检索结果-内蒙古大学图书馆

A Co Concentration Prediction Method for Electronic Nose Based on Trellisnet with Gated Recurrent Unit and Dilated Convolution

学校读者我要写书评

暂无评论

SSRN 2023年

作者： Zhu, Zhengyang Jiang, Qingming Wang, Mingxiang Xu, Min Zhang, Yiyi Shuang, Feng Jia, Pengfei School of Electrical Engineering Guangxi University Nanning China College of Food Science and Bioengineering Xihua University Chengdu China Guangxi Key Laboratory of Power System Optimization and Energy Technology China Guangxi Key Laboratory of Intelligent Control and Maintenance of Power Equipment Guangxi University Nanning China

As one of the harmful substances in the atmosphere, carbon monoxide (CO) is harmful to human beings. With the wide application of gas sensors and machine learning algorithms, the accuracy of concentration predicting of various gas such as CO is constantly improving. Now we apply TrellisNet, a network utilizing both recurrent and convolutional techniques, to gas concentration prediction, which is also a time-series prediction task, with the aim of improving its performance. To enhance TrellisNet’s ability to retain long time series information, we replaced the activation units in each layer of the model with GRU. Compared to using LSTM as the activation unit, our approach has lower computational complexity and offers a more stable model. Additionally, we introduced dilated convolutions in each layer, allowing the model to establish connections with as many past time steps as possible at a given time point, even with fewer layers. This further enhances the preservation of long time series information. We named our improved technique trellis convolutional dilated network (TrelliSense). Due to the injection of the same input values in each layer, TrelliSense also exhibits superior training stability. Experimental results demonstrate that TrelliSense outperforms other time prediction networks, including TCN, LSTM, and GRU, in terms of all error metrics (MAE, RMSE, SMAPE). Therefore, we argue that TrelliSense is a better method for predicting CO concentration. © 2023, The Authors. All rights reserved.

关键词： Convolution

Supervised Drift Sensor Sample Classification Method Based on Meta-Learning

学校读者我要写书评

暂无评论

SSRN 2023年

作者： Zhu, Zhengyang Ling, Haikui Zhang, Yiyi Liu, Jiefeng Shuang, Feng Xu, Min Jia, Pengfei School of Electrical Engineering Guangxi University Nanning China Guangxi Key Laboratory of Intelligent Control and Maintenance of Power Equipment Guangxi University Nanning China College of Food Science and Bioengineering Xihua University Chengdu China Guangxi Key Laboratory of Power System Optimization and Energy Technology China

With the advancements in materials and pattern recognition algorithms, electronic nose (E-nose) has been continuously developing and finding increasingly widespread applications. However, due to factors including but not limited to variations in temperature and humidity, as well as chemical and physical interactions among analytes and gases, E-nose is prone to drift during practical usage, leading to decreased accuracy of pattern recognition algorithms. Currently, the primary methods for addressing the issue of sensor drift in sample classification are semi-supervised learning and unsupervised learning. Supervised learning has been limited in the past due to the need for large amounts of up-to-date samples to retrain the models. But now, this challenge can be overcome with the help of meta-learning. Meta-learning, also known as "learning to learn", aims to equip models with the ability to acquire a learning capability and automatically learn some meta-knowledge. To tackle the challenge of few-shot learning, we applied the model-agnostic meta-learning (MAML) architecture from meta-learning and designed a residual classifier with learnable per-parameter learning rates. Additionally, we present a Gini value-based approach to enhance the certainty of model output labels in a multi-task setting. As a result, we achieved high classification accuracy on a publicly available long-term drift dataset. © 2023, The Authors. All rights reserved.

