Abstract:Free Space Optical（FSO） communications offer high speed, low cost, and strong anti-interference ability. However, the atmospheric turbulence-induced fading causes deterioration in the performance of FSO communication systems. The conventional solution is to use Radio Frequency（RF） links as parallel communication links to improve the system’s performance. On the other hand，Reconfigurable Intelligent Surfaces（RIS） can be employed to further improve the received signal-to-noise ratio of the RF link due to its advantages of low loss, easy deployment, and no need for complex coding and decoding. In this paper, a RIS-assisted hybrid RF-FSO transmission system is proposed to improve the communication quality of service. Based on this hybrid model，the exact expressions for the outage probability, average Bit Error Ratio（BER），and channel capacity are derived, and Monte Carlo simulations are presented to verify the accuracy of the analytical results. The results show that the performance of the proposed system is significantly improved compared with the conventional hybrid RF-FSO system.
Abstract:Although Multiple-Input Multiple-Output（MIMO） technology can improve the utilization rate of the spectrum, multi-dimensional signal processing brings great challenges to the detection of MIMO signals. Based on the analysis of various MIMO detection algorithms, the nonlinear QR decomposition algorithm is selected as the research object. In order to obtain a higher performance of detection, the sorted QR decomposition is further studied and the sorting scheme based on the L1-norm is proposed. Using Matlab for performance simulation, the L1-norm sorting strategy and the L2-norm sorting strategy have the same impact on the MIMO system, but the L1-norm sorting strategy reduces the computational complexity. On this basis, the hardware structure of the improved sorted QR decomposition by Givens rotation on FPGA is proposed. Compared with the solution of the L2-norm, the L1-norm strategy reduces at least 29.2% combinational logic resources and 32.4% register resources when calculating a single-column norm in the realization of 4×4 channel matrix decomposition. By comparing the design of the integral structure and similar dimension one, the frequency of the operating clock is significantly improved.
Abstract:A terahertz fundamental up-conversion mixer with a high Local Oscillator（LO）/Radio Frequency （RF） and local oscillator /Intermediate Frequency（IF） port isolation was presented based on IHP 0.13μm SiGe BiCMOS process. The mixer adopted Gilbert’s double-balanced structure, and the local oscillator signal was transmitted through the Coplanar Waveguide（CPW） to suppress the transmission asymmetry caused by the strong parasitic coupling effect in the transmission process, reducing the characteristic of LO/RF port isolation deterioration caused by the asymmetry. By adopting an asymmetric switch interconnection structure, the imbalance of the parasitic coupling of the local oscillator signal at the collectors of the switching transistors was reduced, and the cancellation efficiency of the local oscillator signal at the collectors of the switching transistors was improved. The leakage of the local oscillator signal was suppressed at the port of intermediate frequency by arranging the position of the transconductance transistors in a reasonable layout. The post-simulation results show that，under the power supply voltage of 2.2 V，the local oscillator signal is 230 GHz，and the intermediate frequency signal is 2-12 GHz. When the up-conversion mixer works at 218-228 GHz, the LO/RF port isolation is greater than 24 dB, LO/IF port isolation is greater than 20 dB，and the conversion gain is -4 dB to-3.5 dB. The output 1 dB compression point is -14.8 dBm when the intermediate frequency signal is 10 GHz. The DC power consumption is 42.4 mW, and the core area of the chip is 0.079 mm2. The up-conversion mixer can be applied to a wireless transmission system using a high-order quadrature amplitude modulation method.
