Pub Date : 2018-06-01DOI: 10.1109/ICIST.2018.8426140
B. Dong, Shuxiang Wang, Fan Zhou, Yan Li, Shenquan Wang, Keping Liu, Yuan-chun Li
This paper presents a decentralized robust optimal control method for modular robot manipulators (MRMs) via a novel critic-identifier (CI) structure-based adaptive dynamic programming (ADP) scheme. The robust control problem of MRMs is transformed into an optimal compensation control approach, which combines model-based compensation control, identifier-based learning control and ADP-based optimal control. The dynamic model of MRMs is formulated based on a torque sensing technique that is deployed for each joint module, where the local dynamic information is utilized effectively to design the model compensation controller. A neural network (NN) identifier is established to approximate the dynamics of the interconnected dynamic coupling (IDC). Based on the ADP algorithm, the Hamiltonian-Jacobi-Bellman (HJB) equation can be solved by constructing a critic NN, and the approximate optimal control policy is derived. The closed-loop robotic system is guaranteed to be asymptotic stable by the implementation of a set of decentralized control policies that have been developed. Finally, simulations verify the effectiveness of the proposed method.
{"title":"Critic-Identifier Structure-Based ADP for Decentralized Robust Optimal Control of Modular Robot Manipulators","authors":"B. Dong, Shuxiang Wang, Fan Zhou, Yan Li, Shenquan Wang, Keping Liu, Yuan-chun Li","doi":"10.1109/ICIST.2018.8426140","DOIUrl":"https://doi.org/10.1109/ICIST.2018.8426140","url":null,"abstract":"This paper presents a decentralized robust optimal control method for modular robot manipulators (MRMs) via a novel critic-identifier (CI) structure-based adaptive dynamic programming (ADP) scheme. The robust control problem of MRMs is transformed into an optimal compensation control approach, which combines model-based compensation control, identifier-based learning control and ADP-based optimal control. The dynamic model of MRMs is formulated based on a torque sensing technique that is deployed for each joint module, where the local dynamic information is utilized effectively to design the model compensation controller. A neural network (NN) identifier is established to approximate the dynamics of the interconnected dynamic coupling (IDC). Based on the ADP algorithm, the Hamiltonian-Jacobi-Bellman (HJB) equation can be solved by constructing a critic NN, and the approximate optimal control policy is derived. The closed-loop robotic system is guaranteed to be asymptotic stable by the implementation of a set of decentralized control policies that have been developed. Finally, simulations verify the effectiveness of the proposed method.","PeriodicalId":331555,"journal":{"name":"2018 Eighth International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114847542","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-06-01DOI: 10.1109/ICIST.2018.8426077
N. Zhang, Jiang Xiong, Jing Zhong, Keenan Leatham
We present a Gaussian process regression (GPR) algorithm with variable models to adapt to numerous pattern recognition data for classification. The algorithms of the Gaussian process regression (GPR) models including the rational quadratic GPR, squared exponential GPR, matern 5/2 GPR, and exponential GPR are described. The response plot, predicted vs. actual plot, and residuals plot of these GPR models are demonstrated. In addition, a comprehensive comparison of classification performance among rational quadratic GPR, squared exponential GPR, matern 5/2 GPR, and exponential GPR is presented in terms of various model statistics. Furthermore, the classification error rates of these four GPR based models are in comparison to the extended nearest neighbor (ENN), classic k-nearest Neighbor (KNN), naive Bayes, linear discriminant analysis (LDA), and the classic multilayer perceptron (MLP) neural network. The excellent experimental results demonstrated that the Gaussian process regression models provide a very promising feature selection solution to numerous pattern recognition problems. The algorithm is able to learn from the global distribution, therefore improving pattern recognition performance.
