Here is a snippet of my previous works.

Machine Learning Projects

Feature Extraction Based Thermoacoustic Instability Detection In Premixed Combustor Using Machine Learning

This project was completed in the partial fulfillment of the Requirements for my Master's degree at IIT Kanpur.

This project aims to classify the type of combustion (stable, transition or, unstable) based on pressure data collected from the premixed combustion chamber wall.

Dataset Description: The data contains pressure signal from the combustor with varying air flow rate and combustor lengths.

Features : Performed feature extraction in time domain and time-frequency domain. End up with some time domain Statistical features and time-frequency domain wavelet packet coefficients to train ML models.

ML Algorithms: Used classification algorithms like Logistic regression, decision Tree, SVM and Ensemble algorithms.

Evaluation: Used Accuracy, F1 Score and Confusion Matrix metrics to evaluate the performance of each model with different types of combustion chambers.

Following are some images of data visualization, extracted features and confusion matrices of different models.

Diagnosis of Rolling Element Bearing

This project aims to classify the type of fault in bearing based on vibration data.

Dataset: I have used IMS dataset which is available on NASA's prognostic-data-repository(link)

There are four classes(normal, inner, roller and outer) of continuous time series vibration data.

we need to first segment the vibration data then calculate the feature from each segment. and generate a feature matrix.

This feature matrix will be used for the training and testing of our model.

I have used LogisticRegression and DecisionTree classifier and got good accuracy with decision tree.

Then we can generate confusion matrix to see the performance of model.

Logistic Regression with NumPy and Python

This is a guided project from coursera

The aim is to classify whether a student will be passed or not based on their score in two different exams.

In this project I Performed classification using logistic function by computing cost with the help of NumPy . Used gradient descent for optimization of cost with writing code from scratch.


University Admission Prediction Using Multiple Linear Regression

This is a guided project from coursera

This is a regression problem. Aim of this project is to predict change of a student to get admission in a university based on their scores and few more parameters.

In this project I Performed exploratory data analysis, feature scaling, standardization. Training and evaluation of linear regression model.

Also trained ANN with ReLU as activation function using Adam optimizer.

Iris Species Classification

This is a self project I have done to get better understanding of data visualization and machine learning libraries.

Performed exploratory data analysis with the help of pandas, matplotlib and seaborn library for visualisation of data (scatter plot, pair plot, contour plot, histogram, pdf, cdf and box plot with wishkers)

Trained logistic regression, kNN and SVM models using sklearn library.

Academic Projects

3-D Model of F-1 car chassis

Stress simulation after impact

Design and simulation of formula one car chassis

These are some images of my project that I developed for the completion of my Bachelor Degree in 2018.


Chassis is a major component in a vehicle, in which different types of failure occur due to static loading condition. In the present work, static characteristics are analyzed “Location of high stress area " is identified with front, side, and rear impact testing using finite element method.

Modal updating of car chassis will be done by comparing the material’s properties hardness, stiffness, compressive strength, coefficient of thermal expansion, torsional rigidity, modulus of elasticity for various materials. The computer aided design geometry of Formula-1 car chassis is generated in Solidworks 2016 and model analysis is done in Ansys workbench 18.0.

SEM Image of Fe2O3/rGo

Laser engraved on coated glass surface

Ferric Oxide(Fe2O3) Nanoplate/Graphene Composite Material for Super capacitor

These are the photos that I taken for my course project in 2019

Supercapacitor is an electrochemical capacitor which stores charge electrostatically and faradaically on the electrode surface by charge transfer between electrolyte and electrode. Used in applications requiring high power density and rapid charge/discharge cycles rather than long term compact energy storage. Supercapacitor can store 10-100 times higher energy density than conventional capacitor and 5- 20 times higher power density than battery.

Engraving on sample prepared (GO deposited on glass plate) at different power. In the air the laser treatment of the GO films occurs with a partial destruction of oxygen-containing groups on the surface and CO and CO2 release. This results in the formation of rGO which has higher conductivity.

Fe2O3 Nanoplates/Graphene composite was synthesized by the Hydrothermal process . Laser treatment increased the conductivity of GO deposited on glass slide by reducing it to r-GO .

Laser textured over SS304 surface. Normal View (Top) and Microscopic View (Bottom)

Laser Texturing on Stainless steel (SS304) using Epilog Fiber Laser

These are the photos that I taken for my course project in 2019

Laser beam machining of engineering materials has become a viable alternative to conventional methods of Creating microstructures on the surface improves load capacity, wear rates, lubrication lifetime and reduces friction coefficient. A regular array of ellipsoidal dimples with different scanning speeds has been created. Entire dimensional, shape accuracy and quality of produced depressions is significantly affected by presence of recast layers and molten material formation on the walls and on the bottoms of these depressions. Formation of recast layer is more prominent at lower laser scanning speed since there is higher material removal. More material removal leads to formation of recast layer and less precision.

Assemblies

Here are some project I did during training at CETPA

•Piston Cylinder Assembly

•Shaper Tool Head Assembly

• Cross Head Assembly

• Plumber Block Assembly

• Fixture Assembly

• Press Tool Assembly

• Pneumatics vehicle chassis


3D Models

These are some 3-D models that I created while practicing Solidworks

  • Spur gear

  • Bevel gear

  • Lenovo laptop

  • Radial engine

  • Rough model of my home