Fujian Electric Power Company
Fujian Electric Power Company
Ye Y.,Fujian Electric Power Company
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2016
Over the past years, a large family of algorithms has been designed to provide different solutions to the problem of dimensionality reduction, such as discriminant neighborhood embedding (DNE), marginal fisher analysis (MFA) and double adjacency graphs-based discriminant neighborhood embedding (DAG-DNE). In this paper, we investigate the effect of data distribution for face recognition. We conduct three settings to investigate the performance when we have different numbers of the training samples. One is randomly select 20% samples as training set and the remaining face images are used for testing. One is randomly select 40% samples as training set and the last one is randomly select 60% samples as training set. In the end, we find as interesting observation is that when the training sample size is large enough to sufficiently characterize the data distribution, all algorithms we discussed in this work can achieve good performance. © Springer International Publishing AG 2016.
Yin Z.,Fujian Electric Power Company |
Zhang W.,Jicheng Electronics Co.
IOP Conference Series: Materials Science and Engineering | Year: 2017
In order to solve the problem of distribution fault diagnosis in case of misreporting or failed-report of fault indicator information, the characteristics of the fault indicator are analyzed, and the concept of the minimum fault judgment area of the distribution network is developed. Based on which, the mathematical model of fault indicator fault diagnosis is evaluated. The characteristics of fault indicator signals are analyzed. Based on two-in-three principle, a probabilistic fault indicator combination signal processing method is proposed. Based on the combination of the minimum fault judgment area model, the fault indicator combination signal and the interdependence between the fault indicators, a fault diagnosis method based on maximum probability is proposed. The method is based on the similarity between the simulated fault signal and the real fault signal, and the detailed formula is given. The method has good fault-tolerance in the case of misreporting or failed-report of fault indicator information, which can more accurately determine the fault area. The probability of each area is given, and fault alternatives are provided. The proposed approach is feasible and valuable for the dispatching and maintenance personnel to deal with the fault. © Published under licence by IOP Publishing Ltd.
Kang R.,Fujian Electric Power Company |
Yang M.,Fuzhou University
Gaoya Dianqi/High Voltage Apparatus | Year: 2017
For power supply of wireless temperature measuring device of high-voltage equipment, a new on-line temperature monitoring system powered by low-voltage LED light was designed. A solar cell was used to convert LED light energy to electric energy for reliable operation of the high-voltage wireless temperature measuring device. A two-stage PWM dimming constant current LED driver was designed to provide constant current for LED and ensure high power factor at input terminal. Temperature signals are collected with thermistor PT1000 and transmitted with ANT wireless communication module in the wireless temperature measuring device, which has the advantages of high precision and low power. Experiments show that the average current consumption of the wireless temperature measuring device is 7.24 mA under 3.3 V voltage, and the output current of the driver is 584 mA with PWM's full duty cycle. The working condition with 70% of PWM's duty cycle is the best condition, in which reliable operation of high-voltage circuit and power saving are balanced. © 2017, Xi'an High Voltage Apparatus Research Institute Co., Ltd. All right reserved.
Zou B.,Fujian Electric Power Company
Proceedings - 2016 16th IEEE International Conference on Computer and Information Technology, CIT 2016, 2016 6th International Symposium on Cloud and Service Computing, IEEE SC2 2016 and 2016 International Symposium on Security and Privacy in Social Networks and Big Data, SocialSec 2016 | Year: 2017
Discriminant Neighborhood Embedding (DNE) is one of the most popular methods for dimensionality reduction, which constructs an adjacency graph to preserve the local structure of original data in the subspace. However, there exist two shortcomings, first, DNE just constructs an adjacency graph, which may not achieve the goal of balancing the within-class and between-class samples. Second, DNE cannot identify the detail location of neighbors, and thus cannot preserve the neighborhood structure well. To remedy this, this paper proposes a novel supervised subspace learning method, named heterogeneous and homogeneous samples with different weights based discriminant neighborhood embedding (HHW-DNE). In HHW-DNE, we construct two adjacency graphs, heterogeneous and homogeneous adjacency graphs, respectively. Moreover, we further give different weights to heterogeneous and homogeneous samples. We evaluate performance of the proposed method over Yale dataset and compare the results with DNE method. The experimental results demonstrate that our proposed method consistently achieve superiors performances. © 2016 IEEE.
Xu Z.,Fuzhou University |
Wang S.,Fujian Electric Power Company
ICREPEC 2017 - Proceedings of the 6th International Conference on Reliability of Electrical Products and Electrical Contacts | Year: 2017
AC contactor with the rated current of 12A is to be the research object in this paper. The simulation module of electromagnetic mechanism of AC contactor with dynamic characteristic calculation and analysis is designed by experiment and simulation research, so the best and worst closing phase angle is determined, and the foundation of speeding up electrical life experiment is laid. The phase angle selection device is made and the control program based on the software of LabVIEW is programmed, as a result, it has the function of automated acquisition, data classification of the waveform, alarm and so on. The test method of electrical life with fixed angle is adopted, the electrical life experiment and data acquisition of 120,000 times is accomplished by life test system. The data processing module is established, which can read the characteristic parameters about electrical life, so the efficiency of life test is improved. By analyzing the trend of arcing time, no-load make duration of contact with the increasing of electrical life times, the prediction model of electrical life of AC contactor is established according to the prediction method of time-series.
