IP Library › Granted Patent US 12,534,994
Granted Patent B2
US 12,534,994 · App. 18/477,656 · Granted Jan 27, 2026

Drilling control

Inventors: Yingwei Yu (Katy, TX); Velizar Vesselinov (Sugar Land, TX); Richard Meehan (Houston, TX); Qiuhua Liu (Houston, TX); Wei Chen (Houston, TX); Minh Trang Chau (Sugar Land, TX); Yuelin Shen (Houston, TX); Sylvain Chambon (Clamart, FR)
Assignee: Schlumberger Technology Corporation
E21B44/00E21B7/04E21B47/024
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 12,534,994
App. No.
18/477,656
Granted
Jan 27, 2026
Kind
B2
Abstract

A system and method that include receiving sensor data during drilling of a portion of a borehole in a geologic environment. The system and method also include selecting a drilling mode from a plurality of drilling modes based at least on a portion of the sensor data. The system and method additionally include simulating drilling of the borehole using the selected drilling mode in a multi-dimensional spatial environment to generate a simulated state of the borehole in the geologic environment. The system and method further include analyzing the simulated state of the borehole and generating a reward using the simulated state of the borehole and a planned borehole trajectory and using the reward to train an agent to provide automated directional drilling with transitions between at least one of: a plurality of drilling modes and a plurality of toolface settings.

Claims (56)

1 . A method, comprising:

receiving sensor data by a computing system during drilling of a portion of a borehole in a geologic environment;

selecting a drilling mode from a plurality of drilling modes based at least on a portion of the sensor data, the plurality of drilling modes including: a rotary drilling mode and a sliding drilling mode;

simulating, by the computing system, drilling of the borehole using the selected drilling mode in a multi-dimensional spatial environment to generate a simulated state of the borehole in the geologic environment;

analyzing, by the computing system, the simulated state of the borehole;

generating, by the computing system, a reward using the simulated state of the borehole and a planned borehole trajectory;

by the computing system, using the reward to train an agent to provide automated directional drilling with transitions between at least one of: the plurality of drilling modes and a plurality of sliding drilling mode downhole motor toolface settings; and

by the computing system, responsive to the provided automated directional drilling, drilling an additional portion of the borehole in accordance with the provided automated directional drilling with the transitions.

2 . The method of claim 1 , wherein selecting the drilling mode includes defining a coordinate system for a portion of a drillstring using at least a portion of the sensor data.

3 . The method of claim 1 , wherein simulating drilling of the borehole includes simulating using a hole propagation method based on commands using at least one of the plurality of drilling modes.

4 . The method of claim 1 , wherein:

the sliding drilling mode includes at least one of: a sliding up drilling mode and a sliding down drilling mode; and

the agent is trained to provide automated directional drilling with transitions between the sliding up drilling mode, the sliding down drilling mode, and the rotary drilling mode.

5 . The method of claim 1 , wherein:

the plurality of sliding drilling mode downhole motor toolface settings includes a plurality of toolface angles; and

the agent is trained to provide automated directional drilling with transitions between the plurality of toolface angles.

6 . The method of claim 1 , wherein the reward is determined at each interval as to a measured distance along an axis of a trajectory of the planned borehole trajectory.

7 . The method of claim 1 , wherein the reward:

is evaluated based on accuracy with respect to the planned borehole trajectory; and

is output as a positive reward or a negative reward based on the accuracy.

8 . The method of claim 7 , wherein the generating the reward includes receiving a planned survey as an input and comparing it with actual drilled point locations that are output as part of the simulated state of the borehole to return a scalar that pertains to the accuracy with respect to the planned borehole trajectory.

9 . The method of claim 1 , wherein the reward penalizes transitions in sliding drilling mode downhole motor toolface settings.

10 . A system, comprising:

one or more processors;

memory accessible by the one or more processors; and

processor-executable instructions stored in the memory and executable by the one or more processors to instruct the system to:

receive sensor data by a computing system during drilling of a portion of a borehole in a geologic environment;

select a drilling mode from a plurality of drilling modes based at least on a portion of the sensor data, the plurality of drilling modes including: a rotary drilling mode and a sliding drilling mode;

simulate, by the computing system, drilling of the borehole using the selected drilling mode in a multi-dimensional spatial environment to generate a simulated state of the borehole in the geologic environment;

analyze, by the computing system, the simulated state of the borehole;

generate, by the computing system, a reward using the simulated state of the borehole and a planned borehole trajectory;

by the computing system, using the reward to train an agent to provide automated directional drilling with transitions between at least one of: the plurality of drilling modes and a plurality of sliding drilling mode downhole motor toolface settings; and

by the computing system, responsive to the provided automated directional drilling, drill an additional portion of the borehole in accordance with the provided automated directional drilling with the transitions.

