SAURABH PANDEY
@iiitkota.ac.in
Assistant Professor, Department of Electronics and Communication Engineering
Indian Institute of Information Technology Kota, MNIT Campus, Jaipur, Rajasthan, India
EDUCATION
Postdoctoral Fellow, Dalian University of Technology, Dalian, P. R. China (November 2018 - October 2020)
Doctor of Philosophy from Indian Institute of Technology Guwahati, Assam, India in 2018
Bachelor of Technology from Gautam Buddh Technical University Lucknow, Uttar Pradesh, India in 2011
RESEARCH INTERESTS
System Identification, Relay Control Systems, Adaptive Parameter Estimation and Control, Data Driven Control, Non-linear Systems, Time Delay MIMO Systems, Modelling of Lithium-Ion Battery and SOC Estimation for EVs, HEVs, FCEVs Applications, Power Electronics and Process Control Applications.
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Prasenjit Ghorai, Somanath Majhi, Azim Eskandarian, Saurabh Pandey, and Venkata Ramana Kasi
Springer Science and Business Media LLC
Saurabh Pandey and Somanath Majhi
IEEE
Using the frequency domain approach, a relay based modeling of industrial processes in terms of dead time models with non-minimum phase and integrating dynamics is addressed in this paper. An asymmetrical relay is fedback to bring sustained oscillatory responses from the unknown process dynamics broadly known as limit cycle. Considering the relay settings and the limit cycle information, an explicit set of mathematical expressions for identification of non-minimum phase integrating first order plus dead time, integrating first order plus dead time and pure integrating plus dead time processes is deduced which does not need an appropriate guess in the solution of nonlinear equations to yield process model parameters. During the identification test, effect of measurement noise over the process output information is minimized by using a Fourier series based curve fitting algorithm. Wellknown examples from literature are simulated to illustrate the advantage of proposed relay based identification algorithm. Finally, the comparison between identified and true process models is carried out through integral of absolute error criterion and frequency response plots.
Saurabh Pandey, Somanath Majhi, and Prasenjit Ghorai
Informa UK Limited
ABSTRACT In this paper, the conventional relay feedback test has been modified for modelling and identification of a class of real-time dynamical systems in terms of linear transfer function models with time-delay. An ideal relay and unknown systems are connected through a negative feedback loop to bring the sustained oscillatory output around the non-zero setpoint. Thereafter, the obtained limit cycle information is substituted in the derived mathematical equations for accurate identification of unknown plants in terms of overdamped, underdamped, critically damped second-order plus dead time and stable first-order plus dead time transfer function models. Typical examples from the literature are included for the validation of the proposed identification scheme through computer simulations. Subsequently, the comparisons between estimated model and true system are drawn through integral absolute error criterion and frequency response plots. Finally, the obtained output responses through simulations are verified experimentally on real-time liquid level control system using Yokogawa Distributed Control System CENTUM CS3000 set up.
S. Pandey and S. Majhi
Institution of Engineering and Technology (IET)
A set of explicit expressions for an accurate identification of unstable first order plus time delay (FOPTD) process using single relay feedback test are deduced. Thereafter, tuning rules for a feedforward and feedback combination of proportional-integral and proportional controllers are proposed to enhance the closed loop transient performance. When the process output is distorted by measurement noise, a Fourier series based curve fitting technique is utilised for the reconstruction of true process output. Benchmark examples from literature are illustrated and thereafter comparisons are made to validate the effectiveness of the proposed identification and control scheme.
Prasenjit Ghorai, Saurabh Pandey, and Somanath Majhi
IEEE
The paper presents a time domain based modelling and identification of real-time control system in terms of stable dynamics of first order plus dead time (FOPDT). An asymmetrical relay is commissioned for yielding limit cycle output from unknown plants. Thereafter, the limit cycle information is utilized in equations for real-time plant dynamics to be identified. Mathematical expressions based on the solution of state equation are utilized for bringing the accuracy in the identification of real plant dynamics. Computer Simulations are conducted on well known examples from literature to demonstrate the efficiency of the proposed scheme. Yokogawa Distributed Control System (DCS) with CENTUM CS3000 is considered here as an experimental platform for the realization of the relay feedback test and estimation of an unknown level control plant model followed by comparison through Nyquist plots with Zeigler Nichols method.
Prasenjit Ghorai, Somanath Majhi, and Saurabh Pandey
ASME International
The paper presents a real-time system modeling and identification scheme for estimation of plant model parameters using a single asymmetrical relay test. A modified set of analytical expressions for unknown plant models under nonzero setpoint and non-negative relay settings is derived. Thereafter, the unknown parameters of three different stable plant models are identified as first-order plus dead time, overdamped, and critically damped second-order plus dead time. The well-known examples from literature are included to show the accuracy of the proposed method through computer simulations. Yokogawa distributed control system centum CS3000 is considered as a design platform for an experimental setup for the realization of asymmetrical relay feedback test. Finally, the transfer function models derived from successive identification of plant dynamics are compared with the literature through Nyquist plots.
