Research Portfolio

Brushless Direct Current (BLDC) motors are extensively used in modern applications such as electric vehicles, robotics, and industrial automation due to their high efficiency, precision control, and low maintenance. However, achieving consistent and accurate speed regulation under varying operating conditions remains a challenge with traditional fixed-gain PID controllers. This study proposes an intelligent control framework that integrates Adaptive PID (APID) and Artificial Neural Network (ANN) based control techniques for BLDC motor speed regulation. The APID controller employs a hybrid strategy combining Model Reference Adaptive Control (MRAC) and Self-Tuning Control (STC) to dynamically tune the controller gains based on real-time system parameters and reference tracking. Additionally, an ANN-based controller is implemented to learn and predict optimal control actions based on motor dynamics. A comparative simulation study using MATLAB/Simulink evaluates the performance of four control strategies: traditional PID, adaptive PID, ANN-based control, and a baseline without PID. Key performance metrics were analyzed. The results demonstrate that the APID controller significantly outperforms the traditional PID, achieving the fastest settling time and improved system stability under dynamic load variations. The ANN controller also delivers superior performance compared to conventional PID, but slightly lags behind APID in responsiveness. These findings suggest that adaptive control and intelligent learning-based strategies offer robust, efficient, and scalable solutions for advanced BLDC motor control systems.

Electric vehicles (EVs) are becoming increasingly popular, due to their beneficial environmental effects and low operating costs. However, one of the main challenges with EVs is their short battery life. This study presents a comprehensive approach for predicting the Remaining Useful Life (RUL) of Nickel Manganese Cobalt-Lithium Cobalt Oxide (NMC-LCO) batteries. This research utilizes a dataset derived from the Hawaii Natural Energy Institute, encompassing 14 individual batteries subjected to over 1000 cycles under controlled conditions. A multi-step methodology is adopted, starting with data collection and preprocessing, followed by feature selection and outlier elimination. Machine learning models, including XGBoost, BaggingRegressor, LightGBM, CatBoost, and ExtraTreesRegressor, are employed to develop the RUL prediction model. Feature importance analysis aids in identifying critical parameters influencing battery health and lifespan. Statistical evaluations reveal no missing or duplicate data, and outlier removal enhances model accuracy. Notably, XGBoost emerged as the most effective algorithm, providing near-perfect predictions. This research underscores the significance of RUL prediction for enhancing battery lifecycle management, particularly in applications like electric vehicles, ensuring optimal resource utilization, cost efficiency, and environmental sustainability.

Brushless DC (BLDC) motors have a wide range of applications in these modern days, such as electric vehicles, industrial robots, washing machines, pumps, and blowers. The brushless DC motors have many advantages when compared to induction motors and conventional DC motors, such as better speed control, noiseless operation, high efficiency, less maintenance, and a long life. Along with these benefits, there is one major disadvantage known as cogging, which causes undesirable effects in the motor such as noises and vibrations. BLDC motors have been widely used in automation and industrial applications due to their attractive features. There are certain parameters to be considered while designing a BLDC motor, such as its dimensions, number of windings turns, type of magnetic materials used, required torque, output current, slot-to-depth ratio, efficiency, temperature rise, etc.

Multiple-level inverters are often used in extreme voltages and high-energy applications because of their excellent efficiency lately. Because of the considerable quantity of semiconductor switches involved in powers conversion, multilevel inverters (MLIs) are more likely to experience switch faults. Therefore, it is essential to find and discover defects as soon as possible. In this context, a technique depending on the level-shifted pulse width modulations (PWM) methodology is developed for diagnosing open-circuits errors in H-bridges multiple-level inverters (HMLI). Instead of utilizing an algorithm to detect faults, the proposed method uses subtle fault features and only requires a little amount of logical judgment. The loads current and H-bridges outputs voltages are utilized to diagnose faults.

One of the most useful renewable energy sources is wind, from which electrical power can be generated using a turbine system for long periods. The reliability of a wind turbine mainly depends on the maintenance work carried out at the site. The Status Surveillance System (SSS) is an important factor for wind turbines to guarantee uninterrupted power supply to the end user. However, the condition monitoring system based on the Wireless Sensor Node (WSN), housed with the current sensor node, is more vulnerable to failure due to circumstantial faults. Due to sensor faults, the data used for decision-making on maintenance are corrupted. This paper devises a robust and reliable mechanism called Sensor Failure Identification and Segregation (SFIS) to detect and detach corrupted data to effectively perform work related to wind turbine failure detection. The short-circuit fault is addressed by a wavelet transient approach to restore the corrupted data, while the invariable anomaly fault is analyzed with the help of the cross-correlation method. Hence, the interference fault can be analyzed using a dynamic time-warping approach. The proposed mechanism is compared with the existing Adaptive Neuro-Fuzzy Inference System (ANFIS) method that uses Supervisory Control and Data Acquisition (SCADA) to prove its reliability and robustness. SFIS offers a reliable and cost-effective solution for wind turbine maintenance work.