11 . The system of claim 10 , wherein the instructions to select the drilling mode include defining a coordinate system for a portion of a drillstring using at least a portion of the sensor data.

12 . The system of claim 10 , wherein the instructions to simulate drilling of the borehole include simulating using a hole propagation method based on commands using at least one of the plurality of drilling modes.

13 . The system of claim 10 , wherein:

the sliding drilling mode includes at least one of: a sliding up drilling mode and a sliding down drilling mode; and

the agent is trained to provide automated directional drilling with transitions between the sliding up drilling mode, the sliding down drilling mode, and the rotary drilling mode.

14 . The system of claim 10 , wherein:

the plurality of sliding drilling mode downhole motor toolface settings includes a plurality of toolface angles; and

the agent is trained to provide automated directional drilling with transitions between the plurality of toolface angles.

15 . The system of claim 10 , wherein the reward is determined at each interval as to a measured distance along an axis of a trajectory of the planned borehole trajectory.

16 . The system of claim 10 , wherein the reward is evaluated based on accuracy with respect to the planned borehole trajectory and is output as a positive reward or a negative reward based on the accuracy.

17 . The system of claim 16 , wherein the generating the reward includes receiving a planned survey as an input and comparing it with actual drilled point locations that are output as part of the simulated state of the borehole to return a scalar that pertains to the accuracy with respect to the planned borehole trajectory.

18 . The system of claim 10 , wherein the reward penalizes transitions in sliding drilling mode downhole motor toolface settings.

19 . A non-transitory computer-readable storage medium storing instructions that, when executed by a computer, which includes one or more processors, perform a method, the method comprising:

receiving sensor data by a computing system during drilling of a portion of a borehole in a geologic environment;

selecting a drilling mode from a plurality of drilling modes based at least on a portion of the sensor data, the plurality of drilling modes including: a rotary drilling mode and a sliding drilling mode;

simulating, by the computing system, drilling of the borehole using the selected drilling mode in a multi-dimensional spatial environment to generate a simulated state of the borehole in the geologic environment;

analyzing, by the computing system, the simulated state of the borehole;

generating, by the computing system, a reward using the simulated state of the borehole and a planned borehole trajectory;

by the computing system, using the reward to train an agent to provide automated directional drilling with transitions between at least one of: the plurality of drilling modes and a plurality of sliding drilling mode downhole motor toolface settings; and

by the computing system, responsive to the provided automated directional drilling, drilling an additional portion of the borehole in accordance with the provided automated directional drilling with the transitions.

20 . The non-transitory computer-readable storage medium of claim 19 , wherein the reward:

is evaluated based on accuracy with respect to the planned borehole trajectory; and

is output as a positive reward or a negative reward based on the accuracy.