{"title":"Gaussian Process Regression Method for Classification for High-Dimensional Data with Limited Samples","authors":"N. Zhang, Jiang Xiong, Jing Zhong, Keenan Leatham","doi":"10.1109/ICIST.2018.8426077","DOIUrl":"https://doi.org/10.1109/ICIST.2018.8426077","url":null,"abstract":"We present a Gaussian process regression (GPR) algorithm with variable models to adapt to numerous pattern recognition data for classification. The algorithms of the Gaussian process regression (GPR) models including the rational quadratic GPR, squared exponential GPR, matern 5/2 GPR, and exponential GPR are described. The response plot, predicted vs. actual plot, and residuals plot of these GPR models are demonstrated. In addition, a comprehensive comparison of classification performance among rational quadratic GPR, squared exponential GPR, matern 5/2 GPR, and exponential GPR is presented in terms of various model statistics. Furthermore, the classification error rates of these four GPR based models are in comparison to the extended nearest neighbor (ENN), classic k-nearest Neighbor (KNN), naive Bayes, linear discriminant analysis (LDA), and the classic multilayer perceptron (MLP) neural network. The excellent experimental results demonstrated that the Gaussian process regression models provide a very promising feature selection solution to numerous pattern recognition problems. The algorithm is able to learn from the global distribution, therefore improving pattern recognition performance.","PeriodicalId":331555,"journal":{"name":"2018 Eighth International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114426410","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-06-01DOI: 10.1109/ICIST.2018.8426122
Lin Zhang, Gang Li, Y. Han, Yuxin Zhao, Ziqian Chen
This paper studies robust adaptive tracking control for a family of multi-mode nonlinear systems. Adaptive laws using tuning functions are proposed to address the parametric uncertainties and unknown disturbances, which can avoid the problem of over-parameterization. In addition, the variation law of the modes is developed based on the mode-dependent sojourn time scheme, which exploits the information of each subsystem, i.e., sojourn time is realized in a subsystem sense. Based on the proposed time-constraint scheme, variation signals that are less conservative than those based on sojourn time are designed. Globally uniformly ultimately bounded stability of the closed-loop multi-mode system is guaranteed. Furthermore, the steady-state performance characterized by an ultimate bound of the tracking error is presented. A numerical simulation demonstrates the effectiveness of the proposed method.
{"title":"Robust Adaptive Tracking Control of a Class of Uncertain Multi-Mode Nonlinear Systems with Unknown Disturbances","authors":"Lin Zhang, Gang Li, Y. Han, Yuxin Zhao, Ziqian Chen","doi":"10.1109/ICIST.2018.8426122","DOIUrl":"https://doi.org/10.1109/ICIST.2018.8426122","url":null,"abstract":"This paper studies robust adaptive tracking control for a family of multi-mode nonlinear systems. Adaptive laws using tuning functions are proposed to address the parametric uncertainties and unknown disturbances, which can avoid the problem of over-parameterization. In addition, the variation law of the modes is developed based on the mode-dependent sojourn time scheme, which exploits the information of each subsystem, i.e., sojourn time is realized in a subsystem sense. Based on the proposed time-constraint scheme, variation signals that are less conservative than those based on sojourn time are designed. Globally uniformly ultimately bounded stability of the closed-loop multi-mode system is guaranteed. Furthermore, the steady-state performance characterized by an ultimate bound of the tracking error is presented. A numerical simulation demonstrates the effectiveness of the proposed method.","PeriodicalId":331555,"journal":{"name":"2018 Eighth International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122803943","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-06-01DOI: 10.1109/ICIST.2018.8426130
Xiupin Lv, Nankun Mu, X. Liao
This paper proposes a new algorithm of generating pseudo-random numbers where delay coupled map lattice is utilized as a pseudo-random function. k-order Chebyshev map embedded time-varying delay is introduced as the dynamic function of delay coupled map lattice to improve random performance of the system. The proposed pseudo-random number generator is subjected to statistical tests which is the well-known NIST 800–22 and TestU01 test in the field of security and other related properties are also investigated. The result shows that the proposed pseudo-random number generator holds better pseudo-random characteristics and suggests strong candidate for cryptographic applications.
