Comparison of the Impact of Symbol Aggregation Approximation Encoding and Multiscale Encoding on Lempel-Ziv Complexity Results

Lempel-Ziv complexity was widely used in multiple fields due to its ability to effectively reflect patterns change in signals. However, the traditional 0–1 encoding method limited Lempel-Ziv complexity representation of detailed pattern change in the signal, reducing the reflection accuracy of Lempel-Ziv complexity of signal complexity. By changing the encoding method of Lempel-Ziv complexity, the accuracy of signal complexity representation can be effectively improved. Symbol aggregation approximation encoding and multiscale encoding were used to improve the encoding of Lempel-Ziv complexity. And the two encoding methods had similar steps. This paper focused on the impact of two encoding methods on Lempel-Ziv complexity calculation and specifically discussed the impact of parameter setting in both encoding methods on the results. The main contribution of this paper was to discuss the impact of different parameters of the symbol aggregation approximation encoding and multiscale encoding on the Lempel-Ziv complexity calculation, providing theoretical support for improving the characterization of bearing fault severity by Lempel-Ziv complexity.