Abstract:

Abstract: To promote the theoretical development and practical application of tendon /cable driven continuum robots，the relevant studies were analyzed and summarized from the aspect of modeling and control． Firstly，the structure design methods and development trend of tendon /cable driven continuum robots were introduced． On this basis，the kinematic and dynamic modeling methods of tendon /cable driven continuum robots were summarized． The modeling methods were compared from the view of modeling principles，model accuracy，complexity and computational efficiency，which would provide a guideline to select the modeling methods in different application scenarios． Then，the reason of model mismatch，parameter perturbation and dynamic response in the control of tendon / cable driven continuum robots was discussed and the existing control strategies were divided into three categories including model-based control，model-free control and hybrid control． The implementation methods，advantages and disadvantages and the development trends of control strategies in open loop and closed loop application scenarios were surveyed． Finally，the challenges in modeling and control of tendon /cable driven continuum robots were concluded and the future development direction was prospected

Abstract: Based on the analysis of the current situation and problems of water resources utilization in the Yellow Ｒiver Basin，this study examined the related concepts of utilization rate of water resources，and put forward the index of " net-utilization ratio of water resources" to represent the development and utilization of water resources in the Yellow Ｒiver Basin． Super-SBM model was further used to quantify utilization rate of water resources in the Yellow Ｒiver Basin，and an adaptation model was constructed by combining Tapio theory to explore the decoupling state between water resources utilization and economic development． The restrictive factors and ways to improve the utilization level of water resources in the Yellow Ｒiver basin were discussed． The results showed that: ① In 2020，the utilization rate of water resources in the Yellow Ｒiver Basin was 65%，and the net-utilization rate of water resources was 58%． Although the actual water resources utilization intensity is not as high as imagined，the development and utilization of water resources in the basin was still in an excessive state． ② The overall level of water resources utilization in the Yellow Ｒiver Basin showed an upward trend during the study period，and the spatial pattern of " higher in the lower reaches than in the middle reaches than in the upper reaches" was generally presented in the nine provinces． Different development patterns and characteristics led to the differences of water resources utilization in different regions． ③ The influencing factors of water resources utilization level were mainly divided into industrial structure，scientific and technological level，policy regulation，economic development and water conservancy facilities construction． Finally，corresponding improvement approaches were proposed for the influencing factors.

Abstract: In order to clarify the spatial distribution characteristics of non-point source nitrogen (N) and phosphorus (P) nutrient pollution risk in the watershed, the key non-point source pollution source areas in the watershed were identified. This study took the Yiluo River Basin as an example. The output coefficient method was used to quantify the N and P pollution loads generated by different land use types, residents′ lives and livestock breeding in the watershed. With the improved potential non-point pollution indicator (PNPI) model and SWAT (soil and water assessment tool) model, the spatial distribution characteristics of N and P pollution risk were described, and the key source areas of N and P pollution were identified. Pearson correlation coefficient method was used to calculate the correlation between the simulation results of the two models, and the reliability of the simulation results of the improved PNPI model was evaluated. The results showed that in 2020, the spatial distribution of N and P non-point source pollution risks in the Yiluo River Basin was similar. It showed the spatial distribution characteristics of lower pollution risks in the upper reaches of the Yi River and Luo River tributaries, and higher pollution risks in the middle and lower reaches. N and P very low, low, medium, high and extremely high risk areas accounted for 48.85%, 14.61%, 9.68%, 14.64%, 12.22% and 55.48%, 8.76%, 11.14%, 13.25%, 11.38% of the total area of Yiluo River, respectively. The output of land use was the main source of N and P pollution in the basin, with the load of 20 643.62 t/a and 3 033.31 t/a, respectively. Among different land use types, the output of N and P pollution from arable land was the most, which were 13 000.07 t/a and 1 956.44 t/a, respectively; the output of grassland produced the least N pollution, with a pollution load of 1 322.99 t/a. The P pollution of residential land was the least, and the pollution load was 113.61 t/a. The Pearson correlation coefficients between the N and P pollution loads simulated by the improved PNPI model and the SWAT model reached 0.6, indicating that the improved PNPI model was suitable for the study area.

