In this paper we study the degree of approximation by superpositions of a sigmoidal function.We mainly consider the univariate case.If f is a continuous function,we prove that for any bounded sigmoidal function σ,d_...In this paper we study the degree of approximation by superpositions of a sigmoidal function.We mainly consider the univariate case.If f is a continuous function,we prove that for any bounded sigmoidal function σ,d_(n,σ)(f)≤‖σ‖ω(f,1/(n+1)).For the Heaviside function H(x),we prove that d_(n,H)(f)≤ω(f,1/(2(n+1))). If f is a continuous funnction of bounded variation,we prove that d_(n,σ)(f)≤‖σ‖/(n+1)V(f)and d_(n,H)(f)≤ 1/(2(n+1))V(f).For he Heaviside function,the coefficient 1 and the approximation orders are the best possible.We compare these results with the classical Jackson and Bernstein theorems,and make some conjec- tures for further study.展开更多
In this paper, a constructive theory is developed for approximating func- tions of one or more variables by superposition of sigmoidal functions. This is done in the uniform norm as well as in the L^p norm. Results fo...In this paper, a constructive theory is developed for approximating func- tions of one or more variables by superposition of sigmoidal functions. This is done in the uniform norm as well as in the L^p norm. Results for the simultaneous approx- imation, with the same order of accuracy, of a function and its derivatives （whenever these exist）, are obtained. The relation with neural networks and radial basis func- tions approximations is discussed. Numerical examples are given for the purpose of illustration.展开更多
In this paper, we introduce a type of approximation operators of neural networks with sigmodal functions on compact intervals, and obtain the pointwise and uniform estimates of the ap- proximation. To improve the appr...In this paper, we introduce a type of approximation operators of neural networks with sigmodal functions on compact intervals, and obtain the pointwise and uniform estimates of the ap- proximation. To improve the approximation rate, we further introduce a type of combinations of neurM networks. Moreover, we show that the derivatives of functions can also be simultaneously approximated by the derivatives of the combinations. We also apply our method to construct approximation operators of neural networks with sigmodal functions on infinite intervals.展开更多
Using some regular matrices we present a method to express any multivariate algebraic polynomial of total order n in a normal form. Consequently, we prove constructively that, to approximate continuous target function...Using some regular matrices we present a method to express any multivariate algebraic polynomial of total order n in a normal form. Consequently, we prove constructively that, to approximate continuous target functions defined on some compact set of Rd, neural networks are at least as good as algebraic polynomials.展开更多
Reynolds-averaged Navier-Stokes(RANS)turbulence modeling can lead to the excessive turbulence level around the interface in two-phase flow,which causes the unphysical motion of the interface in sloshing simulation.In ...Reynolds-averaged Navier-Stokes(RANS)turbulence modeling can lead to the excessive turbulence level around the interface in two-phase flow,which causes the unphysical motion of the interface in sloshing simulation.In order to avoid the unphysical motion of the interface,a novel eddy-viscosity eliminator based on sigmoid functions is designed to reduce the excessive turbulence level,and the eddy-viscosity eliminator based on polynomials is extracted from the cavitation simulations.Surface elevations by combining the eddy-viscosity eliminators and classical two-equation closure models are compared with the experiments,the ones by using the adaptive asymptotic model(AAM)and the ones by using the modified two-equation closure models.The root-mean-squared error(RMSE)is introduced to quantify the accuracies of surface elevations and the forces.The relation between the turbulence level in the transition layer and RMSEs of surface elevations is studied.Besides,the parametric analysis of the eddy-viscosity eliminators is carried out.The studies suggest that(1)the excessive turbulence level in the transition layer around the interface has a significant influence on the accuracies of surface elevations and the forces;(2)the eddy-viscosity eliminators can effectively reduce the excessive turbulence level in the transition layer to avoid the unphysical motion of the interface;(3)the k-ωSST model combined with the eddy-viscosity eliminators is appropriate for predicting surface elevations and forces in RANS simulations of sloshing flow.展开更多
Geometrically nonlinear oscillations are investigated on sigmoid functionally graded material （S-FGM） plates with a longitudinal speed. The material properties of the plates obey a sigmoid distribution rule along th...Geometrically nonlinear oscillations are investigated on sigmoid functionally graded material （S-FGM） plates with a longitudinal speed. The material properties of the plates obey a sigmoid distribution rule along the thickness direction. Based on the D＇Alembert＇s principle, a nonlinear equation of motion is derived for the moving S-FGM plates, where the von K^rm^n nonlinear plate theory is adopted. Utilizing the Galerkin method, the equation of motion is discretized and solved via the method of harmonic bal- ance. The approximate analytical solutions are validated through the adaptive step-size fourth-order Runge-Kutta method. Besides, the stability of the steady-state solutions is examined. The results reveal that the mode interaction behavior can happen between the first two modes of the moving S-FGM plates, leading to a complex nonlinear frequency response. It is further found that the power-law index, the longitudinal speed, the exci- tation amplitude, and the in-plane pretension force can significantly affect the nonlinear frequency-response characteristics of longitudinally traveling S-FGM plates.展开更多
