Search

The precise estimate of remaining useful life (RUL) is vital for the prognostic analysis and predictive maintenance that can significantly reduce failure rate and maintenance costs. The degradation-related features extracted from the sensor streaming data with neural networks can dramatically improve the accuracy of the RUL prediction. The Temporal deep degradation network (TDDN) model is proposed to make the RUL prediction with the degradation-related features given by the one-dimensional convolutional neural network (1D CNN) feature extraction and attention mechanism. 1D CNN is used to extract the temporal features from the streaming sensor data. Temporal features have monotonic degradation trends from the fluctuating raw sensor streaming data. Attention mechanism can improve the RUL prediction performance by capturing the fault characteristics and the degradation development with the attention weights. The performance of the TDDN model is evaluated on the public C-MAPSS dataset and compared with the existing methods. The results show that the TDDN model can achieve the best RUL prediction accuracy in complex conditions compared to current machine learning models. The degradation-related features extracted from the high-dimension sensor streaming data demonstrate the clear degradation trajectories and degradation stages that enable TDDN to predict the turbofan-engine RUL accurately and efficiently.

Hydrogen is one of the prime candidates for clean energy source with high energy density. However, current industrial methods of hydrogen production are difficult to provide hydrogen with high purity, thus hard to meet the requirements in many application scenarios. Consequently, the development of large-scale and low-cost hydrogen separation technology is urgently needed. In this work, the gas separation properties of a newly synthesized two-dimensional nanoporous graphene (NPG) membrane material with patterned dumbbell-shape nanopores are investigated. The permeation energy barriers of different gases through this membrane are calculated using the density functional theory (DFT) calculations. Molecular dynamics (MD) simulations are also employed to study the permeation behavior of H2 in binary mixtures with O2, CO2, CO, and CH4. Both the DFT and MD calculation results show that this newly synthesized NPG membrane material can provide a high permeability as well as an ultrahigh selectivity simultaneously, making it a prospective H2 separation membrane with superior performance., Comment: 22 pages,6 figures

Featuring light weight and high specific capacity, lithium (Li) metal anode has been regarded as the “Holy Grail” for lithium-ion based batteries. The barriers in terms of mossy and dendritic Li formation and infinite volume expansion should be overcome before its application in practice. Herein, we propose a three-dimensional (3D) CNT host on Cu foil, serving as an integrated electrode (C-host@Cu) to stabilize Li metal cyclic behavior. The 3D host is prepared by growing CNT sponge on the Cu substrate directly and seamlessly. Stable cycling with high Coulombic efficiency of ~ 99% is achieved at a current density of 1 mA/cm2 over 250 cycles. In contrast, the bare Cu electrode presents a mossy-like Li morphology with a low Coulombic efficiency of ~ 60% after 150 cycles. The results reveal that the effective manipulation of Li deposition in this unique 3D host successfully suppresses the growth of Li dendrites.

Effective control of sulfate-reducing bacteria (SRB) will help reduce economic losses and prevent threats to human health during the oil production. In this study, the responses of SRB to environmental variables and some functional microorganisms (including denitrifying bacteria (DNB), methanogenic bacteria (MGB), saprophytic bacteria (SPB), zymophyte bacteria (ZPB), and iron bacteria (IB)), during oil production processes were investigated to improve our understanding of how to control SRB abundance. Correlation analyses demonstrated that nitrate and redox potential exihibited significant inhibitory effects on the growth and reproduction of SRB and redox potential (0.175, P

The understanding of electrical breakdown in atmospheric air across micrometer gaps is critically important for the insulation design of micro & nano electronic devices. In this paper, planar aluminum electrodes with gaps ranging from 2 μm to 40 μm were fabricated by microelectromechanical system technology. The influence factors including gap width and surface dielectric states were experimentally investigated using the home-built test and measurement system. Results showed that for SiO2 layers the current sustained at 2–3 nA during most of the pre-breakdown period, and then rose rapidly to 10–30 nA just before breakdown due to field electron emission, followed by the breakdown. The breakdown voltage curves demonstrated three stages: (1) a constantly decreasing region (the gap width d 10 μm). And the surface dielectric states including the surface resistivity and secondary electron yield were verified to be related to the propagation of discharge due to the interaction between initial electrons and dielectrics.

