IP Library › Granted Patent US 12,276,953
Granted Patent B2
US 12,276,953 · App. 17/826,335 · Granted Apr 15, 2025

Multi-market resource dispatch optimization

Inventors: Peeyush Kumar (Seattle, WA); Lucien Werner (Pasadena, CA); Shivkumar Kalyanaraman (Bangalore, IN); Srinivasan Iyengar (Bangalore, IN); Weiwei Yang (Redmond, WA); Tanuja Hrishikesh Ganu (Bangalore, IN); Ranveer Chandra (Kirkland, WA); Riyaz Pishori (Sammamish, WA); Upendra Singh (Redmond, WA)
Assignee: MICROSOFT TECHNOLOGY LICENSING, LLC
G05B19/042G05B2219/2639
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,276,953
App. No.
17/826,335
Granted
Apr 15, 2025
Kind
B2
Abstract

The techniques disclosed herein enable systems to enable multi-market optimization of renewable energies using data-driven models. To achieve this, a model retrieves a current state from a resource generation system and associated resource markets. The model can then compute a policy based on the state as well physical and technical constraints. The policy defines various actions that direct operation of the resource generation system such as resource production and dispatch to markets. Applying the policy to the resource generation results in a modified state which the model extracts along with a measure of optimality which quantifies the success of the policy. Based on these metrics, the model can generate an updated iteration of the policy defining a different set of actions. In this way, the model can gradually develop an optimal policy for controlling the resource generation system.

Claims (36)

1. A method comprising:

retrieving, using a processing system comprising one or more processing units, data defining a current state from an environment containing a resource generation system, a resource purchase agreement market operating on a first timescale comprising years, and a frequency regulation service market operating on a second timescale comprising seconds;

determining, using a model, a policy that is applied to the resource generation system in the environment based on the current state;

configuring the resource generation system using the policy;

dispatching a resource of the resource generation system to the resource purchase agreement market and the frequency regulation service market in accordance with the policy and one or more external variables affecting a timing and a quantity of the resource that is dispatched to the resource purchase agreement market and the frequency regulation service market;

retrieving update data defining a modified state from the environment, the modified state resulting from dispatching the resource to the resource purchase agreement market and the frequency regulation service market in accordance with the policy;

calculating an optimality based on the modified state, the optimality quantifying a level of success of the policy with respect to the resource purchase agreement market and the frequency regulation service market; and

modifying, by the model, one or more actions of the policy to increase the optimality.

2. The method of claim 1 , wherein the resource is electrical energy and wherein the resource generation system is a renewable energy system.

3. The method of claim 1 , wherein a rate at which the resource generation system generates the resource varies over time.

4. The method of claim 1 , wherein the resource purchase agreement market and the frequency regulation service market comprise a first service type and a second service type, respectively.

5. The method of claim 1 , wherein the one or more external variables comprises a current resource price and a resource generation rate.

6. A system comprising:

a processing system comprising one or more processing units; and

computer storage media having encoded thereon computer-readable instructions that when executed by the processing system, causes the system to:

retrieve data defining a current state from an environment containing a resource generation system, a resource purchase agreement market operating on a first timescale comprising years, and a frequency regulation service market operating on a second timescale comprising seconds;

determine, using a model, a policy that is applied to the resource generation system in the environment based on the current state;

configure the resource generation system using the policy;

dispatch a resource of the resource generation system to the resource purchase agreement market and the frequency regulation service market in accordance with the policy and one or more external variables affecting a timing and a quantity of the resource that is dispatched to the resource purchase agreement market and the frequency regulation service market;

retrieve update data defining a modified state from the environment, the modified state resulting from dispatching the resource to the resource purchase agreement market and the frequency regulation service market in accordance with the policy;

calculate an optimality based on the modified state, the optimality quantifying a level of success of the policy with respect to the resource purchase agreement market and the frequency regulation service market; and

modify, by the model, one or more actions of the policy to increase the optimality.

7. The system of claim 6 , wherein a rate at which the resource generation system generates the resource varies over time.

8. The system of claim 6 , wherein the resource purchase agreement market and the frequency regulation service market comprise a first service type and a second service type, respectively.

9. The system of claim 6 , wherein the one or more external variables comprises a current resource price and a resource generation rate.

10. A computer-readable storage media, having encoded thereon computer-readable instructions that when executed by one or more processing units, cause a system to:

retrieve data defining a current state from an environment containing a resource generation system, a resource purchase agreement market operating on a first timescale comprising years, and a frequency regulation service market operating on a second timescale comprising seconds;

determine, using a model, a policy that is applied to the resource generation system in the environment based on the current state;

configure the resource generation system using the policy;

dispatch a resource of the resource generation system to the resource purchase agreement market and the frequency regulation service market in accordance with the policy and one or more external variables affecting a timing and a quantity of the resource that is dispatched to the resource purchase agreement market and the frequency regulation service market;

retrieve update data defining a modified state from the environment, the modified state resulting from dispatching the resource to the resource purchase agreement market and the frequency regulation service market in accordance with the policy;

calculate an optimality based on the modified state, the optimality quantifying a level of success of the policy with respect to the resource purchase agreement market and the frequency regulation service market; and

modify, by the model, one or more actions of the policy to increase the optimality.

