Abstract
Classic reinforcement learning algorithms generate experiences by the agent's constant trial and error, which leads to a large number of failure experiences stored in the replay buffer. As a result, the agents can only learn through these low-quality experiences. In the case of multi-agent systems, this problem is more serious. MADDPG (Multi-Agent Deep Deterministic Policy Gradient) has achieved significant results in solving multi-agent problems by using a framework of centralized training with decentralized execution. Nevertheless, the problem of too many failure experiences in the replay buffer has not been resolved. In this paper, we propose HMADDPG (Hindsight Multi-Agent Deep Deterministic Policy Gradient) to mitigate the negative impact of failure experience. HMADDPG has a hindsight unit, which allows the agents to reflect and produces pseudo experiences that tend to succeed. Pseudo experiences are stored in the replay buffer, so that the agents can combine two kinds of experiences to learn. We have evaluated our algorithm on a number of environments. The results show that the algorithm can guide agents to learn better strategies and can be applied in multi-agent systems which are cooperative, competitive, or mixed cooperative and competitive.
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs13042-022-01505-x/MediaObjects/13042_2022_1505_Fig1_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs13042-022-01505-x/MediaObjects/13042_2022_1505_Fig2_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs13042-022-01505-x/MediaObjects/13042_2022_1505_Fig3_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs13042-022-01505-x/MediaObjects/13042_2022_1505_Fig4_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs13042-022-01505-x/MediaObjects/13042_2022_1505_Fig5_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs13042-022-01505-x/MediaObjects/13042_2022_1505_Fig6_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs13042-022-01505-x/MediaObjects/13042_2022_1505_Fig7_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs13042-022-01505-x/MediaObjects/13042_2022_1505_Fig8_HTML.png)
Similar content being viewed by others
References
Luo F, Dong Z, Liang G, Murata J, Xu Z (2019) A distributed electricity trading system in active distribution networks based on multi-agent coalition and blockchain. IEEE Trans Power Syst 34:4097–4108
Sallab AE, Abdou M, Perot E, Yogamani SK (2017) Deep reinforcement learning framework for autonomous driving. Electron Imaging 2017(19):70–76
Silver D, Huang A, Maddison CJ, Guez A, Sifre L, van den Driessche G, Schrittwieser J, Antonoglou I, Panneershelvam V, Lanctot M, Dieleman S, Grewe D, Nham J, Kalchbrenner N, Sutskever I, Lillicrap TP, Leach M, Kavukcuoglu K, Graepel T, Hassabis D (2016) Mastering the game of go with deep neural networks and tree search. Nature 529(7587):484–489
Yang W, Wang X, Farhadi A, Gupta A, Mottaghi R (2019) Visual semantic navigation using scene priors. In: 7th international conference on learning representations
Wu C, Kreidieh A, Vinitsky E, Bayen AM (2017) Emergent behaviors in mixed-autonomy traffic. In: 1st annual conference on robot learning, vol 78, pp 398–407
Liu S, Lever G, Merel J, Tunyasuvunakool S, Heess N, Graepel T (2019) Emergent coordination through competition. In: 7th international conference on learning representations
Lowe R, Wu Y, Tamar A, Harb J, Abbeel P, Mordatch I (2017) Multi-agent actor-critic for mixed cooperative-competitive environments. In: 30th conference on neural information processing systems. pp 6379–6390
Lin LJ (1992) Self-improving reactive agents based on reinforcement learning, planning and teaching. Mach Learn 8:293–321
Mnih V, Kavukcuoglu K, Silver D, Rusu AA, Veness J, Bellemare MG, Graves A, Riedmiller MA, Fidjeland A, Ostrovski G, Petersen S, Beattie C, Sadik A, Antonoglou I, King H, Kumaran D, Wierstra D, Legg S, Hassabis D (2015) Human-level control through deep reinforcement learning. Nat 518(7540):529–533
