WO2022156440A1 - 基于时间预估模型的agv调度方法 - Google Patents

基于时间预估模型的agv调度方法 Download PDF

Info

Publication number
WO2022156440A1
WO2022156440A1 PCT/CN2021/138631 CN2021138631W WO2022156440A1 WO 2022156440 A1 WO2022156440 A1 WO 2022156440A1 CN 2021138631 W CN2021138631 W CN 2021138631W WO 2022156440 A1 WO2022156440 A1 WO 2022156440A1
Authority
WO
WIPO (PCT)
Prior art keywords
time
agv
estimation
static
stage
Prior art date
Application number
PCT/CN2021/138631
Other languages
English (en)
French (fr)
Inventor
李永翠
刘耀徽
陈强
张晓�
刘长辉
张雪飞
丛安慧
柳璠
孙秀良
Original Assignee
青岛港国际股份有限公司
青岛新前湾集装箱码头有限责任公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 青岛港国际股份有限公司, 青岛新前湾集装箱码头有限责任公司 filed Critical 青岛港国际股份有限公司
Priority to JP2023543308A priority Critical patent/JP2024503141A/ja
Publication of WO2022156440A1 publication Critical patent/WO2022156440A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B19/00Programme-control systems
    • G05B19/02Programme-control systems electric
    • G05B19/418Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
    • G05B19/4189Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM] characterised by the transport system
    • G05B19/41895Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM] characterised by the transport system using automatic guided vehicles [AGV]
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00Program-control systems
    • G05B2219/30Nc systems
    • G05B2219/32Operator till task planning
    • G05B2219/32252Scheduling production, machining, job shop
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02PCLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/60Electric or hybrid propulsion means for production processes

