Automated deep learning system for power line inspection image analysis and processing: architecture and design issues

The continuous growth in the scale of unmanned aerial vehicle (UAV) applications in transmission line inspection has resulted in a corresponding increase in the demand for UAV inspection image processing. Owing to its excellent performance in computer vision, deep learning has been applied to UAV inspection image processing tasks such as power line identification and insulator defect detection. Despite their excellent performance, electric power UAV inspection image processing models based on deep learning face several problems such as a small application scope, the need for constant retraining and optimization, and high R&D monetary and time costs due to the black-box and scene data-driven characteristics of deep learning. In this study, an automated deep learning system for electric power UAV inspection image analysis and processing is proposed as a solution to the aforementioned problems. This system design is based on the three critical design principles of generalizability, extensibility, and automation. Pre-trained models, fine-tuning (downstream task adaptation), and automated machine learning, which are closely related to these design principles, are reviewed. In addition, an automated deep learning system architecture for electric power UAV inspection image analysis and processing is presented. A prototype system was constructed and experiments were conducted on the two electric power UAV inspection image analysis and processing tasks of insulator self-detonation and bird nest recognition. The models constructed using the prototype system achieved 91.36% and 86.13% mAP for insulator self-detonation and bird nest recognition, respectively. This demonstrates that the system design concept is reasonable and the system architecture feasible .

Numerical analysis of hydraulic fracture propagation in deep shale reservoir with different injection strategies

Deep shale reservoirs are characterized by elevated breakdown pressures,diminished fracture complexity,and reduced modified volumes compared to medium and shallow reservoirs.Therefore,it is urgent to investigate particular injection strategies that can optimize breakdown pressure and fracturing efficiency to address the increasing demands for deep shale reservoir stimulation.In this study,the efficiency of various stimulation strategies,including multi-cluster simultaneous fracturing,modified alternating fracturing,alternating shut-in fracturing,and cyclic alternating fracturing,was evaluated.Subsequently,the sensitivity of factors such as the cycle index,shut-in time,cluster spacing,and horizontal permeability was investigated.Additionally,the flow distribution effect within the wellbore was discussed.The results indicate that relative to multi-cluster simultaneous fracturing,modified alternating fracturing exhibits reduced susceptibility to the stress shadow effect,which results in earlier breakdown,extended hydraulic fracture lengths,and more consistent propagation despite an increase in breakdown pressure.The alternating shut-in fracturing benefits the increase of fracture length,which is closely related to the shut-in time.Furthermore,cyclic alternating fracturing markedly lowers breakdown pressure and contributes to uniform fracture propagation,in which the cycle count plays an important role.Modified alternating fracturing demonstrates insensitivity to variations in cluster spacing,whereas horizontal permeability is a critical factor affecting fracture length.The wellbore effect restrains the accumulation of pressure and flow near the perforation,delaying the initiation of hydraulic fractures.The simulation results can provide valuable numerical insights for optimizing injection strategies for deep shale hydraulic fracturing.

Flexible Fixture Design for the Top and End Surface Processing of NP2_NP4 Cylinder Heads

This paper focuses on the design of fixtures for NP2 and NP4 cylinder heads on a horizontal machining center of flexible machining automatic lines.It began with an analysis of the diagrams of part processing and working procedure which formed the basis for the design of the processing technology scheme,a selection of suitable machine tools,and the setting of processing parameters.Fixtures tailored to the chosen machine tools were then designed to meet the processing requirements.Additional aspects of the project included the design of part drawings,calculation of working time quota,design of auxiliary guides,support clamping,and hydraulic circuits,all aimed at fulfilling practical production requirements.

