针对面向智能交通的行业信息物理系统（CPS），以公路冰雪灾害防治作为应用目标，基于信息物理融合理论设计了公路主动式防冰处置的CPS模型，形成由一系列跨越信息世界、物理世界的节点相互耦合而成的巨型分布式反馈控制系统。围绕该CPS模型，提出了基于热力学规律的 路面温度短时预测算法，并基于风险理论设计了主动式防冰控制策略。对实测数据的实验分析和实际路段的系统应用表明，由于在预测算法中考虑了空气温度与路面温度的耦合变化关系，该 CPS 具有较高的预测精度，30min 的平均路面温度预测误差约为±0.2℃；该CPS 的融冰处置能够覆盖高速公路约3/4 的有效行车区域，可以预先避免凝冰和积雪的形成或是主动降低其危害性。

Utilized with the intelligent transportation-oriented industrial Cyber-Physical System (CPS), and aiming at the prevention of snow and ice disasters for highway, a CPS model of proactive anti-icing disposal for highway was established based on information & physics fusion theory. This model was composed by a series of computational and physical nodes crossing cyber world and physical world, thus forming a giant, distributive feedback control system. On the basis of the CPS model, a short-term prediction algorithm of pavement temperature based on thermodynamic exchange laws and a proactive antiicing control strategy based on risk theory were devised respectively. The experimentation with actual meteorological data and the demonstration application of this CPS system show that: due to the coupling variation of air temperature and pavement temperature in the prediction algorithm, this system can predict icing event with rather high precision, the average prediction error of which is around ±0.2℃ within 30-min period; the ice-melting results that almost 3/4 of the major traffic lanes area are free of icing or slightly affected are also appreciable.