Operation Circuit for Ensemble Empirical Mode Decomposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing Ensemble Empirical Mode Decomposition (EEMD) methods require significant computational resources, making them inconvenient for users, especially when applied to various fields that require real-time processing.
Innovation Solution
An operation circuit integrated into an integrated circuit, comprising a normalization unit, extreme value processing unit, curve processing module, component unit, accumulation unit, and denormalization unit, which simplifies matrix operations using LU decomposition and noise mixing to reduce computational burden.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional EEMD methods are used, then decomposition accuracy is maintained, but computational complexity and resource consumption increase significantly
Solution Approach 1:
The patent segments the complex EEMD computation into distinct functional modules: normalization unit, extreme value processing unit, curve fitting unit, and denormalization unit. Each module handles a specific aspect of the decomposition process, allowing parallel processing and reducing overall computational complexity while maintaining decomposition accuracy.
Solution Approach 2:
The patent transforms the computational parameters by working with normalized data representations and using simplified matrix operations. By changing the parameter space through normalization and using efficient linear algebra techniques, the computational burden is reduced while preserving the mathematical integrity of the decomposition.
2Measurement precision
If traditional EEMD methods are used, then decomposition accuracy is maintained, but processing time increases
Solution Approach 1:
The patent applies normalization preprocessing to the input data before the main decomposition process. This preliminary action simplifies subsequent computations by working with standardized data, thereby reducing the time required for extreme value detection, curve fitting, and envelope calculation while maintaining accuracy.
Solution Approach 2:
The patent replaces traditional mechanical computation methods with optimized algorithms and circuit-based processing. By substituting software-based EEMD implementation with a dedicated hardware circuit that uses efficient mathematical operations, processing time is dramatically reduced while preserving decomposition precision.
3Adaptability or versatility
If traditional EEMD methods are used, then comprehensive analysis capability is maintained, but memory consumption increases
Solution Approach 1:
The patent extracts and processes only the essential components needed for EEMD decomposition: extreme values, envelope curves, and mean calculations. By taking out only the necessary data elements and discarding redundant intermediate results, memory consumption is reduced while maintaining comprehensive analysis capability.
Solution Approach 2:
The patent transforms the problem into a more efficient dimensional representation by using normalized coordinate systems and compact matrix formulations. This dimensional transformation allows the same analytical capability to be achieved with fewer stored parameters and less memory footprint.
Data Source
AI summary
An operation circuit and an operation method thereof are revealed. The operation circuit includes an extreme value processing unit, a curve processing module, and a component unit. The extreme value processing unit receives and processes a plurality of input data to get maximum values and minimum values. The curve processing module constructs a first matrix and a second matrix according to the maximum and minimum values and then decomposes the first matrix and the second matrix into first submatrices and second submatrices respectively. According to these submatrices, the curve processing module gets at least one mean value function corresponding to the maximum and the minimum values. The computation of a single matrix is reduced by matrix decomposition and operations of the operation circuit. Compared with conventional Gauss matrix manipulations that run by computer systems, the present invention can be applied to simpler circuits by simplifying matrix operation processes.