Continuity (3)
Continuation 16776373 · Jan 29, 2020
Provisional Application 62850865 · May 21, 2019
Related Publication 20240018864A1 · Jan 18, 2024
References Cited (179)
US 2671346A · Banning, Jr. · 1954 [cited by applicant]
US 5305836A · Holbrook · 1994 [cited by applicant]
US 5312163A · Hanamoto · 1994 [cited by applicant]
US 5794720A · Smith · 1998 [cited by applicant]
US 6109368A · Goldman et al. · 2000 [cited by applicant]
US 6233498B1 · King · 2001 [cited by applicant]
US 6269892B1 · Boulton · 2001 [cited by applicant]
US 6353799B1 · Meany · 2002 [cited by applicant]
US 6408953B1 · Goldman · 2002 [cited by applicant]
US 6424919B1 · Moran · 2002 [cited by applicant]
US 6438495B1 · Chau · 2002 [cited by applicant]
US 6516293B1 · Huang et al. · 2003 [cited by applicant]
US 6612382B2 · King · 2003 [cited by applicant]
US 6749029B2 · Alft · 2004 [cited by applicant]
US 6785641B1 · Huang · 2004 [cited by applicant]
US 6873947B1 · Huang et al. · 2005 [cited by applicant]
US 6968909B2 · Aldred · 2005 [cited by applicant]
US 7020597B2 · Oliver · 2006 [cited by applicant]
US 7085696B2 · King · 2006 [cited by applicant]
US 7139689B2 · Huang · 2006 [cited by applicant]
US 7464013B2 · Huang et al. · 2008 [cited by applicant]
US 7546884B2 · Veeningen · 2009 [cited by applicant]
US 7693695B2 · Huang et al. · 2010 [cited by applicant]
US 7844426B2 · Huang · 2010 [cited by applicant]
US 7953586B2 · Chen · 2011 [cited by applicant]
US 8099266B2 · Burge · 2012 [cited by applicant]
US 8121971B2 · Edwards et al. · 2012 [cited by applicant]
US 8401831B2 · Tang et al. · 2013 [cited by applicant]
US 8849460B2 · Ye · 2014 [cited by applicant]
US 9482055B2 · Huang · 2016 [cited by applicant]
US 10221671B1 · Zhang · 2019 [cited by applicant]
US 10400547B2 · Chen · 2019 [cited by applicant]
US 10662763B2 · Alshawaf · 2020 [cited by applicant]
US 11396804B2 · Madasu · 2022 [cited by applicant]
US 11499409B2 · Suryadi · 2022 [cited by applicant]
US 20010042642A1 · King · 2001 [cited by applicant]
US 20020050989A1 · Sanstrom · 2002 [cited by applicant]
US 20020138240A1 · Jelley · 2002 [cited by applicant]
US 20020177955A1 · Jalali · 2002 [cited by applicant]
US 20030015351A1 · Goldman et al. · 2003 [cited by applicant]
US 20040000430A1 · King · 2004 [cited by applicant]
US 20040143427A1 · Huang · 2004 [cited by applicant]
US 20040149431A1 · Wylie · 2004 [cited by applicant]
US 20040256152A1 · Dashevskiy · 2004 [cited by applicant]
US 20050096847A1 · Huang · 2005 [cited by applicant]
US 20050197777A1 · Rodney · 2005 [cited by applicant]
US 20050205300A1 · Mathiszik · 2005 [cited by applicant]
US 20050273302A1 · Huang · 2005 [cited by applicant]
US 20060195307A1 · Huang · 2006 [cited by applicant]
US 20060272861A1 · Hutchinson · 2006 [cited by applicant]
US 20070067147A1 · Huang · 2007 [cited by applicant]
US 20070192071A1 · Huang · 2007 [cited by applicant]
US 20080255817A1 · Pabon · 2008 [cited by applicant]
US 20080262810A1 · Moran et al. · 2008 [cited by applicant]
US 20090132458A1 · Edwards et al. · 2009 [cited by applicant]
US 20090204245A1 · Sustaeta · 2009 [cited by applicant]
US 20100032165A1 · Bailey · 2010 [cited by applicant]
US 20110214878A1 · Bailey · 2011 [cited by applicant]
US 20120024606A1 · Pirovolou · 2012 [cited by applicant]
US 20130127900A1 · Pena · 2013 [cited by applicant]
US 20130186687A1 · Snyder · 2013 [cited by applicant]
US 20140351183A1 · Germain et al. · 2014 [cited by applicant]
US 20150198014A1 · Servaes · 2015 [cited by applicant]
US 20150226052A1 · Samuel et al. · 2015 [cited by applicant]
US 20150330209A1 · Panchal · 2015 [cited by examiner]
US 20150337640A1 · Huang · 2015 [cited by applicant]
US 20160047206A1 · Vempati · 2016 [cited by applicant]
US 20170306740A1 · Song · 2017 [cited by applicant]
