How and why I built Artifical Intelligence algorithms for Digital Signal Processing library and testbench

Published Dec 01, 2017
How and why I built Artifical Intelligence algorithms for Digital Signal Processing library and testbench

About me

I'm Francisco José Solís Muñoz, a C++ senior developer and researcher. I came up with the idea of using multiparameter estimation for electric voltage signals using Artificial Intelligence algorithms.

The problem I wanted to solve

Power quality assessment is a critical task in any kind of environment. The lifespan of devices plugged in depends on it. Sag, swells, and other kind of electric disturbances need to be analyzed in order to give a score to an electric source.

What are Artifical Intelligence algorithms for Digital Signal Processing library and testbench?

A electric sinewave signal library has been created where the modelling of different disturbances and synthetic signal generation is deployed. A testbench application is included in the package where AI algorithms are used in order to estimate important parameters needed to reduce the error of the Quality Assessment of the power source.

Tech stack

I used C++, CUDA (in a previous phase) and OpenCL using Boost::Compute contrib library. CMake was used for building.

The process of building Artifical Intelligence algorithms for Digital Signal Processing library and testbench

I started reading about the tools that were going to be implemented. My parallel programming background was ready for development, so the first task was to develop the library and the testbench later.

Challenges I faced

This was not difficult.

Key learnings

Documentation of the library before implementation of it was key to not have to rewrite it.

Tips and advice

Learn to use CMake wisely. It saves so much time for using different IDEs and OSs'.

Final thoughts and next steps

This project will be updated in order to add an User Interface.

Discover and read more posts from Francisco José Solís Muñoz
get started