應用智慧型手機檢測道路平整度之演算邏輯

Abstract

近年來，道路平整度受到越來越多公務機關及道路養護單位之重視，但主要檢測儀器如三米直規量測速度過慢，耗費人力物力，因此使用者已經普遍放棄使用三米直規作為檢測工具。而現階段普遍使用之平整度檢測儀器以慣性式剖面儀為主，可以在任何地區、任何車型下量測得到穩定均一之國際糙度指標IRI。但受限於其量測過程繁瑣，需要專業車輛與設備才可進行，同時檢測費用亦相對高昂，因此無法對所有道路均進行IRI量測。在此背景下，隨著智慧型手機之使用越來越普及，同時智慧型手機兼顧使用者多用途需要，搭載之各類感測器功能日趨完善，因此使用手機進行各類探測亦成為世界各大學之研究方向。以手機搭載之三軸加速度規、陀螺儀與GPS等感測器作為數據蒐集方法，通過手機跟隨車輛震動方式，對路面震動狀況進行量測。 而目前針對手機之研究均發現其與國際糙度指標IRI有較好之相關性，因此本文延續各類研究中普遍所使用之加速度均方根指標（ARI）作為與IRI比較之指標，同時以IRI作為基準指標進行比較，以尋找ARI與IRI相關性提升之方法。通過對加速度數據進行濾波處理，同時進行速度校正，可以使ARI與IRI之相關性達到最佳0.86以上。在此基礎上，考量手機日後之用途為開放給民眾使用之問題，通過使用手機三軸加速度之方法，解決不同民眾手機擺放角度之問題，確保手機在任何擺放狀態下均可得到均一結果。而在後期大量數據蒐集處理過程中，由於手機廠牌、車型、擺放位置等各類組合均會導致量測結果不一，針對此情況進行大數據處理邏輯研擬，確定量測道路起始點GPS後再進行數據分析之邏輯，並對不同類別數據進行分類處理，最終根據其數據差異進行線性調整，使不同手機之數據可以得到統一結果。在數據調整過程中，亦考慮異常數據處理與非連續路段數據整合之問題，使手機數據經過大數據處理後會得到與IRI更好之相關性。

In recent years, the measurement of road roughness has received more and more attention from public agencies. However, the main testing equipment such as the initial profiler, which can measured the international roughness index IRI, is commonly used at this stage, However, due to the cumbersome measurement process, which requires professional vehicles and equipment to be carried out, and the relatively high cost of testing, it is impossible to measure the IRI on all roads. thus, the use of mobile phones for road roughness detection has also become a new tendency. The smartphone is equipped with three-axis accelerometers, gyroscopes, and GPS detectors to collect the vehicle vibration data relate to the road roughness, to measure the road vibration conditions. The current research on mobile phones has found that it has a good correlation with the IRI. Therefore, this article continues to use the acceleration root mean square index (ARI) which is commonly used in various researches as an index to compare with IRI. In this research, the relevance of ARI and IRI has been improved to 0.86 through filtering. Based on this, considering the future use of smartphones as crowdsourcing for the public, tri-axial acceleration of mobile phones are used to solve the problem of different people's mobile phone placement angles. In the process of large-scale data collection, big data processing logic is developed to determine the start of measuring roads., the logic of data analysis is performed with GPS location, and different types of data are classified, then the finally linear adjustment is performed according to the ANOVA analysis of different data, so that the data from different smartphones can be unified. In the process of data adjustment, the data integration of abnormal data processing and discontinuous road sections is also considered, so that after the big data processing, the smartphones’ data will get a better correlation with IRI.