起重機(jī)路徑規(guī)劃的摘要

摘 要工程實(shí)踐中吊裝方案設(shè)計(jì)仍然停留在傳統(tǒng)模式，采用手工校核方式選擇吊索具、通過查閱性能表選擇起重機(jī)、利用AutoCAD等常規(guī)二維繪圖軟件設(shè)計(jì)吊裝過程等。這種傳統(tǒng)設(shè)計(jì)方法嚴(yán)重依賴設(shè)計(jì)者的經(jīng)驗(yàn)，需要多次迭代，設(shè)計(jì)周期長，吊裝前無法全面、直觀地查看動(dòng)態(tài)的吊裝過程。在此背景下，計(jì)算機(jī)輔助吊裝方案設(shè)計(jì)（CALPAD）技術(shù)應(yīng)運(yùn)而生。其中，起重機(jī)選型可快速選擇合適的起重機(jī)，而路徑規(guī)劃及吊裝仿真可直觀、快捷地設(shè)計(jì)吊裝過程并進(jìn)行預(yù)演。然而，通過對比工程需求和現(xiàn)有研究發(fā)現(xiàn)：對應(yīng)用廣泛的桁架臂履帶起重機(jī)選型較少；單機(jī)吊裝路徑規(guī)劃未考慮起重機(jī)的行走；雙機(jī)吊裝仿真研究較少，現(xiàn)有方法待設(shè)置參數(shù)多且難以確定。由于大多現(xiàn)有研究缺乏對某些工程約束的考慮，致使其難以真正應(yīng)用到實(shí)際吊裝工程。為此，本文在多重約束下起重機(jī)智能選型、考慮行走的單機(jī)吊裝路徑規(guī)劃、雙機(jī)吊裝仿真三方面開展研究并實(shí)現(xiàn)相應(yīng)的軟件，所形成的理論成果和軟件已成功應(yīng)用到實(shí)際吊裝工程。本文主要研究工作如下：（1）給出了一種面向移動(dòng)式起重機(jī)的多重約束起重機(jī)選型算法。首先構(gòu)建了多重約束的起重機(jī)選型數(shù)學(xué)模型，基于此給出了選型算法的總體框架，然后以桁架臂履帶起重機(jī)為例詳細(xì)闡述了算法框架中起重性能、被吊物與臂架間距、接地比壓約束處理的實(shí)現(xiàn)，最后通過實(shí)例驗(yàn)證了算法的可用性和有效性。與現(xiàn)有選型算法相比，技術(shù)上該算法將復(fù)雜三維空間距離計(jì)算問題轉(zhuǎn)化為二維幾何計(jì)算，降低了間距約束處理難度；該算法由于無需起重機(jī)、被吊物三維建模，在應(yīng)用上更加便捷。（2）提出了基于先驗(yàn)信息的路徑規(guī)劃新算法RRT-Connect+。針對RRT-Connect所得路徑不優(yōu)的問題，利用先驗(yàn)信息引導(dǎo)生成樹向高質(zhì)量區(qū)域生長，具體給出了基于回歸分析的可變多步擴(kuò)展策略、基于采樣池和未探索區(qū)的隨機(jī)點(diǎn)選擇策略。多個(gè)仿真實(shí)驗(yàn)的結(jié)果表明，RRT-Connect+大幅提升路徑質(zhì)量的同時(shí)還保持了原RRT-Connect的高效性，并且總體性能優(yōu)于現(xiàn)有基于RRT的其它算法。此外，該算法保持了RRT處理微分約束的簡便性，無需重新設(shè)計(jì)類似Steer操作的局部規(guī)劃器，可應(yīng)用于實(shí)時(shí)路徑規(guī)劃。（3）提出了一種考慮行走的單臺(tái)履帶起重機(jī)吊裝路徑規(guī)劃算法。首先對規(guī)劃問題進(jìn)行數(shù)學(xué)建模，并設(shè)計(jì)了基于RRT-Connect+的吊裝路徑規(guī)劃算法，給出了位形空間的定義、兩位形間的距離度量、履帶起重機(jī)非完整運(yùn)動(dòng)學(xué)約束的表達(dá)，最后通過三個(gè)仿真實(shí)驗(yàn)驗(yàn)證算法的有效性和性能，結(jié)果表明該算法能在各種復(fù)雜吊裝環(huán)境中找到一條無超載、無碰撞可行路徑。算法將履帶起重機(jī)行走的非完整運(yùn)動(dòng)學(xué)約束融入算法中，使路徑更自然、平滑。此外，其中的距離度量將長度量綱和角度量綱巧妙地統(tǒng)一起來，較好地表達(dá)吊裝路徑長度，并且賦予距離度量直觀物理意義，避免了為每個(gè)分量設(shè)置權(quán)重系數(shù)。（4）提出了一種基于空間幾何約束的雙機(jī)協(xié)同吊裝仿真方法。針對典型吊裝工況的雙機(jī)吊裝仿真問題，研究了雙機(jī)之間的協(xié)同，設(shè)計(jì)了雙機(jī)協(xié)同吊裝仿真模型，給出了雙機(jī)系統(tǒng)基本動(dòng)作的表達(dá)與設(shè)計(jì)。通過實(shí)例驗(yàn)證仿真方法的可用性和有效性，結(jié)果表明該方法可容易地模擬典型工況的雙機(jī)吊裝過程。由于雙機(jī)的協(xié)同策略已嵌入基本動(dòng)作的實(shí)現(xiàn)，因此，在典型雙機(jī)吊裝中該方法吊裝過程仿真更準(zhǔn)確、仿真操作更簡便。此外，該方法將兩臺(tái)起重機(jī)和被吊物看成一個(gè)完整的復(fù)雜系統(tǒng)（稱為雙機(jī)系統(tǒng)），該概念為雙機(jī)協(xié)同吊裝的其它研究提供了一個(gè)全新的視角。（5）提出了一種基于正向運(yùn)動(dòng)學(xué)的雙機(jī)吊裝仿真通用方法。首先從靜力學(xué)的角度探究雙機(jī)吊裝中起升系統(tǒng)部分的運(yùn)動(dòng)學(xué)規(guī)律，利用最小勢能原理將起升系統(tǒng)部分的運(yùn)動(dòng)學(xué)抽象為帶約束的數(shù)學(xué)優(yōu)化問題，提出基于最小勢能原理的起升繩偏擺角及起升力求解算法，通過與ADAMS仿真結(jié)果對比驗(yàn)證了求解算法的正確性。然后在此基礎(chǔ)上設(shè)計(jì)了基于正向運(yùn)動(dòng)學(xué)的雙機(jī)吊裝仿真通用方法。最后通過實(shí)例驗(yàn)證方法的可用性和有效性，結(jié)果表明該方法是一種設(shè)計(jì)和預(yù)演雙機(jī)吊裝過程的有效手段。相比現(xiàn)有基于動(dòng)力學(xué)的雙機(jī)吊裝仿真方法，該方法僅需被吊物的重心相對位置及重量即可準(zhǔn)確求得吊裝過程中被吊物位姿及起升力，為雙機(jī)吊裝仿真提供有力的支撐，可容易地嵌入吊裝仿真軟件中，實(shí)現(xiàn)實(shí)時(shí)的雙機(jī)吊裝作業(yè)仿真，具有參數(shù)少、實(shí)時(shí)等特點(diǎn)。