超音波水位計於道路側溝排水設施監測與預警技術開發

Abstract

水利署近年來建構韌性臺灣，並積極強化早期預警以及即時應變、防災避災等功能。為了達到上述目標，水利署推動水情監控之網路系統，可於第一手時間獲取即時的水情資訊，以便做出應變。淹水感測器以及影像辨識等都是這幾年水利署推動之防災作為，其中淹水感測器更是精準掌握淹水災情的絕佳利器。然目前水利署採用的淹水感測器為路面淹水感測器，往往偵測到淹水時已緩不濟急，無法達到未雨綢繆、防患未然之功能。 鑒於側溝水深為最即時且最優先的排水設施，本研究以防水超音波測距模組作為主體，結合國產DSI 2598+開發板，且通過溫度校正、探測試驗、干擾及耐久度試驗，了解超音波測距於側溝水位探測的可行性後，將儀器架設於大安區雲和街之側溝，以大安森林公園之雨量站數據為主，國立臺灣大學雨量站數據為輔，觀察側溝水位與即時10min降雨強度間之關聯性。分析觀測到的五組數據，得知側溝之集流時間約為7分鐘，同時水位與雨量呈線性關係，水深約為當下10min降雨強度之2.6倍(僅適用本研究採用之側溝)。將此一關係式套回實際10min雨量以及水文設計應用手冊的SSGM一小時北二區設計雨型，所預估之峰值水位與實測水位峰值誤差均小於4公分，為可接受範圍內。同時根據當下水深以及水位上升率，訂定四種警戒分級供參考，並提出了預警系統之發布構想，解決了目前淹水感測無法提供預警及警報之功能。

In recent years, Water Resources Agency has built a resilient Taiwan and actively strengthened early warning system, immediate response, disaster prevention and avoidance. In order to achieve the above goals, Water Resources Agency promotes the network system of water regime monitoring, which can obtain real-time water regime information at first-hand time, so as to respond accordingly. Flooding sensors and image recognition are all disaster prevention measures promoted by Water Resources Agency in recent years. Among them, the flooding sensor is an outstanding tool for accurately grasping the situation of flooding. However, the flooding sensor currently used by Water Resources Agency is a road water level sensor, which can only detect water level while flooding, and cannot prevent disasters before flooding occur. In view of the fact that the water depth of the road gutter is the most immediate and the most intuitive physical quantity that reflects the rainfall intensity, this research mainly based on waterproof ultrasonic sensor， combined with the demo board DSI 2598+ which is fully made in Taiwan. The ultrasonic ranging system has also passed the temperature calibration test, detection range test, interference and durability test. After evaluating the feasibility of the ultrasonic ranging system road gutter water level detection, which is feasible, the instrument was set up in the road gutter in Yunhe Street, Daan District. Based on the data from the rainfall station in Daan Forest Park and supplemented by the data from the rainfall station in National Taiwan University. By analyzing the five sets of observed data, it is known that the concentration time of the road gutter is about 7 minutes, while the relationship between road gutter water level and 10min rainfall intensity is linear, which is about 2.6 times the current 10-minute rainfall intensity. We can also apply the relationship between road gutter water level and 10min rainfall intensity into the actual 10-minute rainfall intensity and one-hour SSGM design storm pattern, the error between the estimated peak water level and the measured peak water level are both less than 4 cm, which is within an acceptable range. Meanwhile, according to the current water depth and water level rising rate, four warning levels are applied, and an early warning system concept is proposed, which solves the problem that current road flooding sensor cannot provide early warnings and flooding alarm.