@vsbec.com
Department of Electrical and Electronics Engineering
VSB Engineering College, Karur, Tamilnadu
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
S. Meenakshi, G. Prabhakar, N. Ayyanar, and P. Nedumal Pugazhenthi
IEEE
Seaweeds are large algae that grow on rocky shorelines, in shallow coastal waters, and in marine environments. It plays a major role in marine ecosystems and is abundantly available in the Gulf of Mannar Biosphere region. However, some species of seaweed clash with coral reefs and damage them severely due to the release of hydrophobic allelochemicals. Conventional methods are often labor-intensive, which results in adverse environmental effects. The solution to this problem is to involve underwater vehicles for seaweed farming. But the existing underwater vehicles need some soft materials to handle the targeted seaweeds carefully, without affecting the coral reefs and other species in the sea.This study aims to develop a soft robotic gripper with pneumatic to grasp seaweed, which is abundantly available under the sea water nearer to coral reefs. The novel soft pneumatic robotic gripper design has been done for 8 chambers using Ansys Finite Element Analysis software. Gripper deformation analysis is carried out for different pressure levels to understand the behavior of the hyperelastic material.
M Monica Dhana Ranjini, M Paul Jeyaraj, M Senthil Kumar, T Arun Prasath, and G Prabhakar
IEEE
Face recognition is one of the most promising applications of image analysis that has gained a lot of attention recently. One may say that a sizable amount of facial recognition tasks involves face detection. Depending on how powerful it is, it can concentrate processing resources on the area of a picture that contains a face. Human faces vary in attitude, expression, location, orientation, and skin tone, making it difficult to identify faces in pictures. spectacles wearers or people with facial hair, camera gain, illumination, and image resolution. The main objective of face detection algorithms is to determine whether a face is visible in an image. The sophistication and accuracy of face detection and recognition have recently been the subject of numerous studies, but the Real-Time Face Detector created by the Viola-Jones Algorithm, which can accurately recognize faces in real-time, has caused a revolution in this area. Numerous detecting problems, including object, face, emotion, and face identification, have been successfully solved using neural networks. It is efficient to recognize objects using cascade classifiers based on Haar features. Face, eye, and smile detection using the Haar cascade classifier and facial recognition using OpenCV will both be implemented in this study utilizing the LBPH algorithm. If the individual is known, the accuracy of the image is predicted as well as face, eye, and grin detection using bounding boxes in the camera stream. The results are then displayed according to whether the person is known or one in the dataset.
M. Ramachandran, G. Prabhakar, V. Nirmal Kannan, and A. Mariya Chithra Mary
IEEE
In recent years, the need for electricity for each home is rapidly increasing due to the number of inventions in electric appliances and electric vehicles being so high. The fast-growing gated community apartments and villas want their own microgrids to meet the needs of their own residents. In Electric Vehicles, Bi-directional power flow is made feasible by the power electronics used in the rechargeable batteries. Therefore, electric vehicles are considered as grid’s power bank Storing excess power in electric vehicle (EV) batteries through Grid-To-Vehicle (G2V) technology and returning the energy back to the grid through Vehicle-To-Grid (V2G) technology while demand is at its maximum can aid in the management of micro-grid energy. This research review analysis the architectural framework and control mechanisms to develop hierarchical energy management for energy sharing among V2G and G2V in terms of simulation. It uses fast EV charging in a microgrid. A dc speedy charging point is designed as part of a trial micro-grid system for connecting EVs. V2G-G2V power transfer, DC Bus voltage, Battery SOC and Battery Voltage and Current is demonstrated through simulation.
S. Dharsini, G. Prabhakar, S. Rajaram, and J. Shanthi
IEEE
A Vehicle-following system is a system that controls a following vehicle to maintain a fixed distance and acceleration behind a preceding vehicle. Cyber attack is an issue on Vehicle following systems due to its wireless connectivity. Cyber dangers to physical systems result from this. As a result, controlling the action of the vehicle is necessary to stop the cyber attack, which causes the development of cyber physical system(CPS). CPS can be accessed and controlled remotely. Their features make them more easily vulnerable to cyber attacks. Fault Data Injection (FDI) attack is an attack that is mathematically fed into the altered Vehicle Following System and modeled to study the attack's effects. The Scope of this work is to develop a machine learning model by generating the cyberattack dataset by analyzing the Physical Model of Vehicle following system. Decision tree algorithm is used to detect the attack. It considerably raises FDI detection accuracy.
