Hybrid neural network for determining at least one parameter of a charging plan for a vehicle

1 . A method of executing a charging plan for at least one vehicle in a transportation system, the method comprising:

receiving, at a charging infrastructure control system, operational state and energy consumption information for a plurality of network-enabled vehicles, wherein the operational state and the energy consumption information that is received at the charging infrastructure control system includes vehicle battery level and vehicle energy consumption rate;

predicting a geolocation of one or more vehicles of the plurality of network-enabled vehicles in a geographic region;

allocating the one or more vehicles of the plurality of network-enabled vehicles to a charging infrastructure in the geographic region; and

optimizing in real-time, at an artificial intelligence system, a parameter of the charging plan based on the prediction of the geolocation of the one or more vehicles in the geographic region,

wherein the optimizing the parameter of the charging plan comprises:

predicting near-term charging needs for the one or more vehicles based on operational status;

evaluating charging infrastructure availability and capacity information within a geofenced region;

detecting accumulation of vehicles at a charging location within the geofenced region; and

adjusting charging infrastructure control parameters locally to accommodate the detected accumulated vehicles queued for charging, wherein the adjusting is performed in real-time responsive to the detecting, and

wherein the artificial intelligence system uses a machine learning algorithm to optimize the parameter of the charging plan.

2 . The method of claim 1 , wherein the predicting the geolocation of the one or more vehicles includes predicting the geolocation based on a planned route of each of the one or more vehicles.

3 . The method of claim 1 , wherein the allocating the one or more vehicles to the charging infrastructure includes allocating based on the predicted geolocation of the one or more vehicles and based on an availability and a capacity of the charging infrastructure in the geographic region.

4 . The method of claim 1 , wherein the optimizing the parameter of the charging plan comprises:

determining charging priority for vehicles based on collective routing information;

adjusting charging rates based on available supply capacity within the geofenced region; and

adjusting charging duration based on detected charge state and requested demand, wherein the determining charging priority, the adjusting charging rates, and the adjusting charging duration are performed simultaneously as part of an integrated optimization process.

5 . The method of claim 1 , further comprising sending a notification to a driver of the one or more vehicles with information about the optimized parameter of the charging plan.

6 . The method of claim 1 , wherein the charging infrastructure includes at least one of: a charging station, a charging pad, or a charging vehicle.

7 . The method of claim 1 , wherein the plurality of network-enabled vehicles is a fleet of electric vehicles.

8 . A transportation system that implements a charging plan for a vehicle, the transportation system comprising:

a charging infrastructure control system that receives operational state and energy consumption information for a plurality of network-enabled vehicles, wherein the operational state and the energy consumption information that is received at the charging infrastructure control system includes vehicle battery level and vehicle energy consumption rate; and

an artificial intelligence system that is connected with the charging infrastructure control system and that, responsive to the operational state and the energy consumption information that is received, optimizes at least one charging plan parameter upon which the charging plan for at least a portion of the plurality of network-enabled vehicles is dependent,

wherein the artificial intelligence system comprises:

a first neural network that processes information including one or more vehicle routes and one or more stored energy states to predict one or more target energy renewal regions for at least the portion of the plurality of network-enabled vehicles; and

a second neural network that processes vehicle energy renewal infrastructure usage and demand information within the one or more target energy renewal regions predicted by the first neural network, wherein the second neural network processes the vehicle energy renewal infrastructure usage and demand information to determine one or more charging infrastructure operational parameters, and

wherein the artificial intelligence system uses a machine learning algorithm to optimize the at least one charging plan parameter in real-time based on one or more energy renewal regions and the determined one or more charging infrastructure operational parameters.

9 . The transportation system of claim 8 , wherein the operational state and the energy consumption information that is received further includes at least one of: a vehicle speed, a vehicle location, or a vehicle direction.

10 . The transportation system of claim 8 , wherein the artificial intelligence system predicts a geolocation of one or more vehicles of the plurality of network-enabled vehicles in a geographic region.

11 . The transportation system of claim 10 , wherein the artificial intelligence system allocates the one or more vehicles of the plurality of network-enabled vehicles to a charging infrastructure in the geographic region based on the predicted geolocation of the one or more vehicles and based on an availability and a capacity of the charging infrastructure in the geographic region.

12 . The transportation system of claim 8 , wherein the charging infrastructure control system sends a notification to a driver of one or more vehicles of the plurality of network-enabled vehicles with information about the optimized at least one charging plan parameter.

13 . The transportation system of claim 11 , wherein the charging infrastructure includes at least one of: a charging station, a charging pad, or a charging vehicle.

14 . The transportation system of claim 8 , wherein the artificial intelligence system adjusts the at least one charging plan parameter based on changes in the operational state and the energy consumption information.

15 . The method of claim 1 , wherein the optimizing the parameter of the charging plan includes optimizing a charging time or a charging rate for the one or more vehicles.

16 . The method of claim 1 , wherein the adjusting charging infrastructure control parameters locally further comprises:

temporarily increasing charging rates at the charging location within the geofenced region for a defined period to accommodate the accumulated vehicles.

17 . The method of claim 1 , wherein the artificial intelligence system:

processes social media data to identify events affecting transportation needs;

predicts transportation demand associated with the identified events; and

adjusts charging infrastructure parameters based on the predicted transportation demand.

18 . The method of claim 1 , wherein the optimizing the parameter of the charging plan comprises:

processing at least one of: an energy-related consumption, a demand, an availability, or access information;

optimizing vehicle electricity usage based on the processed at least one of: the energy-related consumption, the demand, the availability, or the access information; and

optimizing at least one of: a recharging time, a location, or an amount of energy based on the optimization of the vehicle electricity usage.

19 . The transportation system of claim 8 , wherein the artificial intelligence system transitions to a deep reinforcement learning mode after initial training, wherein the deep reinforcement learning mode:

attempts variations in charging approaches through genetic programming techniques; and

tracks outcomes with feedback to exceed initial training performance, wherein the genetic programming techniques modify at least one of the first neural network or the second neural network based on the feedback.

20 . The transportation system of claim 8 , wherein the optimizing the at least one charging plan parameter comprises:

processing at least one of: an energy-related consumption, a demand, an availability, or access information;

optimizing vehicle electricity usage based on the processed at least one of: the energy-related consumption, the demand, the availability, or the access information; and

optimizing at least one of: recharging time, location, or amount based on the optimization of the vehicle electricity usage.

21 . The method of claim 16 , wherein temporarily increasing charging rates further comprises:

calculating an optimal temporary rate based on at least one of: a number of the accumulated vehicles, battery states of the accumulated vehicles, or predicted arrival times of additional vehicles; and

automatically reverting to standard charging rates when the accumulation dissipates as determined by continued real-time monitoring.

22 . The transportation system of claim 19 , wherein the genetic programming techniques comprise:

generating a plurality of variations in charging parameter configurations;

evaluating the plurality of variations based on charging infrastructure performance metrics including at least one of: charging throughput, energy efficiency, or wait time reduction; and

selecting configurations that improve performance relative to the initial training performance.