Abstract:The classical Maximum Eigenvalue Detection (MED) algorithm has an excellent performance in detecting correlated signals. However, with the increasing signal dimensionality, the MED algorithm faces serious problems in the calculation efficiency and implementation of the test statistic and decision threshold, which greatly limits the further application of the algorithm in modern cognitive communication systems. To this end, a low-implementation complexity MED algorithm based on a numerical analysis theoretical framework is proposed. The new algorithm uses the Rayleigh quotient accelerated power method to iteratively compute the test statistic, which has a fast convergence rate in detecting high-dimensional signals compared with the classical power method. Meanwhile, different from the classical look-up table method, the new threshold calculation method based on the cubic spline interpolation method is proposed, which can quickly determine the decision threshold corresponding to any given target false-alarm probability. The proposed MED algorithm effectively improves the computational efficiency and reduces the complexity of algorithm implementation while maintaining the detection performance of the original algorithm, which is particularly attractive for spectrum sensing problems in high-dimensional conditions. Finally, the simulation results demonstrate the effectiveness of the proposed algorithm.
Abstract:This paper proposes a state-sensitive based event-triggered H∞ control strategy to solve the problem of unmanned ground vehicle (UGV) path tracking control under communication restriction. Firstly, a corresponding path-tracking control model is established according to the dynamics of the connected vehicle. Secondly, a novel state-sensitive event-triggered communication (SS-ETC) strategy according to the state perception of path tracking in real time is proposed. Then, an event-triggered H∞ controller is designed by combining the time delay system modeling method and Lyapunov stability theory. The proposed dynamic event-triggered communication strategy based on state perception can dynamically adjust the communication threshold according to the state measurements of the control system, and effectively realize the adaptive co-design of UGV communication and control. Finally, the effectiveness of the proposed dynamic event-triggered control strategy is verified by simulation experiments.
Abstract:Accurate detection of leakage is the key to reduce the leakage rate of water distribution networks. This paper proposes a leakage detection method based on a semi-fixed-length sliding window. The method uses a sliding window to detect leakages by analyzing the time series flow data and improves the information quality of the acquired data with a variable-length window, in which the length of the window is limited. Based on Clustering by Fast Search and Find of Density Peaks（CFSFDP），the entropy function is introduced to adaptively select cutoff distance parameters according to the data distribution characteristics. In this way, the detection rate of leakage events is improved. The experimental results show that the proposed algorithm can effectively detect the leakage data in the four simulated scenarios, and obtain a higher leakage detection rate and a lower false alarm rate.
Abstract:Sonar image is seriously polluted by noise, which leads to the problem of low precision in underwater multi-target segmentation.?Therefore，this paper proposes an underwater multi-object segmentation technique based on self-adjusting spectrum clustering，combined with the entropy weight method.?The technology firstly clusters through self-tuning spectral clustering of sonar image pixel clustering processing, so that the image is divided into multiple independent areas. According to the complementarity and redundancy of multiple features, the information entropy, brightness, contrast and narrow length of each region are calculated. The entropy weight method is used to weight multiple features and select the optimal target region. Then，the optimal target region is matched with all regions by multi-feature similarity. Finally, all target regions are segmented automatically by the adaptive threshold iterative method according to the matching results of similarity. Experimental results show that there is no over-segmented of noise interference regions, and target regions segmented have higher accuracy, which verifies the effectiveness of the proposed method.
Abstract:As affected by suspended particles such as haze in the atmosphere, images taken outdoors often suffer from low contrast and low visibility. The existing dehazing methods fail to make full use of the local feature information of the image，and there are problems such as incomplete dehazing and loss of image details. For this reason, this paper proposes a T-shaped image dehazing network based on wavelet transform and attention mechanism. Specifically, the proposed network obtains the edge detail features of the hazy image by performing multiple discrete wavelet decomposition and reconstruction on the image and proposes a feature attention module that takes into account both the global feature and the local information extraction of the image, which strengthens the network’s learning in image visual perception and detail texture. Secondly, in the process of feature extraction, a T-shaped method is proposed to obtain multi-scale image features, which expands the network’s representation ability. Finally, color balance is performed on the reconstructed clear image to obtain the final restored image. A large number of experimental results in synthetic data sets and real data sets show that the network proposed has superior performance when compared with other existing network models.
