The invention provides a load prediction and diagnosis method based on power distribution network protection and equipment real-
time data, which is characterized by comprising the following steps of: setting and calculating equipment real-
time data based on a local
power grid distribution network, and storing the equipment real-
time data in a
database; based on a KMP
algorithm, multi-level equipment cooperative calculation is realized, time of a corresponding fixed value is analyzed to facilitate cooperative relationship matching, and the time is displayed on the
system through a graphic
visualization technology; image-
text recognition is carried out, then load prediction is carried out, and a protection diagnosis strategy is provided; by means of the image-
text recognition technology, images are converted into characters, structural data are output and automatically recorded to the background, labor can be greatly saved, and efficiency is improved; the method comprises the following steps of:
processing, analyzing and understanding images by using a computer through an integrated image-
text recognition technology so as to recognize targets and objects in various different
modes, realize recognition of basic parameters and setting value sheets of equipment into characters, and store the characters in a
database; and the speed and effectiveness of the load prediction and diagnosis method are greatly improved.