<p>Shiv Nadar University is a comprehensive, multidisciplinary, research-focused, and student-centric University offering a full range of academic programs at the undergraduate, postgraduate and doctoral level. With state-of-the-art infrastructure, the University comprises of academic wings, sophisticated labs, international standard sports facilities, amphitheaters, auditorium, conference rooms and smart classrooms. The University&rsquo;s goal is to become internationally recognized for the quality of its research and creative endeavors and their applicability to improving quality of life, generating new insights and expanding the boundaries of human knowledge creativity. Committed to excellence in teaching, research and service, the University aims to serve the higher education needs of India and the world beyond.</p>

Shiv Nadar University

The paper is concerned with the design of a hybrid controller structure, consisting of the adaptive control law and a neural-network-based learning scheme for adaptation of time-varying controller parameters. The target error vector for weight adaptation of the neural networks is derived using the Lyapunov-function approach. The global stability of the closed-loop feedback system is guaranteed, provided the structure of the robot-manipulator dynamics model is exact. Generalisation of the controller over the desired trajectory space has been established using an online weight-learning scheme. Model learning, using a priori knowledge of a robot arm model, has been shown to improve tracking accuracy. The proposed control scheme has been implemented using both MLN and RBF networks. Faster convergence, better generalisation and superior tracking accuracy have been achieved in the case of the RBF network. © IEE, 1996.

World University root node

Electrical Engineering

(SoE) School of Engineering

IEE Proceedings: Control Theory and Applications

Neuro-adaptive hybrid controller for robot-manipulator tracking control

A robust hybrid control method that integrates a neural controller, and a H∞ controller is proposed. The neural controller is designed for feedback linearization and the H∞ controller is used to minimize the effect of parameter uncertainties of the model on the tracking performance. Simulations are provided to show that the disturbances generated from the model error will be quickly compensated and hence more robust performance can be achieved.

IEEE International Conference on Neural Networks - Conference Proceedings

Robust hybrid controller for robotic manipulators using neural and H∞ techniques

This paper is concerned with the design of a neuroadaptive trajectory tracking controller. The paper presents a new control scheme based on inversion of a feedforward neural model of a robot arm. The proposed control scheme requires two modules. The first module consists of an appropriate feedforward neural model of forward dynamics of the robot arm that continuously accounts for the changes in the robot dynamics. The second module implements an efficient network inversion algorithm that computes the control action by inverting the neural model. In this paper, a new extended Kalman filter (EKF) based network inversion scheme is proposed. The scheme is evaluated through comparison with two other schemes of network inversion: gradient search in input space and Lyapunov function approach. Using these three inversion schemes the proposed controller was implemented for trajectory tracking control of a two-link manipulator. Simulation results in all cases confirm the efficacy of control input prediction using network inversion. Comparison of the inversion algorithms in terms of tracking accuracy showed the superior performance of the EKF based inversion scheme over others. © 1996 IEEE.

Fulltext

IEEE Transactions on Neural Networks

On adaptive trajectory tracking of a robot manipulator using inversion of its neural emulator

A query based learning algorithm is proposed in this paper to obtain a valid neural emulator of the robot manipulator using radial basis function networks. This algorithm is centred around the concept of network inversion for which we have proposed an extended Kalman filtering (EKF) based algorithm. Initially the training data set is generated by tracking various random trajectories in robot work space. During this process, control input (or joint actuating torque) vector is computed using a fixed PD control scheme with added dither signal. The RBF network thus learned is tested for its validity using the inversion based recall process. In case the neural network is partially trained, additional input/output data pairs are generated. The data so generated are used to retrain the RBFN model of the robot manipulator. Experiments showed significant improvement in the neural model after retraining. © 1996 Taylor & Francis Group, LLC.

