| 主权项 |
1. A method comprising:
embedding, using a phase space embedding module, time series data relating to a system within a reconstructed phase space to provide embedded data, wherein the phase space embedding module comprises instructions stored on a non-transitory computer-readable medium that are executable by a processor; predicting, using a prediction module, at least one future state of the system based on a rate of separation of trajectories of the embedded data within the reconstructed phase space; and evaluating a fitness of solutions associated with a plurality of individuals of a population for controlling the system using a fitness function, each individual comprising a dominant genotype and a recessive genotype, wherein each genotype represents a solution of the solutions associated with the plurality of individuals for controlling the system, wherein the fitness function is based on the at least one future state of the system and is applied to the dominant genotype of each individual; wherein predicting the at least one future state of the system based on the rate of separation of trajectories of the embedded data within the reconstructed phase space comprises predicting the at least one future state of the system based on a maximum Lyapunov exponent associated with the embedded data within the reconstructed phase space, and wherein the prediction module comprises instructions stored on the non-transitory computer-readable medium that are executable by the processor; wherein predicting the at least one future state of the system based on the maximum Lyapunov exponent comprises:
selecting a current state point from the reconstructed phase space;identifying a current close point in a different trajectory than the current state point, wherein the current close point is within a predefined distance of the current state point in the reconstructed phase space;calculating a first distance between the current state point and the current close point;calculating a second distance between a previous state point and a previous close point, wherein the previous state point is a point previous to the current state point on the trajectory of the current state point, wherein the previous close point is a point previous to the current close point on the trajectory of the current close point; andpredicting at least one future state point based on the first distance, the second distance, and the maximum Lyapunov exponent. |