Research on the pulsed flashover characteristics of vacuum insulation material is significant for the design and manufacture of pulse power devices. In view of the voltage increase rate of the fast pulse (pulse steepness), the vacuum flashover characteristics of pure epoxy under different steepnesses is investigated by using a pulse generator with parameters of a rise time of 40 ns and a full width at half maximum of 2.5 μs. Pulses with six levels of steepness were achieved by changing the charging voltage of the generator. Based on the linear equation and electron emission equation, the relationship between the flashover voltage and pulse steepness was fitted. By virtue of the fitted formula, it was possible to predict the flashover voltage under near DC or higher steepness conditions. Based on the electron emission equation, the relationship between the time delay and flashover field was also fitted. Result shows that F-N electron emission dominates the flashover process.

The statistical parameters of phase resolved partial discharge (PD) distribution and ultra-wideband (UWB) characteristics of PD pulse are proposed for aging assessment of large generator insulation. Multi-stress aging tests of the model generator stator bar specimens were performed and PD measurements were conducted using both digital PD detector with frequency ranging from 40 kHz to 400 kHz and UWB PD detector with bandwidth from 10 MHz to 3 GHz at different aging staged. The test results show that the skewness and UWB frequency characteristics of PD can be taken as the characterization parameters for aging assessment of generator insulation. Furthermore, the measurement results of real generator stator bars show that these methods based on statistical parameters and UWB characteristics of PD are prospective for aging assessment and residual lifetime estimation of large generator stator insulation. *Supported by National Natural Science Foundation of China (Grant No. 59837260), China Postdoctora...

The circulating current method can judge whether the multipoint earthing fault happens by monitoring the variety of metal sheath's circulating current value. This article introduce an on-line monitoring system which use the circulating current method to monitor the multipoint earthing fault of the metal sheath. In the hardware layer, we use distributed current sensors and microcontroller's acquisition units to complete the signal acquisition and signal conditioning of the circulating current value. After the distributed acquisition and conditioning, the signals, through the RS485 bus, will be sent to the Ethernet Serial Device Server network system and then connected to the console cabinet. The computer of the console cabinet could read the circulating current value on-line by the Real Com Mode of the Ethernet Serial Device Server. This distributed on-line monitoring system has a simple structure, excellent real-time performance and high reliability proved by the practical operation.

Six kinds of thin films (TiO2, Al2O3, Cr2O3, TiO2/Al2O3, TiO2/Cr2O3) were prepared on epoxy substrate using the method of reactive magnetron sputtering. Experiments on the single film, coextruded film, and partial oxidation film were carried out under fast pulse power supply (30 ns/400 ns, rise time and full width at half maximum time) in vacuum, and the flashover characteristics of six kinds of oxide films were studied.

Research on vacuum surface flashover mechanisms of insulating material under nanosecond pulses is important for the design and manufacture of high power pulse generators. This paper studies the vacuum surface flashover characteristic of Al2O3/epoxy nanocomposite filled with different weight proportion of nano-filler. By using a unipolar pulse generator (20 ns rise time and 800 ns full width at half maximum), the flashover characteristic of Al2O3/epoxy nanocomposite is investigated. In addition, the thermally stimulated depolarization current (TSDC) is measured. The experimental results can be explained in terms of the different effects of shallow and deep traps.

The solid dielectrics such as epoxy, high density polyethylene(HDPE), polytetrafluoroethylene (PTFE), nylon and polymethylmethacrylate(PMMA) are often used as vacuum insulator in pulsed power device, in which they were applied under pulse voltage with risetime from several to several tens nanoseconds. Flashover characteristics must be paying more attention because they are the major breakdown models in their applications. A pulse source with output voltage of 50 ns risetime, 320ns pulse width and peak voltage up to 200 kV has been developed in this research. The flashover characteristics including flashover voltage and flashover current of the materials mentioned above under plane electrode were tested under 5E-3Pa vacuum. In order to get the relationship between surface charge deposition and the flashover characteristics, the surface potential distribution before flashover and after once, seventh and thirty-fifth flashovers are presented in this paper.

A hybrid computing intelligence approach was used in the study of an engineering diagnosis problem in this paper. Aimed at the problems of small samples and multi-collinearity of variables in complicated data modeling, RBF neural network was embedded into the regression framework of Partial Least Square (PLS) method. The PLS method was used to extract variable components from sample data and the dimension of input variables was then reduced. Moreover, RBF neural network was used to fit the non-linearity between input and output variables in projection space, and the disadvantages of traditional modeling method were overcome. Finally, this modeling method was applied to the prediction of residual breakdown voltage of the large generator stator insulation. The test results show that the hybrid model has better prediction ability than traditional modeling method.