Abstract:The development of ship intelligence puts forward higher demands on the real-time object detection capability of ship vision perception systems. YOLOv5, the latest achievement of the YOLO (You Only Look Once) series of algorithms, is widely used for object detection at sea with good speed and accuracy. However, in actual sea navigation, it is often accompanied by variable natural conditions and complex activity scenarios, which makes its ability to detect small objects and multi-target classification in complex waters unsatisfactory. Therefore, to improve the target detection capability of YOLOv5 in complex seas, this paper proposes a Multi-Path Aggregation Network (MPANet) structure. MPANet enhances multi-scale localization capability by fusing multi-level feature information in the bottom-up feature transfer process，and enhances higher-order feature semantic information by combining the SimAM attention module and Transformer structure. The experimental results of the custom dataset show that MPANet-YOLOv5 improves AP by 5.4%, recall by 3.3%, AP0.5 by 3.3%, and AP0.5:0.95 by 2.2%，compared with the YOLOv5 model. The results of different sea area tests show that MPANet-YOLOv5 has significantly better detection capability for small objects on the sea surface than YOLOv5.
Abstract:With the rapid development of the Internet, virtual communities are constantly emerging. While these communities provide innovative resources, there are also problems such as the low willingness of users to share and a lack of good incentive mechanisms. Blockchain can better solve these problems and promote community knowledge sharing. This paper constructs an online community knowledge sharing scheme based on Multi Chain, proposes the resource access and storage mode of “metadata & cloud storage”, designs the metadata information table in detail, and designs the overall framework of the knowledge sharing scheme and the key processes of some businesses. Based on the key process, the consensus mechanism of “Nominated Proof of Stake (NPOS)” is proposed to design the blockchain network, which realizes some functions of online community knowledge sharing. Through analysis and experiment, the scheme of this paper has good scientific rationality, safety and execution efficiency, and has good reference value for the development of other related projects.
Abstract:：Experts can provide authoritative answers for community question answer (CQA). Efficient and accurate expert discovery can help improve the service quality of CQA. The expert discovery accuracy of the supervised learning model is reduced by the noise label data existing in the community user data as well as the unbalanced classification data due to the small number of experts. A semi-supervised expert discovery method based on feature perturbation is proposed to solve the mentioned problems. In this method, a feature perturbation strategy for unlabeled data is constructed, using the Sharpening algorithm to label the pseudo-label of unlabeled data. Based on the ADASYN algorithm, expert sample data is expanded by constructing neighbor samples of expert users to alleviate the imbalance of classification data. A joint loss function is constructed, which trains the classifier by both the labeled and pseudo-labeled data to enhance the generalization performance of the method. The experimental results show that this method is superior to the existing models and methods in several evaluation indexes.
Abstract:In the Internet economy, the takeaway has become a popular way of consumption. However, the current takeaway route optimization model and algorithm do not consider the rider’s goal and the disturbance factors they faced, which makes the rider trapped in the system. Most studies optimize the takeaway route as a static problem, generating routes by pairwise insertion of pickup and delivery nodes. However, the takeaway route optimization is dynamic and real-time, and the cross of pickup and delivery is the staple mode of delivery, which means that riders can go to multiple nodes to pick up before delivery. Therefore, this paper studies the delivery route optimization under the cross of pickup and delivery and various interference factors. Firstly, the empty cost of riders and the cost of riders waiting are added to the target function, and an optimization model for takeaway delivery routes is established. Secondly, four interference factors are considered, including mid-way orders, traffic control, the abnormal delivery time of merchants, and the abnormal time of customer pickup. An improved adaptive large neighborhood search algorithm is designed to achieve efficient route optimization. Finally, a simulation example is generated based on the Ele.me platform to verify the effectiveness of the model and algorithm.
Abstract:The Internet of Things（IoT） carries the safe transmission and storage of a large amount of sensitive information. Since IoT devices are resource-constrained, which have expensive communication, slow mission velocity and need to store sensitive information security primitives （such as public key algorithm and digital signature），they are not suitable for the authentication of lightweight IoT devices. This paper proposes a lightweight anonymous key sharing security authentication protocol for IoT devices, which generates a shared key by the Physical Unclonable Function（PUF） and uses security primitives such as the MASK algorithm and the Hash function. The security analysis and verification are accomplished by Ban logic and ProVerif to prove that the protocol ensures security attributes such as anonymity, non-repudiation, and forward/backward confidentiality. Compared with other protocols, this protocol has the characteristics of low computing cost, small communication overhead and storage capacity, and high security performance, which is suitable for the secure communication transmission of resource-constrained devices.
