The invention relates to the technical field of
civil engineering material detection, and particularly discloses a performance detection method of self-repairing concrete for
civil engineering, which comprises the following steps: acquiring micro-deformation
field data of a crack area in real time in a crack induction process, and constructing a
joint simulation model; predicting an initial contact state between a crack and a
repair material based on the model, extracting
interface pressure distribution and effective bonding
area change information, and calculating an
interface bonding failure risk characteristic value through an improved fuzzy C-means clustering
algorithm; a periodic load is applied to a preset crack area, a closed response
signal is collected, and a multi-scale crack expansion
coupling degree characteristic value is constructed to quantify coordination between microscopic filling and macroscopic
recovery; fusing the two feature values into a composite repair quality
feature vector, inputting the composite repair quality
feature vector into a recognition model based on an improved deep residual network, recognizing and repairing a blind area, and generating a
spatial positioning tag map; and finally, dynamically adjusting the monitoring path and the loading frequency according to an identification result.