Definitions

  • the invention belongs to the technical field of automated terminal transportation, and in particular relates to an AGV scheduling method based on a time estimation model.
  • the operation belt in front of the container terminal refers to the area between the front line of the yard and the front line of the terminal. Its function is to serve the loading and unloading operations of the terminal quay bridge and the operation of containers entering and leaving the yard.
  • manual terminals and semi-automatic terminals manual driving is usually used.
  • Internal collection card In recent years, with the continuous rise of labor costs, more and more terminals have converted their horizontal transportation equipment from internal trucks to AGVs (Automated Guided Vehicles). The use of AGVs has greatly improved production efficiency and saved costs. .
  • the interaction between the cargo ship and the automatic guided vehicle is realized through the bridge crane
  • the interaction between the automatic guided vehicle and the yard is realized through the rail crane
  • the wharf and the stack are realized through AGV, L-AGV, Auto Shuttle, etc. Automated job handover between sites.
  • the transportation equipment scheduling of some automated terminals can only ensure the stability of the operation of the entire scheduling system by reducing the number and speed of transportation equipment, but this makes the operation efficiency of the entire terminal very low.
  • AGV operation monitoring has not yet been implemented. Precedents can be found. Based on this, how to invent a method that can improve the efficiency of AGV container transportation is the main technical problem solved by the present invention.
  • the present invention proposes an AGV scheduling method based on a time estimation model, which can solve the above problem.
  • the present invention adopts the following technical scheme to realize:
  • An AGV scheduling method based on a time estimation model comprising the following steps:
  • the static estimation step is to estimate the static estimation value of the operation efficiency of the bridge crane and the rail crane; respectively estimate the static estimation time of the container in each operation stage, including: the static estimation time of the AGV and the bridge crane and the rail crane in the operation interaction stage, the AGV driving The static estimation time of the stage, the static estimation time of the bridge crane operation stage, and the static estimation time of the rail crane operation stage;
  • the dynamic estimation step is to estimate the dynamic estimation value of the operation efficiency of the bridge crane and the rail crane; respectively estimate the dynamic estimation time of the container in each operation stage, including: the dynamic estimation time of the AGV and the bridge crane and the rail crane in the interaction stage, the AGV driving The dynamic estimation time of the stage, the dynamic estimation time of the bridge crane operation stage, and the dynamic estimation time of the rail crane operation stage;
  • the static estimated time and the dynamic estimated time of each operation stage calculate the final estimated time of the container in each operation stage and the total estimated time of operation, and the total estimated time of operation is the sum of the final estimated time of each operation stage;
  • the AGV job scheduling instruction is generated according to the final estimated time of the container in each operation stage and the total estimated time of the operation.
  • the static estimation values for estimating the operating efficiency of the bridge crane and the rail crane respectively include:
  • the estimation methods for the static estimation time of the AGV driving phase include:
  • the static estimated time of the AGV in the interaction phase with the bridge crane and the rail crane, respectively is estimated.
  • weighted average method is used to calculate the final estimated time and the total estimated time of the container in each operation stage respectively according to the static estimated time and the dynamic estimated time of each operation stage.
  • the dynamic estimation step also includes the step of training the efficiency prediction model, including:
  • the efficiency prediction model is a multiple linear regression equation, and the efficiency prediction model is:
  • k is the position number of the container, which is a positive integer
  • f(x) is the output of the efficiency prediction model.
  • Dynamic estimation time w is the parameter matrix of x, b is a constant parameter;
  • the method for determining w and b in the efficiency prediction model is:
  • y i represents the actual operation time of the container at position i in the current operation stage
  • the dispatching system input the dispatching system to generate the dynamic estimated time of the bridge crane operation phase and the dynamic estimate of the rail crane operation phase, respectively. time.
  • the step of training the AGV travel time prediction model is also included.
  • the AGV travel time prediction model includes:
  • the AGV travel time prediction model is trained according to the historical data and location.
  • the AGV travel time prediction model is a polynomial regression equation, and the polynomial regression equation is:
  • k is the position number of the container, which is a positive integer
  • g(x) is the dynamic estimated value of the driving stage output by the AGV driving time prediction model
  • w j is the parameter of x j
  • b2 is a constant parameter.
  • the AGV job scheduling instruction before generating the AGV job scheduling instruction, it also includes determining the job priority of the container, and generating the AGV job scheduling instruction according to the job priority;
  • AGV job scheduling instructions are generated by using mixed integer programming, combinatorial optimization methods and collaborative filtering algorithms.
  • the AGV scheduling method based on the time estimation model of the present invention adopts a combination of static estimation and dynamic estimation to estimate the time used by the container in each operation stage, and combines The estimated time is more reasonable for AGV scheduling, reducing AGV waiting time, shortening AGV no-load running time and distance, and improving AGV box transportation efficiency.
  • FIG. 1 is a flowchart of an embodiment of an AGV scheduling method based on a time prediction model proposed by the present invention.
  • the process of loading and unloading goods at the container terminal is as follows: first, the container is transferred from the freighter to the AGV by the bridge crane, and then the AGV transports the container to the yard, and then the rail crane stacks the container to the designated position in the yard, limited by the number of AGVs, if Unreasonable scheduling arrangement will lead to long AGV waiting time and low transportation efficiency.
  • this embodiment proposes an AGV scheduling method based on a time estimation model, as shown in Figure 1, including the following steps:
