Proceedings of the 2001 American Control Conference. (Cat. No.01CH37148), 2001
This paper applies multiple methods of solving general optimization problems to a time-optimal co... more This paper applies multiple methods of solving general optimization problems to a time-optimal control problem. By applying both constrained optimization and linear programming techniques to a particular problem, the benefits and drawbacks of each method can be easily described. While some of the results will be relative to the problem of interest, this paper describes some guidelines for solving general optimization problems. Linear programming techniques are efficient for finding good initial guesses into a constrained optimization routine. This hybrid approach can often improve efficiency in solving systems where it is difficult to find a good initial guess. Solutions of the time-optimal control problem for a standard two-mass spring-damper system show some interesting properties that illustrate the difficulty of applying constrained optimization exclusively. These solutions also show how the number of optimal switches changes as system parameters change. For those unwilling to apply linear programming to their own problems, the solutions of the time-optimal control problem given here can provide better insight in developing good initial guesses.
A scanning tunneling microscope (STM) uses a piezoelectric actuator to perform constant-velocity ... more A scanning tunneling microscope (STM) uses a piezoelectric actuator to perform constant-velocity scanning motion. Many feedback strategies have been proposed, but their achievable scan rate is substantially limited by the turnaround transients in the scan path. Therefore, a robust time-optimal command shaping technique with an iterative search procedure is introduced in this paper to improve the scan speed of piezoactuators, and is applicable to a general class of systems without rigid-body mode. Furthermore, a time-energy-optimal formulation is presented to reduce the in-maneuver oscillation. The hysteresis nonlinearity of piezoactuators is compensated using the proposed continuous numerical inversion algorithm. Finally, the closed-loop simulation shows the performance robustness in the presence of hysteresis cancellation error and natural frequency perturbation.
The control of vibration in large structures is complicated by the presence of many closely-space... more The control of vibration in large structures is complicated by the presence of many closely-spaced modes within the contoller bandwidh. Problem of stability robustness arise since accurate models for the entire structure are difficult to generate. This paper investigae stability issues associated with a decentralized control strategy applied to subsystem models of the entire structure. A novel coupling parameter which characterizes the degree of subsystem interaction in a 2�?2 system is introduced and defined. This parameter is used to determine stability bounds and performance limitations for decentralized active vibration control of large structures. A simple cantilever beam structure with attached disk is used to experimentally confirm the theoretical development when subsystem coupling is introduced by the actuators. The stability analysis suggests that the set of coupled system can be split into two regions by a critical coupling value, c*. One region can be controlled by a diagonal controller and the other can be controlled by an off-diagonal controller. Should the degree of coupling change from one side of c* to the other during operation, no single controller is capable of maintaining stability.
Abstract Feature selection has always been an important as-pect of statistical model identificat... more Abstract Feature selection has always been an important as-pect of statistical model identification and pattern classification. In this paper we introduce a novel information-theoretic index called the compensated quality factor (CQF) which selects the important features from a large ...
[1993] Proceedings IEEE International Conference on Robotics and Automation, 2000
A robot controller based on artificial neural networks (ANNs) is presented which is capable of co... more A robot controller based on artificial neural networks (ANNs) is presented which is capable of compensating for changing payload masses. Two different feedforward (multilayer) neural networks are used to generate the inverse dynamics and to estimate the payload mass of a two-link planar manipulator. The inverse dynamics ANN receives the same input signals as a conventional computed torque controller as well as the payload mass estimate. By using a separate ANN to generate the payload mass estimate, both ANNs can be trained off-line. The proposed neural network architecture is implemented on actual hardware using a neurocomputer. Experimental results indicate that the ANN-based controller is able to capture the nonlinear dynamics of the actual manipulator. The ANN mass estimator responds very quickly to changing payloads
The primary factor governing response time in SMA actuators is heat transfer rate which is repres... more The primary factor governing response time in SMA actuators is heat transfer rate which is represented by a nonlinear differential equation for cooling by free convection. To use classical control theory for controlling the temperature of the SMA wire, the nonlinear model must be linearized. In this paper, the heat transfer rate model is linearized using Chebyshev polynomials. This method of linearization has the same desirable local stability properties as the more traditional Taylor series expansion. It is shown that the maximum errors between the actual and linearized responses and their derivatives using Chebyshev linearization are smaller in magnitude than the errors obtained using Taylor series expansion, making Chebyshev linearization more desirable for this application.
Proceedings of the 2000 American Control Conference. ACC (IEEE Cat. No.00CH36334), 2000
Page 1. Proceedings of the American Control Conference Chicago, Illinois June 2000 Using Mutual I... more Page 1. Proceedings of the American Control Conference Chicago, Illinois June 2000 Using Mutual Information to Pre-Process Input Data for a Virtual Sensor Paul B. Deignan Jr. Peter H. Meckl Matthew Franchek John Abraham ...
