Abstract:A comfortable operating environment is important for the maintenance of operational capability. An evaluation method driven by cognitive efficiency of oceanauts was proposed to quantitatively evaluate the operation comfort of manned submersible operating system, which was based on the comfort characteristics of the display control interface, operation panel, and communication equipment in the system. The cognitive process of oceanauts was analyzed based on the practical operation, and the relationship between the cognitive performance, efficiency, and comfort of oceanauts was established. The cognitive efficiency model of oceanauts was constructed by using utility function according to the comfort experiment data of operating system. Support vector machines (SVM) were adopted to comprehensively investigate the influence of sub-system attributes, cognitive efficiency, age, educational background, and other factors of oceanauts on sub-system comfort perception. Through the expert weight survey method, the sub-system comfort coupling was realized and the operating system comfort was determined. Results show that the operating system could maintain a higher comfort level under the conditions that the brightness, contrast, and natural saturation of the display control interface were within [-5,0], [-50,-25], and [-5,0], respectively, the operation panel design had reasonable layout parameters and centroid position, the environmental noise frequency was low, and the communication voice sampling frequency was 48 000 Hz in communication equipment. The study indicates that the cognitive efficiency and subjective comfort have the same variation trend, and it is feasible to judge the comfort of the cabin operating system based on the cognitive efficiency of the submarine oceanauts.

Abstract:In view of the problem that the attitude estimation of UAV navigation system in dynamic environment is easily interfered by sensor noise and motion acceleration, a new attitude estimation algorithm of UAV considering motion acceleration interference was proposed. First, an acceleration estimation model was established. The acceleration error model based on Kalman filter and the velocity information provided by the external sensor were combined to accurately estimate the motion acceleration. The estimated motion acceleration was used to correct the original value of accelerometer, so as to reduce the interference of motion acceleration in the attitude estimation of navigation system in dynamic environment. Then, an attitude estimation model based on complementary filter was built. The gyroscope correction value was obtained by using magnetometer information and modified acceleration information, and the original gyroscope value was corrected. The complementary filter was designed to filter the high-frequency noise from accelerometer and magnetometer and the low-frequency noise from gyroscope, so as to avoid the interference of sensor noise signal in attitude estimation. Finally, the sensor information collected during flight test was used to simulate and verify the proposed algorithm. Experimental results show that the algorithm could accurately estimate the motion acceleration, reduce the interference of sensor noise and motion acceleration in attitude estimation, and effectively improve the accuracy and anti-interference ability of UAV navigation system in dynamic environment.

Abstract:In view of the threat of enemy air attack and the efficiency of solving the task assignment problem of medium-scale air defense firepower, a chainlike multi-population genetic algorithm (CMPGA) with superior performance was proposed. First, an improved air defense firepower task allocation model was established, which comprehensively investigates the threat degree of target and the interceptability judgment. The target threat degree was studied in terms of the threat factors such as the height, speed, range, and relative distance of the target. The time constraints, space constraints, and performance constraints were considered in the interceptability judgment, which was integrated into the kill probability to simplify the constraints of the model. Then, the CMPGA algorithm was proposed to solve the optimal allocation scheme of medium-scale air defense firepower. The algorithm utilized the strategy of limiting the number of repetitive individuals in the population, the cross mutation strategy of individuals with similar fitness, the deletion strategy of partial optimal solution when falling into local extremum, and the transfer strategy of the current optimal solution in the chain link. Combining the advantages of multi-population parallel search, the algorithm could speed up convergence speed, maintain the diversity of population, and avoid falling into local extremum. In the simulation of standard test function and the application to air defense firepower task allocation problem, the proposed algorithm was compared with several typical optimization algorithms. Results show that the CMPGA algorithm had advantageous performance and could quickly find the optimal solution with high probability, which indicates the effectiveness and superiority of the algorithm.