  • the static estimation step is to estimate the static estimation value of the operation efficiency of the bridge crane and the rail crane; respectively estimate the static estimation time of the container in each operation stage, including: the static estimation time of the AGV and the bridge crane and the rail crane in the interaction stage, the AGV driving The static estimation time of the stage, the static estimation time of the bridge crane operation stage, and the static estimation time of the rail crane operation stage;
  • the dynamic estimation step is to estimate the dynamic estimation value of the operation efficiency of the bridge crane and the rail crane; respectively estimate the dynamic estimation time of the container in each operation stage, including: the dynamic estimation time of the AGV and the bridge crane and the rail crane in the interaction stage, the AGV driving The dynamic estimation time of the stage, the dynamic estimation time of the bridge crane operation stage, and the dynamic estimation time of the rail crane operation stage;
  • the static estimated time and the dynamic estimated time of each operation stage calculate the final estimated time of the container in each operation stage and the total estimated time of operation, and the total estimated time of operation is the sum of the final estimated time of each operation stage;
  • the AGV job scheduling instruction is generated according to the final estimated time of the container in each operation stage and the total estimated time of the operation.
  • This AGV scheduling method based on the time estimation model uses a combination of static estimation and dynamic estimation to estimate the time used by the container in each operation stage, and combines the estimated time to schedule the AGV more reasonably, reducing the waiting time of the AGV and shortening the The no-load running time and distance of AGV are improved, and the efficiency of AGV container transportation is improved.
  • the combination of static setting and dynamic adjustment is used to estimate the time of each operation stage. Not only can manual adjustment be made according to the actual production operation situation, but also the dynamic adjustment method can be used to improve and perfect the sequence of AGV scheduling, so as to improve the overall dock level. work efficiency.
  • the static estimation values for respectively estimating the working efficiency of the bridge crane and the rail crane include:
  • the estimation of the operation efficiency of the bridge crane and the rail crane is mainly based on two variables: 1. The current actual operation situation. 2. Historical data.
  • the central controller combines the two variables to make a static estimate of operational efficiency. Estimates can be performed using a weighted average method.
  • the estimation methods for the static estimation time of the AGV driving phase include:
  • the static estimated time of the bridge crane operation stage and the static estimated time of the rail crane operation stage are estimated; the AGV travel time estimation is based on two variables: 1. The current actual operation situation . 2. A fixed time matrix, which is based on historical data and makes an AGV travel time estimate by combining two variables.
  • the static estimated time of the AGV in the interaction phase with the bridge crane and the rail crane, respectively is estimated.
  • the weighted average method is preferably used to calculate the final estimated time and the total estimated time of the container operation in each operation stage respectively according to the static estimated time and the dynamic estimated time of each operation stage.
  • algorithms related to big data processing are used to dynamically estimate the operating efficiency of bridge cranes and rail cranes, the estimated AGV travel time, and the estimated interaction time between AGVs and bridge cranes and rail cranes.
  • the dynamic estimation step also includes the step of training the efficiency prediction model, including:
  • the efficiency prediction model is a multiple linear regression equation, and the efficiency prediction model is:
  • k is the position number of the container, which is a positive integer
  • f(x) is the output of the efficiency prediction model.
  • Dynamic estimation time w is the parameter matrix of x, b is a constant parameter;
  • the dynamic estimation of the operating efficiency of bridge cranes and rail cranes uses the multiple linear regression method to calculate the historical time-consuming of each box during the suspension bridge operation based on historical data, which is the dependent variable; combined with the historical position data corresponding to each box, it is the independent variable.
  • the independent variable has a linear relationship with the dependent variable under the condition that the efficiency of the remote operator of the suspension bridge is relatively stable.
  • the determination method of w and b in the efficiency prediction model is:
  • y i represents the actual operation time of the container at position i in the current operation stage
  • the dynamic estimation time of the bridge crane operation stage and the dynamic estimation time of the rail crane operation stage are respectively generated.
  • the step of training the AGV travel time prediction model is also included.
  • AGV travel time prediction models include:
  • the AGV travel time prediction model is trained according to the historical data and location.
  • the AGV travel time prediction model is a polynomial regression equation, and the polynomial regression equation is:
  • k is the position number of the container, which is a positive integer
  • g(x) is the dynamic estimated value of the driving stage output by the AGV driving time prediction model
  • w j is the parameter of x j
  • b2 is a constant parameter.
  • the polynomial regression method is used to estimate the travel time of the AGV.
  • the premise is that the travel distance between the two points of the AGV is determined, and the information on whether the AGV needs to turn around has been given.
  • the training goal is to find a set of parameters minimize losses.
  • the time data in the historical samples can be directly calculated, denoted as y, and the starting point, end point and U-turn information are taken as x, and the parameters are updated by minimizing the direction of the loss function.
  • Estimated interaction time between AGV and rail crane According to the ASC operation efficiency in S21 and the AGV travel time estimation in S22, combined with the historical interaction time and the distance of the interaction area and other data information, the Gaussian process regression model is used to estimate the time.
  • the AGV job scheduling instruction Before generating the AGV job scheduling instruction, it also includes determining the job priority of the container, and generating the AGV job scheduling instruction according to the job priority;
  • the dynamic adjustment method of job priority not only takes into account short jobs, but also overcomes the hunger state of jobs, so that long jobs can also be processed in time.