關(guān)鍵詞：吊裝方案；起重機(jī)；起重機(jī)選型；吊裝仿真；路徑規(guī)劃Engineering practice of hoisting scheme design still remain in the traditional mode, manual calibration rigging selection, through consulting performance table select crane, using AutoCAD and conventional two-dimensional mapping software design of hoisting process.The traditional design methods depend heavily on the experience of the designer, need many iterations, long design period, before lifting, is unable to visualize the dynamic hoisting process. Under this background, the computer aided design of hoisting scheme (CALPAD) technology emerge as the times require. Among them, the crane can quickly select suitable selection of crane, and the path planning and hoisting simulation can directly, quickly design process of hoisting and rehearsal. However, by comparing the engineering needs and existing research findings: on the application of a wide range of truss boom crawler crane selection less; single lifting path planning without considering the crane walking; double hoisting simulation study is less, the existing method for setting parameters are difficult to determine. Since most existing studies are lacking for some engineering constraints are considered, which is difficult to truly applied to actual hoisting engineering.Therefore, based on the multiple constraints crane intelligent selection, consider walking single lifting path planning, double hoisting simulation three aspects of the research and the realization of the corresponding software, by the formation of the theoretical results and the software has been successfully applied to the actual hoisting engineering. In this paper, the main research work is as follows:(1) presents a mobile crane crane selection algorithm with multiple constraints. First constructed a multiple constrained crane type mathematical model, based on the given selection algorithm framework, and then to the truss boom crawler crane as an example, describes the algorithm framework of lifting performance, the hanging object and arm frame spacing, ground pressure constraint processing is realized, finally the example proves that the method is available and the validity of the. With the existing selection algorithm, the algorithm will technically complex 3D spatial distance computation into two-dimensional geometric calculation, reduces the spacing constraint processing difficulty; the algorithm without crane, the hanging object modeling, in the application of more convenient.