Schaul T, Quan J, Antonoglou I, Silver D (2016) Prioritized experience replay. In: 4th international conference on learning representations
Horgan D, Quan J, Budden D, Barth-Maron G, Hessel M, van Hasselt H, Silver D (2018) Distributed prioritized experience replay. In: 6th international conference on learning representations
Luo J, Li H (2019) Dynamic experience replay
Andrychowicz M, Crow D, Ray A, Schneider J, Fong R, Welinder P, McGrew B, Tobin J, Abbeel P, Zaremba W (2017) Hindsight experience replay. In: 30th conference on neural information processing systems. pp 5048–5058
Fang M, Zhou T, Du Y, Han L, Zhang Z (2019) Curriculum-guided hindsight experience replay. In: 32nd conference on neural information processing systems. pp 12602–12613
Liu H, Trott A, Socher R, **ong C (2019) Competitive experience replay. In: 7th international conference on learning representations
Bai C, Liu P, Zhao W, Tang X (2019) Guided goal generation for hindsight multi-goal reinforcement learning. Neurocomputing 359:353–367
Lai Y, Wang W, Yang Y, Zhu J, Kuang M (2020) Hindsight planner. In: AAMAS
Ren Z, Dong K, Zhou Y, Liu Q, Peng J (2019) Exploration via hindsight goal generation. In: NeurIPS
de Villiers B, Sabatta D (2020) Hindsight reward sha** in deep reinforcement learning. In: 2020 international SAUPEC/RobMech/PRASA conference, pp 1–7
Prianto E, Kim M, Park JH, Bae JH, Kim JS (2020) Path planning for multi-arm manipulators using deep reinforcement learning: soft actor-critic with hindsight experience replay. Sensors (Basel, Switzerland) 20:5911
Zuo G, Zhao Q, Lu J, Li J (2020) Efficient hindsight reinforcement learning using demonstrations for robotic tasks with sparse rewards. Int J Adv Robot Syst 17
Littman ML (1994) Markov games as a framework for multi-agent reinforcement learning. In: Machine learning, proceedings of the eleventh international conference. pp 157–163
Tampuu A, Matiisen T, Kodelja D, Kuzovkin I, Korjus K, Aru J, Aru J, Vicente R (2017) Multiagent cooperation and competition with deep reinforcement learning. PLoS ONE 12
Sukhbaatar S, Szlam A, Fergus R (2016) Learning multiagent communication with backpropagation. In: 29th conference on neural information processing systems. pp 2244–2252
Pesce E, Montana G (2020) Improving coordination in small-scale multi-agent deep reinforcement learning through memory-driven communication. Mach Learn 109(9–10):1727–1747
Iqbal S, Sha F (2019) Actor-attention-critic for multi-agent reinforcement learning. In: Proceedings of the 36th international conference on machine learning, vol 97, pp. 2961–2970
Foerster JN, Farquhar G, Afouras T, Nardelli N, Whiteson S (2018) Counterfactual multi-agent policy gradients. In: Proceedings of the thirty-second AAAI conference on artificial intelligence, pp 2974–2982
Omidshafiei S, Pazis J, Amato C, How JP, Vian J (2017) Deep decentralized multi-task multi-agent reinforcement learning under partial observability. In: Proceedings of the 34th international conference on machine learning, vol 70, pp 2681–2690
Lillicrap TP, Hunt JJ, Pritzel A, Heess N, Erez T, Tassa Y, Silver D, Wierstra D (2016) Continuous control with deep reinforcement learning. In: Bengio Y, LeCun Y (eds) 4th international conference on learning representations, ICLR 2016, San Juan, Puerto Rico, May 2–4, 2016, conference track proceedings
Acknowledgements
This work is supported by the National Natural Science Foundation of China (Grant no. 61872260).
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Rights and permissions
About this article
Cite this article
Li, C., Wang, L. & Huang, Z. Hindsight-aware deep reinforcement learning algorithm for multi-agent systems. Int. J. Mach. Learn. & Cyber. 13, 2045–2057 (2022). https://doi.org/10.1007/s13042-022-01505-x
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s13042-022-01505-x