CN-121659176-B - Grain dryer fault detection method and system

Abstract

The invention relates to the technical field of intelligent agricultural equipment monitoring and discloses a fault detection method and system for a grain dryer. The method comprises the steps of collecting multidimensional time sequence data through a sensor network chip and dividing the multidimensional time sequence data to obtain a segmented data sequence, carrying out frequency domain filtering and distribution analysis on the segmented sequence, separating and removing environmental noise to obtain a purified data sequence, carrying out principal component analysis on the purified data sequence to obtain a main feature set, separating short-term fluctuation signals from the main feature, detecting abnormal deviation and generating an abnormal fluctuation index, accumulating and calculating an abnormal deviation value when the index exceeds an early warning threshold value, calculating average accumulated deviation in a unit window as abnormal trend intensity, comparing the trend intensity with a historical reference range, and generating and outputting a fault prediction report when the trend intensity exceeds the historical reference range. According to the method, through deep purification, feature extraction and trend quantitative analysis of multidimensional operation data, accurate early warning of early weak faults of the grain dryer is achieved.

Inventors

• LI JIGANG

Assignees

• 山东华利艾维机械有限公司

Dates

Publication Date

20260512

Application Date

20260205

Claims (5)

1. 1. A method for detecting faults of a grain dryer, comprising the steps of: The method comprises the steps of collecting multidimensional time sequence data of grain drying equipment during operation through a sensing network chip arranged on the grain drying equipment, and dividing the multidimensional time sequence data to obtain a segmented data sequence; separating and removing environmental noise and interference signals from the segmented data sequence to obtain a purified data sequence; performing principal component analysis on the purification data sequence to obtain a main feature set; separating a short-term fluctuation signal from the main characteristic set, detecting abnormal deviation in the short-term fluctuation signal and generating an abnormal fluctuation index; when the abnormal fluctuation index exceeds a preset early warning threshold value, carrying out accumulated calculation on the abnormal deviation in a continuous time window, and calculating the average accumulated deviation in a unit window as abnormal trend intensity based on the accumulated calculation result; comparing the abnormal trend intensity with a predetermined reference range, and generating and outputting a fault prediction report when the abnormal trend intensity exceeds the reference range; Wherein the performing cumulative calculation of the abnormal deviation in the continuous time window, calculating an average cumulative deviation in a unit window as an abnormal trend intensity based on a result of the cumulative calculation, includes: determining a starting time window in which the abnormal fluctuation index exceeds a preset early warning threshold for the first time; starting from the initial time window, carrying out weighted summation on abnormal deviation values in N continuous time windows to obtain a cumulative deviation value, wherein N is an integer greater than 1; calculating the ratio of the accumulated deviation value to the number N of time windows as the abnormal trend intensity; Wherein detecting an abnormal deviation in the short-term fluctuation signal and generating an abnormal fluctuation index includes: calculating a statistical characteristic value of the short-term fluctuation signal in a continuous time window; comparing the statistical characteristic value of each time window with the normal fluctuation range of the corresponding window determined based on the historical data; when the statistical characteristic value exceeds the corresponding normal fluctuation range, marking the corresponding window as an abnormal window; generating the abnormal fluctuation index based on the statistical characteristic values of all abnormal windows; The abnormal fluctuation index is the proportion of the number of abnormal windows to the total number of windows in the last fixed observation time.
2. 2. The method for detecting the failure of a grain dryer according to claim 1, wherein the steps of collecting multi-dimensional time series data of grain drying equipment during operation through a sensor network chip disposed on the grain drying equipment, and dividing the multi-dimensional time series data to obtain a segmented data sequence include: acquiring temperature data, humidity data and vibration data in the running process of the equipment through a sensing network chip arranged on grain drying equipment to obtain multidimensional time sequence data; and dividing the multi-dimensional time sequence data by adopting a sliding window with fixed time length to obtain a segmented data sequence.
3. 3. The method of claim 1, wherein said separating and removing ambient noise and interference signals from said segmented data sequence to obtain a clean data sequence comprises: converting the segmented data sequence from a time domain to a frequency domain by adopting Fourier transformation, filtering specific frequency components corresponding to a pre-acquired environmental noise mode, and carrying out inverse transformation on the frequency domain data after noise filtering to obtain a preliminary filtering signal; And carrying out data distribution analysis on the preliminary filtering signals, and identifying and removing residual interference signals with fluctuation ranges exceeding a first preset threshold value in the preliminary filtering signals to obtain the purified data sequence.
4. 4. The grain dryer malfunction detection method according to claim 1, wherein the generating and outputting a malfunction prediction report includes: integrating abnormal mode information related to the abnormal trend intensity, wherein the abnormal mode information comprises an abnormal type, occurrence time and equipment position; And generating a fault prediction report based on the integrated abnormal mode information, and sending the fault prediction report to the user terminal.
5. 5. A grain dryer malfunction detection system, for implementing a grain dryer malfunction detection method according to any one of claims 1 to 4, comprising: the data acquisition module is used for acquiring multidimensional time sequence data of the grain drying equipment during operation through a sensor network chip arranged on the grain drying equipment, and dividing the multidimensional time sequence data to obtain a segmented data sequence; The data purifying module is used for separating and removing environmental noise and interference signals from the segmented data sequence to obtain a purified data sequence; The feature extraction module is used for carrying out principal component analysis on the purified data sequence to obtain a main feature set; The fluctuation detection module is used for separating short-term fluctuation signals from the main characteristic set, detecting abnormal deviation in the short-term fluctuation signals and generating abnormal fluctuation indexes; the trend quantization module is used for carrying out cumulative calculation on the abnormal deviation in the continuous time window when the abnormal fluctuation index exceeds a preset early warning threshold value, and calculating the average cumulative deviation in the unit window as abnormal trend intensity based on the result of the cumulative calculation; and the fault early warning module is used for comparing the abnormal trend intensity with a predetermined reference range, and generating and outputting a fault prediction report when the abnormal trend intensity exceeds the reference range.

