Venkatesh Madyastha - Academia.edu (original) (raw)

Papers by Venkatesh Madyastha

Proceedings of SPIE, Oct 6, 2011

ABSTRACT

In this paper, we present a joint state and adaptive parameter identification scheme for the case... more In this paper, we present a joint state and adaptive parameter identification scheme for the cases when all the states of the system are measured and when only some states of the system are measured. When all the states are measured, we show that, in the presence of process and measurement noise, the state and parameter estimation errors are bounded. To this end, we show that this is possible only through the appropriate design of a virtual input which ensures that the system error signals are bounded. As a special case of all the states being measured, we show that in the case of a noise free system, the state estimation errors converge to the origin. For the case when only some states are measured, we show that for a linear system with n states, m inputs and p measurements, we can estimate at most p 2 entries of the system matrix and pm entries of the input matrix.

A method is presented for augmenting an extended Kalman filter with an adaptive element. The resu... more A method is presented for augmenting an extended Kalman filter with an adaptive element. The resulting estimator provides robustness to parameter uncertainty and unmodeled dynamics. The design of the adaptive element employs a linearly parameterized neural network. The network weights are adjusted on line using the filter error residuals. Boundedness of signals is proven using Lyapunov's direct method and a backstepping argument. Simulations illustrate the theoretical results.

This paper discusses estimation and guidance strategies for vision-based target tracking. Specifi... more This paper discusses estimation and guidance strategies for vision-based target tracking. Specific applications include formation control of multiple unmanned aerial vehicles (UAVs) and air-to-air refueling. We assume that no information is communicated between the aircraft, and only passive 2-D vision information is available to maintain formation. To improve the robustness of the estimation process with respect to unknown target aircraft acceleration, the nonlinear estimator (EKF) is augmented with an adaptive neural network (NN). The guidance strategy involves augmenting the inverting solution of nonlinear lineof-sight (LOS) range kinematics with the output of an adaptive NN to compensate for target aircraft LOS velocity. Simulation results are presented that illustrate the various approaches.

Asia-Pacific Journal of Chemical Engineering, Mar 25, 2011

Growing environmental concerns and shrinking water resources require methods beyond conventional ... more Growing environmental concerns and shrinking water resources require methods beyond conventional wastewater treatment. Membrane bioreactor (MBR) is a technology that has become a ubiquitous choice for high quality treatment and reuse of wastewater. One of the key challenges in wastewater treatment is the high energy cost associated with aeration. MBR systems use feedback control to regulate the measured dissolved oxygen level at a predetermined set point by manipulating the blower throughputs. However, for high dynamic loads, feedback control may not result in the best performance and energy efficiency. Any attempt to optimize performance and power consumption beyond a simple controls strategy requires a proper trade‐off analysis between investments on additional sensors and the long‐term benefits. This article proposes a joint state and parameter estimation methodology, which measures and controls the MBR system using available measurements. Thus, limitations of feedback strategies can be overcome by predicting the impact of time varying disturbances on the outputs. The novelty in this approach is the ability to reconstruct the unknown states and parameters with available measurements. The unknown parameters are adaptively estimated online. Lyapunov's direct method is employed to show boundedness of state and parameter estimation errors. Simulation results illustrate the efficacy of the approach. Copyright © 2011 Curtin University of Technology and John Wiley & Sons, Ltd.

Water Science and Technology, Oct 1, 2011

Activated sludge treatment is one of the most widely used processes for wastewater treatment (WWT... more Activated sludge treatment is one of the most widely used processes for wastewater treatment (WWT). These systems are built with sufficient design margin to allow changes in loading and process conditions. This is necessary and prudent to overcome limitations in measurement, monitoring and controlling of WWT process parameters at the desired frequency. Online sensors for mixed liquor suspended solids, chemical oxygen demand (COD), nitrogen, phosphorus, and other parameters available today are limited in application due to high cost and low reliability. Hence, many of the parameters are measured off-line when needed. This paper provides a framework to estimate parameters on-line using limited and delayed measurements. The proposed approach is based on the design of a Bayesian filter such as an extended Kaiman filter (EKF), which measures and controls membrane bioreactor system using limited and delayed measurements. The objective is to estimate the states and parameters with limited and delayed measurements. Simulations show the efficacy of the proposed approach.