Analysis of a Cold Wave Weather Process in Chengdu in March 2010

[Objective] The aim was to analyze one cold wave weather process in Chengdu in March in 2010.[Method] Based on the NCEP 1°×1° 6 h interval reanalysis data and daily observation data,using synoptic analysis and diagnosis methods,and combining with the cold wave forecast index in spring of Sichuan,a cold wave event covering the whole region between March 21 and 24,2010 was analyzed from the aspects of circulation background,influencing weather systems and weather causation.[Result] Results showed that the 500 high-altitude cold vortex,700-850 hPa low layer shear,and ground cold front were the main systems that influenced this cold wave;there was a ridge from Lake Balkhash across Lake Baikal at 500 hPa.The early stage of the process was controlled by the high pressure ridge and the temperature was increasing obviously.The daily mean temperature was high.The range of cold high pressure was large and the central intensity was 1 043.0 hPa;the cold air was strong and deep which was in accordance with the strong surface temperature reduction center.The strong north airstream of Lake Balkhash to Lake Baikal,ground cold high pressure center intensity changes,north and south ocean pressure and temperature differences,850 hPa temperature changes,cold advection movement route and intensity were considered as reference factors for the forecast of cold wave intensity.[Conclusion] The study provided theoretical basis for improving the forecast ability of cold wave weather.

Analysis of the Flow Field Characteristics Associated with the Dynamic Rock Breaking Process Induced by a Multi-Hole Combined External Rotary Bit

The characteristics of the flow field associated with a multi-hole combined external rotary bit have been studied by means of numerical simulation in the framework of an RNG k-εturbulence model,and compared with the results of dedicated rock breaking drilling experiments.The numerical results show that the nozzle velocity and dynamic pressure of the nozzle decrease with an increase in the jet distance,and the axial velocity of the nozzle decays regularly with an increase in the dimensionless jet distance.Moreover,the axial velocity related to the nozzle with inclination angle 20°and 30°can produce a higher hole depth,while the radial velocity of the nozzle with 60°inclination can enlarge the hole diameter.The outcomes of the CFD simulations are consistent with the actual dynamic rock breaking and pore forming process,which lends credence to the present results and indicates that they could be used as a reference for the future optimization of systems based on the multi-hole combined external rotary bit technology.

Data processing and preliminary results of the Chang'e-3 VIS/NIR Imaging Spectrometer in-situ analysis

The Chang＇e-3 Visible and Near-infrared Imaging Spectrometer （VNIS） is one of the four payloads on the Yutu rover. After traversing the landing site during the first two lunar days, four different areas are detected, and Level 2A and 2B ra- diance data have been released to the scientific community. The released data have been processed by dark current subtraction, correction for the effect of temperature, radiometric calibration and geometric calibration. We emphasize approaches for re- flectance analysis and mineral identification for in-situ analysis with VNIS. Then the preliminary spectral and mineralogical results from the landing site are derived. After comparing spectral data from VNIS with data collected by the Ma instrument and samples of mare that were returned from the Apollo program, all the reflectance data have been found to have similar absorption features near 1000 nm except lunar sample 71061. In addition, there is also a weak absorption feature between 1750-2400nm on VNIS, but the slopes of VNIS and Ma reflectance at longer wavelengths are lower than data taken from samples of lunar mare. Spectral parameters such as Band Centers and Integrated Band Depth Ratios are used to analyze mineralogical features. The results show that detection points E and N205 are mixtures of high-Ca pyroxene and olivine, and the composition of olivineat point N205 is higher than that at point E, but the compositions of detection points S3 and N203 are mainly olivine-rich. Since there are no obvious absorption features near 1250 nm, plagioclase is not directly identified at the landing site.

