Abstract：With the deepening of the theory and application of artificial intelligence technology, industrial manufacturing is expected to become the key field of its application.In order to provide reference and guidance for formulating industrial intelligence development policies in China,this paper analyses the technical mechanism of industrial (artificial) intelligence in solving industrial problems and establishes an algorithm-application analysis system.Focusing on the representative vertical industries,we also study the application examples and characteristics of industrial intelligence.And finally,our team expound the current situation and development trend of industrial intelligence-related industries.

Abstract：Aiming at the complexity and particularity of the production environment of the steel industry, this paper designs a NB-IoT metallurgical node temperature acquisition and remote monitoring system. The system is based on the AD8495 amplifier processing K-type thermocouple cold junction compensation to complete the on-site temperature and other parameters acquisition, STM32 node linearization algorithm processing, communication with remote computer through Modbus and NB-IoT network data transmission, and based on LabVIEW host computer remote monitoring and safety management of blast furnace channel special environment. Experiments show that compared with the traditional single local monitoring, the system makes full use of the advantages of NB-IoT technology, with local and remote synchronous real-time monitoring and low cost, wide coverage, wireless multi-connection and so on.

Abstract：Embedded authentication technology is the author′s patented technology.The innovative product based on embedded certification is a new method for the protection of the core area of industrial control systems. Following the PKI and IPK cryptosystems, embedded authentication is implemented on the industrial switch, and the SM2 asymmetric algorithm is applied to the terminal access authentication process. This paper with the main source code introduces the implementation method and technical details of the switch embedded authentication server and Windows/Linux authentication client.

Abstract：With the rapid development of China′s manufacturing industry, intelligent and networked, a scientific protection program is particularly important. This paper introduces the current situation of network use of a company from the perspectives of risk status, risk identification, communication security hazards, etc., and analyzes its characteristics and existing security risks. A well-designed information security protection scheme is proposed, which makes the network protection of the entire company more secure through the design of a well-established network architecture and a reasonable supervision mechanism.

Abstract：With the rapid development of communication technology, the "best" service capability provided by traditional network can no longer meet the demands of various vertical industries for high-performance networks such as ultra-low delay, deterministic delay and high reliability, and 5G deterministic network emerges at the historic moment.5G deterministic network builds higher quality network service capability mainly through 5G key technologies such as network slicing, mobile edge computing and time-sensitive network. Starting from the capability of personalization and customization of 5G network, this paper analyzes the requirements of four typical industrial application scenarios on the determination of 5G network performance indicators, puts forward the deterministic network implementation architecture and key technologies, and presents some typical cases.

Abstract：5G industrial Internet is the hot spot and focus in the development of 5G and industrial Internet in China. Edge computing is introduced in both industrial Internet and 5G, but the concern of it is different. This paper analyzes the industrial Internet application based on 5G edge computing and its development, summarizes the industrial Internet edge computing technical architecture, and presents the cloud edge collaborative technical architecture of 5G industrial Internet edge computing.

Abstract：The current status and vision of 3GPP V2X communication security technology are briefly introduced. Corresponding security requirements are given based on the security threatens of V2X communication. This paper briefly introduces the principal method of implementing V2X communication security, and proposes the service architecture, PKI architecture and multi-PKI system mutual trust mechanism for V2X communication security. Finally, the status of domestic V2X security standardization and industrial development prospects are introduced.

Abstract：This paper presents the Quality of Service(QoS) mechanism and the development process of the mobile network. The ability and deficiency of both 3rd-Generation(3G) gold, silver and cooper and the 4rd-Generation(4G) QoS Class Identifier(QCI) in the implementation of customized network are analyzed, and the QoS design ideas, implementation principles and outstanding advantages of 5rd-Generation(5G) network are also introduced. Combined with the productized experience of 4G QoS capability operation in To B mode, the QoS capability of 5G network is end-to-end productized encapsulating for the industry customers to achieve the 5G customized network capability based on the QoS flow. Meanwhile, combined business characteristics with business requirements, e.g. industrial Internet, augmented reality, autonomous driving, intelligent transportation and other industries and applications, the typical application solutions and productized ideas of 5G QoS mechanism are designed.

