Abstract:Gear bending fatigue test is an important means to evaluate gear bending fatigue performance. In this work, the effects of test process errors on test results were investigated to ensure the accuracy of test results. According to the method B in the national standard GB/T 14230-1993, the bending fatigue test scheme was made with symmetrical loading across five teeth. The finite element simulation and strain gauging were carried out to verify the correctness of the test scheme. The bending fatigue test of 18CrNiMo7-6 shot peened gear was performed by D-M method. The effects of gear geometric accuracy, clamping error and equipment loading accuracy on the bending fatigue test results were analyzed. The results show that the maximum error between the calculation of bending stress of tooth root and the results of finite element simulation and strain gauge stress analysis was 8.7%, which meets the engineering requirements. Under 99% reliability, the bending fatigue limit of 18CrNiMo7-6 shot peened gear was 642 MPa, which is 28% higher than 500 MPa of MQ grade in the GB standard, suggesting that the standard design is conservative. The geometric accuracy of the gear and the accuracy of the test equipment have great effects on the bending fatigue test results. Considering the comprehensive effects of the errors, the effect on the test results is 2.43%.

Abstract:To deal with the problems of scarce data of machine tool wear and low recognition accuracy of tool wear status, a tool status recognition method based on sample expansion and improved domain adversarial training of neural networks (SE-IDANN) was proposed. First, to solve the problem of scarce machine tool wear data, two feature extractions on the machine tool data were performed, and the sample was expanded through the Smooth algorithm. Secondly, a residual block was added to the domain adversarial training of neural networks (DANN) feature extractor to further extract effective feature information and solve the problem of weak tool wear characteristics. Finally, to realize the accurate identification of tool wear, the Wasserstein distance used as the data distribution similarity standard between the target domain and the source domain was introduced into the DANN model. Through the analysis and test verification of machine tool data, it is proved that this method can better identify tool wear.

Abstract:The vine copula model provides a new way to describe the nonlinear and non-Gaussian dependence of high-dimensional data and has attracted more and more attention in the field of chemical process modeling. In this article, a novel chemical process fault detection method, LASSO-R-vine copula (LRVC), is proposed by introducing LASSO (least absolute shrinkage and selection operator) regression into R-vine copula. LRVC determines the position of the variables in the R-vine matrix according to the strength of the relationship between the variables, using regression to analyze the regularization path and select the R-vine copula matrix structure. The R-vine structure matrix model is determined to obtain a sparse matrix model related to variables' independence by following the R-vine matrix construction rules and regression process. The matrix structure constructed by this method is independent of the copula function type and parameters. When dealing with high-dimensional complex industrial process data, sparse models and penalties could simplify the copula function type's selection process, shorten the modeling time, and make the statistical modeling more flexible. This method shows an excellent predictive effect in TE and the acetic acid dehydration process fault monitoring, proving its effectiveness in nonlinear and non-Gaussian processes.

Abstract:With the integration of renewable energy resources, electrical vehicles, multiple energy coupling equipment and other new elements, distribution network has new characteristics, including diversification, high uncertainty and complexity. This paper reviews the research on the planning methods for new equipment integrated distribution network. Firstly, the simulation approaches of generation power and load power used in distribution network planning are investigated. Secondly, the distribution network planning models considering different factors are introduced. Several optimization strategies and solution algorithms for distribution network planning are also reviewed. Finally, with considering the state-of-the-art and the developments of AI technology and energy internet, the prospect of distribution network planning theories and technologies under new conditions are presented.

Abstract:After the strong wind passed through the windshield wall along the Lanzhou-Xinjiang high-speed railway, the catenary positive feeder was galloping. In order to analyze the influence of the surface roughness of the stranded wire on the aerodynamic characteristics of the positive feeder under the action of the wake flow of the wind-break wall, a windshield wake wind tunnel experimental device and five wire models with different degrees of surface roughness were established based on fluid mechanics. Due to the large gap between the size of the wire and the overall computing domain, the fluid grid of the overall computing domain is divided into blocks. The fluid simulation software is used to study the aerodynamic characteristics of the forward feeder under different inlet wind speeds. The results show that the lower the surface roughness of the positive feeder is, the more obvious the lift and drag increase. When the inlet wind speed is 1m/s, the basic values of the lift and resistance coefficients of positive feeders with different degrees of surface roughness remain stable. When the inlet wind speed is greater than 5 m/s, the lift and drag coefficients of the positive feeder increase as the surface roughness of the positive feeder decreases. The vortices generated in the tail flow field of positive feeder with different degrees of surface roughness are different, and different small vortices are generated at the concave-convex of the wire. The air flow characteristics near the wall of the positive feeder have changed greatly, and the influence on the aerodynamic characteristics of the wire is obvious.