To get simpler operation in modified fuzzy adaptive learning control network （FALCON） in some engineering application, sigmoid nonlinear function is employed as a substitute of traditional Gaussian membership functi...To get simpler operation in modified fuzzy adaptive learning control network （FALCON） in some engineering application, sigmoid nonlinear function is employed as a substitute of traditional Gaussian membership function. For making the modified FALCON learning more efficient and stable, a simulated annealing （SA） learning coefficient is introduced into learning algorithm. At first, the basic concepts and main advantages of FALCON were briefly reviewed. Subsequently, the topological structure and nodes operation were illustrated; the gradient-descent learning algorithm with SA learning coefficient was derived; and the distinctions between the archetype and the modification were analyzed. Eventually, the significance and worthiness of the modified FALCON were validated by its application to probability prediction of anode effect in aluminium electrolysis cells.展开更多
Road extraction based on deep learning is one of hot spots of semantic segmentation in the past decade.In this work,we proposed a framework based on codec network for automatic road extraction from remote sensing imag...Road extraction based on deep learning is one of hot spots of semantic segmentation in the past decade.In this work,we proposed a framework based on codec network for automatic road extraction from remote sensing images.Firstly,a pre-trained ResNet34 was migrated to U-Net and its encoding structure was replaced to deepen the number of network layers,which reduces the error rate of road segmentation and the loss of details.Secondly,dilated convolution was used to connect the encoder and the decoder of network to expand the receptive field and retain more low-dimensional information of the image.Afterwards,the channel attention mechanism was used to select the information of the feature image obtained by up-sampling of the encoder,the weights of target features were optimized to enhance the features of target region and suppress the features of background and noise regions,and thus the feature extraction effect of the remote sensing image with complex background was optimized.Finally,an adaptive sigmoid loss function was proposed,which optimizes the imbalance between the road and the background,and makes the model reach the optimal solution.Experimental results show that compared with several semantic segmentation networks,the proposed method can greatly reduce the error rate of road segmentation and effectively improve the accuracy of road extraction from remote sensing images.展开更多
In this paper, the technique of approximate partition of unity is used to construct a class of neural networks operators with sigmoidal functions. Using the modulus of continuity of function as a metric, the errors of...In this paper, the technique of approximate partition of unity is used to construct a class of neural networks operators with sigmoidal functions. Using the modulus of continuity of function as a metric, the errors of the operators approximating continuous functions defined on a compact interval are estimated. Furthmore, Bochner-Riesz means operators of double Fourier series are used to construct networks operators for approximating bivariate functions, and the errors of approximation by the operators are estimated.展开更多
To realize the high precision and real-time interpolation of the NURBS （non-uniform rational B-spline） curve, a kinetic model based on the modified sigmoid function is proposed. The constraints of maximum feed rate,...To realize the high precision and real-time interpolation of the NURBS （non-uniform rational B-spline） curve, a kinetic model based on the modified sigmoid function is proposed. The constraints of maximum feed rate, chord error, curvature radius and interpolator cycle are discussed. This kinetic model reduces the cubic polynomial S-shape model and the trigonometry function S-shape model from 15 sections into 3 sections under the precondition of jerk, acceleration and feedrate continuity. Then an optimized Adams algorithm using the difference quotient to replace the derivative is presented to calculate the interpolator cycle parameters. The higher-order derivation in the Taylor expansion algorithm can be avoided by this algorithm. Finally, the simplified design is analyzed by reducing the times of computing the low-degree zero-value B-spline basis function and the simplified De Boor-Cox recursive algorithm is proposed. The simulation analysis indicates that by these algorithms, the feed rate is effectively controlled according to tool path. The calculated amount is decreased and the calculated speed is increased while the machining precision is ensured. The experimental results show that the target parameter can be correctly calculated and these algorithms can be applied to actual systems.展开更多
The above-ground net primary production(ANPP) and the precipitation-use efficiency(PUE) regulate the carbon and water cycles in grassland ecosystems, but the relationships among the ANPP, PUE and precipitation are sti...The above-ground net primary production(ANPP) and the precipitation-use efficiency(PUE) regulate the carbon and water cycles in grassland ecosystems, but the relationships among the ANPP, PUE and precipitation are still controversial. We selected 717 grassland sites with ANPP and mean annual precipitation(MAP) data from 40 publications to characterize the relationships ANPP–MAP and PUE–MAP across different grassland types. The MAP and ANPP showed large variations across all grassland types, ranging from 69 to 2335 mm and 4.3 to 1706 g m^(-2), respectively. The global maximum PUE ranged from 0.19 to 1.49 g m^(-2) mm^(-1) with a unimodal pattern. Analysis using the sigmoid function explained the ANPP–MAP relationship best at the global scale. The gradient of the ANPP–MAP graph was small for arid and semi-arid sites(MAP <400 mm). This study improves our understanding of the relationship between ANPP and MAP across dry grassland ecosystems. It provides new perspectives on the prediction and modeling of variations in the ANPP for different grassland types along precipitation gradients.展开更多
