Forecast Of Rescue Force Demand For Cruise Ship Accident Based On PCA-SVR

In recent years,with the rapid development of the global cruise industry,the issue of cruise safety has attracted the attention of scholars in the society and related fields.As the main function of cruise ship is to carry passengers,once an accident occurs,the number of people involved in the risk is often large,and the loss caused is often catastrophic.Therefore,how to accurately predict the demand of ship rescue force in the stage of rescue and evacuation is one of the important issues to be considered in the field of maritime rescue.The demand of ship rescue force is the premise and basis for formulating the rescue action plan,and it is also an important basis for the coordination and command of ship rescue force and task allocation on the scene of accident rescue.Timely prediction of ship rescue force demand can help the rescue operation make timely response in the initial stage,and facilitate the dispatch of rescue forces.Based on the analysis and summary of previous shipwrecks,this paper makes an in-depth study on the prediction of ship rescue force demand for cruise ship accidents:(1)Based on the research and summary of relevant literature and Shipwreck accidents,this paper determines that the main influencing factors of ship rescue force demand in cruise ship accident rescue action are ship damage degree,number of people in distress,ship search and rescue speed,and establishes the corresponding influencing factor index system by analyzing the index influence coefficient through grey correlation degree.(2)Considering that the ship rescue force demand problem of cruise ship accident is a small sample statistical problem,and its influencing factors index is relatively complex,so the support vector regression theory is established Regression(SVR)method as the theoretical basis of the prediction model,combined with the characteristics of SVR theory and the relevant characteristics of ship rescue force demand,the SVR prediction model was optimized,and the ship rescue force demand forecasting model based on principal component analysis(PCA)and support vector regression machine theory was constructed.(3)By preprocessing the collected data of shipwrecks,the optimization model of relevant parameters is adjusted.Through the model training,testing and analysis and comparison with other prediction models(Bayesian Regression Model,Decision Tree Regression Model),it is found that the change curve between the predicted value and the real value of PCA-SVR model is the best.Therefore,it is considered that PCA-SVR model has better fitting effect and higher prediction accuracy.PCA-SVR model is used to predict cruise ship accidents,so as to verify the validity of the model in this paper.The number of ships dispatched by the model is slightly lower than the actual value.First,the accident data used for training and testing in this paper have a good rescue effect.Therefore,the predicted output result of the model in this paper also conforms to this feature.The second reason is that there is no prediction of the ship rescue force demand in the actual rescue,which leads to the delayed dispatch of rescue force and low rescue efficiency.If we make a timely response when an accident occurs,the number of rescue forces can be predicted in the first time and deployed in a timely manner,and the efficiency of rescue is also relatively greatly improved.The above prediction results verify the validity and feasibility of PCA-SVR model.