Abstract:

Abstract: An examination on studies of the control of a two-degree-of-freedom motor can improve its performance, and is of high theoretical and practical value. This study reviewed the control strategies for various types of two-degree-of-freedom motors, including induction, permanent magnet, switched reluctance, and other types, based on their structures. The study also classified the control strategies used in these motors as open-loop or closed-loop, analyzed the advantages and disadvantages of different strategies, and summarized the application development of two-degree-of-freedom motor control based on the requirements of relevant fields such as automobiles, medical, industrial, robotics, new energy, and military. Finally, based on the theoretical research and practical needs, this study discussed the future development trends of two-degree-of-freedom motor control from four aspects: decoupling, systematization, intelligence, and low-speed high-precision control.

Abstract: Code clone detection, based on deep learning, is often applied to use the model to extract features in the sequence of tokens or the entire AST. That may lead to the missing of important semantic information and induce gradient disappearance. Aiming at these problems, a method of code clone detection based on Transformer and CNN was proposed. First of all, source code was parsed into AST. Then the AST was cut into statement subtrees, which were input into the neural network. Statement subtrees were composed of a sequence of statement nodes obtained by pre-traversal, which contained the structure and hierarchical information. In terms of neural network design, Encoder of Transformer was used to extract global information of the code. CNN was used to capture the local information. Fusion of features were extracted from two different networks. Finally, a vector containing lexical, syntax, and structural information could be learned. The Euclidean distance was used to represent the degree of semantic association. A classifier is trained to detect code clone. Experimental results showed that on OJClone dataset, the Precision, Recall, and F1 values could reach 98. 9%, 98. 1%, and 98. 5%, respectively. On BigCloneBench dataset, the Precision, Recall, and F1 values could reach 99. 1%, 91. 5%, and 94. 2%, respectively. Compared with the relevant methods, the Precision, Recall, and F1 values were all improved. This method could effectively detect code clone.

Abstract: In order to solve the privacy leakage problem caused by adaptive clustering by fast search and find of density peaks(AdDPC) when calculating the local density and the primary allocation strategy, a differential privacy protection scheme of an improved density peak clustering algorithm was proposed. In this scheme, the Laplace random noise was added in the process of calculating the local density of the algorithm. In this way, even if the attacker had the maximum background knowledge, it could not obtain the corresponding information by adding or deleting a point in the dataset, thereby, differential attack was used to obtain the information of the target data point, and to achieve the purpose of protecting the privacy data. In addition, the reachability definition was introduced to improve the allocation strategy of AdDPC when assigning non-clustered center points, so as to avoid the problem of data point allocation error caused by the one-time allocation strategy. The experiment compared F-Measure and ARI values of DP-rcCFSFDP, AdAPC-rDP, IDP K-means, and results showed that: when the privacy budget was greater than 1. 5, the F-Measure and ARI values of the proposed algorithm were better than those of other algorithms, and this algorithm could protect sensitive data and data availability at the same time.

Abstract: In order to overcome the disadvantage that the redirected least squares regression model might destroy the structure of the regression target, a low-rank sparse representation based elastic least squares regression learning ( LRSR-eLSR)model was proposed. The model based on the least squares regression, but did not use the strict 0-1 label matrix as the target matrix. Instead, it introduced the margin constraint to directly learn the regression objective from the data, which could increase the flexibility of the regression model while maintaining the regression target structure. Moreover, in order to capture the structure information of the data, a low rank representation of the data was used to maintain the structure of the data. In the process of calculation, considering the complexity of the problem, the kernel norm regularization was used instead of the rank function. In addition to this, the model introduced a sparse error term with a L2,1 -norm to compensate for regression errors, which facilitates learning more flexible transformations. The model also imposed additional regularization terms on the projection matrix to avoid overfitting. The experimental results showed that the recognition accuracy of the model in this paper is better than that of other methods on four published datasets. The recognition rate could be up to 98% in the COIL-20 dataset.

Abstract: The outlier detection algorithm based on the nearest neighbor is sensitive to the selection of the nearest neighbor. Too small neighborhood range will increase the complexity of the model, resulting in over-fitting; Too much neighborhood will make the model too simple and ignore a lot of available information. In order to reduce the influence and achieve higher accuracy, a voting decision algorithm was designed based on the neighbor relationship. This algorithm consisted of two steps: density estimation and simulated voting. The density estimation was used to accelerate the density of convergent data to obtain the steady-state density, so that the simulated voting of different strategies could be carried out according to the steady-state density. Simulated voting strategy was the core algorithm of outlier detection based on the improvement of community discovery algorithm, and the importance of data points and the similarity of their neighbors to vote were taken into account. The importance of data points was positively correlated with their steady-state density. The data points with greater importance would have priority to vote actively, transmit their own information to the data with the greatest similarity in the neighborhood, and accumulate the voting ranking of the voted data. After each data has took the initiative to vote, the algorithm stopped iteration and obtained the voting ranking of each data point. The data with lower voting ranking was regarded as outlier. The experimental results on 11 real data sets showed that the average accuracy of the voting simulation detection algorithm based on the nearest neighbor was 79%, which could prove the effectiveness of the algorithm.

