Submarine Cable Vibration Signal Recognition Based on Sparrow Search Algorithm Optimized Support Vector Machine

Abstract

Online status monitoring and fault recognition of photoelectric composite submarine cables (hereinafter referred to as submarine cables) offers an effective approach for early warning of faults in submarine cables. In order to improve the speed and accuracy of this practice, this paper proposes a submarine cable vibration signal recognition method based on the sparrow search algorithm (SSA) optimized support vector machine (SSA-SVM). Firstly, the submarine cable fault signal was decomposed by the ensemble empirical module decomposition (EEMD) method, allowing for extraction of the kurtosis and energy entropy combination of each component as the training feature set, so as to avoid any potential signal distortion caused by direct noise reduction that may affect the extraction of target features. Then, the SSA was used to optimize the penalty factor and kernel function parameters of the SVM, thereby improving the recognition accuracy. Finally, through the submarine cable vibration signal simulation system based on Brillouin optical time domain analysis (BOTDA), 500 groups of submarine cable vibration signals under anchor smashing, erosion and friction conditions were obtained, and 3 types of noisy signals were decomposed by EEMD, to extract the feature data set of each component, in which 80% is taken as the training set, and the remaining 20% as the test set. The EEMD-SSA-SVM algorithm proposed in this paper was compared with the EEMD-PSO-SVM and SVM algorithms. The results show that the EEMD-SSA-SVM algorithm exhibits higher accuracy and superior optimization ability. Specifically, the accuracy of the test set reaches 95%, surpassing that achieved by other algorithm models.

Keywords

• pattern recognition
• submarine cable
• sparrow search algorithm(ssa)
• support vector machine(svm)
• signal decomposition