Ice Resistance Data Correction via Subdomain Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
There is a lack of effective guidance methods for ship ice resistance experiments in broken ice fields, leading to discrepancies between actual and design conditions, which affects the reliability of ice resistance measurements.
Innovation Solution
A method involving uniform distribution of ice floes, repeated experiments, segmentation of the broken ice field into subdomains, outlier data judgment, image correction, and calculation of actual coverage rates to modify ice resistance data, ensuring consistency with design conditions and reducing experimental uncertainty.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If ice floes are uniformly distributed in the broken ice field, then the experimental setup conforms to design working conditions, but water surface perturbance causes the actual distribution to differ from the uniform distribution
Solution Approach 1:
The patent applies preliminary action by uniformly distributing ice floes before each experiment and eliminating mutually overlapped ice floes in advance. This preparatory arrangement ensures that the ice field is properly configured before the ship begins its experiment, minimizing the impact of water surface perturbance during the actual experiment.
Solution Approach 2:
The patent implements periodic action by repeating the ice floe distribution and experiment process at least three times. This periodic repetition allows for multiple measurements under similar conditions, enabling statistical analysis and reducing the impact of random perturbances on the overall experimental results.
2Speed
If the ship passes through the broken ice field at constant speed, then the experiment can be conducted smoothly, but the actual ice resistance measurement may be affected by distribution deviations
Solution Approach 1:
The patent applies segmentation by dividing the broken ice field into multiple subdomains along the ship's track. This segmentation allows for localized analysis of ice resistance in different sections, enabling the identification and elimination of outlier data from abnormal subdomains while retaining data from normal subdomains.
Solution Approach 2:
The patent uses parameter changes by calculating the actual coverage rate of ice floes in each subdomain and using this parameter to modify the ice resistance measurements. This approach adjusts the measured data to account for deviations from the ideal uniform distribution, improving measurement accuracy.
3Reliability
If repeated experiments are conducted at least three times, then the impact of water surface perturbance is reduced, but the experimental time and resource consumption increase
Solution Approach 1:
The patent applies partial action by conducting exactly three repeated experiments, which is sufficient to reduce the impact of perturbances without excessive time consumption. Additionally, the segmentation approach allows for partial use of experimental data by eliminating only the outlier subdomains while retaining valid data from other subdomains.
Data Source
AI summary
In a method of ice distribution and data processing for ship ice resistance experiment in broken ice field, the ice floe is uniformly distributed in the broken ice field before each ship ice resistance experiment with eliminating the mutually overlapped broken ice to make it close to the experimental design working condition, and independent repeated experiments are carried out on each ship ice resistance experiment to reduce the influence of accidental factors; after finishing the experiment, the experimental data are divided into intervals and the experimental data corresponding to unreasonable subdomains are eliminated by mathematical statistics, and the uncertainty analysis is carried out; image correction is carried out for pictures of in broken ice field in the pre-experimental data, and the pictures are divided to obtain subdomain pictures to calculate the actual coverage rate of each subdomain; the ice resistance in the experimental data corresponding to each remaining subdomain is modified.