Abstract：In order to solve the problem of secure storage and exchange of manufacturing big data, a storage and exchange mode oriented to manufacturing big data is proposed. The security risks in storage and exchange of manufacturing big data are mainly analyzed. Key technologies such as searchable encryption technology, attribute-based access control technology and process measurement method are proposed to solve the leakage risks in data storage and exchange. The current situation and shortcomings of these key technologies are summarized and analyzed, and the future development is prospected.

Abstract：Aiming at the problem of low correct recognition rate and relying on experience to select parameters in gear box fault diagnosis by using neural network, a fault diagnosis method of gear box based on particle swarm optimization BP network is proposed. In this paper, a fault model is established by extracting characteristic parameters from gear vibration principle. The model takes eigenvector of gear box as input and fault type as output. The fault diagnosis of gear box is realized by BP neural network, probabilistic neural network and particle swarm optimization BP neural network. The simulation results show that the convergence speed of BP neural network for gear box fault diagnosis is slow, and the recognition rate of fault diagnosis is 82%. The recognition rate of probabilistic neural network model fault diagnosis is determined by selecting spreads based on experience, and the maximum recognition rate is 98%. The recognition rate of BP neural network fault diagnosis based on particle swarm optimization is 100% and adaptive ability is strong.

Abstract：In order to serve the customer service intelligent dialogue system of the State Grid Customer Service Center, it is necessary to extract knowledge from a large number of documents and traditional knowledge base as well as dialog data. This paper proposes a new knowledge graph framework that integrates fact graph and event evolutionary graph, which can be based on multiple sourcea data. The constructed knowledge graph has good performance in the vertical domain of accurate question and answer, customer service system knowledge support, dialogue management guidance, knowledge reasoning and so on. New knowledge graph was put into use in the customer service center question and answer system, which changed the working mode of the customer service and greatly improved efficiency of the customer service.

Abstract：Image classification is to distinguish different types of images based on image information. It is an important basic issue in computer vision, and is also the fundamental for image detection, image segmentation, object tracking and behavior analysis. Deep learning is a new field in machine learning research. Its motivation is to simulate the neural network of the human brain for analytical learning. Like the human brain, deep learning can interpret the data of images, sounds and texts. The system is based on the Caffe deep learning framework. Firstly, the data set is trained and analyzed, and a model based on deep learning network is built to obtain the image feature information and corresponding data classification. Then the target image is expanded based on the bvlc-imagenet training set model. And finally，"search an image with an image" Web application is achieved.

Abstract：A novel high-efficiency and wide bandwidth rectifying circuit is proposed. By connecting a microstrip line in series at the front end of the rectifying diode, and adjusting the length of the line within its inductive range, the capacitive reactance generated by the diode at the fundamental frequency can be offset, and thereby a good impedance matching is realized by using the 1/4 wavelength impedance conversion method rather than the single branch matching method. In this way, the circuit is able to achieve a compact structure and meanwhile an enhanced working bandwidth. Moreover, the design of the sector branch in the direct current(DC) filter has been improved, e.g. its length of branch is adjusted according to its effect to the DC voltage waveform, so that the output DC voltage ripple can become more stable, and hence the rectifying efficiency of the circuit can be improved. A rectifying circuit working at 5.8 GHz has been designed. Measurement results show, when the input microwave power is in a range of 0～16 dBm, the rectifying efficiency of the circuit is over 55% with a maximum value up to 78.7%; at an input microwave power of 13 dBm, the circuit realizes a bandwidth of 16.6% for the rectifying efficiency more than 70%.

Abstract：For researching the transient dose rate effect of the H-type NMOS based on submicro SOI technology，the 3D modeling and numeric simulation of the 0.13 μm SOI NMOS is processed using TCAD tool. The relationship between the transient dose rate and the currents of drain and body-contact of the NMOS is released by using seven points which the γ-transient dose rate ranges from 1×108 to 1×1010 (Gy(Si)/s). From the simulation results, we can concluded that there is little influence on the device when the γ-transient dose rate is less than 5×109 Gy(Si)/s. The H-type NMOS device has strong ability of anti-transient dose rate radiation. The results of H-type NMOS through 3D numeric simulation provide the reference for rad-hard design in VLSI.

Abstract：With the development of attack technology, modeling attacks based on ML, DL and other technologies threaten the security of PUF seriously. In view of the flaw in Glitch PUF static output, this paper first proposes the multilayer perceptron algorithm for Glitch PUF machine learning to analysis nonlinear output data. The experimental results show that the MLP algorithm proposed in this paper can significantly reduce the prediction error rate compared with the logistic regression algorithm and the random forest classification algorithm.

