Advanced Materials Research Vols. 433-440

Abstract: In order to resolve the problems of real-time data mining, a dynamic fast mining algorithm is proposed in this paper based on association rules. The algorithm do not need to scan the original database again, and overcome the weaknesses of Apriori Algorithm, which is the most representative mining approach using association rules, such as time-consuming, no online digging, and so on. The algorithm meets the demand of online and progressive mining. With the properties of multilayer processing simultaneously and updating promptly, as well as the definition of sensitivity parameter, the algorithm can be used to mine useful real-time information for the decision makers.

Abstract: In the radar imaging mechanism different phenomena are present, as a result the radar image is not a direct representation of the sea state. In analyzing radar image spectra, it can be realized that all of these phenomena produce distortions in the wave spectrum. The main effects are more energy for very low frequencies. This work investigates the structure of the sea clutter spectrum, and analysis the low wave number energy influence on determining sea surface current. Then the radar measure current is validated by experiments. By comparing with the in situ data, we know that the radar results reversed by image spectrum without low wave number spectrum have high precision. The low wave number energy influent determining current seriously.

Abstract: This paper presents a new index to voltage collapse based on the system physical quantities such as nodes' voltages and injected powers.The index has a critical value. It is physically meaningful, varies smoothly when the system changes (i.e. does not exhibit discontinous changes on value for small system changes such as oLTCs´action). This new index does not need to build any models and not be limited with load patterns.Moreover it can be used to real time system monitoring because it can be easily calculated , and can alert operators to take appropriate actions. The simulate result is very promising.

Abstract: With the widely use of the non-linear elements, the unreasonable parameters setting or interruption to HVDC system, converting stations may produce a kind of non-characteristic harmonics. In order to study its impact to the system, traits of these harmonics are analyzed based on the power system simulation tool—EMTDC. Contents and amplitudes of the harmonics in the DC line are detected by the expansion Prony algorithm, after that a basic analysis is given. Analysis of non-characteristic harmonics on the DC line is helpful to study the harmonic losses, the influence to electrical equipments and power qualities, especially to the design of the DC filters in HVDC system.

Abstract: The accuracy of voltage and current transformer (VT and CT) is a key factor in electric power metering. Ratio error and phase displacement calibration of VT and CT is traditionally carried out off-line. As a result of the expansion of the power grid and the increasing metering quality requirements from customer side, the off-line metering of VT and CT becomes more and more difficult. In this paper, a VT accuracy on-line evaluation system is developed, which can evaluate the accuracy of VT under on-line operating status. The principle and adopted key techniques in the development of the VT accuracy on-line evaluation system is introduced. Some on-site test results are also presented and analyzed as application demonstration.

Abstract: In order to have a better understanding of the vibration characteristics, micro tremor measurements are carried out on Yingxian Wooden pagoda. Horizontal vibration modes of the tower and the torsional vibration modes on each floor are attained by analysis of the test results. Based on transfer function method, elastic displacements of the tower under the different seismic waves are derived. Trough theoretical analysis, the significative conclusions are received as a basis for researching and protecting the ancient building.

Abstract: In order to enhancing the accuracy of fault diagnosis system, an improved method based on the probabilistic neural network (PNN) is proposed, in which the synthetic attribute weights of faults are introduced that are obtained by integrating algebra view and information theory view of rough set. The synthetic attribute weights are utilized to training the classical PNN and dealing with the classification of faults so as to improving the PNN model. The new model is more accurate and can represent expertise. This novel approach is applied in digital data network to diagnose failures, and the results of the experiment verify that the method is practical and effective in raising accuracy of diagnosis as well as avoiding misdirection in fault remedy.

Abstract: Predictive current method as a control strategy of the STATCOM is more and more applied to practical production. The traditional predictive current method can only predict the reactive current in the broad sense, and not detailed to predict for any harmonic; and the delay caused by sampling, analysis and calculation, prediction do not led to a good predictor for control requirements during the traditional prediction. In this paper, the above issues are appropriately improved. And the simulation results verify its operability.

Abstract: Taking 1000MW air-cooled hydro-generator as an example, the solving region and mathematical model are given. Based on the theories of electromagnetic field, the stator iron loss of large generator under non-load is studied. Considering the structural characteristics of generator, the iron loss was calculated by using magnetic circuit method and the finite element method, respectively. First, the iron loss in generator stator is calculated basing on the theory of electromagnetic field. Then the steady state and transient fields were used to calculate and analyze the 2D electromagnetic field by finite element method. And the exact value and distribution of iron loss are obtained. Last, we obtain some useful conclusions through comparing the iron loss calculated by the two methods above. That will provide a theoretical basis for further study of the physical fields of new generation giant hydro-generators.

Abstract: Harmonic detection is an important step in grid harmonic status evolution and suppression measures implementation. While it’s accuracy is interference to the noise in the power grid. The different characteristics of signal and noise in the wavelet transform are proposed firstly, and the method of soft threshold de-noising based on wavelet multiresolution is analyzed in the paper. Using the method, the noisy harmonic signals in the power grid are processed and the de-noisy signals are analyzed by Fourier Transformation (FT) to obtain the contents of each harmonics. The simulation results show that the harmonic containing rate is close to the original signals after de-noising by WT. That is the method reduces the interference of noise and the accuracy of harmonic detection is improved.