Papers by Farzad Pourboghrat
Summary form only given, as follows. A feedforward neural network was used to solve the problem o... more Summary form only given, as follows. A feedforward neural network was used to solve the problem of inverse kinematics for the redundant robots. A learning algorithm was also developed for the training of the network. The convergence of the training process was guaranteed according to Liapunov's stability theory. Moreover, the speed of training can be increased by increasing a learning
IEEE power engineering review, Nov 1, 2001
A novel concept for power quality study is proposed. The concept integrates the power system mode... more A novel concept for power quality study is proposed. The concept integrates the power system modeling, classifying, and characterizing of power quality events, studying equipment sensitivity to the event disturbance and locating the point of event occurrence into one unified frame. Both Fourier and wavelet analyses are applied for extracting distinct features of various types of events as well as for characterizing the events. A new fuzzy expert system for classifying power quality events based on such features is presented with improved performance over previous neural network-based methods. A novel simulation method is outlined for evaluating the operating characteristics of the equipment during specific events. A software prototype implementing the concept has been developed in MATLAB. The voltage sag event is taken as an example for illustrating the analysis methods and software implementation issues. It is concluded that the proposed approach is feasible and promising for real world applications.
ABSTRACT The feasibility of utilizing neural networks for the design of learning controllers is i... more ABSTRACT The feasibility of utilizing neural networks for the design of learning controllers is investigated. The design of both regulators and trackers for such controllers is considered. The relationship of these learning controllers to conventional adaptive controllers is investigated. This enables a learning methodology that guarantees the stability of the training process under mild conditions to be developed for these networks, with minimal information about the plant dynamics required
IEEE Transactions on Power Systems, Feb 1, 2003
This paper describes an integrated control and protection scheme for the power conditioner used b... more This paper describes an integrated control and protection scheme for the power conditioner used by the medium rating grid-connected distributed generators and storage devices. The proposed control scheme consists of two loops: a steady-state loop that achieves optimum harmonic output by selective elimination of low-order harmonics and a transient loop based on space vector methods that enhances the transient response
w. the problem or posltlon and force control f o r t h e compliant motion or zhe manipulators 1s ... more w. the problem or posltlon and force control f o r t h e compliant motion or zhe manipulators 1s consldered. The external force and the posltlon of tho end-effector a r e related by a seoond order lmpedrnce functlon. The force control problem 1s then translated llito a posltlon
Collision-free trajectory planning for robotic manipulators is investigated. The task of the mani... more Collision-free trajectory planning for robotic manipulators is investigated. The task of the manipulator is to move its end-effector from one point to another point in an environment with polyhedral obstacles. An on-line algorithm is developed based on finding the required joint angles of the manipulator, according to goals with different priorities. The highest priority is to avoid collisions, the second priority is to plan the shortest path for the end effector, and the lowest priority is to minimize the joint velocity for smooth motion. The pseudo-inverse of the Jacobian matrix is applied for inverse kinematics. When a possible collision is detected, a constrained inverse kinematic problem is solved such that the collision is avoided. This algorithm can also be applied to a time-variant environment.
Electro International, 1991, 1991
ABSTRACT Not Available
An adaptivecontrolalgorithm isproposed fora classofnonlinear systems,such as robotic manipulators... more An adaptivecontrolalgorithm isproposed fora classofnonlinear systems,such as robotic manipulators,which iscapableofimproving itsperformance in repetitive motions. When thetaskisrepeated,theerrorbetween thedesiredtrajectory and thatofthesystem is guaranteed todecrease.The designisbased on thecombinationofa direct adaptivecontrol and a learning process. This method does not requireany knowledge ofthe dynamic parameters ofthesystem. l. lntroducUon
In this paper the application of DSPs for motion control of robotic manipulators is considered. I... more In this paper the application of DSPs for motion control of robotic manipulators is considered. It is shown that DSPs are able to implement advanced control algorithms in real time, and hence, can provide enhanced motion control for robotic manipulators
2011 ASEE Annual Conference & Exposition Proceedings, Sep 4, 2020
Farzad Pourboghrat received his Ph.D. degree in Electrical Engineering from the University of Iow... more Farzad Pourboghrat received his Ph.D. degree in Electrical Engineering from the University of Iowa in 1984. He has since been with the department of Electrical and Computer Engineering (ECE) at Southern Illinois University, Carbondale (SIUC) where he is currently a Professor and director of the Embedded Control Systems (ECS) Lab. He is a senior member of IEEE. His research interests include control theory, real-time embedded control, mechatronics and distributed robotic systems.