Description

Grain dryer fault detection method and system Technical Field The invention relates to the technical field of intelligent monitoring of agricultural mechanical equipment, in particular to a fault detection method and system for a grain dryer. Background At present, grain drying is a vital post-production treatment link in agricultural production, and the stable operation of equipment directly relates to the safety and economic value of grain storage, and has irreplaceable effects on guaranteeing grain quality and reducing mildew loss. Along with the development of agricultural modernization and intellectualization, higher requirements are put forward on real-time accurate monitoring of the running state of the dryer. In one prior art, fault detection of grain dryers has relied on single parameter alarm mechanisms based on fixed thresholds and periodic manual field inspection. According to the method, a single-function temperature sensor or a single-function humidity sensor is deployed at a key part of equipment, operation data are collected and compared with a preset static safety threshold, and an alarm is triggered when the operation data exceed the preset static safety threshold. Meanwhile, maintenance personnel are relied on to check the equipment on site according to fixed periods to confirm the state. This monitoring method, which relies on static rules and human experience, has inherent limitations. Because the method lacks intelligent extraction and analysis capability of multi-dimensional time sequence data characteristics, environmental noise cannot be effectively filtered to capture weak early signals representing progressive degradation of equipment performance, deep association and trend analysis of multiple indexes such as temperature, humidity and vibration are difficult, fusion analysis requirements of multi-source heterogeneous time sequence data generated in a complex monitoring network formed by sensor network chips are difficult to meet, and deep sensing and early warning capability of equipment operation states is lacking. As a result, the system has insufficient sensitivity to identify complex failure modes and a high false alarm rate, and is often discovered after the failure has become apparent or caused a shutdown. Therefore, the core technical problem faced in the prior art is how to carry out deep cleaning, feature extraction and trend analysis on multidimensional operation data through an intelligent algorithm based on rich data sources provided by modern sensor network chips, so that early and accurate early warning of potential faults of the grain dryer is realized, and the defects of lag reaction and insensitivity to weak anomalies of the traditional detection method are overcome. Disclosure of Invention The invention provides a fault detection method and a fault detection system for a grain dryer, which are used for solving the technical problems that early and weak potential faults of the grain dryer are early-delayed and abnormal trends are difficult to accurately identify in multidimensional complex data due to the fact that fixed thresholds and manual experience are relied on in the prior art. In order to solve the technical problems, the present invention provides a method for detecting faults of a grain dryer, comprising: The method comprises the steps of collecting multidimensional time sequence data of grain drying equipment during operation through a sensing network chip arranged on the grain drying equipment, and dividing the multidimensional time sequence data to obtain a segmented data sequence; separating and removing environmental noise and interference signals from the segmented data sequence to obtain a purified data sequence; performing principal component analysis on the purification data sequence to obtain a main feature set; separating a short-term fluctuation signal from the main characteristic set, detecting abnormal deviation in the short-term fluctuation signal and generating an abnormal fluctuation index; when the abnormal fluctuation index exceeds a preset early warning threshold value, carrying out accumulated calculation on the abnormal deviation in a continuous time window, and calculating the average accumulated deviation in a unit window as abnormal trend intensity based on the accumulated calculation result; And comparing the abnormal trend intensity with a predetermined reference range, and generating and outputting a fault prediction report when the abnormal trend intensity exceeds the reference range. In an optional embodiment, the collecting the multidimensional time sequence data of the grain drying equipment during operation through the sensor network chip disposed on the grain drying equipment, and dividing the multidimensional time sequence data to obtain a segmented data sequence includes: acquiring temperature data, humidity data and vibration data in the running process of the equipment through a sensing network chip arranged on grai