摘 要
隨著工業(yè)的飛躍發(fā)展，人類對(duì)于能源的需求消耗越來(lái)越多，陸地和近海領(lǐng)域能源被開采殆盡，而深海領(lǐng)域蘊(yùn)藏著豐富的資源，深海鉆井平臺(tái)在這種背景下應(yīng)運(yùn)而生。面對(duì)深海復(fù)雜的海況鉆井平臺(tái)要保持其定位精度以完成采油任務(wù)，傳統(tǒng)的錨泊技術(shù)已經(jīng)無(wú)法滿足深海作業(yè)的定位要求，越來(lái)越多的海洋平臺(tái)采用動(dòng)力定位系統(tǒng)。
動(dòng)力定位系統(tǒng)利用自身推進(jìn)器產(chǎn)生的推力，有效產(chǎn)生力矩和反力矩以抗衡外界力和力矩以保持平臺(tái)定位。推力分配問題實(shí)質(zhì)上是求解非線性方程組最優(yōu)解的問題，即在滿足一定的約束條件基礎(chǔ)上，在眾多組合中尋找每個(gè)推進(jìn)器最適當(dāng)?shù)耐屏Ψ岛头较颍蛊溥_(dá)到推進(jìn)系統(tǒng)能耗最優(yōu)和誤差最優(yōu)的目標(biāo)。本文以半潛式深海平臺(tái)為研究對(duì)象，主要對(duì)推力分配策略進(jìn)行研究和比較。
首先針對(duì)鉆井平臺(tái)具體對(duì)象，對(duì)外界風(fēng)浪流載荷、平臺(tái)運(yùn)動(dòng)模型等進(jìn)行分析。同時(shí)根據(jù)半潛式海洋平臺(tái)的動(dòng)力定位系統(tǒng)的特殊性，闡述深海鉆井平臺(tái)推進(jìn)器的選型、布置并分析其原因，并描述推進(jìn)器與平臺(tái)之間的推力特性。
然后結(jié)合半潛式深海平臺(tái)推進(jìn)器配置方案，重點(diǎn)分析在此配置方案下的推力分配策略。采用序列二次規(guī)劃法進(jìn)行推力最優(yōu)分配，推力分配目標(biāo)旨在最小化推力系統(tǒng)的能耗和推力誤差，同時(shí)考慮推進(jìn)器的推力區(qū)域、推力極限等約束，并實(shí)現(xiàn)對(duì)推力區(qū)域的線性化處理，通過(guò)仿真結(jié)果指出序列二次規(guī)劃算法的不足，其對(duì)初值的依賴性強(qiáng)，結(jié)果不穩(wěn)定，易陷入局部最優(yōu)。
最后根據(jù)動(dòng)力定位系統(tǒng)推力系統(tǒng)的基本要求，以及推力分配問題在近年來(lái)研究現(xiàn)狀，深入了解比較各種優(yōu)化算法的優(yōu)劣，在此基礎(chǔ)上提出基于量子粒子群算法的推力分配策略，并采用量子粒子群算法制定了詳盡的辦法解決推力分配問題。在相同的推力指令要求下，分別基于能耗最優(yōu)和誤差最優(yōu)兩種不同的海況要求，比較兩種算法的推力分配結(jié)果并分析。仿真結(jié)果表明量子粒子群算法具有更好的尋優(yōu)結(jié)果，但尋優(yōu)時(shí)間相對(duì)序列二次規(guī)劃法長(zhǎng)，所以對(duì)于量子粒子群算法應(yīng)用于工程應(yīng)用還需要進(jìn)一步改進(jìn)。
關(guān)鍵詞 動(dòng)力定位系統(tǒng)；推力分配；序列二次規(guī)劃法；量子粒子群算法；



Abstract
With the development of industry in the world, more and more resources are required and consumed. As a result, resources on land and the sea close to the shore will be mined out in the near future. There are extraordinary abundant resources in the deep sea, so deepwater drilling platform arises at the historic moment. How to operate the drilling platform in the complex deepwater surroundings to fulfill drilling cannot afford to ignore. Traditional way to anchor the platforms cannot meet the orientation require in the deepwater surroundings. More and more drilling rigs have adopted dynamic positioning system (DPS).
Compared to the traditional way to anchor the platforms, DPS totally depends on its own thrust system to generate force and moment to orient the drilling rig. Thrust allocation is defined to search the best thrust of every propeller in order to meet the objective, such as minimize the energy and thrust error, by using the platform’s own propellers. Also the thrust must satisfy some restriction.
Firstly according to the particularity of the dynamic positioning system, the selection and arrangement of the propellers on the drilling rig is expatiated and the reason is explained as well. What’s more, the thrust characteristics of the propeller are depicted. Integrating the semi-submersible offshore platform’s thruster configuration scheme of this topic, the thrust allocation problem in this case is focused on the analysis. Semi-submersible drilling rigs in this thesis possess eight azimuth propellers. The objective of thrust allocation is to minimize the energy consumption of thrust system and the thrust error, considering the restriction such as the limit of the propellers and singularity and so on. This thesis adopted SQP (Sequence quadratic programming) to solve the optimal solution of thrust allocation at first. The shortcomings using SQP methods are illustrated, for example, it strongly depends on the initial value and is easy to fall into the local optimum. Then by means of analyzing the essential request and the development of the thrust allocation, and in-depth understanding on the various kinds of optimization algorithms, based on which, this thesis put forward a new method using QPSO (Quantum particle swarm optimization) method to solve the thrust allocation. Then QPSO combined with disjunctive programming techniques are used to present a more elaborate solution to optimal thrust allocation problems. The simulation results showed that the method using QPSO possess good effect to search for the optimal result for the thrust allocation at last.
Keywords: Dynamic position system; thrust allocation; Sequence quadratic programming; Quantum particle swarm optimization.


目錄
摘 要 I
Abstract III
第1章 緒論 1
1.1 課題研究背景及意義 1
1.2 半潛式平臺(tái)及其動(dòng)力定位系統(tǒng) 2
1.3 國(guó)內(nèi)外推力分配算法研究現(xiàn)狀 5
1.4 量子粒子群算法的發(fā)展及其優(yōu)點(diǎn) 6
1.5 本論文主要研究?jī)?nèi)容 7
第2章 半潛式平臺(tái)動(dòng)力定位系統(tǒng)數(shù)學(xué)建模 10
2.1 半潛式深海平臺(tái)主要參數(shù) 10
2.2 環(huán)境載荷模型 11
2.2.1 風(fēng)載荷 11
2.2.2 流載荷 13
2.2.3 波浪二階力 14
2.3 半潛式鉆井平臺(tái)的數(shù)學(xué)模型 14
2.3.1 平臺(tái)運(yùn)動(dòng)模型 14
2.3.2 平臺(tái)控制模型 15
2.3.3 Matlab模型仿真 17
2.4 本章小結(jié) 19
第3章 半潛式鉆井平..