VMVishal Mehra

Vishal Mehra

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Vishal Mehra has a total of 12 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Radial basis function neural network based higher order sliding mode control

    by Vishal Mehra, Dipesh Shah, Axaykumar Mehta

    This paper presents a Radial Basis Function neural network based higher-order Sliding Mode Control for robust control of a dynamical system. A conventional sliding mode controller is suffering from a chattering problem, and the Super Twisting Algorithm is a special kind of higher-order Sliding Mode Control that has the capability of minimizing the chattering problem. There are also unknown model parameters and external disturbances that exert a negative influence on the control performance. To address the issues of model uncertainties and chattering, a Radial Basis Function (RBF) neural network based Super Twisting Algorithm is designed. The RBF neural network evaluates the model parameters and uncertainties, while the Super Twisting approach mitigates chattering, hence improving the controller’s overall performance. Lyapunov stability based adaptive laws are derived for online updating of the parameters of the neural network. The proposed control algorithm was tested on a 2-degree-of-freedom serial flexible joint robotic arm to investigate its efficacy. The controller has a lower control chattering amplitude, lower control energy consumption, and a good tracking response, when compared to the RBF based conventional Sliding Mode controller and simple STA controller, as shown by the results.

  • Constrained GA Optimized Super Twisting Sliding Mode Controller for 2-DoF Robotic Arm

    by Vishal Mehra, Dipesh Shah, Axaykumar Mehta, Siddharth Mehta

    This paper presents a constrained optimization method for the Super-Twisting Algorithm using a Genetic Algorithm. The Super-Twisting Algorithm is an important class of higher-order sliding mode strategy commonly employed in designing robust controllers and observers that effectively mitigate the chattering effect observed in conventional sliding mode control. The performance and stability of the dynamical system critically rely on the appropriate selection of gain parameters, so it is necessary to choose the optimal gains for the controller. A novel approach to optimize the control gain parameters using a constrained Genetic Algorithm is proposed to achieve the desired dynamic response. The constraints are derived from a strict Lyapunov function, thereby ensuring the finite-time stability of the closed-loop system through Lyapunov stability criteria. The performance of the proposed optimized gains is evaluated through real-time experiments on a 2-degree-of-freedom serial flexible joint robotic arm. The experiment results show that the proposed technique performs better regarding dynamic response, disturbance rejection, and chattering reduction.

  • Estimation of Evaporation Using Long Short Term Memory and Gated Recurrent Unit Based Neural Network

    by Vishal I. Mehra, Arvind N. Nakiya, J. Sravan Kumar

    This study presents the comparison of conventional Artificial Neural Network (ANN) and advanced neural networks to predict weekly potential evaporation for Anand, Gujarat, India, which comes under the subtropical climatic zone. Recently, many advanced deep neural structures like Long Short Term Memory (LSTM) and Gated Recurrent Unit (GRU) have been introduced that have excellent prediction accuracy. Climate data such as bright sunshine hours, rainfall, wind speed, maximum and minimum temperature, and maximum and minimum relative humidity have been used to train and test the conventional and advanced neural network models. A comparison was made for the estimation of evaporation predicted by these models. The performance results show that deep neural network models with advanced structures like LSTM and GRU have performed better in terms of Root Mean Square Error and correlation coefficient and are able to learn the events very well in comparison to conventional neural network structures.


  • Open Access

    Adaptive STA based SOSMC using Gated Recurrent Unit blended RBFNN structure

    by Vishal Mehra, Dipesh Shah, Axaykumar Mehta

    In this paper, an adaptive Super Twisting Algorithm (STA) based Second Order Sliding Mode Controller (SOSMC) design using a novel Gated Recurrent Unit (GRU) blended Redial Basis Function Neural Network (RBFNN) structure is proposed. In this GRU blended RBFNN structure, the GRU is placed in 1st hidden layer and the RBF is placed in the 2nd hidden layer to enhance the estimation of unknown function f (x) of the nonlinear dynamical system. The estimated function is used to design an adaptive STA of the SOSMC with adaptive laws for updating STA gains. The closed loop asymptotic stability and finite time convergence of the system is also ensured using the Lyapunov theory. Finally, the controller is implemented on two link Robotic Arm with serial flexible joints to show the performance and efficacy. The experimental results exhibit an excellent tracking performance and robustness of the controller.

  • Predictive Machine Learning Models for DDoS Attack Mitigation

    by Jyoti Bansal, Vishal Mehra, Mohd Sadim

    This research provides a comprehensive method for enhanced Distributed Denial of Service (DDoS) detection and mitigation by utilizing ensemble learning, namely the Random Forest, Gradient Boosting, and AdaBoost algorithms. This may be done by combining these three strategies of learning. These algorithms were chosen for their demonstrated capacity to analyze high-dimensional data, understand complex connections, and adapt to new forms of attack. The dataset is then divided into training, validation, and testing sets to assess the model's resilience. Predictions obtained because of independent training of each basic technique on the training dataset were labeled RFi(X), GBi(X), and ABi(X), respectively. The weights for this approach are determined by a procedure known as cross-validation. The ensemble forecast is used to establish a threshold, and when a DDoS attack reaches that level, mitigation is activated. One method for dealing with the problem is traffic rerouting. Other methods include rate limiting, temporary IP address bans, and the adoption of new machine learning models for more in-depth research. The ensemble learning model's efficiency is reviewed on a regular basis, and changes are made as needed to account for developing trends in DDoS assaults. The suggested technique has several advantages and benefits over more standard DDoS protection systems. These enhancements include higher rates of true positives, lower rates of false alarms, and increased accuracy and precision. This illustrates its agility in both responding to and defending against a wide range of DDoS attack routes.