Landscapes

  • Engineering & Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Manufacturing & Machinery (AREA)
  • Quality & Reliability (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Ship Loading And Unloading (AREA)
  • Control And Safety Of Cranes (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种基于时间预估模型的AGV调度方法,包括以下步骤:静态估算步骤,分别估算桥吊和轨道吊作业效率的静态估算值;分别估算集装箱在各作业阶段的静态估算时间;动态估算步骤,分别估算桥吊和轨道吊作业效率的动态估算值;分别估算集装箱在各作业阶段的动态估算时间;根据各作业阶段的静态估算时间和动态估算时间,计算集装箱分别在各作业阶段的最终估算时间以及总作业估算时间;根据集装箱分别在各作业阶段的最终估算时间以及总作业估算时间生成AGV作业调度指令。本发明的基于时间预估模型的AGV调度方法,结合所估算的时间对AGV的调度更加合理,减少了AGV等待时间,缩短了AGV空载运行时间和距离,提高了AGV运箱效率。

Description

基于时间预估模型的AGV调度方法 技术领域
本发明属于自动化码头运输技术领域,具体地说,涉及一种基于时间预估模型的AGV调度方法。
背景技术
集装箱码头前方作业带是指堆场前边线至码头前沿线之间的区域,其功能是服务于码头岸桥装卸船作业以及集装箱进出堆场作业,在人工码头和半自动化码头通常使用人工驾驶的内集卡。近年来,随着人工成本的不断攀升,越来越多的码头将水平运输设备由内集卡转为了AGV(Automated Guided Vehicle,自动引导车),AGV的使用大大提高了生产效率,节约了成本。
在全自动化集装箱码头的堆场海侧,通过桥吊实现货船与自动引导车的交互,通过轨道吊实现自动引导车与堆场的交互,通过AGV,L-AGV,Auto Shuttle等实现码头与堆场间的自动化作业交接。
现阶段,AGV合理调度、调度系统的稳定性、最大化AGV的利用率是自动化码头生产效率提升的关键。AGV的作业状态也是影响码头作业中重要的一环,通过对AGV作业状态的实时监控可以让操作人员更快发现生产过程中的问题,从而对AGV的调度算法进行改进和完善,提高整个自动化码头的工作效率。
目前一些自动化码头的运输设备调度只能通过减少运输设备数量、降低运输设备速度来保证整个调度系统运行的稳定性,但这样使得整个码头的作业效率很低,现阶段,AGV的作业监控还没有先例可寻,基于此,如何发明一种能够提高AGV运箱效率的方法,是本发明主要解决的技术问题。
发明内容
本发明针对现有技术中AGV运箱效率低的技术问题,提出了一种基于时间 预估模型的AGV调度方法,可以解决上述问题。
为实现上述发明目的,本发明采用下述技术方案予以实现:
一种基于时间预估模型的AGV调度方法,包括以下步骤:
静态估算步骤,分别估算桥吊和轨道吊作业效率的静态估算值;分别估算集装箱在各作业阶段的静态估算时间,包括:AGV分别与桥吊和轨道吊作业交互阶段的静态估算时间、AGV行驶阶段的静态估算时间、桥吊作业阶段的静态估算时间、轨道吊作业阶段的静态估算时间;
动态估算步骤,分别估算桥吊和轨道吊作业效率的动态估算值;分别估算集装箱在各作业阶段的动态估算时间,包括:AGV分别与桥吊和轨道吊作业交互阶段的动态估算时间、AGV行驶阶段的动态估算时间、桥吊作业阶段的动态估算时间、轨道吊作业阶段的动态估算时间;
根据各作业阶段的静态估算时间和动态估算时间,计算集装箱分别在各作业阶段的最终估算时间以及总作业估算时间,所述总作业估算时间为各作业阶段的最终估算时间之和;
根据集装箱分别在各作业阶段的最终估算时间以及总作业估算时间生成AGV作业调度指令。
进一步的,静态估算步骤中,分别估算桥吊和轨道吊作业效率的静态估算值包括:
获取作业效率的历史数据,分别估算桥吊和轨道吊作业效率的静态估算值;
AGV行驶阶段的静态估算时间的估算方法包括:
获取AGV行驶阶段的历史时间,估算AGV行驶阶段的静态估算时间;
根据桥吊和轨道吊作业效率的静态估算值,估算桥吊作业阶段的静态估算时间和轨道吊作业阶段的静态估算时间;
根据桥吊和轨道吊作业效率的静态估算值和AGV行驶阶段的静态估算时间,估算AGV分别与桥吊和轨道吊作业交互阶段的静态估算时间。
进一步的,采用加权平均法根据各作业阶段的静态估算时间和动态估算时 间,计算集装箱分别在各作业阶段的最终估算时间以及总作业估算时间。
进一步的,动态估算步骤中,还包括训练效率预测模型的步骤,包括:
获取各集装箱在桥吊作业阶段的历史时间,以及所对应的集装箱的位置;
根据所述历史时间和位置训练作业效率预测模型,所述效率预测模型为多元线性回归方程,所述效率预测模型为:
f(x)=w Tx+b
其中,x=[x 1,x 2,…,x k] T为各集装箱的在当前作业阶的历史时间,k为集装箱的位置编号,为正整数,f(x)为效率预测模型输出的动态估算时间,w为x的参数矩阵,b为常参数;
