Availability Prediction Model Segmentation for System Maintenance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for predicting the availability of information processing systems require significant effort to maintain and update models when system configurations change, as they need to account for complex interactions between various constituent elements.
Innovation Solution
An information processing device and method that includes a configuration storage unit for identifying constituent elements, a rule storage unit for influencing elements during failures, and an availability prediction model generation unit, which generates a prediction model based on this information to minimize maintenance efforts by separating system configuration updates from influence rule updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a model for predicting availability is constructed to account for complex interactions between constituent elements, then the prediction accuracy is improved, but the model complexity and maintenance effort increase
Solution Approach 1:
The patent segments the availability prediction model into two independent parts: (1) system configuration information describing constituent elements and their relationships, and (2) influence rules describing failure propagation patterns. This segmentation allows each part to be maintained independently, reducing overall model complexity while preserving prediction accuracy through the separate management of structural and behavioral aspects of the system.
Solution Approach 2:
The patent extracts the influence rules from the system configuration, separating the static structural description from the dynamic failure propagation logic. By taking out the influence rules as a distinct component, the system can update configuration information without altering the influence rules, thereby reducing maintenance effort while maintaining accurate predictions of availability.
2Measurement precision
If the model accounts for complex interactions between constituent elements, then the prediction accuracy is improved, but the maintenance effort increases
Solution Approach 1:
By segmenting the model into configuration information and influence rules, the patent enables independent updates of each component. When system configuration changes occur, only the configuration information needs updating, while the influence rules remain unchanged. This segmentation significantly reduces maintenance effort while preserving the accurate prediction capability derived from the comprehensive interaction modeling.
Solution Approach 2:
The patent performs preliminary action by pre-defining the influence rules that capture failure propagation patterns among constituent elements. These pre-established rules can be applied repeatedly without modification when configuration changes occur, reducing the time and effort required for maintenance while maintaining accurate prediction of system availability under various failure scenarios.
3Reliability
If the model is updated when system configuration changes, then the prediction accuracy is maintained, but the maintenance effort increases
Solution Approach 1:
The patent segments the model into configuration information and influence rules, allowing updates to be applied selectively. When system configuration changes, only the configuration information portion needs updating, while the influence rules remain intact. This segmentation maintains prediction reliability by ensuring the model reflects current system state, while simultaneously improving maintenance ease by limiting the scope of required updates.
Solution Approach 2:
The influence rules are established in advance through preliminary action, capturing the essential failure propagation relationships. These pre-defined rules can be reused across different system configurations, meaning that when configuration changes occur, the system can maintain reliable predictions by simply updating the configuration information portion, without needing to redesign the entire model.
Data Source
AI summary
A configuration storage unit (110) stores the constituent element identification information of each constituent element, in association with the type information indicating the type of the constituent element, the constituent element identification information of another constituent element related to the constituent element, and the type information of the another constituent element. For each piece of the type information, a rule storage unit (120) stores the type information of another constituent element, which is influenced when failures occur in the constituent element corresponding to the type information, in association with influence information indicating a content of the influence. An availability model generation unit (130) generates an availability prediction model for an information processing system, on the basis of the information stored in the configuration storage unit (110) and the information stored in the rule storage unit (120).