Analysis on the Heavy Snow Weather Process in Benxi Area

[Objective] The aim was to study the formation and development of a heavy snow in Benxi area. [Method] Based on conventional meteorological data, the formation and development of a heavy snow in Benxi area from December 4th to 5th in 2009 were analyzed from the aspects of weather situation evolution and physical quantity field feature. [Result] The heavy snow was caused by upper trough and North China cyclone. In this process, there was upper level divergence and lower level convergence over Benxi area, and it was warm at low attitude and cold at high attitude; southwest jet at low attitude transported water vapor from Bohai Sea to eastern Liaoning, which provided good water vapor condition for snow, but it didn’t reach heavy snow due to inadequate ascending force. The development of Ural Mountains high ridge played an important role in the snow process and the strengthened high ridge moving northward was beneficial to the southward movement of cold air and deepening of upper trough. Analysis on physical quantity field could provide reference for predicting beginning and ending time and strength of heavy snow. [Conclusion] The study could provide basis for the forecast of heavy snow.

Analysis of a Rare Continuous Rainy Process in Midsummer

[Objective] The research aimed to analyze a rare low temperature and rainy weather process which happened in Anhui Province from July 22 to August 14,2009.[Method] Based on the data of conventional observation,NCEP analysis field and automatic station,a rare low temperature and rainy weather process which occurred in Anhui Province from July 22 to August 14,2009 was analyzed.The formation reason of continuous rainy process in midsummer was discussed.The circulation characteristics and influence systems of continuous rainy process were revealed.On the base,the influences of configuration of circulation fields and difference of physical quantity fields at high and low layers on range and intensity of precipitation were analyzed.[Result] According to the circulation situation and influence system,the continuous rainy process could be divided into four stages:July 22-24,from July 27 to August 1,August 4-8 and August 9-14.Moreover,it was respectively affected by northeast low vortex,cold and warm air,high-level low trough,typhoon and periphery of subtropical high at four stages.The maintenance of big specific humidity zone provided sufficient water vapor condition for the continuous rainy weather.The rainstorm appeared in dense zone of specific humidity line,where the specific humidity >13 g/kg in the humidity front zone.A temperature trough maintained at 850 hPa.The cold air which continued to diffuse and go south was main reason of the abnormally low temperature during the continuous rainy period.Moreover,it provided ascending motion condition for precipitation maintenance.[Conclusion] The research provided references for actual forecast of continuous rainy weather.

Stress Analysis of the Subsea Dynamic Riser BaseProcess Piping

Thesubsea dynamic riser base （SDRB） is an important piece of equipment for the floating production platform mooring system.One end is connected to the rigid pipeline, carrying a rigid pipeline thermal expansion load and the other end is connected to a flexible riser, carrying the dynamic load of the flexible riser, so its function is a transition connection between the flexible riser and the rigid pipeline which fixes the flexible riser on the seabed. On the other hand. as a typical subsea product, the design will satisfythe requirements of the standards for subsea products. By studying the stress analysisphilosophy of the topside piping and subsea pipeline, a physical model and procedure for piping stress analysis of the SDRB have been established.The conditions of the adverse design load have been considered, and a combination of the static load from the rigid pipeline and the dynamic load flexibility has also been optimized. And a comparative analysis between the AMSE, DNV and API standards for piping stress with the checking rules has been done.Because theSDRB belongs to the subsea pipeline terminal product, the use of DNV standards to check its process piping stress is recommended. Finally, the process piping stress of the SDRB has been calculated, and the results show that the jacket pipe and the carrier pipe stress of the SDRB process piping satisfy the DNV standards as a whole.The bulkhead cannot be accurately simulated by the AutoPIPE software which uses the FEA software ANSYS inthe detailed analysis, but the checking results will still meet the requirements of the DNV standards.

