Method for the computer-aided forecasting of future operating states of machine components
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for predicting maintenance requirements and impending damage events in machine components rely heavily on empirical data and operator interpretation, leading to inconsistent and potentially unreliable predictions.
Innovation Solution
A method utilizing automated data processing to identify influencing factors and interactions specific to each application, feeding this information into a database to generate precise recommendations for maintenance actions based on measured parameters, with a self-learning system for continuous improvement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If automated data processing is implemented to eliminate operator interpretation, then prediction reliability is improved, but system complexity increases
Solution Approach 1:
The patent replaces the mechanical system of manual operator interpretation with an automated data processing system that uses algorithms and computational models to analyze measured values and generate maintenance recommendations, thereby eliminating human subjectivity while managing system complexity through automation
Solution Approach 2:
The system performs self-service by automatically processing measured values, comparing them against stored reference data and models, and generating maintenance recommendations without requiring operator intervention or interpretation, thus improving reliability through consistent automated decision-making
2Measurement precision
If application-specific influencing factors and interactions are systematically identified and stored, then measurement interpretation accuracy is improved, but data processing complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-identifying and storing application-specific influencing factors, their interactions, and reference measured values in a database before actual measurement interpretation occurs. This preparation enables accurate real-time analysis without complex processing during operation
Solution Approach 2:
The system introduces an intermediary layer of pre-processed knowledge (stored influencing factors and interactions) that mediates between raw measured values and interpretation results, simplifying the actual measurement processing while improving accuracy through contextual information
3Measurement precision
If a self-learning system is implemented for continuous improvement, then long-term prediction accuracy is improved, but initial system complexity increases
Solution Approach 1:
The patent implements feedback mechanisms where measured values and their interpretations are continuously fed back into the system to refine influencing factor models and update reference data, enabling continuous improvement of prediction accuracy while managing initial complexity through structured feedback loops
Data Source
Figure 1
Figure 2
AI summary
The invention relates to a method for the computer-aided forecasting of future operating states of machine components that makes it possible to predict reproducibly and with a high degree of certainty a forthcoming event in the life cycle of a machine component. For this purpose, a parameter that is representative of the state of a machine component under consideration is selected for the component and is acquired as a measured value in the course of the use of the machine component. While taking into consideration the application concerned for which the machine component is being used, the acquired measured value is fed into an automatically running process to find a recommendation for action. Properties and the influencing variables determining these properties, the applications coming into consideration for them, the influencing variables to which the machine component is exposed in the respective application and the resultant influencing variables have in this case been formulated and determined for the machine component under consideration, and these influencing variables should be considered in the interpretation of the acquired measured value and the recommendation for action deduced from it. Findings obtained during the use of the machine component with respect to the influencing variables or findings pertaining to them that have been obtained by systematic experimental investigations are used to ascertain the interactions and effects of the influencing variables and bring them into a relationship with the damage events determined. The items of information ascertained in relation to the interactions and the damage events influenced by them are fed into a database. The application, the acquired measured value and the assigned influencing variables are fed as input variables into a selection algorithm running on a computer which, taking into consideration the input variables and the acquired measured value, selects the recommendation for action that is optimally appropriate for the use in the respective application on the basis of the information stored in the database.