Proceedings of the 1997 American Control Conference (Cat. No.97CH36041), 1997
In this paper, a model of a central fuel injected natural gas engine with transmission is develop... more In this paper, a model of a central fuel injected natural gas engine with transmission is developed and linear system identification is carried out to identify key model parameters that could lead to automated identification of transmission dynamics. The paper has two major components. First, the natural gas engine is modeled with an extension of the mean value engine model
Proceedings of 1995 American Control Conference - ACC'95, 1995
A control architecture that incorporates a radial basis function (RBF) neural network is proposed... more A control architecture that incorporates a radial basis function (RBF) neural network is proposed for the control of a particular nonlinear system, in this case the continuously stirred tank reactor (CSTR). The RBF network consists of Gaussian activation functions and is trained online to learn the inverse dynamics of the CSTR plant with and without parameter variation. Using Lyapunov stability
Proceedings of the 1999 American Control Conference (Cat. No. 99CH36251), 1999
A multivariable proportional-integral (PI) tuning strategy is developed and the advantage of the ... more A multivariable proportional-integral (PI) tuning strategy is developed and the advantage of the design is illustrated in an automotive application. The multivariable control design technique developed has disturbance rejection as its main objective rather than set-point tracking, which is the focus of most of the MIMO PI tuning techniques so far in the literature. The benefit of the design is that it does not try to minimize interactions in a multivariable plant and instead uses the interactions to achieve better disturbance rejection. The application of the control design technique is in multivariable speed and air-to-fuel ratio (A/F) control of a lean-burn natural gas engine to achieve less speed drop during load transients at idle
Proceedings of the 1999 IEEE International Conference on Control Applications (Cat. No.99CH36328), 1999
Many manufacturing devices must execute motions as quickly as possible to achieve profitable high... more Many manufacturing devices must execute motions as quickly as possible to achieve profitable high-volume production. This paper develops shaped torque inputs that minimize move time while avoiding overshoot and oscillation. These inputs are applied to an actual semiconductor manufacturing machine, utilizing both feedforward and feedback strategies for minimizing the degree of vibration generated during fast motions
Proceedings of the 2001 American Control Conference. (Cat. No.01CH37148), 2001
This paper applies multiple methods of solving general optimization problems to a time-optimal co... more This paper applies multiple methods of solving general optimization problems to a time-optimal control problem. By applying both constrained optimization and linear programming techniques to a particular problem, the benefits and drawbacks of each method can be easily described. While some of the results will be relative to the problem of interest, this paper describes some guidelines for solving general optimization problems. Linear programming techniques are efficient for finding good initial guesses into a constrained optimization routine. This hybrid approach can often improve efficiency in solving systems where it is difficult to find a good initial guess. Solutions of the time-optimal control problem for a standard two-mass spring-damper system show some interesting properties that illustrate the difficulty of applying constrained optimization exclusively. These solutions also show how the number of optimal switches changes as system parameters change. For those unwilling to apply linear programming to their own problems, the solutions of the time-optimal control problem given here can provide better insight in developing good initial guesses.
A scanning tunneling microscope (STM) uses a piezoelectric actuator to perform constant-velocity ... more A scanning tunneling microscope (STM) uses a piezoelectric actuator to perform constant-velocity scanning motion. Many feedback strategies have been proposed, but their achievable scan rate is substantially limited by the turnaround transients in the scan path. Therefore, a robust time-optimal command shaping technique with an iterative search procedure is introduced in this paper to improve the scan speed of piezoactuators, and is applicable to a general class of systems without rigid-body mode. Furthermore, a time-energy-optimal formulation is presented to reduce the in-maneuver oscillation. The hysteresis nonlinearity of piezoactuators is compensated using the proposed continuous numerical inversion algorithm. Finally, the closed-loop simulation shows the performance robustness in the presence of hysteresis cancellation error and natural frequency perturbation.
The control of vibration in large structures is complicated by the presence of many closely-space... more The control of vibration in large structures is complicated by the presence of many closely-spaced modes within the contoller bandwidh. Problem of stability robustness arise since accurate models for the entire structure are difficult to generate. This paper investigae stability issues associated with a decentralized control strategy applied to subsystem models of the entire structure. A novel coupling parameter which characterizes the degree of subsystem interaction in a 2�?2 system is introduced and defined. This parameter is used to determine stability bounds and performance limitations for decentralized active vibration control of large structures. A simple cantilever beam structure with attached disk is used to experimentally confirm the theoretical development when subsystem coupling is introduced by the actuators. The stability analysis suggests that the set of coupled system can be split into two regions by a critical coupling value, c*. One region can be controlled by a diagonal controller and the other can be controlled by an off-diagonal controller. Should the degree of coupling change from one side of c* to the other during operation, no single controller is capable of maintaining stability.