Analysis on "96. 8" Heavy Rainstorm Process in Handan

[ ObjEtive] The research aimed to analyze ＂96.8＂ heavy rainstorm process causing flood disaster in Handan. [ Method] Based on ac- tual situation data, satellite cloud data and NCEP reanalysis data in the first dekad of August in 1996, ＂96.8＂ heavy rainstorm process causing flood disaster in Handan was analyzed to understand occurrence reason of the flood disaster. [ Result] Two meso-scale convective cloud clusters which developed and went north in turn caused ＂96.8＂ heavy rainstorm in Handan. Typhoon and inverted trough were main weather systems induced flood disaster in Handan. In going north process of the low-level jet, due to blocking of the subtropical high, water vapor and energy accumulated in Handan, providing material basis for formation of the heavy rainstorm. Development and eastward movement of the short-wave trough at middle lati- tude and continuous invasion of the reflux weak cold air at the low layer were direct reason for triggering generation and development of the convec- tive cloud cluster, and further causing continuous rainstorm. Wet layer over the rainstorm zone was deep and thick. Meridional distribution of the wet zone was wider than latitudinal distribution. South China Sea and Bay of Bengal were water vapor sources for the rainstorm zone. In the whole rain- storm period, it was convergence at low layer and divergence at high layer in the rainstorm zone. It was positive vorticity at low layer and negative vorticity at high layer. Precipitation intensity changed as convergence and divergence. Rainstorm zone had strong ascending motion. As strengthe- ning and uplifting of the ascending motion strong center, strong precipitation also strengthened. Rainstorm center was near the biggest vertical ve- locity center. Strong precipitation changed as vertical ascending motion. [ Conclmion] The research provided scientific basis for disaster prevention and reduction and decision-making service.

An aligned mixture probabilistic principal component analysis for fault detection of multimode chemical processes

A novel approach named aligned mixture probabilistic principal component analysis(AMPPCA) is proposed in this study for fault detection of multimode chemical processes. In order to exploit within-mode correlations,the AMPPCA algorithm first estimates a statistical description for each operating mode by applying mixture probabilistic principal component analysis(MPPCA). As a comparison, the combined MPPCA is employed where monitoring results are softly integrated according to posterior probabilities of the test sample in each local model. For exploiting the cross-mode correlations, which may be useful but are inadvertently neglected due to separately held monitoring approaches, a global monitoring model is constructed by aligning all local models together. In this way, both within-mode and cross-mode correlations are preserved in this integrated space. Finally, the utility and feasibility of AMPPCA are demonstrated through a non-isothermal continuous stirred tank reactor and the TE benchmark process.

Comparative Proteomic Analysis of the Fiber Elongating Process in Cotton

A comparative proteomic analysis was performed to explore the mechanism of cell elongation in developing cotton fibers.The temporal changes of global proteomes at five representative

RFPA Realistic Failure Process Analysis

RFPA is a numerical testing tool for realistic failure process analysis(RFPA)of rock,concrete,composites,and engineering structures.The RFPA solution offers perfect simulation tools for robust modelling of brittle material failure and engineering structural damage.The RFPA family of 2D and 3D core products offers the full depth of analysis tools—from a conceptual simulation to advanced

Analysis on a Severe Convective Weather Process in Summer in Beihai

[Objective] The research aimed to analyze a strong convective weather process in Beihai in summer. [Method] By using Micaps conventional data and the single station site information of Beihai, radar data, a strong convective weather process in Beihai City in August, 2010 was analyzed. [Result] 850 and 700 hPa cyclonic low-pressure circulation in the north of Vietnam and Beibu Gulf coast was favorable for the transportation of Bengal Bay southwest airflow, which provided the sufficient water vapor condition for the strong precipitation. The interaction of weak cold air and southeast warm wet airflow was the main cause of the strong rainfall process. The falling zone of strong precipitation related to the water vapor convergence. The maintenance of water vapor flux divergence field in southeast-northwest direction was favorable for the duration of strong precipitation in the south of Guangxi. Before the strong precipitation, the positive and negative vorticity in the high and low levels in Beihai station both abruptly strengthened, which was very favorable for generating the convection. The vertical velocity in the strong precipitation process turned from the positive value at the beginning into the negative value as the height variation. The turning had the indicated and early-warning roles on the generation of strong convective weather. [Conclusion] The research provided the theory basis for the strong convective weather forecast in Guangxi coast in summer.