关键词： Electronic nose

An Electronic Nose for Harmful Gas Early Detection Based on a Hybrid Deep Learning Method H-Crnn

学校读者我要写书评

暂无评论

SSRN 2023年

作者： Mao, Guosheng Zhang, Yanmei Xu, Yang Li, Xiaoyu Xu, Min Zhang, Yiyi Jia, Pengfei School of Electrical Engineering Guangxi University Nanning China Guangxi Key Laboratory of Intelligent Control and Maintenance of Power Equipment Guangxi University Nanning China College of Food Science and Bioengineering Xihua University Chengdu China Guangxi Key Laboratory of Power System Optimization and Energy Technology China

Among harmful gases, carbon monoxide (CO) has posed a significant threat to human safety due to its concealment and harmfulness. Therefore, it’s imperative for the preservation of human safety to implement early detection for CO concentration. However, traditional methods are plagued by the limitations of insufficient accuracy and the incapacity to make early prediction. To address this issue, a novel technique, namely H-CRNN, is presented for electronic nose (E-nose) to perform early prediction of CO concentration, leveraging the strengths of convolutional and recurrent neural network to efficiently capture long-term dependencies within data. Also, shortcut connection and gated attention mechanism is proposed to enhance the capacity to capture crucial features. Furthermore, a linear process unit has been incorporated to handle the scale insensitivity issue. The experimental results show that the proposed H-CRNN outperforms RNN with a notable average reduction of 50.51% in the root relative squared error and 53.8% in the relative absolute error, and exhibits varying degrees of improvement when compared to state-of-the-art algorithms (TCN, TPA-LSTM, STCN, LSTNet, LSTM). Additionally, the H-CRNN achieved an average accuracy rate of 96.42%, surpassing all other algorithms and demonstrating the highest level of performance. Therefore, our work substantiates the proposed H-CRNN as an applicable method for early prediction of CO concentration. © 2023, The Authors. All rights reserved.

关键词： Carbon monoxide

Coo-Snse Monolayer: A High Potential Candidate for Sf6 Characteristic Decomposition Gas Adsorption and Detection

学校读者我要写书评

暂无评论

SSRN 2023年

作者： Cao, Jianjun Wang, Mingxiang Chen, Dachang Zhang, Yiyi Liu, Jiefeng Jia, Pengfei School of Electrical Engineering Guangxi University Nanning China School of Electrical and Electronic Engineering Wuhan Polytechnic University Wuhan China Guangxi Key Laboratory of intelligent Control and Maintenance of Power Equipment Guangxi University Nanning China Guangxi Key Laboratory of Power System Optimization and Energy Technology China

Insulation performance testing of gas-insulated switchgear (GIS) and treatment of SF6 decomposition products under partial discharge are two crucial studies. In this paper, the adsorption behavior and sensor properties of five characteristic decomposition gases of SF6 (HF, H2S, SO2, SOF2, SO2F2) on intrinsic and CoO-doped SnSe monolayers are investigated in detail based on density-functional theory (DFT). The results showed that the adsorption behaviors of HF, H2S, SOF2 and SO2F2 on SnSe monolayers were all physisorption (van der Waals force), among which the best adsorption was SO2 with an adsorption energy of -0.618eV, which is a weak chemisorption. After doping with CoO, the adsorption capacity of SnSe monolayer for five gases was significantly enhanced, in which the adsorption energies for SO2 and SOF2 reached -1.356eV and -1.175eV, respectively. In addition, the conductivity of the system is greatly improved, with the bandgap changing from 1.129eV to 0eV. The microscopic mechanism of the interaction of gas molecules with CoO-SnSe monolayers has been revealed by energy band structure (Eg), density of states (DOS), Milligan charge analysis (Q) and electrostatic potential (ESP). Finally, the sensitivity (S) and desorption time () of the five adsorption systems were calculated to illustrate the macroscopic gas-sensitive properties of the system. This work will help to explore the application of CoO-SnSe monolayers in SF6 decomposition gas sensing detection and adsorption treatment. © 2023, The Authors. All rights reserved.