Saurabh Pandey and Somanath Majhi
Informa UK Limited
ABSTRACT A set of novel explicit expressions for the identification of stable, unstable and integrating first-order plus dead time process dynamics is presented. In the absence of sensor noise/static load disturbances, an autonomous relay control system with an asymmetrical relay induces a smooth limit cycle at the process output. Dynamic model parameters are estimated using the set of proposed expressions which depend on the parameters of limit cycle output and its derivatives. However, in practice, the process output is generally corrupted by the measurement noise, thereby rendering an erroneous identification of process dynamics. Furthermore, static load disturbance during an identification test also induces an asymmetrical limit cycle output resulting in inaccurate measurements. Hence, a fast Fourier transform technique and biased relay feedback methods are implemented to obviate the problem of asymmetries and chattering in the limit cycle output yielding the original limit cycle and its subsequent derivatives. The proposed method has been validated, considering five typical examples from literature. An extensive comparison study with existing approaches based on Nyquist plots demonstrate the efficacy of the method presented.
Prasenjit Ghorai, Saurabh Pandey, and Somanath Majhi
IEEE
In this paper, attempt is made to explore real time plant modelling based on asymmetrical relay feedback test with step input as an excitation signal. The simulated and experimental investigations result in a set of modified analytical expressions for identification of stable first order plus dead time, non-minimum phase first order plus dead time and over-damped non-minimum phase second order plus dead time processes. These expressions are further utilized for identification of processes in presence of measurement noise using curve fitting tool. Well known examples from literature are included for validation of proposed modelling and identification scheme. Yokogawa distributed control system centum CS3000 with field instruments are utilized for conducting real time experiment on level control system.
Saurabh Pandey and Somanath Majhi
IEEE
General equations for the identification of stable and unstable second order plus dead time (SOPDT) processes are proposed in this paper. During feedback test, a nonlinear device (relay) is connected to process with a phase lag of n radians yielding sustained oscillations known as limit cycle. With the help of describing function (DF) approximation, equivalent gain of an asymmetrical relay is derived and thereafter used in the derivation of mathematical equations for the identification of process dynamics. When a process output is subjected to measurement noise, relay output shows spurious switching which brings chattering in the limit cycle output. In order to retrieve a clean limit cycle output, Fourier series based curve fitting method is implemented. Using frequency response plots, the efficacy of the proposed methodology is drawn through comparisons between actual and estimated process models.
Saurabh Pandey and Somanath Majhi
IEEE
In this paper, general expressions for the identification of various processes in terms of a stable and unstable first order plus dead time (FOPDT) models are derived. During the identification test, a nonlinear device (relay) is connected to an unknown process through a feedback path which yields sustained oscillations known as limit cycle. The equivalent gain of an asymmetrical relay is approximated by describing function (DF) approach and further used in the deduction of mathematical expressions for process identification. When the process output is subjected to measurement noise, multiple switching at the relay output brings chattering at the output. Therefore, a Fourier series based curve fitting method is utilized in order to retrieve a clean limit cycle output. Further, to show the supremacy of the proposed method comparisons between estimated and actual process models are drawn using frequency response plots.
Raghunath Bajarangbali, Somanath Majhi, and Saurabh Pandey
Elsevier BV
Saurabh Pandey and Somanath Majhi
Elsevier BV
Abstract In this paper, a set of novel expressions for the identification of stable and unstable first order plus time delay (FOPTD) processes are derived. Using feedback compensation scheme a biased relay is connected to process which yields limit cycle output. Subsequently, from a limit cycle and its first derivative information explicit expressions for the process model parameters are developed. When the process output is subjected to measurement noise, chattering in relay output becomes the source of noisy limit cycle. The consequences of measurement noise are suppressed by discrete cosine transform (DCT) with a proper threshold parameter selection for the reconstruction of original limit cycle. Two well known examples are included to illustrate the effectiveness of the proposed method. Also, a comparison is drawn between estimated and actual system using frequency response plots.
Saurabh Pandey and Somanath Majhi
IEEE
Using an asymmetrical relay based approach, a set of generalized expressions for the identification of integrating first order plus dead time (IFOPDT) and second order plus dead time (ISOPDT) processes are derived. From derived expressions, calculation for unknown process model parameters with small and large dead time is performed. When the process output is subjected to measurement noise, a well known method is attempted in order to extract the important information from the limit cycle output. Examples are illustrated to show the effectiveness of method which successfully recovers the limit cycle under noisy environment. Subsequently, from denoised limit cycle, identified parameters are used for estimation of process model dynamics. Finally, a comparison between identified and actual process is demonstrated through simulation by Nyquist plot.
Saurabh Pandey and Somanath Majhi
IEEE
A technique for identification of system dynamics under the influence of measurement noise is presented. When output signal from sensor is corrupted by noise, identification of system parameters become an arduous task. Therefore, denoising methods can be employed in order to achieve a clean limit cycle under noisy environment. A Fast Fourier Transform (FFT) method has been reported as an effective technique for the mitigation of noise by an appropriate thresholding. Generalized explicit expressions are examined to obtain unknown plant model parameters for stable and unstable First Order Plus Dead Time (FOPDT) processes. Examples are illustrated to show the effectiveness of method during parametric identification of process models. Subsequently, a comparison is drawn between estimated and actual system through simulation by step test and Nyquist plot.
RECENT SCHOLAR PUBLICATIONS
MOST CITED SCHOLAR PUBLICATIONS
Publications
Saurabh Pandey, Tao Liu, Qing-Guo Wang, “Parametric identification of sampled systems with time delay subject to load disturbance with unknown dynamics”, IET Control Theory & Applications (Wiley), vol. 15, issue 15, pp. 1942-1955, 2021. DOI: . [SCI].