@imsec.ac.in
Assistant Professor, Electronics & Communication Engineering Department
IMS Engineering College, Ghaziabad
Chandan Choubey received B.Tech degree from Uttar Pradesh Technical University, Lucknow in 2008, M.Tech degree from National Institute of Technology Kurukshetra in 2010. Presently, he is working as Assistant Professor in Dronacharya Group of Institution, Greater Noida and pursuing his Doctorate degree in the field of Robotics and its Control from NIT Kurukshetra. His areas of interest includes Robotic System and its control, Optimization, Soft Computing and its application for the Dynamics and Control of Parallel Manipulators.
PhD from NIT Kurukshetra
M.Tech from NIT Kurukshetra
B.Tech from Uttar Pradesh Technical University, Lucknow
Control system
Optimization Techniques
Optimal Trajectory generation of robotic arm
Optimal Trajectory tracking of robotic arm
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Abhas Kanungo, Chandan Choubey, Varun Gupta, Pankaj Kumar, and Neeraj Kumar
Springer Science and Business Media LLC
G. Chandramowleeswaran, Chandan Choubey, Sireesha Pendem, Shaunak Pal, and Amit Verma
IEEE
By merging autonomous systems and human-robot interaction (HRI), Industry 4.0 has caused a paradigm shift in manufacturing and production processes. To maximize productivity and ensure smooth machine-human collaboration in this environment, motion planning, and control parameters must be effectively integrated with autonomous robots. With the help of hybrid deep learning models including Convolutional Neural Networks (CNNs), Long Short-Term Memory (LSTM), and Recurrent Neural Networks (RNNs), the implementation of HRI with motion planning and control parameters in autonomous systems is proposed in this research. To increase productivity, safety, and flexibility in industrial settings, the suggested approach intends to enable robust and adaptive interaction between humans and robots. CNNs are used to analyze and comprehend the surroundings by the autonomous system through object identification and visual perception. LSTM and RNN models are used to combine this perception with the temporal components of motion planning and control. The suggested method allows the autonomous system to learn from sizable datasets of human-robot interactions, enabling the real-time prediction of human intentions and behavior. The models' ability to respond to changing settings and modify the motion planning and control parameters as necessary makes it easier to complete tasks quickly and work together. By solving important issues with human-robot collaboration, the adoption of this hybrid deep learning-based HRI system advances Industry 4.0. The suggested method improves the general effectiveness and safety of industrial operations by seamlessly integrating motion planning and control parameters with autonomous systems. Additionally, the hybrid deep learning models' adaptive nature enables ongoing learning and performance improvement over time.
Varun Gupta, Abhas Kanungo, Pankaj Kumar, Neeraj Kumar, and Chandan Choubey
Springer Science and Business Media LLC
Chandan Choubey and Jyoti Ohri
Informa UK Limited
This paper presents the mathematical modeling and optimal trajectory tracking control of a 3-degree-of-freedom parallel manipulator, commonly known as Maryland manipulator. Three unlike sequential ...
Chandan Choubey and Jyoti Ohri
Springer Science and Business Media LLC
The paper presents mathematical modeling and optimal path control of a 3-DOF Maryland manipulator. Three dissimilar and consecutive paths are taken under observation, and control action is performed by linear quadratic regulator (LQR)-based proportional–integral–derivative (PID) controller. To achieve optimal path tracking control, tuning of PID gain parameters is necessary and it is done by optimal selection of weighting matrices of LQR. The evolutionary optimization algorithms like GA and PSO are used in the past for the optimal selection of LQR parameters for tuning of PID gain parameter, but both the methods fail to achieve accurate trajectory tracking control because the simulation results show higher values of fitness function (J), sum square error, integral square error, and integral absolute error between the desired and the actual trajectory for all joint angles, as discussed in the result analysis section. The gray wolf optimizer (GWO) algorithm is then proposed; this method provides optimal values of gain parameters. The simulation results show that the values of all types of error and fitness function for all joint angles are quite lesser than GA and PSO. Hence, the proposed GWO proves better among all and it provides high accuracy in trajectory tracking control with better performance indices. To demonstrate the proposed algorithm, the mathematical simulations are performed as well as the conduction of experimental work.
Chandan Choubey∗ and Jyoti Ohri∗∗
Abul Hassan Laiq, Bheem Kant Kaushal, Mohan Kumar, and Chandan Choubey
IOP Publishing
Abstract Fire occurrences are the disaster that can cause many humans as well as animal death, property mutilation and disabilities to affected humans. The main motive to develop this robot is to overcome the loss of human lives and get best of this disaster. The idea is to develop a robot that will be used for fire extinguishing purpose and save a huge number of human lives. Keeping in mind that the human and animal lives are important to work upon. A low-cost robot will be developed for fire extinguisher. Arduino programming will be done wisely so that the flame sensor can work accordingly. Small and low-cost robot will definitely reduce the chances of fire to get to its extreme level.
Chandan Choubey and Jyoti Ohri
Cambridge University Press (CUP)
SUMMARYIn this paper we designed an optimal trajectory generation (OTG) method to generate easy and errorless continuous path motion with quick converging using Grey Wolf Optimization (GWO) method. The proposed OTG method finds the trajectory path with minimum tracking error, combined speed, joint increasing speed wrinkle, as well as joint lurching move to follow an error-free smooth continuous path.
S.S. Ghosh, C. Choubey, and A. Sil
Elsevier BV
Chandan Choubey and Jyoti Ohri
IEEE
In this paper, Maryland parallel manipulator is being examined, having 3-degrees-of-freedom (DOF). For the dynamic analysis, three different sequential trajectories are considered. The major concern is the nonlinearities and disturbances, which are always associated with the parallel manipulators and it is observed that the process and measurement noise can be rejected by using Linear-Quadratic-Gaussian (LQG) controller. For the designing and to control the trajectory of the examined manipulator, the weighting matrices Q & R must be optimally selected. Three different methods for tuning Q & R weighted matrices are proposed i.e. (i) Trial and Error, (ii) Algebraic Approach and (iii) Genetic Algorithm (GA). The performance analysis of LQG is being simulated by MATLAB. Finally, after designing and tuning of LQG the results of all three different methods of tuning are compared, and it are found that in case of LQG-GA the trajectory tracking is better and efficient as compared to other two methods.