检索结果-内蒙古大学图书馆

Identifying optimal ground feature classification and assessing leaf nitrogen status based on UAV multispectral images in an apple orchard

引用

PLANT AND SOIL 2024年 1-20页

作者： Sun, Guangzhao Chen, Shuaihong Hu, Tiantian Zhang, Shaowu Li, Hui Li, Aoqi Zhao, Lu Liu, Jie Northwest A&F Univ Key Lab Agr Soil & Water Engn Arid Semiarid Areas Minist Educ Xianyang Shaanxi Peoples R China Northwest A&F Univ Coll Water Resources & Architectural Engn Xianyang Shaanxi Peoples R China

Aims In the case of uneven vegetation coverage, it is still facing many problems that using UAV-remote sensing to assess the canopy nutrient status. The research objective is to determine the optimal ground feature classification and establish the leaf nitrogen concentration (LNC) inversion model. Methods UAV-Phantom 4 multispectral platform was used to acquire apple orchard images. The ground features of remote sensing image were classified with the minimum distance, maximum likelihood, and object-oriented feature extraction classifications (MDC, MLC, OFEC), while spectral vegetation indices were used to perform LNC inversion using the backpropagation neural network (BP) and extreme learning machine (ELM). Further, genetic algorithm (GA) and particle swarm optimization (PSO) were used to optimize two inversion models. Results Compared with the MDC, the overall accuracy of MLC increased 20.30% and 26.69% in 2021 and 2022, while the OFEC increased 30.48% and 31.04%, respectively. The LNC inversion model of BP and ELM produced an acceptable performance (R-c (2)> R(2)p > 0.60, RRMSE < 15%). The use of GA and PSO algorithms did not improve the prediction performance of the BP inversion model. Compared with ELM, the RRMSE of GA_ELM and PSO_ELM inversion models decreased 2.63% and 20.39%. Furthermore, the PSO_ELM inversion model decreased the RRMSE 10.95% compared to the BP model. ConclusionThe combination of reverse thinking iterative approach with the PSO algorithm significantly enhances the predictive performance of the ELM inversion model, which can present a rapid assessment method for leaf nitrogen nutrition diagnosis based on the relationship between LNC and fruit yield.

关键词： UAV-remote sensing Vegetation index optimization algorithm Foliar nitrogen Apple trees

来源：评论

学校读者我要写书评

暂无评论

The safety evaluation method of steel enterprises based on deep learning

引用

JOURNAL OF INTELLIGENT & FUZZY SYSTEMS 2023年第1期44卷 1337-1348页

作者： Li, Jingmin Xu, Shuzhen Beijing Beiye Funct Mat Corp Beijing Peoples R China Yantai Univ Sch Comp & Control Engn Yantai 264005 Shandong Peoples R China

As an important basic industry of national economy, the iron and steel industry has provided an important raw material guarantee for a long time. However there are a large number of hazard sources which are prone to safety accidents in the production process. Then safety evaluation in the production system is highly needed to effectively prevent the occurrence of accidents in iron and steel enterprises. Hence we introduce a method based on deep learning model to evaluate safety of the enterprises. Firstly, the risk factors and casualties in production process are investigated, and a set of safety evaluation index system is ***, since deep neural network model has the characteristics of strong feature extraction ability and simple model structure, we design a safety evaluation model based on deep neural network. The 25-dimensional evaluation index value is the input of the network, and the network output corresponds to five risk levels. On this basis, the optimization algorithm of deep neural network model is designed to explore the mapping relationship between risk characteristics and safety level. Tensorflow deep learning framework is used to build the evaluation model, classification loss function and network optimization method are designed to train the model. Finally, through experiments, the optimal model structure is determined by comparing the influence of different parameter optimization strategies, different hidden layer structures, and different activation functions on the safety evaluation performance. A three hidden layer structure with the Adam back propagation algorithm and LeakyRelu activation function is adopted to obtain higher accuracy and faster convergence rate. The experiments show that our evaluation model provides an operational method for evaluating the safety management status of iron and steel enterprises.

关键词： Iron and steel enterprises safety assessment neural network optimization algorithm deep learning

来源：评论

学校读者我要写书评

暂无评论

Optimized supervised learning approach to predict Parkinson's disease with minimal attributes using PPMI Datasets

引用

MULTIMEDIA TOOLS AND APPLICATIONS 2023年第16期83卷 48499页

作者： Kanagaraj, S. Hema, M. S. Guptha, M. Nageswara Kumaraguru Coll Technol Dept Informat Technol Coimbatore India Sri Venkateshwara Coll Engn Dept Comp Sci & Engn Vidya Nagar Bengaluru 562157 Karnataka India Sri Venkateshwara Coll Engn Vidya Nagar Bengaluru 562157 Karnataka India