Recently, various control methods represented by proportional-integral-derivative （PID） control are used for robotic control. To cope with the requirements for high response and precision, advanced feedforward contr...Recently, various control methods represented by proportional-integral-derivative （PID） control are used for robotic control. To cope with the requirements for high response and precision, advanced feedforward controllers such as gravity compensator, Coriolis/centrifugal force compensator and friction compensators have been built in the controller. Generally, it causes heavy computational load when calculating the compensating value within a short sampling period. In this paper, integrated recurrent neural networks are applied as a feedforward controller for PUMA560 manipulator. The feedforward controller works instead of gravity and Coriolis/centrifugal force compensators. In the learning process of the neural network by using back propagation algorithm, the learning coefficient and gain of sigmoid function are tuned intuitively and empirically according to teaching signals. The tuning is complicated because it is being conducted by trial and error. Especially, when the scale of teaching signal is large, the problem becomes crucial. To cope with the problem which concerns the learning performance, a simple and adaptive learning technique for large scale teaching signals is proposed. The learning techniques and control effectiveness are evaluated through simulations using the dynamic model of PUMA560 manipulator.展开更多
In this paper we present a new version of Chen＇s system： a piecewise linear （PWL） Chert system of fractional-order. Via a sigmoid-like function, the discontinuous system is transformed into a continuous system. By...In this paper we present a new version of Chen＇s system： a piecewise linear （PWL） Chert system of fractional-order. Via a sigmoid-like function, the discontinuous system is transformed into a continuous system. By numerical simulations, we reveal chaotic behaviors and also multistability, i.e., the existence of small pararheter windows where, for some fixed bifurcation parameter and depending on initial conditions, coexistence of stable attractors and chaotic attractors is possible. Moreover, we show that by using an algorithm to switch the bifurcation parameter, the stable attractors can be numerically approximated.展开更多
A new vehicle steering control algorithm is presented. Unlike the traditional methods do, the algorithm uses a sigmoid function to describe the principle of the human driver＇s steering strategy. Based on this functio...A new vehicle steering control algorithm is presented. Unlike the traditional methods do, the algorithm uses a sigmoid function to describe the principle of the human driver＇s steering strategy. Based on this function, a human simulating vehicle steering model, human-simulating steering control（HS） algorithm is designed. In order to improve the adaptability to different environments, a parameter adaptive adjustment algorithm is presented. This algorithm can online modify the value of the key parameters of the HS real time. HS controller is used on a vehicle equipped with computer vision system and computer controlled steering actuator system, the result from the automatic vehicle steering experiment shows that the HS algorithm gives good performance at different speed, even at the maximum speed of 172 km/h.展开更多
We analyzed the relationship between internode number and intemode length for one of the largest bamboo, Phyllostachys pubescens Mazel ex Houz. For 50 sample culms with various sizes felled in a pure stand of P. pubes...We analyzed the relationship between internode number and intemode length for one of the largest bamboo, Phyllostachys pubescens Mazel ex Houz. For 50 sample culms with various sizes felled in a pure stand of P. pubescens, the intemode number was assigned from base to tip and the length for each internode was directly measured. The result indicated that the intemode length should be cumulated from base to tip, and then the cumulated internode length should be relativized by the total culm length. It was inappropriate to relativize the internode length by the maximum intenode length. In addition, the relationship between the relative internode number （the intemode number relativized by the total number of intemodes） and the relative cumulated internode length should be described not by a power function but by a sigmoid function such as the third-order function. The determined function enabled us to estimate the actual internode length, with the root mean squared error being 4 cm. In conclusion, the mathematical expression presented here, i.e., the relativization of the cumulated internode length by the total culm length and the application of the sigmoid function, will be useful in describing the relationship between internode number and internode length for P. pubescens.展开更多