根据w和b估算桥吊和轨道吊作业效率的动态估算值。
进一步的,所述效率预测模型中w和b的确定方法为:
构建数组D={(x 1,y 1),…(x k,y k)};
计算时间损耗
Figure PCTCN2021138631-appb-000001
其中,y i表示位置为i的集装箱在当前作业阶段的实际作业时间;
确定w和b,使得J最小。
进一步的,根据所述效率预测模型输出的桥吊和轨道吊作业效率的动态估算值以及需要运输的集装箱数量输入调度系统,分别生成桥吊作业阶段的动态估算时间和轨道吊作业阶段的动态估算时间。
进一步的,动态估算步骤中,还包括训练AGV行驶时间预测模型的步骤。
进一步的,所述AGV行驶时间预测模型包括:
根据所述历史数据和位置训练AGV行驶时间预测模型,所述AGV行驶时间预测模型为多项式回归方程,所述多项式回归方程为:
Figure PCTCN2021138631-appb-000002
其中,x=[x 1,x 2,…,x k] T为AGV在所述桥吊和轨道吊之间的行驶时间 历史数据,k为集装箱的位置编号,为正整数,g(x)为AGV行驶时间预测模型输出的行驶阶段的动态估算值,w j为x j的参数,b2为常参数。
进一步的,生成AGV作业调度指令之前,还包括确定集装箱的作业优先级,并根据所述作业优先级生成AGV作业调度指令;
所述作业优先级的设定基于时间响应比:
Figure PCTCN2021138631-appb-000003
时间响应比越大,优先级越低。
进一步的,利用混合整数规划、组合优化方法和协同过滤的算法生成AGV作业调度指令。
与现有技术相比,本发明的优点和积极效果是:本发明的基于时间预估模型的AGV调度方法,采用静态估算和动态估算相结合的方法估算集装箱分别在各作业阶段所用时间,结合所估算的时间对AGV的调度更加合理,减少了AGV等待时间,缩短了AGV空载运行时间和距离,提高了AGV运箱效率。
结合附图阅读本发明的具体实施方式后,本发明的其他特点和优点将变得更加清楚。
附图说明
为了更清楚地说明本发明实施例中的技术方案,下面将对实施例中所需要使用的附图作一简单地介绍,显而易见地,下面描述中的附图是本发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1是本发明提出的基于时间预估模型的AGV调度方法的一种实施例流程图。
具体实施方式
为了使本发明的目的、技术方案及优点更加清楚明白,以下将结合附图和实施例,对本发明作进一步详细说明。
需要说明的是,在本发明的描述中,术语“上”、“下”、“左”、“右”、“竖”、“横”、“内”、“外”等指示的方向或位置关系的术语是基于附图所示的方向或位置关系,这仅仅是为了便于描述,而不是指示或暗示所述装置或元件必须具有特定的方位、以特定的方位构造和操作,因此不能理解为对本发明的限制。此外,术语“第一”、“第二”仅用于描述目的,而不能理解为指示或暗示相对重要性。
实施例一
集装箱码头装卸货物的流程是:首先通过桥吊将集装箱从货轮转移至AGV,然后AGV将集装箱输送至堆场,再由轨道吊将集装箱码放至堆场的指定位置,受AGV的数量限制,如果调度安排不合理,将会导致AGV等待时间长,运输效率低,为了解决上述问题,本实施例提出了一种基于时间预估模型的AGV调度方法,如图1所示,包括以下步骤:
静态估算步骤,分别估算桥吊和轨道吊作业效率的静态估算值;分别估算集装箱在各作业阶段的静态估算时间,包括:AGV分别与桥吊和轨道吊作业交互阶段的静态估算时间、AGV行驶阶段的静态估算时间、桥吊作业阶段的静态估算时间、轨道吊作业阶段的静态估算时间;
动态估算步骤,分别估算桥吊和轨道吊作业效率的动态估算值;分别估算集装箱在各作业阶段的动态估算时间,包括:AGV分别与桥吊和轨道吊作业交互阶段的动态估算时间、AGV行驶阶段的动态估算时间、桥吊作业阶段的动态估算时间、轨道吊作业阶段的动态估算时间;
根据各作业阶段的静态估算时间和动态估算时间,计算集装箱分别在各作业阶段的最终估算时间以及总作业估算时间,所述总作业估算时间为各作业阶段的最终估算时间之和;
根据集装箱分别在各作业阶段的最终估算时间以及总作业估算时间生成AGV作业调度指令。
本基于时间预估模型的AGV调度方法,采用静态估算和动态估算相结合的 方法估算集装箱分别在各作业阶段所用时间,结合所估算的时间对AGV的调度更加合理,减少了AGV等待时间,缩短了AGV空载运行时间和距离,提高了AGV运箱效率。
采用静态设置和动态调节相结合的方法估算各作业阶段的时间,不仅可以根据实际生产作业情况进行手动调节,同时也可以运用动态调节的方法对AGV调度的先后顺序进行改进和完善,提高码头整体作业效率。
作为一个优选的实施例,静态估算步骤中,分别估算桥吊和轨道吊作业效率的静态估算值包括:
获取作业效率的历史数据,分别估算桥吊和轨道吊作业效率的静态估算值,桥吊和轨道吊作业效率的估算主要依据变量有两方面:1、当前实际作业情况。2、历史数据。中控人员将两个变量相结合,做出作业效率静态估算。可采用加权平均法施行估算。
AGV行驶阶段的静态估算时间的估算方法包括:
