The invention discloses a
bulk density analysis-based container
coke quality early warning method, which comprises the following steps of: acquiring data such as container size, weighing gross weight /
tare weight, historical
net weight of suppliers, size fraction and the like, calculating volume and
bulk density of a container body, counting average
net weight in a short period of time according to the suppliers, and generating a
net weight early warning value; the method comprises the following steps: calculating an actually measured net weight in real time and comparing the actually measured net weight with an early warning value during factory weighing, automatically giving an alarm when the net weight exceeds a threshold value, classifying and grouping, issuing targeted sampling detection, writing a detection result back to a
database to correct an
early warning model, and forming an alarm-detection-feedback
closed loop. Full-coverage rapid
primary screening can be achieved for each box in the weighing link, the problems of insufficient coverage and detection
lag of traditional sampling are remarkably solved,
granularity abnormity can be recognized more accurately, suppliers are prevented from avoiding inspection through a boxing means, the
response time is shortened, and the quality risk and the inspection cost are reduced. The method is suitable for a container
coke scene with full container delivery as a main part, and early warning parameters can be dynamically adjusted according to historical data or a more complex model.