Construction Machine Abnormality Detection via Sensor Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing abnormality detection methods for construction machines, such as hydraulic shovels, face challenges in accurately judging component failures due to temperature differences between engine cylinders and the need for complex control strategies, leading to false detections and inefficient maintenance.
Innovation Solution
An abnormality detecting device that uses correlation coefficient calculation and comparison to analyze sensor information from multiple sensors, determining the degree of difference in physical state information to judge component abnormalities, eliminating the need for learning values or judgment thresholds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If average temperature is used as reference for abnormality judgment, then the judgment process is simplified, but false detection occurs due to temperature differences between cylinders and complex engine control
Solution Approach 1:
The patent changes the judgment parameter from absolute temperature values to correlation coefficients that measure the relationship between temperature changes of different cylinders. This transformation allows the system to ignore actual temperature levels and focus on whether cylinders are behaving consistently relative to each other, thereby avoiding false detections caused by environmental temperature variations and complex control strategies.
Solution Approach 2:
The patent introduces correlation coefficients as an intermediary parameter between raw temperature sensor data and abnormality judgment. Instead of directly comparing temperature values or using average temperature as reference, the system uses correlation coefficients to mediate the judgment process, capturing the relationship between cylinder temperature variations while filtering out common environmental influences.
2Measurement precision
If multiple sensor information is analyzed to improve judgment accuracy, then detection precision improves, but calculation complexity and processing time increase
Solution Approach 1:
The patent extracts only the essential relationship information from multiple sensor readings by calculating correlation coefficients. Instead of analyzing all possible combinations of sensor data or using complex multivariate analysis, the system extracts the key feature - the correlation between temperature changes of different cylinders - which is sufficient for detecting abnormalities while keeping calculations simple.
3Measurement precision
If learning values and judgment thresholds are calculated for different operating conditions, then judgment accuracy improves, but the system becomes more complex and requires more data processing
Solution Approach 1:
The patent creates a universal abnormality detection method using correlation coefficients that works across different operating conditions without requiring condition-specific learning values or thresholds. The correlation-based approach is inherently adaptive to various operating states because it measures relative relationships rather than absolute values, making the system universally applicable while eliminating the need for extensive data processing to establish condition-specific criteria.
Data Source
AI summary
Provided is an abnormality detecting device for a construction machine that can estimate an abnormality occurring to a component (engine, pump, etc.) of the construction machine based on the relationship among a plurality of pieces of sensor information and thereby prevent machine failure. A correlation coefficient calculation unit 102 calculates correlation coefficients between time-series sensor values acquired by a plurality of sensors 101. A correlation coefficient comparison unit 103 compares the correlation coefficients and calculates the degree of difference between each correlation coefficient and other correlation coefficients. An abnormality judgment unit 104 judges that an abnormality has occurred to a part related to a sensor when the degree of difference calculated in regard to the sensor exceeds a preset value.