基于改進(jìn)K近鄰算法的船舶通航環(huán)境智能識(shí)別

王壯　李嘉源　黃連忠　王凱　姜雅喬　馬冉祺

摘要：為實(shí)現(xiàn)船舶能效的智能優(yōu)化，從船舶通航環(huán)境出發(fā)，研究船舶通航環(huán)境的智能識(shí)別方法。基于所獲得的船舶通航環(huán)境大數(shù)據(jù)，建立基于改進(jìn)K均值聚類算法的船舶通航環(huán)境類別知識(shí)庫(kù)，設(shè)計(jì)相關(guān)系數(shù)加權(quán)的K近鄰算法，實(shí)現(xiàn)船舶通航環(huán)境的智能識(shí)別。實(shí)例分析結(jié)果表明，此基于改進(jìn)K近鄰算法的智能識(shí)別方法的識(shí)別準(zhǔn)確率可達(dá)97.25%，相對(duì)于未改進(jìn)的K近鄰算法，準(zhǔn)確率提高7.81%。所提出的智能識(shí)別方法可為基于通航環(huán)境智能識(shí)別的船舶分段航速智能優(yōu)化方法的研究奠定基礎(chǔ)。

關(guān)鍵詞：通航環(huán)境; K均值聚類; K近鄰算法; 智能識(shí)別

中圖分類號(hào)：? U676.3

文獻(xiàn)標(biāo)志碼：A

Intelligent identification of ship navigation environment

based on improved K-nearest neighbor algorithm

WANG Zhuang， LI Jiayuan， HUANG Lianzhong， WANG Kai，

JIANG Yaqiao， MA Ranqi

（Marine Engineering College， Dalian Maritime University， Dalian 116026， Liaoning， China）

Abstract：

In order to realize the intelligent optimization of ship energy efficiency， starting from the ship navigation environment， the intelligent identification method of the ship navigation environment is studied. Based on the obtained ship navigation environment big data， a knowledge base of ship navigation environment categories based on the improved K-means clustering algorithm is established， and a K-nearest neighbor algorithm weighted by correlation coefficients is designed to realize the intelligent identification of ship navigation environment. The example analysis results show that， the identification accuracy of the intelligent recognition method based on the improved K-nearest neighbor algorithm can reach 97.25%， which can increase the accuracy rate by 7.81% compared with the unimproved K-nearest neighbor algorithm. The proposed intelligent identification method can lay the foundation for the research on the intelligent optimization method of ship segmentation speed based on the intelligent identification of navigation environment.

Key words：

navigation environment; K-means clustering; K-nearest neighbor algorithm; intelligent identification

0 引 言

船舶在航行期間所處的通航環(huán)境對(duì)船舶能效有重要影響，在燃油價(jià)格高居不下[1]、水運(yùn)能源消耗量大、船舶排放對(duì)環(huán)境造成污染的背景下[2-4]，節(jié)能減排一直是航運(yùn)領(lǐng)域的熱點(diǎn)話題。當(dāng)船舶在水面航行時(shí)，風(fēng)、浪、流等通航環(huán)境要素會(huì)對(duì)船舶運(yùn)行狀態(tài)產(chǎn)生影響，通航環(huán)境能夠引起船舶的附加阻力，如風(fēng)阻、波浪增阻等，從而使得船舶的總阻力增加，進(jìn)而影響船舶的航行和能效水平。IMO對(duì)某型船的研究表明，船舶推進(jìn)力中有三分之一左右用于克服由興波、水流和風(fēng)速等環(huán)境因素引起的增阻[5]。在通航環(huán)境研究方面：張寶吉[6]和KURODA等[7]提出了船舶波浪增阻的簡(jiǎn)化算法;王典[8]通過(guò)在非結(jié)構(gòu)網(wǎng)格上模擬由水和空氣構(gòu)成的自由界面問(wèn)題，對(duì)水面波浪進(jìn)行數(shù)值模擬，并分析了其對(duì)船舶航行的影響;WANG等[9]考慮多變的通航環(huán)境對(duì)船舶能效進(jìn)行了優(yōu)化;YAN等[10]考慮通航環(huán)境對(duì)內(nèi)河船舶航速進(jìn)行了優(yōu)化。……