Graphene oxide (GO) was modified by 1,4−cyclohexanediamine (CDA), and polyvinyl alcohol (PVA) was crosslinked by 4−sulfophthalic acid (SPTA). Hydrophilized polyvinylidene fluoride (PVDF) ultrafiltration membrane as support, SPTA−PVA/CDA−GO/PVDF composite membrane was prepared through vacuum filtration and drop coating. The composite membrane was characterized by ATR−FTIR, SEM and contact angle, and the pervaporation desalination performance of composite membrane was tested. The results show that the modification of CDA extended the interlayer spacing of GO nanosheets, thus improving the water permeability of CDA−GO layer, and the crosslinked PVA layer enhanced the surface hydrophilicity and stability of composite membrane, consequently, improved the water absorption and water transport properties of composite membrane. The composite membrane obtained a flux of 15.6 kg/(m²•h) and a rejection of 99.99% when treating with concentration of 3.5% NaCl solution at 70 ℃.

For a kind of gas burner, the appropriate size of the combustion chamber is selected to match, and then a complete calculation area was formed. Natural gas fuel and combustion air were injected into the combustion chamber through the gas burner to realize the complete combustion process. By simplifying the gas burner to complete geometric modeling, heat flow and combustion analysis software were used for numerical simulation of the calculation area, and the Realizable k-ε turbulence model, ED combustion model, DO radiation model and thermal NO_x generation model were selected to explore the influence of excess air coefficient α on the combustion process of the gas burner. The cloud image and distribution of temperature field and NO_x concentration field were obtained and analyzed. When the excess air coefficient α = 1.1, the NO_x concentration reached the maximum. The research results can provide reference for relevant research experiments and design of low NO_x burners.

An oil-gas pulse flushing method was proposed to remove contaminants in the pipe by changing the charging time to generate a pulse oscillating flow. The oil-gas pulse dynamic model and the mathematical governing equation were established, user defined function (UDF) secondary development based on Ansys Fluent module and simulation test were carried out. k−ε two equation turbulence model and volume of  fluid (VOF) model for gas-liquid two-phase flow were used, and Simple algorithm was carried to solve iteratively, the effects of radial velocity, turbulence intensity and dynamic pressure on z = 0 cross-section at different positions from the inlet, as well as wall shear force were studied. The results show that the flow velocity, turbulence intensity, dynamic pressure and wall shear force of oil-gas two-phase flushing are higher than that of single oil flushing. Pulse flushing compresses air to act on oil in the form of pulse, which enhances turbulence pulsation and increases inertia shear stress. The cleaning effect is better than continuous charging method. The result can provide theoretical basis and engineering practice method for pipeline oil-gas pulse flushing.

The summer air in the area with hot summer and cold winter is humid, so dehumidification is the key task of air-conditioning. Aiming at the problem of insufficient dehumidification capacity of the fresh air unit in the temperature and humidity independent control air-conditioning system, the double heat pipe fresh air unit was adopted to enhance dehumidification capacity of the air-conditioning system. The system was designed the fresh air volume of 300 m³/h, the theoretical dehumidification capacity of 662.4 g/h, and three-stage filter was used to purify the fresh air, which can meet needs of dehumidification and sanitation. A metal radiant ceiling was used to treat the indoor air temperature in the system, and the risk of metal ceiling condensation could be avoided through the monitoring system. It will provide a reference for temperature and humidity control and building energy conservation.

In view of the high cost of current underground pipeline trajectory measurement system based on inertial navigation principle, the cost of the system was effectively reduced by reducing the number of inertial sensors. Firstly, the formula of using only uniaxial angular velocity and biaxial acceleration data to restore pipeline trajectory was derived. Then, the complete ensemble empirical mode decomposition with adaptive noise was used to process the original data of inertial sensors. Finally, the pipeline trajectory was reconstructed by using the derived formula and the processed data. In the 75 m long test pipeline, the maximum deviation of the reconstruction track is less than 0.2% of the total length, and the cost of the sensor is reduced by half, which has strong practical value.