Researchers can examine numerous ailments and forecast improved treatments using a huge number of medical databases. The Michael Fox PPMI data sets provide a baseline evaluation of the disease using the Unified Parkinson's Disease Rating Scales, the most prevalent sequential scales for identifying Parkinson's conditions. Existing research uses a Gaussian mixture model to predict PD disease using min-max normalization and dimensionality reduction based on principal component analysis. It significantly improved the findings, but it required more data to make predictions. This may be developed by decreasing the numbers of features utilized for prediction utilizing optimization algorithms. As a result, the ant-colony optimizations approach is suggested in this paper to enhance the classifier with few features. The ant colony approach uses this information to select the lowest features to use in training the regression neural network for disease prediction. When compared to various dimensionality reductions methods including Fast-ICA, PCAs, Kernel-PCA, as well as NMF, the findings suggest that the suggested optimization approaches performing-well. The neural regression network also reveals that the suggested strategy outperforms the Gaussian mixture model with the least patient information on the UPDRSs dataset.

关键词： Parkinson's disease (PD) Machine learning Normalization Dimensionality reduction optimization algorithm

来源：评论

学校读者我要写书评

暂无评论

Lung cancer detection and classification using deep neural network based on hybrid metaheuristic algorithm

引用

SOFT COMPUTING 2023年第15-16期28卷 8579页

作者： Prasad, Umesh Chakravarty, Soumitro Mahto, Gyaneshwar Birla Inst Technol Dept Comp Sci & Engn Ranchi Jharkhand India Birla Inst Technol Dept Management Ranchi Jharkhand India Swami Vivekanand Sch Ranchi Jharkhand India

Lung cancer causes millions of deaths annually, and CT scans and lung X-rays are common diagnostic tools. However, large datasets can contain noisy and irrelevant features that can affect the effectiveness of deep learning classification systems. Feature selection is a preprocessing step that can minimize database dimensionality and improve classification accuracy by selecting essential features. In addition, to solve the disadvantage of expensive solution set analysis in the wrapper technique, this study offers an efficient evaluation methodology that significantly decreases the evaluation cost and enhances the efficacy of the feature selection method. Hence, the study proposed a Hybrid Spotted Hyena optimization with Seagull algorithm to solve feature selection problems, which efficiently obtained the optimum subset with the largest number of relevant attributes. Biomedical lung data are collected from the LIDC/IDRI and chest X-ray datasets, and the bicubic-interpolation approach is employed to eliminate noise present in the data. We used generative modeling technique, DCGAN, to perform data augmentation. The chosen lung characteristics are evaluated using a hybrid CNN-LSTM, which identified normal and abnormal features in biomedical lung data. The system's efficacy is determined by examining its accuracy, precision, recall, specificity, and sensitivity using a Python experimental design. In the LIDC/IDRI database, the proposed classifier achieved a sensitivity of 99.8%, specificity of 99.3%, precision of 99.14%, and accuracy of 99.6%. In the chest X-ray dataset, the classifier achieved a sensitivity of 99.62%, specificity of 97.8%, precision of 97.5%, and accuracy of 99.7%.

关键词： CNN Feature selection Lung cancer LSTM optimization algorithm Preprocessing Segmentation SH-SOA

来源：评论

学校读者我要写书评

暂无评论

Optimal design of main girder structure of bridge crane based on equal life concept driven by data

引用

JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY 2023年第9期37卷 4767-4786页

作者： Yu, Yannan Liu, Zhiping Lu, Yao Zhang, Peng Liu, Hui Wuhan Univ Technol Coll Transportat & Logist Engn Wuhan 430063 Peoples R China

In traditional structural crane design, the fatigue strength of the structure is qualitatively evaluated, only depending on the steel type, working level and joint form, which leads to the excessive service life of some parts of the existing main girder structure, resulting in a waste of industrial resources and manufacturing costs. Based on the concept of equal life design, this paper obtained the service information of the crane in the scheduled inspection period through the rapid prediction method of "field collection+machine learning." Combined with the bearing capacity evaluation method and the fatigue life assessment method of "S-N curve+fracture mechanics," a nonlinear mixed integer equal life optimization model of crane girder structure was established, and the multi-specular reflection optimization algorithm was used to optimize the life distribution at different positions and overall weight of the main girder. The structural design parameters of the crane main girder which meet the bearing capacity and have the characteristics of "equal life+lightweight" were obtained. The performance of the optimized main girder structure was verified by finite element simulation. The results show that this method can realize the quantitative control of the life distribution of each position of the crane main girder structure and the lightweight design of the whole structure, and provides a scientific reference for the equal life and lightweight design of large construction machinery products.

