Search

Distribution network is the key link that affects the level of power supply service. In order to realize first-class modern distribution network and ensure the high quality electric power for residents, it is necessary to change the original working mode of operation and maintenance “passive repair” and the investment mode”treating the symptoms and not the disease” in distribution network. In such a huge distribution network structure, in order to achieve active maintenance and accurate investment construction, load forecasting of distribution network transformers scientifically will play a vital role. This paper proposes a method that realized the reasonable prediction of load in distribution network by using the self-learning and prediction function of artificial neural network and combing with the change characteristics of the load, which has important reference value to guide the construction investment and operation and maintenance of distribution network.

SMBI (supersonic molecular beam injection), as an effective fueling method for fusion plasmas, has been widely used on the HL-2A tokamak and other fusion devices. Two different types of SMBI system are now installed on HL-2A. One is an electromagnetic valve injector on the low field side (LFS), and the other is a pneumatic valve injector on the high field side (HFS). A new electron density record n?e = 4.7 ? 1019 m?3, which exceeds both the Greenwald density limit and the maximum density obtained by gas puffing (GP), was obtained on HL-2A with single-null-divertor operation. The HFS injection system is still under test, however, its outstanding fueling characteristics have already been observed, e.g. it has higher fueling efficiency compared to the LFS system. This excellent feature is still preserved during the process of ECRH.

Density limits with different fuelling methods have been compared in HL-2A, i.e. direct gas puffing and supersonic molecular beam injection (SMBI) from outer midplane, and divertor gas fuelling. The maximum densities for low current discharges are 3.4×1019 m−3, 4.3×1019 m−3 and 4.7×1019 m−3 for the 3 kinds of fuelling methods. The corresponding density ratios to Greenwald density limit are 0.9, 1.1, 1.2, respectively. The behavior of density limit disruption is analyzed as well.

Geometry sensitivity of magnetohydrodynamics (MHD) duct flow and some abnormal phenomena deviating from the classical MHD laws were measured in an experimental investigation of the MHD effect of the flow channel inserter (FCI) duct flow. The results showed that the velocity distribution deviated from the M-type profile on the centre-plane of the cross section of the duct with small slots (in short ?S-duct). The electric potential difference in the duct wall was much lower than that given by the classical electromagnetic laws. The MHD pressure drop was much higher than that from the classical magneto-hydrodynamics theory. These results may help us find some new methods to reduce the MHD pressure drop, or explore other potential applications such as flow structure control.

A new multi-channel far infrared (FIR) laser interferometer was built up and applied to HL-2A. The unique feature of real-time heterodyne interferometer is the combination of high power radiation source (300 mW), lower noise room temperature detector (noise temperature below 6000 K) with good spatial resolution of 7 cm. Various parameters are optimized for maximum laser output power. Zero crossings of the signals are counted with field programmable gate array (FPGA) digital circuitry yielding the resolution of 1/1000 fringe. The newly measured results including density fluctuation are also presented.

The electron density profile peaking and the impurity accumulation in the HL-2A tokamak plasma are observed when three kinds of fuelling methods are separately used at different fuelling particle locations. The density profile becomes more peaked when the line-averaged electron density approaches the Greenwald density limit nG and, consequently, impurity accumulation is often observed. A linear increase regime in the density range ne 0.6nG are obtained. There is no significant difference in achieved density peaking factor fne between the supersonic molecular beam injection (SMBI) and gas puffing into the plasma main chamber. However, the achieved fne is relatively low, in particular, in the case of density below 0.7nG, when the working gas is puffed into the divertor chamber. A discharge with a density as high as 1.2nG, i.e. ne = 1.2nG, can be achieved by SMBI just after siliconization as a wall conditioning. The metallic impurities, such as iron and chromium, also increase remarkably when the impurity accumulation happens. The mechanism behind the density peaking and impurity accumulation is studied by investigating both the density peaking factor versus the effective collisionality and the radiation peaking versus density peaking.

