告别亏电焦虑：AI能源管理或将重塑增程式电动车
增程式电动车在全球电动化进程中长期被视为“折中方案”：它既具备电驱的平顺性与低成本补能优势，又依赖发动机弥补续航焦虑。然而，当电池电量不足时，用户对油耗上升、动力迟滞以及噪声与振动骤增的集体抱怨，成为这一架构最突出的短板。
问题的核心在于能量管理的局限。现有大多数增程式系统依赖电池荷电状态（SOC）的阈值触发发动机介入，缺乏预测能力，只能被动响应实时需求。这使发动机频繁运行在低效区间，能量回收也常常错失最佳时机，导致整体体验显著下降。
广汽近日公布的“星源”增程技术尝试通过AI能源管控平台改善这一局面。依托中央计算平台，该系统将导航数据、实时路况、坡道路形与用户驾驶习惯整合至同一模型，实现预测性调度和个性化优化。例如，车辆可在长上坡前提前储能，或在长下坡前预留能量回收空间，从而保持发动机在高效区间运转。同时，系统能够通过学习驾驶者的通勤规律，形成差异化的补电策略。
在油电切换环节，AI平台预设超过300种功率分配方案，并以微秒级运算匹配车速、功率需求与NVH（噪声、振动、声振粗糙度）参数，使切换过程几乎不可察觉，车辆表现更接近纯电动车。
业内普遍认为，这一尝试凸显了增程式技术的发展方向——其未来突破不再依赖硬件性能，而在于算法和系统整合。随着AI能量管理的引入，增程式电动车有望从“过渡技术”转向可持续的市场选择。
AI Energy Management Could Redefine Range-Extended EVs
Range-extended electric vehicles have long been seen as a compromise—offering smooth electric driving and low charging costs while relying on combustion engines to alleviate range anxiety. Yet when batteries run low, drivers often face higher fuel consumption, sluggish response, and intrusive noise and vibration—persistent drawbacks that have limited wider acceptance.
At the root of the issue lies conventional energy management. Most REEV systems still rely on state-of-charge (SOC) thresholds to trigger the engine, reacting passively to immediate demand. This often pushes the engine into inefficient operating zones and prevents effective energy recovery, eroding overall performance.
China’s GAC recently unveiled its “Xingyuan” range-extender platform, designed to address these shortcomings through AI-enabled energy management. The system integrates navigation data, real-time traffic, road gradients, and driver behavior into predictive and personalized models. For example, it can pre-charge before steep climbs or reserve battery capacity ahead of descents, while tailoring charging strategies to individual commuting patterns.
To ensure seamless transitions, the platform applies over 300 power distribution strategies and conducts microsecond-level calculations of speed, torque demand, and NVH parameters. This allows nearly imperceptible shifts between engine and battery, delivering a driving feel closer to that of a battery-electric vehicle while preserving efficiency.
Industry analysts note that the evolution of REEVs may depend less on hardware gains and more on intelligent systems. With AI now taking a central role in power management, range-extended architectures could move beyond their “transitional” label toward becoming a viable long-term option in global EV markets.