An Optimal Algorithm for Renewable Energy Generation Based on Neural Network

Algorithm 1

IRANSAC algorithm.

(1) The feature point matching pairs whose number of intersection points with other matching pairs is greater than 3 are eliminated from the matching pair set.

(2) Sort the remaining feature point matching pairs in an ascending order according to the ratio to form a feature point matching pair set . Among them, is the minimum value of the Euclidean distance in the feature point matching pair; is the Euclidean distance of the remaining point pairs.

(3) Divide into four parts, , and in sequence; extract five sample data from in turn; and solve the transformation matrix .

(4) If the error between the remaining matching point pairs in and is less than the agreed threshold (take 0.6), it is a consistent set with the sample drawn.

(5) Use the obtained transformation matrix to traverse the corresponding matching points in , calculate the proportion of matching point pairs in the set that satisfy the transformation matrix under the threshold, repeat the above steps, select the transformation matrix with the largest , and record it as as the final transformation matrix.