A methodology is presented whereby a neural network is used to learn the inverse kinematic relationships of the position and orientation of a six joint manipulator. The arm solution for the orientation of a manipulato...A methodology is presented whereby a neural network is used to learn the inverse kinematic relationships of the position and orientation of a six joint manipulator. The arm solution for the orientation of a manipulator using a self organizing neural net is studied in this paper. A new training model of the self organizing neural network is proposed by thoroughly studying Martinetz, Ritter and Schulten′s self organizing neural network based on Kohonen′s self organizing mapping algorithm using a Widrow Hoff type error correction rule and closely combining the characters of the inverse kinematic relationship for a robot arm. The computer simulation results for a PUMA 560 robot show that the proposed method has a significant improvement over other methods documented in the references in self organizing capability and precision by training process.展开更多
neural network model based on backbone propagation was applied to Learn-ing and predicting the interaction between antiparallelly interactive peptide seg-ments in proteins.Hydrophobic properties pf residues were found...neural network model based on backbone propagation was applied to Learn-ing and predicting the interaction between antiparallelly interactive peptide seg-ments in proteins.Hydrophobic properties pf residues were found dominant in in-terpeptides.Weights of each kind of residues, obtained by this work,suggestedsome different scales for the hydrophobicity of the residue.These will be helpful in understanding polypeptide structure and protein folding.展开更多
In this paper, by using Liapunov functional, some sufficient conditions are obtained for the stability of the equilibrium of a neural network model with delay of the type u ′ i(t)=-b iu i(t)+∑nj=1T ij f ...In this paper, by using Liapunov functional, some sufficient conditions are obtained for the stability of the equilibrium of a neural network model with delay of the type u ′ i(t)=-b iu i(t)+∑nj=1T ij f j(μ ju j(t-τ j))+c i, τ j≥0, i=1,2,…,n.展开更多
The global asymptotic stability for Hopfield neural networks with time delay was investigated, A theorem and two corollaries were obtained, in which the boundedness and differentiability of f(j) on R in some articles ...The global asymptotic stability for Hopfield neural networks with time delay was investigated, A theorem and two corollaries were obtained, in which the boundedness and differentiability of f(j) on R in some articles were deleted. Some sufficient conditions for the existence of global asymptotic stable equilibrium of the networks in this paper are better than the sufficient conditions in quoted articles.展开更多
In this paper a novel class of neural networks called generalized congruence neural networks (GCNN) is proposed. All neurons in the neural networks are activated in the form of congruence. The architectures, learnin...In this paper a novel class of neural networks called generalized congruence neural networks (GCNN) is proposed. All neurons in the neural networks are activated in the form of congruence. The architectures, learning rules and two algorithms are presented. Simulation results indicate that such network has satisfactory generalization properties near the sample points. Since this kind of neural nets can be easily operated and implemented, it is appropriate to make further research concerning the theory and applications of GCNN.展开更多
The robust attitude control for a novel coaxial twelve-rotor UAV which has much greater payload capacity,higher drive capability and damage tolerance than a quad-rotor UAV is studied. Firstly,a dynamical and kinematic...The robust attitude control for a novel coaxial twelve-rotor UAV which has much greater payload capacity,higher drive capability and damage tolerance than a quad-rotor UAV is studied. Firstly,a dynamical and kinematical model for the coaxial twelve-rotor UAV is designed. Considering model uncertainties and external disturbances,a robust backstepping sliding mode control( BSMC) with self recurrent wavelet neural network( SRWNN) method is proposed as the attitude controller for the coaxial twelve-rotor. A combinative algorithm of backstepping control and sliding mode control has simplified design procedures with much stronger robustness benefiting from advantages of both controllers. SRWNN as the uncertainty observer is able to estimate the lumped uncertainties effectively.Then the uniformly ultimate stability of the twelve-rotor system is proved by Lyapunov stability theorem. Finally,the validity of the proposed robust control method adopted in the twelve-rotor UAV under model uncertainties and external disturbances are demonstrated via numerical simulations and twelve-rotor prototype experiments.展开更多
