Trajectory generation of a robot using a neural network

A device for generating a trajectory of a robot from a first configuration to a second configuration within an environment while steering away from obstacles, the device comprising:

• a memory configured to store instructions;
• and
• a processor configured to execute the instructions to:
• obtain physical workspace information associated with the environment in which the robot is configured to operate;
• obtain, using a first neural network, a set of weights of a second neural network that is configured to generate a set of values associated with a set of configurations of the robot with respect to the second configuration, based on the physical workspace information;
• obtain, by applying the set of weights to the second neural network, the set of values associated with the set of configurations of the robot with respect to the second configuration;
• generate the trajectory of the robot from the first configuration to the second configuration within the environment, based on a gradient of the set of values;
• generate a configuration space map that partitions the environment into collision regions and non-collision regions, based on the physical workspace information;
• and
• reshape the set of weights using the second neural network that is a radial basis function network.

A non-transitory computer-readable medium storing instructions, the instructions comprising:

• one or more instructions that, when executed by one or more processors of a device for generating a trajectory of a robot from a first configuration to a second configuration within an environment while steering away from obstacles, cause the one or more processors to:
• obtain physical workspace information associated with the environment in which the robot is configured to operate;
• obtain, using a first neural network, a set of weights of a second neural network that is configured to generate a set of values associated with a set of configurations of the robot with respect to the second configuration, based on the physical workspace information;
• obtain, by applying the set of weights to the second neural network, the set of values associated with the set of configurations of the robot with respect to the second configuration;
• generate the trajectory of the robot from the first configuration to the second configuration within the environment, based on a gradient of the set of values;
• generate a configuration space map that partitions the environment into collision regions and non-collision regions, based on the physical workspace information;
• and
• reshape the set of weights using the second neural network that is a radial basis function network.