摘 要
隨著科技的不斷進(jìn)步，我國疏浚作業(yè)能力已取得長足發(fā)展，疏浚設(shè)備也隨之得到更新。我國疏浚設(shè)備朝著大型化和自動化方面發(fā)展，但是對疏浚高效化方面還缺少研究，在現(xiàn)有的疏浚設(shè)備基礎(chǔ)上提高疏浚性能、提升疏浚效率是我國疏浚事業(yè)亟待研究和發(fā)展的方向。基于此開展耙吸挖泥船疏浚機(jī)理研究，探討疏浚優(yōu)化工況方法，對提高挖泥船疏浚效率具有重大意義。
本課題從耙吸挖泥船疏浚機(jī)理出發(fā)，建立基于控制的耙吸挖泥船數(shù)學(xué)模型，通過對疏浚過程和生產(chǎn)效率的智能評估與分析，采用模型預(yù)測控制（MPC）技術(shù)，獲取疏浚優(yōu)化的最佳策略，實現(xiàn)在不同土質(zhì)和不同疏浚裝備工況條件下的優(yōu)化控制。
首先在考慮土壤等因素的影響條件下，論文建立基于控制的耙吸挖泥船挖掘裝艙過程的數(shù)學(xué)模型，對耙頭挖掘過程和泥艙沉積過程進(jìn)行了機(jī)理分析與數(shù)學(xué)建模，并采用挖泥船實測數(shù)據(jù)對模型進(jìn)行驗證。驗證結(jié)果表明模型具有很高的準(zhǔn)確性，可用于MPC控制設(shè)計。
其次在系統(tǒng)分析和研究挖泥船疏浚過程的基礎(chǔ)上，提出一種基于模型預(yù)測控制的在線疏浚優(yōu)化的方法，優(yōu)化的目的是使挖泥船在完整疏浚周期內(nèi)的產(chǎn)量最大化。MPC控制器由數(shù)學(xué)模型、目標(biāo)函數(shù)和優(yōu)化器三個部分構(gòu)成。
疏浚優(yōu)化過程是一個復(fù)雜的多系統(tǒng)耦合，多約束條件問題。本文將遺傳算法運(yùn)用于MPC控制器的優(yōu)化器，使優(yōu)化器能在大搜索空間中以相對較少的時間達(dá)到最優(yōu)值，尋找到最佳的可控疏浚參數(shù)。然后采用“新海鳳”號自航耙吸式挖泥船工程實測數(shù)據(jù)對該MPC控制器及其優(yōu)化算法進(jìn)行了仿真驗證，并與現(xiàn)有控制技術(shù)進(jìn)行性能比較，結(jié)果表明MPC優(yōu)化方法能夠使挖泥船的疏浚周期縮短10%～18%，周期產(chǎn)能提高10%。
最后采用LabVIEW開發(fā)了基于MPC的疏浚優(yōu)化控制系統(tǒng)人機(jī)界面。系統(tǒng)根據(jù)周期產(chǎn)量和時間效率對疏浚性能進(jìn)行評估，給出歷史最優(yōu)疏浚周期以及與疏浚工況條件相適應(yīng)的最優(yōu)可控參數(shù)，以達(dá)到疏浚產(chǎn)量最大化，提高挖泥船的疏浚效率和性能。


關(guān)鍵字：自航耙吸挖泥船；模型預(yù)測控制；疏浚優(yōu)化；遺傳算法；性能評估












































Abstract
With the rapid development of science and technology, our dredging cause has made great progress and dredging equipment has also got updated. Great progress has made in dredging equipment in the aspect of its large scale and automation, while the high efficiency of dredging is still lack of study. On the basis of dredging equipments, the improvement of dredging performance and elevating dredging efficiency are the aspects what our dredging cause should be researched and developed. Research on the mechanism of trailing suction hopper dredger dredging and search of the optimal ways will mean to the improvement of the dredger’s efficiency of construction and the market competitiveness.
On the foot of the mechanism of trailing suction hopper dredger dredging, a control-based trailing suction hopper dredger dredging mathematics model was researched and built, and intelligent analysis and eva luation was made to dredging process and production efficiency. Model-based predicative control strategy (MPC) was adopted to obtain the best strategy for dredging optimization and to achieve dredging optimization control under different soil conditions and different dredging equipment.
Firstly, taking the soil and other factors into account, the thesis presents a mathematics model of hopper process. Besides, the thesis models and simulates drag-head and hopper and verifies these models by real measured data. The result indicates the model is of high accuracy and can be used in MPC control design.
Secondly, on the basis of system analysis and study of dredgers’ dredging process, an online dredging optimization method developed by model predictive control was brought up. The purpose of optimization is to maximize the dredgers’ production within a complete dredging cycle. MPC controller consists of three parts: mathematics model, target function and optimizer.
Dredging optimization process is a complex, multi-system coupling and a multi-constraint conditions problem. Genetic Algorithm was adapted to MPC controllers’ optimizer in the thesis. Then the optimized values, the best controllable dredging parameters can be found by optimizer in a large search room and within less time. The optimization algorithm and MPC controller were stimulated and validated, using the project data measured on “Xinhai feng” trailing suction hopper dredger, and performance contrast with updated control technology. The result shows that MPC optimization shortens the dredging time of 10% --18%, and increases dredging efficiency of 10%.

Lastly, man-machine interface is developed based on LabVIEW Software. According to cycle production and time efficiency, dredging performance was eva luated by system. Then the optimal dredging cycle and the best controllable parameters suitable to dredging conditions were given in record to maximize the high dredging production and to improve the dredging efficiency and its performance.

Key words: Trailing Suction Hopper Dredger; Model Predictive Control; Dredging Optimization; Genetic Algorithm; Performance Assessment















-b..