Fault diagnosis of agricultural sprinkler irrigation machinery equipment based on machine vision

Abstract The traditional fault diagnosis of agricultural sprinkler irrigation machinery and equipment has the disadvantages of low accuracy and time-consuming. To solve these problems, the study designs a machine vision (MV)-based fault diagnosis model for agricultural sprinkler irrigation machinery and equipment. The study first investigates MV in sprinkler irrigation equipment fault diagnosis, then constructs a fault diagnosis model using improved convolutional neural network, and finally compares it with other methods to verify the performance of the model. The results showed that in the Iris dataset, the accuracy of the proposed algorithm was 95.13%, the training accuracy was 0.95, and the recall rate was 89.7%, which were better than the comparison methods. In the mechanical fault diagnosis dataset, the highest accuracy of the proposed model could reach 98.45%. This indicates that the MV model constructed in the study has higher accuracy and efficiency in the fault diagnosis of agricultural sprinkler irrigation mechanical equipment, which provides convenience for maintenance and repair.