MATLAB Consulting services

We can consult you about the MATLAB Simulink Stateflow project and algorithm. 

MATLAB is our proficiency

matlab simulink stateflow consulting services project expert

Expert MATLAB Simulink consultant

Kazem is our chairman. He has a master of science in electrical engineering from a top university in 2009.

He can develop any type of algorithm in MATLAB. He has more than 18 years of experience in writing code in this software. He is an expert in MATLAB/Simulink/Stateflow. He can do anything in MATLAB !!. He has implemented more than 900 projects in MATLAB/Simulink. Some people called him the God of MATLAB! He is famous for this software and founder of several websites about MATLAB, Python, and C: (matlab1.ir, iran-matlab.ir, matlab1.com). Furthermore, He had more than 500 international students from different countries such as Turkey, Italy, Germany, and India. He also had a position at a university where had taught and consulted the Master and PhD students in programming. He liked teaching other people and was interested in knowledge sharing in a team.

 

blank

He can integrate the simulation models in simulation environments (Model in Loop (MIL), Hardware in Loop (HIL), Software in Loop (SIL)) and create simulation models for the entire vehicle, powertrain, and vehicle components. In addition it, He can maintain and extend the current simulation libraries and can ensure the operation of the HIL systems as well as support and implementation of technical changes and extensions of the HIL systems. He can check and calibrate the HIL functions. He knows about PHS bus, AUTOBOX, SCALEXIO, Automotive Simulation Models (ASMs), Real-Time Interface (RTI).

He also worked with the HIL simulator from the dSPACE GmbH company. One of the most important Simulink toolboxes in HIL simulation is Simscape. He worked with different blocks in this toolbox. He also worked with Simulink Real-Time Explorer. He is familiar with the V-model (Automotive Software Development Process). He can accompany the complete development process taking into account the V-model.

He is familiar with several topics in the automotive industry and has read some parts AUTOSAR, Adaptive AUTOSAR and ISO 26262. Furthermore, He has a lot of information about autonomous cars and various sensors in this area, and He has experience with Model Advisor (ISO 26262, MISRA, IEC 62304, IEC 61508-3) in MATLAB / Simulink.

He has worked with different data sets such as Excel, text, HTML, Image (JPEG, PNG, Tiff), video, and sound. They were from different research areas such as Iris Data Set, Time Series Analysis Dataset, websites, digit recognition dataset, medical data set, MRI image dataset.

He worked for several years as a mechatronics engineer in the oil and gas industry. So, He is familiar with measurement and control technology and many types of sensors.

His first work with image processing and Artificial Neural networks was twenty years ago. He uses ANN to classify the fingerprint image into four classes. The results of this work were published in an international journal. He worked with many Artificial Neural Networks such as Multi-Layer Perceptron, Radial Bias Function, Support Vector Machine, Recurrent Neural Network, Reinforcement Learning, Convolutional Neural Network (CNN), R-CNN.

He has also experience in Python, machine learning, artificial intelligence (AI), deep learning, many types of artificial neural networks (MLP, RBF, ANFIS, CNN, R-CNN, DBN), optimization algorithms (GA, PSO, BCO, etc.) and ML toolkit (Keras, TensorFlow). I have experience in machine learning, Artificial Neural Networks, image processing and artificial intelligence in Python.

In addition to it, I have some experiments in data reduction methodologies in many projects. For example, in one of my previous projects, there were 715 features for each sample. It was impossible to apply this large feature vector to a small machine learning model, so I tested some data reduction methodologies to reduce the number of features and applied some feature selection algorithms to find the best feature, finally, there were 193 features.

One of my previous projects in Python was OCR post-correction techniques to automatically correct OCR output texts. Our algorithm only requires a small amount of relatively clean training data, but previous methods need a large dataset. We tested the model on a noisy dataset and had a good performance. The performance of the final model was 93 %. My model was Bi-directional LSTM. Many experiments were performed to identify the optimal model. One of the challenges was special characters in the Turkish language. Loss in the training steps was calculated with TensorFlow’s sparse softmax cross-entropy function. I used Python, Keras, TensorFlow, Numpy, codes, sys, Anaconda, Spyder.

 

 

Some companies that use MATLAB and Simulink :

Scania AB

Robert Bosch

Siemens

Ford Motor Company

ZF Group

Altran

Jaguar Land Rover

AVL

Continental Engineering Services

Trinamics

Porsche Engineering

Volvo Group

Zenuity

ZF Group

Danfoss

ARTEMIS GmbH

Tesla

ABB

Veoneer

Bose Corporation

 

 

Kazem’s email :

blank