Abstract: Satisfiability ( SAT) problem is the basic problem of artificial intelligence, which is also a hard problem of NP. It has practical applications in machine learning, pattern recognition and natural language processing. However, with the development of artificial intelligence, more and more complex problems popped up. The original algorithms were no longer applicable and need to be further optimized or improved, which put forward higher requirements for basic research. In order to study the inherent nature of the difficult SAT problem, the structural characteristics of the problem were studied, and then the efficient algorithm to solve the SAT problem were found out. Phase transition, tree width, structural entropy, and DNA origami were four metric models for studying the structural characteristics of SAT problems, which attracted the attention of researchers in recent years. In order to clarify the research progress on the structural characteristics of SAT problems, based on the above four measurement models, the structural characteristics of SAT problems were summarized, and the challenges and future directions of the research on structural characteristics of SAT problems were pointed out. Although some research achievements were made in phase change analysis, tree decomposition algorithm, structural entropy and DNA origami in solving SAT problems, breakthroughs are still needed in solving the accurate upper bound of phase change points, guiding the design of SAT solvers by structural metric model, and improving the efficiency of tree decomposition algorithm, which will become the focus of future research on structural characteristics of SAT problems.

Abstract: Regarding to the low efficiency, a bypass structure was set to improve the internal fluid of a traditional ejector. Computational fluid dynamics was conducted to study the influence of the structure and the working parameters of the bypass on the flow field, including the bypass inlet angle, width and the flow pressure. The optimum matching parameters was also investigated. Based on the simulation results, when all other parameters of the ejector were fixed, the entrain ratio increased firstly and then decreased with the increase of the inlet angle, the width and the flow pressure of the bypass, respectively, and there existed the optimum parameters. In this investigation, the optimum inlet angle, width and the flow pressure of the bypass was 15°, 3 mm and 618. 5 kPa. Compared with the traditional ejector, the entrainment ratio of the bypass ejector could be increased by 25. 7% - 56. 8% in the same conditions.

Abstract: To address the precision problems in wind turbine fault diagnosis and maintenance processes, the lack of management of fault domain knowledge, and the large amount of historical fault data records left behind, a wind turbine fault diagnosis system was proposed to be constructed in the form of a knowledge graph. Firstly, knowledge extraction of fault texts was carried out by an improved named entity recognition model BERT-BiLSTM-CRF. The data set used text data of wind turbine fault cases and accident analysis in the past 10 years. And it was proved through experiments that the improved entity recognition method was 2. 54% more effective compared to the traditional model in the wind turbine fault domain. The extracted knowledge entities were then structurally analysed. As the traditional fault tree lacked purpose in actual fault reasoning, and each bottom event had different levels of importance to the top event, it was proposed that the characteristic attributes of the fault was introduced, as branching conditions, into the fault tree reasoning, to complete the fault tree qualitative and quantitative analysis, and the fault mode impact and hazard analysis ( FMECA) were combined to refine the fault domain knowledge model. Then Protg development tools were use to complete the ontology modelling of the fault tree structure based on the six-tuple concept, so that the constructed ontology knowledge base could meet the prerequisites for inference. Finally, the visualization of knowledge model was realized by Neo4j, and the ability of reading and writing knowledge data was improved.

Abstract: To accurately model and simulate the multibody system including large deformation beams, a 2D large deformation curved beam element, which had few element parameters and can be integrated with the CAD geometric model, was proposed based on cubic spline interpolation. Firstly,by taking the position vector at each node and the axial position gradient vector at the two-end nodes as the global parameters, the motion of the beam centroid line was approximated using the cubic spline interpolation. Secondly, on the basis of fully considering the deformation of the slender beam, the rotation of the beam cross-section was determined by the centroid tangent according to the deformation assumption of Euler-Bernoulli beam. Finally, the axial strain and curvature of the beam were derived based on the geometrically exact beam theory, the mass matrix, nodal force and generalized external force of the planar curved beam element were derived based on the virtual power principle, and the tangent stiffness matrix was obtained. Compared with the existing large deformation planar beam element, the number of the global parameters of the proposed beam element was greatly reduced. In addition, because the vertical relationship between the beam cross-section and the tangent direction of the centroid line was guaranteed in the proposed coupling deformation field, the shear locking phenomenon in the application could be avoided. In the numerical example, through the calculation and comparison of typical examples including the static geometric nonlinear problems and dynamic rigid flexible coupling problems, it was shown that the calculation efficiency of the proposed geometrically exact curved beam element could be greatly improved while not losing the calculation accuracy.