IETE Journal of Research

A Query Based Learning Algorithm for Modelling Robot Manipulator Dynamics using Network Inversion

Electrochemical noise measurements were taken in a 7.5 cm diameter, 10 m long multiphase oil/water/gas flow system. Full pipe flow studies were conducted for liquid velocities of 0.8, 1.0 and 1.5 m/s and slug flow for Froude numbers 4, 6, 9 and 12, using two phase (ASTM standard saltwater and carbon dioxide gas) and three phase (oil of 2cP viscosity, saltwater and gas) flow. The pressure was maintained at 134 KPa and temperature at 40 C. Noise data was measured with a Zero Resistance Ammeter and analyzed using software provided with the instrument. The noise fluctuations show trends which can be related to the type of How. Full pipe flow studies show random fluctuations around the mean for both potential and current noise.This implies the occurrence of uniform corrosion. The Fast Fourier Transform (FFT) shows roll-off greater than -40 dB/decade in magnitude, characteristic of uniform or general corrosion. In slug flow the potential noise fluctuations take the form of characteristic transients. The frequency of these transients correspond to the observed frequency of slugs. The FFT of this data shows roll-off slopes less than -20 dB/decade shown to be characteristic of pitting corrosion. A second slope of magnitude greater than -30 dB/decade is observed at higher frequencies and this might signify the presence of a second time constant controlling the corrosion reaction. This is due to the unique mechanisms associated with the mixing zone of slugs. Peaks in the Fast Fourier Transform (FFT) are also seen at particular frequencies. One of these corresponds to the slugging frequency. An increase in the percent change in standard deviation of current with Froude number implies enhanced corrosion activity on the surface of the electrode. The electrochemical noise technique is extremely useful for studying corrosion mechanisms and flow parameters in multiphase flow pipelines. Its ability to isolate dynamic incidents and study them both in the time and frequency domain makes it a powerful tool for gaining an insight into mechanistic information on corrosion in multiphase pipelines. © 1996 by NACE International.

NACE - International Corrosion Conference Series

Use of electrochemical noise to monitor multiphase flow and corrosion

A novel ultrasonic flow metering system been developed that has the capability of measuring liquid film thicknesses of the order of 0.1 mm and mist contents in gas flow of 4-5%. The principle of this nonintrusive technique is based on recording the transit time and attenuation of ultrasonic signals in both axial and radial directions using sets of upto eight transducers flush-mounted on the outside pipe wall at intervals around the pipe circumference and at two locations along the pipeline. The data from the meter is then used in existing multiphase flow models developed at the Corrosion in Multiphase Systems Center, to predict the quality of the gas and flow rate of the individual phases as a function of time.

American Society of Mechanical Engineers, Fluids Engineering Division (Publication) FED

Development of a novel non-intrusive, ultrasonic flow meter for wet gas pipelines

Experiments have been performed in a 10 cm ID horizontal pipeline to observe the effect of temperature, carbon dioxide partial pressure, flow, and oil/water composition, on corrosion in horizontal multiphase slug flow, using a low viscosity (2 cp) oil. Corrosion rates have been measured for water cuts ranging form 20 to 100%, carbon dioxide partial pressures of 0. 27, 0. 45 and 0. 79 MPa, temperatures of 40. 60, 70. 80 and 90 C and Froude numbers of 6, 9 and 12. The corrosion rates are found to increase with an increase in temperature, Froude number and carbon dioxide partial pressure. The corrosion rates are found to decrease with increase in oil composition. The average void fraction as well as the pressure drop across the slug is found to increase with an increase in Froude number and carbon dioxide partial pressure. A model to predict corrosion rates for low viscosity oils has been established which relates the corrosion rate to the pressure gradient, temperature, carbon dioxide partial pressure and water cut. © 1996 by NACE International.

Predictive model for sweet corrosion in horizontal multiphase slug flow

Mathematical models are developed tor the prediction of corrosion rates in large diameter, high pressure, horizontal multiphase oil/water/gas pipelines. The models quantify the effects of multiphase flow characteristics, such as phase distribution, turbulent intensity, wall shear stress, and pressure drop, on the corrosion processes at the metal surface. Mechanisms are presented for both full pipe flow and slug flow. The models incorporate the effects of fluid properties, pressure, temperature, and pipe diameter, on corrosion rate. In full pipe oil/water flows, the corrosion rate can be predicted as a function of the total wall shear stress for a given carbon dioxide partial pressure and temperature. In slug flow, the effect of the liquid composition becomes significant, and the total shear stress is difficult to calculate. In this case, the pressure gradient across the slug can be correlated with the corrosion rate for a given liquid composition, at a particular carbon dioxide partial pressure and system temperature.