Integrating wind energy forecasting into urban city energy management systems offers significant potential for optimizing energy usage, reducing the carbon footprint, and improving overall energy efficiency. This article focuses on developing a wind power forecasting model using cutting-edge technologies to enhance urban city energy management systems. To effectively manage wind energy availability, a strategy is proposed to curtail energy consumption during periods of low wind energy availability and boost consumption during periods of high wind energy availability. For this purpose, an LSTM-based model is employed to forecast short-term wind power, leveraging a publicly available dataset. The LSTM model is trained with 27,310 instances and 10 wind energy system attributes, which were selected using the Pearson correlation feature selection method to identify crucial features. The evaluation of the LSTM-based forecasting model yields an impressive R2 score of 0.9107. The model’s performance metrics attest to its high accuracy, explaining a substantial proportion of the variance in the test data. This study not only contributes to advancing wind power forecasting, but also holds promise for sustainable urban energy management, enabling cities to make informed decisions in optimizing energy consumption and promoting a greener, more resilient future.

The characteristics of a PV (photovoltaic) module is non-linear and vary with nature. The tracking of maximum power point (MPP) at various atmospheric conditions is essential for the reliable operation of solar-integrated power generation units. This paper compares the most widely used maximum power point tracking (MPPT) techniques such as the perturb and observe method (P\&O), incremental conductance method (INC), fuzzy logic controller method (FLC), neural network (NN) model, and adaptive neuro-fuzzy inference system method (ANFIS) with the modern approach of the hybrid method (neural network + P\&O) for PV systems. The hybrid method combines the strength of the neural network and P\&O in a single framework. The PV system is composed of a PV panel, converter, MPPT unit, and load modelled using MATLAB/Simulink. These methods differ in their characteristics such as convergence speed, ease of implementation, sensors used, cost, and range of efficiencies. Based on all these, performances are evaluated. In this analysis, the drawbacks of the methods are studied, and wastage of the panel’s available output energy is observed. The hybrid technique concedes a spontaneous recovery during dynamic changes in environmental conditions. The simulation results illustrate the improvements obtained by the hybrid method in comparison to other techniques.

The cogging torque is the most significant issue in permanent magnet applications, since it has a negative impact on machine performance. In this article, the impact of magnetic materials on cogging torque is analyzed on brushless DC motors (BLDC). The effect of neodymium magnets (NdFeB), compression molded magnet, and samarium cobalt (SmCo) magnet on the cogging torque is analyzed to the BLDC motor designed for hybrid electric vehicle traction that has the peak power rating of 50 kW motor with 48 stator slots and 8 rotor poles. With the presence of these three magnetic materials, the cogging torque is estimated independently using multiposition simulation. The multiposition is simulated using a transient application that runs at constant speed. The results of cogging torque, rotational speed, angular position of BLDC motor, and magnetic flux density distribution have been presented. Also, the maximal, mean, minimal, rectified mean, and rms values of cogging torque were provided.

The maximum electrical energy conversion efficiency of the Solar PV panel is up to 22\% in normal conventional roof- top system under the temperature of 25˚C on Standard Test Condition (STC). In Indian climatic conditions, the atmospheric temperature is mostly above 35˚C to 45˚C, it incites 35˚C to 80˚C temperature on the PV panel. The black body of the PV panel absorbs more heat. This temperature affects the electrical efficiency of the panel significantly. This paper proposes the mathematical modelling of the solar PV panel for different solar irradiation and the temperature. The experimental evaluation is conducted in the latitude of 11.36 (N) and longitude 77.82 (E). The testing and monitoring was done with LabVIEW based National Instruments hardware such as NI cDAQ-9178, NI DAQ - 9227 and NI DAQ 9225. The comparative study between the simulated result and real time hardware results are discussed in this paper. The test result shows that the output of the proposed model mismatches with the experimental output of the solar PV panel due to the negative correlation between the efficiency and temperature for variable irradiation condition. It shows a power difference of 9.41W between the output of the proposed model and the experimental setup.

