Search

The current disturbance classification of power quality data often has the problem of low disturbance recognition accuracy due to its large volume and difficult feature extraction. This paper proposes a hybrid model based on distributed compressive sensing and a bi-directional long-short memory network to classify power quality disturbances. A cloud-edge collaborative framework is first established with distributed compressed sensing as an edge-computing algorithm. With the uploading of dictionary atoms of compressed sensing, the data transmission and feature extraction of power quality is achieved to compress power quality measurements. In terms of data transmission and feature extraction, the dictionary atoms and measurements uploaded at the edge are analyzed in the cloud by building a cloud-edge collaborative framework with distributed compressed sensing as the edge algorithm so as to achieve compressed storage of power quality data. For power disturbance identification, a new network structure is designed to improve the classification accuracy and reduce the training time, and the training parameters are optimized using the Deep Deterministic Policy Gradient algorithm in reinforcement learning to analyze the noise immunity of the model under different scenarios. Finally, the simulation analysis of 10 common power quality disturbance signals and 13 complex composite disturbance signals with random noise shows that the proposed method solves the problem of inadequate feature selection by traditional classification algorithms, improves the robustness of the model, and reduces the training time to a certain extent.

Objective: Bar code recognition is widely used. This paper introduces a method of using digital image algorithm to complete the identification of stain bar code. Methods: firstly, the barcode image was binarized and positioned; secondly, the pixels in the barcode image were stacked on the abscissa axis by projection technology to remove the barcode stains; finally, the barcodes were effectively identified by the center distance algorithm. Results: the experiments show that the barcodes with different degrees of pollution are well recognized by our method, which is better than the traditional recognition algorithm. Conclusion: This paper proposes a new barcode recognition algorithm, which can quickly and accurately identify the bad barcode

The paper proposes a novel self-reference voltage pre-regulator for LED driver chip. The system modeling of the self-reference voltage pre-regulator is established, which is composed of pre-regulator circuit and reference voltage generating circuit. The reference voltage generating circuit generates a reference voltage, in which the circuit returns the reference voltage as the reference voltage of pre-regulator circuit, therefore, the preregulator outputs a stable voltage for the LED driver chip. In this paper, the circuit implementation is described in detailed, in order to enhance the stability of the system, the miller compensation is used to the circuit, which makes the system more stable. Simulations of the circuit show that the pre-regulator not only can output a stable voltage for the chip, but also the load regulation, line regulation and power support ripple rejection of the pre-regulator exhibit high performance.

To study the damage problem of reconnaissance satellite under the information countermeasures soft kill conditions, fault tree of satellite performance has been built and destructive level to distinct the degree of the satellite damage has been proposed linked by the satellite reconnaissance information link. Satellite damage assessment based on the multi-state fault tree was proposed and has been applied into some reconnaissance satellite under the conditions of information countermeasures.

This paper presents a CMOS low-dropout (LDO) regulation with a dynamic-biased gain stage. The dynamic-biased gain stage which has the dynamic adjustable output resistance, current and gain is designed as a second stage driving the power transistor. A LDO regulator with the proposed gain stage has been implemented in a 0.18mm standard CMOS process and occupies 0.11mm2 chip area. The experimental results show that the proposed LDO regulator can be stable without any additional on-chip compensation capacitors or complex circuit. The LDO regulator provides a full load transient response of settling time less than 15ms with a 2.2mF load capacitor and the voltage variation is less than 40mV. Furthermore, the load regulation and line regulation are 321mV/mA and 1.8mV/V, respectively.

The security defect detection technology based on source code usually makes use of static analysis methods to detect security vulnerabilities in the test software, which does not consider the runtime information of program execution and the interaction information with the program runtime surrounding environment, so the security defect detection technology based on source code usually results in higher false positive rate. The binary program vulnerability detection methods based upon dynamic analysis usually have lower false positive rate, but their effectiveness depends entirely on test case generation, the detection efficiency of dynamic analysis based on binary program is lower. Moreover, the research of source code vulnerability detection technique and binary executable vulnerability detection technique are essentially fragmented, without considering the relationship between the two classes of methods. In order to improve the efficiency for source code and binary executables vulnerability detection, we presents a collaborative analysis method for vulnerability detection of source code and binary executables based upon the unified defect mode set, meanwhile, we verify the effectiveness of our method using a concrete example in this paper.

For RFID mutual authentication security protocols which need back-end server have poor mobile performance, it cannot be applied to environments which have poor wireless network. The RFID mutual authentication security protocol based on key dispersion algorithm is designed in this paper, which proposes a new method for managing and distributing keys for a large number of tags. This protocol which does not need the back-end database can be achieved security mutual authentication between the reader and the tag, the reader only need to store a pair of root keys, the key dispersal algorithms can disperse the corresponding mutual authentication keys with the tag identification number. Compared with the other of the same type RFID mutual authentication security protocols, results indicate that the proposed protocol in this paper has higher efficiency and security.

In this work, the proportional navigation (PN) law with gravity compensated based on GPS/INS combination is formulated at first, and a guidance law is designed via Variable Structure Control (VSC). Secondly, as the primary reason of chattering in the traditional VSC, the input Relay-characteristic is improved in order to reduce chattering at the end time, and three continuous functions are given. To evaluate the performance of the two guidance solutions, a numerical simulation of the missile's guidance with a two-order dynamics of the autopilot is used. The results show that, guidance law based on VSC is faster than PN law. Moreover, the improvement has a great effect on reducing chattering at the end time, and the third method addressed by this work has a better result.