Prediction method and system for significant wave height of wave of full harbor basin, storage medium and application

Abstract

The invention belongs to the technical field of wave information prediction, and discloses a method and system for predicting the significant wave height of full-basin waves, a storage medium and application. The method for predicting the significant wave height of the wave in the whole harbor basin comprises the steps that firstly, data set construction and preprocessing are conducted, missing value processing and normalization are conducted on data of waves inside and outside the harbor basin and data of a wind field outside the harbor, multi-scale feature coding is conducted on time information corresponding to the data, a conditional generative adversarial network is built after the data set is divided, and the significant wave height of the wave in the harbor basin is predicted. And inputting the wind field and wave features after multi-scale time feature splicing fusion processing into a neural network, continuously iteratively training and updating and adjusting neural network parameters and hyper-parameters of the model, and finally establishing an optimal end-to-end deep learning prediction model. The prediction model is very simple and easy to use, and the effective wave height of the waves in the whole harbor basin can be rapidly predicted as long as preprocessed wind and wave data and multi-scale time coding information outside the harbor basin are provided and input into the established prediction model.

Description

technical field [0001] The invention belongs to the technical field of wave information forecasting, and in particular relates to a method for predicting the effective wave height of waves in the entire harbor basin, a forecasting system for the effective wave height of waves in the harbor basin, and a storage medium for receiving user input programs, and waves in harbor basins, water conservancy projects and near-shore projects. applications in forecasting. Background technique [0002] The importance of rapid and accurate prediction of waves in the harbor basin is: waves have always been an important research topic in harbor basins, hydraulic engineering and near-shore engineering. Waves will not only affect the safe docking, operation and navigation of ships, but may also damage coastal buildings. Therefore, reliable and efficient wave prediction methods are very important. First of all, the prediction of the wave condition in the harbor basin is very important for contr...

AI Technical Summary

Problems solved by technology

[0009] (2) Since the construction of the numerical simulation model involves a large number of external parameters, it takes a lot of time to adjust the parameters to better simulate the real wave motion behavior. The process is quite complicated and the calculation time is high, which cannot meet the real-time forecast need;

[0010] (3) Most of the existing neural network-based wave height prediction models only consider the factor of wave height information outside the harbor, while the wave height in the harbor basin is obviously affected by many related factors, such as tides and seasonal changes. The wave height presents obvious periodic changes, and the external wind field also has a direct impact on the wave height. Therefore, ignoring these important factors that affect the behavior characteristics of the wave itself will affect the accuracy of the model prediction.

[0011] (4) Constrained by the limited monitoring equipment in the harbor basin, most of the existing wave height prediction methods in the harbor basin only predict the wave height at a specific location, and have not realized the wave height prediction at each position of the entire harbor basin. The efficient and precise construction and operation of the port have brought adverse effects

[0013] (1) Construct a high-quality data set. The deep learning model requires a large amount of data for training. The quality of the data set affects the quality of the model, especially in the case of limited monitoring data. It is difficult to construct a high-quality data set;

[0014] (2) The factors affecting wave height are complex and changeable. Select important relevant factors and deal with them effectively to better characterize the behavioral characteristics of waves (such as periodicity, etc.), so that the model can learn these factors more accurately. species characterization is a challenge;

[0015] (3) At present, only one or a few buoys are arranged in the harbor basin to monitor the wave situation, so the monitoring data obtained are often very sparse, and what is actually needed is to predict the wave height information at each position of the harbor basin, which also has certain challenge

Images