An analysis on El Nino processes during 1949-1987

-In this paper, the outset, evolution and intensity of El Nino were analyzed. Different features were found in each El Nino process. The El Nino of 1986 was also analyzed and outlined.

Development of the Data Processing and Analysis System Framework for ICF Experiments

An idea is presented about the development of a data processing and analysis system for ICF experiments, which is based on an object oriented framework. The design and preliminary implementation of the data processing and analysis framework based on the ROOT system have been completed. Software for unfolding soft X-ray spectra has been developed to test the functions of this framework.

Thermal Analysis Simulation of Germanium Zone Refining Process Assuming a Constant Radio-Frequency Heating Source

Three-dimensional thermal a nalysis simulation of a horizontal zone refining system is conducted for germanimn semiconductor materials. The considered geometry includes a g＇ral）hite boat filled with germanium placed in a cylindrical quartz tube. A flow of Ar and H2 gas mixture is purged througll the tube. A narrow section of the, boat is assmned to be exposed to a constant heat rate produced b v an rf coil located outside the quartz tube. The results of this analysis provide essential information about various parameters such as the shape of tile molten zone, required power and temperature gradient in the system.

Process Capability Analysis of Delivering Neonatal Care with Normal Weight （Case Study of Neonatal Weight Data at a Maternity Clinic in Banjarmasin）

Statistical Quality Control （SQC） is used to analyze and monitor quality characteristic measurements of normal neonatal weight in a maternity clinic in Banjarmasin in this paper. The objective of this study is to assist medical practitioners in observing pregnant women to deliver their babies with normal weight. It is also assumed that pregnant women who delivered their babies in the clinic have been monitored during their nine-month pregnancy. Thus, they can manage their own pregnancy to deliver normal weight babies. The use of Statistical Process Control （SPC） tools, such as frequency histogram, probability plot, and the implementation of Shewhart, R, and S control charts as primary techniques, are presented to display the monitoring aspects of the process. In addition, Process Capability Analysis （PCA） is performed to ensure that the process outcomes are capable of meeting certain requirements or specifications. The Process Capability Ratio （PCR） for the process is also presented. This analysis is an essential part of an overall quality improvement program.

Analysis of Heavy Rain Process in South Liaoning Province Caused by No. 1105 Tropical Storm "Meari"

[ Objective] The aim was to study the heavy rainstorm by tropical storm ＂Meari＂ in Liaoning. [ Method] Based on the ground data, radar data, numerical report data, encrypted automatic station and NCEP reanalysis data, the heavy rain process by tropical storm ＂ Meari＂ was ana- lyzed. The changes of each physical quantity were discussed while the tropical storm moving northward. [ Result] The subtropical high pressure and Mainland high pressure jointed and maintained, forced ＂ Meari＂ turned northwest. The low-level southeast jet by ＂Meari＂ provided full water trans- mission and layer unstable conditions. The heavy rain mainly distributed in the tongue area of high energy, and the total energy was a key index to predict the rainfall area. [ Condusion] The study provided reference for the report business.

Analysis of the First Soaking Rain Process in Lanzhou City in 2011

[Objective]The research aimed to analyze the first soaking rain process in Lanzhou in 2011.[Method]By diagnostic analyses on weather scale,meso-scale system and physical quantity when the first soaking rain weather happened in Lanzhou City in 2011,formation reason and physical quantity characteristics of the soaking rain weather were discussed.[Result]Combination of the plateau low vortex and westerly cold trough was influence system of the first soaking rain.There was a southwest-southeast airflow from Sichuan Basin to Hexi in front of the trough at 700 hPa.It was commonly referred to as ＂ inverted-buckle wind＂,and provided sufficient water vapour for occurrence and maintaining of the precipitation.Convergence at the low layer and divergence at the high layer were favourable for vertical motion.Analyses on physical quantity and meso-scale system also proved important role of the ＂ inverted-buckle wind＂ at the low layer.[Conclusion]The research provided reference for real-time forecast business of the long drought turning rain in spring.