关键词： Density functional theory

Crn (N=1-4) Clustered (8,0）Single-Walled Cnt: Comparison of Gas-Sensitive Properties to Air Discharge Pollutants (Co, Nox)

学校读者我要写书评

暂无评论

SSRN 2023年

作者： Wang, Mingxiang Cao, Jianjun Zhang, Yiyi Liu, Jiefeng Chen, Dachang Jia, Pengfei School of Electrical Engineering Guangxi University Nanning China School of Electrical and Electronic Engineering Wuhan Polytechnic University Wuhan China Guangxi Key Laboratory of Intelligent Control and Maintenance of Power Equipment Guangxi University Nanning China Guangxi Key Laboratory of Power System Optimization and Energy Technology China

Air discharge pollutants detection is considered to be one of the most reliable methods in fault diagnosis of air discharge . Based on the first-principles, the adsorption and sensing characteristics of Single-walled CNT modified by transition metal clusters (Crn) for air discharge pollutants (CO, NOx) were analyzed. Firstly, through the modeling and analysis of different doping sites, four most stable models of Crn doped carbon nanotubes are obtained, and their resistance is briefly analyzed. On the basis of these four configurations, a variety of adsorption structures were constructed and optimized. The parameters such as transfer charge, adsorption energy and adsorption distance were obtained. Furthermore, the differential charge density, frontier molecular orbitals, selectivity and desorption properties of the system were analyzed. The results show that Crn-SWCNT surface has good adsorption performance for pollutants. At room temperature, the pollutants can’t be desorbed from the surface of Crn-SWCNT. At extremely high temperatures, CO can be rapidly desorbed from the surface of Crn-SWCNT. Therefore, Crn-SWCNT can be used as solid adsorption materials for purifying CO, NOx at room temperature. This work provides a new idea for clean air to treat air discharge pollutants. © 2023, The Authors. All rights reserved.

关键词： Gas adsorption

Research on Path Analysis Method and Damping Characteristics of Subsynchronous Oscillation in Direct-Drive Wind Farm with VSC-HYDC system

学校读者我要写书评

暂无评论

SSRN 2022年

作者： Gao, Benfeng Liu, Yangyang Shao, Bingbing Ran, Huijuan State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources North China Electric Power University Baoding071003 China Anhui Province Key Laboratory of Renewable Energy Utilization and Energy Saving Hefei University of Technology Hefei230009 China

Subsynchronous oscillation (SSO) exists in direct-drive wind farm with VSC-HVDC (DDWFV) system. In the process of mechanism analysis, the existing analysis methods are difficult to reveal the disturbance transfer process and the coupling relationship between direct-drive wind farm and VSC-HVDC. In this paper, a path analysis method is proposed to study the mechanism of SSO. Firstly, the linearization model of the DDWFV is established, and the transfer path of subsynchronous disturbances between direct-drive wind farm and VSC-HVDC is revealed based on the transfer function block diagram. Then, the damping characteristics of the SSO in DDWFV are studied, in which the damping of interaction can be separated by damping reconstruction. Finally, the influence of the controller parameters on the damping characteristics is analyzed, and the correctness of theoretical analysis is verified by time-domain simulation. The results show that the disturbance transfer path between direct-drive wind farm and VSC-HVDC presents a four-loop coupling relationship, which can cause the occurrence of SSO. By optimizing the controller parameters, the SSO mode damping can be increased, thereby improving the stability of the system. © 2022, The Authors. All rights reserved.

关键词： Regression analysis

Unlocking new possibilities in ionic thermoelectric materials: a machine learning perspective

学校读者我要写书评

暂无评论

National Science Review 2025年第1期 195-206页

作者： Yidan Wu Dongxing Song Meng An Cheng Chi Chunyu Zhao Bing Yao Weigang Ma Xing Zhang Key Laboratory for Thermal Science and Power Engineering of Ministry of Education Department of Engineering Mechanics Tsinghua University Key Laboratory of Process Heat Transfer and Energy Saving of Henan Province School of Mechanics and Safety Engineering Zhengzhou University College of Mechanical and Electrical Engineering Shaanxi University of Science and Technology Key Laboratory of Power Station Energy Transfer Conversion and System of Ministry of Education School of Energy Power and Mechanical Engineering North China Electric Power University School of Materials and Chemical Engineering Xuzhou University of Technology

