An effective monitoring method of dynamic compaction construction quality based on time series modeling.

The existing quality monitoring method of dynamic compaction construction is costly and inefficient, and the measurement results for the falling distance of the hammer and the number of tamping times may even become completely ineffective in abnormal situations during the construction process, such as empty hook rise and early release hammer. Therefore, this paper proposes an effective measurement method based on time series model. Simulate the regular periodic motion state transformation in strong ramming construction by motion state discrimination model, combined with the prediction interval of running time estimation model to assist state discrimination, realizes real-time monitoring of anomalies as well as effective measurement of important construction parameters such as the tamping times and the falling distance of the hammer. Experimental results demonstrate that the method achieves accuracy of 100% under normal tamping conditions, and accuracy of no less than 95% even in tamping conditions involving abnormal behavior. • Constructed DCC state discrimination model based on bi-objective parallel features. • Proposed a method measuring tamping times using combined temporal-spatial features. • Established a model for estimating the duration of states based on prior data. [ABSTRACT FROM AUTHOR]