Dynamic terrain slope measurement suffers from strong environmental disturbances and delayed positioning feedback, which lead to coupling errors between measurement and trajectory tracking. These coupling errors directly limit the reliability of dynamic slope acquisition in long-duration autonomous operations and reduce the effectiveness of real-time positioning correction in complex terrain environments. To address this problem, this paper proposes a collaborative optimization framework for slope measurement and positioning tracking based on the Deep Deterministic Policy Gradient algorithm, which performs continuous control through an Actor-Critic architecture driven by multi-source time series state information. The method integrates slope correction, speed adjustment, and sensor parameter regulation into a unified continuous action space, and introduces slope measurement error as a negative feedback signal to guide policy learning under dynamic disturbances. Experimental results under multiple terrain and disturbance conditions show that the proposed method achieves an average MAE of 1.5° and an RMSE of 1.9°, with stable trajectory deviation, limited correction amplitude, and a total control loop delay of 17.5 ms, demonstrating reliable collaborative control accuracy and real-time performance. The proposed framework is oriented toward dynamic terrain surveying, mobile robotic inspection, and autonomous ground measurement systems that require synchronized measurement precision and positioning stability.
Gao et al. (2026) studied this question.