given. The intermediate frequency forwarding interference model is established through quantitative analysis, and the frequency deviation database is established based on the operation of satellite ground station in the network. An intermediate frequency forwarding interference location algorithm based on frequency deviation is designed to accurately locate the earth station with forwarding interference. The simulation results verify the feasibility of the algorithm in the process of intermediate frequency forwarding jamming location, it effectively reduce the jamming location time and improve the probability of accurate positioning of jamming station, and provide a solution for the rapid jamming location when the satellite communication system is faced with such interference, which has great practical significance for the stable operation of satellite communication network.

Abstract：In the wireless multi hop network environment, the data transmission is subject to more interference and the channel is unstable. Therefore, the traditional TCP protocol can not meet the communication requirements in this complex environment. The increase of ARQ retransmission between nodes will lead to a large cache occupancy rate and a low overall transmission performance. In order to solve this problem, Semi-TCP protocol and ARQ protocol are combined to lower congestion control to the link layer without changing the original TCP architecture, change the error correction capability of ARQ in the link layer, and improve the channel utilization. The simulation on Exata platform shows that this method can effectively improve the network transmission performance in wireless multi hop networks.

Abstract： Aiming at the problem that the clothing industry is affected by fashion trends and the material model is updated rapidly, which leads to the confusion of warehouse inventory structure and low efficiency of operation, a storage location optimization method combining Apriori algorithm with improved genetic algorithm (AIGA) is designed. First of all, Apriori algorithm mines the association rules between material groups, combines the materials with strong correlation to form a large class of warehouse area, and dynamically adjusts the location of the warehouse area according to the large class picking frequency. Secondly, based on the material correlation and picking frequency, a storage allocation model is established with the main objective of minimizing the picking distance. The storage allocation search is carried out by genetic algorithm, and the initialization, crossover and mutation operators of genetic algorithm are improved, and the catastrophe mechanism is designed to improve the search performance of the algorithm. The results show that compared with the existing storage allocation scheme, the picking distance is reduced by 23.85% on average, and the warehouse operation efficiency is effectively improved.

Abstract：Template matching algorithm with rotation invariant is widely used in industrial manufacturing. To solve the problems of traditional template matching methods such as target rotation and matching speed, a template matching method based on local variance and posterior probability classification is proposed. To reduce the amount of computation, some candidate windows are filtered out by local variance in the matching process, and window correlation is calculated by comparing the gray levels of stable feature points in different regions in the posterior probability classification module. Using posterior probability classification to calculate window correlation can achieve rotation invariant during preconditioning and ensure that the accuracy is above 95%. Experimental results show that the algorithm can reduce the matching time to less than 10 ms after selecting the appropriate sliding window moving step on 800 000 pixel-level arbitrary angle matching images, which is faster than the existing algorithms.

Abstract： A virtual endpoint-based address allocation scheme for PCIe switches is proposed for the PCIe bus characteristics, combined with the application characteristics of industrial control equipment and the convenience of design, debugging and problem finding. By analyzing the enumeration and resource allocation process of PCIe devices, adding virtual endpoints to the PCIe switch system topology, designing a fixed allocation mechanism for PCIe peripheral address space, and giving a formula for calculating the physical slot number of peripherals, the faulty peripheral can be quickly located at the moment of peripheral interaction failure. Through experimental verification, the address allocation scheme can realize the fixed address space allocated to the valid endpoints accessed under each downstream virtual PCI bridge in the device supporting hot-swapping, regardless of the number, type and location of peripheral board operation changes, and eliminates the traversal query peripheral address link to improve the efficiency of data transmission, which is beneficial to the PCIe bus can be more conveniently applied to the industrial control field.

Abstract： A large number of sensitive data generated by grassroots governance can be desensitized to remove private content, but these data contain more unstructured text data, which is difficult to desensitize directly. Therefore, it is necessary to identify named entities from unstructured text data to extract sensitive data.Firstly, the sensitive entities are divided into 16 categories and the letters and visits are labeled. The input layer is represented by the pre-trained model BERT, and the coding layer uses the bidirectional long short-term memory network to extract the context information. The decoding layer constructs a highly accurate identification model for sensitive entities in grassroots governance through the conditional random field model optimization sequence. According to the needs of desensitization, the loss weight of false negative and false positive is changed, and the Box Selection rate of sensitive entities is used to assist in evaluating the performance of the model.Experiments were conducted on the grassroots governance petition data set and the public data set MSRA. The F1 values were 88.38% and 90.11%, respectively, which were 4.64% and 3.78% higher than the benchmark model. The model can be applied to sensitive entity recognition of unstructured text with high recognition success rate.The existing evaluation indicators fail to better reflect the indirect reasoning relationship of sensitive entities, and a more perfect evaluation system of sensitive entities should be explored.