Abstract:The trapezoid tearing behavior and damage mechanism of envelope materials were investigated. Numerical simulation studies were conducted on a typical warp-knitted envelope fabric under a series of parameters, considering the microscopic structure of the yarns and the nonlinear characteristics of the envelope material. First, the numerical simulation method was verified by comparing the obtained results with experimental data. Then, the damage morphology and tearing damage process under typical bias angles were analyzed by the numerical simulation results. Next, the effects of the slit length and bias angle on the mechanical behavior and tearing strength of the envelope material were discussed in detail. Finally, the shearing effect was introduced and the prediction theory of off-axis trapezoid tearing strength of envelope material was proposed. Results show that the shape and extent of the airfoil-shaped stress concentration zone in the vicinity of the slit were vital factors influencing the propagation direction and tearing strength of slit. The evolution of the airfoil-shaped zone of stress concentration was obviously affected by the slit length and initial bias angle of yarns. Slit length could interfere with the out-of-plane deformation of envelope surface, the shape of slit tip, and the stress level of yarns. As the slit length increased, the fracture displacement decreased significantly, whereas the on-axis tearing strength increased. Since the yarn direction could exert complex effects on the airfoil-shaped zone and the stress level of the yarn, an inverted “V”-shaped relationship was presented between tearing strength and yarn bias angle. The research findings can provide reference for the slit arrest analysis and safety assessment of pneumatic membrane structures.

Abstract:A full lifetime capacity prediction method for vehicle power batteries was proposed, so as to accurately quantify the aging degree of automotive power batteries, improve the utilization rate of batteries, and achieve accurate prediction of the remaining useful life (RUL) in the whole life cycle of batteries. By integrating the traditional empirical exponential model and the improved polynomial regression model, the proposed method could track the degradation trend of battery life cycle based on the analysis of experimental data. The particle filter (PF) was adopted to adjust the model parameters online. Experiments were carried out to predict the RUL of power batteries with different states and capacities. The model was evaluated by comparing the prediction accuracy of different models. Experimental results show that the proposed model had a stronger ability in battery capacity attenuation tracking than that of the traditional empirical exponential model and the improved polynomial regression model. Combined with particle filter algorithm, the method achieved high-precision prediction results for both in-service and retired batteries. Besides, the method could accurately predict the failure time of power batteries with different capacities, which has a wide applicability in battery cascade utilization.

Abstract:To remove space garbage effectively and simulate the recycle process of abandoned satellites in space, we conducted ground test assessments on dragging retractable device by means of semi-physical experimental technology to verify the performance of the dragging system. Semi-physical simulation is a technology that uses physical objects or a combination of physical models and mathematical models for simulation. Ground-based semi-physical simulation experiment can be a priori study for space tasks, which can greatly reduce the cost of space experiments and effectively guide the design and control of space tasks. First of all, the dynamics modeling of the dragging process of tethered satellite was carried out, the mission goal of the dragging process was clarified, and the mechanical composition of the semi-physical dragging system was established. Secondly, the full physical simulation unit was designed as a passive loading unit. The satellite mass was simulated by using an equivalent inertia turntable. Then, the simulation loading unit was designed by a semi-physical simulation method to simulate the dynamic states of the target under the dragging action of the tethering system. Finally, in view of the disturbance of tension measurement error in the actual test process, a control strategy based on model predictive control was designed. To accurately control the tension and tether length of the dragging process, the experiment adopted double closed-loop motor joint control and combined tension position dual motor simulation through model predictive control. Results show that the tension error accuracy of model predictive control was 5%, which verifies the effectiveness of the tether retracting control system on the dragging process.

Abstract:To solve the test problem of degradation state in testability design, this paper proposes a multi-state system test modeling and test point selection method based on structure and function. Firstly, considering the description of performance states in multi-state system, the general idea of multi-state system test modeling based on structure and function was defined, and the related descriptions of structure, function, and test in the model were analyzed. Secondly, the problem of state testing was discussed, and the calculation methods for the state sets that can be detected by a single test or by a combination of test sets were analyzed. In addition, the state detection rate was proposed to quantify the test indexes of the system states. Finally, given that the fault testability problem is included in system testability design and analysis, a test point selection method of multi-state system integrating fault detection rate, fault isolation rate, and state detection rate was proposed under test cost constraint. Results show that the multi-state system test modeling method could analyze the testability of fault as well as the testability of degraded state. Compared with that obtained by the test selection method of binary system, the optimal test set obtained by the test selection method of multi-state system achieved higher state detection rate under test cost limitation. In the case of unlimited test cost, the optimal test set obtained by the test selection method of multi-state system not only achieved higher state detection rate, but also obtained the same maximum fault detection rate and fault isolation rate as those of binary system.