Abstract: The H_∞ filtering for cyber physical systems with incomplete information transmission was studied. This phenomenon mainly included quantization error caused by communication equipment defects and continuous random multiple packet loss caused by inherent network factors. The state space model of cyber physical systems and H_∞ filter model were constructed by using Delta operator discretization method. Through analysis and modeling, the filter error system was obtained by combining quantization error and packet loss model with filter model. Then, by using Lyapunov stability theory and Schur complement lemma, two theorems guaranteeing asymptotic stability and H_∞ performance were deduced. The filter was verified by simulation based on target tracking system. Assuming that the parameters of quantizer were fixed，with the increase of the maximum number of packet loss，the H_∞ performance index γ increased gradually. Which indicate that the more serious the packet loss，the greater the impact on the stability of the system，but the designed filter still could meet the stability condition. Simulation results showed that the proposed filtering method was effective.

Abstract: In order to solve the problems of low recognition accuracy and easy omission of existing target detection methods in foggy scenes, an improved vehicle detection method based on YOLOv5s was proposed. Firstly, based on VisDrone data set, LightFogVisDrone and ThickFogVisDrone were generated by atmospheric scattering model,and the MixFogData was composed of real fog scene pictures. Secondly, the Mosaic data enhancement method of the original model was improved from the original 4 pictures to 9 pictures randomly, which reduced the gray background area, accelerated the convergence of the model and improved the training efficiency, and the CBAM attention mechanism module was added before the prediction end to improve the feature extraction ability of the network to tackle the problem of missed detection of occluded targets and small targets. Finally, the prior frame of NMS non-maximum suppression value was optimized to improve the problem of missing detection of vehicle targets. The experimental results showed that, compared with the original YOLOv5s, the average accuracy of the improved YOLOv5s in light fog, dense fog and mixed fog was increased by 16.14, 16.16 and 15.05 percentage points, respectively, which proved that the improved YOLOv5s was effective and practical for vehicle target detection in foggy environment.

Abstract: To address the problem of efficient detection of noncooperative communication signals in large bandwidth scenarios, an intelligent detection and identification method for broadband signals based on time-frequency map cutting was proposed in combination with machine learning techniques. In this study, the method adopted Mobilenet network instead of CSPdarknet53 network in YOLOv4 for feature extraction, and constructed a lightweight YOLOv4 model. At the same time, the model introduced the Focal-EIOU cost function and an improved weighted box fusion algorithm (WBF), which could effectively improve the training efficiency and the detection and recognition accuracy. The experimental results showed that the method in this study could quickly detect the continuous and burst signals in the communication acquisition data with large bandwidth, as well as the moment of appearance, frequency range, modulation mode and other related parameters, and its performance is better than the traditional energy detection methods. Compared with other similar methods, the average detection accuracy (mAP) of this study was greater than 81%, and the detection speed of YOLOv4-MobilenetV1 model reached 77.60 frames/s, which was better for both detection accuracy and real-time requirements and was more conducive to engineering deployment.

Abstract: Aiming at the problem of low accuracy and positioning divergence in the UWB TDOA positioning mode with the positioning scenarios such as severe noise environments or tags near the base stations, a new TOF and TDOA joint positioning method was constructed by introducing TOF ranging and a weighted centroid algorithm. Firstly, using TOF ranging, the TDOA hyperbolic positioning equation was obtained, which did not require strict time synchronization compared with the traditional TOA ranging. In order to solve the problem that multiple positioning circles did not intersect and hence the positioning equation has no solution in severe noise environment, the TOF weighted centroid was introduced to obtain the tag initial rough position, which would be substituted into the TDOA hyperbolic equation. Finally, the Taylor iteration was adapted to calculate the tag final coordinates. Besides, the positioning accuracy, complexity and tag positioning scatter plot were analyzed and compared with the existing methods by simulation experiment, the corresponding simulation results have shown that the proposed positioning algorithm could provide the higher positioning accuracy and the larger positioning range even in severe noise environment, for the noise of variance 0.75, the positioning accuracy of the proposed algorithm was improved by 30.1%, 25.6% and 25.8%, respectively, compared with the existing TOF、TOF-TDOA and TOA-TDOA. For the tags away from hyperbolic asymptotes or near the base stations, the proposed algorithm could provide more stable positioning performance and meet the needs of real-time positioning in UWB systems.