An IDS(intrusion detection system)provides a foremost front line mechanism to guard networks,systems,data,and information.That’s why intrusion detection has grown as an active study area and provides significant cont...An IDS(intrusion detection system)provides a foremost front line mechanism to guard networks,systems,data,and information.That’s why intrusion detection has grown as an active study area and provides significant contribution to cyber-security techniques.Multiple techniques have been in use but major concern in their implementation is variation in their detection performance.The performance of IDS lies in the accurate detection of attacks,and this accuracy can be raised by improving the recognition rate and significant reduction in the false alarms rate.To overcome this problem many researchers have used different machine learning techniques.These techniques have limitations and do not efficiently perform on huge and complex data about systems and networks.This work focused on ELM(Extreme Learning Machine)technique due to its good capabilities in classification problems and dealing with huge data.The ELM has different activation functions,but the problem is to find out which function is more suitable and performs well in IDS.This work investigates this problem.Here,Well-known activation functions like:sine,sigmoid and radial basis are explored,investigated and applied to measure their performance on the GA(Genetic Algorithm)features subset and with full features set.The NSL-KDD dataset is used as a benchmark.The empirical results are analyzed,addressed and compared among different activation functions of the ELM.The results show that the radial basis and sine functions perform better on GA feature set than the full feature set while the performance of the sigmoid function is almost equal on both features sets.So,the proposal of GA based feature selection reduced 21 features out of 41 that brought up to 98%accuracy and enhanced overall efficiency of extreme learning machine in intrusion detection.展开更多
Incessant fire-outbreak in urban settlements has remained intractable especially in developing countries like Nigeria. This is often characterized by grave socio-economic aftermath effects. Urban fire outbreak in Nige...Incessant fire-outbreak in urban settlements has remained intractable especially in developing countries like Nigeria. This is often characterized by grave socio-economic aftermath effects. Urban fire outbreak in Nigerian cities has been on increase in recent times. The major problem faced by fire fighters in Nigerian urban centres is that there are no mechanisms to detect fire outbreaks early enough to save lives and properties. They often rely on calls made by neighbours or occupants when an outbreak occurs and this accounts for the delay in fighting fire outbreaks. This work uses Artificial Neural Networks (ANN) with backpropagation method to detect the occurrence of urban fires. The method uses smoke density, room temperature and cooking gas concentration as inputs. The work was implemented using Java programming language and results showed that it detected the occurrence of urban fires with reasonable accuracy. The work is recommended for use to minimize the effect of urban fire outbreak.展开更多
This paper investigates the problem of parameter identification for ship nonlinear Nomoto model with small test data,a nonlinear innovation-based identification algorithm is presented by embedding sigmoid function in ...This paper investigates the problem of parameter identification for ship nonlinear Nomoto model with small test data,a nonlinear innovation-based identification algorithm is presented by embedding sigmoid function in the stochastic gradient algorithm.To demonstrate the validity of the algorithm,an identification test is carried out on the ship‘SWAN’with only 26 sets of test data.Furthermore,the identification effects of the least squares algorithm,original stochastic gradient algorithm and the improved stochastic gradient algorithm based on nonlinear innovation are compared.Generally,the stochastic gradient algorithm is not suitable for the condition of small test data.The simulation results indicate that the improved stochastic gradient algorithm with sigmoid function greatly increases its accuracy of parameter identification and has 14.2%up compared with the least squares algorithm.Then the effectiveness of the algorithm is verified by another identification test on the ship‘Galaxy’,the accuracy of parameter identification can reach more than 95%which can be used in ship motion simulation and controller design.The proposed algorithm has advantages of the small test data,fast speed and high accuracy of identification,which can be extended to other parameter identification systems with less sample data.展开更多
In this paper, spatial patterns of a diffusive predator-prey model with sigmoid （Holling type III） ratio-dependent functional response which concerns the influence of logistic population growth in prey and intra-spe...In this paper, spatial patterns of a diffusive predator-prey model with sigmoid （Holling type III） ratio-dependent functional response which concerns the influence of logistic population growth in prey and intra-species competition among predators are investigated. The （local and global） asymptotic stability behavior of the corresponding non- spatial model around the unique positive interior equilibrium point in homogeneous steady state is obtained. In addition, we derive the conditions for Turing instability and the consequent parametric Turing space in spatial domain. The results of spatial pat- tern analysis through numerical simulations are depicted and analyzed. ~rthermore, we perform a series of numerical simulations and find that the proposed model dynamics exhibits complex pattern replication. The feasible results obtained in this paper indicate that the effect of diffusion in Turing instability plays an important role to understand better the pattern formation in ecosystem.展开更多
文摘In this paper we study the degree of approximation by superpositions of a sigmoidal function.We mainly consider the univariate case.If f is a continuous function,we prove that for any bounded sigmoidal function σ,d_(n,σ)(f)≤‖σ‖ω(f,1/(n+1)).For the Heaviside function H(x),we prove that d_(n,H)(f)≤ω(f,1/(2(n+1))). If f is a continuous funnction of bounded variation,we prove that d_(n,σ)(f)≤‖σ‖/(n+1)V(f)and d_(n,H)(f)≤ 1/(2(n+1))V(f).For he Heaviside function,the coefficient 1 and the approximation orders are the best possible.We compare these results with the classical Jackson and Bernstein theorems,and make some conjec- tures for further study.