Abstract:To detect the wake characteristics of a moving target in water and obtain the information of the target such as trajectory direction and geometric, we proposed a wake detection method by imitating the mechanism of seals using beards to detect prey. First, a whisker model similar to the shape of seal whisker was fabricated with 3D printing technology. Combined with piezoresistive elements and other materials, a sensor array that can sense the wake of a moving target in the water was fabricated. Then, the experimental scheme of wake detection was designed, and simulation analysis was performed under uniform incoming flow of 0.5,1.0,1.5 m/s and non-uniform flow with slight fluctuations to obtain the wake velocity field distribution in different conditions. The lift coefficient and spectral distribution of S₁₁ were analyzed. The wake vortex shedding frequency obtained by theoretical calculation was compared to verify the validity of the experimental scheme. Finally, under laboratory conditions, the sensor array was used to measure the wake characteristics of a target, and the influence of the imitation seal whisker angle on the wake detection was analyzed. Results show that the lift coefficient of the whisker model fluctuated significantly under different incoming flow conditions, and the main frequency obtained by the fast Fourier transform was consistent with the theoretical value of the wake vortex shedding frequency. The average voltage value of the wake signal detected in the wake area was more than 3 times the average voltage value of the non-wake area, so that the wake area and the non-wake area could be accurately identified. On this basis, the gradient direction of the average voltage value attenuation was analyzed to estimate the trajectory direction of the measured target. The maximum response frequency of the wake signal measured by the sensor array was close to the main frequency of the numerical simulation. By inverting the response frequency, the diameter size of the measured target could be estimated. The seal whisker angle could affect the vortex-induced vibration of the sensor and increase the disturbance signal, but it had no significant effect on the estimation of the direction of the motion track and the estimation of the measured target size.

Abstract:To improve the circuit security of lightweight cipher algorithm (LWCA) and reduce power consumption, we proposed a look-up table (LUT) circuit with magnetic tunnel junction (MTJ)/CMOS hybrid structure, which can realize the complete PRESENT-80 encryption algorithm circuit by combining with sense amplifier based logic (SABL) cells. MTJ cells were introduced into the protection circuits for the design of the LUT circuit with hybrid MTJ/CMOS structure. Firstly, on the basis of 40 nm CMOS process and MTJ simulation model, the proposed LUT architecture was applied to design the S-box circuits which are essential in the operation process of cryptographic algorithm circuits, and results were verified through simulation. Secondly, a complete PRESENT-80 algorithm circuit was designed with the combination of the proposed circuit and SABL cells. Finally, all the circuits were tested by correlation power attack (CPA). Identical simulation and power consumption analysis were conducted on conventional CMOS single-rail and SABL dual-rail circuit structures. Results show that the proposed circuit possessed excellent power consumption attack resistance ability, which was capable of protecting against the CPA attack from 10 000 samples. Besides, the average power consumption of the proposed circuit was apparently reduced compared with that of the traditional SABL circuit.

Abstract:The second-order finite volume method (FVM) Godunov scheme was adopted to investigate the hydraulic transients in pumped storage power station. Firstly, the governing equations of mathematical model were discretized according to FVM, and the flux was calculated by Riemann solver. Three slope limiters were introduced and compared to avoid spurious oscillations during data reconstruction. The whole characteristic curves of the unit were transformed by using improved Suter approach. The virtual-boundary approach combined with the governing equations of the unit was presented to achieve a unified computation scheme for all the control volumes at internal domain and boundaries. The results calculated by the proposed scheme were compared with those of method of characteristics (MOC) scheme and measured data, and parameter sensitivity was discussed. Results show that when Courant number was equal to 1, the second-order FVM and MOC had the same accuracy; when Courant number was less than 1, the second-order FVM was more accurate and stable than MOC. During load rejection process of pumped storage station, the unit speed calculated by second-order FVM was basically consistent with the measured value and more accurate than that predicted by MOC. For the calculation of hydraulic transients in pumped storage power station, the values of wave speed in pipe sections with different properties should be adjusted to realize MOC simulations, resulting in computation complexity and numerical errors. While FVM is simpler in computation with higher accuracy, which only requires reducing the Courant number. Therefore, the proposed numerical approach is stable and accurate for hydraulic transients.