(2) Based on a priori information about the new path planning algorithm for RRT-Connect+. The path for RRT-Connect and problem, the use of prior information to guide the generation tree to high quality regional growth, concrete is presented based on the regression analysis of variable multi step expansion strategy, based on the sampling pool and unexplored area of the random selection strategy. A number of simulation results show that, RRT-Connect+ significantly enhance the quality of route while maintaining the original RRT-Connect efficiency, and overall performance is better than other algorithms based on RRT.In addition, this algorithm keeps the RRT differential constraints is simple, no need to design similar to the Steer operation for local programming, can be applied to real-time path planning.(3) This paper presents a method for considering walking single crawler crane hoisting path planning algorithm. The planning problem in mathematical modeling, and design based on the RRT-Connect+ lifting path planning algorithm, gives the configuration space is defined, two form distance metric, crawler crane nonholonomic kinematic constraint expression, through three finally simulation results verify the effectiveness of the algorithm and performance, the results show that the algorithm can a variety of complex lifting environment to find a feasible path without overload, without collision.Algorithm crawler crane walking nonholonomic kinematic constraints into the algorithm, so that a more natural, smooth path. In addition, the distance measure the length dimension and angle dimension ingenious united, better expression of hoisting the length of the path, and give the distance metric intuitive physical significance, avoid for each component set weight coefficient.(4) Proposed one kind based on the space geometry constraint dual cooperative lifting simulation method. Aiming at the typical hoisting condition double hoisting simulation problems, study the dual synergy between, designed dual cooperative lifting simulation model, given the dual system of basic action expression and design. Through the example to validate the simulation method is feasible and effective, the results show that the method can be easily simulated typical operating conditions of double hoisting process.Due to dual coordination strategy has embedded basic action is achieved, thus, in a typical double hoisting method of the hoisting process simulation simulation more accurate, more simple operation. In addition, the two crane and the lifted object as a whole complex system (known as the dual system), the concept of dual cooperative lifting other research provides a new angle of view.(5) Proposed one kind based on the forward kinematics of double hoisting simulation method. First from the static point of view to explore dual lifting hoisting system part of the kinematic law, using the principle of minimum potential energy will rise system part of the kinematic abstract as a constrained mathematical optimization problems, is proposed based on the principle of minimum potential energy of the hoisting rope swinging angle and lifting algorithm, through the ADAMS simulation results validated the solution the correctness of the algorit