Abstract: In view of the defects of accuracy and robustness caused by the uncertainty of noise and model parameters in the process of generator dynamic state estimation, a robust dynamic state estimation method for generators—H-infinity unscented particle filter (HUPF) was proposed. Firstly, a fourth-order dynamic state space model of generator was established. Secondly, the uncertainty constraint criterion of model was constructed based on the H-infinity theory to define the uncertainty boundary range. By effectively combining robust control theory and particle filtering, and using unscented transformation to calculate the important density function, the particle swarm would be closer to the actual posterior probability distribution. Finally, a novel estimation error covariance update strategy was designed, which could be dynamically adjusted based on model uncertainty. In IEEE 39-bus system, the effectiveness of the proposed method was verified. The simulation results demonstrated that the minimum root mean square error (RMSE) of the proposed HUPF method was 0.006 and the maximum was 0.045 8. Compared with UKF, UPF, and AUKF methods, the HUPF method had the smallest RMSE and could significantly improve the state estimation accuracy of the generator with model uncertainty and stronger robustness.

Abstract: In the optimization operation of electric heating networks based on energy hubs, the neglect of the characteristic of energy efficiency of energy hub equipment changing with its input or output power leads to a significant deviation between the optimization operation results and the actual results. To solve this problem, firstly, the piecewise linearization method was used to approximate the variable energy efficiency of the equipment, and a standardized matrix model was established to characterize the energy hub considering the variable energy efficiency of the equipment. Secondly, with the goal of minimizing the total cost of system energy procurement, a linear model for optimizing the operation of the electric heating network was established by coupling the regional heating network and distribution network considering the energy hub with variable energy efficiency of equipment. Finally, numerical examples were used verifying that the optimized operation of the electric heating network is related to the variable energy efficiency of the energy hub equipment. The operation situation of the electric heating network based on the variable energy efficiency of the equipment,compared with the operation of the electric heating network based on constant energy efficiency of the equipment, the relative error in operating cost dropped from 12.71% to 0.03%, the wind power consumption rate increased by 4.29%。

Abstract: It was difficult to estimate the state of charge of LiFePO4 battery due to a large voltage platform area and voltage and current measurement errors. In order to improve the accuracy of estimation of the state of charge of lithium iron phosphate batteries in voltage platform area, an improved Kalman filter algorithm based on amper-hour integral method was proposed. Firstly, amper-hour integration method and AUKF algorithm were used to estimate SOC of lithium iron phosphate batteries. Secondly, the increment of the estimated value of the two algorithms was calculated. The characteristics of the two estimation algorithms were used to determine the optimal estimated value by comparing the increment relationship, and the estimated result of the AUKF algorithm was corrected. Finally, the effectiveness of the proposed method had been verified in various operating conditions of lithium iron phosphate batteries. Experimental results showed that the proposed method could keep the SOC estimation error less than 0.02 under the condition of voltage deviation, and achieve accurate SOC estimation.

Abstract: To study the mechanical properties of magnetorheological damper (MRD), 11 groups of MRD mechanical properties tests with axially cyclic loading were completed. The effects of current size, peak displacement, loading rate and current control mode were considered. The variation law of damping force-displacement curve and characteristics of each stage of MRD were analyzed. The results showed that the damping force at the peak point of MRD increased linearly with the current. When the peak displacement was less than the initial displacement, the MRD peak damping force was small, and when the peak displacement was greater than the initial displacement, the MRD peak damping force was relatively stable; the damping force at the peak point of MRD increased with the increase of loading rate; when the current was closed, the damping force of MRD would be rapidly reduced to 0 A. If the power was continued, the damping force of MRD could still reach a stable state after a short loading displacement. Finally, the calculation model of MRD damping force-displacement curve considering the influence factors such as current size, peak displacement, loading rate and current control mode was established. It was verified that the calculation curve was in good agreement with the experimental curve.

