Open Access

ARTICLE

The Design and Implementation of a Biomechanics-Driven Structural Safety Monitoring System for Offshore Wind Power Step-Up Stations

Ruigang Zhang^1,*, Qihui Yan², Jialiang Wang¹, Hao Wang¹, Jie Sun², Junjiao Shi²

1 Xi’an Thermal Power Research Institute Co., Ltd., Xi’an, 710054, China
2 Huaneng Jiangsu Clean Energy Branch, Nanjing, 210015, China

* Corresponding Author: Ruigang Zhang. Email:

(This article belongs to the Special Issue: Advances in Grid Integration and Electrical Engineering of Wind Energy Systems: Innovations, Challenges, and Applications)

Energy Engineering 2025, 122(9), 3609-3624. https://doi.org/10.32604/ee.2025.066880

Received 19 April 2025; Accepted 04 July 2025; Issue published 26 August 2025

Abstract

As the core facility of offshore wind power systems, the structural safety of offshore booster stations directly impacts the stable operation of entire wind farms. With the global energy transition toward green and low-carbon goals, offshore wind power has emerged as a key renewable energy source, yet its booster stations face harsh marine environments, including persistent wave impacts, salt spray corrosion, and equipment-induced vibrations. Traditional monitoring methods relying on manual inspections and single-dimensional sensors suffer from critical limitations: low efficiency, poor real-time performance, and inability to capture millinewton-level stress fluctuations that signal early structural fatigue. To address these challenges, this study proposes a biomechanics-driven structural safety monitoring system integrated with deep learning. Inspired by biological stress-sensing mechanisms, the system deploys a distributed multi-dimensional force sensor network to capture real-time stress distributions in key structural components. A hybrid convolutional neural network-radial basis function (CNN-RBF) model is developed: the CNN branch extracts spatiotemporal features from multi-source sensing data, while the RBF branch reconstructs the nonlinear stress field for accurate anomaly diagnosis. The three-tier architectural design—data layer (distributed sensor array), function layer (CNN-RBF modeling), and application layer (edge computing terminal)—enables a closed-loop process from high-resolution data collection to real-time early warning, with data processing delay controlled within 200 ms. Experimental validation against traditional SOM-based systems demonstrates significant performance improvements: monitoring accuracy increased by 19.8%, efficiency by 23.4%, recall rate by 20.5%, and F1 score by 21.6%. Under extreme weather (e.g., typhoons and winter storms), the system’s stability is 40% higher, with user satisfaction improving by 17.2%. The biomechanics-inspired sensor design enhances survival rates in salt fog (85.7% improvement) and dynamic loads, highlighting its robust engineering applicability for intelligent offshore wind farm maintenance.

---

Keywords

---