Flexible Alternating Current Transmission System devices have numerous applications in electrical transmission lines like improvement of voltage stability, reactive power compensation, congestion management, Available Transfer Capacity enhancement, real power loss reduction, voltage profile improvement and much more. The effectiveness of these FACTS devices is enhanced by the placement of these devices in the transmission lines. The placement is based on transmission line sensitivity factors such as Bus voltage stability index and line voltage stability index. This research article focuses on optimizing the location, number and ratings of FACTS devices using Evolutionary Algorithms like Bacterial Foraging Algorithm and Gravitational search algorithm. FACTS devices such as Static Var Compensator, Thyristor Controlled Series Capacitor and Unified Power Flow Controller are placed on IEEE 14 bus and IEEE 30 bus systems for reducing the real power loss in the transmission system. The results show that the performance of the transmission lines is enhanced more using Bacterial Foraging Algorithm than Gravitational Search Algorithm.

The separation of waste materials into degradable and non-degradable waste is one of the major issues in our country. As the plastic materials cause major harm to the humans as well as the environment it should be separated and effectively recycled. The non-degradable plastic wastes do not allow the rain water to pass through it thereby decreasing the ground water level. If these wastes are burnt, it emits carbon monoxide which is harmful to the humans. Our project is to encourage the people to dump and segregate the waste materials in their households using smart bins. The separation of the plastic waste is done by using capacitive type proximity sensor and ultrasonic sensor. In order to encourage the people to collect and segregate the waste materials, in large amount, they are rewarded for the amount of waste that they collect and segregate. The amount of waste collected is displayed in the website and in their mobile phones with the amount of reward points earned. The customers redeem their reward points in the shops mentioned in the website.

This paper proposes the voltage balance implementing multilevel inverter in renewable energy system. Multilevel inverters are used in high power and medium voltage applications. Using multilevel inverter with renewable energy alone, voltage balance cannot be made because the number of level increases in multilevel inverter the control gets complexity and therefore voltage unbalance problem is being introduced. So then we introduce a new technique where the diode clamped multilevel inverter is being combined with multilevel boost converter. Therefore, in this case voltage unbalance problem is reduced where the entire system is simulated and executed in PSIM software. By implementing this technique high conversion efficiency can be provided.

{\textless}p{\textgreater}Multilevel inverters are used in high power and medium voltage applications. Employing multilevel inverter with renewable energy alone, the voltage balance cannot be made because the number of level increases in multilevel inverter the control gets complexity. So voltage imbalance problems are introduced. The voltage imbalance problems can be classified into two types; Midpoint unbalance and the central capacitor discharge. These problems can be solved by using voltage balancing solutions. The solutions are hardware based; software based, and combined solutions. By using these types of solutions the voltage balancing problems can be solved and the efficiency of multilevel inverter could be high. This paper reviews about various voltage balancing solutions in multilevel inverter.{\textless}strong{\textgreater}{\textless}/strong{\textgreater}{\textless}/p{\textgreater}

There will be unexpected crowds in certain places, such as shopping malls, gaming venues, or other spots with seasonal traffic. We observe that many people arrive quickly without warning, which makes it difficult to manage resources, creates imbalances because of crowding, and manages parking and traffic, among other issues. IR sensors are the greatest option for counting people because they are affordable and simplify and improve the system. Controlling the entry and exit gates with servo motors is all that is required to keep an eye on the throng. The system also includes a buzzer-based exit gate control method for schools and other locations that require exit gate control. With advance knowledge about the crowd in a given region, the system keeps an eye on and manages the crowd, providing necessary individuals with information or a warning. In addition to controlling and monitoring the crowd without allowing for human interaction, the system also uses a bell to ensure that people leave the designated area. The system can be expanded to meet needs like managing student arrivals and departures from schools, keeping crowds within a set number of people in busy situations, parking lots, etc.

A Vehicle-to-Grid (V2G) integrated microgrid system, which represents a potential community-scale energy network, is modelled and simulated in this article. A diesel generator for base power generation, renewable energy sources (wind farms and photovoltaic cells), residential loads, and a V2G system with 100 electric cars make up the simulated microgrid. By controlling car battery charging and discharging, the V2G system provides grid support on a day of low demand for around 1,000 homes. While the PV farm and wind farm offer renewable energy based on real-time solar irradiance and wind profiles, the diesel generator maintains power balance by making up for variations in generation and consumption. Electric car charging and grid stability during power outages are the two functions of the V2G technology. Realistic charging patterns and energy availability are reflected in the modelling of five different vehicle-user profiles. Dynamic events like asynchronous machine starting, partial solar shading, and wind farm disconnection are all included in the 24-hour simulation. The system’s reaction to these occurrences is examined in the article, with particular attention paid to grid frequency stability, renewable power variability, and V2G’s function in load balancing and frequency control. The findings demonstrate how V2G integration may improve grid stability and lessen the erratic nature of renewable energy sources in energy systems of the future.

