A novel algorithm for high compression rates focalized on electrical power quality signals

Area of Science:

• Electrical Engineering
• Signal Processing
• Data Compression

Background:

• Massive data acquisition in telecommunications necessitates efficient data compression techniques.
• Existing methods may alter signal amplitude and introduce shifts post-compression.
• Reducing bandwidth usage and storage costs is crucial for network infrastructure.

Purpose of the Study:

• To develop a high-performance algorithm for compressing electrical power quality signals.
• To address amplitude differences and signal shifting issues after wavelet-based compression.
• To introduce new performance metrics: compression time and recovery time.

Main Methods:

• Application of a biorthogonal wavelet level six transform.
• Signal normalization for amplitude correction.
• Moving average filtering to eliminate signal ripple.
• Shift correction by aligning signal peaks.

Main Results:

• Achieved a compression rate of 99.803%.
• Reconstruction fidelity metrics: RTE 99.9479%, NMSE 0.000434, Cross-Correlation 0.999925.
• Introduced and evaluated compression and recovery times for real-time applicability.

Conclusions:

• The proposed algorithm effectively compresses electrical power quality signals with high fidelity.
• The integrated normalization and filtering steps successfully restore signal accuracy.
• The new time-based metrics offer valuable insights into the algorithm's practical performance speed.