Abstract: In recent years, intelligent services in various fields have been paid more and more attention and achieved rapid development. There are a lot of user comment data in the use feedback of each application software. It is of great significance to mine meaningful user demands from these large numbers of user comment data with varying quality. At present, the existing requirements intelligent classification methods do not improve the reusability and application value from the perspective of service computing. Therefore, it is an important direction to study a method of intelligent mining of user requirements and design it as reusable service to help the iterative update of application software. This paper focuses on the design method of an intelligent service to mine user needs, selects user review data of multiple apps on IOS platform, conducts intelligent mining and classification on them, and analyzes the potential demands. First, analyze the user demand category, define the demand category Then, the user data is crawled, cleaned and annotated to form a software classification data set. Then, the effect of mainstream deep learning (TextCNN, BiLSTM_Attention and BERT) on intelligent classification of user comment data is explored through experiments, and the classification results are prioritized. Finally, the intelligent service is designed and packaged as a Python package that can be invoked. Through experimental comparison, it is found that BERT model performs better in Precision, Recall and F1-Measure.

Abstract: Aiming at the problem of substation equipment point cloud recognition, this study proposed a new recognition method. Firstly, a local coordinate system of equipment point cloud was established. The symmetry and distribution density of equipment point cloud were used to determine the x and y axes of the local coordinate system. The coordinate system was invariant to translate and rotate, and was robust to noise. Then, a new feature descriptor was defined based on the difference between the shape and view of substation equipment point cloud, which was used to describe and recognize the point cloud of equipment. Also, a template library containing 54 kinds of electrical equipment such as lightning arrester, circuit breaker, and disconnecting switch was established, which contained the information of type, number and feature descriptor of each template device. The feature descriptor of the equipment point cloud to be identified was calculated and was used for matching the most similar model in the template library to realize the recognition of equipment point cloud. Finally, the method proposed in this study and another two substation point cloud recognition algorithms were tested on 90 equipment point clouds to be identified. The results showed that our method could achieve 90% recognition accuracy, and the average time to identify a device was 3.2 s, which could balance the recognition accuracy and recognition efficiency. And our method slightly outperformed the other two methods when equipment point cloud with noise and occlusion. Moreover, when the density of the point cloud to be identified was not uniform, our method could still maintain the recognition accuracy of more than 70%.

Abstract: It was necessary for the virtual synchronous generator (VSG) to be synchronized with the grid when it switched from off grid mode to grid connected mode. In this study, a modified pre-synchronization control method based on power matching and self-adaptive inertia for the VSG was proposed to improve its stability after grid connection. The proposed method took the amplitude and frequency of grid voltage as the reference value for the VSG output voltage, and then adjusted its output power reference value according to the local load power, thus the voltage frequency and amplitude of the VSG could be regulated to be synchronized with the grid. Besides, feedback control of phase deviation with PI controller was applied in the frequency control loop to achieve phase synchronization. With this method, the output power corresponding to the steady-state operation point of the VSG in grid connected mode was consistent with that before grid connection, and the output power step after the VSG connected to the grid could be avoided. While, when the VSG connected to grid in non-ideal situation, power and frequency oscillations would occur in the synchronization process, so time-varying inertia was introduced in this study to accelerate the VSG to stable state. A simulation model was constructed based on MATLAB / Simulink and simulation results indicate that the proposed method can maintain the same steady-state output power before and after the VSG connected to grid, besides, oscillation time after the VSG connected to grid in non-ideal situation could be reduced by about 60%, which proved the proposed method effective.