A system in accordance with the invention uses an adaptive element to augment a filter for tracki... more A system in accordance with the invention uses an adaptive element to augment a filter for tracking an observed system. The adaptive element only requires a single neural network and does not require an error observer. The adaptive element provides robustness to parameter uncertainty and unmodeled dynamics present in the observed system for improved tracking performance over the filter alone. The adaptive element can be implemented with a linearly parameterized neural network, whose weights are adapted online using error residuals generated from the Filter. Boundedness of the signals generated by the system can be proven using Lyapunov' s direct method and a backstepping argument. A related apparatus and method are also disclosed.

SPIE Proceedings, 2011

ABSTRACT

2004 43rd IEEE Conference on Decision and Control (CDC) (IEEE Cat. No.04CH37601), 2004

This paper implements several methods for performing vision-based formation flight control of mul... more This paper implements several methods for performing vision-based formation flight control of multiple aircraft in the presence of obstacles. No information is communicated between aircraft, and only passive 2-D vision information is available to maintain formation. The methods for formation control rely either on estimating the range from 2-D vision information by using Extended Kalman Filters or directly regulating the size of the image subtended by a leader aircraft on the image plane. When the image size is not a reliable measurement, especially at large ranges, we consider the use of bearing-only information. In this case, observability with respect to the relative distance between vehicles is accomplished by the design of a time-dependent formation geometry. To improve the robustness of the estimation process with respect to unknown leader aircraft acceleration, we augment the EKF with an adaptive neural network. 2-D and 3-D simulation results are presented that illustrate the various approaches.

IEEE Conference on Decision and Control and European Control Conference, 2011

In this paper, we present a joint state and adaptive parameter identification scheme for the case... more In this paper, we present a joint state and adaptive parameter identification scheme for the cases when all the states of the system are measured and when only some states of the system are measured. When all the states are measured, we show that, in the presence of process and measurement noise, the state and parameter estimation errors are bounded. To this end, we show that this is possible only through the appropriate design of a virtual input which ensures that the system error signals are bounded. As a special case of all the states being measured, we show that in the case of a noise free system, the state estimation errors converge to the origin. For the case when only some states are measured, we show that for a linear system with n states, m inputs and p measurements, we can estimate at most p 2 entries of the system matrix and pm entries of the input matrix.

AIAA Guidance, Navigation, and Control Conference, 2011

ABSTRACT

Proceedings of the 2005, American Control Conference, 2005.

This paper discusses estimation and guidance strategies for vision-based target tracking. Specifi... more This paper discusses estimation and guidance strategies for vision-based target tracking. Specific applications include formation control of multiple unmanned aerial vehicles (UAVs) and air-to-air refueling. We assume that no information is communicated between the aircraft, and only passive 2-D vision information is available to maintain formation. To improve the robustness of the estimation process with respect to unknown target aircraft acceleration, the nonlinear estimator (EKF) is augmented with an adaptive neural network (NN). The guidance strategy involves augmenting the inverting solution of nonlinear lineof-sight (LOS) range kinematics with the output of an adaptive NN to compensate for target aircraft LOS velocity. Simulation results are presented that illustrate the various approaches.

Proceedings of the 2005, American Control Conference, 2005.

A method is presented for augmenting an extended Kalman filter with an adaptive element. The resu... more A method is presented for augmenting an extended Kalman filter with an adaptive element. The resulting estimator provides robustness to parameter uncertainty and unmodeled dynamics. The design of the adaptive element employs a linearly parameterized neural network. The network weights are adjusted on line using the filter error residuals. Boundedness of signals is proven using Lyapunov's direct method and a backstepping argument. Simulations illustrate the theoretical results.

Proceedings of the 41st IEEE Conference on Decision and Control, 2002.