US 20170328181A1 · Kristjansson · 2017 [cited by applicant]
US 20180087321A1 · Johnston · 2018 [cited by applicant]
US 20190003297A1 · Brannigan · 2019 [cited by applicant]
US 20190048706A1 · Benson et al. · 2019 [cited by applicant]
US 20190048707A1 · Benson · 2019 [cited by applicant]
US 20190102676A1 · Nazari · 2019 [cited by applicant]
US 20190145240A1 · Benson · 2019 [cited by applicant]
US 20190284909A1 · Spencer · 2019 [cited by applicant]
US 20190330968A1 · Boone · 2019 [cited by applicant]
US 20200104544A1 · Suryadi · 2020 [cited by applicant]
US 20200157887A1 · Alonso · 2020 [cited by applicant]
US 20200370409A1 · Yu · 2020 [cited by applicant]
US 20210082129A1 · Nazari · 2021 [cited by applicant]
US 20210148213A1 · Madasu · 2021 [cited by applicant]
US 20210230998A1 · Chen · 2021 [cited by applicant]
US 20210312332A1 · Panchal · 2021 [cited by examiner]
US 20230124120A1 · Yu · 2023 [cited by applicant]
AU 2011283192A1 · 2013 [cited by applicant]
CA 3026315A1 · 2017 [cited by applicant]
CA 3061860A1 · 2020 [cited by applicant]
CA 3106973C · 2023 [cited by applicant]
CN 104040376A · 2014 [cited by applicant]
CN 104508240A · 2015 [cited by applicant]
CN 106437513A · 2017 [cited by applicant]
EA 013694B1 · 2010 [cited by applicant]
EP 1146200A1 · 2001 [cited by applicant]
EP 1335108A2 · 2003 [cited by applicant]
EP 1644800A2 · 2006 [cited by applicant]
EP 2222937A2 · 2010 [cited by applicant]
EP 2222937B1 · 2014 [cited by applicant]
EP 2951720A1 · 2015 [cited by applicant]
EP 3334897A1 · 2018 [cited by applicant]
EP 3465282A1 · 2019 [cited by applicant]
EP 3715579A1 · 2020 [cited by applicant]
GB 2290330A · 1995 [cited by applicant]
GB 2367843A · 2002 [cited by applicant]
WO 0050735A1 · 2000 [cited by applicant]
WO 03072907A1 · 2003 [cited by applicant]
WO 2009058635A2 · 2009 [cited by applicant]
WO 2010101473A1 · 2010 [cited by applicant]
WO 2011010099A1 · 2011 [cited by applicant]
WO 2012015894A1 · 2012 [cited by applicant]
WO 2012170273A1 · 2012 [cited by applicant]
WO 2013192524A2 · 2013 [cited by applicant]
WO 2015200259A1 · 2015 [cited by applicant]
WO 2016022388A1 · 2016 [cited by applicant]
WO 2016032640A1 · 2016 [cited by applicant]
WO 2016154723A1 · 2016 [cited by applicant]
WO 2017030756A1 · 2017 [cited by applicant]
WO 2017206157A1 · 2017 [cited by applicant]
WO 2017206158A1 · 2017 [cited by applicant]
WO 2018117890A1 · 2018 [cited by applicant]
WO 2020028298A1 · 2020 [cited by applicant]
WO 2020046351A1 · 2020 [cited by applicant]
WO 2020172447A1 · 2020 [cited by applicant]
WO 2020191360A1 · 2020 [cited by applicant]
DeepMind, “Safety-first Al for autonomous data centre cooling and industrial control,” retrieved at https://deepmind.com/blog/safety-first-ai-autonomous-data-centre-cooling-and-industrial-control/ on Jan. 27, 2020. [cited by applicant]
Winih et al., “Asynchronous Methods for Deep Reinforcement learning,” Proceedings of the 33rd International Conference on Machine Learning, New York, NY, USA, 2016. [cited by applicant]
Winih et al., “Human-level control through deep reinforcement learning,” Nature vol. 518, pp. 529, Feb. 26, 2015. [cited by applicant]
OpenAI, “Learning Dexterity,” retrieved at https://blog.openai.com/learning-dexterity/ on Jan. 27, 2020. [cited by applicant]
Pollock et al., “Machine Learning for Improved Directional Drilling,” Offshore Technology Conference held in Houston, TX USA, Apr. 30-May 3, 2018. [cited by applicant]
Yu et al., Approaches to Directional Drilling Using Machine Learning, U.S. Appl. No. 62/850,865, filed May 21, 2019. [cited by applicant]
Office Action received in U.S. Appl. No. 16/776,373 dated Mar. 18, 2021, 20 pages. [cited by applicant]
International Search Report and Written Opinion of International Patent Application No. PCT/US2020/015529 mailed on Jun. 1, 2020, 10 pages. [cited by applicant]