The permanent magnet synchronous motor (PMSM) has a high requirement on the smoothness of torque output. Due to influences of groove torque, reluctance torque pulsation, electromagnetic torque pulsation and other factors, the output torque pulsation of the interior PMSM is relatively large, which increases electromagnetic noise and output instability. A finite element model was established based on a 20 kW PMSM, and the accuracy of the model was verified. Taking the maximum output torque, the minimum torque pulsation and the minimum slot torque as optimization objectives, the sensitivity analysis of rotor structure parameters was carried out, and the stratified optimization was carried out by combining NSGA−Ⅱ algorithm and single variable parameter method. Compared with single-layer optimization, stratified optimization has a better effect on the target which is susceptible to variables. The acoustic simulation shows that the optimized rotor structure can improve electromagnetic noise significantly.

Based on the Volterra series, a hybrid Π type equivalent circuit was used to model the common-emitter amplification circuit according to the characteristics of the Class AB power amplifier. The nonlinear components in the amplifier were described and analyzed to obtain a model of the system behavior of the amplifier circuit. An optimized design method was proposed based on Volterra series and Kirchhoff's current law (KCL) to find the optimal bias parameters quickly. A RF amplifier operating at 850 MHz was designed and the optimal input bias size for operation taking into account both the efficiency and linearity of the amplifier was   calculated to be 6 ohms, and the output P1dB was 23.4dBm (the reference power is 1 mW). Multisim circuit simulation proves the results of the theoretical analysis.

The analytical inverse solution of 2UPR/UPS/UP+AC pendulum hybrid mechanism was derived. With 2UPR/UPS/UP mechanism as the research object, the single-joint error model and branched-chain error model were systematically expounded , and the related geometric error sources were given. The error mapping models among the passive error, the pose error of the moving platform and the error sources of the motion pairs of each branch chain were established, and the error sensitivity of the error sources of each motion pair was simulated based on the mechanism error model. The research results can provide theoretical support for motion error distribution and precision design of 5-DOF hybrid robot engineering design based on 2UPR/UPS/UP parallel mechanism, as well as reference for error modeling and sensitivity analysis of other hybrid mechanisms and parallel mechanisms.

Location and mapping is one of the key technologies for autonomous driving. With limitations of lidar sensors or vision sensors, the advantages of different sensors can be brought into play through multi-sensor fusion and the accuracy and robustness of location and mapping can be improved. The Harris algorithm was optimized for corner extraction, the key frame was used to optimize the feature point matching algorithm, and then the nonlinear least square method was used for back-end optimization. The location and mapping experiments were carried out on the test platform to verify the algorithm, and the positioning error was analyzed with the EVO tool. The result shows that the error of the proposed back-end optimization algorithm is 13% less than that of a single sensor.

In order to improve the adaptability and tracking stability of autonomous vehicles to different roads, a lateral and longitudinal trajectory tracking method based on road boundary constraints and dual proportional-integral-derivative (PID) was proposed. Based on the road boundary constraints, the steering curvature for safe driving was obtained, and then the safe driving steering angle was calculated by combining the two-degree-of-freedom dynamic model. The method is simple to calculate, and the farthest preview point is determined autonomously by the width and curvature information of the road. Through the simulation test, the calibration relationship table of vehicle acceleration, speed, accelerator and brake were made, and a dual-PID speed tracking controller was designed based on the calibration table. Finally, through the Carsim-Simulink co-simulation, it is proved that the lateral and longitudinal controller can safely drive within the road range, and has good tracking accuracy and stability.

The pantograph-catenary offline has many adverse effects on the safe operation of the subway. Firstly, the simulation model of pantograph-catenary offline traction system was established, and then the offline conditions of different durations were simulated. By analyzing the transient process of the DC side voltage, IGBT voltage and motor rotor current of the traction inverter circuit during the offline period, the differential effects of different offline time lengths on the electrical characteristics of the train traction process were obtained. The results have a good theoretical guiding role for the fault diagnosis of the traction system and the formulation of maintenance strategies.

