Research On Prediction And Evaluation Method Of Continuous Icebreaking Capability Of Polar Ships

Global warming has further exacerbated sea ice melting,and seaworthiness of the Arctic has been improved year by year.As an important equipment carrier for studying the Arctic,polar vessels are the basis for the implementation of polar ocean engineering equipment.It plays an important role on the polar stage.It is particularly important to design,predict,evaluate and build polar ships with strong icebreaking capabilitiesContinuous icebreaking capability is an important indicator of the performance of polar ships.The purpose of this paper is to establish a method for predicting and evaluating the continuous icebreaking ability of polar ships.A method of equilibrium between ice resistance and net thrust is proposed to predict the continuous ice-breaking capability of polar ships at different speeds.Firstly,the ice resistance is divided into ice-breaking resistance and underwater ice resistance,and the ice resistance prediction is carried out at speeds of 2,3 and 4 knots.The ice-breaking resistance of ship is simulated numerically by the peridynamic method,and the underwater ice resistance is solved by the Lindqvist empirical method.At the same time,with reference to the concept of net thrust in the Finnish-Swedish ice class rules,the net thrust estimation at different speeds is completed,so as to realize the prediction of continuous ice-breaking capacity.The prediction results are consistent with the actual icebreaking ability,which verifies the feasibility of the continuous ice breaking ability prediction methodIn addition,research on continuous ice-breaking ability evaluation method based on radial basis neural network.The principle of RBF neural network is briefly introduced.The parameters affecting continuous icebreaking capacity are selected:icebreaker bow,ship length L,ship width B,draft T,displacement,power P,and speed V as training input parameters,and the corresponding ice-breaking thickness Hice is used as the training output.The data is processed in dimensionless,and establishing an ice-breaking ability evaluation model based on RBF neural network.The final training parameters are as follows:icebreaker bow,B/L,T/L,CB,P/Pmax,V/Vmax,and the corresponding output Hice/Hunit.Fifty-eight training data are determined.Six training samples were selected as forecast samples to carry out model verification work.Finally,the icebreaking capacity of the target ship is evaluated and compared with the actual icebreaking ability results.The icebreaking capacity evaluation value is within a reasonable range,and the evaluation results under the training data-intensive 3kn speed are close to the prediction results,which validates the effectiveness of the RBF neural network model to evaluate the continuous icebreaking capacity of polar ships.It can further improve the parameters affecting the ice breaking ability and expand the training sample size to improve the accuracy of the evaluation model.