A Support Data-Based Core-Set Selection Method for Signal Recognition

In recent years,deep learning-based signal recognition technology has gained attention and emerged as an important approach for safeguarding the electromagnetic environment.However,training deep learning-based classifiers on large signal datasets with redundant samples requires significant memory and high costs.This paper proposes a support databased core-set selection method(SD)for signal recognition,aiming to screen a representative subset that approximates the large signal dataset.Specifically,this subset can be identified by employing the labeled information during the early stages of model training,as some training samples are labeled as supporting data frequently.This support data is crucial for model training and can be found using a border sample selector.Simulation results demonstrate that the SD method minimizes the impact on model recognition performance while reducing the dataset size,and outperforms five other state-of-the-art core-set selection methods when the fraction of training sample kept is less than or equal to 0.3 on the RML2016.04C dataset or 0.5 on the RML22 dataset.The SD method is particularly helpful for signal recognition tasks with limited memory and computing resources.

Analysis of the interaction between bolt-reinforced rock and surface support in tunnels based on convergence-confinement method

To investigate the interaction of the bolt-reinforced rock and the surface support,an analytical model of the convergence-confinement type is proposed,considering the sequential installation of the fully grouted rockbolts and the surface support.The rock mass is assumed to be elastic-brittle-plastic material,obeying the linear Mohr-Coulomb criterion or the non-linear Hoek-Brown criterion.According to the strain states of the tunnel wall at bolt and surface support installation and the relative magnitude between the bolt length and the plastic depth during the whole process,six cases are categorized upon solving the problem.Each case is divided into three stages due to the different effects of the active rockbolts and the passive surface support.The fictitious pressure is introduced to quantify the threedimensional(3D)effect of the tunnel face,and thus,the actual physical location along the tunnel axis of the analytical section can be considered.By using the bolt-rock strain compatibility and the rocksurface support displacement compatibility conditions,the solutions of longitudinal tunnel displacement and the reaction pressure of surface support along the tunnel axis are obtained.The proposed analytical solutions are validated by a series of 3D numerical simulations.Extensive parametric studies are conducted to examine the effect of the typical parameters of rockbolts and surface support on the tunnel displacement and the reaction pressure of the surface support under different rock conditions.The results show that the rockbolts are more effective in controlling the tunnel displacement than the surface support,which should be installed as soon as possible with a suitable length.For tunnels excavated in weak rocks or with restricted displacement control requirements,the surface support should also be installed or closed timely with a certain stiffness.The proposed method provides a convenient alternative approach for the optimization of rockbolts and surface support at the preliminary stage of tunnel design.

Comparative study of different machine learning models in landslide susceptibility assessment: A case study of Conghua District, Guangzhou, China

Machine learning is currently one of the research hotspots in the field of landslide prediction.To clarify and evaluate the differences in characteristics and prediction effects of different machine learning models,Conghua District,which is the most prone to landslide disasters in Guangzhou,was selected for landslide susceptibility evaluation.The evaluation factors were selected by using correlation analysis and variance expansion factor method.Applying four machine learning methods namely Logistic Regression(LR),Random Forest(RF),Support Vector Machines(SVM),and Extreme Gradient Boosting(XGB),landslide models were constructed.Comparative analysis and evaluation of the model were conducted through statistical indices and receiver operating characteristic(ROC)curves.The results showed that LR,RF,SVM,and XGB models have good predictive performance for landslide susceptibility,with the area under curve(AUC)values of 0.752,0.965,0.996,and 0.998,respectively.XGB model had the highest predictive ability,followed by RF model,SVM model,and LR model.The frequency ratio(FR)accuracy of LR,RF,SVM,and XGB models was 0.775,0.842,0.759,and 0.822,respectively.RF and XGB models were superior to LR and SVM models,indicating that the integrated algorithm has better predictive ability than a single classification algorithm in regional landslide classification problems.

Small-time scale network traffic prediction based on a local support vector machine regression model

In this paper we apply the nonlinear time series analysis method to small-time scale traffic measurement data. The prediction-based method is used to determine the embedding dimension of the traffic data. Based on the reconstructed phase space, the local support vector machine prediction method is used to predict the traffic measurement data, and the BIC-based neighbouring point selection method is used to choose the number of the nearest neighbouring points for the local support vector machine regression model. The experimental results show that the local support vector machine prediction method whose neighbouring points are optimized can effectively predict the small-time scale traffic measurement data and can reproduce the statistical features of real traffic measurements.

