Edge Data Analysis Apparatus Using Correlation Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In information processing systems, efficiently specifying and acquiring analysis target data from moving edges while updating configurations frequently, without increasing the acquisition frequency of feature amounts, is challenging due to the need to specify data within allowed times.
Innovation Solution
The system determines analysis target data satisfying specific conditions by calculating correlation coefficients of feature amounts and specifying ranges of detection positions and times, allowing for the acquisition of relevant data from edges without frequent updates, thereby optimizing data acquisition and reducing communication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the system frequently updates edge configurations to track moving edges, then the accuracy of specifying analysis target data is improved, but the communication overhead and processing time increase
Solution Approach 1:
The system performs preliminary actions by calculating correlation coefficients between feature amounts and specifying ranges of detection positions and times in advance, before the actual data acquisition. This allows the management apparatus to efficiently determine which edges have relevant analysis target data without needing to frequently update and process all edge configurations in real-time, thus improving accuracy while reducing processing time
2Reliability
If the system acquires feature amounts from all edges to ensure complete data coverage, then the reliability of data selection is improved, but the communication overhead between edges and management apparatus increases
Solution Approach 1:
The system extracts only the necessary feature amounts from edges based on calculated correlation coefficients and specified ranges, rather than acquiring all feature amounts from all edges. The management apparatus calculates correlation coefficients between feature amounts and uses this information to identify and acquire only those feature amounts that are relevant to the analysis conditions, thereby maintaining reliable data selection while reducing communication overhead
Solution Approach 2:
The system changes parameters by using correlation coefficients as a filtering mechanism. Instead of uniformly acquiring data from all edges, the system calculates correlation coefficients between feature amounts and uses these coefficients to determine which edges' data should be acquired, thus reducing communication overhead while maintaining reliability through selective parameter-based filtering
3Manufacturing precision
If the system specifies analysis target data based on exact detection positions and times, then the precision of data acquisition is improved, but the difficulty of detecting and measuring increases due to moving edges
Solution Approach 1:
The system transitions from tracking edges in a single spatial dimension to a multi-dimensional approach by calculating correlation coefficients between feature amounts and specifying ranges that include both detection positions and detection times. This dimensional expansion allows the system to account for edge movement across space and time, maintaining precision in data acquisition while reducing the difficulty of tracking moving edges through correlation-based identification
Data Source
AI summary
An analysis method includes acquiring target data collected at edges; determining first analysis target satisfying a first condition, and specifying a first detection position indicating a position at which the first analysis target is detected at the edges and a first detection time; calculating a correlation coefficient of the feature amount; specifying a first range of the first detection position and a second range of the first detection time of the analysis target for which the correlation coefficient satisfies a predetermined relationship; determining second analysis target satisfying a second condition, and specifying a second detection position and a second detection time; determining whether the analysis target in which the second detection position is included in the first range and the second detection time; and acquiring any one of the analysis target from the edges when it is determined that the analysis target is included.