In this paper we address the problem of augmenting a linear observer with an adaptive element. Th... more In this paper we address the problem of augmenting a linear observer with an adaptive element. The design of the adaptive element employs two nonlinearly parameterized neural networks, the input and output layer weights of which are adapted on line. The goal is to improve the performance of the linear observer when applied to a nonlinear system. The networks' teaching signal is generated using a second linear observer of the nominal system's error dynamics. Boundedness of signals is shown through Lyapunov's direct method. The approach is robust to unmodeled dynamics and disturbances. Simulations illustrate the theoretical results.

This paper presents three recursive Bayesian filtering approaches in phase shifting interferometr... more This paper presents three recursive Bayesian filtering approaches in phase shifting interferometry involving a piezoelectric device (PZT). The three approaches are extended Kalman filtering, unscented Kalman filtering, and particle filtering. These approaches estimate the phase ...

Water Science and Technology, 2011

Activated sludge treatment is one of the most widely used processes for wastewater treatment (WWT... more Activated sludge treatment is one of the most widely used processes for wastewater treatment (WWT). These systems are built with sufficient design margin to allow changes in loading and process conditions. This is necessary and prudent to overcome limitations in measurement, monitoring and controlling of WWT process parameters at the desired frequency. Online sensors for mixed liquor suspended solids, chemical oxygen demand (COD), nitrogen, phosphorus, and other parameters available today are limited in application due to high cost and low reliability. Hence, many of the parameters are measured off-line when needed. This paper provides a framework to estimate parameters on-line using limited and delayed measurements. The proposed approach is based on the design of a Bayesian filter such as an extended Kalman filter (EKF), which measures and controls membrane bioreactor system using limited and delayed measurements. The objective is to estimate the states and parameters with limited ...

Optics Letters, 2009

We present an unscented Kalman filter to identify the phase step imparted to a piezoelectric tran... more We present an unscented Kalman filter to identify the phase step imparted to a piezoelectric transducer in phase shifting interferometry in the presence of Gaussian noise. The advantage of the proposed algorithm lies in its ability to determine the phase step values between -pi and pi rad without any prior calibration of the piezoelectric device. The algorithm is tested rigorously by using the simulated data in the presence of Gaussian distributed noise. Experimental validations are also performed in a holographic interferometry optical setup to verify the proposed approach. Once the phase step is identified, the interference phase can be estimated by using the least-squares fitting approach.

Proceedings of SPIE, Oct 6, 2011

ABSTRACT

In this paper, we present a joint state and adaptive parameter identification scheme for the case... more In this paper, we present a joint state and adaptive parameter identification scheme for the cases when all the states of the system are measured and when only some states of the system are measured. When all the states are measured, we show that, in the presence of process and measurement noise, the state and parameter estimation errors are bounded. To this end, we show that this is possible only through the appropriate design of a virtual input which ensures that the system error signals are bounded. As a special case of all the states being measured, we show that in the case of a noise free system, the state estimation errors converge to the origin. For the case when only some states are measured, we show that for a linear system with n states, m inputs and p measurements, we can estimate at most p 2 entries of the system matrix and pm entries of the input matrix.

A method is presented for augmenting an extended Kalman filter with an adaptive element. The resu... more A method is presented for augmenting an extended Kalman filter with an adaptive element. The resulting estimator provides robustness to parameter uncertainty and unmodeled dynamics. The design of the adaptive element employs a linearly parameterized neural network. The network weights are adjusted on line using the filter error residuals. Boundedness of signals is proven using Lyapunov's direct method and a backstepping argument. Simulations illustrate the theoretical results.

This paper discusses estimation and guidance strategies for vision-based target tracking. Specifi... more This paper discusses estimation and guidance strategies for vision-based target tracking. Specific applications include formation control of multiple unmanned aerial vehicles (UAVs) and air-to-air refueling. We assume that no information is communicated between the aircraft, and only passive 2-D vision information is available to maintain formation. To improve the robustness of the estimation process with respect to unknown target aircraft acceleration, the nonlinear estimator (EKF) is augmented with an adaptive neural network (NN). The guidance strategy involves augmenting the inverting solution of nonlinear lineof-sight (LOS) range kinematics with the output of an adaptive NN to compensate for target aircraft LOS velocity. Simulation results are presented that illustrate the various approaches.

