Sound Localization, MC, and Neural Network

[tian]

Currently I am working on my final project. My final project title is “Design and Implementation of Mobile Robot Sound Tracker Based on Artificial Neural Network”.

I want to make it embedded with battery as the supply.

Any idea where should I learn first? Where should I start first?
Should I use Microcontroller Atmel AVR or Microchip PIC?
What Neural Network type I work to?
Should I use Fast Fourier Transfom (FFT), Discrete Fourier Transform (DFT), or not use at all?
What is motor type I working with? Is DC Motor is OK?

[torbjorn]

I would recommend that you first develop your algorithms in a programming environment that you already are used to (for example, writing the code in C for running under Windows or Linux on a PC). Then, at a later stage in the project, port the code to a suitable microcontroller.
Regarding Atmel or Microchip, both manufacturers have a wide range of microcontrollers. I think you should first put down a requirement specification when you know how much you need of I/O, communication ports, analog inputs, PWM, code memory, RAM memory etc, then match this specification across the selection guides of the microcontroller manufacturers to choose a suitable part.
In the past, Atmel has been known for having development tools (compilers, assemblers, debuggers etc) of better quality than Microchip has, but this maybe has changed now.
The choice of motor technology is not critical, just make sure that you know how to build control circuitry for the chosen motor.

[Guest]

Thank you for your comment.

[torbjorn]

The C compiler will probably take care of how to store float type data. Regarding the details of which transform to use etc, i think it is better to study textbooks on the subject as well as code examples that might be found at the microcontroller manufacturer's web sites. This is not really my field of profession.

[nikomaster]

You should try a DSP Microcontroller, or try to find another system because a Neural Network uses a lot of processing power.
_________________
Electronics Engineer
The Neural Networks project.