基金supported, in part, by the GNAMPA and the GNFM of the Italian INdAM
文摘In this paper, a constructive theory is developed for approximating func- tions of one or more variables by superposition of sigmoidal functions. This is done in the uniform norm as well as in the L^p norm. Results for the simultaneous approx- imation, with the same order of accuracy, of a function and its derivatives （whenever these exist）, are obtained. The relation with neural networks and radial basis func- tions approximations is discussed. Numerical examples are given for the purpose of illustration.
基金Supported by National Natural Science Foundation of China(Grant No.10901044)Qianjiang Rencai Program of Zhejiang Province(Grant No.2010R10101)+1 种基金Scientific Research Foundation for the Returned Overseas Chinese Scholars,State Education MinistryProgram for Excellent Young Teachers in Hangzhou Normal University
文摘In this paper, we introduce a type of approximation operators of neural networks with sigmodal functions on compact intervals, and obtain the pointwise and uniform estimates of the ap- proximation. To improve the approximation rate, we further introduce a type of combinations of neurM networks. Moreover, we show that the derivatives of functions can also be simultaneously approximated by the derivatives of the combinations. We also apply our method to construct approximation operators of neural networks with sigmodal functions on infinite intervals.
文摘Using some regular matrices we present a method to express any multivariate algebraic polynomial of total order n in a normal form. Consequently, we prove constructively that, to approximate continuous target functions defined on some compact set of Rd, neural networks are at least as good as algebraic polynomials.
基金financially supported by the National Natural Science Foundation of China(Grant Nos.11802176,11802301)。
文摘Reynolds-averaged Navier-Stokes(RANS)turbulence modeling can lead to the excessive turbulence level around the interface in two-phase flow,which causes the unphysical motion of the interface in sloshing simulation.In order to avoid the unphysical motion of the interface,a novel eddy-viscosity eliminator based on sigmoid functions is designed to reduce the excessive turbulence level,and the eddy-viscosity eliminator based on polynomials is extracted from the cavitation simulations.Surface elevations by combining the eddy-viscosity eliminators and classical two-equation closure models are compared with the experiments,the ones by using the adaptive asymptotic model(AAM)and the ones by using the modified two-equation closure models.The root-mean-squared error(RMSE)is introduced to quantify the accuracies of surface elevations and the forces.The relation between the turbulence level in the transition layer and RMSEs of surface elevations is studied.Besides,the parametric analysis of the eddy-viscosity eliminators is carried out.The studies suggest that(1)the excessive turbulence level in the transition layer around the interface has a significant influence on the accuracies of surface elevations and the forces;(2)the eddy-viscosity eliminators can effectively reduce the excessive turbulence level in the transition layer to avoid the unphysical motion of the interface;(3)the k-ωSST model combined with the eddy-viscosity eliminators is appropriate for predicting surface elevations and forces in RANS simulations of sloshing flow.
基金supported by the National Natural Science Foundation of China(Nos.11672071,11302046,and 11672072)the Fundamental Research Funds for the Central Universities(No.N150504003)
文摘Geometrically nonlinear oscillations are investigated on sigmoid functionally graded material （S-FGM） plates with a longitudinal speed. The material properties of the plates obey a sigmoid distribution rule along the thickness direction. Based on the D＇Alembert＇s principle, a nonlinear equation of motion is derived for the moving S-FGM plates, where the von K^rm^n nonlinear plate theory is adopted. Utilizing the Galerkin method, the equation of motion is discretized and solved via the method of harmonic bal- ance. The approximate analytical solutions are validated through the adaptive step-size fourth-order Runge-Kutta method. Besides, the stability of the steady-state solutions is examined. The results reveal that the mode interaction behavior can happen between the first two modes of the moving S-FGM plates, leading to a complex nonlinear frequency response. It is further found that the power-law index, the longitudinal speed, the exci- tation amplitude, and the in-plane pretension force can significantly affect the nonlinear frequency-response characteristics of longitudinally traveling S-FGM plates.
