The trend of globalization in the world is becoming increasingly frequent, and people from different regions are communicating more closely. Therefore, the demand for a second language is constantly ...expanding, accelerating the development of the field of English oral evaluation and also accelerating the development of online education. The study proposes a text priori based oral evaluation model, which is based on the Transformer model and uses target phonemes as input to the Decoder. The model successfully predicts the relationship between actual pronunciation and error labels. At the same time, a self-supervised oral evaluation model with accent is constructed, which simulates the training process of misreading data by calculating semantic distance. The experimental results show that when the training set ratio reaches its maximum in the Speed Ocean dataset and the L2 Arctic dataset, the F1 values of the proposed method are 0.612 and 0.596, respectively; the length of the target phoneme has a smaller impact on this model compared to other models. Experiments have shown that the proposed deep learning method can alleviate deployment difficulties, directly optimize the effectiveness of oral evaluation, provide more accurate feedback, and also provide users with a better learning experience. This has practical significance for the development of the field of oral evaluation.
Compared to conventional technologies based on low-frequency thyristor-controlled power-supply units, the high-frequency power-supply units of electrostatic precipitators, based on a new cutting-edge ...technology, ensure more efficient gas conditioning at thermal power plants and industrial enterprises. Particular attention in the design of high-frequency power-supply units should be paid to designing a step-up transformer with high nominal output voltage and capacity at an operational frequency from several to tens of kilohertz. The design features of this transformer significantly influence the characteristics of the whole high-frequency power-supply unit. This article proposes an approach to forming a broadband design transformer model which correctly works in a broad range of low and high frequencies. This model allows imitating processes in the transformer together with an external circuit; in this case, it is possible to analyze the influence of the transformer design on the characteristics of the whole unit, as well as the operating conditions of the transformer itself, which is especially important in the device design phase.
The vital parameter condition prediction of aero-engine gas-path performance is increasingly becoming an important part of aero-engine fault diagnosis. The rapid and accurate parameter forecast plays ...a significant role in the early detection of the aero-engine faults. Comparing to the time series structure of traditional RNN or LSTM, the mechanism of self-attention and positional encoding approach are applied to avoid the long-term training process, as well as the training complexity of each layer is also improved and simplified. Moreover, the wavelet threshold denoising method is previously employed in data preprocessing, and then the Transformer model based on self-attention mechanism is constructed for time series prediction, and the prediction results are compared with the traditional regression algorithm. Finally, it is concluded that the model can accurately predict the development trend of aero-engine gas-path parameters. Therefore, the model is applied and validated in practice on site data, which implies the accuracy and efficiency of the approach. Transformer model based on self-attention mechanism has broad application prospects in the field of aero-engine parameter prediction.
Electrical-Model of ADITYA-U Tokamak Kumawat, Devilal; Kumar, Rohit; Macwan, Tanmay ...
Indian journal of physics,
03/2021, Volume:
95, Issue:
3
Journal Article
Peer reviewed
An Electrical-Model is developed for ADITYA-U Tokamak to study the generation of different shapes and duration of LVs out of the total available flux, which is 0.6 Vs in positive convertor supply. ...Loop voltage (LV) is required to be controlled during plasma discharge to optimize and predict the breakdown of plasma, proper impurity burn-through time, plasma current magnitude and duration of the plasma as per the experimental requirement. Loop voltage waveform is governed by selecting different magnitudes of resistance values, switching time and duration of external resistance in the wave shaping primary circuit. Loop voltage control system has been modelled using simulation code (Electrical-Model). Same results have been validated with the modelling of ohmic wave shaping circuit in MATLAB-Simulink. Temporal profile of loop voltage has been incorporated for ADITYA-U Tokamak for several plasma discharges, and detailed observations and results are presented in the paper.