A simple new BP algorithm named circle BP algorithm is introduced.With this algorithm,local minimums can be completely got rid of and learning speed can improve dramatically.It can be easily designed into the circuitr...A simple new BP algorithm named circle BP algorithm is introduced.With this algorithm,local minimums can be completely got rid of and learning speed can improve dramatically.It can be easily designed into the circuitry and advance further the application of MLP neural network .展开更多
he pattern recognition method and artificial neural network method to predict the composition of epilayer of GaInAsSb by MOCVD. It is concluded that a neural network with the composition of the vapor phase and growth ...he pattern recognition method and artificial neural network method to predict the composition of epilayer of GaInAsSb by MOCVD. It is concluded that a neural network with the composition of the vapor phase and growth temperature as training data can predict the composition of the epilayers. Satisfactory pattern recognition and artificial neural network classification results were obtained by using four technical parameters as characteristic features and the epilayers composition as classification criteria.展开更多
The field of neural network has found solid application in the past ten years and the field itself is still developing rapidly. Neural network is composed of many simple elements operating in parallel. A neural netwo...The field of neural network has found solid application in the past ten years and the field itself is still developing rapidly. Neural network is composed of many simple elements operating in parallel. A neural network can be trained to perform a particular mapping and this is the basis of its application to practical problems. In this paper, new methods for predicting the strong earthquakes are presented based on neural network. Neural network learns from existing earthquake sequences or earthquake precursors how to make medium and short term prediction of strong earthquakes. This paper describes two neural network prediction models. One is the model based on earthquake evolution sequences, which is applied to the modeling of the magnitude evolution sequences in the Mainland of China, the other is based on earthquake precursors, which is applied to the modeling of the occurrence time of strong earthquakes in North China. Test results show that the prediction methods based on neural networks are efficient, and convenient. They would find more application in the future.展开更多
Sufficient conditions are obtained for the existence, uniqueness and stability of T-periodic solutions far the Hopfield neural network equations with delay [GRAPHICS]
FAM(Fuzzy Associative Memory) Network Model, FAM Adaptive Learning Algorithm and Principal of FAM Inference Machine are introduced, and successfully application to ″New Generation Expert System for Earthquake Predict...FAM(Fuzzy Associative Memory) Network Model, FAM Adaptive Learning Algorithm and Principal of FAM Inference Machine are introduced, and successfully application to ″New Generation Expert System for Earthquake Prediction″ (NGESEP). This system has good function for knowledge learning without disadvantages of neural network, which the learned knowledge implied in network is difficult to be understood or interpreted by expert system.展开更多
A dynamic hysteresis model based on neural networks is proposed for piezoelectric actuator.Neural network has been widely applied to pattern recognition and system identification.However,it is unable to directly model...A dynamic hysteresis model based on neural networks is proposed for piezoelectric actuator.Neural network has been widely applied to pattern recognition and system identification.However,it is unable to directly model the systems with multi-valued mapping such as hysteresis.In order to handle this problem,a novel hysteretic operator is proposed to extract the dynamic property of the hysteresis.Moreover,it can construct an expanded input space to transform the multi-valued mapping of hysteresis into one-to-one mapping.Then neural networks can directly be used to approximate the behavior of dynamic hysteresis.Finally,the experimental results are presented to illustrate the potential of the proposed modeling method.展开更多
Based on wave digital filter(WDF) principles, this paper presents a digital model of cellular neural networks(CNNs). The model can precisely simulate the dynamic behavior of CNNs.