Ahmed Hamza, Mohamed Shamlooh, Ibnelwaleed A. Hussein, Mustafa Nasser, Saeed Salehi, Polymeric formulations used for loss circulation materials and wellbore strengthening applications in oil and gas wells: A review, Jou… [cited by applicant]
Cayeux, Eric, Daireaux, Benoit, and Erik Wolden Dvergsnes. “Automation of Drawworks and Topdrive Management to Minimize Swab/Surge, and Poor-Downhole_Condition Effects”, SPE Drilling & Completions 26 (2011): 557-568. do… [cited by applicant]
Hegde, Chiranth, Wallace, Scott, and Ken Gray. “Real Time Prediction and Classification of Torque and Drag During Drilling Using Statistical Learning Methods”, Paper presented at the SPE Eastern Regional Meeting, Morgan… [cited by applicant]
Lashari, S. et al., “Drilling performance monitoring and optimization: a data-driven approach”, Journal of Petroleum Exploration and Production Technology, 2019, 9(4), pp. 2747-2756. [cited by applicant]
Tompkins, M. F. “Optimization Techniques for Task Allocation and Scheduling in Distributed Multi-Agent Operations”, Jun. 2003, MIT, 107 pages. [cited by applicant]
Liu, H. et al., “A Reinforcement Based 3D Guided Drilling Method: Beyond Grund Control”, in the Proceedings of the 2018 VII International Conference on Network Communication and Computing (ICNCC), Association for Comput… [cited by applicant]
Kim et al., “Recent Advances in Multi-agent Reinforcement Learning”, Communications of the Korean Institute of Information Scientists and Engineers, 2019, 37(11), pp. 8-17. [cited by applicant]
Office Action issued in U.S. Appl. No. 17/177,897 dated Jan. 11, 2024, 29 pages. [cited by applicant]
International Search Report and Written Opinion for the equivalent International Patent Application PCT/CN2017/082848 mailed on Jan. 19, 2018. [cited by applicant]
Office Action issued in U.S. Appl. No. 16/497,792 dated Jun. 24, 2021, 18 pages. [cited by applicant]
Guria, C. et al., “Multi-objective optimization of oil well drilling using elitist non-dominated sorting genetic algorighm”, Petroleum Science, 2014, 11(1), pp. 97-110. [cited by applicant]
Office Action issued in U.S. Appl. No. 16/497,792 dated Mar. 22, 2022, 36 pages. [cited by applicant]
Advisory Action issued in U.S. Appl. No. 16/497,792 dated Jul. 25, 2022, 5 pages. [cited by applicant]
Search Report and Written Opinion of International Patent Application No. PCT/US2022/045200 dated Jan. 12, 2023, 10 pages. [cited by applicant]
Booth, SPE/IADC 67816 Meeting Future Drilling Planning and Decision Support Requirements Copyright 2001, Feb. 27-Mar. 1, 2001. [cited by applicant]
Bilgesu, H.I., et al., A New Approach for the Prediction of Rate of Penetration (ROP) Values, Society of Petroleum Engineers, Oct. 1997, pp. 175-179, SPE 39231. [cited by applicant]
Sinor, A. and Warren, T.M., Drag Bit Wear Model, SPE Drilling Engineering, Jun. 1989, pp. 128-136. [cited by applicant]
Irrangang, R., et al., A Case-Based System to Cut Drilling Costs, Society of Petroleum Engineers, Oct. 1999, pp. 1-17, SPE 56504. [cited by applicant]
Luo, Y, et al., Flow-Rate Predictions for Cleaning Deviated Wells, IADC/SPE, Feb. 1992, pp. 367-376, SPE 23884. [cited by applicant]
Bourgoyne, Jr., A., et al., Drilling Hydraulics, Applied Drilling Engineering, 1986, pp. 113-189, First Printing, Society of Petroleum Engineers, Richardson, TX. [cited by applicant]
International Search Report issued in PCT Application No. PCT/US2005/009119 dated Jul. 6, 2005, 3 pages. [cited by applicant]
Written Opinion issued in PCT Application No. PCT/US2005/009119 dated Jul. 6, 2005, 5 pages. [cited by applicant]
Office Action issued in U.S. Appl. No. 10/802,545 dated May 22, 2006, 6 pages. [cited by applicant]