Abstract:To improve the accuracy of particle method in simulating the mechanical properties of ice, especially the plastic deformation characteristics of ice in the failure process, this paper proposes an elastoplastic constitutive model of ice bending failure based on the peridynamic theory of meshless particle method. Peridynamics is a newly proposed non-local theory, which is formulated in an integral form and can be applied to predict failures without extra assumptions. In addition, peridynamics is effective in deformed bodies, even in discontinuous objects. First, on the basis of Von-Mise criterion, the elastoplastic constitutive model of ice was established according to the relationship of incremental plastic strain and incremental plastic stress. Then, the application method of boundary conditions and the criterion of ice failure were analyzed. Next, the numerical calculation method of the constitutive model in FORTRAN was introduced. Lastly, the four-point ice bending process was simulated by using the ordinary state-based peridynamics method, and the force curve over time was predicted. The simulation results were compared with experimental results. Results show that ice failure process, crack propagation mode, and force curve predicted by the numerical model were in good agreement with the experimental results. Therefore, the numerical model established in this paper can be applied to the prediction of elastoplastic bending failure process of ice. The proposed constitutive model improves the basic numerical strategy for simulating the actual ice breaking process of ships sailing in horizontal ice.

Abstract:To effectively improve the electromagnetic absorption strength and effective absorption bandwidth of absorbing materials, the spherical Fe-Si-Cr was flattened by ball milling, and carbonyl iron powder (CIP) with different morphologies was compounded with Fe-Si-Cr by mechanical blending method. The effects of mass ratio on the electromagnetic parameters and microwave absorption properties of the composites were studied. Experimental results show that high aspect ratio could improve the dielectric constant and permeability of the material, and make the reflectivity peak shift to low frequency. The low-frequency absorption properties of flaky Fe-Si-Cr and CIP were better than those of spherical particles. Among the composite materials prepared in the experiment, the maximum reflection loss of sample I was -45.92 dB, and the effective absorption bandwidth was 1.5 GHz. The maximum reflection loss of sample L was -22.11 dB, and the effective absorption bandwidth was greater than 6.2 GHz. The morphology of the absorber had a significant influence on the electromagnetic absorption performance of the material. After compounding different morphologies of CIP and Fe-Si-Cr according to different proportions, absorbers with excellent performance at different frequencies were obtained. The impedance matching function could theoretically predict the thickness and frequency of the absorber corresponding to the peak of the reflection loss. Compared with single-layer absorbing materials, the effective absorption bandwidth of double-layer absorbing materials was larger, and the absorbers with excellent absorbing properties at different frequencies could be obtained by changing the thickness of matching layer and absorption layer, which can easily meet the requirements of new absorbing materials of “thin, light, wide, and strong”.

Abstract:The permeability and porosity vary significantly with different formation pressure in the gas injection process of the underground gas storage reconstructed from low-permeability gas reservoirs. It is necessary to understand the variation of permeability and porosity with formation pressure so as to simulate the injection and production process of underground gas storage. Firstly, the variogram theory in geostatistics was introduced, and the initial distribution of the reservoir physical parameters was determined based on the known information of the well points. Next, on the basis of the inverse problem theory, the differences between the measured values and calculated values of formation pressure were used to construct objective function, thereby transforming the inversion identification problem of reservoir physical parameters into the optimization problem. Then, the change rate of formation pressure to porosity and permeability was solved, and the physical parameters of reservoir were inversed by using conjugate gradient method. The correctness of the model was proved by a case study. A certain area of gas storage was taken as the research object for inversion analysis. Results show that although the distribution of permeability and porosity in the reservoir was basically the same at the beginning, the change rate of permeability and porosity was different at the end of gas injection, and the correlation between them was not consistent. The biggest change in permeability occurred near the 5# injection production well, and the value increased from 1.66×10^-3 μm² to 2.81×10^-3 μm², with an increase of nearly 70%. The function relationship between permeability, porosity, and formation pressure was built with the least square method. The research results can provide a theoretical basis for the injection and production simulation of low-permeability gas reservoirs converted to underground gas storage.

Abstract:To solve the problems of low thermal efficiency and hard utilization of condensing heat in traditional low and medium temperature heat recovery systems, a parallel type organic Rankine cycle (ORC) and ejector heat pump (EHP) combined cycle (referred to as PCHP) was proposed. The heat engine cycle and heat pump cycle were organized in parallel and an external heat exchanger (EHE) was added. All the condensing heat could be recovered for the production of radiant heating water, and the thermoelectric ratio could be adjusted within a certain range. The energy analysis model and exergy analysis model of the system were established. The thermal performance of the combined cycle and ORC were compared and analyzed. The characteristics of component exergy destruction and system exergy flow were investigated, and the effects of working fluid flow ratio and evaporating temperature on the thermal performance of the system were analyzed. Results show that the heat recovery capacity and exergy efficiency of the combined cycle were increased by 60.83% and 30.76% compared with those of ORC. The two components with the highest exergy destruction were the generator and the EHE. The main reason for the low exergy efficiency of the EHE was the high temperature difference of internal heat transfer. The working fluid flow ratio had a significant effect on the thermal performance of the system. When the evaporating temperature was lower than 25.2 ℃, the adjustable range of the working fluid flow ratio was the largest. The research results can provide guidance for the research and development of high-performance low and medium temperature heat recovery systems.