On the basis of the characteristic parameters selected from the fault sonic signals of cracking hammer with artificial diamond,by means of with time series analysis and time domain statistics,three layer artificial n...On the basis of the characteristic parameters selected from the fault sonic signals of cracking hammer with artificial diamond,by means of with time series analysis and time domain statistics,three layer artificial neural network is trained by an improved BP algorithm.The results state that the fault sonic signals can be identified by trained network system precisely.展开更多
In this paper, we first make a brief review on the fundamental properties of artificial neural networks (ANN) and the basic models, and explore emphatically some potential application of artificial neural networks in ...In this paper, we first make a brief review on the fundamental properties of artificial neural networks (ANN) and the basic models, and explore emphatically some potential application of artificial neural networks in the area of product quality diagnosis, prediction and control, state supervision and classification, factor recognition, and expert system based diagnosis, then set up the ANN models and expert system for quality forecasting, monitoring and diagnosing. We point out that combining ANN with other techniques will have the broad development and application of perspectives. Finally, the paper gives out some practical applications for the models and the system.展开更多
The analytical simulation relationship has been found between energy of a Hopfield back error propagation neural network model and the conventional mechanical mass system model. Since the energy expression is in qua...The analytical simulation relationship has been found between energy of a Hopfield back error propagation neural network model and the conventional mechanical mass system model. Since the energy expression is in quadratic form, which is corresponding to a steady state of energy distribution among processing unit of the neural network, and it is proved as a positive definite problem. Through simulation, a “Hamilton principle like” energy expression is introduced and an additional condition of the steady state of neural network system can be formulated through certain transformations. These results can be served for speeding the convergence of machine learning and identification processes of the neural network systems.展开更多
文摘A methodology is presented whereby a neural network is used to learn the inverse kinematic relationships of the position and orientation of a six joint manipulator. The arm solution for the orientation of a manipulator using a self organizing neural net is studied in this paper. A new training model of the self organizing neural network is proposed by thoroughly studying Martinetz, Ritter and Schulten′s self organizing neural network based on Kohonen′s self organizing mapping algorithm using a Widrow Hoff type error correction rule and closely combining the characters of the inverse kinematic relationship for a robot arm. The computer simulation results for a PUMA 560 robot show that the proposed method has a significant improvement over other methods documented in the references in self organizing capability and precision by training process.
文摘neural network model based on backbone propagation was applied to Learn-ing and predicting the interaction between antiparallelly interactive peptide seg-ments in proteins.Hydrophobic properties pf residues were found dominant in in-terpeptides.Weights of each kind of residues, obtained by this work,suggestedsome different scales for the hydrophobicity of the residue.These will be helpful in understanding polypeptide structure and protein folding.
文摘In this paper, by using Liapunov functional, some sufficient conditions are obtained for the stability of the equilibrium of a neural network model with delay of the type u ′ i(t)=-b iu i(t)+∑nj=1T ij f j(μ ju j(t-τ j))+c i, τ j≥0, i=1,2,…,n.
文摘The global asymptotic stability for Hopfield neural networks with time delay was investigated, A theorem and two corollaries were obtained, in which the boundedness and differentiability of f(j) on R in some articles were deleted. Some sufficient conditions for the existence of global asymptotic stable equilibrium of the networks in this paper are better than the sufficient conditions in quoted articles.
文摘In this paper a novel class of neural networks called generalized congruence neural networks (GCNN) is proposed. All neurons in the neural networks are activated in the form of congruence. The architectures, learning rules and two algorithms are presented. Simulation results indicate that such network has satisfactory generalization properties near the sample points. Since this kind of neural nets can be easily operated and implemented, it is appropriate to make further research concerning the theory and applications of GCNN.
基金Supported by the National Natural Science Foundation of China(No.11372309,61304017)Science and Technology Development Plan Key Project of Jilin Province(No.20150204074GX)the Science and Technology Special Fund Project of Provincial Academy Cooperation(No.2017SYHZ00024)
文摘The robust attitude control for a novel coaxial twelve-rotor UAV which has much greater payload capacity,higher drive capability and damage tolerance than a quad-rotor UAV is studied. Firstly,a dynamical and kinematical model for the coaxial twelve-rotor UAV is designed. Considering model uncertainties and external disturbances,a robust backstepping sliding mode control( BSMC) with self recurrent wavelet neural network( SRWNN) method is proposed as the attitude controller for the coaxial twelve-rotor. A combinative algorithm of backstepping control and sliding mode control has simplified design procedures with much stronger robustness benefiting from advantages of both controllers. SRWNN as the uncertainty observer is able to estimate the lumped uncertainties effectively.Then the uniformly ultimate stability of the twelve-rotor system is proved by Lyapunov stability theorem. Finally,the validity of the proposed robust control method adopted in the twelve-rotor UAV under model uncertainties and external disturbances are demonstrated via numerical simulations and twelve-rotor prototype experiments.
