With the rapid development of environmental technologies such as renewable energy, smart grid, and electric power transportation, future power generation and power supply will show new features. The ...design of energy consumption characteristics of modern power system is more flexible and easy to control, which will also affect the scale of power generation system. This paper presents a combined capacity optimization method for a typical independent microgrid including solar photovoltaic, wind turbines, diesel generators and battery energy storage system. The mathematical models of photovoltaic, wind turbines, diesel generation system, battery energy storage system and electric vehicle charging loads are developed to improve the capacity optimization method. Research objectives include: 1) minimizing costs; 2) reducing greenhouse gas emissions; 3) reducing waste energy. This research can provide strong support for decision-making, analysis and strategy-making of multi-agent joint microgrid capacity.
In this paper, a new staged parameter identification method of energy storage system is proposed in different stages of Low Voltage Ride Through (LVRT). Firstly, according to the mathematical model ...of grid connected converter under LVRT process, the time-domain response formula of current is established, and then the particle swarm optimization algorithm is used to identify the Proportional-Integral (PI) control parameters of converter and the control parameters of LVRT. Secondly, considering the sensitivity difference between different parameters, the control parameters of LVRT are identified in phases, and an improved fitness function of particle swarm optimization algorithm is proposed to increase the identification accuracy. Finally, the effectiveness of the parameter identification method is verified by simulation and experiments.
Electric vehicles (EVs) are being increasingly integrated into the electric grid as a result of their popularity and people's travel patterns. Especially in Universities, the campuses have a large ...number of people in a short period of time, which increases the electric vehicles' charging load. In this paper, an optimized charging and discharging method of EVs within a campus microgrid is proposed, which would significantly improve grid operation stability. This article used the disorderly charging of EVs using probability density functions and the Monte Carlo (MC) method. By analyzing the EVs owners' willingness to charge in the campus area, the distribution of start-of-charge (SOC) of the vehicles, and the EVs' stay time on campus, the amount of charge is estimated. To achieve a suitable charging scenario for a campus microgrid, the analysis parameters are coded and optimized using the Genetic Algorithm (GA). According to the characteristics of university class-times, it is expected that the charging of vehicles entering campus will cause peak-load periods. During the time period from 6:00 to 24:00, 1,000 EVs that follow these plans will exist. The GA algorithm is used to allocate the best charging time for the EVs, thereby adjusting the peak-to-valley gap of the campus power grid. Further, four types of GA are compared and it is realized that the input parameters have a significant impact on the outcomes. The results demonstrate the impact of optimized charging on improving microgrid stability and reducing charging costs.
ROI Detection of Hand Bone Based on YOLO V3 Shi, Yingcong; Qiao, Jiaxin; Song, Jian ...
2021 IEEE International Conference on Artificial Intelligence and Computer Applications (ICAICA),
2021-June-28
Conference Proceeding
China-05 method is the standard method for bone age assessment (BAA) in China. BAA is a typical target detection research project. In this study, a target detection algorithm was proposed combining ...image processing and deep learning based on the regions of interest (ROI) of 13 hand bones concerned by RUS-CHN method in China-05 method for bone age assessment. We sorted 13 actual ROIs blocks into 4 ROI collection regions (ROI-C), and detected ROI-C through image processing methods such as image equalization, binarization, rotation and positioning. After that, the YOLO V3 model is used to train the new data set composed of ROI-C to detect the actual ROI. This detection algorithm solves the problems of poor robustness, inaccurate detection area positioning, target detection error or detection duplication when we use traditional single image processing or deep learning algorithm. This method greatly improves the detection accuracy of ROI and has great significance for accurate assessment of bone age.
In the power system, wind power generation plays a significant role. Research on the frequency regulation capability of wind turbines is very important. MATLAB is a commonly used simulation platform ...for control method research, however, the simulation speed of a large-scale power system containing wind power generation is quite slow. To improve the research efficiency as much as possible while meeting research needs, an integrated model of the doubly-fed induction generator (DFIG) model and system frequency response (SFR) model is proposed in this paper. First, a simplified DFIG model is established which only retains the wind turbine, swing equation of rotor, speed regulator with frequency regulation control, and pitch controller. Then, the integrated method of the simplified DFIG model and SFR model is introduced. The simulation result shows that the frequency response of the integrated model is consistent with the complete power system model containing a detailed DFIG model, and the simulation speed is increased by more than 130 times. In addition, since the wind speed input-related part is preserved, the integrated model can be used for frequency regulation control research of wind turbine generators considering the continuous change in wind speed.