According to shortcomings that the evaluation methods used in emergency drills are single in urban rail transit, there is strong subjectivity in the weights of evaluation indicators and it neglects of the objective links between them, an evaluation model based on AHP−risk network was proposed. The objective weights obtained from the risk network were used to correct the subjective weights obtained from the analysis hierarchy process, so that the weights of the indicators of the emergency drill evaluation were both professional and objective. Finally, it was applied to the evaluation of a large passenger flow of emergency drill in a station of Shanghai metro, and the evaluation results had verified the effectiveness of the evaluation model. The model can provide a combination of subjective and objective assessment method that considering objective links between evaluation indicators for urban rail transit emergency drills.

Under the normalization of the epidemic, the additional health detection link in the airport terminal has made the check-in process longer and inefficient. On the basis of analysis for current situation of check-in process of an airport terminal in China, a process model based on Petri network check-in process was established, and the correctness of the model was verified by reliability test. It is found that: 1) Self-service check-in, online check-in and counter check-in all make passengers check in their baggage for too long; 2) Compared with counter check-in, the other two check-in methods have a shorter waiting time for passengers to serve and more passengers choose. Finally, it is proposed to reduce the waiting time of passengers in line by adding self-service check-in machines or check-in counters, so as to alleviate and improve the bottleneck in the check-in process.

Nickel-based superalloy Inconel718 can maintain excellent mechanical properties at high temperatures above 600 ℃, so it is widely used in aerospace and other important scientific and industrial fields. It is a difficult-to-process material for situations of large drilling force, high temperature in drilling area and large tool wear during drilling process. A step drill was designed based on twist drill and the concept of diameter ratio (the ratio of the diameter of the first step to the second step) was introduced, and finite element simulations of Inconel718 drilling with two kinds of drill bits were carried out by using Deform−3D software. The results show that the step drilling can reduce the axial force by 25.5% on average compared with twist drilling, the maximum tool wear can reduced by 33.9%, the temperature of the workpiece drilling area was more lower and the bit temperature distribution under the same drilling environment was more better.

Thermal deformation behavior of 304 stainless steel was investigated by utilizing Gleeble−3800 thermo-mechanical simulator. The true stress-strain curves of 304 stainless steel were obtained through the hot compression experiment with temperature range of 800~1200 ℃ and the strain rate range of 0.001~1 s⁻¹. The results show that the flow stress of 304 stainless steel increases with the reduction of temperature and the growth of strain rate during the hot deformation. According to the true stress-strain curves, the Arrhenius constitutive equation of 304 stainless steel was constructed. The hot deformation activation energy is 490 kJ/mol and the linear correlation coefficient between lnZ and ln

is 0.93, which shows that the stress values predicted by the model are in good agreement with the experimental ones.

Using the interferograms obtained by Mach-Zehnder interferometry, it was introduced how to use Interferometrical data evaluation algorithms (IDEA) software to realize processes of Mask, 2D Fast Fourier Transformation and Phase Unwrapping, thus phase shift distribution diagrams corresponding from the interferograms were obtained. The results show that the phase shift of the interferograms can be extracted by simple menu commands in IDEA. Meanwhile the optical noise and shadows in the interferograms can be effectively reduced. Therefore, the phase extraction based on IDEA has certain significance in the application of optical interferometry.

In order to improve the pressure comfort of indoor climbing tight sportswear under different technical actions, four tight-fitting climbing sportswear were designed through different structural divisions and fabric splicing, and the dress comparison experiment was carried out with the popular two in the market. The pressure value of each test point was analyzed, and the point that exceeded the pressure comfort threshold was determined. Through correlation analysis, it was found that the main pressure parts were the shoulders, waist, knees and hips. The human rock climbing action was captured by video software, and the principal component analysis of the test points of the main pressure parts was carried out to determine the 4 main technical actions that affected the pressure. Finally,  weight analysis of the pressure value of each test point of the 6 experimental garments under the 4 main technical actions was carried out. The results show that these methods mentioned can effectively improve the pressure comfort of climbing sportswear, including the use of raglan sleeve design and dividing line design on the front chest, back width and other parts, appropriately increasing the bending amount of the front knee and reducing the looseness of the back of the knee.