Abstract: To address the security problems caused by malicious nodes in wireless sensor networks, in this study, based on the Bayesian trust models, the adaptive reputation maintenance function was introduced to reduce the influence of the previous node and number of interaction, and the abnormal weakening factor was introduced to reduce the false detection of node by the abnormal behaviors caused by network faults, and combined with the fuzzy evaluation mechanism, to calculate direct trust. In order to improve the reliability of recommendation trust evaluation, the similarity measure theory was adopted to assign weight to different recommendation nodes and redistribute to obtain indirect trust. In order to improve the detection accuracy of the trust model, a weighted factor was adopted to determine the size of the comprehensive trust value jointly by variables in direct and indirect trust. Using the adaptive weighting dynamic updating comprehensive trust value, it could effectively avoid the rapid promotion of trust in a short time, and use the sliding time window to predict the comprehensive trust value. The WSN dynamic trust evaluation and prediction model integrating multiple indicators FSEPM was built. The difference between the predicted trust value and the actual trust value was compared with the trust threshold to judge the node property. Simulation results showed that the trust evaluation model could accurately and reliably evaluate the trust relationship between nodes, detect malicious nodes effectively, and improve the security of wireless sensor networks.

Abstract: Aim to the traditional artificial potential field algorithm was easy to fall into the local optimum, this study proposed an artificial potential field method which could improve the repulsive force model. According to the relative relationship between vehicle velocity direction and obstacle position and the distance between obstacle and road boundary, the repulsive deflection direction and deflection angle were determined to avoid vehicle falling into local optimum. The repulsive field function was redesigned to introduce the safety distance of vehicle obstacle avoidance into the scope of repulsive force. Logarithmic function and longitudinal relative distance adjustment factor were introduced into the repulsive force function to reduce the curvature and total rotation angle of the planned path. The simulation results showed that the selection of repulsive deflection angle had a direct influence on the stability and safety of the local path when the artificial potential field method of the improved repulsive model is used for local path planning, and the appropriate repulsive deflection angle could avoid the occurrence of local optimum in path planning. When the vehicle need continuous obstacle avoidance in multi-obstacle environment, the total path angle and curvature peak planned by the improved repulsive force model decreasd significantly, which could effectively improve the driving safety index.

Abstract: A detection method that could integrat field signature method (FSM) and electromagnetic ultrasonic transducer (EMAT) was proposed,aiming at the problem that the online monitoring data of metal pipe corrosion in the petrochemical industry was not reliable. Through the theoretical analysis of the measurement principle of FSM technology, the source of data error in the measurement process was identified, that was, the expression of offset voltage. The measurement results of FSM were compared with the measurement results of EMAT, and the measurement data of EMAT at a certain moment was taken as the true value of FSM data to obtain the corrected probe voltage value. Thus, the problem of data drift caused by zero offset voltage and data processing unit magnification change during long-term measurement was corrected, and the large-scale and long-term stable online monitoring of pipeline was realized. The experimental results showed that after two years of practical application, the proposed method could reduce the maximum error of FSM measurement data from 10.09% to 0.49%, and the maximum error was reduced by 9.6 percentage points.

Abstract: To address the difficult problem of coordinated control of body height and damper damping force during air suspension body height adjustment control, a hierarchical control strategy of air suspension body height and adjustable damping based on the mixed logic dynamic (MLD) model was proposed. Considering the hybrid characteristics of the air suspension in the process of inflating and deflating, the MLD modelling method was used to establish a nonlinear air suspension hybrid model with solenoid valves and magnetorheological dampers. The hybrid automaton describing the switching state of the solenoid valves was designed for the upper level control of body height adjustment, and the input current of the magneto rheological dampers was controlled at the lower level based on the predictive control method of the hybrid model. And then hierarchical control of body height and adjustable damping was realized by changing the switching state of the solenoid valves and the input current of the magneto rheological dampers. Through the simulation verification in random road excitation conditions, it could be concluded that the proposed control method could effectively track vehicle height while reducing vehicle acceleration by 34.33% and 34.34% compared to passive suspension and hybrid model predictive control, respectively, which could not only improve the ride comfort of the vehicle, but also directly prevent the frequent switching of solenoid valves and extending the service life of solenoid valves.