Abstract: Aiming at the problems of cracking and the durability reduction in hot and humid areas of southern China, basalt fibers were selected as concrete reinforcement materials. In simulated standard, hot, and humid curing environments indoors, the plastic shrinkage test, drying shrinkage test, humidity distribution test and circular ring restrained test were designed to explore the early cracking behavior of basalt fiber bridge deck concrete in hot and humid environment. The results showed that plastic cracking of basalt fiber bridge concrete was restrained in hot and humid environment. Compared with the reference concrete, its total cracking area per unit area was significantly reduced by 76.5%. The change of relative humidity in basalt fiber bridge deck concrete was the driving force for its development of drying shrinkage deformation. The decrease of relative humidity in the concrete slab could be slowed down by adding fiber, so as to inhibit its drying shrinkage cracking. The 28 d drying shrinkage deformation at the center and corner of the fiber reinforced concrete slab was 27.5% and 25.5% lower than that of the reference concrete. Besides, the incorporation of basalt fibers could inhibit the restrain cracking of concrete rings and improve cracking resistance of concrete.

Abstract: In order to establish an accurate and reasonable probability model for extreme value of vehicle load (VL) effect, based on the existing extreme value theories and models of VL effect, a novel prediction model named block over-threshold model was proposed. Firstly, the over-threshold model was established by making full use of the existing monitoring information according to POT method; and then the parameter correlation between the overthreshold model and the block maximum model was studied. A bridge was constructed for their conversion through a point process,afterward the block maximum model was derived from established GPD model. Thus the maximum distribution of VL effect in any T period was predicted. Finally, the established block over-threshold model was tested, and the probability density of VL strain in 10, 40, 70 and 100 a was extrapolated by using the model in combination with bridge design service life. The data collected by five steel longitudinal sensors in mid-span section of a cable-stayed bridge were analyzed and modeled. The results showed that the P-P diagram and Q-Q diagram in the model diagnostic plots all had very good linearity, the points described by the empirical recurrence level were within the 95% confidence interval of the recurrence level, and the empirical probability density histogram also fitted perfectly with the corresponding GP distribution. They all proved that the model could well simulate and predict the extreme value of VL effect with actual traffic flow.

Abstract: In order to study the influence of weathering degree on the shear strength characteristics such as shear strength and shear deformation of sawtooth structural planes with different undulating angles, four kinds of regular sawtooth structural planes with undulating angles of 15°, 25°, 35° and 45° were made. The rock samples with different weathering degrees were prepared by indoor weathering simulation test, and the direct shear test of indoor structural plane under different normal stresses was carried out to explore the variation law of shear strength and parameters of limestone sawtooth structural plane. The test results showed that shear stress-shear displacement curve of serrated structural plane showed peak shear type, which increased approximately linearly at first, then decreased after reaching the peak shear stress, and finally tended to be stable and maintained a certain residual shear strength. With the increase of weathering degree, the shear strength of limestone sawtooth structural plane showed a decreasing trend, and the shear strength parameters had different degrees of deterioration, and the deterioration degree of cohesion was greater than that of friction angle, while the fluctuation angle of structural plane was positively correlated with the shear strength. Based on the model test results, the shear strength estimation model of weathered sawtooth structural plane was established to reveal the quantitative relationship between shear strength and weathering degree and fluctuation angle.

Abstract: To discuss the plastic deformation ability of UHPC-NC hybrid pier reinforced by high-strength steel bars, a ductility analysis model was established and calibrated by the corresponding pseudo-static test of concrete piers strengthened with different UHPC reinforcement heights and UHPC box piers with different loading angles, with the involving in the constitutive model of Mander and double line, the constitutive relationship of UHPC material and the damage plastic model of concrete in the finite element program ABAQUS. Based on the validation of the test results, the method on determining the length of the plastic zone was proposed, and the parameter analysis varying in the axial compression ratio, the diameter and yield strength of the longitudinal reinforcement, and the height of the specimen was conducted. On the basis of the applicability of the proposed formula in the evaluating code, the formula for calculating the equivalent plastic hinge length of hybrid pier was established. The results showed that, the length of plastic zone for UHPC-NC hybrid pier reinforced by high-strength steel bars would decrease with the increase of axial compression ratio, increased firstly and then decreased with the diameter of the longitudinal reinforcement and longitudinal reinforcement yield strength, and increased with the increase of the specimen height. When the axial compression ratio was close to 0. 5, the failure of UHPC occurred simultaneously with the yield of longitudinal bars in the tensile zone. When the longitudinal reinforcement diameter and yield strength were 16 mm and 500 MPa, the energy dissipation capacity of the plastic zone reached the optimum. By comparing with the numerical analysis results, it was further shown that the empirical formulas underestimated the plastic deformation ability of UHPC-NC hybrid pier. The established equivalent plastic hinge regression formula could provide a reference for the seismic research of UHPC-NC hybrid pier reinforced by high-strength steel bars.