关键词： Crane Equal life Fatigue life Structural optimization optimization algorithm Finite element

来源：评论

学校读者我要写书评

暂无评论

Decentralized Observation Allocation for a Large-Scale Constellation

引用

JOURNAL OF AEROSPACE INFORMATION SYSTEMS 2023年第8期20卷 447-461页

作者： Parjan, Shreya Chien, Steve A. CALTECH Jet Prop Lab Pasadena CA 91109 USA

Increased space sensing enables new measurements of a wide range of Earth science phenomena including volcanism, flooding, wildfires, and weather. Large-scale observation constellations of hundreds of assets have already been deployed (for example, Planet Labs's Dove satellites), and several constellations of tens of thousands of assets are planned. New challenges exist to rapidly assimilate available data and to optimize measurements by directing spacecraft assets to best observe complex Earth science phenomena. Centralized approaches to managing request allocation in these large constellations are constrained by 1) the need to assign/elect a central node to assign requests to spacecraft and 2) reliance on a single agent communicating with potentially thousands of dependent agents. On the other hand, entirely decentralized approaches to request allocation and observation are prone to oversatisfaction of some requests and undersatisfaction of others due to a lack of communication among agents. In large constellations, an intermediary method is necessary to solve the request allocation problem in a distributed manner. We present distributed artificial intelligence/multiagent methods that leverage existing work on distributed constraint optimization to allocate observations in a satellite constellation. We compare their performance to centralized and highly decentralized approaches using realistic orbits and observation request distributions. Our distributed algorithms can find approximate solutions to the large-scale constellation request allocation problem with low data volume for agent coordination and extend to continuous planning problems with varying request sets and availability of spacecraft agents.

关键词： Distributed Artificial Intelligence algorithms and Data Structures Computing and Informatics optimization algorithm Satellite Remote Sensing Earth Observation Satellite Multi-Agent System

来源：评论

学校读者我要写书评

暂无评论

Enhanced network lifespan in future wireless communication using machine learning based convolution neural networks

引用

OPTICAL AND QUANTUM ELECTRONICS 2024年第4期56卷 579-579页

作者： Sheela, S. V. Radhika, K. R. BMS Coll Engn Dept Informat Sci & Engn Bangalore Karnataka India

A collection of sensor nodes called a wireless sensor network is used to track and document the physical parameters of the surrounding area. The design of network clustering approaches has a big problem when it comes to extending the lifetime of wireless sensor networks (WSNs) and improving energy usage by making sure Both the processing speed and the batteries have a lengthy lifespan. This study presents a framework for machine learning-based channel property variation tracking and learning that is based on a Convolutional Neural Network (CNN)-Long Short-Term Memory (Convolutional-LSTM) network. Our hybrid technique improves sensor connectivity and lowers power consumption, Wireless sensor network longevity is increased. These algorithms are evaluated in a wireless sensor network: Harris Hawks Optimisation (HHO), Coyote Optimisation algorithm (COY), Support Vector Machine (SVM), and Genetic algorithm (GA). As for nodes analysis and energy consumption, the article concludes demonstrates the CNN-LSTM technique under consideration outperforms other algorithms. The learning authentication system's robustness and detection performance are thoroughly examined, and exhaustive simulations and testing reveal a notable improvement in the detection accuracy in time-varying scenarios.

关键词： Wireless sensor networks Energy efficiency optimization algorithm Convolutional neural network

来源：评论

学校读者我要写书评

暂无评论

Maximizing the lender's profit: profit-oriented loan default prediction based on a weighting model

引用

ANNALS OF OPERATIONS RESEARCH 2024年 1-34页

作者： Cui, Huiyu Zhang, Lifang Yang, Hufang Wang, Jianzhou Liu, Zhenkun Dongbei Univ Finance & Econ Sch Finance & Taxat Dalian 116025 Peoples R China Nanjing Univ Finance & Econ Sch Finance Dalian 116025 Peoples R China Nanjing Univ Posts & Telecommun Sch Econ 9Wenyuan St Nanjing 210000 Jiangsu Peoples R China Macau Univ Sci & Technol Inst Syst Engn Taipa St Macau 999078 Peoples R China Nanjing Univ Posts & Telecommun Sch Management Dalian 116025 Peoples R China

Loan default risk prediction is necessary in credit risk assessment, as it helps financing institutions and investors make decisions. However, existing prediction models focus more on using individual classifiers to obtain higher prediction accuracy, which is far from the core purpose of business (i.e., maximizing profit) and leaves opportunities to explore profit-oriented and interpretable weighting models. This study proposes a profit-oriented weighting model for loan default prediction. The model consists of three stages: constructing multiple profit-oriented sub-classifiers, determining profit-oriented weight coefficients, and providing interpretable analysis. Five lending datasets are examined based on accuracy and profit-based metrics. The empirical results demonstrate that the proposed weighting prediction system helps lenders achieve higher profits and provides concise and intuitive interpretability. Thus, it can help practitioners make better decisions and manage risk.

