To address the rapid growth of data exchange volume in onboard computing chips driven by the rapid development of intelligent connected vehicles, while satisfying the high-level security protection requirements for in-vehicle data as the most valuable core asset of automotive manufacturers, the paper discusses the impact of current information security standards and algorithms on in-vehicle information security chips and analyzes the replacement path and development trend of national cryptographic algorithms in the field of in-vehicle chips. The results show that as relevant standards for in-vehicle chips are successively implemented, national cryptography level certification will become an industry norm, and in-vehicle information security chips that pass such certification will occupy the mainstream domestic market.

This study investigates the selection of SoC chips and platform architecture design for intelligent assisted driving systems. A systematic analysis is conducted on key factors affecting chip selection, including hardware interfaces, computing requirements, cost optimization, software ecosystems, functional safety, and supply chain risks. Based on this analysis, a low-cost, engineering-oriented intelligent driving system architecture centered on domestic intelligent driving SoC chips is proposed. The proposed architecture achieves an optimized balance between cost efficiency and resource utilization while ensuring system performance and reliability. It provides valuable references for the design of intelligent driving domain controllers, product development, and the localization substitution of automotive semiconductor solutions.

From the perspective of domestic patent applications, a patent analysis is conducted on the technological evolution of the perception module in intelligent driving, based on 2 branches: sensor hardware technology and multi-sensor fusion technology. The study finds that multi-sensor combinations of cameras, LiDAR, and millimeter-wave radar remain the mainstream configuration. Feature-level and decision-level fusion technologies, due to their lower computational resource demands, have become the preferred solutions for the current large-scale deployment of intelligent driving. In the future, performance improvements in perception systems will advance along 3 directions: self-supervised spatiotemporal alignment, lightweight deployment of end-to-end models, and robustness against extreme weather conditions.

Aiming at the low accuracy of State of Charge (SOC) estimation and poor adaptability of control strategies under dynamic working conditions for power batteries of new energy vehicles, a multi-model fusion SOC estimation framework integrating Kalman Filtering (KF), Particle Filtering (PF) and Long Short-Term Memory (LSTM) neural network is constructed. Dynamic charging and discharging management strategies based on genetic algorithm and particle swarm optimization are designed. A power battery test platform covering multiple temperatures and charge-discharge rates is built, and simulations and physical tests are carried out under standard UDDS and WLTC working conditions. The test results show that the root mean square error of SOC estimation of the proposed method is reduced to 0.013 5, the battery charge-discharge efficiency is increased by 9.3%, and the maximum discharge current is reduced by 23.7%. The proposed method can slow down battery capacity attenuation and reduce thermal runaway risk, which provides theoretical support and engineering reference for the optimization of battery management systems.

To address the issue of quantitative data during the concept phase, this paper proposes a quantitative safety goal assessment method based on the ISO 26262 standard. By establishing a mapping model between fault scenario dynamics and injury characteristics, simulations were conducted to quantitatively determine the Severity (S), Controllability (C), and Fault Tolerant Time Interval (FTTI). The results indicate that the FTTI increases with rising time headway but decreases as driver reaction time extends; moreover, high-speed conditions exhibit higher sensitivity to variations in these parameters. Without requiring physical prototypes, this method efficiently generates quantitative metrics through parameter configuration, thereby supporting the precise definition of safety goals in the concept phase.

To address the challenges posed by the rapid evolution of new energy vehicle (NEV) electrical systems toward higher voltages, higher power density, and greater integration, and to mitigate the critical impacts of in-network ripple on overall vehicle electrical safety, Electromagnetic Compatibility (EMC), and core component reliability, this study investigates a mass-produced NEV. Based on the ISO 21498-2: 2024 international standard and an extended testing protocol, a high-voltage ripple test system covering full operating conditions and multiple states was established. Through dynamic real-vehicle testing, the time-domain and frequency-domain characteristics, distribution patterns, and key influencing factors of the ripple were analyzed in depth. The results indicate that the safety factor level of the high-voltage in-network ripple for this vehicle model is high, verifying the excellent stability and robustness of the vehicle's high-voltage system. Furthermore, the study reveals the coupling mechanisms of operating conditions, battery State of Charge (SOC), power electronics switching actions, and system parasitic parameters on the ripple. Systematic ripple suppression and optimization schemes are proposed from 3 dimensions: control strategy, hardware design, and test verification. The research findings provide important theoretical basis and engineering practice references for the forward design, performance optimization, and standard improvement of NEV high-voltage systems.