Different defects in gas insulation substation (GIS) cause several types of partial discharge (PD). A novel on-line PD monitoring technique for GIS is described in this paper. An external sensor is adopted to detect the PD signal and has a wide frequency band. Two kinds of measurement method are used for PD signal acquisition, and high performance hardware units are also employed in order to reflect the true PD characteristics. Based on the idea of combining statistical analysis and high frequency analysis, the analysis level is improved. The developed PD on-line monitoring system has been operating at one 110 kV GIS substation. The results show that the system is effective

It is an important means for partial discharge (PD) on-line monitoring of oil-immersed power transformers in order to evaluating the insulation condition. Radio frequency current transducers (RFCT) are applied to couple the PD pulse current signal, and the difference-balance pairs (one of three-phase bushing-tap grounding terminals respectively corresponded with the neutral or iron core terminal) are selected, and high speed data acquisition card is adopt to sample the PD signal (>10MS/s) from the balance pair, and further amplitude modulation, phase modulation and difference arithmetic is designed, then external electromagnetic interference is effectively suppressed and it is propitious to extracting PD pulses and judge the PD occurrence place. The wideband PD couplers based on microstrip antenna are also applied to detect the high frequency PD pulse signal. Four sensors are installed at the special parts on the transformer where PD electromagnetic wave signal can be received, then the signal is transmitted to the input analog channel of the oscilloscope (>IGS/s). So single PD waveform analysis throughout more wide frequency band can be done, and further PD fault location can also be realized. The PD on-line monitoring system based on the above two monitoring methods for in-service transformer has also been developed and installed at one main transformer in a plant (220/18RV, 360MVA, three phase and dual winding)

It is very important that a GIS on-line detecting system is built availably, and it is also the key to acquiring successful detection. And it is of great significance to find out which parts are liable to faults as well as their real discharge characteristics. By using an ultra-wideband partial discharge detecting system, the discharge signal from disk-type spacer in key part of 110 kV GIS is acquired and studied. These key parts are the part nearby PT, the part nearby CT, the part nearby CB, the part nearby bus, the part nearby the LA. According to the different machine structure and electromagnetic field non-uniform distribution, these disk-spacers have different partial discharge characteristics. It is found that spacers in different parts of GIS have their own partial discharge characteristics, and their discharge frequencies concentrate below 300 MHz, which provide numerical reference for the on-line monitoring of GIS.

Based on the characteristics of partial discharge signals occurring in XLPE power cable and the theory of electromagnetic coupling effect, we developed a wideband sensor for detecting partial discharge signals occurring in XLPE power cable. The experimental result indicated that the sensor was fit to apply in wideband frequency with high signal-to-noise ratio (SNR). It was the foundation for partial discharge online monitor of XLPE power cable.

In recent years, great progresses have been made in technique for partial discharge detection, such as high frequency technique and ultra-wideband technique. Using these kinds of detective technique, we can get abundant information of partial discharge with high frequency. But it is rather difficult to analyze characteristics of nanosecond-order discharge signals in different frequency-band and to express those in quantities. In this paper, a kind of new method is introduced into analysis of the nanosecond-order discharge signals. First, an nanosecond-order discharge signal is analyzed into different frequency components based on wavelet analysis, then the fractal dimensions of the signal in each frequency-band is calculated, and the characteristics of the partial discharge signals can be expressed in quantities with fractal dimensions. By this method, the fractal characteristics of nanosecond-order discharge signals in different frequency-band can be analyzed clearly. And then, by using the analyzing method, the discharge characteristics of epoxy sample with different kinds of electrode system are analyzed, and the feature peaks of nanosecond-order discharge for these samples can be founded clearly. It is the base for the further study of the nanosecond-order discharge.

The partial discharge characteristics in the XLPE insulation during electrical treeing ageing are studied in this paper. The condition of electrical tree initiation is well established under ac stress. Specimens using point-plane electrode geometries in XLPE insulated samples have been investigated. Using the partial discharge detection system, which includes traditional detecting method and ultra-wideband(UWB) partial discharge detecting method, the partial discharge characteristics of XLPE insulation during treeing are studied. We can find that the characteristics of UWB partial discharge signals have a change observably in the time-domain and frequency-domain during electrical treeing. The high energy density area of the partial discharge signals of the XLPE insulation specimens moves to the higher frequency during electrical treeing. Thus, the status of XLPE insulation can be evaluated by detecting the partial discharge signals occurring in them. It is important of this research for evaluating the status of XLPE power cables in service.