Iyappan Murugesan, Prabhakar Gunasekaran, Suresh Muthusamy, and Ponarun Ramamoorthi
Informa UK Limited
ABSTRACT Through transmission lines (TLs), an electric power transmission system has been able to transmit power from generating stations to consumers. During transmission, various kinds of malfunctions take place and they are termed as a fault. Although fault is undesirable, it is unavoidable event hampering the smooth functioning of the power system. In power transmission systems, and a large number of voltage and current signal, distortions take place due to faults. Faults occur in power TL causing power supply interruption. Several fault detection techniques have been presented by researchers to detect a fault in TL. However, the time required to locate the fault remained higher and the power loss rate (PLR) was not reduced. To overcome these issues and identify faults in electrical power TL, Haar wavelet feature extraction-based firefly optimized fault detection (HWFE-FFOFD) method has been introduced. The power TL signal sample has been taken as an input. Zero-mean normalization is the pre-processing approach that converts the transmission signal sample into a specified range. To extort features (i.e. voltages and current values) with higher accuracy, the normalized signal has been given to Haar Wavelet Transform. Then, the extracted features at different time instants have been given to the firefly optimized fault detection (FFOFD) algorithm. In the FFOFD algorithm, extracted features have been considered as firefly populations. The FFOFD algorithm functions with the flashing behavior of a firefly. At last, the firefly position has been updated and ranked according to light intensity to detect a fault in electrical power TL. In this manner, the fault detection time (FDT) gets reduced using HWFE-FFOFD method. HWFE-FFOFD method is evaluated in FEA, FDT, and PLR. From the experimental results obtained, it can be confirmed that the HWFE-FFOFD method has been able to enhance accuracy by 14% and minimize time by 26% and PLR by 62% when compared to conventional methods.
B. Koushik Aryan, O. Sobhana, G.C. Prabhakar, and N. Amarnadh Reddy
IEEE
This paper presents a method for detecting faults in a micro grid using Artificial intelligence (AI). As we know fault detection is very important for microgrids and by using AI we can make the microgrid as smart grid as it can detect the faults itself by using artificial intelligence methods. Here the microgrid model consists of a diesel generator, a solar photovoltaic system, and a wind generator. Grid integration of solar and wind energy sources with diesel generation system can have faults which need to be detected and cleared as soon as possible as it is combination of three generation systems and lots of parameters included. In Simulink model, the microgrid's normal functioning and fault scenarios are simulated. The fault conditions simulated represent faults encountered by a distribution line. The AI based RBF classifier is trained using voltage and current samples. To foresee the performance of algorithm classification rate is stated in terms of the statistical metrics.
M. Balamurugan, G. Prabhakar, G. Amsaveni, M. Karthikumar, J. Jasmin Shifa, and E. Sharmila
IEEE
Shopping is an essential part of human life. Nowadays, when we want to buy something from a store, it’s critical for us because we have to wait for two reasons. The first step is to inspect the purchased item. The second step is to make the payment. Nowadays, shopping is essential, and it becomes even more complicated during festival seasons. People find it difficult to shop physically because all of the stores are overcrowded, and also people have to wait in line for a long time to verify the items purchased and pay bills. This research study proposes a novel system to easily handle these complexities. IoT connects the system to the user’s mobile device and a computer available in the shop. All the details are updated in both the mobile application and server once an item is added in the cart with the price and total amount. The payment link is added once they complete shopping, so user can pay with ease and there will be no waiting time in the queue to verify the amount and the list of items purchased.
M. L. Ramamoorthy, S. Selvaperumal, and G. Prabhakar
Computers, Materials and Continua (Tech Science Press)
Iyappan M., G. Prabhu, S. Suganth, S. Vigneshwaran, and G. Prabhakar
Elsevier BV
Abstract The power management scheme has implemented in microgrid system. (IOT) Internet of Things with electronic microcontroller is used to overcome the electricity demand. Let the fastest growing of the electronic devices and that elements load need an entire electricity demand. The proper preposition method is also one of the possible solutions to control the energy demand by using an IOT (Internet of Things). This project describes the smart energy management system using IOT Based system consumer load management. Grid based applications integrated with EV application. Consumer based service implemented with grid applications integrated IOT system for analysis the demand of the system.