Abstract:An unsteady numerical heat transfer model was proposed to study the operation stability and heat-affected radius of medium-deep borehole heat exchanger (MDBHE). The governing equations were discretized and solved based on the finite difference method (FDM). The numerical results were verified through the measured data of a project. On this basis, the outlet water temperature of MDBHE and the temperature of rock and soil under long-term operation were studied. The sensitivity analysis of the operating parameters affecting the thermal disturbance of MDBHE was also carried out. Results show that after operation of 15 years, the outlet temperature of MDBHE was relatively stable, and the outlet temperature of MDBHE was basically in a quasi-steady state from the 11th year, and the maximum temperature drop ratio of the surrounding rock and soil was only 6.5%. With the increase in the thermal diffusivity of rock and soil, the increase rate of heat-affected radius of MDBHE was faster in the early stage and slower in the later stage. When the thermal diffusivity a increased from 1.43×10^-6 m²/s to 2.01×10^-6 m²/s, the heat-affected radius r increased from 82.69 m to 99.23 m, with an increase rate of 20%. The relationship between them changed exponentially and satisfied the equation r=-151.99×exp[-a/(5.14×10^-7)]+98.14, where R²=0.97. This research has certain reference significance for the design of MDBHE.

Abstract:A direct correction method for estimating the outdoor mean radiant temperature (MRT) with black globe thermometer method was proposed to solve the observation defects of the traditional black globe thermometer method and improve the estimation accuracy of the outdoor radiant environment in hot summer and warm winter zone. First, the six-directional radiation method was used to observe and analyze the radiation characteristics of three different underlying surface sites in Guangzhou, and the effect of the underlying surface albedo on the radiation environment was clarified. Then, the estimation results of the six-directional radiation method and the black globe thermometer method were compared, and the suitable response periods of 50 mm and 150 mm black globe thermometers were determined to be 5 min and 20 min respectively. Finally, the regional applicability of the direct correction method was verified. On the basis of the statistical results of the two observation methods for estimating MRT, a polynomial correction model was constructed for two specifications of the black globe thermometer method under different weather conditions, and the accuracy of the model was verified in the corresponding weather. Research results show that the direct correction method was suitable for regions with the same climatic conditions, and the fitting equation, the coefficient of determination (R²), and the root mean square error (RMSE) between the six-directional radiation method and the black globe thermometer method modified according to the regression function of different weather conditions were greatly improved. The deviation ranges of the two specifications of black globe thermometers were improved by 61.35% and 68.71% respectively. This study can provide a theoretical basis for the estimation of outdoor radiation environment in hot summer and warm winter zone represented by Guangzhou.

Abstract:The effects of thermal adaption and dynamic demands for thermal comfort of occupants on heating and cooling energy consumption in residential buildings were explored. First, taking Chongqing as a typical city in the hot summer and cold winter zone, a thermal environment database for residences (sample size 2 151) was analyzed, and the S-shaped variation trend of annual comfort temperatures of occupants with outdoor temperatures was revealed. Then, the annual heating and cooling periods and the dynamic temperature settings for residences were determined based on the thermal adaption and dynamic comfort of occupants. Lastly, a typical reference residence was established by using EnergyPlus, with a family number of three and building area of 105 m². According to the design requirements of the JGJ 134—2010 standard, the annual energy saving potential for heating and cooling in residential buildings was analyzed based on the thermal adaption and dynamic demands of occupants. Results show that under annually natural ventilation condition, only 32.2% of thermal environment was distributed in the comfort zone (18-26 ℃). With the traditional heating and cooling temperature design (18-26 ℃), the proportion was increased to 62.87%, and it was further increased by 5.45% under dynamic temperature settings. The annual energy consumptions per area for traditional heating and cooling temperature settings were nearly 8.93 kWh/m² and 17.18 kWh/m². While the dynamic temperature settings significantly reduced the annual energy consumptions for heating and cooling, with the values of 7.78 kWh/m² and 12.96 kWh/m² respectively. The annual energy saving was up to 20.6%, and the energy saving potential for cooling was greater.

