Multi-Sensor Machine Diagnostics for Accurate Fault State Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current diagnostic systems for machines are inadequate as they only measure exceeding limit values and do not allow for precise evaluation of fault states, leading to incorrect fault recognition and inability to identify new or unknown fault states.
Innovation Solution
A diagnostic system using at least two sensors (vibration, strain, and position sensors) with a control part that stores model fault-free and fault states, processes data to pair sensor measurements, and performs temperature compensation and filtering to accurately evaluate machine states, identifying both known and unknown faults.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple sensors are used to measure different phenomena, then the ability to detect fault states is improved, but the complexity of data processing and evaluation increases
Solution Approach 1:
The evaluation process is segmented into distinct steps: data acquisition from multiple sensors, pairing of measured data in time, comparison with model states, and fault identification. This segmentation makes the complex data processing manageable and systematic, allowing multiple sensors to be integrated without overwhelming complexity
Solution Approach 2:
Model states serve as an intermediary between raw sensor data and fault diagnosis. The measured data from multiple sensors are compared against pre-established model states (fault-free and various fault conditions), which mediates the complexity by providing a structured framework for evaluation rather than direct complex analysis of raw multi-sensor data
2Device complexity
If fault recognition is based on data from a single sensor type, then the system complexity is reduced, but the accuracy of fault identification deteriorates
Solution Approach 1:
Data from multiple sensor types (vibration, strain, position, distance) are merged and paired in time within the evaluation process. This combining of multi-sensor data provides comprehensive information for fault identification, achieving high measurement precision while managing complexity through systematic data integration
Solution Approach 2:
The evaluation process is designed to be universal, handling data from different sensor types through a common framework of pairing and comparison with model states. This multi-functional approach allows the same evaluation mechanism to process diverse sensor inputs, maintaining simplicity while achieving accurate fault identification
3Measurement precision
If the diagnostic system uses comprehensive model states for all possible faults, then the evaluation precision is improved, but the memory requirements and data storage needs increase
Solution Approach 1:
Model states are prepared in advance and stored in the control part's memory before actual diagnostic operations. This preliminary action includes creating fault-free models and various fault condition models based on historical data and expert knowledge, allowing rapid comparison during operation without real-time complex analysis
Solution Approach 2:
Instead of storing all possible raw measurement data for every fault condition, the system uses simplified model states that capture the essential characteristics of different fault conditions. These models are copies or representations of fault patterns that require less storage space while maintaining diagnostic precision
Data Source
AI summary
A method for performing technical diagnostics of machines is carried our by means of a diagnostic system of machines that employs at least two sensors to be placed on the machines, wherein the sensors are selected from the group of vibration sensors, strain sensors, position sensors, and distance sensors, and wherein measured data is evaluated by an evaluation process comprising a step of pairing the measured data and a step of comparing processed data with model states.