Asia-Pacific Journal of Chemical Engineering, Mar 25, 2011

Growing environmental concerns and shrinking water resources require methods beyond conventional ... more Growing environmental concerns and shrinking water resources require methods beyond conventional wastewater treatment. Membrane bioreactor (MBR) is a technology that has become a ubiquitous choice for high quality treatment and reuse of wastewater. One of the key challenges in wastewater treatment is the high energy cost associated with aeration. MBR systems use feedback control to regulate the measured dissolved oxygen level at a predetermined set point by manipulating the blower throughputs. However, for high dynamic loads, feedback control may not result in the best performance and energy efficiency. Any attempt to optimize performance and power consumption beyond a simple controls strategy requires a proper trade‐off analysis between investments on additional sensors and the long‐term benefits. This article proposes a joint state and parameter estimation methodology, which measures and controls the MBR system using available measurements. Thus, limitations of feedback strategies can be overcome by predicting the impact of time varying disturbances on the outputs. The novelty in this approach is the ability to reconstruct the unknown states and parameters with available measurements. The unknown parameters are adaptively estimated online. Lyapunov's direct method is employed to show boundedness of state and parameter estimation errors. Simulation results illustrate the efficacy of the approach. Copyright © 2011 Curtin University of Technology and John Wiley & Sons, Ltd.

Water Science and Technology, Oct 1, 2011

Activated sludge treatment is one of the most widely used processes for wastewater treatment (WWT... more Activated sludge treatment is one of the most widely used processes for wastewater treatment (WWT). These systems are built with sufficient design margin to allow changes in loading and process conditions. This is necessary and prudent to overcome limitations in measurement, monitoring and controlling of WWT process parameters at the desired frequency. Online sensors for mixed liquor suspended solids, chemical oxygen demand (COD), nitrogen, phosphorus, and other parameters available today are limited in application due to high cost and low reliability. Hence, many of the parameters are measured off-line when needed. This paper provides a framework to estimate parameters on-line using limited and delayed measurements. The proposed approach is based on the design of a Bayesian filter such as an extended Kaiman filter (EKF), which measures and controls membrane bioreactor system using limited and delayed measurements. The objective is to estimate the states and parameters with limited and delayed measurements. Simulations show the efficacy of the proposed approach.

A system in accordance with the invention uses an adaptive element to augment a filter for tracki... more A system in accordance with the invention uses an adaptive element to augment a filter for tracking an observed system. The adaptive element only requires a single neural network and does not require an error observer. The adaptive element provides robustness to parameter uncertainty and unmodeled dynamics present in the observed system for improved tracking performance over the filter alone. The adaptive element can be implemented with a linearly parameterized neural network, whose weights are adapted online using error residuals generated from the Filter. Boundedness of the signals generated by the system can be proven using Lyapunov' s direct method and a backstepping argument. A related apparatus and method are also disclosed.

SPIE Proceedings, 2011

ABSTRACT

2004 43rd IEEE Conference on Decision and Control (CDC) (IEEE Cat. No.04CH37601), 2004

This paper implements several methods for performing vision-based formation flight control of mul... more This paper implements several methods for performing vision-based formation flight control of multiple aircraft in the presence of obstacles. No information is communicated between aircraft, and only passive 2-D vision information is available to maintain formation. The methods for formation control rely either on estimating the range from 2-D vision information by using Extended Kalman Filters or directly regulating the size of the image subtended by a leader aircraft on the image plane. When the image size is not a reliable measurement, especially at large ranges, we consider the use of bearing-only information. In this case, observability with respect to the relative distance between vehicles is accomplished by the design of a time-dependent formation geometry. To improve the robustness of the estimation process with respect to unknown leader aircraft acceleration, we augment the EKF with an adaptive neural network. 2-D and 3-D simulation results are presented that illustrate the various approaches.

IEEE Conference on Decision and Control and European Control Conference, 2011

In this paper, we present a joint state and adaptive parameter identification scheme for the case... more In this paper, we present a joint state and adaptive parameter identification scheme for the cases when all the states of the system are measured and when only some states of the system are measured. When all the states are measured, we show that, in the presence of process and measurement noise, the state and parameter estimation errors are bounded. To this end, we show that this is possible only through the appropriate design of a virtual input which ensures that the system error signals are bounded. As a special case of all the states being measured, we show that in the case of a noise free system, the state estimation errors converge to the origin. For the case when only some states are measured, we show that for a linear system with n states, m inputs and p measurements, we can estimate at most p 2 entries of the system matrix and pm entries of the input matrix.