Donor specific human leukocyte antigen (HLA) antibodies (DSA) are a significant cause of allograft failure. However, it has been reported that some DSA negative patients still experience allograft ...failure. In addition, some DSA positive patients maintain good graft function for >20 years. These findings suggest that while DSA is a cause of failure, it is not the sole risk factor for graft dysfunction and that the presence of DSA alone may not predict the time course of graft failure. Here, we report the predictive value of a proprietary panel of four biomarkers in long-term renal allograft outcome. A total of 310 consecutive patients, who received kidney transplants between 1999 and 2012, were included in this study. Recipient sera was tested for HLA antibodies and biomarkers at 3, 6, 12, 24, and 36 months post-transplant. HLA antibodies were identified using Labscreen single antigen beads. The biomarker combination (BMC) test consisted of a proprietary panel of 4 biomarkers and was performed using Luminex. Sera were defined as positive when any one of the 4 biomarkers became detectable. Sera of normal healthy people were used as negative controls. Graft survival analyses were performed and compared between different patient groups based on the positivity of DSA and BMC. Our results indicate that 57% of DSA negative patients and 54% of DSA positive patients had detectable biomarkers. There was no significant difference in BMC positive patients between the DSA positive and negative groups, which suggests that presence of BMC is not associated with HLA DSA. DSA positive patients had a 10% lower 10-year graft survival rate than patients without DSA, while BMC positive patients had a 25% lower 10-year graft survival rate than patients without detectable BMC. When DSA negative patients were divided into two groups based on the positivity of BMC, BMC positive patients had a 20% lower 10-year graft survival rate compared to BMC negative patients (p<0.05). Similarly, when DSA positive patients were divided into two groups based on the positivity of BMC, BMC positive patients had a 30% lower 10-year graft survival rate compared to BMC negative patients (p<0.01). When both DSA and BMC testing results were considered, DSA and BMC double positive patients had the lowest and double negative patients had the highest graft survival rates. The survival rates for the BMC alone and DSA alone positive groups were in between (p<0.001). Multivariate Cox models confirmed that BMC was an independent risk factor for graft failure, with a higher hazard ratio than DSA (BMC=2.60 versus DSA=1.64). In conclusion, serum BMC is an independent predictor of graft failure. BMC was more significantly associated with graft failure than DSA. In combination with DSA, BMC better predicted graft outcome than DSA or BMC alone.
Earlier-stage evaluations of a new AI architecture/system need affordable AI benchmarks. Only using a few AI component benchmarks like MLPerf alone in the other stages may lead to misleading ...conclusions. Moreover, the learning dynamics are not well understood, and the benchmarks' shelf-life is short. This paper proposes a balanced benchmarking methodology. We use real-world benchmarks to cover the factors space that impacts the learning dynamics to the most considerable extent. After performing an exhaustive survey on Internet service AI domains, we identify and implement nineteen representative AI tasks with state-of-the-art models. For repeatable performance ranking (RPR subset) and workload characterization (WC subset), we keep two subsets to a minimum for affordability. We contribute by far the most comprehensive AI training benchmark suite. The evaluations show: (1) AIBench Training (v1.1) outperforms MLPerf Training (v0.7) in terms of diversity and representativeness of model complexity, computational cost, convergent rate, computation, and memory access patterns, and hotspot functions; (2) Against the AIBench full benchmarks, its RPR subset shortens the benchmarking cost by 64%, while maintaining the primary workload characteristics; (3) The performance ranking shows the single-purpose AI accelerator like TPU with the optimized TensorFlow framework performs better than that of GPUs while losing the latter's general support for various AI models. The specification, source code, and performance numbers are available from the AIBench homepage https://www.benchcouncil.org/aibench-training/index.html.
A methodology is proposed to estimate the actual influence of important factors during steady-state Temperature Humidity aging tests on SiN passivated MEMSs: temperature-humidity effects, protection ...by epoxy resins and SiN passivated layer. The developed packages become a hybrid thin-film hermetic encapsulation consisting of an internal shell using PE-CVD SiO, a seal layer coating with resin, and an external protective layer formed by PE-CVD SiN. The process is fully compatible with standard low-cost back-end-of-the-line (BEOL) technologies for LSIs package. This hybrid structure was very effective for protecting the MEMS device from external moisture. In this work, accelerated aging tests on MEMS were carried out, with or without SiN passivated layer.