关键词： Loan default risk prediction Weighting model Profit-driven prediction Interpretable analysis optimization algorithm

来源：评论

学校读者我要写书评

暂无评论

Tracking performance of the coaxial counter-paddle flight system using fractional active disturbance rejection controller

引用

APPLIED MATHEMATICAL MODELLING 2023年第1期121卷 800-827页

作者： Gao, Yong Zhang, Hao Chen, Xiao Lu, Tingting Tan, Shizhe Yang, Hua Gulliver, T. aaron Ocean Univ China Fac Informat Sci & Engn Sch Elect Engn Qingdao 266100 Peoples R China Univ Victoria Dept Elect & Comp Engn Victoria BC V8W 2Y2 Canada

Stable aerial flight capability is crucial for amphibious vehicles to perform cross-domain motion, yet it is often influenced by unpredictable cross-domain environments and model uncertainties. A fractional-order active disturbance rejection controller based on finite-time convergent extended state observer is proposed for the anti-interference control problem in the attitude trajectory tracking of bullet-shaped trans-domain amphibious vehicle in complex environments. Firstly, a complete dynamic model of the airborne flight system of the bullet-shaped trans-domain amphibious vehicle is established. A hierarchical control scheme based on the proposed controller is designed assuming that all states of the system are available, in order to improve the control reliability and robustness of the flight system in multi-source interference environments. Then, based on a novel hybrid particle swarm search technique called the Particle Swarm optimization with Levy Flight and Wavelet Mutation algorithm, which combines Levy flight and wavelet mutation, a new parameter identification and tuning algorithm is proposed for the proposed controller to obtain optimal control parameters for attitude and position. To provide a comprehensive comparison, the stability of the optimization algorithm and the robustness performance of the controller were evaluated. Finally, all controllers were tested under three different disturbance scenarios. The results show that the proposed controller can achieve better stability in flight attitude and trajectory tracking tasks compared to Proportion Integration Differentiation controller, Fractional-order Proportion Integration Differentiation controller, and Active Disturbance Rejection controller in all scenarios. & COPY;2023 Elsevier Inc. All rights reserved.

关键词： Coaxial counter-paddle Trajectory tracking control Anti-disturbance control Bullet-shaped amphibious vehicle optimization algorithm

来源：评论

学校读者我要写书评

暂无评论

Layered Soil Remote Sensing With Multichannel Passive Microwave Observations Using a Physics-Embedded Artificial Intelligence Framework: A Theoretical Study

引用

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING 2023年 61卷 1页

作者： Bai, Xuyang Tan, Shurun Univ Illinois Urbana Champaign Inst Zhejiang Univ Haining 314400 Peoples R China Zhejiang Univ Coll Informat Sci & Elect Engn State Key Lab Modern Opt Instrumentat Hangzhou 310027 Peoples R China Univ Illinois Dept Elect & Comp Engn Urbana IL 61801 USA

The vertical distribution of soil properties is crucial in accurately representing various hydrological and ecological processes such as freeze-thaw cycles and diurnal variations. In this article, considering the complexity of the multiparameter features of layered soil, we evaluate the potential to retrieve the vertical distribution of the moisture and temperature of soil using multichannel passive microwave observations. To enhance the inversion efficiency and accuracy, a novel physics-embedded artificial neural network (P-ANN) inversion algorithm combining multiangle (30 degrees-50 degrees), multifrequency (L-, C-, and X-bands), and multipolarization (horizontal and vertical polarization) passive observations is proposed. In this approach, the multichannel physical brightness temperature simulations corresponding to the predicted soil state parameters are integrated into the loss function of a standard fully connected feedforward neural network (NN), enabling efficient convergence with limited sampling data in the training process. Testing results exhibit that the inversion performance of P-ANN is superior to that of conventional NN approaches, which only adopts errors in soil states in the loss function to train the network. The test also shows that the proposed P-ANN approach outperforms traditional optimization algorithms in dealing with layered soil retrieval. In order to further improve the retrieval accuracy, an advanced local optimization scheme is also proposed, where the output from P-ANN is further treated as the initial value to a local optimization algorithm, achieving even closer results to the ground truths without excessive computational costs. In addition, to estimate the reliability of the model predictions, this article also establishes, in the testing process, a statistical relationship between the soil inversion error and the error of the corresponding brightness temperatures. When the trained NN is in operation, the error of brightness temper

关键词： Soil Artificial neural networks Neural networks optimization Brightness temperature Reliability Training Layered soil remote sensing optimization algorithm physics-driven artificial neural network (ANN) reliability

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：