{"title":"A Pseudo-Random Number Generator Based on Delay Coupled Map Lattice","authors":"Xiupin Lv, Nankun Mu, X. Liao","doi":"10.1109/ICIST.2018.8426130","DOIUrl":"https://doi.org/10.1109/ICIST.2018.8426130","url":null,"abstract":"This paper proposes a new algorithm of generating pseudo-random numbers where delay coupled map lattice is utilized as a pseudo-random function. k-order Chebyshev map embedded time-varying delay is introduced as the dynamic function of delay coupled map lattice to improve random performance of the system. The proposed pseudo-random number generator is subjected to statistical tests which is the well-known NIST 800–22 and TestU01 test in the field of security and other related properties are also investigated. The result shows that the proposed pseudo-random number generator holds better pseudo-random characteristics and suggests strong candidate for cryptographic applications.","PeriodicalId":331555,"journal":{"name":"2018 Eighth International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127829545","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-06-01DOI: 10.1109/ICIST.2018.8426169
Huaqiang Qiu, Baoran An, Shen Yin
An industrial process system plays a crucial role in the economic development of a country or region, process monitoring is effective in ensuring the safety and reliability of industrial processes, and has received much attention. For complex nonlinear systems, the traditional model-based methods and knowledge-based methods are difficult to apply, and data-driven methods provide a new solution. However, for the complex nonlinear systems with deterministic disturbances, the existing data-driven approaches also exhibit defects because they no longer satisfy the Gauss distribution. To solve this problem, a method called JITL-DD for process monitoring of nonlinear systems with deterministic disturbances is proposed. The JITL-DD combines the JITL model and the DD fault diagnosis method, the JITL model is used to predict the output of the local model, then the residual is processed as the input of the DD, and the fault information is obtained by analyzing the residual. The continuous stirred tank heater process is used as a simulation of the nonlinear system to illustrate the effectiveness of the proposed method.
{"title":"Research on Method of Process Monitoring with Deterministic Disturbances Based on Just-in-Time Learning","authors":"Huaqiang Qiu, Baoran An, Shen Yin","doi":"10.1109/ICIST.2018.8426169","DOIUrl":"https://doi.org/10.1109/ICIST.2018.8426169","url":null,"abstract":"An industrial process system plays a crucial role in the economic development of a country or region, process monitoring is effective in ensuring the safety and reliability of industrial processes, and has received much attention. For complex nonlinear systems, the traditional model-based methods and knowledge-based methods are difficult to apply, and data-driven methods provide a new solution. However, for the complex nonlinear systems with deterministic disturbances, the existing data-driven approaches also exhibit defects because they no longer satisfy the Gauss distribution. To solve this problem, a method called JITL-DD for process monitoring of nonlinear systems with deterministic disturbances is proposed. The JITL-DD combines the JITL model and the DD fault diagnosis method, the JITL model is used to predict the output of the local model, then the residual is processed as the input of the DD, and the fault information is obtained by analyzing the residual. The continuous stirred tank heater process is used as a simulation of the nonlinear system to illustrate the effectiveness of the proposed method.","PeriodicalId":331555,"journal":{"name":"2018 Eighth International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130325957","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-06-01DOI: 10.1109/ICIST.2018.8426085
M. Wang, G. Feng, Jianbin Qiu
This paper tackles the problem of piecewise affine (PWA) memory filtering design for discrete-time uncertain T-S fuzzy affine systems in finite frequency domain. It is assumed that the frequency of the disturbances is in a finite frequency domain. The objective is to design an admissible filter using both current and past output measurements of the system to guarantee the asymptotic stability of the filtering error system with a given finite frequency H) performance index. Based on piecewise fuzzy Lyapunov functions, a new sufficient condition for finite frequency H) filtering performance analysis is first derived, and then, the PWA memory filter synthesis is obtained. Finally, simulation studies are given to show the advantages and effectiveness of the proposed approach.
