Search

It is proposed that an energy optimization model of multi energy interaction in thermal power plants with wind power, photovoltaic and hydrogen production and hydrogen fuel cell system (HPHFCS). Considering that the wind power, photovoltaic and HPHFCS are connected to the auxiliary power system of thermal power plant, according to the output characteristics of the multiple power sources and auxiliary power load, an energy optimization model of multi energy interaction at the auxiliary power is proposed and analysed. This paper quantitatively analyses the role of wind power, photovoltaic and HPHFCS connected to the auxiliary power system of thermal power plants in promoting the local consumption of renewable energy sources and reducing the power consumption rate of thermal power plants. Numerical cases and simulation results validate the correctness and effectiveness of the energy optimization model in this paper.

The current status of the regional 10kV distribution network has problems such as the unreasonable capacity declaration, capacity redundancy, and line margin reduction due to a large number of distributed PV and electric vehicle access. In this regard, a redundancy margin analysis method of a 10kV regional distribution network based on demand-side response is proposed. The customer distribution substation power consumption data and operation data are analyzed to establish a 10kV regional distribution network redundancy margin analysis model, and the model is constrained. In the experiments, the proposed method is verified for analysis accuracy. The analysis of the experimental results shows that the proposed method has a high analytical accuracy for analyzing the redundancy margin of the 10kV regional distribution network.

With the rapid development of smart power terminals equipped with embedded operating systems, business APPs have emerged rapidly along with them. Since various business APPs have excessive access to system resources, this will lead to the waste of system resources, and at the same time, the system security is not guaranteed. Therefore, in this paper, based on the context-aware program access control hardware resource framework, an auxiliary tool is designed to control the access of business APPs with excessive authority. The auxiliary tool trains the model through the steps of training sample collection, sample data processing, feature extraction, parameter selection, and model training. By conducting simulation experiments, the results show that by using the auxiliary tool proposed in this paper, the device can well identify the business APPs with excessive privileges and reduce the abuse of related resources with a small load, which has almost negligible impact on the users’ usage.

To solve the problem of the low recovery rate of critical loads in the capacity efficiency margin calculation model of the distribution substation area, a data-driven capacity efficiency margin calculation model for the distribution substation area is designed by identifying the data characteristics of customer-side resources in the distribution substation area, approximating the functional relationship between mechanical and electrical energy using segmentation functions, finding the equipment regulation strategies corresponding to historical load sequences from the database, designing a dynamic optimal tide framework using data-driven technology, discovering the shortcomings that limit the performance of network carrying capacity, and constructing a capacity efficiency margin calculation model. Test results: Under three different simulation scenarios, the average values of key load recovery rates of the distribution substation capacity efficiency margin calculation model and the other two models are 64.946%, 56.223%, and 55.024% respectively, indicating that the designed distribution substation capacity efficiency margin calculation model is more effective when fully combined with data-driven technology.

YOLO v5 was born when Joseph Redmon, the father of YOLO, announced his retirement from the field of computer vision, and its structure was further optimized on the basis of YOLO v4. After the sort target tracking algorithm was proposed, the deep sort algorithm added the optimization of apparent features on the basis of the sort algorithm, which enables the model to be more suitable for practical situations and more mature for vehicle detection on traffic roads. Focusing on the new direction of road traffic detection, a single camera is used to detect and track dual-channel vehicles. In the process of experimental exploration, a “double vertical line” algorithm is proposed by introducing the attention mechanism into the YOLO v5 structure and combining with the deep sort target tracking algorithm, which provides a new thinking angle and flow chart for vehicle repeat counting. Finally, the detection results of vehicle identity and traffic flow are obtained through experiments, enabling vehicles to be detected and tracked in real time even on roads with a certain speed. In addition, the calculation law avoids the miscalculation of roads with forks, obtains excellent results on vehicle types and traffic flow, and uses id identity to give car information, which significantly improves the possibility that the model can be applied to practical problems.

