Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Jul 2;15(1):23005.
doi: 10.1038/s41598-025-05904-2.

Deep reinforcement learning enhanced PID control for hydraulic servo systems in injection molding machines

Xiaoxi Hao  1 , Zengmiao Xin  2 , Weizhuo Huang  1 , Sicheng Wan  3 , Guangfan Qiu  4 , Tianlei Wang  5 , Zhu Wang  6
Affiliations

Deep reinforcement learning enhanced PID control for hydraulic servo systems in injection molding machines

Xiaoxi Hao et al. Sci Rep. .

Abstract

To address the issue of insufficient position control accuracy in the servo-hydraulic system of injection molding machines under nonlinear characteristics and external disturbances, this paper proposes a novel adaptive PID control strategy enhanced by the Deep Deterministic Policy Gradient (DDPG) algorithm. An auxiliary servo valve is introduced to improve flow capacity and enhance the system's dynamic response performance. Meanwhile, the DDPG algorithm is utilized to adjust the PID parameters in real time based on tracking errors and system state feedback, thereby improving the controller's adaptability to time-varying operating conditions. Compared with traditional control methods, the proposed strategy demonstrates superior tracking accuracy, faster convergence, and stronger robustness. In particular, this work innovatively integrates the DDPG algorithm with an auxiliary servo valve structure for PID parameter optimization and dynamic performance enhancement, offering new ideas and technical pathways for adaptive control of complex hydraulic systems.

Keywords: Adaptive control; DDPG; Intelligent control; Position control; Reinforcement learning; Servo hydraulic system.

PubMed Disclaimer

Conflict of interest statement

Declarations. Competing interests: The authors declare no competing interests.

Figures

Fig. 1
Fig. 1
Schematic diagram of a dual-valve position servo hydraulic system.
Fig. 2
Fig. 2
Main block diagram of the DDPG-PID control strategy.
Fig. 3
Fig. 3
Full workflow of the DDPG-PID approach.
Algorithm 1
Algorithm 1
Pseudo-code of the DDPG-PID control strategy.
Fig. 4
Fig. 4
AMESim co-simulation model of the servo hydraulic system.
Fig. 5
Fig. 5
Simulink model of the DDPG-PID control strategy.
Fig. 6
Fig. 6
The neural network structure layer defined by DDPG.
Fig. 7
Fig. 7
Training convergence plot of the DDPG algorithm.
Fig. 8
Fig. 8
Fuzzy logic control.
Fig. 9
Fig. 9
Hydraulic cylinder step displacement signal response.
Fig. 10
Fig. 10
Hydraulic cylinder sine displacement signal response.
Fig. 11
Fig. 11
Tracking error density comparison over 10 s simulation.
Fig. 12
Fig. 12
Multi-stage displacement signal response.
Fig. 13
Fig. 13
Tracking performance under parameter uncertainty.
Fig. 14
Fig. 14
Comparison of DDPG-tuned PID and fixed-gain PID controllers: ablation study under sinusoidal trajectory.
Fig. 15
Fig. 15
Sinusoidal signal ablation experiment with and without auxiliary valve.
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Le, V. D. et al. Computational study on the clamping mechanism in the injection molding machine. Int. J. Adv. Manuf. Technol.121(11), 7247–7261 (2022).
    1. Ren, Z. et al. A learning-based model predictive control scheme for injection speed tracking in injection molding process. Complex Intell. Syst.10(6), 7845–7861 (2024).
    1. Wang, W. et al. Output feedback control for energy-saving asymmetric hydraulic servo system based on desired compensation approach. Appl. Math. Model.101, 360–379 (2022).
    1. Ren, Z. et al. Dynamic optimal control of flow front position in injection molding process: A control parameterization-based method. J. Process Control132, 103125 (2023).
    1. Hu, H. et al. Practical adaptive robust tracking control of the dual-valve parallel electro-hydraulic servo system with reduced-order model. Trans. Inst. Meas. Control.46(3), 501–512 (2024).

LinkOut - more resources