{"title":"A Novel Piecewise Affine Memory Filtering Design for T-S Fuzzy Affine Systems in Finite Frequency Domain","authors":"M. Wang, G. Feng, Jianbin Qiu","doi":"10.1109/ICIST.2018.8426085","DOIUrl":"https://doi.org/10.1109/ICIST.2018.8426085","url":null,"abstract":"This paper tackles the problem of piecewise affine (PWA) memory filtering design for discrete-time uncertain T-S fuzzy affine systems in finite frequency domain. It is assumed that the frequency of the disturbances is in a finite frequency domain. The objective is to design an admissible filter using both current and past output measurements of the system to guarantee the asymptotic stability of the filtering error system with a given finite frequency H) performance index. Based on piecewise fuzzy Lyapunov functions, a new sufficient condition for finite frequency H) filtering performance analysis is first derived, and then, the PWA memory filter synthesis is obtained. Finally, simulation studies are given to show the advantages and effectiveness of the proposed approach.","PeriodicalId":331555,"journal":{"name":"2018 Eighth International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125262345","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-06-01DOI: 10.1109/ICIST.2018.8426121
Zhenyu Liu, Huanhuan Chen
Judicial cases can be modeled with thetextual frequency vectors under the Bag-of-Words assumption to predict the decision outcome. However, such models are often with much more numbers of features than training samples, which usually leads to the over fitting problem. In this paper, we conduct an empirical investigation on linear dimensionality reduction of the high-dimensional judicial predictive models via the wide spread principal component analysis approach. The experimental results show that these high-dimensional models do not suffer from the overfitting problem, but the under fitting problem. Moreover, the higher-order dependency in the textual frequency data cannot be decorrelated by the linear dimensionality reduction approach, which restrains the performance of judicial classification models subject to the unchanged level of signal-noise ratio in the derived low-dimensional features.
{"title":"An Empirical Study of Linear Dimensionality Reduction for Judicial Predictive Models","authors":"Zhenyu Liu, Huanhuan Chen","doi":"10.1109/ICIST.2018.8426121","DOIUrl":"https://doi.org/10.1109/ICIST.2018.8426121","url":null,"abstract":"Judicial cases can be modeled with thetextual frequency vectors under the Bag-of-Words assumption to predict the decision outcome. However, such models are often with much more numbers of features than training samples, which usually leads to the over fitting problem. In this paper, we conduct an empirical investigation on linear dimensionality reduction of the high-dimensional judicial predictive models via the wide spread principal component analysis approach. The experimental results show that these high-dimensional models do not suffer from the overfitting problem, but the under fitting problem. Moreover, the higher-order dependency in the textual frequency data cannot be decorrelated by the linear dimensionality reduction approach, which restrains the performance of judicial classification models subject to the unchanged level of signal-noise ratio in the derived low-dimensional features.","PeriodicalId":331555,"journal":{"name":"2018 Eighth International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129414691","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-06-01DOI: 10.1109/ICIST.2018.8426142
Donovan L. Welsford, C. Pretorius, M. D. du Plessis
Inverse kinematics refers to determining the forces that must be applied to a particular system to result in a desired configuration of the system. In robotics, inverse kinematics means calculating the robot actuator movements necessary to make a robot perform a specific task. Calculating the inverse kinematics using traditional methods is a complex and time consuming task. This paper reports on a novel approach to predicting the inverse kinematics of a mobile robot using Neural Networks (NNs). The main advantage of using artificial intelligence to determine inverse kinematics is that minimal human input and intervention is required. This research makes use of Feedforward NNs to predict the motor velocities and the time that they must be maintained to make the robot reach a specified destination. Inertia and friction are automatically incorporated into the NN predictions. Experimental evidence is presented which shows that the proposed approach can successfully produce commands which allow the robot to traverse a given path.
{"title":"Neural Networks for Mobile Robot Inverse Kinematics","authors":"Donovan L. Welsford, C. Pretorius, M. D. du Plessis","doi":"10.1109/ICIST.2018.8426142","DOIUrl":"https://doi.org/10.1109/ICIST.2018.8426142","url":null,"abstract":"Inverse kinematics refers to determining the forces that must be applied to a particular system to result in a desired configuration of the system. In robotics, inverse kinematics means calculating the robot actuator movements necessary to make a robot perform a specific task. Calculating the inverse kinematics using traditional methods is a complex and time consuming task. This paper reports on a novel approach to predicting the inverse kinematics of a mobile robot using Neural Networks (NNs). The main advantage of using artificial intelligence to determine inverse kinematics is that minimal human input and intervention is required. This research makes use of Feedforward NNs to predict the motor velocities and the time that they must be maintained to make the robot reach a specified destination. Inertia and friction are automatically incorporated into the NN predictions. Experimental evidence is presented which shows that the proposed approach can successfully produce commands which allow the robot to traverse a given path.","PeriodicalId":331555,"journal":{"name":"2018 Eighth International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126094514","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}