Abstract:The stock price is nonstationary and volatile, the investors are easily influenced by their own sentiments, and their investment decision is irrational. Thus, the stock price is difficult to predict. Aiming at the problem of an unbalanced distribution of text labels in the sentiment analysis method based on the CNN neural network, this paper proposes a stock price prediction method based on sentiment analysis and a generative adversarial network. First, a sentiment dictionary database is established in the financial field. Then, the dictionary-based sentiment analysis method is used to calculate the sentiment polarity of financial text data and the overall sentiment trend of investors every day, that is, the sentiment index. Finally, the generative adversarial network is used to predict the stock market volatility, where the generator generates stock sequence data, and the discriminator uses a convolutional neural network to distinguish the generated data from the real data. This method can dynamically update the prediction results of stocks and obtain smaller error values.
Abstract:Money laundering in cryptocurrency transactions is differentiated from traditional financial crimes due to its strong anonymity and decentralization. The existing anti-money laundering techniques cannot be directly applied to cryptocurrency transactions. Considering the traceability, interpretability, and measurability of money laundering crime forensics, this paper designs a four-stage money laundering detection approach: (1) defining a set of transactions of a user in a period as a transaction behavior; (2) constructing a set of features to characterize transaction behaviors; (3) adopting outlier detection and small cluster detection methods to find out loud and subtle anomalous transactions; (4) analyzing the suspicious score distributions of users and calculating a suspected-launderer value for each of them. To evaluate the performance of our proposed method, a real-world money laundering dataset is obtained and experimentally evaluated. The experiment results show that our approach obtains 96.02%, 95.05%, 95.83%, and 95.81% accuracy in terms of abnormal transaction behaviors, suspected money launderers, loud abnormal transactions, and subtle abnormal transactions, respectively, all better than benchmark algorithms. Moreover, the carefully-designed features of transaction behaviors can offer supportive interpretations for the detection results and help exchange security officers to carry on further investigations and crime forensics.
Abstract:To improve the safety and stability of the wheeled off-road vehicles, the impact of the key element balance valve in the drive system on the performance of the off-road vehicles hydraulic driving system is studied. The hydraulic driving system of the off-road vehicle is taken as an example to establish the theoretical analysis model of the balance valve. By means of establishing the hydraulic drive system model of the off-road vehicle through AMESim software, the pressure and speed of the pump，hydraulic motor, and balance valve in the drive system under flat road driving conditions and downhill driving conditions are analyzed. The correctness of the simulation model is verified through experiments. The results indicate that the balance valve has the functions of balancing negative loads, preventing insufficient fuel supply and hydraulic motor stalls under downhill conditions. In order to improve the shock in the downhill process, damping groove depth and orifice diameter at the control end of the balance valve spool are optimized. Through the analysis results, it is found that changing the damping groove depth can effectively reduce the pressure shock, and when the groove depth is reduced from 0.55 mm to 0.35 mm, the back pressure can be reduced by 12.5%. The theoretical model and simulation model provides a reliable basis for the further optimization of the hydraulic driving system of the off-road vehicle.