Abstract：With the rapid development of smart city and smart terminal, security monitoring system is widely used, and higher requirements have been raised on power management components. LDO(low dropout linear regulator) is an ideal power management solution with low noise, high PSRR(power supply rejection ratio) and simple peripherals. It can meet the demand of security monitoring system which is very sensitive to noise. This paper introduces the application of high performance LDO in security monitoring, key technical parameters of LDO and its impact on security monitoring system.

Abstract：In this paper, a three-level(3-level) envelope modulator structure is proposed. The conventional two-level(2-level) switching converter is replaced by the 3-level switching converter. The switching current slew-rate can adapt to the amplitude of the input envelope by the three-level fitting method. The output current of the linear stage can be reduced effectively, and the dynamic power consumption of the linear stage can be reduced, thus the overall efficiency of the system can be improved. The circuit is designed with 55 nm CMOS technology. The simulation results show that the output power of the power amplifier is 27.5 dBm and the efficiency is about 85% under 3.3 V power supply and 10 MHz LTE input signal. Compared with the conventional structure, the efficiency is improved by 2%.

Abstract：Four-rotor aircraft physical mathematical model is established for the high difficulty in the Four-rotor aircraft flight control. This paper analyzes the shortcomings of several current attitude algorithm schemes, and eliminates the rotation noncommutativity error in the operation by selecting quaternion attitude algorithm. According to predictability of angle data at the next moment, the Kalman filter is designed to eliminate the noise signal caused by environmental electromagnetic interference, temperature drift and high frequency oscillation when motor running in high speed. A micro control system circuit and a cascade PID regulator are built to make the angular velocity of the system returning to its equilibrium position is directly proportional to the angle, it solves the oscillation caused by large angle error and the lack of strength caused by small angle error. The test results outdoors show that the problem of poor stability of aircraft is solved and can meet control requirement under the model of flight, including stable hovering, forward, backward, yawing and so on.

Abstract：In this paper, a calibration method using a corner reflector as the marker is designed for the calibration of the coordinate system of the LiDAR and inertial measurement unit(IMU). Firstly, the pre-processing methods of region segmentation, ground filtering and marker point extraction are used to extract the marker points. And then the three-dimensional point cloud matching method is used to convert the coordinate system joint calibration into the point cloud matching problem, using the iterative closest point(ICP) algorithm to obtain the coordinate transformation matrix of the two coordinate systems. The calibration results are compared with results obtained by the traditional least squares algorithm, which shows that the calibration method using 3D point cloud matching is feasible.

Abstract：The zero-drift phenomenon often occurs in piezoelectric accelerometers when performing high-g impact measurements, which seriously affects the accuracy and reliability of test results. In order to solve this problem, the influence of the RC time constant on the zero-drift in the measurement circuit is analyzed by establishing the mathematical response model of the piezoelectric sensor impact measurement. At the same time, a three-op amp differential charge amplifier circuit is designed for the common mode noise of the transmission cable. The test proves that the circuit can detect and amplify the small charge of ±20 pC while suppressing the common mode noise interference. Finally, the contrast shock test before improvement was compared with that after improvement by the Endevco impact table in order to verify that the improved design is effective in suppressing zero-drift in the impact measurement, and provides certain reference value for engineering use.

Abstract：AD9371 is a 5G mobile transceiver system chip which covered all sub-6G frequency bands newly provided by ADI corporation. Because the bandwidth and data rate are very high，when this AD9371 receiver worked under the cross-correlation peak algorithm, if a burst packet reached, for CAZAC synchronize word sequence′s cross-correlation defect, this synchronize detector will output a value jumping，causing the synchronize detected threshold beyond the truly threshold, so that all cross-relation peak detect algorithm lost its synchronize status. To solve this burst problem, this paper provides an unite burst frame edge and bit synchronize detect algorithm based on AD9371 transceiver chip with a independence frame edge detect function added. The practical test result proves it is suitable for AD9371 burst channel.

Abstract：Aiming at the loading imbalance of subdomain controller in multi-domain deployment in SDN, this paper proposes an adaptive switch migration strategy for balancing controller loads. By detecting the load of controller in each subdomain and choosing the immigration domain and emigration domain based on adaptive general algorithm, this method could achieve the global optimal selection of migration domain. In order to balance the controller load, it implements survival time and elimination mechanism to migrate the high load degree switch from the emigration domain to immigration domain. The simulation shows that the efficiency of migration has increased by 19.7% when compared with the existing migration algorithm. The controller load and the number of switches in each subdomain have been balanced and enhance stability of network.