SOFT SENSING MODEL BASED ON SUPPORT VECTOR MACHINE AND ITS APPLICATION

Soft sensor is widely used in industrial process control. It plays animportant role to improve the quality of product and assure safety in production. The core of softsensor is to construct soft sensing model. A new soft sensing modeling method based on supportvector machine (SVM) is proposed. SVM is a new machine learning method based on statistical learningtheory and is powerful for the problem characterized by small sample, nonlinearity, high dimensionand local minima. The proposed methods are applied to the estimation of frozen point of light dieseloil in distillation column. The estimated outputs of soft sensing model based on SVM match the realvalues of frozen point and follow varying trend of frozen point very well. Experiment results showthat SVM provides a new effective method for soft sensing modeling and has promising application inindustrial process applications.

A Knowledge Model-and Growth Model-Based Decision Support System for Wheat Management

By applying the system analysis principle and mathematical modeling technique to knowledge expression system for crop cultural management, the fundamental relationships and quantitative algorithms of wheat growth and management indices to variety types, ecological environments and production levels were analysed and extracted, and a dynamic knowledge model with temporal and spatial characters for wheat management(WheatKnow)was developed. By adopting the soft component characteristics as non language relevance , re-utilization and portable system maintenance. and by further integrating the wheat growth simulation model(WheatGrow)and intelligent system for wheat management, a comprehensive and digital knowledge model, growth model and component-based decision support system for wheat management(MBDSSWM)was established on the platforms of Visual C++ and Visual Basic. The MBDSSWM realized the effective integration and coupling of the prediction and decision-making functions for digital crop management.

Dynamics Modeling and Numerical Analysis of Rotor with Elastic Support/Dry Friction Dampers

The elastic support/dry friction damper is a type of damper which is used for active vibration control in a rotor system.To establish the analytical model of this type of damper,a two-dimensional friction model-ball/plate model was proposed.By using this ball/plate model,a dynamics model of rotor with elastic support/dry friction dampers was established and experimentally verified.Moreover,the damping performance of the elastic support/dry friction damper was studied numerically with respect to some variable parameters.The numerical study shows that the damping performance of the elastic support/dry friction damper is closely related to the stiffness distribution of the rotor-support system,the damper location,the pressing force between the moving and stationary disk,the friction coefficient,the tangential contact stiffness of the contact interface,and the stiffness of the stationary disk.In general,the damper should be located on an elastic support which has a large vibration amplitude in order to achieve a better damping performance,and the more vibration energy in this elastic support concentrates,the better performance of the damper will be.The larger the tangential contact stiffness of the contact interface,and the stiffness of the stationary disk are,the better performance of the damper will be.There will be an optimal value of the friction force at which the damper performs best.

FAULT DIAGNOSIS APPROACH BASED ON HIDDEN MARKOV MODEL AND SUPPORT VECTOR MACHINE

Aiming at solving the problems of machine-learning in fault diagnosis, a diagnosis approach is proposed based on hidden Markov model （HMM） and support vector machine （SVM）. HMM usually describes intra-class measure well and is good at dealing with continuous dynamic signals. SVM expresses inter-class difference effectively and has perfect classify ability. This approach is built on the merit of HMM and SVM. Then, the experiment is made in the transmission system of a helicopter. With the features extracted from vibration signals in gearbox, this HMM-SVM based diagnostic approach is trained and used to monitor and diagnose the gearbox＇s faults. The result shows that this method is better than HMM-based and SVM-based diagnosing methods in higher diagnostic accuracy with small training samples.

A novel excitation controller using support vector machines and approximate models

This paper proposes a novel excitation controller using support vector machines （SVM） and approximate models. The nonlinear control law is derived directly based on an input-output approximation method via Taylor expansion, which not only avoids complex control development and intensive computation, but also avoids online learning or adjustment. Only a general SVM modelling technique is involved in both model identification and controller implementation. The robustness of the stability is rigorously established using the Lyapunov method. Several simulations demonstrate the effectiveness of the proposed excitation controller.