To address the safety validation challenges faced during the large-scale deployment and industrial application of Automated Driving System (ADS), and to bridge the gaps in existing methods, particularly regarding the systematic behavioral competency modeling, the semantic completeness of test scenarios, the scenario coverage of the Operational Design Domain (ODD), and the validation of boundary capabilities, a framework for behavioral competencies and scenario modeling of ADS is presented. Based on investigations of international technical regulations and automated driving pilot practices, the framework follows the main line of"ODD-behavioral competencies-test scenarios-validation criteria-risk assessment", adopts a functional, abstract, logical and concrete scenario representation, and integrates knowledge-driven, data-driven, and goal-oriented modeling mechanisms to achieve structured and traceable validation paths. For behavioral competency modeling, it covers 4 scenario types: nominal, critical, failure as well as ODD-boundary, and establishes mappings among competencies, scenarios, as well as validation criteria through object-event-response analysis. At the validation level, the formulation logic of performance metrics and validation criteria is clarified to support the assessment of validation coverage and residual risks. The framework can provide methodological references for ADS safety declarations and product compliance assessment.

To address safety challenges arising from dynamic variations in operational capability boundaries, shifts in driving responsibility, and the activation of risk mitigation strategies in automated driving systems under multiple operational modes, this paper proposes a safety argumentation method for operational mode transitions. The method constructs a three-layer finite state machine model of"operational mode-operational state-behavioral response"to formally describe multi-mode operation and transition logic. Furthermore, a three-dimensional verification model of"state machine path, scenario type, safety evidence chain" is proposed to systematically link regulatory safety requirements with engineering design, testing and validation activities, and safety evidence. Typical scenarios are used to demonstrate the complete process from safety requirement decomposition and model construction to validation implementation and safety argumentation, providing an engineering reference for the safety design, regulatory review, and compliance decision-making of multi-mode automated driving systems.

To address the challenge that the traditional static type approval model struggles to accommodate the full lifecycle safety governance of Automated Driving Systems (ADS), which are characterized by continuous software iteration, complex operational environments, and dynamic evolution of AI models, the paper constructs an In-Service Monitoring and Reporting (ISMR) mechanism and its implementation pathway based on institutional analysis and pilot practices. Employing normative analysis and mechanism design methods, a enterprise-side implementation mechanism covering operational data acquisition, data governance, safety analysis, monitoring reporting and safety improvement is proposed. Furthermore, a government-enterprise collaborative governance framework featuring "rule supply, operational execution, and evidence feedback" is established, alongside an event-triggered post-deployment safety assessment framework. The results demonstrate that the proposed mechanism, centered on operational evidence, effectively supports the continuous safety assurance of ADS. It enables the ongoing identification, analysis, and closed-loop mitigation of operational risks while providing a basis for the dynamic updating of safety cases.

To address the safety challenges posed by the global accelerated deployment of Driver Control Assistance Systems (DCAS), this study compares the regulatory paradigms of the European Union, China, and the United States, and reveals that UN R171 transforms general safety principles into quantifiable engineering constraints, thereby driving the evolution of DCAS from discrete feature sets into systematically controlled systems. The findings indicate that high-standard compliance drives firms to develop three core engineering capabilities: deterministic modeling of system boundaries, evidentiary governance across the lifecycle, and data-driven closed-loop optimization. These capabilities constitute the engineering foundation for migration toward Automated Driving Systems (ADS). Case studies reveal that enterprises differentiate into 2 strategic pathways: systematic integration and agile breakthroughs. The proposed regulation-driven capability-building model elucidates the transmission mechanism from regulatory constraints to strategic assets, providing a theoretical framework for transforming compliance costs into long-term competitive advantages. Accordingly, the study offers policy recommendations focusing on implementing capability-oriented compliance within enterprises and fostering a dynamically balanced regulatory governance system.