Municipalities can seize this fast-growing urbanization trend to build efficient transportation systems that meet climate targets and benefit urban life. The increasing adoption of EVs prompts a pressing urge to improve the sustainable framework by combining various solutions. Electric vehicles can be delivered as a service (eMaaS) to enhance public transportation in intelligent cities. A substantial business model that connects all parties and technologies missing drives back the complete rollout of eMaaS. This research intends to close this gap by introducing a data-driven digital platform that enables eMaaS in intelligent cities. By performing a thorough review of existing studies we examine eMaaS’s current status and suggest a business approach rooted in system design. Our research brings forward major difficulties like data regulation and tech integration. We also suggest a policy structure to support the transition to electric shared mobility. This investigation stresses the vital role of cutting back on private automobile use and building an interconnected digital framework for sustainable urban transport. To help cities fulfill their goals for sustainability of society and the environment priorities emphasize expanding eMaaS integration.

The inclusion of renewable energy sources into the nominal circuit of residential microgrids poses several issues due to the stochastic nature of renewable resources. This paper examines a full-scale DSM plan for a grid-integrated residential microgrid environment focusing on improved energy usage profiles, cost-efficiency, and integration of renewables. However, in contrast to the conventional load management, this approach consists of real time demand response and energy storage system, which makes the grid more flexible and reliable. One of the main results of calculations, based on data collected from living lab environments within the GSBP in Benguerir Morocco and performed in Matlab, is the range of a monthly energy saving of about 59\% coupled with a monthly use of renewable energy of about 23\%. The study goes further in explaining a more generalized application of AI predictive models to demand response and non-storage techniques for reliability. Overall, the results suggest that it is still possible to gain additional levels of energy savings and grid stability – proving that such an approach can be considered as highly scalable and more universally applicable to other residential and urban microgrids. Future work will analyse how cybersecurity measures can be implemented and how the system can be adjusted according to various energy markets.

Analysis and Simulation of Boost-Flyback Converter for Renewable Energy Integration is mainly focusing on boosting and decreasing the voltages coming from the renewable energy sources. The proposed methodology combines the advantages of both Boost and Flyback topologies, providing enhanced efficiency, reduced voltage stress, and improved transient response. The Boost-Flyback Converter employs a two-stage topology, where the Boost stage is responsible for stepping up the input voltage, and the flyback stage facilitates energy transfer and output voltage regulation. The analysis includes a detailed examination of the converter's operating principles, voltage and current waveforms, and control strategies. A comprehensive simulation study is conducted using advanced simulation tools to validate the converter's performance under various operating conditions and load profiles.

Currently, there is a growing prominence on using switched capacitor and switched inductor techniques in high-power boost converters to achieve higher voltages. This is accomplished by employing reactive elements, where the inductor discharges while the capacitor charges. The switched capacitor and switched inductor can extremely attain dc voltage obtain with require few quantities like inductors, capacitors, diodes, and a switch. Modifications were made to the switched inductor converter, resulting in a reduction in the voltage stress on the active switch. The converter now operates based only on the duty ratio. This study suggests adjustments to the switched capacitor and switched inductor converter to decrease the stress on the switch by altering the duty ratio closer to unity. The paper covers the converter's operation, waveforms, design equations, and simulation results to illustrate this modified converter setup.

The depletion of fossil fuels worldwide is driving the acceleration of renewable power energy systems, particularly those based on photovoltaics (PV). To enhance the voltage output from PV electricity systems for grid-linked inverters, a dc-dc boost converter can be employed. However, traditional boost converters face various challenges, leading to the proposal of several configurations aimed at improving their boosting capabilities. One such configuration is the Quasi impedance (Z)-source dc-dc converter, derived from the conventional impedance (Z)-source converter by incorporating a switch and diode at the output terminals. Compared to existing Z-source configurations, this converter operates with a small duty cycle (less than 0.25), resulting in a higher boosting factor and avoiding instability caused by inductor saturation. Moreover, it requires fewer passive components, leading to reduced losses, increased power density, and cost reduction. This paper discusses the features of the proposed converter and its functioning in continuous current conduction mode.

In this paper the idea of virtual instrument and its legitimacy against traditional instrument will be presented. Here we will utilize the graphical programming idea of National Instrument's Lab VIEW. In this paper we will present the idea of a temperature control framework which is a PC-based virtual instrument application. Accepting these musings as standards, a temperature control framework will be planned utilizing Lab view dataflow visual programming language and climate to understand the information of the temperature of the articles. The last examination included the improvement of a relative fundamental subordinate (PID) wind speed control framework for a little subsonic air stream. Notwithstanding the course, understudies and analysts can get to the air stream speed control framework through the Internet to explore the impacts of various addition settings on the PID air stream control framework.