Abstract：To detect overlapping communities more conveniently and accurately, an integrated overlapping community detection(IOCD) algorithm is proposed, which considers inferring overlapping communities from multiple non-overlapping community structures. Firstly, an eigenvector is generated for each vertex according to the results of the basic community detection(CD) algorithm. These features are used to detect the overlapping regions with dense connections in an unsupervised learning manner. That is to say, non-overlapping CD solutions are used to extract the implicit feature information associated with each vertex. Then, some iterations are made to maximize the possibility that each vertex belongs to its own community. Experiments on synthetic network and real community network datasets show that the proposed IOCD algorithm is superior to other similar algorithms under three standard metrics, and is almost unaffected by the basic CD algorithm.

Abstract：Extracting valuable information from a large number of report data is a very important step in the enterprise market analysis. And how to show a large amount of data to the enterprise managers in the short time with the form of diagramming is the significance of this study. With the support of Ajax, the page uses the graphical reporting technology ECharts, making the rigid report data become clear and visualized to the enterprise managers, and provide them a solid foundation for their precise market judgment.

Abstract：A design method based on dual-machine hot standby is proposed. The dual-machine hot standby master-slave working mode is adopted, and a full-duplex working mode of dual-system hot standby is simulated by a certain voting algorithm. The two-machine hot standby design is carried out from the aspects of fault detection, control voting, data synchronization and listening guard. On the basis of not increasing hardware, the system reliability is improved as much as possible to ensure the uninterrupted operation of the system. It has been verified that the standby computer can perform function control quickly and "seamlessly" when all or part of the main control computer functions fail.

Abstract：The processing board is the core of the miniaturizing IR imaging module. Every task in the designing process of the miniaturizing processing board is introduced, including the designing of overall scheme,the selection of main components, the measures in the designing of power-distributed-system(PDS)，signal integrality(SI) and power integrality(PI). The institution of wiring constraint rules of the interfacing signals between the LPDDR2 and FPGA is also introduced.The calculation method of the memory interfacing bandwidth and the improving in memory interfacing bandwidth by hard memory controllers in Cyclone V FPGA is discussed.

Abstract：A power controller for wearable system is designed based on MCU control. It has charge/power supply management of double batteries; built-in single chip computer: soft control of power switch and remote shutdown; real-time monitoring and indicating the working status of the controller, which can indicate the battery power information; communication with the host computer to upload the working status and battery power information; full protection, over-current protection, short circuit protection, input reverse protection, parallel operation of double batteries and other protective work. The utility model has the advantages of high efficiency, small size, simple operation, full function, flexible configuration of double batteries, good electromagnetic compatibility of metal case and high reliability. Solar power or hand-driven generator can be used to charge the battery through the controller to improve its field adaptability.

Abstract：A multi-mode control strategy for full-bridge buck-boost(FBBB) converter is proposed to solve the problems of high heat dissipation requirement, low efficiency and poor reliability in wide input voltage range. This strategy determines the working mode of the converter according to the relationship between the input voltage and the output voltage. By combining the average current control with dual-edge modulation and phase-shifting control, the FBBB converter can achieve a smooth transition between step-down, step-up and step-down modes. The converter works in the effective duty cycle region and conducts inductive current. Effective control is taken to ensure the safe and reliable operation of the converter. Type-III compensation network is used to improve the control loop and achieve high precision output and fast dynamic response. Finally, a simulation model is built to verify its correctness and effectiveness.

Abstract：In order to convert the low-voltage direct current of photovoltaic panels and fuel cells in distributed power generation system into high-voltage direct current，a high step-up DC/DC converter with pump-up capacitor based on switch capacitor and switch inductance network is proposed. This converter further increases the voltage gain and reduces the voltage stress of the power device.Using magnetic integration technology, two inductors are wound on the same magnetic core to form coupled inductor, which improves the power density of the converter.Finally, the basic working principle of the topology is analyzed, and a prototype with input voltage of 30 V, output voltage of 380 V and rated power of 200 W is made. The experimental results verify the correctness of the theoretical analysis.