G. Prabhakar, S. Selvaperumal, P. Nedumal Pugazhenthi, K. Umamaheswari, and P. Elamurugan
Elsevier BV
Abstract The current work explores the modelling, control and realistic execution of nonlinear segway system in both simulation and real-time. The Segway vehicle has been considered as a physical system that is nonlinear and unstable. Equations of motions have been derived using Lagrangian dynamics by considering kinetic energy and potential energy. Then, the nonlinear differential equations have linearised using Taylor series expansion of functions to acquire a state space model of the system and convert it into two transfer functions that corresponds to tilt angle and yaw angle. These are the control parameters to stabilize the Segway system. Effective simulation analysis of nonlinear control is made up with PID, GA Tuned PID and Model Predictive controller (MPC) by MATLAB. And also, an electro-mechanical model of Segway vehicle has developed for real-time control analysis. It is autonomously sensed and actuated through embedded processor to meet the desired response. An online optimization technique has involved solving dynamic optimization problems in MPC for superior performance. This innovative realisation paves the approach for the young researchers to consider the essence of analysis in nonlinear applications.
Shyamala Balakumar, Selvaperumal Sundaramoorthy, Ramasubramanian Bhoopalan, and G. Prabhakar
Springer Science and Business Media LLC
Detection of moving object from a visual sequence plays a vital role for the tracking of object. The main objective of this proposed work is to detect and classify the various video sequences with the help of different classification algorithms. The input video sequences from the publicly available datasets are collected and the individual frames are extracted. These frames are pre-processed and then applied to the novel background subtraction process. Important features based on the Local Binary Pattern (LBP) and grey level co-efficient are extracted. Finally these features are classified by three different classifiers like SVM, PLS, and PNN. The performance of these different classifiers are evaluated and compared. It is found that PLS classifier produces more classification accuracy but with more computation time.
Prabhakar Gunasekaran, Selvaperumal Sundaramoorthy, and Nedumal Pugazhenthi Pulikesi
Institution of Engineering and Technology (IET)
Cyber defence mechanism is started with modelling the accurate car-following behaviour including cyber attack. The creation of finest models made the path of control action easier. The connection between the vehicles is mathematically formulated with the help of car-following behaviour, incorporating the derived acceleration function from the cruise control physical system. The modified car-following model is simulated as closed-loop control system to analyse its behaviour in terms of acceleration and distance. Fault data injection cyber attack is mathematically injected into the modified car-following model and simulated to analyse the impact of attack. Initially, the impact of fault data injection attack is detected and mitigated with the help of parallel proportional-integral-derivative controller and genetic algorithm tuned proportional-integral-derivative controller. Interval type-2 fuzzy proportional-integral-derivative controller is introduced to mitigate the cyber attack and to overcome the uncertainty. The integral square error and integral absolute error are used to compare the performance of the controllers. Inbuilt Wi-Fi connected car like mobile robots are used in real-time model. This model is designed and developed based on the Node MCU processors, real-time operating system, sensors and actuators.
G. Prabhakar, S. Selvaperumal, and P. Nedumal Pugazhenthi
Informa UK Limited
ABSTRACT An effort is made to design the fuzzy proportional-derivative (PD) plus I controller for a nonlinear cruise control system in automobiles, which provides adaptive capability in set-point tracking performance. A cruise control system has been considered as a nonlinear first order plus delay time model to exhibit the control behaviour involved in both conventional proportional-integral-derivative (PID) control and fuzzy control. The paper demonstrates the design of fuzzy PD plus I controller including comparative investigation with control structures like PID, I – PD, and PI – D using Simulink modelling. Real-time implementation has been carried out on a robotic prototype using Arduino UNO micro-controller. Based on the performance measures such as integral absolute error (IAE) and integral squared error (ISE), the proposed fuzzy PD plus I structure shows superior performance on servo and regulatory problems in the cruise control system. This control structure avoids integral windup issues and to suppress the derivative kick in the cruise control system.
K. Annaraja, S. S. Sundaram, S. Selvaperumal, and G. Prabhakar
Bentham Science Publishers Ltd.