Modeling of corrosion in high pressure horizontal multiphase flow pipelines

The paper investigates the application of inversion of a radial basis function network (RBFN) to nonlinear control problems for which the structure of the nonlinearity is unknown. Initially, the RBF network is trained to learn the forward dynamics of the plant. Two different controller structures are then proposed based on this identified RBFN model. In one scheme, a feedback control law is derived based on the input prediction by inversion of the RBFN model so that the system is Lyapunov stable. The second kind of controller structure predicts the feedforward control action, while the fixed controller actuates the feedback stabilizing signal. An extended Kalman filtering based algorithm is employed to carry out the network inversion during each sampling interval. Two examples are presented to verify the proposed scheme. Simulation results show that the performance of the controller based on the proposed network inversion scheme is efficient.

Inversion of RBF networks and applications to adaptive control of nonlinear systems

Fast and accurate trajectory tracking of a robot arm primarily depends on the knowledge of its explicit inverse dynamics model. On line learning of inverse dynamics using supervised learning algorithm is difficult in the absence of a priori knowledge of command error. On the other hand, self-organizing neural network employing unsupervised learning scheme does not depend on the command error. These networks are suitable for both off-line and on-line schemes of learning the inverse dynamics. The present paper proposes both the schemes based on two unsupervised learning algorithms, namely, Kohonen's self-organizing topology conserving feature map and 'neural gas' algorithm. Simulation results on a single link manipulator confirms the efficacy of the proposed schemes.

IEEE/IAS International Conference on Industrial Automation and Control, Proceedings

Self-organizing neural networks for learning inverse dynamics of robot manipulator

This paper presents a neural network based control scheme for robot tracking applications. Gaussian networks are used to model both the forward and the inverse dynamics of a robot arm. The feedforward torque is actuated by the output of inverse dynamic mapping while the feedback control law is derived using the linear perturbation model of the identified forward dynamics along the desired trajectory. The gradient based learning algorithm is used and the effectiveness of the proposed scheme is highlighted through simulation studies. © 1994.

Robotics and Autonomous Systems

Trajectory tracking of robot manipulator using Gaussian networks

A unified study of adaptive control and neural network based control schemes for the trajectory tracking problem of robot manipulators is presented. Efficacy of parametrized adaptive algorithms in compensating the structured uncertainties in robot dynamics is verified through extensive simulation. The ability of neural networks to provide a robust adaptive framework in the presence of both structured and unstructured uncertainties is investigated. A case study is carried out in support of a parametrized adaptive scheme using neural networks. Simulation results clearly indicate that the neural network based adaptive controller achieves better tracking in the presence of parametric uncertainties as well as unmodelled effects compared to the simple direct adaptive scheme. © 1994 Taylor & Francis Group, LLC.

International Journal of Systems Science

Adaptive manipulator trajectory control using neural networks

For the trajectory-tracking problem of robot manipulators various decentralized control schemes are investigated in this paper. These control schemes result in poor tracking performance in case of large payload variations. A compensation scheme using an error-data knowledge base is developed to improve the performance robustness. © 1993.

On robustness of decentralized control for robot manipulators

For representing the input-output behaviour of a robot manipulator by a linear time-invariant model, four direct linearization schemes are: (i) state linearization, (ii) linearization based on an identification method, (iii) linearization based on neglecting velocity-dependent and gravity terms and (iv) linearization based on neglecting the velocity-dependent term only (rate linearization). In order to make an appropriate choice of linear model for the development of real-time control, these schemes are extensively studied in this paper. It is shown that the rate linearization method leads to a satisfactory tradeoff between computation, accuracy, and stability. In the case of high velocity motions, a combination of state linearization and rate linearization is proposed. © 1993 Kluwer Academic Publishers.