11. The computer-readable storage media of claim 10 , wherein a rate at which the resource generation system generates the resource varies over time.

12. The computer-readable storage media of claim 10 , wherein the resource purchase agreement market and the frequency regulation service market comprise a first service type and a second service type, respectively.

13. The computer-readable storage media of claim 10 , wherein the one or more external variables comprises a current resource price and a resource generation rate.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 12, 2022
From: KUMAR, PEEYUSH; WERNER, LUCIEN; KALYANARAMAN, SHIVKUMAR; IYENGAR, SRINIVASAN; YANG, WEIWEI; GANU, TANUJA HRISHIKESH; CHANDRA, RANVEER; PISHORI, RIYAZ; SINGH, UPENDRA
To: MICROSOFT TECHNOLOGY LICENSING, LLC
Reel/Frame 060486/0622 →
Priority Claims (1)
IN 202241003868 · Jan 24, 2022 · national
Continuity (1)
Related Publication 20230236559A1 · Jul 27, 2023
References Cited (22)
US 8311681B1 · Marcus · 2012 [cited by examiner]
US 20120278220A1 · Chassin · 2012 [cited by examiner]
US 20160273518A1 · Chen · 2016 [cited by examiner]
US 20190325533A1 · Perry · 2019 [cited by examiner]
US 20190385182A1 · Price · 2019 [cited by examiner]
“Wärtsilä is delivering a 70MW energy storage system in California to optimise renewable energy system performance and maximise ROI”, Retrieved from: https://www.wartsila.com/media/news/29-04-2020-wartsila-is-delivering… [cited by applicant]
Arulkumaran, et al., “Deep reinforcement learning: A brief survey”, In Journal of IEEE Signal Processing Magazine, vol. 34, Issue 6, Nov. 2017, pp. 26-38. [cited by applicant]
Cao, et al., “Deep Reinforcement Learning-Based Energy Storage Arbitrage With Accurate LithiumIon Battery Degradation Model”, In Journal of IEEE Transactions on Smart Grid, vol. 11, Issue 5, Sep. 2020, pp. 4513-4521. [cited by applicant]
Chen, et al., “Enforcing Policy Feasibility Constraints through Differentiable Projection for Energy Optimization”, In Repository of arXiv:2105.08881v1, May 19, 2021, 12 Pages. [cited by applicant]
Chen, et al., “Reinforcement learning for decision-making and control in power systems: Tutorial, review, and vision”, In Repository of arXiv:2102.01168v1, Jan. 27, 2021, pp. 1-16. [cited by applicant]
Mbuwir, et al., “Battery Energy Management in a Microgrid Using Batch Reinforcement Learning”, In Journals of Energies, vol. 10, Issue 11, Nov. 12, 2017, 19 Pages. [cited by applicant]
Oh, et al., “Reinforcement-Learning-Based Energy Storage System Operation Strategies to Manage Wind Power Forecast Uncertainty”, In Journal of IEEE Access, vol. 8, Jan. 23, 2020, pp. 20965-20976. [cited by applicant]
Powell, Warren B. , “Clearing the Jungle of Stochastic Optimization”, In Journal of Informs, Sep. 2014, pp. 109-137. [cited by applicant]
Powell, Warren B. , “Reinforcement learning and stochastic optimization”, In Publication of A John Wiley & Sons, Inc, Jul. 22, 2021, 51 Pages. [cited by applicant]
Powell, et al., “Tutorial on Stochastic Optimization in Energy—Part I: Modeling and Policies”, In Journal of IEEE Transactions on Power Systems , vol. 31, Issue 2, Mar. 2016, pp. 1459-1467. [cited by applicant]
Shang, et al., “Stochastic dispatch of energy storage in microgrids: An augmented reinforcement learning approach”, In Journal of Applied Energy, vol. 256, Issue 4, Mar. 1, 2020, pp. 1-11. [cited by applicant]
Wang, et al., “Energy storage arbitrage in real-time markets via reinforcement learning”, In Proceedings of IEEE Power & Energy Society General Meeting (PESGM), Aug. 5, 2018, 6 Pages. [cited by applicant]
Willuhn, Marian, “RWE abandons plans for 1.1 GW lignite-fired power station in Germany”, Retrieved from: https://www.pv-magazine.com/2019/04/26/rwe-abandons-plans-for-1-1-gw-lignite-fired-power-station-in-germany/, Apr.… [cited by applicant]
Yang, et al., “A deep reinforcement learning method for managing wind farm uncertainties through energy storage system control and external reserve purchasing”, In International Journal of Electrical Power & Energy Syst… [cited by applicant]
Yang, et al., “Reinforcement learning in sustainable energy and electric systems: a survey”, In Journal of Annual Reviews in Control, vol. 49, Apr. 9, 2020, pp. 1-19. [cited by applicant]
Zhang, et al., “Deep reinforcement learning for power system applications: An overview”, In CSEE Journal of Power and Energy Systems, vol. 6, Issue 1, Mar. 30, 2020, pp. 213-225. [cited by applicant]
Zheng, et al., “Safe Reinforcement Learning of Control-Affine Systems with Vertex Networks”, In Proceedings of the 3rd Conference on Learning for Dynamics and Control, May 29, 2021, pp. 1-12. [cited by applicant]