Traditional target detection methods have poor tracking effect for targets with excessive displacement or occlusion between adjacent frames, and it is difficult to deal with high-speed moving targets. In view of the low detection rate of the current system in detecting long-distance weak and small targets, false targets and occlusion in the air, the working mode of convolutional neural network (CNN) and its advantages in target detection are analyzed, and a target detection method based on CNN is proposed, which can ensure the target detection accuracy and meet the real-time requirements of combat scenes, Finally, compared with other algorithms, the proposed method achieves the best results on four types of data. While the center error is low, it has high tracking success rate and low time complexity, which meets the requirements of high precision and real-time of the system.

This paper improves the core functions of the real-time video analysis system for edge computing which we proposed before. The improvement of this paper including distributed cluster status management and two-level horizontal expansion. The three-level task scheduling mechanism is realized, and the complete data link processing is realized. At the same time, RK3399pro is proposed to replace RK3399 as the hardware computing unit of the AI business. On the premise of satisfying real-time processing, the computing unit resources are saved. Compared with our previous system, the processing capacity of the new system is improved by 6-8 times.

Aiming at a series of issues on traditional boost converter which commonly used in photovoltaic maximum power point tracking (MPPT) devices, such as the limited output voltage, the low boost efficiency, the excessive heat generating, this paper proposed a novel z-source converter. Compared with traditional boost converter, several researches have done among topology structure, operational principle, boost factor. Based on the theoretic analysis, this paper given two schemes of photovoltaic MPPT system which using boost converter and z-source converter, the models of them are established based on Matlab. The simulation results indicate that the duty-ratio of photovoltaic MPPT system based on z-source converter is smaller, the stability of MPPT is better. The novel z-source converter can fundamentally solve the issues caused by traditional boost converter in photovoltaic MPPT devices, and it has incomparable advantages than traditional boost converter in the field of photovoltaic power generation.

In order to study the effect of blade structures on the induced excitation characteristics, two different structures are investigated, which are the blade with special trailing edge and the staggered blade. Firstly, based on the steady numerical simulation, the time-averaged velocity and pressure distribution are obtained at different Re numbers (5×103 and 1×104). Secondly, the unsteady numerical simulation method LES is applied to acquire the transient and complex flow field around the blade. It is found that both two blade structures significantly change the flow characteristics around the blade, especially in the wake flow region. The evident frequency is observed from the frequency spectrums. It is clear that the dominant frequency is originated from the vortex shedding, which is similar to the Karmen vortex street. Finally, the results will be useful for the design of blade with low noise purpose.

In order to improve the performance of electricity information collection system, this paper studies and proposes a new method of low-voltage power line carrier signal frequency domain noise reduction. It includes frequency domain decomposition of carrier signal, filtering noise reduction and signal frequency domain restoration. Then compared our method with the wavelet packet noise reduction method based on multi-resolution. The experimental results show that the proposed algorithm is better than the comparison algorithm under different modulation modes and carrier frequencies, especially suitable for carrier signal noise reduction under bad channel conditions. The algorithm proposed in this paper has good anti signal attenuation performance, and it can effectively enhance the SNR, and ensure the transmission performance and system reliability of the electricity information collection system.

This paper presents a broadband LNA using 70nm GaAs MHEMT process operating from 1.5GHz to 20GHz. The multi-feedback technique and input/output impedance match are applied to compensate the transistor gain roll-off with frequency and optimize the stability and reflection coefficients. The effect of the multi-feedback technique for stability, noise figure, gain flatness, linearity, and S-parameter performance are analyzed and discussed. Among the whole bandwidth, this LNA features gain of average 29.5dB increasing with frequency and noise figure of average 1.15dB. This LNA also achieves good reflection and isolation performance. The consuming of this LNA is blow 150mW and the chip area is less than 1.4mm*0.8mm. The figure-of-merit (FOM) is about 362.6 (GHz/mW). This LNA could be found its application in high data rate transmission, software defined radio, 5G communication, test instrumentation, wideband radar and other wideband systems.