Computers & Electrical Engineering, Jul 1, 2002
This paper presents adaptive control rules, at the dynamics level, for the nonholonomic mobile ro... more This paper presents adaptive control rules, at the dynamics level, for the nonholonomic mobile robots with unknown dynamic parameters. Adaptive controls are derived for mobile robots, using backstepping technique, for tracking of a reference trajectory and stabilization to a fixed posture. For the tracking problem, the controller guarantees the asymptotic convergence of the tracking error to zero. For stabilization, the problem is converted to an equivalent tracking problem, using a time varying error feedback, before the tracking control is applied. The designed controller ensures the asymptotic zeroing of the stabilization error. The proposed control laws include a velocity/acceleration limiter that prevents the robotÕs wheels from slipping.
Farzad Pourboghrat received his Ph.D. degree in Electrical Engineering from the University of Iow... more Farzad Pourboghrat received his Ph.D. degree in Electrical Engineering from the University of Iowa in 1984. He has since been with the department of Electrical and Computer Engineering (ECE) at Southern Illinois University, Carbondale (SIUC) where he is currently a Professor and director of the Embedded Control Systems (ECS) Lab. He is a senior member of IEEE. His research interests include control theory, real-time embedded control, mechatronics and distributed robotic systems.
Journal of Intelligent and Robotic Systems, 1990
time, the exploration of chemical laboratory robotics can trace backward in the 1980s. [2] The ro... more time, the exploration of chemical laboratory robotics can trace backward in the 1980s. [2] The robotics technology at that time lacks broadening intelligence and only can process limited tasks. Today, motived by the breakthrough of the AI and intelligent robots, it is right time emerging to revisit and transfer these day's advanced technology to the fundamental research areas. [3] The preciseness of computer language and robot operation benefits a lot on solving the reproducibility problem, [4] which has been proved both in molecular synthesis, [5] 2D materials assembling, [6] and nanocrystal growth. [7] Recently, the robotics technology starts to apply in an automatic chemical laboratory vigorously widely. [8] Big chemistry data and AI algorithms endow robots with the ability to search reaction paths and optimize experimental parameters by itself, called autonomous discovery. [9] However, these scaling-out methods by exploring ample parameter space are more focusing on working efficiency, less on understanding and development of scientific knowledge. [10] The future materials discovery needs not only the robot-assisted workforce but upgrade of methodology to make synthesis really "predictable." [11] The "Dial a Molecule" program in the UK and the DARPA project "Make it" in the USA are typical ongoing cases for "On-Demand" molecule discovery. [12] The discovery process should start from materials design, synthesis planning, automatic experiments, and characterization, to parameter self-adjusting autonomously to achieve a closing-loop process. [13] It needs an intelligent brain that well coupled neuron networks, physical-chemistry theory and scientists-machine interactions. [14] Since yet, most of the algorithms-assisted synthesis system was utilized for optimizing organic molecular reaction, for example, the SNOBFIT-assisted robotic platform in Massachu-setts Institute of Technology (MIT), [15] and the cloud synthesis system based on Complex Method in Cambridge. [16] For higher demand, brains of these self-driving laboratories are expected to extract the common concepts in experimental science, having wider compatibility, and more expansion interface. Here, we report our intelligent robotics system Materials Acceleration Operation System (MAOS), for "On-Demand" materials synthesis and scientific discovery. MAOS is of the E-commerce typical like eBay or Alibaba, the user ordered the materials online, and then MAOS deliver it. A virtual lab was developed to provide a safe and practical way for man-machine interaction to train MAOS for the new experimental operations. Collaborative A Materials Acceleration Operation System (MAOS) is designed, with unique language and compiler architecture. MAOS integrates with virtual reality (VR), collaborative robots, and a reinforcement learning (RL) scheme for autonomous materials synthesis, properties investigations, and selfoptimized quality assurance. After training through VR, MAOS can work independently for labor and intensively reduces the time cost. Under the RL framework, MAOS also inspires the improved nucleation theory, and feedback for the optimal strategy, which can satisfy the demand on both of the CdSe quantum dots (QDs) emission wavelength and size distribution quality. Moreover, it can work well for extensive coverages of inorganic nanomaterials. MAOS frees the experimental researchers out of the tedious labor as well as the extensive exploration of optimal reaction conditions. This work provides a walking example for the "On-Demand" materials synthesis system, and demonstrates how artificial intelligence technology can reshape traditional materials science research in the future.
