沥青路面养护智能检测与决策综述

Review on intelligent detection and decision-making of asphalt pavement maintenance

通过疏理沥青路面养护智能检测与决策应用领域、研究现状和发展趋势,分析沥青路面检测与监测技术工作原理与研究进展,总结目前各种路面病害智能识别技术特点和存在的问题,针对沥青路面智能检测和智能养护决策手段所面临的挑战和发展趋势进行探究与展望。分析结果表明:可视化方法、雷达、光纤光栅、声学技术和振动技术等手段在沥青路面检测与监测方面得到广泛应用,提高了路面检测的效率和精度,但相关传感器在耐久性、系统性和智能化等方面仍存在一些问题;人工智能和大数据分析已广泛应用于沥青路面智能养护决策,基于人工智能算法的路面平整度、抗滑性能、车辙和裂缝等问题的智能识别也得到长足发展;将来多种病害智能识别集成技术可用于对路面进行全过程、全时空的病害监测,还可以进一步对道路服役性能、道路养护策略优化等领域进行全自动高效分析,但在智能决策训练样本、深度学习算法等方面仍需开展大量研究。

Through research on the application,research status and development trend of intelligent detection and decision-making of asphalt pavement maintenance,the working principle and research progress of asphalt pavement detection and monitoring technology were summarized,the characteristics and existing problems of intelligent identification technology of various pavement diseases were analyzed,and the challenges and development of intelligent detection and decision-making methods of asphalt pavement maintenance in the future were pointed out.The results show that the visualization method,radar,fiber Bragg grating,acoustic technology,vibration technology et al are widely used in the detection and monitoring of asphalt pavement,which improves the efficiency and accuracy of pavement detection,but the related sensors need to be more durable,compatible and smart.The intelligent recognition of common diseases such as pavement flatness,skid resistance,rutting and cracks based on artificial intelligence algorithm have also made great progresses.In the future,the integrated technology of intelligent recognition of various diseases is expected to perform disease monitoring in the whole process and all of the time.Artificial intelligence and large data analysis methods can be widely used in intelligent maintenance decision-making of asphalt pavement,which can be used for road disease recognition and road maintenance,road service performance evaluation,road maintenance strategy optimization and so on,but a lot of research work still needs to be carried out in intelligent decision-making training samples,deep learning algorithm and so on.