文摘To get simpler operation in modified fuzzy adaptive learning control network （FALCON） in some engineering application, sigmoid nonlinear function is employed as a substitute of traditional Gaussian membership function. For making the modified FALCON learning more efficient and stable, a simulated annealing （SA） learning coefficient is introduced into learning algorithm. At first, the basic concepts and main advantages of FALCON were briefly reviewed. Subsequently, the topological structure and nodes operation were illustrated; the gradient-descent learning algorithm with SA learning coefficient was derived; and the distinctions between the archetype and the modification were analyzed. Eventually, the significance and worthiness of the modified FALCON were validated by its application to probability prediction of anode effect in aluminium electrolysis cells.
基金supported by National Natural Science Foundation of China(No.61864025)2021 Longyuan Youth Innovation and Entrepreneurship Talent(Team),Young Doctoral Fund of Higher Education Institutions of Gansu Province(No.2021QB-49)+4 种基金Employment and Entrepreneurship Improvement Project of University Students of Gansu Province(No.2021-C-123)Intelligent Tunnel Supervision Robot Research Project(China Railway Scientific Research Institute(Scientific Research)(No.2020-KJ016-Z016-A2)Lanzhou Jiaotong University Youth Foundation(No.2015005)Gansu Higher Education Research Project(No.2016A-018)Gansu Dunhuang Cultural Relics Protection Research Center Open Project(No.GDW2021YB15).
文摘Road extraction based on deep learning is one of hot spots of semantic segmentation in the past decade.In this work,we proposed a framework based on codec network for automatic road extraction from remote sensing images.Firstly,a pre-trained ResNet34 was migrated to U-Net and its encoding structure was replaced to deepen the number of network layers,which reduces the error rate of road segmentation and the loss of details.Secondly,dilated convolution was used to connect the encoder and the decoder of network to expand the receptive field and retain more low-dimensional information of the image.Afterwards,the channel attention mechanism was used to select the information of the feature image obtained by up-sampling of the encoder,the weights of target features were optimized to enhance the features of target region and suppress the features of background and noise regions,and thus the feature extraction effect of the remote sensing image with complex background was optimized.Finally,an adaptive sigmoid loss function was proposed,which optimizes the imbalance between the road and the background,and makes the model reach the optimal solution.Experimental results show that compared with several semantic segmentation networks,the proposed method can greatly reduce the error rate of road segmentation and effectively improve the accuracy of road extraction from remote sensing images.
基金Supported by the National Natural Science Foundation of China(61179041, 61101240)the Zhejiang Provincial Natural Science Foundation of China(Y6110117)
文摘In this paper, the technique of approximate partition of unity is used to construct a class of neural networks operators with sigmoidal functions. Using the modulus of continuity of function as a metric, the errors of the operators approximating continuous functions defined on a compact interval are estimated. Furthmore, Bochner-Riesz means operators of double Fourier series are used to construct networks operators for approximating bivariate functions, and the errors of approximation by the operators are estimated.
基金The Doctoral Fund of Ministry of Education of China(No.20090092110052)the Natural Science Foundation of Higher Education Institutions of Jiangsu Province(No.12KJA460002)College Industrialization Project of Jiangsu Province(No.JHB2012-21)
文摘To realize the high precision and real-time interpolation of the NURBS （non-uniform rational B-spline） curve, a kinetic model based on the modified sigmoid function is proposed. The constraints of maximum feed rate, chord error, curvature radius and interpolator cycle are discussed. This kinetic model reduces the cubic polynomial S-shape model and the trigonometry function S-shape model from 15 sections into 3 sections under the precondition of jerk, acceleration and feedrate continuity. Then an optimized Adams algorithm using the difference quotient to replace the derivative is presented to calculate the interpolator cycle parameters. The higher-order derivation in the Taylor expansion algorithm can be avoided by this algorithm. Finally, the simplified design is analyzed by reducing the times of computing the low-degree zero-value B-spline basis function and the simplified De Boor-Cox recursive algorithm is proposed. The simulation analysis indicates that by these algorithms, the feed rate is effectively controlled according to tool path. The calculated amount is decreased and the calculated speed is increased while the machining precision is ensured. The experimental results show that the target parameter can be correctly calculated and these algorithms can be applied to actual systems.
