Real-time safety risk assessment of in-service marine engineering structures is one of the key components for the intelligent development of marine resources. Floating platforms typically have weak modes and time-varying characteristics, and hence, traditional structural damage identification and safety assessment methods cannot be directly applied to floating platforms. The objective of this study was to achieve the real-time safety risk assessment of in-service floating platforms. First, a structural damage identification method for floating platforms was proposed based on virtual loads, and the physical parameters of the platforms were obtained through inversion of measurable data. The identification result was verified based on the lengths and stiffnesses of polyester cables, which affected the dynamic response of the floating platform. The results showed that the proposed method effectively detected the parameter changes of the platform during long-term service. Then, the neural network was used to construct a load–stress surrogate model at key positions of the floating offshore platform. This model could calculate the stresses at key positions in real time, thereby evaluating the safety risks of the platform. Lastly, a real-time safety risk assessment method for floating offshore platforms was proposed, which provides a technical solution for ensuring the safety of offshore platforms.

中文翻译：

---

基于现场监测的海上浮动平台实时安全风险评估

在役海洋工程结构实时安全风险评估是海洋资源智慧开发的关键内容之一。浮动平台通常具有弱模态和时变特性，传统的结构损伤识别和安全评估方法不能直接应用于浮动平台。本研究的目的是实现在役浮动平台的实时安全风险评估。首先，提出一种基于虚拟载荷的浮动平台结构损伤识别方法，通过实测数据反演获得平台物理参数。根据影响浮动平台动态响应的聚酯缆的长度和刚度对识别结果进行了验证。结果表明，该方法有效检测了平台在长期使用过程中的参数变化。然后，利用神经网络在海上浮动平台的关键位置构建了载荷-应力代理模型。该模型可以实时计算关键位置的应力，从而评估平台的安全风险。最后提出了一种海上浮动平台实时安全风险评估方法，为确保海上平台安全提供了技术方案。