Abstract:In order to reduce volatile organic compounds (VOCs) in a vehicle and find the solution, this study proposes a new mathematical model and method of quantitative analyzing and tracing VOCs at multi-levels. In the present work, it is the first time to establish a multi-level VOCs mathematical analysis model and tracing method from vehicle parts to a whole vehicle by synthetically considering both the interactive effects of VOCs compositions emitted from different vehicle parts, and the hazard level of each composition of VOCs based on the national standard limits. This study also applies the multi-objective decision method with the triangular fuzzy number semantics definition theory to construct a pairwise comparison matrix for quantitatively analyzing the influence weights of different levels and the combined weights of multi-levels. The proposed VOCs tracing method can not only accurately analyze the contribution sequence of different VOCs compositions to the VOCs in a whole vehicle, but also quantitatively calculate the contribution of each typical vehicle part to the VOCs in a whole vehicle considering the interaction effects of different VOC compositions coming from different parts, and thus providing a more accurate theoretical method for the traceability and control of VOCs in vehicles. The case study shows that the contribution sequence of different VOC compositions is ranked as formaldehyde > acetaldehyde > toluene > xylene > ethylbenzene. The contribution sequence of vehicle parts to the VOCs in a whole vehicle is ranked as front seats > rear seats > carpet > dashboard and ceiling. The case study proves that the contribution of a vehicle part to the VOCs in a whole vehicle is not a simple superposition of all VOCs compositions, and it is the interaction results coming from different vehicle parts.
Abstract:In order to achieve precise and stable control of the robot joint under the influence of nonlinear friction and unknown external disturbance moment, the LuGre friction model is modified to describe the nonlinear friction characteristics of the system, and an adaptive algorithm is used to compensate the friction to approximate the change of friction. The fuzzy neural network is used to approximate the influence of unknown external disturbance moment on the system. In this paper, the tangent barrier Lyapunov function is introduced to constrain the output signal, so that the error is limited within a given range. A hyperbolic sine function tracking differentiator is used to solve the “differential explosion” caused by virtual input differentiation and the poor accuracy of the first-order filter. A fuzzy adaptive backstepping control method with friction compensation is proposed by combining the adaptive control method with the backstepping control theory. Lyapunov criterion is used to prove that all the errors of the closed-loop system are uniformly bounded. Simulation results show that，compared with the traditional PID control and RBFDSC, the position tracking error of the proposed control method is improved by nearly 7.5% and 3%, respectively. Moreover, when the parameters of the LuGre model are changed, the adaptive algorithm can accurately track and compensate for the friction force, thus verifying the effectiveness and robustness of the proposed control strategy.
Abstract:The software Fluent was used to simulate the full flow field of the submersible pump, to solve the bulky volume of submersible pumps. Based on the experience impeller structure of CFturbo software, the steady simulation analysis was conducted on the flow field characteristics of a submersible pump. The parametric modeling and simulation process of the impeller structure was realized by the ISIGHT optimization platform with CFTurbo and Pumplinx. The impeller structure was optimized by using the Multi-island genetic algorithm for the three optimization objectives: head, efficiency, and volume. The optimized results show that the head is increased by 5.1%, the hydraulic efficiency is increased by 2.1%, and the impeller diameter is reduced by 1.9% under the rated flow condition. The performance of the submersible pump under different flow conditions is generally better than that before optimization. According to the experimental data measured in the field, under the rated flow condition, the head error is 0.3%, which proves that the optimization results are reliable and submersible pump performance is good.
Abstract:The relative motion between the gear teeth of harmonic drive belongs to the spatial conjugate motion under the spatial elastic deformation condition of the flexspline shell, and its spatial conjugate theory is the core factor that determines its motion and force transmission and comprehensive performance. For this purpose, a spatial conjugate motion model for harmonic drive is proposed based on adjoint approach. By means of the semi-moment theory of the shell and the kinematics between the components, the ruled surface motion equation of the neutral layer generatrix of the shell is established. According to the quasi-fixed line condition and the ruled surface motion of neutral layer generatrix, the axode equation of the gear teeth in harmonic drive is derived. The internal relationship between the axodes and the conjugate tooth surfaces is studied by the adjoint approach. Taking the axodes as the original surfaces and the tooth surfaces of the circular spline / flexspline as the concomitant surfaces, the conjugate condition formula of the spatial concomitant motion of the rigid tooth surface is derived, and the conjugate model of the space concomitant motion is formed. The relative motion between the meshing points is transformed into the spiral motion around the instantaneous axis, and the relationship between the instantaneous axis and the normal vector of the meshing point is analyzed. The spatial conjugate motion is degenerated into the planar conjugate motion, and the constraint properties of the spatial conjugate and planar conjugate are analyzed. The plane of action is constrained as a quasi-fixed surface, and the spatial conjugate motion characteristics under the condition of quasi-fixed surface are analyzed. Finally, through the simulation analysis of an example, it can be seen that the plane motion of the rigid gear teeth of harmonic drive is a special motion after the degradation of spatial motion, and the spatial motion after the degradation of harmonic drive is consistent with the plane motion, which verifies the correctness of the spatial conjugate model and motion characteristics described in this paper.