Abstract Feature selection has always been an important as-pect of statistical model identificat... more Abstract Feature selection has always been an important as-pect of statistical model identification and pattern classification. In this paper we introduce a novel information-theoretic index called the compensated quality factor (CQF) which selects the important features from a large ...
[1993] Proceedings IEEE International Conference on Robotics and Automation, 2000
A robot controller based on artificial neural networks (ANNs) is presented which is capable of co... more A robot controller based on artificial neural networks (ANNs) is presented which is capable of compensating for changing payload masses. Two different feedforward (multilayer) neural networks are used to generate the inverse dynamics and to estimate the payload mass of a two-link planar manipulator. The inverse dynamics ANN receives the same input signals as a conventional computed torque controller as well as the payload mass estimate. By using a separate ANN to generate the payload mass estimate, both ANNs can be trained off-line. The proposed neural network architecture is implemented on actual hardware using a neurocomputer. Experimental results indicate that the ANN-based controller is able to capture the nonlinear dynamics of the actual manipulator. The ANN mass estimator responds very quickly to changing payloads
The primary factor governing response time in SMA actuators is heat transfer rate which is repres... more The primary factor governing response time in SMA actuators is heat transfer rate which is represented by a nonlinear differential equation for cooling by free convection. To use classical control theory for controlling the temperature of the SMA wire, the nonlinear model must be linearized. In this paper, the heat transfer rate model is linearized using Chebyshev polynomials. This method of linearization has the same desirable local stability properties as the more traditional Taylor series expansion. It is shown that the maximum errors between the actual and linearized responses and their derivatives using Chebyshev linearization are smaller in magnitude than the errors obtained using Taylor series expansion, making Chebyshev linearization more desirable for this application.
Proceedings of the 2000 American Control Conference. ACC (IEEE Cat. No.00CH36334), 2000
Page 1. Proceedings of the American Control Conference Chicago, Illinois June 2000 Using Mutual I... more Page 1. Proceedings of the American Control Conference Chicago, Illinois June 2000 Using Mutual Information to Pre-Process Input Data for a Virtual Sensor Paul B. Deignan Jr. Peter H. Meckl Matthew Franchek John Abraham ...
Proceedings of the 1997 American Control Conference (Cat. No.97CH36041), 1997
In this paper, a model of a central fuel injected natural gas engine with transmission is develop... more In this paper, a model of a central fuel injected natural gas engine with transmission is developed and linear system identification is carried out to identify key model parameters that could lead to automated identification of transmission dynamics. The paper has two major components. First, the natural gas engine is modeled with an extension of the mean value engine model
Proceedings of 1995 American Control Conference - ACC'95, 1995
A control architecture that incorporates a radial basis function (RBF) neural network is proposed... more A control architecture that incorporates a radial basis function (RBF) neural network is proposed for the control of a particular nonlinear system, in this case the continuously stirred tank reactor (CSTR). The RBF network consists of Gaussian activation functions and is trained online to learn the inverse dynamics of the CSTR plant with and without parameter variation. Using Lyapunov stability
Proceedings of the 1999 American Control Conference (Cat. No. 99CH36251), 1999
A multivariable proportional-integral (PI) tuning strategy is developed and the advantage of the ... more A multivariable proportional-integral (PI) tuning strategy is developed and the advantage of the design is illustrated in an automotive application. The multivariable control design technique developed has disturbance rejection as its main objective rather than set-point tracking, which is the focus of most of the MIMO PI tuning techniques so far in the literature. The benefit of the design is that it does not try to minimize interactions in a multivariable plant and instead uses the interactions to achieve better disturbance rejection. The application of the control design technique is in multivariable speed and air-to-fuel ratio (A/F) control of a lean-burn natural gas engine to achieve less speed drop during load transients at idle
Proceedings of the 1999 IEEE International Conference on Control Applications (Cat. No.99CH36328), 1999
Many manufacturing devices must execute motions as quickly as possible to achieve profitable high... more Many manufacturing devices must execute motions as quickly as possible to achieve profitable high-volume production. This paper develops shaped torque inputs that minimize move time while avoiding overshoot and oscillation. These inputs are applied to an actual semiconductor manufacturing machine, utilizing both feedforward and feedback strategies for minimizing the degree of vibration generated during fast motions
Uploads