文摘A simple new BP algorithm named circle BP algorithm is introduced.With this algorithm,local minimums can be completely got rid of and learning speed can improve dramatically.It can be easily designed into the circuitry and advance further the application of MLP neural network .
文摘he pattern recognition method and artificial neural network method to predict the composition of epilayer of GaInAsSb by MOCVD. It is concluded that a neural network with the composition of the vapor phase and growth temperature as training data can predict the composition of the epilayers. Satisfactory pattern recognition and artificial neural network classification results were obtained by using four technical parameters as characteristic features and the epilayers composition as classification criteria.
文摘The field of neural network has found solid application in the past ten years and the field itself is still developing rapidly. Neural network is composed of many simple elements operating in parallel. A neural network can be trained to perform a particular mapping and this is the basis of its application to practical problems. In this paper, new methods for predicting the strong earthquakes are presented based on neural network. Neural network learns from existing earthquake sequences or earthquake precursors how to make medium and short term prediction of strong earthquakes. This paper describes two neural network prediction models. One is the model based on earthquake evolution sequences, which is applied to the modeling of the magnitude evolution sequences in the Mainland of China, the other is based on earthquake precursors, which is applied to the modeling of the occurrence time of strong earthquakes in North China. Test results show that the prediction methods based on neural networks are efficient, and convenient. They would find more application in the future.
文摘Sufficient conditions are obtained for the existence, uniqueness and stability of T-periodic solutions far the Hopfield neural network equations with delay [GRAPHICS]
文摘FAM(Fuzzy Associative Memory) Network Model, FAM Adaptive Learning Algorithm and Principal of FAM Inference Machine are introduced, and successfully application to ″New Generation Expert System for Earthquake Prediction″ (NGESEP). This system has good function for knowledge learning without disadvantages of neural network, which the learned knowledge implied in network is difficult to be understood or interpreted by expert system.
基金supported by the National Natural Science Foundation of China(No.61273184)the Program for Changjiang Scholars and Innovative Research Team in University(No.IRT13097)the Natural Science Foundation of Zhejiang Province(Nos.LY15F030022, LY13E050025,LZ15F030005)
文摘A dynamic hysteresis model based on neural networks is proposed for piezoelectric actuator.Neural network has been widely applied to pattern recognition and system identification.However,it is unable to directly model the systems with multi-valued mapping such as hysteresis.In order to handle this problem,a novel hysteretic operator is proposed to extract the dynamic property of the hysteresis.Moreover,it can construct an expanded input space to transform the multi-valued mapping of hysteresis into one-to-one mapping.Then neural networks can directly be used to approximate the behavior of dynamic hysteresis.Finally,the experimental results are presented to illustrate the potential of the proposed modeling method.
文摘Based on wave digital filter(WDF) principles, this paper presents a digital model of cellular neural networks(CNNs). The model can precisely simulate the dynamic behavior of CNNs.
文摘On the basis of the characteristic parameters selected from the fault sonic signals of cracking hammer with artificial diamond,by means of with time series analysis and time domain statistics,three layer artificial neural network is trained by an improved BP algorithm.The results state that the fault sonic signals can be identified by trained network system precisely.
文摘In this paper, we first make a brief review on the fundamental properties of artificial neural networks (ANN) and the basic models, and explore emphatically some potential application of artificial neural networks in the area of product quality diagnosis, prediction and control, state supervision and classification, factor recognition, and expert system based diagnosis, then set up the ANN models and expert system for quality forecasting, monitoring and diagnosing. We point out that combining ANN with other techniques will have the broad development and application of perspectives. Finally, the paper gives out some practical applications for the models and the system.
文摘The analytical simulation relationship has been found between energy of a Hopfield back error propagation neural network model and the conventional mechanical mass system model. Since the energy expression is in quadratic form, which is corresponding to a steady state of energy distribution among processing unit of the neural network, and it is proved as a positive definite problem. Through simulation, a “Hamilton principle like” energy expression is introduced and an additional condition of the steady state of neural network system can be formulated through certain transformations. These results can be served for speeding the convergence of machine learning and identification processes of the neural network systems.