Abstract:Aiming at the burning loss of smart electricity meters, after correlation analysis of various factors, this paper proposes a burning fault prediction for smart meter based on XGBoost algorithm. Taking the data of a province from 2019 to 2020 as an example, the proposed method is tested and verified. Using the basic information data, operation data and environmental data, the proposed method is compared with the traditional algorithms such as KNN, naive Bayes and support vector machine. The results show that the burning fault prediction of the XGBoost algorithm is better than the traditional algorithms. The precision of the XGBoost is 91%, the recall is 66%, and F1-score is 76.51%. In the process of system deployment, LSTM algorithm is used to fill some missing values. The experimental results show that the model can accurately predict the burning fault of smart meter in low-voltage platform area.
Abstract:In order to simulate and optimize the Electron Beam Lithography（EBL） process, and to improve the manufacturing quality of EBL layout, our team in Hunan University（HNU） developed a set of Electronic Design Automation（EDA） software toolkit named “HNU-EBL”. In this software，the following functionalities have been implemented：1）Calculation of the scattering process and trajectory of the electron beam in the resist and substrate based on Monte Carlo method；2）Calculation and fitting of the point spread function of electron beam scattering based on the multi-Gaussian plus exponential function models；3）Correction of the proximity and fogging effects and optimization of the incident electron dose distribution based on the GDSII lithography layout；4）Calculation of the energy deposition density under a given incident electron dose distribution based on convolution, and evaluation of the key lithography pattern fidelity metrics such as edge placement error. Using an Exclusive OR（XOR） integrated circuit layout as the lithography target pattern，the EBL process of a 10 kV electron beam in Polymethyl Methacrylate（PMMA） resist and silicon substrate layers is calculated. The functionalities and validity of the HNU-EBL is demonstrated by comparing the developed layout patterns with and without the proximity effect correction. Using exactly the same computing hardware and calculation settings，it is shown that the proposed HNU-EBL EDA software’s efficiency is better than some of the imported mainstream EBL EDA software. The website http://www.ebeam.com.cn has been established, and the HNU-EBL software is licensed to EBL users for free.
Abstract:IGBT junction temperature fluctuation is one of the important factors affecting the accuracy of IGBT life evaluation of photovoltaic inverters. This paper proposes a lifetime evaluation method for photovoltaic inverters that considers the influence of fundamental frequency/low frequency junction temperature. Firstly, the junction temperature profile is divided into different time scales. The physical characteristics of the fundamental frequency junction temperature and the low frequency junction temperature were compared and analyzed. Secondly, the calculation range of the junction temperature fluctuation frequency in the IGBT lifetime evaluation process was expanded, and the corresponding photovoltaic inverter IGBT electro-thermal model and lifetime model were established to quantitatively analyze the influence of fundamental frequency/low frequency junction temperature on the lifetime evaluation of photovoltaic inverters. Finally, taking different latitude regions as examples, the life damage of photovoltaic inverters under different mission profiles and different sampling periods is evaluated. The results show that the method comprehensively considering the influence of fundamental frequency/low frequency junction temperature can effectively improve the accuracy of lifetime evaluation of photovoltaic inverters, which is helpful to guide the operation and maintenance of photovoltaic inverters, and reduce economic losses caused by unplanned shutdowns.
Vol. , No.,