一种带多约束的船体分段点集数据分析方法

为了达到船体分段建造的精度要求,船体分段的测量点集与设计点集的数据分析技术是关键。在船舶领域中,传统的点集数据分析方法没有考虑多种船舶约束,因此数据分析出来的结果与实际结果相差较大。针对此情况,提出一种考虑多种船舶约束的数据分析方法。首先采用基于高斯混合模型(GMM)的相干点漂移法(CPD)算法,获得数据分析初值;然后利用权值向量实现对不同方向上精度要求的误差分配,同时把垂直度、平面性和水平度等多种工程约束引入数据分析的多优化目标函数中,通过调整多约束权值,求解出更合理的数据分析结果。实例表明,该方法在满足多种船舶约束的情况下,可获得更合理的数据分析结果,为后续的合拢搭载提供一定依据。

Point registration algorithm for hull blocks with engineering constraints

In order to achieve the accuracy requirements of the hull blocks construction, the point set data analysis technique of measurement point set and design point set for hull blocks is very important. In the ship"s field, various constraints are not considered in the traditional point set data analysis technique, so the data analysis results are very different from the actual one. To solve the problem, a point set data analysis algorithm which considered a variety of shipbuilding constraints was presented in this article. Firstly, the initial data analysis values were obtained by using the coherent point drift (CPD) algorithm based on the Gaussian mixture model (GMM). Secondly, the error distribution in different directions was determined by using weight vectors, and the various engineering constraints (e.g. verticality, flatness and levelness) were introduced into the multi - optimization objective function, after adjusting multiple constraint weights, the more reasonable data analysis results could be obtained. The results obtained by this method prove that the method can obtain the more reasonable data analysis results which meet the various shipbuilding constraints. It can provide the basis for the subsequent assembly.