Apparatus and method for operating energy storage system

Patent metadata

Field	Value
Publication number	US-12449778-B2
Application number	US-202217994071-A
Country	US
Kind code	B2
Filing date	Nov 25, 2022
Priority date	Dec 30, 2021
Publication date	Oct 21, 2025
Grant date	Oct 21, 2025

Abstract

The present disclosure relates to an operating device and method of an ESS. The ESS operating method may include forecasting electricity information during a predetermined period using a deep learning model generated based on data about an electricity price and an electricity demand, deriving an ESS operating policy by a reinforcement learning model based on the forecasted electricity information and state information of an energy storage device included in the ESS, and controlling the ESS based on the derived ESS operating policy.

First claim

What is claimed is:

1. An ESS operating method performed by an operating device of an energy storage system (ESS), comprising:
forecasting electricity information for a first period using a deep learning model generated based on data about an electricity price and an electricity demand;
deriving an ESS operating policy by a reinforcement learning model based on the electricity information and state information of an energy storage device included in the ESS; and
controlling the ESS based on the derived ESS operating policy,
wherein the deriving of an ESS operating policy includes:
as the ESS operating policy, deriving the electricity to be supplied to the load of the electricity customer by buying an electricity from a grid operator in a period other than an electricity demand peak period according to the forecasted electricity information to charge the energy storage device and selling the electricity from the energy storage device to an electricity customer during the peak period to discharge an electricity stored in the energy storage device, and
as the ESS operating policy, further deriving to buy an electricity transaction stimulus together when the electricity is bought from the grid operator and sell the electricity transaction stimulus to the grid operator whenever the electricity is sold to the electricity customer.

2. The ESS operating method according to claim 1 , wherein the electricity information includes an electricity demand forecasted during the first period, an electricity price, and elasticity of the electricity demand with respect to the electricity price.

3. The ESS operating method according to claim 1 , wherein the state information of the energy storage device includes a state of charging, a charging efficiency, and a discharging efficiency of the energy storage device.

4. The ESS operating method according to claim 1 , wherein the deriving of the electricity to be supplied to the load of the electricity customer includes:
buying the electricity with a price below a first set value in a period other than the peak period and selling the electricity with a price above a second set value during the peak period.

5. The ESS operating method according to claim 1 , wherein a price of the electricity transaction stimulus is determined based on the elasticity of the electricity demand with respect to the electricity price at the time of transaction.

6. The ESS operating method according to claim 1 , wherein the deriving of an ESS operating policy includes:
as the ESS operating policy, deriving to perform the charging or discharging of the energy storage device such that the charging amount of the energy storage device of the ESS does not deviate from the predetermined range.

7. A non-transitory computer readable recording medium in which a program which is executed by at least one processor to causes the at least one processor to perform the method according to claim 1 is recorded.

8. An operating device of an energy storage system (ESS), comprising:
at least one processor; and
a memory which is operably connected to the processor and stores at least one code executed in the processor,
wherein the memory stores a code which is executed by the processor to cause the processor to forecast electricity information for a first period using a deep learning model generated based on data about an electricity price and an electricity demand, derive an ESS operating policy by a reinforcement learning model based on the forecasted electricity information and state information of an energy storage device included in the ESS, and
control the ESS based on the derived ESS operating policy,
wherein the memory stores a code which causes the processor to as the ESS operating policy, derive the electricity to be supplied to the load of the electricity customer by buying an electricity from a grid operator in a period other than an electricity demand peak period according to the forecasted electricity information to charge the energy storage device, and selling the electricity from the energy storage device to an electricity customer during the peak period to discharge an electricity stored in the energy storage device, and
wherein the memory stores a code which causes the processor to as the ESS operating policy, further derive to buy an electricity transaction stimulus together when the electricity is bought from the grid operator, and sell the electricity transaction stimulus to the grid operator whenever the electricity is sold to the electricity customer.

9. The ESS operating device according to claim 8 , wherein the electricity information includes an electricity demand forecasted during the first period, an electricity price, and elasticity of the electricity demand with respect to the electricity price.

10. The ESS operating device according to claim 8 , wherein the state information of the energy storage device includes a state of charging, a charging efficiency, and a discharging efficiency of the energy storage device.

11. The ESS operating device according to claim 8 , wherein a price of the electricity transaction stimulus is determined based on the elasticity of the electricity demand with respect to the electricity price at the time of transaction.

Assignees

• Korea Inst Energy Res

Inventors

• Han Gwang Woo
• Kim Jong Kyu

Classifications

• G06N3/092

  Reinforcement learning · CPC title

• G05B2219/2639

  Energy management, use maximum of cheap power, keep peak load low · CPC title

• G06N3/0442

  characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU] · CPC title

• Y04S20/222

  Demand response systems, e.g. load shedding, peak shaving · CPC title

• G06Q50/06

  Energy or water supply · CPC title