High voltage XLPE power cable is a very important part of power system. The degree of electric-tree aging in the cable can be estimated effectually by measuring partial discharge. An on-line partial discharge monitor system of XLPE power cable was developed based on the virtual instrument designing conception. This system has the advantage of wide analog bandwidth, excellent anti-interference ability, good stability, powerful analysis, friendly interface, convenient operation, etc, it can detect the quantity of partial discharge (PD) in XLPE power cable. It holds utility value in power system.

The stator winding insulation is the important part of a large generator. The stability of its service depends on the performance of stator winding insulation. The characteristics of partial discharge of 300 MW, 18 kV turbo-generator stator insulation in Yaomeng Power Plant Co. Ltd of China are studied by the developed ultra-wideband (UWB) partial discharge measurement system. The characteristics of partial discharge in the bus room, the ground-wire room and the end of excitation are studied. The partial discharge signals can be detected correctly and sensitively in these three places. The interfering signals are detected in the field too. The PD signals and the interfering signals are analyzed in the time domain and the frequency domain. The result shows that the characteristics of time-domain and frequency-domain depend on the position of partial discharge signals. It is useful for UWB partial discharge detection to be used in field test.

The technique of partial discharge (PD) detection in the range of ultra-wideband (UWB) is a new technological field and greatly developed recently. There are two different methods to measure the PD signal in the range of UWB. One is the method of matched impedance, and the other is the method of micro-strip coupling antenna. A sampler of matched impedance and a coupling antenna are developed in this paper. Based on the sensors, a measurement system, which can utilize the two methods, is developed. PD signals of several kinds of samples and signals of interference are measured by the system. The PD waveforms of a same sample in air are similar and the amplitudes of the PD waveforms are proportional. The shape of the waveforms is also similar, but there is phase-angle difference in the PD waveforms of a same sample in insulating oil. The experimental results show that it is comparable between the two methods.

The statistical operators of partial discharge, namely fingerprint parameters, are some important parameters of PD, which can be used to estimate discharge condition of insulation. The statistical operators are produced based on three phase and two intensity distributions by using statistical method, so they are in good connection with the shapes of these distributions and contain most significant information of discharge. Recently they are used widely to estimate the aging condition of insulation system for its advantages. In order to study the aging laws of Epoxy/mica insulation of stator coil, the multi-stress accelerated aging experiment based on the stator bar of 780 kW/6 kV large electric motor has been conducted in the laboratory environment. Partial discharge is tested at the end of each aging cycle, and then several fingerprints of PD parameter distributions are analyzed and studied in this paper. The results of the experiment show that the correlation factor and asymmetry of H/sub qmax/(/spl phi/), H/sub qn/(/spl phi/) and H/sub n/(/spl phi/) at the rated voltage always increase with the aging time. The increase of correlation factor shows that the shapes of phase-resolved distributions of PD parameters become more and more similar between the positive and negative halves of voltage cycle. And the increase of asymmetry operator shows that the discharge in negative half becomes more and more serious with respect to that in positive half. It is concluded that the special fingerprint parameters can be used to evaluate the aging condition of epoxy/mica insulation.

The digital expression of the characteristics of the ultrawideband waveform in time-domain is always difficult in the research of the ultra-wideband partial discharge detection. This digital expression is very important to popularize the technique of the ultra-wideband partial discharge detection. A quantitative digital analysis method is established in this research. The calculation, analysis and display are realized by MATLAB program. The characteristics of the ultra-wideband partial discharge waveform are clearly and accurately expressed by using this analyzing method. It is useful for building the fingerprint bank of the ultra-wideband partial discharge signals.

