A Glider Simulation Model Based on Optimized Support Vector Regression for Efficient Coordinated Observation

Fangjie Yu,Ge Chen,Zhiyuan Zhuang,Jie Yang

doi:10.3389/fmars.2021.671791

Abstract

Multi-gliders have been widely deployed as an array in nowadays ocean observation for fine and long-term ocean research, especially in deep-sea exploration. However, the strong, variable ocean currents and the delayed information feedback of gliders are remaining huge challenges for the deployment of glider arrays which may cause that the observed data cannot meet the study needs and bring a prohibitive cost. In this paper, we develop a Glider Simulation Model (GSM) based on the support vector regression with the particle swarm optimization (PSO)-SVR algorithm to integrate the information feedback from gliders and ocean current data for rapid modeling to effectively predict the gliders’ trajectories. Based on the real-time predictive information of the trajectories, each glider can select future movement strategies. We utilize the in-suit datasets obtained by sea-wing gliders in ocean observation to train and test the simulation model. The results show that GSM has an effective and stable performance. The information obtained from the modeling approaches can be utilized for the optimization of the deployment of the glider arrays.

Highlights

Underwater gliders are characterized as a type of persistent, long operation range, small power consumption, and unmanned marine vehicle which are propelled by buoyancy (Rudnick et al, 2004)
We introduce three observation datasets obtained by sea-wing gliders into the simulation experiments; the results demonstrate that the Glider Simulation Model (GSM) has great stability and effectiveness on sea-wing gliders
A sea-wing glider (g-002) with multiple sensors is deployed for the inner observation of eddies, and the trajectory is shown in Figure 1B; the black line represents the trajectory of g-002

Summary

INTRODUCTION

Underwater gliders are characterized as a type of persistent (can operate continuously for weeks to months even to years), long operation range (up to thousand kilometers), small power consumption, and unmanned marine vehicle which are propelled by buoyancy (Rudnick et al, 2004). Based on the predictive information, each glider can determine future motion strategies In this model, we introduce support vector regression based on PSO-SVR as the supporting algorithm. Three sets of in-suit observation data collected by sea-wing gliders are employed to train and test the simulation model A sea-wing glider (g-002) with multiple sensors is deployed for the inner observation of eddies, and the trajectory is shown in Figure 1B; the black line represents the trajectory of g-002. If the experiment meets the iteration steps or the value of fitness function reaches the extent predefined, we stop updating and choose gbest as the best hyperparameters (C,σ); otherwise, we return to Third

EXPERIMENTS

The Results of Parameter Optimization

DATA AVAILABILITY STATEMENT