Spiking neural network controller for flight stabilization of a microrobotic bee

The Microrobotic Bee is an 80-milligrams micro aerial vehicle designed by Harvard Microrobotics Lab. Inspired on the flight apparatus of Diptera (flies) the Microrobotic Bee relies on flapping wings to generate lift and torques necessary for flying. The current project illustrates the design of a Neural Network controller based on spiking neurons, for the flight stabilization of Microrobotic Bee. The neurocontroller presents a continuous time and space Actor-Critic architecture where the Temporal Difference error, driven by dopaminergic neurons, modulates weights change within plastic synapses to learn the correct control law.