Abstract Electricity places a vital role in the development of economy of a country. Electrical power grids are the most complex of all human engineered systems. This paper presents the progress made in the energy grid and outlines the power scenario in India. It reports the impact created by the event “Switching off lights” on grid frequency. Role of POSOCO in the management of the event has been vividly described. The paper provides us the most important input for prioritizing measures or steps to be taken to cope up with the situation and possible action plan have been summarized.

Abstract This paper proposes the design of a novel Modified Counter based Digital Pulse Width Modulation generator to analyze the performance of the closed-loop DC-DC buck converter. The closed-loop DC-DC buck Converter uses the Proportional Integral controller to regularize the output voltage and the bio-inspired algorithms namely Particle Swarm Optimization and Ant Colony Optimization are considered for the generation of the optimal values for the PI gains namely KP and KI. Conventionally, the switching of the buck converter is controlled by the PWM signals that exhibit complexity in design. The updated Digital Pulse Width Modulation techniques were suitable for voltage regulation at the cost of high clock frequency requirement, increase in the design area for real implementation, and the trade-off between the switching frequency and the component size of the buck converter. To overcome these, the modified Counter based Digital Pulse width modulation that generates the high switching frequency DPWM is developed in HDL. Then the proposed technique is validated in the closed-loop PI-based DC-DC buck converter using the System Generator MATLAB SIMULINK. To regularize the voltage output, the PI controller is included along with the optimization algorithms such as Particle Swarm Optimization and Ant Colony Optimization to optimize the PI gains. The time transient analysis of the proposed method exhibits improvement in the ACO based design compared to the PSO based method. The power and area are manipulated by using the Cadence and Xilinx tools

This paper presents a boost DC-DC converter topology with non - isolated high gain and output midpoint, to boost the voltage obtained from solar photovoltaic panels. The three-level boost converter is coupled to the output port of the single-switch quadratic boost converter to derive the proposed converter topology. The voltage gain of the proposed converter is greater than that of the classical boost converter. The voltage stress on the switches of the proposed converter is equal to half of the converter output voltage. Static analysis, operating modes, experimental waveforms in continuous current conduction and discontinuous current conduction modes are shown. A 520 W prototype converter was implemented in the laboratory and its results are presented.

The growing popularity of 6G-enabled internet of things (IoT) is addressing persistent challenges in long-term applications. Large-scale knowledge analytics requires real-time, accurate data analysis from AI. When utilizing artificial intelligence for in-depth data analysis, challenges emerge in safety, confidentiality, teaching data, and integrated architecture. This study recommends combining blockchain and artificial intelligence for internet of thinking applications. The suggested architecture is evaluated qualitatively and quantitatively. The authors measure blockchain and AI's problem-solving abilities using the AI-centered B.C. and BC-destined AI framework. After comparing qualitative methodologies, the AI-BC architecture outperforms sophisticated methods.

Academia has recently focused on taxi demand prediction, seeing its potential in intellectual transference systems. Older methodologies often overlooked nuanced journey conditions, mainly forecasting from origin locations. This approach lacks efficiency, disregarding demand dynamics between origins and destinations. The research introduces taxi origin-destination demand prediction, leveraging mechanical automation. The authors aim to anticipate future demand across all potential area pairings, acknowledging complex location interplay. A crucial challenge is efficiently collecting diverse contextual data for effective analysis. They employ a sophisticated mechanical automation system integrating deep neural networks (DNNs) to classify journey starting and ending points, outperforming traditional methods in accuracy and performance. Through extensive testing on large-scale datasets, the DNN-based system excels in predicting taxi demand. Leveraging advanced technologies like mechanical automation, the authors pave the way for more efficient transportation systems.

The exponential rise of internet users accelerated the global spread of social media. This study proposed a data-mining-based methodological approach for forecasting the performance metrics of content posted on brands’ Facebook pages. Analysis of Facebook posts provides valuable information for social media users. The main objective of this study is to predict social media performance metrics using Facebook data. Machine learning (ML) will recognise patterns and then make predictions with better accuracy. The dataset for Facebook metrics utilised in this study was obtained from the UCI ML repository. The dataset consists of 500 instances with 19 attributes. The performance of the brand’s Facebook page has been predicted using the attribute “total interactions” as a goal variable. For prediction, the methods Light Gradient Boosting and random forest (RF), AdaBoost and XGBoost ML algorithms have been employed. A heatmap has been used to observe the correlation between the different features. It is observed that based on regression evaluation metrics, the XGBoost -based regression model performs better.