Agricultural implement harrow blade fault identification method based on encoder and decoder

By installing a high-definition camera and a neural network model based on an encoder decoder on agricultural machinery, real-time identification and diagnosis of rake sheet faults is achieved, and the problem of inefficient traditional manual inspections is solved, and the accuracy and efficiency of fault detection are improved.

CN119723182APending Publication Date: 2025-03-28JINLING INST OF TECH
0 Cites 0 Cited by

Patent Information

Application Number
CN202411784037.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-06
Publication Date
2025-03-28

AI Technical Summary

Technical Problem

The traditional method of rake sheet fault detection of agricultural machinery relies on manual inspection, which is inefficient and prone to missed and missed inspections, and cannot meet the needs of modern agricultural mass production.

Method used

A neural network model based on the encoder decoder is adopted to collect rake images in real time through a high-definition camera, perform image preprocessing and feature extraction, and combine deep learning technology to train the fault recognition model to achieve accurate identification of rake failures.

Benefits of technology

It greatly improves the efficiency and accuracy of rake failure detection, reduces the time and cost of manual inspection, can quickly respond to fault conditions in different environments, and provides diagnostic information and maintenance suggestions.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses an agricultural implement disk blade fault identification method based on an encoder and a decoder, and the method comprises the steps: firstly, installing a camera which is high in resolution and adapts to various illumination conditions nearby an agricultural implement disk blade, and collecting image data; secondly, denoising, enhancing, cutting, normalizing and other pre-processing are carried out on the collected image, and high-quality data are provided for the follow-up process; thirdly, an encoder and decoder model is constructed, a convolutional neural network is designed by using a deep learning framework, an encoder extracts features, and a decoder recovers image details and encodes fault features; then, the normal and fault states of the disk blade are accurately recognized by supervising a learning training model, the model is deployed to edge computing equipment, images are collected in real time, data are transmitted, and the edge equipment can preliminarily detect faults to reduce delay; and finally, the fault identification system detects faults after receiving real-time data, gives an alarm and displays positions and types when finding problems, and the system provides diagnosis information and maintenance suggestions according to the fault types and collects and feeds back continuous optimization models, so that the fault identification accuracy and the system reliability are improved.
Need to check novelty before this filing date? Find Prior Art