Shaft Vibration Monitoring with Adaptive Reference Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing shaft vibration monitoring systems struggle to differentiate between normal and abnormal vibrations due to changes in operation conditions, leading to erroneous detections, and require a lengthy data collection period before monitoring can start, especially in turbomachines.
Innovation Solution
A monitoring device and method that acquires process data and shaft vibration measurements, creates a shaft vibration calculation model, and develops a determination model to evaluate vibrations based on operation conditions, allowing for immediate monitoring and updating of models during actual operation, thereby compensating for data shortages and adapting to changes in operation conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a fixed threshold value is used for abnormality determination, then the monitoring system is simple to operate, but it causes erroneous detection when operation conditions change
Solution Approach 1:
The patent implements dynamic threshold values that automatically adjust according to operation conditions such as rotational speed and load. Instead of using a fixed threshold, the system calculates adaptive thresholds based on real-time operational parameters, allowing the monitoring system to remain simple to operate while accurately distinguishing between normal vibrations caused by condition changes and true abnormalities.
Solution Approach 2:
The system changes the threshold parameter from a fixed value to a dynamic value that varies with operation conditions. By establishing a relationship between operation conditions and threshold values, the system adapts to different operational states, preventing erroneous detections while maintaining ease of operation through automated adjustment.
2Measurement precision
If learning data is collected by changing operation conditions in a normal range, then the determination model becomes accurate, but the data collection period becomes excessively long
Solution Approach 1:
The patent performs preliminary system identification to create a shaft vibration calculation model before actual operation begins. This pre-established model can predict shaft vibrations under various operation conditions, allowing the system to start monitoring immediately without requiring extensive data collection. The model is subsequently refined during operation, achieving high precision without the lengthy initial data collection period.
Solution Approach 2:
The system creates a virtual copy of the shaft vibration characteristics through the calculation model. Instead of collecting actual vibration data for all possible operation conditions, the model replicates expected vibration behavior, enabling accurate determination without physically experiencing all operational scenarios during the data collection phase.
3Loss of time
If monitoring starts immediately after introduction, then the loss of time is reduced, but the determination model lacks sufficient learning data
Solution Approach 1:
The patent performs system identification and creates the initial determination model before the machine enters actual operation. This preliminary action ensures that sufficient learning data is obtained during factory testing and commissioning, allowing monitoring to start immediately with a pre-trained model that will be continuously refined during operation.
Solution Approach 2:
The shaft vibration calculation model serves as an intermediary that bridges the gap between limited pre-operation data and the need for accurate determination. The model is trained on available data and then used to generate additional learning data through simulation, enabling immediate monitoring with adequate precision.
4Loss of time
If a shaft vibration calculation model is created and system identification is performed, then the determination model can be created before actual operation, but additional complexity is introduced
Solution Approach 1:
The shaft vibration calculation model serves multiple functions: it acts as a prediction tool for immediate monitoring, generates synthetic learning data for model refinement, and provides a framework for continuous improvement during operation. This multi-functionality justifies the additional complexity by eliminating the need for lengthy data collection periods.
Solution Approach 2:
The calculation model serves as an intermediary computational layer that processes operation conditions and predicts shaft vibrations. While this adds computational complexity, it enables immediate monitoring by bridging the gap between limited available data and comprehensive determination requirements, ultimately simplifying the overall deployment timeline.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
What is provided is a monitoring device for detecting abnormalities of shaft vibrations on the basis of a reference according to a change in an operation condition. A monitoring device includes a process data acquisition unit configured to acquire process data indicating an operation condition of a machine having a rotating shaft, a shaft vibration value acquisition unit configured to acquire a measurement value of a shaft vibration value of the rotating shaft under the operation condition indicated by the process data, a determination model configured to determine a normal value of the shaft vibration value according to the operation condition created on the basis of the shaft vibration value measured during an operation of the machine and the shaft vibration value calculated on the basis of a predetermined shaft vibration calculation model, and a monitoring unit configured to evaluate the measurement value of the shaft vibration value on the basis of the process data, the measurement value of the shaft vibration value, and the determination model.