The stator winding insulation is one of the most important parts in large generators and electric motors. It is always subjected to the combined thermal, mechanical, electrical and environmental stresses during long-term operation. So the gradual deterioration of mechanical and electrical properties is inevitable. Continuous deterioration of electrical insulation will result in the decreasing of breakdown voltage and failure of generator or electric motor. It is very important for the safe operation of a large generator to inspect insulation condition and evaluate the remaining life of stator winding insulation. In order to assure the validity of evaluation, it is necessary to study the characteristic parameters that can indicate the condition of insulation deterioration. In this paper, a multi-stress aging system has been developed and the accelerated aging experiment based on the stator bars of 780 kW/6 kV electric motor has been conducted in the laboratory in order to simulate the practical aging of electric motor or generator in service. The developing laws of AC current parameters with aging time are analyzed. Moreover, the relations between the breakdown voltage and them are studied. During the process of multi-stress aging of epoxy/mica insulation, the AC current parameters, including the first and second AC current rapid voltage (U/sub pi1/, and U/sub pi2/), equivalent impedance and current increasing factor m/sub 2/, have apparent developing laws with aging of insulation. Moreover, the second AC current rapid increase voltage U/sub pi2/ shows close correlation with the residual breakdown. voltage, which can be used as a characteristic parameter to estimate the remaining life of epoxy/mica insulation. Therefore, the degradation of epoxy/mica insulation can be assessed efficiently by the AC testing method.

The long-term aging test for stator winding insulation (SWI) of large generator is carried out with multi-stresses, including thermal, electrical, mechanical and thermal cycling stresses. During aging periods, the characteristics of SWI are measured for the determination of deterioration degree. Some characteristic parameters, included PD intensity index PDI, PD phase index PDP, ac current rapid increase rate m/sub 2/, ac current rapid increase index /spl Delta//sub 3/ and so on, are proposed to diagnose the degradation condition of large generator SWI subjected to combined stresses. The definitions of PDI, PDP, m/sub 2/ and /spl Delta//sub 3/ are given in this paper. It is theoretically concluded that the PDI and PDP are the functions of dissipation factor tip-up and capacitance tip-up. The m/sub 2/ and /spl Delta//sub 3/ indicate the growth rate of ac current between the first rapid increase point and the second rapid increase point. The experimental results are given and the relationship between the new characteristic parameters and residual breakdown voltage of stator winding insulation are discussed. It has been verified that these parameters, especially PDI and m/sub 2/, have closely relations with the residual breakdown voltage of the specimen insulation.

The characteristics of the partial discharge (PD) in the aging process of insulation are studied in an ultra-wideband range from 0 to 500 MHz in this paper. The electrical or thermal aging of epoxy samples involving cavities with cone-cone electrode or needle-plate electrode systems are conducted. A TE571 PD detector is used to distinguish the discharge classification, and an ultra-wideband PD detecting system is employed for detecting ultra-wideband PD characteristics of insulation in aging process. The experimental results show that the major characteristic peaks of these insulation samples locate in the frequency ranges of such as 0/spl sim/75 MHz, 150/spl sim/200 MHz. In addition, the discharge characteristics of stator bar insulation samples of a 300 MW turbo-generator out of service is also measured, indicating that characteristic peaks locate in the frequency range of 100/spl sim/125 MHz.

Using an ultra-wideband detection system, the characteristics of a partial discharge are studied. The new detection results are different from traditional experimental results in the time-domain. Based on collision ionization theory, a new physical model is put forward for analyzing the new detection results in this paper. In the new model, the second emission resulting from the ion impacting cathode and photoelectron radiation is considered. The electron current and positive ion current of the first ionization, electron current resulting from the ion impacting cathode and photoelectron radiation in the nanosecond discharge process are computed in this paper. Each generation discharge current is obtained, and a computing discharge waveform is obtained. The theoretical computed current waveform is no different from the measured nanosecond discharge waveform. It is shown that the new model can interpret the detection results.

Since the introduction of partial discharge (PD) detection, the majority of conventional electrical measurement of PD has been performed with detectors operating in narrow bandwidths from 10 kHz to 1 MHz for several decades. In the past decade, advances in electronic instrumentation have permitted the measurement of PD with bandwidths up to 1 GHz. Since PD pulses are fundamentally events with duration as low as a few nanoseconds which have a very wide frequency spectrum, comparing with narrow band PD detection, PD measurement with such ultra-wideband (UWB) detection methods can provide more information on the deterioration of stator winding insulation subjected to combined stresses. A UWB PD measuring technique with bandwidths from 20 kHz to 1 GHz is presented, and some waveforms of UWB PD signals of stator bar specimens in different aging periods both in time domain and frequency domain are given in this paper. It has been found that the ultra high frequency components contained in PD pulse increase and the rise-time of the pulse decrease gradually with aging time. It would be suggested that the ultra high components of PD pulse may be used to characterize the aging degree of stator winding insulation.