Guide: Prof. Ranjan K. Malik and Prof. Ravi D. Gudi, IIT Bombay
Designed a conceptual flowsheet for butyl acetate synthesis based on extensive literature survey on industrial manufacturing technologies. Introduced process intensification by reactive distillation for improved conversion in equilibrium-limited reaction and saving capital costs.
Simulated the flowsheet in ASPEN Plus by identifying optimum process parameters. Performed multivariable parametric sensitivity analysis and detailed economic analysis of the plant.
Our project was selected as the best UG Design Project for its novelties and is to be presented at CHEMCON, the annual session of Indian Institute of Chemical Engineers.
Pole Placement Controller for a Single Board Heater System (Jan ’14 - Apr ’14)
Guide: Prof. Kannan M. Moudgalya, IIT Bombay
As part of Digital Control course project, performed numerous virtual experiments on the Single Board Heater System (SBHS) hosted on IIT Bombay servers. SBHS is a Multiple Input Single Output (MISO) system with two inputs - heating rate and fan speed, and a single output - temperature.
Carried out system identification using optimization tool in Scilab, modeling the SBHS as a Single Input Single Output (SISO) system with fan speed kept constant during experiments. Implemented a robust Pole Placement Controller on the SBHS to control its temperature at a desired steady state by manipulating heating rate at constant fan speed.
Multivariate Data Modeling (Feb ’13 - Apr ’13)
Guide: Prof. Mani Bhushan, IIT Bombay
Developed a model for predicting salaries of baseball pitchers in Major League Baseball by incorporating multivariate statistical data handling techniques in MATLAB. Past data of pitchers performance statistics, teams performance statistics and salaries were provided. Processed and split these data for model estimation and model validation.
Used multivariate linear regression to build the model and predict future salaries. Validated performance of the model after removing outliers and demonstrated it with residual plots.