Systems and methods for training a neural network to control an aircraft

A method of training a neural network to control an aircraft, the method comprising:

• providing a first control input to a flight simulator, wherein the flight simulator creates a simulation of use of the aircraft, and wherein the flight simulator generates first state data responsive to the first control input and provides the first state data as input to a first neural network to generate a candidate second control input;
• selecting, based on a random value, a second control input for the flight simulator, wherein the second control input is selected from between the candidate second control input and a randomized offset control input, and wherein the randomized offset control input is based on a random offset applied to the first control input;
• providing the second control input to the flight simulator, wherein the flight simulator generates second state data responsive to the second control input;
• and
• updating weights of the first neural network via a neural network trainer based, at least in part, on the first state data and the second state data, wherein the first neural network is configured to receive sensor data from a plurality of sensors and is configured to control one or more systems of the aircraft based on the sensor data and the weights of the first neural network updated by the neural network trainer.

A non-transitory computer-readable storage device storing instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising:

• providing a first control input to a flight simulator, wherein the flight simulator creates a simulation of use of an aircraft, wherein the flight simulator generates first state data responsive to the first control input and provides the first state data as input to a first neural network to generate a candidate second control input;
• selecting, based on a random value, a second control input for the flight simulator, wherein the second control input is selected from between the candidate second control input and a randomized offset control input, and wherein the randomized offset control input is based on a random offset applied to the first control input;
• providing the second control input to the flight simulator, wherein the flight simulator generates second state data responsive to the second control input;
• and
• updating weights of the first neural network via a neural network trainer based, at least in part, on the first state data and the second state data, wherein the first neural network is configured to receive sensor data from a plurality of sensors and is configured to control one or more systems of the aircraft based on the sensor data and the weights of the first neural network updated by the neural network trainer.