KEYWORD |

#### Visual solution of electric circuits

keywords CIRCUITS, MACHINE LEARNING, OCR, PATTERN RECOGNITION

Reference persons STEFANO GRIVET TALOCIA

Research Groups EMC Group (Electromagnetic Compatibility)

Description The objective of this thesis is the realization of a software package that takes on input an image representing an electric circuit drawing (including component data and a statement of the circuit problem to be solved), and produces on output the complete solution. The input image can be a picture, a scan of a textbook, or even a sketch generated by hand. Possible problem statements include the evaluation of voltages and currents under DC excitation, AC steady-state, symbolic (Laplace- or Fourier- domain) or transient.

Methodologies to be used and skills to be developed include:

- image processing and machine learning/pattern recognition, in order to recognize which circuit elements are included, what is the circuit topology.

- text recognition, in order to identify component names/values and problem statement (existing libraries for OCR will be investigated and used)

- circuit equation formulation using the Modified Nodal Analysis (MNA) framework and solution

- automated generation of a "clean" circuit schematic and problem solution

Required skills Predisposition for mathematical formulation of problems and algorithms. Programming in the Matlab environment. Knowledge of C/C++ programming language is desirable to use existing software libraries

Deadline 07/10/2019
PROPONI LA TUA CANDIDATURA