O-mode Doppler reflectometer has been successfully developed as an important diagnostic system on HL-2A. It can be used to measure the turbulence propagation in both plasma edge and confinement zone. The Doppler reflectometer system consists of two fixed frequency homodyne receivers: 15 GHz (corresponding to cutoff density of 0.3×1019 m−3) and 33 GHz (corresponding to cutoff density of 1.35×1019 m−3). The Doppler reflectometry principle and the experimental arrangements on HL-2A are presented. Meanwhile, the experimental Doppler reflectometric spectra under different discharge conditions, with and without ECRH, were obtained. Furthermore, the turbulence propagation velocity change and the profile were also observed in different discharge conditions.

Density fluctuations were first measured in the core region of HL-2A tokamak plasma using a newly developed multi-channel FIR interferometer system. In divertor ohmic discharges, we measured the radial density fluctuation levels of 5%, which increase to 10–20% during the appearance of MHD activity. Most of the power density in the density fluctuation spectrum is directly associated with m = 2 tearing modes. The fluctuation levels reduce to 1/3 and plasma confinement is improved during off-axis electron-cyclotron-resonance heating (ECRH).

A 1 m vacuum ultraviolet (VUV) spectrometer with temporal and spatial resolution was developed for impurity study of HL-2A tokamak. The instrument is equipped with two concave gratings blazed at 80 nm and 150 nm, respectively, and a windowless back-illuminated charge coupled device (CCD) detector of 256 × 1024 pixels. The total wavelength coverage of spectrometer is 30 ~ 320 nm with a spectral resolution of 0.015 nm at a width of entrance slit of 10 μm. A portion of this range is observed during a plasma discharge with a spectral range of 20 nm. The minimum integration time of the detector system is about 6.7 ms for each frame in a full binning mode. Using a space-resolved slit located between the entrance slit and the grating a radial profile on the vertical direction with a range of 400 mm can be obtained. The primary results were successfully obtained with high signal-to-noise ratio and good spectral resolution, which demonstrated the instrument functions very well.

A new criterion has been proposed to predict the major disruptions caused by tearing mode instabilities. According to the HL-2A experimental results, the statistical analyses are employed to investigate the relationships between MHD activities and the plasma disruptions. Two kinds of the tearing mode activities can finally cause the disruption on HL-2A operations. By introducing a new parameter, i.e. an integral of poloidal magnetic field over time, as the criterion of disruption precursor, almost all of the disruptions can be predicted.

Temporal distribution of the H atom and H2 densities in a pulsed microwave hydrogen plasmas has been measured simultaneously by two-photon absorption laser-induced fluorescence (TALIF). The measurement of the H-atom absolute density obtained by NO2 titration in a flow tube reactor shows that the density of H2 could be determined by the measured effective lifetime of the TALIF signal via the quenching equation. The H-atom density of about 1.5×1015cm-3 in both pulsed and stationary phases does not obviously change. It is found that the gas temperature volume effect plays an important role in governing the distributions of the H-atom density and it is mole fraction. The calculated gas temperature is in good consistent with the rotational temperature of H2 measured by optical emission spectroscopy in pulsed phase.

The non-local transport phenomenon induced by supersonic molecular beam injection (SMBI) was first observed in the HL-2A tokomak. In comparison with the phenomena induced by other methods in various tokamaks, it has its own feature: the effect induced by SMBI in HL-2A lasts much longer than that induced by pellet injection in other similar size tokomaks. Both the bolometer radiation and Hα emission decrease when the non-local effect appears. This suggests that an electron transport barrier has been formed at the position just outside the q = 1 surface when the non-local effect appears.

The particle confinement property of LHCD (lower hybrid current drive) plasma on the HL-1 tokamak is mainly affected by the line-averaged density of electrons (ne). With ne < 2.0 × 1013 cm-3, the particle confinement time (τp) is improved with the suppression of Hα(Dα) fluctuation at the edge, and tends to increase with the power PLH. The peak of τp appears near the critical density (1.0×1013 cm-3). These results are not influenced by the current drive directions.

