近年来，在国家的大力推动下，自动驾驶技术取得了蓬勃的发展，逐步进入示范运行阶段。目前，大部分车企研发精力都聚焦于单车的智能水平提升，将单车智能水平的提升作为提高产品溢价的手段。而“单车智能”仍然有较高成本，无法实现大规模的商业化，同时在可用性方面仍然面临诸多问题。相比较而言，“网联智能”中车路协同等技术能够与单车智能的自动驾驶技术形成很好的互补，不存在研发壁垒和资源浪费的问题，以地域为单位积累的数据和不断更新迭代的算法，应用到所有通行于该地区的智能车辆上，相比车企各自为战的迭代速度，网联智能明显更加有优势。因此，本文从产业研究入手，通过对行业及用户需求的研究来分析车路协同的必要性，分析个人消费者与企业用户在高级别自动驾驶技术方面的实际需求，并从经济学角度尝试解决单车与车路协同的成本评估和成本比较问题，同时基于管理学角度，寻求在车路协同实现自动驾驶的场景下，单车智能与车路协同技术组合模式的评估方法论，给出了单车与车路协同自动驾驶技术组合的成本分析及决策模型，帮助助推更早实现“网联协同感知”和“网联协同决策与控制”来降低单车成本，为高级别自动驾驶汽车产业发展提供参考依据。本文研究内容主要包括以下几点：1．建立了不同用户对高级别自动驾驶技术的功能需求模型。结合消费者需求和使用场景，对比分析单车智能与网联智能车路协同两种模式的发展优势及不足，了解不同用户对高级别自动驾驶技术的需求，对目前存在问题进行了分析，并针对个人消费者和企业用户分别建立了高级别自动驾驶技术的功能需求模型。2．构建了单车智能与车路协同技术的成本等效模型，并利用所构建的相关成本模型评价了不同高级别自动驾驶方案的经济性。3．基于帕累托最优理论提出了综合考虑经济性和安全性的最优组合选择方法，建立了结合单车智能与车路协同成本等效模型的组合方案经济性评价机制。相关方法在是实际案例种进行了应用，为高级别自动驾驶方案的选择提供了理论支撑。

In recent years, under the strong promotion of the state, autonomous driving technology has made vigorous development and gradually entered the demonstration operation stage. At present, most car companies‘ R&D efforts focus on the improvement of the intelligence level of vehicles, taking the improvement of the intelligence level of vehicles as a means to increase the product premium. The "independent autonomous driving" is still too costly to be commercialised on a large scale and still faces many problems in terms of usability. In contrast, " intelligent connected vehicle" complements "independent autonomous driving" very well, and does not have the problem of R&D barriers and waste of resources. The data accumulated in the region and the continuously updated and iterative algorithm are applied to all intelligent vehicles passing in the area. Compared with car companies, The iterative speed of each battle, the cooperative vehicle-highway autonomous driving is obviously more advantageous.Therefore, this article starts from the perspective of industry research, analyzing the necessity of vehicle-to-road cooperation through research on the industry and user needs. It examines the actual needs of individual consumers and corporate users in terms of advanced autonomous driving technology. From an economic perspective, it attempts to address the cost evaluation and cost comparison issues of standalone vehicles and vehicle-road cooperation. Additionally, from a management perspective, it seeks to evaluate the combination of standalone vehicle intelligence and vehicle-road cooperation technology in the context of autonomous driving achieved through vehicle-road cooperation. The article provides a cost analysis and decision-making model for the combination of standalone vehicles and vehicle-road cooperation in autonomous driving technology, aiming to facilitate the earlier realization of "networked cooperative perception" and "networked cooperative decision-making and control" to reduce the cost of standalone vehicles, and provide a reference basis for the development of advanced autonomous driving industry.The main research contents of this article include the following points:1.Established functional demand models for advanced autonomous driving technology for different users. By combining consumer demands and usage scenarios, the article compares and analyzes the development advantages and limitations of standalone vehicle intelligence and networked intelligent vehicle-road cooperation. It understands the demands of different users for advanced autonomous driving technology, analyzes the current problems, and establishes functional demand models for advanced autonomous driving technology for individual consumers and corporate users separately.2.Constructed equivalent cost models for standalone vehicle intelligence and vehicle-road cooperation technology, and used the relevant cost models to evaluate the economic feasibility of different advanced autonomous driving solutions.3.Based on the Pareto optimality theory, proposed an optimal combination selection method that considers both economic feasibility and safety, and established an economic evaluation mechanism for combination schemes that combines the equivalent cost models of standalone vehicle intelligence and vehicle-to-road cooperation. The methods were applied in practical cases, providing theoretical support for the selection of advanced autonomous driving solutions.