Abstract:To address the defect of low catalytic performance due to the two-phase mass transfer resistance in heterogeneous photo-Fenton, a hydrothermal method was used to prepare catalysts α-Fe₂O₃ that exposed different ratios of crystal plane {110}/{113}. α-Fe₂O₃ was characterized by XRD, TEM/HRTEM, FTIR and UV-vis DRS. The effects of H₂O₂ concentration, catalyst dosage, initial concentration of methyl orange and initial pH on α-Fe₂O₃ on methyl orange catalytic degradation were investigated. The results showed that the α-Fe₂O₃ achieved the best catalytic performance when the concentration of precursor FeCl₃·6H₂O was 33 mmol/L, the dosage ratio of FeCl₃:6H₂O:NaH₂PO₄:NaF was 1:1:15, and the reaction temperature was 180℃. Moreover, at the H₂O₂ dosage of 20 mmol/L, catalyst dosage of 200 mg/L, initial methyl orange concentration of 40 mg/L and initial pH of 5, the removal of methyl orange was the best, reaching 97.83% in 60 min. The catalyst exhibited good stability after consecutive degradation cycles. The H₂O₂ consumption analysis and free radical trapping experiment results showed that hydroxyl radicals and photo-generated holes played a key role in the catalytic degradation process, while superoxide radicals had less effect.

Abstract:In the intersecting system around Dianchi Lake, a mixed water purification plant was built with MBR process. To ensure that the nitrogen and phosphorus concentration of the MBR effluent met the discharge standard stably, a pilot-scale system was built on site to explore the effects of supplymentary carbon source (sodium acetate) and phosphorus removal agent (polyaluminum chloride) dosage on the enhanced removal of nitrogen and phosphorus in actual sewage, respectively. Due to perennial low concentration of influent water quality, low carbon-nitrogen ratio and carbon-phosphorus ratio of the agricultural non-point source sewage of Dianchi Lake, as well as large fluctuation of the TN and TP concentrations, the flexible use of the dual control method of external carbon source (sodium acetate) and auxiliary chemical phosphorus removal (polyaluminum chloride) or the single control method of auxiliary chemical phosphorus removal was proposed, which could save operating costs with the precondition that the nitrogen and phosphorus in the effluent could be up-to-standard discharged stably. At the same time, based on the pilot study parameters, the actual operation and regulation efficiency of the Luolong River Purification Plant was effectively improved, the optimization and adjustment time of the entire system was shortened, and the stability of effluent water quality reaching the discharge requirement in the purification plant was finally achieved. Especially the TP concentration was stably lower than the local standard requirement of 0.3 mg/L. It was estimated that the two purification plants with MBR process around Dianchi Lake could reduce TP emission up to 65.3 t per year.

Abstract:To obtain the mechanical properties and stress-strain characteristics of traditional adobe and determine the mathematical expression of its variation relationship, a series of uniaxial compression tests and three-point bending tests were performed for molded adobe and compacted adobe. The mechanical properties of the two kinds of adobe were analyzed and compared by the compressive strength, flexural strength, failure mechanism, and stress-strain relationship. The results show that the test method for compression and bending of bricks can be applied to adobe. The compressive strength and flexural strength of compacted adobe are 3 times and 1.3 times of molded adobe, respectively, but the breaking energy of molded adobe is about 2.5 times higher than that of compacted adobe. In the early stage of compression, the stress-strain curve of compacted adobe has a concave section due to compaction of earth material, while this phenomenon has not been observed in molded adobe. The uniaxial compression constitutive model was proposed according to the stress-strain characteristics. This model shows a good correlation with the experimental data and may be applied to the numerical analysis of adobe structure.

