纵坡场景下自动驾驶与人工驾驶纵向控制特性差异研究

A Comparative Study of Longitudinal Control Characteristics between Autonomous and Manual Driving under Longitudinal Grade Scenarios

摘要: 为探究自动驾驶车辆在纵坡路段的适应性及其与人工驾驶的差异，本文基于CarSim-PreScan-Simulink联合仿真平台与驾驶模拟实验，构建统一道路场景，对1%、2%和4%坡度条件下两种驾驶模式的纵向控制特性进行对比分析。选取速度、纵向加速度指标，并采用卡尔曼滤波方法对加速度信号进行降噪处理。在此基础上，通过统计分析与显著性检验量化两种驾驶模式的差异特征。结果表明：在1%、2%坡段下，自动驾驶表现出更优的速度稳定性与控制平顺性；随着坡度增大，自动驾驶速度保持能力显著下降。研究揭示了自动驾驶在低坡度条件下具有较好的平顺性优势，但在中大坡度场景中受限于基于反馈的控制策略，易出现调节滞后问题；相比之下，人工驾驶凭借前瞻性调节在复杂纵坡环境中表现出更强的适应性。研究结果可为自动驾驶纵向控制策略优化及坡度适应性提升提供参考。

Abstract: To explore the adaptability of autonomous vehicles on inclined road sections and the differences between autonomous and manual driving, this study constructs a unified road scenario based on the CarSim-PreScan-Simulink joint simulation platform and driving simulation experiments. The longitudinal control characteristics of two driving modes under 1%, 2%, and 4% slope conditions are compared and analyzed. Speed and longitudinal acceleration are selected as indicators, and Kalman filtering is applied to denoise the acceleration signal. Based on this, statistical analysis and significance testing are conducted to quantify the differences between the two driving modes. The results indicate that, under 1% and 2% slope conditions, autonomous driving demonstrates better speed stability and control smoothness. However, as the slope increases, the speed maintenance ability of autonomous driving significantly decreases. The study reveals that autonomous driving has a smoothness advantage under low slope conditions, but in medium and high slope scenarios, the feedback-based control strategy leads to a lag in adjustment. In contrast, manual driving, with its anticipatory adjustments, exhibits stronger adaptability in complex longitudinal slope environments. The research findings provide insights for optimizing autonomous driving longitudinal control strategies and improving slope adaptability.