Real-Time System Analysis Using Selective Expense Calculation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing real-time analysis methods for technical systems are inefficient, often requiring unnecessary checks of time intervals, leading to prolonged run times and potential erroneous classification of systems as not real-time capable, especially under high workload conditions or with task systems containing both small and large periods.
Innovation Solution
A method that dynamically adjusts the degree of approximation by initially using approximate system expenses for a subset of tasks and gradually switching to precise expenses, allowing for iterative checks of time intervals, thereby reducing run time and improving classification reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If precise system expenses are calculated for all jobs, then classification accuracy is improved, but run time is prolonged
Solution Approach 1:
The patent applies partial action by calculating precise system expenses only for a subset of jobs (those that may cause deadline misses) rather than all jobs. The method identifies critical jobs using approximate expenses first, then performs precise calculation only where needed, achieving accurate classification without the full computational overhead of universal precise calculation.
Solution Approach 2:
The patent segments the job set into different categories: jobs processed with approximate expenses and jobs processed with precise expenses. This segmentation allows the system to apply different levels of computational effort to different jobs based on their criticality, thereby reducing overall run time while maintaining classification accuracy for critical cases.
2Reliability
If many time intervals are checked to ensure real-time capability, then classification reliability is improved, but run time is prolonged
Solution Approach 1:
The patent performs preliminary classification using approximate expenses to identify potentially problematic time intervals before conducting precise analysis. This preliminary action filters out obviously safe cases, allowing the system to focus computational resources only on intervals that require detailed examination, thus maintaining reliability while reducing run time.
Solution Approach 2:
The method applies partial checking by examining only those time intervals that are identified as potentially problematic through approximate analysis, rather than checking all possible intervals. This selective approach maintains classification reliability for critical intervals while avoiding unnecessary computational overhead.
3Productivity
If approximate values are used for system expenses, then run time is reduced, but classification accuracy deteriorates
Solution Approach 1:
The patent uses approximate values for the majority of jobs to maintain run time efficiency, while switching to precise calculation for critical jobs that may affect classification accuracy. This partial use of approximation achieves a practical balance between speed and accuracy.
Solution Approach 2:
The system uses feedback from approximate analysis to determine when precise calculation is necessary. By monitoring whether approximate expenses indicate potential deadline misses, the system dynamically decides which jobs require precise expense calculation, thereby maintaining accuracy where needed while preserving efficiency overall.
Data Source
AI summary
The invention relates to a method for the real-time analysis of a system, especially a technical system, which is to process tasks (τ). A job that is defined by processing of a task (τ) generates system expenses. In order to create a particularly quick and accurate method, an approximation of the method is cancelled when it is considered that an interval (I, Iτ, Iτnτ) cannot be processed in real time, the system expenses being taken into account instead of the approximate value for at least one job of a task (τ).