The behavior analysis of the transformer is usually achieved by the frequency response analysis (FRA), which is obtained by the application of a very low AC voltage in over a wide frequency range. ...This paper presents a low and high frequency modelling approach of a three-phase transformer. The developed model consists of a cascade of parallel RLC cells, whose parameters are identified using the frequency response analysis data measurements obtained on each transformer phase. Thus, the proposed model can simulate the frequency behavior of the transformer windings without reference to the geometries of the coils which makes it easily usable in the failure diagnosis field. Experimental results on a 300 VA laboratory transformer validate the proposed model.
This paper views waste as a resource and anaerobic digestion (AD) as an established biological process for waste treatment, methane production and energy generation. A powerful simulation tool was ...developed for the optimization and the assessment of co-digestion of any combination of solid waste streams. Optimization was aimed to determine the optimal ratio between different waste streams and hydraulic retention time by changing the digester feed rates to maximize the biogas production rate. Different model nodes based on the ADM1 were integrated and implemented on the Matlab-Simulink
® simulation platform. Transformer model nodes were developed to generate detailed input for ADM1, estimating the particulate waste fractions of carbohydrates, proteins, lipids and inerts. Hydrolysis nodes were modeled separately for each waste stream. The fluxes from the hydrolysis nodes were combined and generated a detailed input vector to the ADM1. The integrated model was applied to a co-digestion case study of diluted dairy manure and kitchen wastes. The integrated model demonstrated reliable results in terms of calibration and optimization of this case study. The hydrolysis kinetics were calibrated for each waste fraction, and led to accurate simulation results of the process and prediction of the biogas production. The optimization simulated 200,000
days of virtual experimental time in 8
h and determined the feedstock ratio and retention time to set the digester operation for maximum biogas production rate.
This study focuses on the analysis of the non-linear behaviour of magnetisation inductance during a transformer frequency response test in open-circuit configuration. For this work, frequency ...response measurements were made on a three-phase transformer and a transformer model was used to simulate the response curve and to estimate the magnetisation inductance curve as a function of frequency. It was observed that the magnetisation inductance increases with increase of the test voltage. Also observed was an initial decrease at a very low frequency of the magnetisation inductance, caused by the displacement of the work point by the first elbow of the magnetisation curve. Additionally, the magnetisation inductance acquires a linear behaviour when the input small signal is sufficiently low. Finally, it was observed that the remanent magnetisation level in the core reduces the value of the magnetisation inductance.
This paper investigates the behavior of power transformers under the occurrence of permanent or intermittent winding insulation faults. For the study of these phenomena, a simple and efficient ...permeance-based electromagnetic transformer model is proposed, which is based on the simultaneous consideration of magnetic and electric equivalent circuits. To incorporate the internal faults in this model, a suitable equivalent circuit of the faulty winding is described. With the aid of this transformer model, the onload exciting current Park's Vector Approach will be applied for diagnosing the occurrence of permanent and intermittent winding faults. Experimental and simulation tests results are presented in this paper, which demonstrate not only the adequacy of the digital transformer model for winding fault studies, but also the effectiveness of the proposed technique for detecting winding interturn insulation faults in operating three-phase transformers.
This paper proposes a computational high-frequency transformer model. The model parameters are obtained by using coupled-circuit-finite-element (FE) nonlinear analysis. The frequency response of the ...transformer was obtained by coupling the transformer FE model and external electric circuits. This technique would allow the physical representation of the nonlinear magnetization behavior of the transformer as well as the strong frequency dependence of the transformer parameters. The self capacitance of each conductor and the mutual capacitances between the turns were calculated by an electrostatic FE analysis. The capacitance order was then reduced to a lower order by shifting the capacitances connected to internal nodes (windings) to the external ones (coil nodes). The resulting reduced capacitances, along with the inductances and resistances, were then used in the circuit domain of the coupled circuit-FE analysis. The transformer frequency response was then obtained from FE analysis. This response was then fitted with rational function approximation. This rational function approximation is then used to construct a frequency-dependent branch (FDB) which is connected in parallel with the nominal frequency transformer model. The FDB branch represents the transformer high-frequency behavior over a wide frequency range. The implementation is performed on a 125-kVA transformer. The developed model terminal behavior was tested under different operating conditions. This includes different switching frequencies and connecting cable length.
PDF