Abstract:To further reveal the heat/mass transfer and pressure drop characteristics of evaporative cooling between staggered tube bundles, considering the formation of water film on the tube bundle wall and the mass transfer process between wet air and spray water, an analytical model was established by using the Euler-Lagrange method based on the coupling of discrete phase model (DPM) and wall film. First, the reliability of the model was verified by the experimental data from literature, and the error was within 1%. Then, the characteristics of heat/mass transfer and pressure drop between staggered tube bundles along the vertical direction were studied by simulation method. Results show that the mass transfer coefficient fluctuated along the coil height due to the variation of the enthalpy differences of wet air and saturated air, but the overall trend was decreasing. Under the same working condition, the influence of the enthalpy difference of wet air was greater on the mass transfer coefficient, and the fluctuation of mass transfer coefficient was mainly caused by the enthalpy variationof wet air. The temperature of the spray water film between staggered tube bundles was not constant, but it decreased with the decrease in the coil height, that is, the temperature of the liquid film gradually decreased along the falling direction. The spray water evaporation in the bottom area of the staggered tube bundle was the largest, and it decreased along the height of the coil. The spraying process mainly occurred on the surface of the tube bundle and the wake area of the tube bundle. By adding baffles or reducing the distance between the tube bundles, the turbulence of the flow field on the surface and the wake area of the tube bundles was strengthened, and the evaporative cooling effect of the spray water was enhanced.The heat transfer coefficient between the staggered tube bundles had little changes in the middle stable region of the heat exchange coil, and it was less affected by the inlet air temperature, spray water temperature, and relative humidity. The pressure loss between the staggered tube bundles had little changes along the coil height, and it was the largest in the air inlet transition area of the heat exchange coil. The above research results provide theoretical basis for the design of heat/mass transfer and pressure drop performance of air-water evaporative cooling between staggered tube bundles.

Abstract:To obtain the suitable hourly methods for key energy-saving meteorological parameters under maritime climate condition, this study analyzed five energy-saving meteorological parameters including temperature, dew point temperature, relative humidity, wind speed, and atmospheric pressure. By comparing representative inland and coastal cities, the applicability and characteristics of linear interpolation, cubic spline interpolation, piecewise cubic Hermite interpolation, Akima interpolation, and radial basis function interpolation in maritime climate conditions were discussed. Results show that under maritime climate conditions, the best interpolation method for winter temperature was cubic spline interpolation, and that for humidity was Akima interpolation; the best interpolation method for summer temperature and humidity was piecewise cubic Hermite interpolation. The best interpolation method for dew point temperature was linear interpolation in both winter and summer, and that for wind speed and atmospheric pressure was Akima interpolation in both winter and summer. The interpolation accuracy of different meteorological parameters had the same overall trend in the same season. The interpolation accuracy in summer from small to large was wind speed, dew point temperature, humidity, temperature, and atmospheric pressure, and that in winter was dew point temperature, wind speed, humidity, temperature, and atmospheric pressure. The interpolation accuracy of dew point temperature, humidity, and wind speed was greatly affected by the region, and they were affected in an ascending order of humidity, dew point temperature, and wind speed. The interpolation accuracy of temperature and atmospheric pressure data was less affected by the region.

Abstract:In view of the problem that the lack of solar radiation data leads to the inaccuracy of building energy consumption simulation and building energy conservation analysis, this paper takes Xian as an example to study the selection of typical meteorological year (TMY) and prediction of hourly radiation when the solar radiation data is missing. First, through correlation analysis, it was found that the easily obtained sunshine duration had the highest correlation with solar radiation. Therefore, on the basis of the traditional Sandia method, TMY was selected according to sunshine duration instead of radiation, and the parameters of the selection results were compared and analyzed to verify the accuracy of TMY selected by the new parameters. Then, reasonable prediction input parameters were selected through the comparison of the main influencing parameters of solar radiation, and hourly radiation prediction was carried out by using neural network and its optimization algorithm, which has strong ability in dealing with generalization problems. The obtained results were compared with statistical model and observation data. Finally, an office building model was established according to the building energy conservation design standards in China. The meteorological data obtained by the proposed method were used to simulate and verify the energy consumption of the building, and the changes in the heating and cooling energy consumption were analyzed respectively. Results show that the proposed TMY selection method could well solve the problem of selecting TMY in areas with radiation data missing, and the neural network algorithm could accurately predict hourly radiation data, which provides a new idea for the study of building energy conservation with radiation data missing.