R&D Areas/Projects
- Sensor Network Design and Retrofit: Choosing appropriate variables for measurement in the process is essential to a variety of tasks, such as process control, diagnosis, inventory management, process safety, and quality control, amongst others. The sensor network design problem is concerned with issues related to deciding which variables to be measured, the spatial location of the measurements (in case of distributed parameter systems), hardware redundancy (number of sensors used to measure any given variable), and sampling frequency of the chosen sensor. Currently, we are focusing on retrofit strategies to audit an existing sensor network. Additional work is towards optimal base case design approaches that ensure cost-effective retrofit in future. With increasing availability of cheap, wireless sensing devices, frequent retrofit of existing sensor networks will be possible in future.
- Nonlinear, Constrained State Estimation: To ensure efficient process operations, we are working on development of computationally efficient but accurate state estimation techniques for nonlinear systems. In particular, we have developed efficient, optimization based deterministic sampling based approaches that lead to estimates of states that satisfy the known constraints on the states. These deterministic sampling based approaches do not require analytical derivatives for linearization of process models and avoid excessive computational burden associated with random sampling. Initially focussed on variations of unscented Kalman filter that intrinsically assumes Gaussianity assumption, of late, we have been extending these ideas to a Gaussian sum unscented Kalman filter, that does not invoke Gaussianity assumption. Applications on several nonlinear processes demonstrates the applicability of the proposed approaches.
- Aerosol Impact on Climate: Causality Analysis: We are using causality analysis tools, such as those based on path analysis, to hypothesize and quantitatively validate mechanims by which aerosols affect key climate variables such as monsoon precipitation over the Indian subcontinent. This analysis is being carried out mainly using observational datasets.
- Data driven approaches for real time monitoring of self powered neutron detectors (SPNDs): Data driven approaches for real time monitoring of self powered neutron detectors (SPNDs): A typical nuclear power reactor can have hundreds of SPNDs spread throughout the reactor core, measuring the neutron flux distribution in the reactor. These detectors fail over a period of time but cannot be easily replaced as that entails reactor shutdown. In collaboration with BARC and NPCIL, we have developed data driven approaches that can (a) detect if a detector is faulty, and (b) provide an estimate of the true flux value corresponding to the faulty detector. In particular, we have used clustering and principal component analysis (PCA) based approaches to develop multiple linear models that enable fault detection, isolation and reconstruction. Proof of concept has been demonstrated in off line studies on real data. Currently, implementation in a real-time environment is being pursued to further test the applicability and performance of the proposed approach.
- Modeling, Simulation, Control and Optimal Operation of Solar Thermal Power Plants: The interest in this area is develop control relevant dynamic models for hybrid solar thermal power plants. Mode of solar energy capture could be through parabolic trough collectors, linear fresnel reflectors and heating of molten salt. Currently we are investigating use of these model for joint state and parameter estimation as well as implementing advanced control strategies.