Background: A novel system for the usage of Maximum Power Point Tracking of an expansive Solar Photo Voltaic (SPV) farm subjected to conceivable incomplete shading is displayed in this paper. The SPV farm being spread over an expansive territory a remote sensor organize is utilized for checking the sun based protection in the region of each board. The motivation behind the remote sensor organize is to screen the sunlight based protection at various areas near each of the PV board from the tremendous region of the photograph voltaic homestead comprising of countless voltaic boards. The observed protection information is utilized by a prepared. Artificial Neural Network to locate the ideal DC terminal voltage to be kept up over the general DC terminals of the photograph voltaic ranch. All the PV boards are associated in arrangement association with the fundamental bye pass diodes. The DC control accessible at the yield terminals of the SPV cultivate is first DC to DC changed over with a Positive Output Luo Converter (POLC) and bolstered to a heap. A MATLAB Simulink based reproduction was created to approve the proposed system. Methods: Maximum Power Point Tracking based on Artificial Neural Network through wireless sensor networks. Results: As the result of the proposed idea and its implementation in MATLAB we have two sets of results. In either case the input is a vector of 40 elements and the output of the first segment of the work is the estimation of the threshold PV terminal voltage that will guarantees maximum power point operation. In the first case we have the MATLAB SIMULINK implementation of the basic configuration of the forty PV panels arranged in series connection and we have provided a facility to edit the solar insulation levels pertaining to the individual PV panels. In this first configuration we have set a continuously variable PV current for all the panels and the PV current for all the panel are the same. Using this setup, for any combination of solar insulation pattern of the forty panels the overall PV curve and the overall VI curve can be drawn in MATLAB. As the simulation runs the PV current is changed from 0 to the maximum or the short circuit current level in a slowly rising manner implemented using a ramp signal. </P><P> During this period the total power output and the terminal voltage of the PV farm are sent to the work space and the data is thus collected in the workspace of MATLAB. Using basic MATLAB commands the maximum power output and the PV terminal voltage corresponding to the maximum power output are obtained. The PV current at maximum power output condition, the corresponding PV farm terminal voltage, the maximum power output recorded at this condition all correspond to the present insulation vector condition. This way, by changing the elements of the insulation for all the forty panels in a random manner we obtain for each case the Ipmax[i], Pmax[i], Vpmax[i] and this corresponds to insulation[n,i]. Where n is the number of panels, in this case 40 and i the ith experiment. In each experiment the solar insulation level of all the forty panels can be changed and the parameters Vpmax[i], Ipmax[i] and Pmax[i] can be obtained. The value of the harvested power as found from the characteristics for any given set of insulation is denoted as the estimated power. The value of power as obtained from the proposed ANN SMC POLC combination is denoted as the Actual Power. Conclusion: A wireless network based insulation monitoring has been done. An ANN based MPPT algorithm has been developed that gives the reference MPP voltage. The sliding mode control scheme uses the reference voltage and produces the switching pulses for the POLC. The ANN had been trained with a number of combinations of different insulation values falling on each of the forty panels and the ANN gives the correct reference voltage for any combination of insulation levels that were not used while training. The sliding mode controller uses this reference voltage and gives the switching pulses to the POLC that harvests the maximum power output to the RL load. The proposed system has been implemented in the MATLAB SIMULINK environment and has thus been validated. The obtained results have been compared against the maximum power output values that could be derived from the characteristic curves obtained for the given combination of insulation levels. The proposed system gives results very close to the values obtained from the characteristics. As a future work the proposed idea can be validated using hardware based experimental setup.
A. Shyamala, S. Selvaperumal, and G. Prabhakar
Bentham Science Publishers Ltd.
