This application belongs to the field of
smart water technology and relates to a pipeline
leak detection method. The method includes: dividing the
water supply network into zones to obtain target monitoring areas; deploying
capsule robots to the target monitoring areas for dynamic inspection to obtain target survey data; the
capsule robots are equipped with at least a
pressure sensor, an ultrasonic flow sensor, and a camera; performing multi-
source data fusion
processing based on the target survey data and the target monitoring areas to obtain the target pipeline leak location; using the
capsule robots to accurately locate the target pipeline leak location to obtain pipeline leak coordinates; performing
leak detection based on the pipeline leak coordinates to obtain leak information; and uploading the leak information to a maintenance platform for leak alarm. This application combines macroscopic monitoring of DMA zones with
microscopic inspection by capsule robots to achieve sub-meter-level positioning of leak points, reducing the
false alarm rate of
leak detection and improving inspection efficiency.