获取AGV行驶阶段的历史时间,估算AGV行驶阶段的静态估算时间;
根据桥吊和轨道吊作业效率的静态估算值,估算桥吊作业阶段的静态估算时间和轨道吊作业阶段的静态估算时间;AGV行驶时间估算,其依据变量有两个:1、当前实际作业情况。2、固定时间矩阵,该矩阵以历史数据为基础,通过将两个变量相结合,做出AGV行驶时间预估。
根据桥吊和轨道吊作业效率的静态估算值和AGV行驶阶段的静态估算时间,估算AGV分别与桥吊和轨道吊作业交互阶段的静态估算时间。
本实施例中优选采用加权平均法根据各作业阶段的静态估算时间和动态估算时间,计算集装箱分别在各作业阶段的最终估算时间以及总作业估算时间。
本步骤中运用大数据处理相关算法,动态预估桥吊、轨道吊作业效率,估算的AGV行驶时间,估算的AGV与桥吊和轨道吊的交互时间。
具体的,动态估算步骤中,还包括训练效率预测模型的步骤,包括:
获取各集装箱在桥吊作业阶段的历史时间,以及所对应的集装箱的位置;
根据所述历史时间和位置训练作业效率预测模型,所述效率预测模型为多元线性回归方程,所述效率预测模型为:
f(x)=w Tx+b
其中,x=[x 1,x 2,…,x k] T为各集装箱的在当前作业阶的历史时间,k为集装箱的位置编号,为正整数,f(x)为效率预测模型输出的动态估算时间,w为x的参数矩阵,b为常参数;
根据w和b估算桥吊和轨道吊作业效率的动态估算值。
桥吊和轨道吊作业效率的动态估算通过使用多元线性回归方法,根据历史数据计算获得进行吊桥作业时每个箱子的历史耗时,为因变量;结合每个箱子对应的历史位置数据,为自变量,在吊桥远程操作司机效率相对稳定的情况下,自变量与因变量具有线性关系。采用多元线性回归模型在通过历史数据样本计算出参数后,可以根据当前箱子的位置信息对预估作业耗时进行预测。
效率预测模型中w和b的确定方法为:
构建数组D={(x 1,y 1),…(x k,y k)};
计算时间损耗
Figure PCTCN2021138631-appb-000004
其中,y i表示位置为i的集装箱在当前作业阶段的实际作业时间;
确定w和b,使得J最小。
根据效率预测模型输出的桥吊和轨道吊作业效率的动态估算值以及需要运输的集装箱数量输入调度系统,分别生成桥吊作业阶段的动态估算时间和轨道吊作业阶段的动态估算时间。
动态估算步骤中,还包括训练AGV行驶时间预测模型的步骤。
AGV行驶时间预测模型包括:
根据所述历史数据和位置训练AGV行驶时间预测模型,所述AGV行驶时间预测模型为多项式回归方程,所述多项式回归方程为:
Figure PCTCN2021138631-appb-000005
其中,x=[x 1,x 2,…,x k] T为AGV在所述桥吊和轨道吊之间的行驶时间历史数据,k为集装箱的位置编号,为正整数,g(x)为AGV行驶时间预测模型输出的行驶阶段的动态估算值,w j为x j的参数,b2为常参数。
AGV行驶时间估算采用多项式回归的方法,其前提是确定的AGV两点间行驶路程,且AGV是否需要掉头的信息已给出。
(w j,b)(j=1,…,n)为待估参数。
如果用z j替代上式中的x j,相当于获得一个关于z=(z 1,z 2,…,z n)的线性函数h(z),用线性函数h(z)去拟合数据并使得损失
Figure PCTCN2021138631-appb-000006
最小。训练目标就是找到一组参数
Figure PCTCN2021138631-appb-000007
使损失最小化。
此外,还可以通过添加L1正则化项来去除其中不重要的非线性关系:
Figure PCTCN2021138631-appb-000008
在该算法中,历史样本中的时间数据可以直接计算获得,记为y,将出发点以及终点和调头信息作为x,通过最小化损失函数的方向更新参数。
AGV与桥吊交互时间预估:根据桥吊作业效率和AGV行驶时间估算,同时结合历史交互时间以及交互区距离等数据信息,采用高斯过程回归模型进行时间预估。
AGV与轨道吊交互时间预估:根据S21中ASC作业效率和S22中AGV行驶时间估算,同时结合历史交互时间以及交互区距离等数据信息,采用高斯过程回归模型进行时间预估。
在使用训练集和测试集进行测试后,将历史数据导入回归模型中,得到一个高斯回归曲线,进而用来预估AGV与轨道吊的交互时间。
在使用训练集和测试集进行测试后,将历史数据导入回归模型中,得到一 个高斯回归曲线,进而用来预估AGV与桥吊的交互时间。
生成AGV作业调度指令之前,还包括确定集装箱的作业优先级,并根据所述作业优先级生成AGV作业调度指令;
作业优先级的设定基于时间响应比:
Figure PCTCN2021138631-appb-000009
时间响应比越大,优先级越低。
利用混合整数规划、组合优化方法和协同过滤的算法生成AGV作业调度指令。
该作业优先级动态调整方法既兼顾了短作业同时也客服了作业的饥饿状态,使得长作业也能得到及时处理。
以上实施例仅用以说明本发明的技术方案,而非对其进行限制;尽管参照前述实施例对本发明进行了详细的说明,对于本领域的普通技术人员来说,依然可以对前述实施例所记载的技术方案进行修改,或者对其中部分技术特征进行等同替换;而这些修改或替换,并不使相应技术方案的本质脱离本发明所要求保护的技术方案的精神和范围。

