Abstract: In the complex and variable deep-sea environment, the compensation control of ship motion ensures the safety and efficiency of equipment installation and transportation in offshore wind farms. However, the ship motion posture compensation control system is severely affected by uncertainties, which significantly impact the accuracy of compensation control. In this paper, we propose a ship three-degree-of-freedom (3-DoF) motion posture stabilization control method based on the DTW-LSTM-MATD3 algorithm. We use the multi-agent twin delayed deep deterministic policy gradient (MATD3) to control a platform with six electric cylinders to achieve stable control. However, owing to random noise affecting the ship’s motion posture, we use a dynamic time warping (DTW) algorithm to distinguish between high-frequency noise and low-frequency tracking signals. Further, we embed a long short-term memory (LSTM) network into the MATD3 network to better align the Critic network’s training with the true Q-value. We use a combined reward function to enhance the agent’s exploration capability in complex dynamic environments. Finally, verification was conducted under sixth-level, abrupt sea conditions with high-frequency noise, as well as under real abrupt sea conditions, and a generalization test was also carried out. Simulation results show that the proposed DTW-LSTM-MATD3 method has great compensation control ability.

基于DTW-LSTM-MATD3的船舶高低频干扰下平台三自由度运动姿态稳定控制

[1]CauzM, WyrschN, PerretL, et al., 2025. Embracing wind power in the solar PV-dominated Swiss landscape. Energy Reports, 13:3341-3350.

[2]HouY, HanGJ, ZhangF, et al., 2024. Distributional soft actor-critic-based multi-AUV cooperative pursuit for maritime security protection. IEEE Transactions on Intelligent Transportation Systems, 25(6):6049-6060.

[3]JimohIA, KüçükdemiralIB, BevanG, 2021. Fin control for ship roll motion stabilisation based on observer enhanced MPC with disturbance rate compensation. Ocean Engineering, 224:108706.

[4]KangJC, SunLP, Guedes SoaresC, 2019. Fault tree analysis of floating offshore wind turbines. Renewable Energy, 133:1455-1467.

[5]LiC, MogollónJM, TukkerA, et al., 2022. Environmental impacts of global offshore wind energy development until 2040. Environmental Science & Technology, 56(16):11567-11577.

[6]LiuXX, YinY, SuYZ, et al., 2022. A multi-UCAV cooperative decision-making method based on an MAPPO algorithm for beyond-visual-range air combat. Aerospace, 9(10):563.

[7]LvYY, LiH, 2023. Strong fixed-time dynamic inverse adaptive LQR integrated control strategy for dynamic positioning of ship. Ocean Engineering, 288:115969.

[8]MöllerströmE, GipeP, OttermoF, 2025. Wind power development: a historical review. Wind Engineering, 49(2):499-512.

[9]MouZY, ZhangY, GaoFF, et al., 2021. Deep reinforcement learning based three-dimensional area coverage with UAV swarm. IEEE Journal on Selected Areas in Communications, 39(10):3160-3176.

[10]NathwaniJ, KammenDM, 2019. Affordable energy for humanity: a global movement to support universal clean energy access. Proceedings of the IEEE, 107(9):1780-1789.

[11]QinYH, ZhangZS, LiXL, et al., 2023. Deep reinforcement learning based resource allocation and trajectory planning in integrated sensing and communications UAV network. IEEE Transactions on Wireless Communications, 22(11):8158-8169.

[12]ShaoS, LiuHW, ZhangL, et al., 2022. Integration of super-resolution ISAR imaging and fine motion compensation for complex maneuvering ship targets under high sea state. IEEE Transactions on Geoscience and Remote Sensing, 60:5222820.

[13]TangG, LeiJM, LiFR, et al., 2023. A modified 6-DOF hybrid serial–parallel platform for ship wave compensation. Ocean Engineering, 280:114336.

[14]WangWX, NingYH, ZhangY, et al., 2025. Linear active disturbance rejection control with linear quadratic regulator for Stewart platform in active wave compensation system. Applied Ocean Research, 156:104469.

[15]WangYF, ZhaoY, 2025. Multiple ships cooperative navigation and collision avoidance using multi-agent reinforcement learning with communication. Ocean Engineering, 320:120244.

[16]WinursitoA, PratamaGNP, 2021. LQR state feedback controller with precompensator for magnetic levitation system. Journal of Physics: Conference Series, 2111(1):012004.

[17]WoodacreJK, BauerRJ, IraniR, 2018. Hydraulic valve-based active-heave compensation using a model-predictive controller with non-linear valve compensations. Ocean Engineering, 152:47-56.

[18]WuLS, ZhangC, ZhangB, et al., 2025. Toward energy-efficiency: integrating MATD3 reinforcement learning method for computational offloading in RIS-aided UAV-MEC environments. IEEE Internet of Things Journal, 12(14):26582-26595.

[19]YanF, FanK, YanXC, et al., 2020. Constant tension control of hybrid active-passive heave compensator based on adaptive integral sliding mode method. IEEE Access, 8:103782-103791.

[20]ZhangQ, DingZY, ZhangMJ, 2020. Adaptive self-regulation PID control of course-keeping for ships. Polish Maritime Research, 27(1):39-45.

[21]ZhangYH, LiGX, TianY, et al., 2025. Model-free reinforcement learning-based transient power control of vehicle fuel cell systems. Applied Energy, 388:125614.

[22]ZhaoEY, ZhouN, LiuCJ, et al., 2024. Time-aware MADDPG with LSTM for multi-agent obstacle avoidance: a comparative study. Complex & Intelligent Systems, 10(3):4141-4155.

[23]ZhouYT, KongXR, LinKP, et al., 2024. Novel task decomposed multi-agent twin delayed deep deterministic policy gradient algorithm for multi-UAV autonomous path planning. Knowledge-Based Systems, 287:111462.