US 12,214,691 B2
Systems for machine learning, optimizing and managing local multi-asset flexibility of distributed energy storage resources
Simon Richard Daniel, London (GB); and Christopher Verity Wright, Surrey (GB)
Assigned to Moixa Energy Holdings Limited, (GB)
Appl. No. 15/734,705
Filed by MOIXA ENERGY HOLDINGS LIMITED, London (GB)
PCT Filed Jun. 20, 2019, PCT No. PCT/EP2019/066382
§ 371(c)(1), (2) Date Dec. 3, 2020,
PCT Pub. No. WO2019/243524, PCT Pub. Date Dec. 26, 2019.
Claims priority of application No. 1810314 (GB), filed on Jun. 22, 2018.
Prior Publication US 2021/0221247 A1, Jul. 22, 2021
Int. Cl. B60L 53/68 (2019.01); B60L 53/67 (2019.01); G05B 15/02 (2006.01); G06N 3/044 (2023.01); G06N 3/08 (2023.01); H02J 3/14 (2006.01); H02J 13/00 (2006.01)
CPC B60L 53/68 (2019.02) [B60L 53/67 (2019.02); G05B 15/02 (2013.01); G06N 3/044 (2023.01); G06N 3/08 (2013.01); H02J 3/144 (2020.01); H02J 13/00002 (2020.01)] 20 Claims
OG exemplary drawing
 
1. A system for managing and optimizing an energy network, the system comprising:
a plurality of distributed end devices, wherein the plurality of distributed end devices includes a first distributed end device and a second distributed end device;
energy resources including electric vehicle (EV) batteries and/or other energy storage batteries;
a hardware processor; and
a memory coupled with the hardware processor, wherein the memory is configured to provide the hardware processor with instructions which when executed cause the hardware processor to:
receive data and monitor usage of the plurality of distributed end devices and the energy resources at plural remote sites in an energy network, wherein the hardware processor and the plurality of distributed end devices and energy resources are configured to exchange data via a communication network;
process external data and market signals;
manage energy usage of the energy resources responsive to predictions of energy usage, comprising to:
input into a model a time series of measurements indicative of energy usage or activity in the system, wherein the model is a recurrent neural network, wherein the recurrent neural network is trained based on real data, wherein for the first distributed end device, the time series of measures is input into a first model, wherein for the second distributed end device, the time series of measures is input into a second model, and wherein the first model is different from the second model:
identify, based on the model, a time or occupancy dependent mode of use of the system;
output a scaler real-time value representing one or more properties associated with the mode of use, being one or more of the device or mode type, start-time of the event or mode, time and power load duration expectation, wherein the outputting of the scaler real-time value comprises to:
validate the scaler real-time value using the recurrent neural network; and
calculate flexibility in the energy network over a time period based at least in part on the scalar value; and
coordinate how the flexibility in the energy resources, can be scheduled, shared or orchestrated to determine a battery charging plan for charging and/or discharging batteries at the remote sites that delivers an identified amount of flexibility to the energy network.