A Hybrid Probabilistic Battery Health Management Approach for Robust Inspection Drone Operations

TL;DR

The paper tackles reliable UAV inspection under harsh environments by predicting Li-Po EOD voltage with uncertainty quantification using a hybrid model that merges physics-based discharge with probabilistic error correction. It employs MC dropout and a negative log-likelihood loss to separately capture epistemic and aleatoric uncertainties, yielding total predictive uncertainty used in health-state diagnostics. Evaluation on 26,156 samples from 33 offshore-wind drone flights shows a 14.8% improvement in probabilistic accuracy over the best baseline and robust health-state inference via prediction-interval-based metrics. This approach enables more robust drone operations by providing both accurate voltage forecasts and uncertainty-aware battery health assessments, with guidance for future enhancements including PINNs and calibration techniques.

Abstract

Health monitoring of remote critical infrastructure is a complex and expensive activity due to the limited infrastructure accessibility. Inspection drones are ubiquitous assets that enhance the reliability of critical infrastructures through improved accessibility. However, due to the harsh operation environment, it is crucial to monitor their health to ensure successful inspection operations. The battery is a key component that determines the overall reliability of the inspection drones and, with an appropriate health management approach, contributes to reliable and robust inspections. In this context, this paper presents a novel hybrid probabilistic approach for battery end-of-discharge (EOD) voltage prediction of Li-Po batteries. The hybridization is achieved in an error-correction configuration, which combines physics-based discharge and probabilistic error-correction models to quantify the aleatoric and epistemic uncertainty. The performance of the hybrid probabilistic methodology was empirically evaluated on a dataset comprising EOD voltage under varying load conditions. The dataset was obtained from real inspection drones operated on different flights, focused on offshore wind turbine inspections. The proposed approach has been tested with different probabilistic methods and demonstrates 14.8% improved performance in probabilistic accuracy compared to the best probabilistic method. In addition, aleatoric and epistemic uncertainties provide robust estimations to enhance the diagnosis of battery health-states.

Figures (11)

• Figure 1: Uncertainty-aware battery health management approach.
• Figure 2: Structure of the proposed uncertainty-aware CNN architecture
• Figure 3: Use of dropout at test stage with MC sampling.
• Figure 4: Health-state inference example based on prediction interval quantification.
• Figure 5: Offshore inspection drone.