Model based decision support system for land use changes and socio-economic assessments

Hydrological models are often linked with other models in cognate sciences to understand the interactions among climate, earth, water, ecosystem, and human society. This paper presents the development and implementation of a decision support system（DSS） that links the outputs of hydrological models with real-time decision making on social-economic assessments and land use management. Discharge and glacier geometry changes were simulated with hydrological model, water availability in semiarid environments. Irrigation and ecological water were simulated by a new commercial software MIKE HYDRO. Groundwater was simulated by MODFLOW. All the outputs of theses hydrological models were taken as inputs into the DSS in three types of links： regression equations, stationary data inputs, or dynamic data inputs as the models running parallel in the simulation periods. The DSS integrates the hydrological data, geographic data, social and economic statistical data, and establishes the relationships with equations, conditional statements and fuzzy logics. The programming is realized in C＋＋. The DSS has four remarkable features：（1） editable land use maps to assist decision-making;（2） conjunctive use of surface and groundwater resources;（3） interactions among water, earth, ecosystem, and humans; and（4） links with hydrological models. The overall goal of the DSS is to combine the outputs of scientific models, knowledge of experts, and perspectives of stakeholders, into a computer-based system, which allows sustainability impact assessment within regional planning; and to understand ecosystem services and integrate them into land and water management.

Support Vector Machine-Based Nonlinear System Modeling and Control

This paper provides an introduction to a support vector machine, a new kernel-based technique introduced in statistical learning theory and structural risk minimization, then presents a modeling-control framework based on SVM. At last a numerical experiment is taken to demonstrate the proposed approach's correctness and effectiveness.

Aeroelastic modeling of wind loading on a cable-net supported glass wall

Wind loading study on a cable-net supported glass wall is conducted by means of wind tunnel tests. An equiva- lent aeroelastic model is designed and constructed. Response of displacements of the wall is measured and analyzed. In order to design a glass wall under wind loading, the ＂wind- vibration factor＂ is estimated and discussed. In fact, the mech- anism of wind acting on the wall is commonly known not only as positive pressure, but also as negative pressure caused by the flow separation on the corners of the building. Due to the diffidence in the mechanism of wind acting, two typi- cal response cases are classified. The results show that the dynamic response of the structure caused by the negative pressure is stronger than that of the positive pressure case. To determine the aerodynamic wind loading on a flexible part of structure on a building, wind tunnel study may be useful and play an important role.

TYRE DYNAMICS MODELLING OF VEHICLE BASED ON SUPPORT VECTOR MACHINES

Various methods of tyre modelling are implemented from pure theoretical to empirical or semi-empirical models based on experimental results. A new way of representing tyre data obtained from measurements is presented via support vector machines （SVMs）. The feasibility of applying SVMs to steady-state tyre modelling is investigated by comparison with three-layer backpropagation （BP） neural network at pure slip and combined slip. The results indicate SVMs outperform the BP neural network in modelling the tyre characteristics with better generalization performance. The SVMsqyre is implemented in 8-DOF vehicle model for vehicle dynamics simulation by means of the PAC 2002 Magic Formula as reference. The SVMs-tyre can be a competitive and accurate method to model a tyre for vehicle dynamics simuLation.

A Multi-case Study of Community Supported Agriculture (CSA) Operation Model in Beijing

As one of the ways to localize foods,community supported agriculture can strengthen food safety and should be reasonably guided and developed.Based on the analysis of the spatial layout of the community supported agriculture organizations in Beijing,this paper described the development framework and course of community supported agriculture in Beijing.Combining 3 differentiated cases with the consumer's participation in community supported agriculture and the trust mechanism,it was found that the drive type has a very significant impact on the characteristics of these organizations.

High-rise building fire pre-warning model based on the support vector regression

Aiming at reducing the deficiency of the traditional fire pre-warning algorithms and the intelligent fire pre-warning algorithms such as artificial neural network,and then to improve the accuracy of fire prewarning for high-rise buildings,a composite fire pre-warning controller is designed according to the characteristic（ nonlinear,less historical data,many influence factors）,also a high-rise building fire pre-warning model is set up based on the support vector regression（ SV R）. Then the wood fire standard history data is applied to make empirical analysis. The research results can provide a reliable decision support framework for high-rise building fire pre-warning.

Support vector machine-based multi-model predictive control

In this paper, a support vector machine-based multi-model predictive control is proposed, in which SVM classification combines well with SVM regression. At first, each working environment is modeled by SVM regression and the support vector machine network-based model predictive control （SVMN-MPC） algorithm corresponding to each environment is developed, and then a multi-class SVM model is established to recognize multiple operating conditions. As for control, the current environment is identified by the multi-class SVM model and then the corresponding SVMN-MPC controller is activated at each sampling instant. The proposed modeling, switching and controller design is demonstrated in simulation results.

