为提升自动驾驶汽车在全天候条件下的弯道行驶安全性，解决现有安全车速模型对复杂环境因素与车辆运动耦合关系考虑不足的问题，提出一种全天候条件下弯道安全车速计算模型。首先，在四轮车辆动力学模型基础上，引入摩擦椭圆约束机制，建立包含道路坡度、超高、摩擦系数及车辆纵向加速度的轮胎横纵向力非线性耦合关系，构建了安全车速计算模型。进一步，通过仿真分析不同道路参数和水膜、冰膜厚度下的安全车速变化规律，并探究了纵向加减速行为对侧滑失稳临界速度的制约机理。最后，将该模型输出的安全车速作为模型预测控制（Model Predictive Control, MPC）的速度规划约束，开展弯道速度规划仿真，进一步验证其在自动驾驶决策中的应用效果。研究结果表明，与传统质点模型、NHTSA模型和Lusetti模型相比，本文模型在低附着及复杂工况下的计算精度显著提升，与仿真实验结果的均方根误差控制在2.84 km/h以内，在此基础上形成的速度规划能够有效降低弯道侧滑风险。研究结果可为自动驾驶汽车的决策规划提供更为精准的理论支撑。

To enhance the cornering safety of autonomous vehicles under all-weather conditions and address the insufficient consideration for the coupling between complex environmental factors and vehicle motion in existing safe speed models, this paper proposes a safe cornering speed calculation model for all-weather conditions. First, building upon a four-wheel vehicle dynamics model, a friction ellipse constraint mechanism is introduced. This paper establishes a nonlinear coupling relationship between tire lateral and longitudinal forces, incorporating road grade, superelevation, friction coefficient, and vehicle longitudinal acceleration, thereby formulating the safe speed calculation model. Furthermore, through simulation analysis, the variation of safe vehicle speed under different road parameters such as curve radius, road surface friction coefficient, and the thickness of water film and ice film is studied, and the restraining mechanism of longitudinal acceleration and deceleration behavior on the critical speed of sideslip instability is explored. Finally, the safe speed output by the proposed model is used as a constraint for MPC-based speed planning, and simulations of speed planning on curved roads are conducted to further validate its application effectiveness in autonomous driving decision-making. The findings indicate that, compared to traditional particle models, the NHTSA model, and the Lusetti model, the proposed model demonstrates significantly improved computational accuracy under low-adhesion and complex conditions. The root mean square errors between the model outputs and simulation results are controlled within 2.84 km/h. Furthermore, the speed planning strategy developed on this basis effectively mitigates the risk of vehicle sideslip on curved roads, thereby enhancing the safety of autonomous vehicles. The all-weather curve speed calculation model proposed in this study can provide more accurate theoretical support for the decision-making and planning of autonomous vehicles.