Claims (10)

  1. 一种基于时间预估模型的AGV调度方法,其特征在于,包括以下步骤:
    静态估算步骤,分别估算桥吊和轨道吊作业效率的静态估算值;分别估算集装箱在各作业阶段的静态估算时间,包括:AGV分别与桥吊和轨道吊作业交互阶段的静态估算时间、AGV行驶阶段的静态估算时间、桥吊作业阶段的静态估算时间、轨道吊作业阶段的静态估算时间;
    动态估算步骤,分别估算桥吊和轨道吊作业效率的动态估算值;分别估算集装箱在各作业阶段的动态估算时间,包括:AGV分别与桥吊和轨道吊作业交互阶段的动态估算时间、AGV行驶阶段的动态估算时间、桥吊作业阶段的动态估算时间、轨道吊作业阶段的动态估算时间;
    根据各作业阶段的静态估算时间和动态估算时间,计算集装箱分别在各作业阶段的最终估算时间以及总作业估算时间,所述总作业估算时间为各作业阶段的最终估算时间之和;
    根据集装箱分别在各作业阶段的最终估算时间以及总作业估算时间生成AGV作业调度指令。
  2. 根据权利要求1所述的基于时间预估模型的AGV调度方法,其特征在于,静态估算步骤中,分别估算桥吊和轨道吊作业效率的静态估算值包括:
    获取作业效率的历史数据,分别估算桥吊和轨道吊作业效率的静态估算值;
    AGV行驶阶段的静态估算时间的估算方法包括:
    获取AGV行驶阶段的历史时间,估算AGV行驶阶段的静态估算时间;
    根据桥吊和轨道吊作业效率的静态估算值,估算桥吊作业阶段的静态估算时间和轨道吊作业阶段的静态估算时间;
    根据桥吊和轨道吊作业效率的静态估算值和AGV行驶阶段的静态估算时间,估算AGV分别与桥吊和轨道吊作业交互阶段的静态估算时间。
  3. 根据权利要求2所述的基于时间预估模型的AGV调度方法,其特征在于,采用加权平均法根据各作业阶段的静态估算时间和动态估算时间,计算集装箱 分别在各作业阶段的最终估算时间以及总作业估算时间。
  4. 根据权利要求1所述的基于时间预估模型的AGV调度方法,其特征在于,动态估算步骤中,还包括训练效率预测模型的步骤,包括:
    获取各集装箱在桥吊作业阶段的历史时间,以及所对应的集装箱的位置;
    根据所述历史时间和位置训练作业效率预测模型,所述效率预测模型为多元线性回归方程,所述效率预测模型为:
    f(x)=w Tx+b
    其中,x=[x 1,x 2,…,x k] T为各集装箱的在当前作业阶的历史时间,k为集装箱的位置编号,为正整数,f(x)为效率预测模型输出的动态估算时间,w为x的参数矩阵,b为常参数;
    根据w和b估算桥吊和轨道吊作业效率的动态估算值。
  5. 根据权利要求4所述的基于时间预估模型的AGV调度方法,其特征在于,所述效率预测模型中w和b的确定方法为:
    构建数组D={(x 1,y 1),…(x k,y k)};
    计算时间损耗
    Figure PCTCN2021138631-appb-100001
    其中,y i表示位置为i的集装箱在当前作业阶段的实际作业时间;
    确定w和b,使得J最小。
  6. 根据权利要求5所述的基于时间预估模型的AGV调度方法,其特征在于,根据所述效率预测模型输出的桥吊和轨道吊作业效率的动态估算值以及需要运输的集装箱数量输入调度系统,分别生成桥吊作业阶段的动态估算时间和轨道吊作业阶段的动态估算时间。
  7. 根据权利要求5所述的基于时间预估模型的AGV调度方法,其特征在于,动态估算步骤中,还包括训练AGV行驶时间预测模型的步骤。
  8. 根据权利要求7所述的基于时间预估模型的AGV调度方法,其特征在于,所述AGV行驶时间预测模型包括:
    根据所述历史数据和位置训练AGV行驶时间预测模型,所述AGV行驶时间预测模型为多项式回归方程,所述多项式回归方程为:
    Figure PCTCN2021138631-appb-100002
    其中,x=[x 1,x 2,…,x k] T为AGV在所述桥吊和轨道吊之间的行驶时间历史数据,k为集装箱的位置编号,为正整数,g(x)为AGV行驶时间预测模型输出的行驶阶段的动态估算值,w j为x j的参数,b2为常参数。
  9. 根据权利要求1-6任一项所述的基于时间预估模型的AGV调度方法,其特征在于,生成AGV作业调度指令之前,还包括确定集装箱的作业优先级,并根据所述作业优先级生成AGV作业调度指令;
    所述作业优先级的设定基于时间响应比:
    Figure PCTCN2021138631-appb-100003
    时间响应比越大,优先级越低。
  10. 根据权利要求9所述的基于时间预估模型的AGV调度方法,其特征在于,利用混合整数规划、组合优化方法和协同过滤的算法生成AGV作业调度指令。
PCT/CN2021/138631 2021-01-25 2021-12-16 基于时间预估模型的agv调度方法 WO2022156440A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2023543308A JP2024503141A (ja) 2021-01-25 2021-12-16 時間推定モデルに基づくagvスケジューリング方法

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110098290.8 2021-01-25
CN202110098290.8A CN112764405B (zh) 2021-01-25 2021-01-25 基于时间预估模型的agv调度方法

Publications (1)

Publication Number Publication Date
WO2022156440A1 true WO2022156440A1 (zh) 2022-07-28

Family

ID=75707173

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/138631 WO2022156440A1 (zh) 2021-01-25 2021-12-16 基于时间预估模型的agv调度方法

Country Status (3)

Country Link
JP (1) JP2024503141A (zh)
CN (1) CN112764405B (zh)
WO (1) WO2022156440A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115049324A (zh) * 2022-08-17 2022-09-13 上海国际港务(集团)股份有限公司 码头agv车调度方法、装置、计算机设备和存储介质
CN115293443A (zh) * 2022-08-17 2022-11-04 上海国际港务(集团)股份有限公司 桥吊和集装箱船舶装卸作业时间的预测方法、系统及介质
CN115685927A (zh) * 2022-10-29 2023-02-03 福州大学 基于进化优化算法的氨纶盘头作业动态调度方法
CN116300775A (zh) * 2023-05-23 2023-06-23 山东曙岳车辆有限公司 一种智能物流agv小车控制系统

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112764405B (zh) * 2021-01-25 2021-11-30 青岛港国际股份有限公司 基于时间预估模型的agv调度方法

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105740979A (zh) * 2016-01-29 2016-07-06 上海海事大学 自动化码头多自动导引车的智能调度系统与方法
CN106485434A (zh) * 2015-08-25 2017-03-08 上海玑智自动化科技有限公司 自动搬运装置的调度方法及系统
CN108845496A (zh) * 2018-05-23 2018-11-20 大连海事大学 一种自动化码头agv调度方法及系统
CN109213163A (zh) * 2018-09-06 2019-01-15 青岛港国际股份有限公司 自动化码头qctp作业中agv路径优化方法和系统
CN109685435A (zh) * 2018-11-07 2019-04-26 上海图森未来人工智能科技有限公司 船舶卸船、装船控制系统及相关系统和装置
CN110516976A (zh) * 2019-08-30 2019-11-29 大连海事大学 一种可选择装卸完工时间的自动化集装箱码头agv调度方法
CN110942203A (zh) * 2019-12-03 2020-03-31 青岛港国际股份有限公司 一种自动化集装箱码头agv路径优化方法
CN111401617A (zh) * 2020-03-09 2020-07-10 兰剑智能科技股份有限公司 一种基于时间预测的agv调度方法及系统
WO2020210986A1 (zh) * 2019-04-16 2020-10-22 西门子股份公司 应用于agv的导航方法、系统、计算设备、介质以及产品
CN112764405A (zh) * 2021-01-25 2021-05-07 青岛港国际股份有限公司 基于时间预估模型的agv调度方法