The high thermopower of ionic thermoelectric(i-TE) materials holds promise for miniaturized waste-heat recovery devices and thermal sensors. However, progress is hampered by laborious trial-and-error experimentations, which lack theoretical underpinning. Herein, by introducing the simplified molecular-input line-entry system, we have addressed the challenge posed by the inconsistency of i-TE material types, and present a machine learning model that evaluates the Seebeck coefficient with an R2of0.98 on the test dataset. Using this tool, we experimentally identify a waterborne polyurethane/potassium iodide ionogel with a Seebeck coefficient of 41.39 m V/K. Furthermore, interpretable analysis reveals that the number of rotatable bonds and the octanol-water partition coefficient of ions negatively affect Seebeck coefficients, which is corroborated by molecular dynamics simulations. This machine learning-assisted framework represents a pioneering effort in the i-TE field, offering significant promise for accelerating the discovery and development of high-performance i-TE materials.

关键词：

Robust Optimal power Flow Model of Electricity-Gas Integrated energy system Considering the Uncertainty of Roof Photovoltaic: ∗note: Sub-titles are not captured in Xplore and should not be used

学校读者我要写书评

暂无评论

Robust Optimal Power Flow Model of Electricity-Gas Integrate...

2020 IEEE Sustainable power and energy Conference, iSPEC 2020

作者： Qin, Zhengfeng Bai, Xiaoqing School of Electrical Engineering Guangxi University Guangxi Key Laboratory of Power System Optimization and Energy Technology Nanning China

ISBN: (纸本)9781728191645

With more and more distributed roof photovoltaic generators connected to the grid and their complementarity with gas turbines, which providing many possibilities for the coordinated operation of electricity-gas integrated energy system (IES). In this paper, aiming at minimize the total generation costs, a robust optimal power flow (ROPF) model was proposed for the electricity-gas IES connecting roof photovoltaic (PV) generators in the distribution network. Then, the proposed ROPF model is transformed into the mixed-integer second-order conic programming (MISOCP) model, which can be solved by Gurobi. The numerical simulations are carried on IG-30 system and IG-118 system to verify the applicability, safety and economy of the proposed ROPF model. © 2020 IEEE.

关键词： Electric load flow

Environmental,social,and economic assessment of energy utilization of crop residue in China

学校读者我要写书评

暂无评论

Frontiers in energy 2021年第2期15卷 308-319页

作者： Yueling ZHANG Junjie LI Huan LIU Guangling ZHAO Yajun TIAN Kechang XIE Key Laboratory of Resources and Environmental System Optimization(Ministry of Education) College of Environmental Sciense and EngineeringNorth China Electic Power UniversityBeijing 102206China School Chemical and Environmental Engineering China University of Mining and TechnologyBeijng 100083China Department of Energy Conversion and Storage Technical University of DenmarkLyngby 2820Denmark National Institute of Clean and Low Carbon Energy Beijing 102211China Chinese Acadamy of Engineering Beijing 100088China

This paper aims to discuss an environmental, social, and economic analysis of energy utilization of crop residues from life cycle perspectives in China. The methodologies employed to achieve this objective are environmental life cycle assessment (E-LCA), life cycle cost (LCC), and social life cycle assessment (S-LCA). Five scenarios are developed based on the conversion technologies and final bioenergy products. The system boundaries include crop residue collection, transportation, pre-treatment, and conversion process. The replaced amounts of energy are also taken into account in the E-LCA analysis. The functional unit is defined as 1 MJ of energy produced. Eight impact categories are considered besides climate change in E-LCA. The investment capital cost and salary cost are collected and compared in the life cycle of the scenarios. Three stakeholders and several subcategories are considered in the S-LCA analysis defined by UNEP/SETAS guidelines. The results show that the energy utilization of crop residue has carbon emission factors of 0.09–0.18 kg (CO2 eq per 1 MJ), and presents a net carbon emissions reduction of 0.03–0.15 kg (CO2 eq per 1 MJ) compared with the convectional electricity or petrol, but the other impacts should be paid attention to in the biomass energy scenarios. The energy utilization of crop residues can bring economic benefit to local communities and the society, but the working conditions of local workers need to be improved in future biomass energy development.

关键词： crop residue life cycle assessment life cycle cost social life cycle assessment energy production