System for evolving controllers for micro-air vehicles
AssigneeHRL Laboratories, LLC
InventorsCharles E. Martin, Heiko Hoffmann
Life of this patent
Estimating the term…
Abstract
Described is a system for optimizing controllers for micro-air vehicles. The system identifies binary rules that lead to a desired behavior of a micro-air vehicle. The binary rules are identified by sampling rule-sets from a probability distribution over a search space. A fitness value of each rule-set for the desired behavior is determined, and top-performing rule-sets are selected. The top-performing rule-sets are used to update the probability distribution over the search space until a convergence criterion is met, resulting in a fittest rule-set. A control signal based on the identified binary rules is transmitted to one or more actuators of the micro-air vehicle.
Claims17
A system for optimizing controllers for a micro-air vehicle, each controller having an electronic circuit of connections configured to receive sensor inputs as a temporal sequence of sparse binary vectors, the system comprising:
- one or more processors and a non-transitory computer-readable medium having executable instructions encoded thereon such that when executed, the one or more processors perform operations of:
- identifying a set of binary rules that leads to a desired behavior of the micro-air vehicle, wherein identifying the set of binary rules comprises:
- (a) sampling a plurality of rule-sets from a probability distribution over a search space;
- (b) determining a fitness value of each rule-set for the desired behavior;
- (c) selecting a set of top-performing rule-sets having fitness values at or above a predetermined percentile;
- (d) updating the probability distribution over the search space with the set of top-performing rule-sets;
- repeating (a) through (d) until a convergence criterion is met, resulting in a fittest rule-set;
- and
- transmitting a control signal based on the identified set of binary rules to one or more actuators of the micro-air vehicle to cause the one or more actuators to physically actuate and cause the micro-air vehicle to perform the desired behavior.
A computer program product for optimizing controllers for a micro-air vehicle, each controller having an electronic circuit of connections configured to receive sensor inputs as a temporal sequence of sparse binary vectors, the computer program product comprising:
- a non-transitory computer-readable medium having executable instructions encoded thereon, such that upon execution of the instructions by one or more processors, the one or more processors perform operations of:
- identifying a set of binary rules that leads to a desired behavior of a micro-air vehicle, wherein identifying the set of binary rules comprises:
- (a) sampling a plurality of rule-sets from a probability distribution over a search space;
- (b) determining a fitness value of each rule-set for the desired behavior;
- (c) selecting a set of top-performing rule-sets having fitness values at or above a predetermined percentile;
- (d) updating the probability distribution over the search space with the set of top-performing rule-sets;
- repeating (a) through (d) until a convergence criterion is met, resulting in a fittest rule-set;
- and
- transmitting a control signal based on the identified set of binary rules to one or more actuators of the micro-air vehicle to cause the one or more actuators to physically actuate and cause the micro-air vehicle to perform the desired behavior.
A computer implemented method for optimizing controllers for a micro-air vehicle, each controller having an electronic circuit of connections configured to receive sensor inputs as a temporal sequence of sparse binary vectors, the method comprising an act of:
- causing one or more processers to execute instructions encoded on a non-transitory computer-readable medium, such that upon execution, the one or more processors perform operations of:
- identifying a set of binary rules that leads to a desired behavior of a micro-air vehicle, wherein identifying the set of binary rules comprises:
- (a) sampling a plurality of rule-sets from a probability distribution over a search space;
- (b) determining a fitness value of each rule-set for the desired behavior;
- (c) selecting a set of top-performing rule-sets having fitness values at or above a predetermined percentile;
- (d) updating the probability distribution over the search space with the set of top-performing rule-sets;
- repeating (a) through (d) until a convergence criterion is met, resulting in a fittest rule-set;
- and
- transmitting a control signal based on the identified set of binary rules to one or more actuators of the micro-air vehicle to cause the one or more actuators to physically actuate and cause the micro-air vehicle to perform the desired behavior.
Dependencies are inferred from the claim text.