Abstract: In order to solve the problems of high toxicity of isocyanate in traditional epoxy polyurethane coatings, high brittleness, poor toughness, poor chemical resistance, and salt spray resistance of the cured coating, a new type of epoxy non-isocyanate polyurethane anticorrosive coating was prepared by using epoxy-cyclocarbonate synthesized by the reaction of silicone modified epoxy resin with carbon dioxide as resin base material (component A), polypropylene oxide diamine as curing agent base material (component B), and proper fillers and auxiliaries. The cured coating was characterized by FTIR and its performance was tested. The results showed that the coating could be dried at 80 ℃ for 20 min, with a smooth appearance and no cracking, blistering or peeling, and then after curing at room temperature for seven days, the hardness could reach 2H, the adhesion was 0 grade, and the salt spray resistance time could reach 1 500 h. The prepared silicone modified NIPU anticorrosive coating had good mechanical properties and corrosion resistance.

Abstract: In order to explore the effect of densification process on the microstructure and properties of C/C-Cu composite, the 2.5D needle-punched carbon felt was densified by chemical vapor infiltration (CVI) and precursor impregnation pyrolysis (PIP) to obtain C/C reinforcements filled with pyrolytic carbon and resin carbon, respectively. C/C-Cu composite was prepared by vacuum infiltration process, and the microstructure analysis and performance test were carried out. The results showed that the TiC interface layer in the C/C-Cu composite prepared by CVI process was thinner, and the pyrolytic carbon had a better protection effect on the carbon fiber. The resistivity and compressive strength in the vertical (parallel) direction were 0.72(0.63)μΩ·m, 367.61(326.87)MPa, respectively.And the tensile strength is 62.54 MPa. Its electrical conductivity, compressive strength, tensile strength, and plasticity were better than those of PIP process. The failure mechanism was fiber pull-out failure. The hardness value of C/C-Cu composite prepared by CVI process was 77.28 HBW, which was slightly lower than that of PIP process (81.59 HBW), but the difference between them was small. In conclusion, the CVI densification process was more suitable for the preparation of C/C reinforcements of C/C-Cu composites.

Abstract: In order to solve the problem of brake drum failure caused by frequent hot and cold cycles during continuous braking of heavy truck, a surface modification method, i.e., friction stir welding, was proposed to improve the thermal cracking resistance of the gray cast iron brake drum. In this experiment, the gray cast iron (HT250) brake drum was selected and the cyclic quenching treatment was used to simulate the service condition of the brake drum. The quenching treatment was set as 900 ℃ and the cooling method was water cooling. The Vickers hardness tester was used to measure the hardness changes of the samples before and after quenching treatment. Thereafter, the microstructure and crack morphology of the samples were characterized by the scanning electron microscope, energy dispersive spectrometer and metallographic microscope. It was found that frequent cold and hot cycling was the main reason for the failure of the brake drum. The experimental results showed that, after 20 cycles of quenching treatment, the cracks originated from the interface between graphite and pearlite, and then propagated mainly along the length of graphite. Besides that, the new phase of iron oxide is formed around the cracks. Moreover, the Vickers hardness of gray cast iron increased after cyclic quenching, but the hardness dispersion also increased. After the same numbers of cycle quenching, the number of cracks in the surface modified sample is less than that of the original sample, and the crack length was shorter. This is mainly because the graphite length width ratio on the surface of the gray cast iron decrease after the surface treatment, and the pearlite lamellar spacing decreased, which improve the resistance of crack initiation and propagation and thus enhanced the thermal cracking resistance of the gray cast iron brake drum.

Abstract: To explore the application of artificial intelligence technology, big data, and user profiling in the intelligent transformation of academic journal, in this study, the development status of digital intelligence and open access of academic journal publishing platforms at home and abroad were analyzed, and suggestions for the intelligent integration and development of academic journal publishing in China were proposed. Firstly, the construction of user profiles was based explor on data correlation, which consisted of four steps: data collection, data processing, label segmentation, and user profile presentation. Secondly, the intelligent reconstruction of the publishing process was proposed, which included efficient topic planning, intelligent review and evaluation, automated editing and proofreading, on-demand production and printing, and precision marketing services. Thirdly, the transformation of knowledge services into intelligence were accelerated, including the graphing of knowledge retrieval services, customization of knowledge recommendation services, intelligence of scientific research decision-making tools, networking of social media services, and sharing of open access services. Finally, a framework design for intelligent academic journal publishing platforms were proposed, providing ideas for the intelligent transformation of academic journal publishing.