Fluid Device Fault Diagnosis Using Suction-Discharge Difference Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fault diagnosis methods for fluid machinery are limited by one-sided reference data selection, failing to comprehensively consider failure characteristics, leading to low accuracy in detection results, particularly for pump valves where overall pressure or specific port pressure is the sole consideration.
Innovation Solution
A fault diagnosis method that includes obtaining a dataset with first characteristic data from the suction end, second characteristic data from the discharge end, and input-output difference data, using a trained fault diagnosis model with an encoder and decoder comprising fully connected layers to determine if the fluid device is in failure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If only overall pressure or specific port pressure data is used for fault diagnosis, then the device complexity is reduced, but the measurement precision and reliability of fault detection deteriorate
Solution Approach 1:
The patent segments the fault diagnosis data into three distinct components: suction end characteristic data, discharge end characteristic data, and input-output difference data. This segmentation allows comprehensive fault detection while maintaining manageable data processing through structured organization of multi-source data.
Solution Approach 2:
The patent merges multiple data sources (suction end pressure data, discharge end pressure data, and their difference data) into a unified fault diagnosis model. This combination of diverse data types enhances the comprehensiveness of fault detection and improves diagnostic accuracy by considering multiple operational aspects simultaneously.
2Reliability
If only single-point pressure data is collected, then the measurement system is simpler, but the reliability of fault diagnosis is insufficient
Solution Approach 1:
The patent applies local quality by collecting data from specific locations (suction end and discharge end) with distinct characteristic meanings. Each location provides localized operational information that reflects different aspects of device health, enabling more reliable fault diagnosis through spatially-distributed measurement quality.
Solution Approach 2:
The patent adds a new dimension to fault diagnosis by introducing input-output difference data alongside absolute pressure data. This dimensional expansion transforms single-point measurements into multi-dimensional diagnostic information, significantly improving reliability without requiring excessive measurement points.
Data Source
AI summary
The embodiments of the present disclosure provide a fault diagnosis method, a method for building a fault diagnosis model, fault diagnosis equipment, electronic device, and non-transitory computer-readable storage medium. The fault diagnosis method, for diagnosing a fluid device, which includes a suction end and a discharge end, includes: obtaining a data set for diagnosing the fluid device, wherein the data set includes first characteristic data about the suction end, second characteristic data about the discharge end, and input-output difference data, and the input-output difference data represents data difference between the suction end and the discharge end; obtaining a fault diagnosis model; and determining whether the fluid device is in failure based on the fault diagnosis model and the data set.