Abstract:The bridging effect of steel fiber and its synergistic effect with rubber particles can significantly improve the performance of concrete, but there are few studies on the static mechanical properties and constitutive relationship of self-compacting concrete mixed with steel fiber and rubber. To prepare self-compacting concrete specimens with steel fiber and rubber, the rubber concrete with 10%, 20%, and 30% rubber powder content were mixed with the steel fiber content of 0.5%, 1.0%, 1.5%, respectively. Prismatic axial compression tests were carried out to investigate the effects of steel fiber content and rubber powder content on the typical mechanical characteristics of specimens and the constitutive relationship of the axial compressive strength of the self-compacting concrete mixed with steel fiber and rubber was put forward. The results show that the failure of self-compacting concrete specimen mixed with steel fiber and rubber presents ductile failure, while its integrity is good. With the increase of rubber powder content, the axial compressive strength of the specimen gradually decreases. When the rubber powder content is high (20%, 30%), steel fiber has no obvious enhancement effect on the axial compressive strength of self-compacting concrete with rubber. The addition of rubber powder increases the peak strain of the specimen. The peak strain of the specimen reaches the maximum (an increase of 10%) when the amount of rubber powder is 10%. Different amounts of steel fiber can generally enhance the peak strain of self-compacting concrete mixed with steel fiber and rubber, but it does not show obvious regularity. The axial compressive constitutive relationship can reflect the mechanical properties of self-compacting concrete mixed with steel fiber and rubber powder.

Abstract:This paper proposes a method to predict the actual cracking of the asphalt pavement under the condition of 5 to 10 years through the material performance test. Based on the relationship between the cracking index and the low-temperature performance of asphalt materials, the 180 s stiffness modulus values of 5 types of commonly used asphalts are obtained by fitting the CAM model, and the cracking indexes of various asphalts under the PAV state are predicted and proposed. The accuracy of this simulation method is verified by actual data. The investigation on the cracking situation of the core samples of the Changji Expressway supports the prediction effect of the CI value on the actual cracking level of the asphalt pavement. The results show that the 180 s modulus calculation method of the asphalt beam obtained under the CAM model is simple and the data reliability is high. The predicted SBS modified asphalt cracking index is consistent with the actual cracking level of the pavement, which also proves the feasibility of the use of material performance to predict the actual crack resistance of the asphalt pavement. According to the standard practice, the CI values of five types of asphalt commonly used in northern China are given.

Abstract:To address the shortcomings of the existing smartphone user authentication methods, this paper proposes a continuous identity authentication method based on adaptive weight feature fusion. A convolution neural network is designed to extract and fuse the deep features of user behavior information data obtained from the built-in sensors (accelerometer, gyroscope, magnetometer) of mobile phones. In the network, three kinds of sensor features are extracted from the three sub-network flows respectively, and weighted fusion is performed in the feature fusion layer. The weight of each feature is adaptively assigned according to the contribution of different features in the network learning process. After feature selection for fused features, one-class support vector machine is used for user classification and authentication. The experimental results show that this method achieves an equal error rate of 1.20% for different user authentication. Compared with other existing authentication methods, the proposed method demonstrates better authentication accuracy.

Abstract:Aiming at the problem of channel mismatch in the field of voiceprint recognition and incomplete acquisition of voiceprint features under short speech or noise conditions,a method that combines traditional methods with deep learning is proposed, and the ResNet model is used as the student model to perform knowledge distillation on the I-Vector model as the teacher model. We construct a ResNet network based on metric learning, introduce an attentive statistics pooling layer, capture and emphasize the important information of voiceprint features, and improve the distinguishability of voiceprint features. The mean square error (MSE) is combined with the loss based on metric learning to reduce computational complexity and enhance model learning capabilities. Finally, the trained model is used for voiceprint recognition test, and compared with the voiceprint recognition model under a variety of deep learning methods. It's found that the equal error rate (EER) is reduced by at least 8%, and the equal error rate has reached 3.229%, indicating that the model can perform speaker verification more effectively.

Abstract:The parameter adjustment of the washout algorithm greatly affects its performance. To address the shortcomings of the current classical washout algorithm when it is applied to flight simulators, an improved washout algorithm based on MOEA/D multi-objective optimization was proposed. Taking the human perception error, washout displacement and simulation acceleration error as the optimization objectives, and using the motion space as the restriction condition, the parameters of the washout algorithm were optimized synchronously by adopting the multi-objective optimization algorithm based on decomposition, and the optimal solution was obtained by using the fuzzy membership function. In order to verify the effectiveness of the improved algorithm, a flight simulator motion test platform was established and the washout algorithms with different optimization methods were applied to the platform for comparative analysis. Simulation and experimental results show that the improved algorithm had the shortest return reaction time. Moreover, the phase delay was reduced by 3.5 s, the stability increased by 30%, the sensory peak value was corrected, and the workspace was optimized by 54.6%, so that the simulator has enough movement space and higher dynamic fidelity when simulating extreme motion.