基金jointly funded by the Strategic Priority Research Program of the Chinese Academy of Sciences(XDA20020401)the Young Foundation of Institute of Mountain Hazard and Environment(SDS-QN-1702)National Natural Science Foundation of China(Grant No.41571205)
文摘The above-ground net primary production(ANPP) and the precipitation-use efficiency(PUE) regulate the carbon and water cycles in grassland ecosystems, but the relationships among the ANPP, PUE and precipitation are still controversial. We selected 717 grassland sites with ANPP and mean annual precipitation(MAP) data from 40 publications to characterize the relationships ANPP–MAP and PUE–MAP across different grassland types. The MAP and ANPP showed large variations across all grassland types, ranging from 69 to 2335 mm and 4.3 to 1706 g m^(-2), respectively. The global maximum PUE ranged from 0.19 to 1.49 g m^(-2) mm^(-1) with a unimodal pattern. Analysis using the sigmoid function explained the ANPP–MAP relationship best at the global scale. The gradient of the ANPP–MAP graph was small for arid and semi-arid sites(MAP <400 mm). This study improves our understanding of the relationship between ANPP and MAP across dry grassland ecosystems. It provides new perspectives on the prediction and modeling of variations in the ANPP for different grassland types along precipitation gradients.
基金supported by Grant-in-Aid for Scientific Research(C) (No. 20560248) of Japan
文摘Recently, various control methods represented by proportional-integral-derivative （PID） control are used for robotic control. To cope with the requirements for high response and precision, advanced feedforward controllers such as gravity compensator, Coriolis/centrifugal force compensator and friction compensators have been built in the controller. Generally, it causes heavy computational load when calculating the compensating value within a short sampling period. In this paper, integrated recurrent neural networks are applied as a feedforward controller for PUMA560 manipulator. The feedforward controller works instead of gravity and Coriolis/centrifugal force compensators. In the learning process of the neural network by using back propagation algorithm, the learning coefficient and gain of sigmoid function are tuned intuitively and empirically according to teaching signals. The tuning is complicated because it is being conducted by trial and error. Especially, when the scale of teaching signal is large, the problem becomes crucial. To cope with the problem which concerns the learning performance, a simple and adaptive learning technique for large scale teaching signals is proposed. The learning techniques and control effectiveness are evaluated through simulations using the dynamic model of PUMA560 manipulator.
基金funded by the European Regional Development Funding via RISC projectby CPER Region Haute Normandie France,the Australian Research Council via a Future Fellowship(FT110100896)Discovery Project(DP140100203)
文摘In this paper we present a new version of Chen＇s system： a piecewise linear （PWL） Chert system of fractional-order. Via a sigmoid-like function, the discontinuous system is transformed into a continuous system. By numerical simulations, we reveal chaotic behaviors and also multistability, i.e., the existence of small pararheter windows where, for some fixed bifurcation parameter and depending on initial conditions, coexistence of stable attractors and chaotic attractors is possible. Moreover, we show that by using an algorithm to switch the bifurcation parameter, the stable attractors can be numerically approximated.
基金This project is supported by Key Technology R & D Program of China during the 10th 5-year Plan Period(No.2002BA404A21)State Key Laboratory of Automobile Safety and Energy, China(No.KF2005-004).
文摘A new vehicle steering control algorithm is presented. Unlike the traditional methods do, the algorithm uses a sigmoid function to describe the principle of the human driver＇s steering strategy. Based on this function, a human simulating vehicle steering model, human-simulating steering control（HS） algorithm is designed. In order to improve the adaptability to different environments, a parameter adaptive adjustment algorithm is presented. This algorithm can online modify the value of the key parameters of the HS real time. HS controller is used on a vehicle equipped with computer vision system and computer controlled steering actuator system, the result from the automatic vehicle steering experiment shows that the HS algorithm gives good performance at different speed, even at the maximum speed of 172 km/h.
基金supported in part by the fiscal 2010 special grant from the president of the Prefectural University of Kumamoto
文摘We analyzed the relationship between internode number and intemode length for one of the largest bamboo, Phyllostachys pubescens Mazel ex Houz. For 50 sample culms with various sizes felled in a pure stand of P. pubescens, the intemode number was assigned from base to tip and the length for each internode was directly measured. The result indicated that the intemode length should be cumulated from base to tip, and then the cumulated internode length should be relativized by the total culm length. It was inappropriate to relativize the internode length by the maximum intenode length. In addition, the relationship between the relative internode number （the intemode number relativized by the total number of intemodes） and the relative cumulated internode length should be described not by a power function but by a sigmoid function such as the third-order function. The determined function enabled us to estimate the actual internode length, with the root mean squared error being 4 cm. In conclusion, the mathematical expression presented here, i.e., the relativization of the cumulated internode length by the total culm length and the application of the sigmoid function, will be useful in describing the relationship between internode number and internode length for P. pubescens.