In this paper, a broadband low noise amplifier (LNA) is presented with 70nm GaAs metamorphic high electron mobility transistor (MHEMT) process. The feedback and input/output impedance match techniques for broadband amplifier are presented. The stability, noise figure, gain flatness, and S-parameter performance of this LNA are analyzed and discussed. The small signal gain of the LNA is 20dB±0.2dB and noise figure of 0.86dB~1.35dB among operating from almost DC to 15GHz. This LNA exhibits reflection below -10dB, power dissipation of 70mW and chip area of 1.0mm*0.6mm. This LNA finds its application in many wideband systems.

The three-dimensional high-resolution imaging sonar uses multi-beam technology combined with synthetic aperture technology to achieve centimeter-level imaging of small underwater targets. It has a good imaging effect on underwater anchoring, sinking mine and buried mine target detection and underwater topography, and has a wide range of military and civil engineering applications. The system has a large amount of data transmission and computation, including multiple functional modules such as synthetic aperture, high resolution array processing, image processing, and target recognition. The software architecture is complex and real-time system development faces serious challenges. The blade servers have been widely used in the telecommunications, financial and other big data processing fields as for the high efficiency, stability and autonomy. This paper introduces a cluster 3D imaging signal processing system based on x240m5 blade server. The message queue communication (ZMQ) is used to realize the stable transmission of large data volume. At the same time, the mixed programming technology of C language and Matlab language combines the convenience of real-time and software integration, which greatly shortens the development cycle. The system runs stably and realizes the imaging recognition of the small spheres and their clump weights in the simulated target water.

Multi-path effecting and Dopplereffecting on missile-borne tester signal transmission based onZigbee application was researched. The maximum transmission distance was derived based on communication channel loss model and multi-path model. The effecting of multi-path on transmission distance was discussed. The effecting of Doppler shift was researched by speed of rocket sled and receiver position. Transmission speed and error rate of missile-borne tester was studied by Doppler effecting. The measure which reduced multi-path fading and Doppler effecting was gained. These can advance signal transmission quality and improve revamping of equipment.

In this paper, we proposed to use step heterojunctions emitter spacer (SHES) and InGaN sub-quantum well in AlGaN/GaN/AlGaN double barrier resonant tunnelling diodes (RTDs). Theoretical analysis of RTD with SHES and InGaN sub-quantum well was presented, which indicated that the negative differential resistance (NDR) characteristic was improved. And the simulation results, peak current density JP=82.67 mA/μm2, the peak-to-valley current ratio PVCR=3.38, and intrinsic negative differential resistance RN=-0.147Ω at room temperature, verified the improvement of NDR characteristic brought about by SHES and InGaN sub-quantum well. Both the theoretical analysis and simulation results showed that the device performance, especially the average oscillator output power presented great improvement and reached 2.77mW/μm2 magnitude. And the resistive cut-off frequency would benefit a lot from the relatively small RN as well. Our works provide an important alternative to the current approaches in designing new structure GaN based RTD for practical high frequency and high power applications.

To understand the failure modes and impact resistance of double-layer plates separated by water, a flat-nosed projectile was accelerated by a two-stage light gas gun against a water-filled vessel which was placed in an air-filled tank. Targets consisted of a tank made of two flat 5A06 aluminum alloy plates held by a high strength steel frame. The penetration process was recorded by a digital high-speed camera. The same projectile-target system was also used to fire the targets placed directly in air for comparison. Parallel numerical tests were also carried out. The result indicated that experimental and numerical results were in good agreement. Numerical simulations were able to capture the main physical behavior. It was also found that the impact resistance of double layer plates separated by water was lager than that of the target plates in air. Tearing was the main failure models of the water-filled vessel targets which was different from that of the target plates in air where the shear plugging was in dominate.