Journal of Intelligent & Robotic Systems

Comparative study on linearized robot models

Design procedures based on exact eigenstructure assignment are not suitable because of very high computational requirements. A computation-effective control design procedure for eigenstructure assignment using aggregation is presented. Dominant eigenvalues are placed at specified locations in the complex plane and the non-dominant eigenvalues are placed in a specified disk. The proposed design procedure is applied to the trajectory tracking problem of a robot manipulator. An error-pattern based payload estimation and compensation scheme is also proposed to improve performance robustness. © 1993 Taylor & Francis Group, LLC.

Eigenstructure assignment in robot tracking applications

To make an appropriate choice of linear model for the development of real‐time control of robot manipulators, four direct linearization schemes have been studied here on the basis of computation and accuracy. Accuracy is assured in the state linearization method, whereas the rate linearization method leads to a satisfactory trade‐off between computational effort and accuracy. In the case of high‐velocity motions, a combination of state and rate linearization is proposed. Copyright © 1992 Wiley Periodicals, Inc., A Wiley Company

Journal of Robotic Systems

Linearized dynamic models for trajectory control of robot manipulators

The tracking problem for a robotic manipulator with uncertain dynamics is considered. The proposed control scheme utilizes manipulator ‚inverse’ as an adaptive feedforward controller, the controller gains are updated to cow with variations in the manipulator inverse due to changes the operating point A feedback conlroller that provides arbitrary assignment of all eigenvalues and certain parts o f 3 closed-loop eigenvector structure: Is also used. Eigenstructure assignment is used to enhance closed-loop stability and to achieve robust tracking. Simulation results are presented in support of the proposed control scheme. The results demonstrate satisfactory performance of the controller despite variations in payload. © 1991 Taylor & Francis Group, LLC.

Robust trajectory control of a robot manipulator

The DC motors used in robotics and other industrial applications are submitted to frequent environment changes. As a consequence the motor model has time-varying parameters resulting in a poor performance when nonadaptive linear controllers are used. In this paper, two adaptive schemes for the position control of DC servomotors have been studied. The first scheme is based on self-tuning technique which uses pole-assignment by linear state feedback as the control strategy The second scheme uses adaptive feedforward in addition to pole-assigning linear state feedback. Experimental results show that a simple self-tuning control scheme for DC servomotors is a better solution to low cost automation problems in robotics and other industrial applications. The results obtained show that the adaptive scheme proposed by Kelly seems to offer no great benefit, rather it seems to increase the amount of overshoot. Standard schemes such as self-tuning control provide cost-effective solution to the control problem. The paper is organized as follows: a DC motor model and linear state feedback controller are discussed in Section II. Section III deals with self-tuning regulator, whereas adaptive feed-forward with linear state feedback is dealt with in Section IV. Experimental results on the DC motor control are given in Section V and some conclusions are drawn in Section VI.

IFAC Proceedings Series

Adaptive control schemes for DC servomotors

Two closed-loop hierarchical control strategies are examined for the control of a two-area hydro-thermal power system model. The control is obtained using the interaction prediction approach and structural perturbation method. The first approach has been developed to provide optimal feedback control using an off-line decentralized calculation procedure. The second method is based on obtaining local feedback controllers for each subsystem on the first level by ignoring the interactions, then a global controller at the second level provides corrective signals to neutralize the interaction effects. In addition, the dynamic behaviour of the power system under the effect of non-linearities is also investigated. Finally, a two-area hydro-thermal power system is used to illustrate the approaches. © 1989 Taylor & Francis Group, LLC.

Multilevel optimization of a hydro-thermal interconnected power system

Industrial robots are multilink, highly nonlinear and coupled dynamic systems, where control is a difficult and challenging task. This paper presents a review of the basic problems involved and their existing control solutions. This review also attempts to give a broadbased classification of control strategies.