There are a variety of requirements for future forecasts in relation to optimizing the production of wave energy. Daily forecasts are required to plan maintenance activities and allow power producers to accurately bid on wholesale energy markets, hourly forecasts are needed to warn of impending inclement conditions, possibly placing devices in survival mode, while wave-by-wave forecasts are required to optimize the real-time loading of the device so that maximum power is extracted from the waves over all sea conditions. In addition, related hindcasts over a long time scale may be performed to assess the power production capability of a specific wave site. This paper addresses the full spectrum of the aforementioned wave modeling activities, covering the variety of time scales and detailing modeling methods appropriate to the various time scales, and the causal inputs, where appropriate, which drive these models. Some models are based on a physical description of the system, including bathymetry, for example (e.g., in assessing power production capability), while others simply use measured data to form time series models (e.g., in wave-to-wave forecasting). The paper describes each of the wave forecasting problem domains, details appropriate model structures and how those models are parameterized, and also offers a number of case studies to illustrate each modeling methodology.

中文翻译：

---

波浪能生产的海洋预测

关于优化波浪能生产的未来预测有多种要求。需要每日预测来计划维护活动并允许电力生产商准确地对批发能源市场进行投标，需要每小时预测来警告即将发生的恶劣条件，可能会将设备置于生存模式，而需要逐波预测以优化设备的实时加载，以便在所有海况下从波浪中提取最大功率。此外，还可以进行长时间范围内的相关后报，以评估特定波场的发电能力。本文讨论了上述波浪建模活动的全部范围，涵盖了各种时间尺度和适合各种时间尺度的详细建模方法，以及在适当情况下驱动这些模型的因果输入。一些模型基于系统的物理描述，包括测深，例如（例如，在评估发电能力时），而其他模型仅使用测量数据来形成时间序列模型（例如，在波到波预测中）。本文描述了每个波浪预测问题域，详细说明了适当的模型结构以及这些模型的参数化方式，还提供了许多案例研究来说明每种建模方法。