A Multiple Model Approach to Modeling Based on Fuzzy Support Vector Machines

A new multiple models(MM) approach was proposed to model complex industrial process by using Fuzzy Support Vector Machines(F -SVMs). By applying the proposed approach to a pH neutralization titration experiment, F -SVMs MM not only provides satisfactory approximation and generalization property, but also achieves superior performance to USOCPN multiple modeling method and single modeling method based on standard SVMs.

State of the art in applications of machine learning in steelmaking process modeling

With the development of automation and informatization in the steelmaking industry,the human brain gradually fails to cope with an increasing amount of data generated during the steelmaking process.Machine learning technology provides a new method other than production experience and metallurgical principles in dealing with large amounts of data.The application of machine learning in the steelmaking process has become a research hotspot in recent years.This paper provides an overview of the applications of machine learning in the steelmaking process modeling involving hot metal pretreatment,primary steelmaking,secondary refining,and some other aspects.The three most frequently used machine learning algorithms in steelmaking process modeling are the artificial neural network,support vector machine,and case-based reasoning,demonstrating proportions of 56%,14%,and 10%,respectively.Collected data in the steelmaking plants are frequently faulty.Thus,data processing,especially data cleaning,is crucially important to the performance of machine learning models.The detection of variable importance can be used to optimize the process parameters and guide production.Machine learning is used in hot metal pretreatment modeling mainly for endpoint S content prediction.The predictions of the endpoints of element compositions and the process parameters are widely investigated in primary steelmaking.Machine learning is used in secondary refining modeling mainly for ladle furnaces,Ruhrstahl–Heraeus,vacuum degassing,argon oxygen decarburization,and vacuum oxygen decarburization processes.Further development of machine learning in the steelmaking process modeling can be realized through additional efforts in the construction of the data platform,the industrial transformation of the research achievements to the practical steelmaking process,and the improvement of the universality of the machine learning models.

System dynamics model of the support-surrounding rock system in fully mechanized mining with large mining height face and its application

Fully mechanized mining with large mining height(FMMLMH)is widely used in thick coal seam mining face for its higher recovery ratio,especially where the thickness is less than 7.0 m.However,because of the great mining height and intense rock pressure,the coal wall rib spalling,roof falling and the instability of support occur more likely in FMMLMH working face,and the above three types of disasters interact with each other with complicated relationships.In order to get the relationship between each two of coal wall,roof,floor and support,and reduce the occurrence probability of the three types of disasters,we established the system dynamics(SD)model of the support-surrounding rock system which is composed of"coal wall-roof-floor-support"(CW-R-F-S)in a FMMLMH working face based on the condition of No.15104 working face in Sijiazhuang coal mine.With the software of Vensim,we also simulated the interaction process between each two factors of roof,floor,coal wall and the support.The results show that the SD model of"CW-R-F-S"system can reveal the complicated and interactive relationship clearly between the support and surrounding rock in the FMMLMH working face.By increasing the advancing speed of working face,the support resistance or the length of support guard,or by decreasing the tipto-face distance,the stability of"CW-R-F-S"system will be higher and the happening probability of the disasters such as coal wall rib spalling,roof falling or the instability of support will be lower.These research findings have been testified in field application in No.15104 working face,which can provide a new approach for researching the interaction relationship of support and surrounding rock.

Improved Ant Colony Algorithm for Vehicle Scheduling Problem in Airport Ground Service Support

Support vehicles are part of the main body of airport ground operations,and their scheduling efficiency directly impacts flight delays.A mathematical model is constructed and the responsiveness of support vehicles for current operational demands is proposed to study optimization algorithms for vehicle scheduling.The model is based on the constraint relationship of the initial operation time,time window,and gate position distribution,which gives an improvement to the ant colony algorithm(ACO).The impacts of the improved ACO as used for support vehicle optimization are compared and analyzed.The results show that the scheduling scheme of refueling trucks based on the improved ACO can reduce flight delays caused by refueling operations by 56.87%,indicating the improved ACO can improve support vehicle scheduling.Besides,the improved ACO can jump out of local optima,which can balance the working time of refueling trucks.This research optimizes the scheduling scheme of support vehicles under the existing conditions of airports,which has practical significance to fully utilize ground service resources,improve the efficiency of airport ground operations,and effectively reduce flight delays caused by ground service support.