Office Action issued in U.S. Appl. No. 10/802,545 dated Nov. 13, 2006, 6 pages. [cited by applicant]
Office Action issued in U.S. Appl. No. 10/802,545 dated Oct. 6, 2008, 8 pages. [cited by applicant]
Notice of Allowance issued in U.S. Appl. No. 10/802,545 dated Apr. 6, 2009, 6 pages. [cited by applicant]
International Search Report and Written Opinion issued in corresponding International Application No. PCT/US2014/034053, mailed Aug. 18, 2014 (18 pages). [cited by applicant]
First Office Action and Search Report issued in Chinese Patent Application No. 201480020646.4 on Jul. 31, 2017, 18 pages. [cited by applicant]
Office Action issued in U.S. Appl. No. 14/252,726 dated May 8, 2017, 16 pages. [cited by applicant]
Office Action issued in U.S. Appl. No. 14/252,726 dated Dec. 13, 2017, 10 pages. [cited by applicant]
Kingma, Diederik et al., Adam: A Method for Stochastic Optimization, arXiv:1412.6980, v9, Jan. 30, 2017, 15 pages. [cited by applicant]
Ba et al., Positive Displacement Motor Modeling: Skyrocketing the Way We Design, Select, and Operate Mud Motors, Society of Petroleum Engineers, SPE-183298-MS, Abu Dhabi International Petroleum Exhibition & Conference, … [cited by applicant]
Evangelatos, G. I. et al., “Advanced BHA-5 Rop Modeling Including Neural Network Analysis of Drilling Performance Data”, IADC/SPE Drilling Conference and Exhibition, 2016, 24 pages. [cited by applicant]
Wardana, R., “Build Rate Prediction Using Artificial Neural Network for Positive Displacement Motor Application in Field X”, Journal Petro, 2018, VII(1), 6 pages. [cited by applicant]
Office Action issued in U.S. Appl. No. 16/408,514 dated Sep. 24, 2021, 24 pages. [cited by applicant]
Office Action issued in U.S. Appl. No. 16/408,514 dated Mar. 21, 2022, 30 pages. [cited by applicant]
Office Action issued in U.S. Appl. No. 16/497,792 dated May 1, 2024, 35 pages. [cited by applicant]
Bello et al., “Application of artificial intelligence methods in drilling system design and operations: a review of the state of the art”, Journal of Artificial Intelligence and Soft Computing Research 5.2 (2015): 121-1… [cited by applicant]
Lowe et al., “Multi-agent actor-critic for mixed coopertive-competitive environments”, Advances in neural information processing systems, 30. (Year: 2017). [cited by applicant]
Iqbal and Sha, “Actor-attention-critic for multi-agent reinforcement learning”, In International conference on machine learning (pp. 2961-2970). PMLR. (Year: 2019). [cited by applicant]
Schlumberger, Energy Glossary, Definition of the term “dogleg”, accessed electronically on Aug. 12, 2025. URL: glossary(dot)slb(dot)com/en/terms/d/dogleg (Year: 2025). [cited by applicant]
Okafor, Z. et al., “Application of Tandem Rotary Steerable-Positive Displacement Motor Bottom Hole Assembly in Drilling Horizontal Wells: Case Study of Three Eastern Siberia Wells”, SPE/IADC 140241, Presented at the SPE… [cited by applicant]
Dvoynikov, M, V., “Research On Technical and Technological Parameters of Inclined Drilling”, Journal of Mining Institute, 2017, 26(223), pp. 86-92. [cited by applicant]
Yu, Y. et al., “Training an automated directional drilling agent with deep reinforcement learning in a simulated environment”, SPE/IADC-204105-MS, prepared for presentation at the SPE/IADC Drilling Conference and Exhibi… [cited by applicant]
Wu, K. et al., “Learn to Navigate Autonomously Through Deep Reinforcement Learning”, IEEE Transactions on industrial Electronics, 2021, 69(5), pp. 5342-5352. [cited by applicant]
Rieker, J. D. et al., “An intelligent agent for optimal river-reservoir system management”, Water Resources Research, 2021, 48(9), 16 pages. [cited by applicant]
Lin, Y.-C. et al., “Tactics of Adversarial Attack on Deep Reinforcement Learning Agents”, arXiv: 1703.06748v4, 2019, 7 pages. [cited by applicant]