AIAA Guidance, Navigation, and Control Conference, 2011

ABSTRACT

Proceedings of the 2005, American Control Conference, 2005.

This paper discusses estimation and guidance strategies for vision-based target tracking. Specifi... more This paper discusses estimation and guidance strategies for vision-based target tracking. Specific applications include formation control of multiple unmanned aerial vehicles (UAVs) and air-to-air refueling. We assume that no information is communicated between the aircraft, and only passive 2-D vision information is available to maintain formation. To improve the robustness of the estimation process with respect to unknown target aircraft acceleration, the nonlinear estimator (EKF) is augmented with an adaptive neural network (NN). The guidance strategy involves augmenting the inverting solution of nonlinear lineof-sight (LOS) range kinematics with the output of an adaptive NN to compensate for target aircraft LOS velocity. Simulation results are presented that illustrate the various approaches.

Proceedings of the 2005, American Control Conference, 2005.

A method is presented for augmenting an extended Kalman filter with an adaptive element. The resu... more A method is presented for augmenting an extended Kalman filter with an adaptive element. The resulting estimator provides robustness to parameter uncertainty and unmodeled dynamics. The design of the adaptive element employs a linearly parameterized neural network. The network weights are adjusted on line using the filter error residuals. Boundedness of signals is proven using Lyapunov's direct method and a backstepping argument. Simulations illustrate the theoretical results.

Proceedings of the 41st IEEE Conference on Decision and Control, 2002.

In this paper we address the problem of augmenting a linear observer with an adaptive element. Th... more In this paper we address the problem of augmenting a linear observer with an adaptive element. The design of the adaptive element employs two nonlinearly parameterized neural networks, the input and output layer weights of which are adapted on line. The goal is to improve the performance of the linear observer when applied to a nonlinear system. The networks' teaching signal is generated using a second linear observer of the nominal system's error dynamics. Boundedness of signals is shown through Lyapunov's direct method. The approach is robust to unmodeled dynamics and disturbances. Simulations illustrate the theoretical results.

This paper presents three recursive Bayesian filtering approaches in phase shifting interferometr... more This paper presents three recursive Bayesian filtering approaches in phase shifting interferometry involving a piezoelectric device (PZT). The three approaches are extended Kalman filtering, unscented Kalman filtering, and particle filtering. These approaches estimate the phase ...

Water Science and Technology, 2011

Activated sludge treatment is one of the most widely used processes for wastewater treatment (WWT... more Activated sludge treatment is one of the most widely used processes for wastewater treatment (WWT). These systems are built with sufficient design margin to allow changes in loading and process conditions. This is necessary and prudent to overcome limitations in measurement, monitoring and controlling of WWT process parameters at the desired frequency. Online sensors for mixed liquor suspended solids, chemical oxygen demand (COD), nitrogen, phosphorus, and other parameters available today are limited in application due to high cost and low reliability. Hence, many of the parameters are measured off-line when needed. This paper provides a framework to estimate parameters on-line using limited and delayed measurements. The proposed approach is based on the design of a Bayesian filter such as an extended Kalman filter (EKF), which measures and controls membrane bioreactor system using limited and delayed measurements. The objective is to estimate the states and parameters with limited ...

Optics Letters, 2009

We present an unscented Kalman filter to identify the phase step imparted to a piezoelectric tran... more We present an unscented Kalman filter to identify the phase step imparted to a piezoelectric transducer in phase shifting interferometry in the presence of Gaussian noise. The advantage of the proposed algorithm lies in its ability to determine the phase step values between -pi and pi rad without any prior calibration of the piezoelectric device. The algorithm is tested rigorously by using the simulated data in the presence of Gaussian distributed noise. Experimental validations are also performed in a holographic interferometry optical setup to verify the proposed approach. Once the phase step is identified, the interference phase can be estimated by using the least-squares fitting approach.