Mathanker S.K.,Fujian Electric Power Company |
Hansen A.C.,University of Illinois at Urbana - Champaign
Applied Engineering in Agriculture | Year: 2014
Previous studies have shown that cutting speed and blade configurations play a critical role in crop harvesting. This study investigated the effect of cutting speed, blade oblique angle, and blade mounting in a field setting. To investigate their effect on Miscanthus harvesting power consumption, a hydraulically-driven single disk cutter head platform was developed. It has the provisions to adjust cutting parameters and was instrumented to measure the bending force on a push bar, torque, and cutting speed of the disk cutter. The cutting energy was determined at three oblique angles (0°, 30°, 40°), two blade mountings (fixed, flexible), and three hydraulic flow settings to drive the developed platform. Three hydraulic flow settings were used that resulted in average recorded cutting speeds of 31.5, 47.3, and 63.0 m s-1. The differences between the blade mountings were found to be negligible. A 40° oblique angle operating at 31.5 m s-1 had the lowest energy consumption, averaging 9.1 MJ ha-1. Similarly, a 30° oblique angle consumed 16.9 MJ ha-1 and a straight blade consumed 23.1 MJ ha-1. The results indicate that the cutting speed and blade oblique angle are directly related to the power requirements and efficiency of Miscanthus harvesting machinery. Information about the bending force of the Miscanthus was also collected and compared to the energy consumption of the machine. The data show that the energy consumption was correlated to the bending force. It is expected that the results of this study would help in modifying existing Miscanthus harvesters. © 2014 American Society of Agricultural and Biological Engineers.
Chen L.-J.,Fuzhou University |
Dai Z.-H.,Fuzhou University |
Liu Z.-W.,Fujian Electric Power Company
Yantu Lixue/Rock and Soil Mechanics | Year: 2011
A series of stress path tests on typical soft soils (silt and organic silty clay) in the influence range of foundation pit in Fuzhou city were conducted by using triaxial apparatus with stress and strain control in order to research the stress-strain behaviours of unloading and loading caused by excavation and retaining structure of foundation pit. Compared with the results of triaxial tests between normal consolidation and K0 consolidation samples in the same confining pressure, it is indicated that there are large differences in deformation and strength. Based on the results of K0consolidated triaxial tests, Duncan-Chang parameters were figured out and the methods of certification of each parameter in the model were put forward. At the same time, comparison with model parameters obtained by normal consolidation samples showed that some parameters had significant difference; and they decided the difference of stress-strain relationship. Duncan-Chang model parameters were suggested to obtained through triaxial loading and unloading test by K0 consolidation samples for the sake of satisfying numerical simulation analysis of foundation pit engineering. Duncan-Chang parameters of typical soft soils offered by the tests can be referred to numerical simulation of foundation pit engineering in Fuzhou city or other similar strata sites.
Zeng H.,Fujian Electric Power Company |
Lin F.,Fujian Electric Power Company |
Kang X.,Fujian Electric Power Company
Dianwang Jishu/Power System Technology | Year: 2014
Utilizing distributed parameter modeling the voltage at fault point is calculated in real time by voltage and current measured at the installation position of distance protection, and according to the relation of hyperbolic tangent function between the voltage drop from installation position of the protection to the fault point and the fault distance the impedance of faulty line is solved. This protection method is suitable to the protection for single-phase earth fault protection and the distance protection for phase-to-phase fault. In the algorithm design of the proposed protection method the voltage at fault point is taken into account, thus the affects of load current and transition resistance on protection performance are eliminated in principle, so the proposed protection method possesses better protection performance than traditional protection, and the effective identification of short-circuit faults under low-, medium- or high-resistance can be implemented. The correctness and validity of the proposed method are proved by results of EMTDC simulation and the recorded field test data.
Ma J.,North China Electrical Power University |
Zeng H.-M.,Fujian Electric Power Company |
Lin X.-H.,Fujian Electric Power Company
Dianli Xitong Baohu yu Kongzhi/Power System Protection and Control | Year: 2012
Using distributed parameter model, a new backup protection for multi-terminal high voltage transmission lines system is proposed based on wide area information. The backup protection consists of a new principle of wide area current differential protection and an algorithm of fault branch selection. In a distributed parameter circuit under normal operation, the algebraic sum of current flow-in and flow-out of the reference point selected in the multi-terminal high voltage transmission system is zero, while in case of fault occurring, the fault current can be reflected accurately by the differential variable of the reference point. Using this character, the wide area current differential protection can detect the fault within the protective zone. Based on the root distribution features of the fault-branch-selection equation, the fault branch selection algorithm is able to select the fault branch and locate the fault position. Calculated and analyzed by the wide area backup protection server, if a fault happens in the protective zone, the master station sends control command to the tripping device of fault line via communication net. The tripping device trips off the breakers of the fault line after receiving the trip-off command, isolating the fault from the system. Simulation test verifies the correctness and effectiveness of this backup protection algorithm.
Lu L.,Zhejiang University |
Wen F.,Zhejiang University |
Xue Y.,NARI Group |
Kang J.,Fujian Electric Power Company
Dianli Xitong Zidonghua/Automation of Electric Power Systems | Year: 2013
Reasonably controlling or dispatching the charging and discharging behaviors of aggregated electric vehicles (PEVs) can provide ancillary services for a power system, and hence improve the security and economics of the system. Based on the driving pattern statistic data from the New Jersey region in USA, an economic analysis model is developed for an agent/aggregator of PEVs in providing regulation and spinning reserve services for the power system concerned, considering the optimal control of the PEVs' charging and discharging. An optimization model is developed to maximize the agent's profit while respecting relevant constraints. The revenue of the agent consists of the incomes from the regulation and spinning reserve markets, as well as the allowances for mitigating emissions, while the cost includes the battery depreciation cost and charging cost. The developed model is demonstrated with actual data from the New York Independent System Operator (NYISO) with 5 000 PEVs specified, and the economic analysis result shows that the agent of PEVs can make profits by providing regulation and spinning reserve services in the electricity environment. © 2013 State Grid Electric Power Research Institute Press.