Abstract： With the rapid development of Low Earth Orbit (LEO) Internet satellite systems represented by “Star Link”, the number of applications of spaceborne phased array antennas has grown rapidly. In the future, LEO Internet satellites are easy to provide space-borne electromagnetic wave direction of arrival (DOA) detection capabilities. Positioning and investigation of illegal interference in the coverage area of the satellite is an important guarantee for the normal operation and maintenance of the future Internet satellite system. At present, the commonly used TDOA / FDOA positioning system has the problem that the interference source in the blind area cannot be located due to the “positioning blind area” where the positioning error increases significantly. This paper proposes a TDOA / DOA dual-satellite interference source positioning technology system based on the weighted least squares constraint optimization model. The positioning error of geometric dilution of precision factor (GDOP) is analyzed. The simulation results show that the positioning method has the advantage of no “blind spot”. The positioning error is less than 0.2 km in the beam range of 4^∘×4^∘latitude and longitude, and the geographical average of positioning error is 0.112 km, which meets the application requirements of illegal interference positioning and troubleshooting. The accuracy of GDOP error is verified by the Monte Carlo method of root mean square error (RMSE). The geographical stability of the positioning error of this positioning method is better than that of the conventional two-satellite TDOA / FDOA positioning algorithm.

Abstract： In passive positioning scenarios of Low Earth Orbit (LEO) satellites, different target radiation sources interfere with each other, and the Time Difference of Arrival (TDOA) parameters of different targets are mixed and difficult to distinguish, so it is difficult to achieve accurate target positioning. In this paper, based on the basic principle of Clustering Algorithm based on Grid Density (CAGD) and the multi-complex characteristics of TDOA parameters, a multi-objective TDOA parameter sorting model is constructed to achieve TDOA parameter sorting. By defining the grid density trough, the model solves the problem that the data among localization targets are clustered into one class, and introduces the principle of position connection to realize the best clustering, and finally realizes the sorting of localization parameters. The simulation results show that compared with the traditional grid and density clustering methods, the proposed method performs better in the multi-target TDOA parameter sorting in the scenario of LEO satellite passive positioning.

Abstract： In recent years, radar detection technology has made rapid development in various fields, and is widely used in military, civil and other aspects. Harmonic radar has attracted people's attention due to its advantages of non-contact, penetrability and portability. In order to improve the detection range resolution and anti-electromagnetic interference ability of the transmitter, a design scheme of a broadband high-power harmonic radar transmitter is proposed, and the specific design process of the power amplifier, antenna and other key devices in the transmitter front-end is given. Through the simulation and analysis of the harmonic radar transmitting front-end system, the physical production and testing, the feasibility of the scheme is verified.

Abstract： Simulation and analysis is given about channel group delay influence for high-order signal receives. In the condition of linear group delay and parabolic group delay, index is given about system group delay design with 8PSK,16APSK and 16QAM. For compensation technology of group delay, CMA and DD equalization algorithm is used, and result is given by simulation, which can be used to subsequent engineering design.

Abstract： With the continuous progress of aerospace technology, especially the rapid development of commercial aerospace in recent years, high-density space launch has become the norm. As one of the important business systems of the space launch site, the operational ability and efficiency of the system will greatly affect the results of the implementation of the space launch mission. Therefore, it is necessary to carry out the research on the decision-making and evaluation of the process of the measurement and control mission. Considering that the rapid assessment of future measurement and control tasks will present the trend of less people, generalization, automation and intelligence. This paper presents a design idea of measurement and control task evaluation system based on domain driven design. Through in-depth analysis of measurement and control task evaluation business, combined with the characteristics of domain-driven design and micro-service architecture, the whole system is designed through domain model design, micro-service identification, micro-service implementation and other steps. Finally, the feasibility of this scheme is verified by combining with typical space launch TT&C mission scenarios, which provides reference and ideas for further improving the efficiency of TT&C mission evaluation in China.