Abstract: In order to improve the peeling rate of cantaloupe and reduce the amount of residue on the two tops of cantaloupe, an improved inner arc type positioning fork structure was designed. Concurrently, in view of the safety challenges facing the fruit holding and clamping mechanism, the clamping block structure was improved to a holding and grasping type to enhance safety in the manual operation process. In order to determine the optimal value of the drive motor speed frequency, cutter motor speed frequency, and cutting angle, SPSS software was used to perform multiple linear regression analysis on these three parameters, establish a multiple linear regression mathematical model, and obtain the experimental factors through regression tests. Regression equation with the test index was used to determine the best value of the test factor parameters whereby: drive fruit motor speed frequency was 32 Hz, cutter motor speed frequency was 25 Hz, and cut-in angle was 26°. The test results show that the peeling rate of the double-knife cantaloupe peeling machine was over 89.1%, and the average peeling rate was 91.1%, which meets the actual processing requirements of the enterprise.