Background: Moving object detection in dynamic environment video is more complex than the static environment videos. In this paper, moving objects in video sequences are detected and segmented using feature extraction based Adaptive Neuro-Fuzzy Inference System (ANFIS) classifier approach. The proposed moving object detection methodology is tested on different video sequences in both indoor and outdoor environments. Methods: This proposed methodology consists of background subtraction and classification modules. The absolute difference image is constructed in background subtraction module. The features are extracted from this difference image and these extracted features are trained and classified using ANFIS classification module. Results: The proposed moving object detection methodology is analyzed in terms of Accuracy, Recall, Average Accuracy, Precision and F-measure. The proposed moving object segmentation methodology is executed on different Central Processing Unit (CPU) processor as 1.8 GHz and 2.4 GHz for evaluating the performance during moving object segmentation. At present, some moving object detection systems used 1.8 GHz CPU processor. Recently, many systems for moving object detection are using 2.4 GHz CPU processor. Hence, CPU processors 1.8 GHz and 2.4 GHz are used in this paper for detecting the moving objects in video sequences. Table 1 shows the performance evaluation of proposed moving object detection on CPU processor 1.8 GHz (100 sequence). Table 2 shows the performance evaluation of proposed moving object detection on CPU processor 2.8 GHz (100 sequence). The average moving object detection time on CPU processor 1.8 GHz for fountain sequence is 62.5 seconds, for airport sequence is 64.7 seconds, for meeting room sequence is 71.6 seconds and for Lobby sequence is 73.5 seconds, respectively, as depicted in Table 3. The average elapsed time for moving object detection on 100 sequences is 68.07 seconds. The average moving object detection time on CPU processor 2.4 GHz for fountain sequence is 56.5 seconds, for airport sequence is 54.7 seconds, for meeting room sequence is 65.8 seconds and for Lobby sequence is 67.5 seconds, respectively, as depicted in Table 4. The average elapsed time for moving object detection on 100 sequences is 61.12 seconds. It is very clear from Table 3 and Table 4; the moving object detection time is reduced when the frequency of the CPU processor increases. Conclusion: In this paper, moving object is detected and segmented using ANFIS classifier. The proposed method initially segments the background image and then features are extracted from the threshold image. These features are trained and classified using ANFIS classification method. The proposed moving object detection method is tested on different video sequences which are obtained from different indoor and outdoor environments. The performance of the proposed moving object detection and segmentation methodology is analyzed in terms of Accuracy, Recall, Average Accuracy, Precision and F-measure.
P. Annapandi, S.P. Rajaram, and G. Prabhakar
Bentham Science Publishers Ltd.
P. V. Mani Maalini, G. Prabhakar, and S. Selvaperumal
IEEE
The ball and beam is a basic reference point system with high nonlinearity and unstable system in its dynamics. Many simple and present day control methods have been used to balance the ball and beam system. The goal of this project is to model and control the ball and beam system. Here considering the beam angle of servo motor and designing controllers to control the ball position. Lagrange approach is used to find the ball position of the system. It is based on energy balance of the system. Based on the transfer function and state space model, open loop system and closed loop system are designed. The system is designed by using two Degrees-of-Freedom. The nonlinear characteristic of the second order system is regulated by using PID controller. The controller controls the ball position in moving the beam using the motor and beaten the disturbances. The parameters of the PID are tuned using PID tuning Algorithm. In order to analyse the accomplishment of PID to learn the effect of simplifying expectation, two control methods are designed and implemented using Proportional Derivative Integral (PID) as non-model based control method, Proportional Derivative and Proportional integral combination of model based and non-model based control methods.
S. Muthuviswadharani, G. Prabhakar, and S. Selvaperumal
IEEE
In Automation era, a real time system needs an effective sensor design. Sensors act as a feedback unit of the control system, which provides measurable input from the environment. Individual sensors in the real time system are affected by the noisy environment which results in the production of messy output. So we go for soft sensor design. In autonomous vehicle, there is a need of gyro sensor, distance measuring sensor, and encoders to attain the desired target. The proposed project is to design the proximity sensor and Micro Electro Mechanical System motion sensor (Three axis digital output gyroscope) in Simulink by analyzing the real time datasheet of the corresponding sensors. These two sensors models are taken as systems in the form of transfer function. Then the outputs of these sensors are analyzed with the presence of noise and the signal is separated from the noise using FFT algorithm.
K. Dhivya, R. Nagarajan, and G. Prabhakar
IEEE
The scope of the project is to model the parallel parking system of nonholonomic vehicles through Embedded Processors. Nonholonomic vehicles are designed by analyzing the nonholonomic constraints and control by using PI controller to achieve the closed loop response. Then the parallel parking of nonholonomic vehicles are established using path planning algorithm. The path planning algorithm is developed based on the data acquired from the sensor. The project clearly explains the Mathematical Modeling of non-holonomic vehicles. There are three important phases involved in order to park the non-holonomic vehicles securely. The first phase is the scanning phase where the vehicle scans the parking atmosphere by the ultrasonic distance sensors. The second phase, which is the positioning phase, will be implemented, where the vehicle will moves forward and backward. The purpose is to adjust the suitable distance from the start point to turn point before continue to maneuvering phase. The parallel parking of nonholonomic vehicles with non-holonomic constraints are implemented through MATLAB programming.