Search

To enable power generation companies to make full use of effective wind energy resources and grid companies to correctly schedule wind power, this paper proposes a model of offshore wind power forecast considering the variation of wind speed in second-level time scale. First, data preprocessing is utilized to process the abnormal data and complete the normalization of offshore wind speed and wind power. Then, a wind speed prediction model is established in the second time scale through the differential smoothing power sequence. Finally, a rolling PSO-LSTM memory network is authorized to realize the prediction of second-level time scale wind speed and power. An offshore wind power case is utilized to illustrate and characterize the performance of the wind power forecast model.

Malware analysis technology has been one of the most important research topics of cyber security. The recent surge in adoption of ransomware is rapidly changing the malware landscape. A large body of researches in security community have given us an understanding of ransomware individuals and families. However, to the best of our knowledge, there are currently few works that explore common and distinct malicious behaviors on large scale ransomware dataset. Our insight is that although the implementation of each ransomware vary widely, its malicious behaviors inevitably interact with the underlying operating system, which will be exposed and captured by system event tracing mechanism. In this paper, we propose a novel ransomware analysis pipeline, a system provenance graph based approach for better understanding the ransomware’s behaviors. Then we leverage the analysis framework to analyze on large scale ransomware dataset and present some interesting findings on diverse ransomware and their families. Furthermore, our analysis on ransomware also reveals that system provenance graph is an ideal tool, with strong abstract expression ability and relatively high efficiency.

When the distributed supervisory control and data acquisition (DSACADA) and micro phasor measurement units (μPMU) are installed in the distribution network, a topology identification method based on multi-source data fusion is proposed. The identification model of topology changing time can be constructed with the voltage angle measurements. Based on the change of voltage angle and amplitude, the topology change trend criterion can be constructed for narrowing range of the possible topology. Then based on WLS, the fusion of μPMUs and DSACADA can be realized and all the node voltage angles in every possible topology can be estimated. Finally, the similarity identification model is constructed to identify the real operation topology. he reliable topology identification can be realized even some errors occur in the measurement of μPMU and DSACAD is shown in the example.

This paper proposes a fault line selection method using zero-sequence current waveform clustering for distribution network with same bus loop which has a high reliability for single-phase grounding faults. The purpose is to improve the adaptability of fault detection method for different line structure and fault conditions. Firstly, a single-phase grounding fault composite sequence (module) network for distribution network with same bus loop is established. It is used to analyze the zero-sequence current waveform difference between fault line/fault loop and healthy line during fault transient time. Furthermore, a fault line selection method using zero-sequence current waveform clustering is proposed. This method can be applied to low and high impedance grounding faults on a single outgoing line, the loop with the same bus, and the bus line. It has a strong applicability and improves the reliability of single-phase grounding fault line selection for resonant grounding systems. Simulation data verify the correctness of the method.

Subsynchronous Oscillation (SSO) may exist in doubly-fed induction generator (DFIG)-based wind farms connected with a series compensated transmission network, also called subsynchronous control interaction (SSCI). Different from thermal power units, the SSO frequency of wind farm varies with the change of system operating conditions, and has the characteristic of time-varying. When the operating conditions change, the existing mitigation strategies may lose the effect. To address this issue, a SSO mitigation strategy based on adaptive band-stop filter (ABSF) is proposed in this paper: When SSO occurs in the system, the ABSF parameter is adjusted according to the oscillation frequency measured by PASTd algorithm to filter out the subsynchronous component in the RSC loop effectively, so as to break the generation path of SSO and suppress SSO. Compared with the existing mitigation strategies, the proposed strategy doesn’t need the mathematical model of the system, and can ensure strong robustness under various operating conditions. Finally, the effectiveness and robustness of the proposed strategy are verified on the Matlab/Simulink.

The coordinated technology control system is one of the most important factors for improving the products accuracy and mill productivity. This paper deals with a system design problem for tandem cold mill. A multi-objective optimization model of rolling schedule based on cost function was proposed to determine rolling parameters for a specified product. Taking stand 3 as pivot stand, the mill speed model is established. The coordinated technology control and sequence control are designed and automatic cutting process based on strip tracking is investigated. Fault handling strategy is introduced to avoid the occurrence of any damage. Finally the self-design coordinated control system based on high-performance Siemens controller is applied to a 1450mm rolling line, the practical application shows that the system possesses high calculation accuracy and achieves the high-speed stable rolling.

Erroneous remote signal in distribution supervisory control and data acquisition (DSCADA) decreases reliability of distribution network topology which is identified based on remote signaling data of DSCADA. A new method of topology identification is proposed to solve this problem, which adopts multi-sampling of power injections by micro synchronous phasor measurement unit (pPMU) to construct the variance model of the branch voltage deviation. The Kruskal algorithm is used to obtain the minimum spanning tree with the variance of the branch voltage deviation as the line weight to achieve the identification of operation topology structure of distribution network, and then the influence of the sampling times of the power injection data and the complexity of the distribution network on the identification errors is analyzed. The example shows that the algorithm has good reliability and practicability for the topology identification in distribution network.

In recent years, with the fast development of various new power supply, the composition of smart distribution grid is increasingly diverse. Thus, the difficulty of network loss calculation is increased. Based on the influence of time — space characteristics of load on distribution network loss, the basic parameters of distribution network structure and load is considered, and the Kmeans clustering algorithm is used to classify the feeder. Based on the obtained data of the feeder network loss, the network loss of other feeders is estimated according to the similarity of the feeder load parameters, and then the whole network loss of the distribution network is calculated. Based on the case of Jiangsu power grid, based on the similarity of feeder loss is analyzed, and its feasibility is verified. The Jiangsu power grid is taken as an example for the calculation of smart distribution grid loss based on the similarity of feeder loss, and the feasibility of the method is verified.

In order to solve the face detection problem with complex lighting conditions, this paper presents a new method for face illumination compensation. The high-pass filtering is used to enhance edge information, the logarithmic transformation and the exponential transformtion are used to adjust the global brightness simultaneously, nonlinear variances are applied to reduce the impact of the local highlights and shadows. The uneven lighting problem is lighten. Comparison experiments are conducted on the Yale B face dataset, Orl face dataset and our own face dataset. The results demonstrate that the illumination compensation method can significantly improve face detection rate.

As for uncertainties and decision-makers’ overconfidence psychological bias, overconfidence has been incorporated into real option decision-making model of metal mineral resources mining to estimate its effect on decision-making of the project and thus a behavioral real option decision-making model of metal mineral resources mining based on overconfidence has been established. Furthermore, numerical simulation and sensitivity analysis have been conducted to verify the practicality of the model. Results show that model in this paper has greatly changed trigger value and option value of mineral resources mining project compared with traditional real option model, thus greatly changing optimal decision results. Incorporating overconfidence into real option decision-making model of metal mineral resources development is a crucial extension of project evaluation theory.

based on the in-depth study of continuous wavelet and discrete wavelet transform theory, this paper proposes a digital watermarking algorithm of the gray-scale images based on discrete wavelet transform (DWT). The results show that this algorithm not only makes the embedded watermark have good invisibility and the ability against various attacks, but also has strong robustness.

Due to the work of Baker, Gill and Solovay [SIAM J. Comput., 4 (1975), pp. 431–442] and others, it has become a paradigm that important open questions about complexity classes do not relativize. This paper develops restrictions of the standard oracle machine model for which, in contrast, positive inclusion relationships do relativize. Our results are obtained by uniform simulation techniques. As a consequence, the new oracle machine models exhibit the same computational power as do the corresponding nonrelativized devices.