基金This project was funded by the Deanship of Scientific Research(DSR)at King Abdulaziz University,Jeddah under grant no.G:656-611-1439The authors,therefore,acknowledge with thanks DSR for technical and financial support.
文摘An IDS(intrusion detection system)provides a foremost front line mechanism to guard networks,systems,data,and information.That’s why intrusion detection has grown as an active study area and provides significant contribution to cyber-security techniques.Multiple techniques have been in use but major concern in their implementation is variation in their detection performance.The performance of IDS lies in the accurate detection of attacks,and this accuracy can be raised by improving the recognition rate and significant reduction in the false alarms rate.To overcome this problem many researchers have used different machine learning techniques.These techniques have limitations and do not efficiently perform on huge and complex data about systems and networks.This work focused on ELM(Extreme Learning Machine)technique due to its good capabilities in classification problems and dealing with huge data.The ELM has different activation functions,but the problem is to find out which function is more suitable and performs well in IDS.This work investigates this problem.Here,Well-known activation functions like:sine,sigmoid and radial basis are explored,investigated and applied to measure their performance on the GA(Genetic Algorithm)features subset and with full features set.The NSL-KDD dataset is used as a benchmark.The empirical results are analyzed,addressed and compared among different activation functions of the ELM.The results show that the radial basis and sine functions perform better on GA feature set than the full feature set while the performance of the sigmoid function is almost equal on both features sets.So,the proposal of GA based feature selection reduced 21 features out of 41 that brought up to 98%accuracy and enhanced overall efficiency of extreme learning machine in intrusion detection.
文摘Incessant fire-outbreak in urban settlements has remained intractable especially in developing countries like Nigeria. This is often characterized by grave socio-economic aftermath effects. Urban fire outbreak in Nigerian cities has been on increase in recent times. The major problem faced by fire fighters in Nigerian urban centres is that there are no mechanisms to detect fire outbreaks early enough to save lives and properties. They often rely on calls made by neighbours or occupants when an outbreak occurs and this accounts for the delay in fighting fire outbreaks. This work uses Artificial Neural Networks (ANN) with backpropagation method to detect the occurrence of urban fires. The method uses smoke density, room temperature and cooking gas concentration as inputs. The work was implemented using Java programming language and results showed that it detected the occurrence of urban fires with reasonable accuracy. The work is recommended for use to minimize the effect of urban fire outbreak.
基金funded by the National Natural Science Foundation of China,grant number 51679024,51909018the Science and Technology Innovation Fundation of Dalian City,grant number 2019J12GX026+1 种基金the Fundamental Research Funds for the Central University,grant number 3132019343,3132021132the University 111 Project of China,grant number B08046.
文摘This paper investigates the problem of parameter identification for ship nonlinear Nomoto model with small test data,a nonlinear innovation-based identification algorithm is presented by embedding sigmoid function in the stochastic gradient algorithm.To demonstrate the validity of the algorithm,an identification test is carried out on the ship‘SWAN’with only 26 sets of test data.Furthermore,the identification effects of the least squares algorithm,original stochastic gradient algorithm and the improved stochastic gradient algorithm based on nonlinear innovation are compared.Generally,the stochastic gradient algorithm is not suitable for the condition of small test data.The simulation results indicate that the improved stochastic gradient algorithm with sigmoid function greatly increases its accuracy of parameter identification and has 14.2%up compared with the least squares algorithm.Then the effectiveness of the algorithm is verified by another identification test on the ship‘Galaxy’,the accuracy of parameter identification can reach more than 95%which can be used in ship motion simulation and controller design.The proposed algorithm has advantages of the small test data,fast speed and high accuracy of identification,which can be extended to other parameter identification systems with less sample data.
文摘In this paper, spatial patterns of a diffusive predator-prey model with sigmoid （Holling type III） ratio-dependent functional response which concerns the influence of logistic population growth in prey and intra-species competition among predators are investigated. The （local and global） asymptotic stability behavior of the corresponding non- spatial model around the unique positive interior equilibrium point in homogeneous steady state is obtained. In addition, we derive the conditions for Turing instability and the consequent parametric Turing space in spatial domain. The results of spatial pat- tern analysis through numerical simulations are depicted and analyzed. ~rthermore, we perform a series of numerical simulations and find that the proposed model dynamics exhibits complex pattern replication. The feasible results obtained in this paper indicate that the effect of diffusion in Turing instability plays an important role to understand better the pattern formation in ecosystem.