Consider the single-input single-output delay system with a single delay X(t) = A(d) X(t) + B u(t... more Consider the single-input single-output delay system with a single delay X(t) = A(d) X(t) + B u(t) (I) y(t) = C X(t) where X(t) = ¿(t), t¿[-h,0] A(d) = Ao + A1d, dX(t) = X(t-h) X(t)¿Rn, u(t)¿R1 and y(t)¿R1. The problem is to find the state variable vector X(t) from the output variable y and the control input u. Unfortunately,
International Journal of Control, Feb 1, 1989
The parameter identification of a class of multiple-input multiple-output continuous linear delay... more The parameter identification of a class of multiple-input multiple-output continuous linear delay systems is considered. Only the input and output variables of the system are used for identification. The identification process is carried out through an intermediate pseudo-equivalent system. A linear modulating operator is introduced to eliminate the problems of input and output differentiation and initial values. It is shown
International Journal of Control, Sep 1, 1986
... FARZAD POURBOGHRATt and DONG HAK CHYUNGt ... Some methods have been proposed as an extension ... more ... FARZAD POURBOGHRATt and DONG HAK CHYUNGt ... Some methods have been proposed as an extension of the Kalman filter, which approximately estimates the state variables of the delay system (Britt and Luecke 1978, Liang 1978). ...
2007 7th International Conference on Power Electronics and Drive Systems, 2007
ABSTRACT This paper proposes an online technique for the H∞ control of speed and flux norm of cur... more ABSTRACT This paper proposes an online technique for the H∞ control of speed and flux norm of current-fed induction motors (IM). Integrals of the speed and flux norm tracking errors are considered as elements of the state variables. A robust H∞, optimal control strategy for this problem is determined by solving the algebraic Riccati equation (ARE), using multilayer recurrent neural networks. The proposed controller allows for the simultaneous and independent control of both speed and flux norms of the induction motors. The control implementation involves estimating the rotor flux, rotor resistance, and speed of the induction motor, using continuous-time extended Kalman filter (EKF). The simulation results show the effectiveness of the proposed controller even in the presence of rotor resistance and load torque disturbances.
Summary form only given, as follows. A feedforward neural network was used to solve the problem o... more Summary form only given, as follows. A feedforward neural network was used to solve the problem of inverse kinematics for the redundant robots. A learning algorithm was also developed for the training of the network. The convergence of the training process was guaranteed according to Liapunov's stability theory. Moreover, the speed of training can be increased by increasing a learning
Two neural learning controller designs for manipulators are considered. The first design is based... more Two neural learning controller designs for manipulators are considered. The first design is based on a neural inverse-dynamics system. The second is the combination of the first one with a neural adaptive state feedback system. Both types of controllers enable the manipulator to perform any given task very well after a period of training and to do other untrained tasks satisfactorily. The second design also enables the manipulator to compensate for unpredictable perturbations.
IEEE Transactions on Control Systems Technology, 2006
In this paper, a novel real-time optimal control is designed for a class of embedded flexible str... more In this paper, a novel real-time optimal control is designed for a class of embedded flexible structures such as a cantilever beam. The proposed strategy is based on a new recursive subspace identification and receeding horizon optimal control. The proposed recursive strategy can be parameterized in terms of recursive approximation of subspace intersections and adaptive estimation of state sequences. The
Uploads
Papers by Farzad Pourboghrat