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106485434A (zh) * 2015-08-25 2017-03-08 上海玑智自动化科技有限公司 自动搬运装置的调度方法及系统
CN105740979A (zh) * 2016-01-29 2016-07-06 上海海事大学 自动化码头多自动导引车的智能调度系统与方法
CN108845496A (zh) * 2018-05-23 2018-11-20 大连海事大学 一种自动化码头agv调度方法及系统
CN109213163A (zh) * 2018-09-06 2019-01-15 青岛港国际股份有限公司 自动化码头qctp作业中agv路径优化方法和系统
CN109685435A (zh) * 2018-11-07 2019-04-26 上海图森未来人工智能科技有限公司 船舶卸船、装船控制系统及相关系统和装置
WO2020210986A1 (zh) * 2019-04-16 2020-10-22 西门子股份公司 应用于agv的导航方法、系统、计算设备、介质以及产品
CN110516976A (zh) * 2019-08-30 2019-11-29 大连海事大学 一种可选择装卸完工时间的自动化集装箱码头agv调度方法
CN110942203A (zh) * 2019-12-03 2020-03-31 青岛港国际股份有限公司 一种自动化集装箱码头agv路径优化方法
CN111401617A (zh) * 2020-03-09 2020-07-10 兰剑智能科技股份有限公司 一种基于时间预测的agv调度方法及系统
CN112764405A (zh) * 2021-01-25 2021-05-07 青岛港国际股份有限公司 基于时间预估模型的agv调度方法

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115049324A (zh) * 2022-08-17 2022-09-13 上海国际港务(集团)股份有限公司 码头agv车调度方法、装置、计算机设备和存储介质
CN115049324B (zh) * 2022-08-17 2022-11-04 上海国际港务(集团)股份有限公司 码头agv车调度方法、装置、计算机设备和存储介质
CN115293443A (zh) * 2022-08-17 2022-11-04 上海国际港务(集团)股份有限公司 桥吊和集装箱船舶装卸作业时间的预测方法、系统及介质
CN115293443B (zh) * 2022-08-17 2024-03-22 上海国际港务(集团)股份有限公司 桥吊和集装箱船舶装卸作业时间的预测方法、系统及介质
CN115685927A (zh) * 2022-10-29 2023-02-03 福州大学 基于进化优化算法的氨纶盘头作业动态调度方法
CN116300775A (zh) * 2023-05-23 2023-06-23 山东曙岳车辆有限公司 一种智能物流agv小车控制系统
CN116300775B (zh) * 2023-05-23 2023-07-21 山东曙岳车辆有限公司 一种智能物流agv小车控制系统

Also Published As

Publication number Publication date
JP2024503141A (ja) 2024-01-24
CN112764405B (zh) 2021-11-30
CN112764405A (zh) 2021-05-07

Similar Documents

Publication Publication Date Title
WO2022156440A1 (zh) 基于时间预估模型的agv调度方法
CN112070296B (zh) 一种基于模糊推理的集卡调度方法
CN110363380B (zh) 一种集装箱堆场双场桥动态协同调度方法
CN111639790B (zh) 保证货运可靠性和安全性的多式联运调度方法和系统
Zhang et al. Cooperative scheduling of AGV and ASC in automation container terminal relay operation mode
CN105523476A (zh) 一种桥式起重机防摇控制方法和系统
CN110942203A (zh) 一种自动化集装箱码头agv路径优化方法
CN107451720B (zh) 面向自动化集装箱码头的船次作业计划评价方法
CN107871179A (zh) 一种基于运到时限的铁路货物列车运行图编制方法
WO2024032376A1 (zh) 基于混合遗传算法的车辆路径优化方法及应用
CN116935616A (zh) 面向港区车辆调度-规控耦合的智能集卡协同调度方法
Lu The optimization of automated container terminal scheduling based on proportional fair priority
CN114408613B (zh) 一种适配装船模式的智能化码头art动态调序方法
Chen et al. Research on integrated scheduling optimization of double-trolley quay crane and AGV in automated terminal
CN113743739B (zh) 一种基于混合整数规划、组合优化算法的agv调度方法
Lu et al. Multi-Depot Vehicle Scheduling Optimization for Port Container Drop and Pull Transport
JP6654040B2 (ja) 作業指示システム
Prayogo Carbon emission modelling in container terminal operations planning using a system dynamics approach
Zakrzewski et al. Changes in logistics processes caused by the implementation of automation in transport
CN116502976B (zh) 一种多式联运枢纽工艺布局方法及系统
Rong et al. A Real Time Vehicle Scheduling Strategy in Container Terminals
Homayouni et al. Energy-Efficient Scheduling of Intraterminal Container Transport
Li A heuristic algorithm for equipment scheduling at an automated container terminal with multi-size containers
CN117035338B (zh) 一种港口生产作业实时监控方法和系统
Zhang et al. Dynamic scheduling optimization of AGVs in automated container terminals under uncertainty

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21920814

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 2023543308

Country of ref document: JP

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 22/11/2024)