Experiments with high-energy deuterium neutral beam (NB) injection were performed on the HL-2A tokamak. To obtain information on NB deposition and the slowing down of beam ions in HL-2A plasmas, very short-pulse deuterium NB injection, or the so-called ‘blip’ injection, was applied to MHD-quiescent ohmic deuterium plasmas. Analysis of neutron decay following the NB ‘blip’ injection indicates that tangentially injected beam ions are well confined, slowing down classically in the HL-2A tokamak. In contrast to the MHD-quiescent plasma, anomalous losses of beam ions were observed when a core-localized mode with a frequency up-chirping from 15 to 40 kHz appeared in the plasma. The core-localized mode was identified as a beta-induced Alfvén acoustic (BAAE) mode by its frequency sweeping behaviour and numerical calculation. Such a high energetic particle driven mode led to fast-ion loss, showing the strong influence of the core-localized fast-ion-driven BAAE mode on the fast-ion transport. Furthermore, a clear frequency splitting was first observed on the Alfvén-acoustic-type mode, and is found to be strongly linked to the effect of resonant wave–particle interaction, providing further insights into how frequency splitting structures are generated in the plasma.

Spatially resolved measurements of ion temperature are crucial for understanding plasma dynamics. A charge exchange recombination spectroscopy diagnostic system based on a heating neutral beam is installed on the HL-2A tokamak. An optical collection system is designed in accordance with the actual geometry of HL-2A and the CVI 529.2 nm (n = 8 ? 7) charge-exchange recombination line. Preliminary measurements show that the CX signals have enough signal-to-noise ratio (SNR). The ion temperature and its profiles are obtained through this system for the first time.

For heating the tokamak plasma effectively, the ion source must be capable of producing ions with high proton ratio. The proton ratio, which is found to be more than 65.6% at the ion current of 19.6 A with the extraction voltage of 39.6 kV, is measured with an image spectrograph by Doppler shift effect of Balmer-α-radiation spectrum emitted from fast hydrogen particles. The tendency of proton ratio with the ion density in experiment is almost the same as the mode devised by Zhang et al. Okumura et al. only gave the affection of the plasma volume and ion loss area on the proton ratio, but the relationship between the ion density in chamber and the proton ratio was not presented. We give the relationship.

An experiment at the HL-2A tokamak with a high-energy deuterium neutral beam (NB) injection (30keV, about 0.6 MW) was performed. The emission of d-d fusion neutrons dominated by beam-plasmas reactions when the deuterium NB was injected into the deuterium target plasma was observed by means of a 235U fission chamber. To obtain information on NB deposition and the slowing down of beam ions in HL-2A plasmas, a very short-pulse deuterium NB injection, or the so-called 'blip' injection, was performed into MHD-quiescent Ohmic deuterium plasmas. Analysis of neutron decay following the NB 'blip' injection indicates that tangentially injected beam ions are well confined, slowing down classically in the HL-2A. In contrast to the MHD-quiescent plasma, anomalous losses of beam ions were observed when tearing mode instabilities were present.

Modulated supersonic molecular beam (SMB) injection is introduced to study transport features of non-local transport phenomenon on HL-2A. Repetitive non-local effect induced by modulated SMBI allows Fourier transformation of the temperature perturbation, yielding detailed investigation of the pulse propagation. Fourier analysis provides evidence for existence of internal transport barriers. Meanwhile, experimental progress of nonlocal effect was made in the HL-2A Tokamak in 2007. The core electron temperature Te rise increases from 18% to more than 40% and the duration of the Te rise could be prolonged by changing the conditions of SMB injection.

A broadband, O-mode sweeping Doppler reflectometry designed for measuring plasma E × B flow velocity profiles is operated in HL-2A. The main feature of the Doppler reflectometry is its capability to be tuned to any selected frequency in total waveband from 26–40 GHz. This property enables us to probe several plasma layers within a short time interval during a discharge, permitting the characterization of the radial distribution of plasma fluctuations. The system allows us to extract important information about the velocity change layer, namely its spatial localization. In purely Ohmic discharge a change of the E × B flow velocity profiles has been observed in the region for 28 < r < 30 cm if only the line average density exceeds 2.2 × 1019 m−3. The density gradient change is measured in the same region, too.