The GSN SNN 4−10−30−2 is a hardware based spiking neural network that can autonomous control a remote control robot vehicle. There are 10 artificial neurons and 30 artificial synapses, and is built on 16 full-size breadboards. Four infrared proximity sensor are used on top of the vehicle to determine how far it is away from objects and walls. The sensor data is used as inputs into the first later of neurons.
A full circuit level diagram of the neural network is provided, as well as an architecture diagram. The weights on the network are set based on the resistance value. The synapses allow the weights to be set as excitatory or inhibitory.
Testing of the network went great and the robot had much smoother control than previous testing as the output now has an analog output.
Consider Supporting the Global Science Network.
Ko-fi (Preferred as there is zero platform fee for the transactions)
https://ko-fi.com/globalsciencenetwork.
Patreon (If you prefer Patreon)
/ globalsciencenetwork.
Full Parts List.
