WO2021056605A1 - 一种基于复合智能体的车间实时调度方法及装置 - Google Patents

一种基于复合智能体的车间实时调度方法及装置 Download PDF

Info

Publication number
WO2021056605A1
WO2021056605A1 PCT/CN2019/109886 CN2019109886W WO2021056605A1 WO 2021056605 A1 WO2021056605 A1 WO 2021056605A1 CN 2019109886 W CN2019109886 W CN 2019109886W WO 2021056605 A1 WO2021056605 A1 WO 2021056605A1
Authority
WO
WIPO (PCT)
Prior art keywords
workpiece
agent
workshop
physical
rfid chip
Prior art date
Application number
PCT/CN2019/109886
Other languages
English (en)
French (fr)
Inventor
唐敦兵
潘俊峰
张泽群
王立平
聂庆玮
Original Assignee
南京航空航天大学
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 南京航空航天大学 filed Critical 南京航空航天大学
Publication of WO2021056605A1 publication Critical patent/WO2021056605A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B19/00—Programme-control systems
    • G05B19/02—Programme-control systems electric
    • G05B19/418—Total 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]
    • G05B19/41865—Total 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 job scheduling, process planning, material flow
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00—Program-control systems
    • G05B2219/30—Nc systems
    • G05B2219/32—Operator till task planning
    • G05B2219/32252—Scheduling production, machining, job shop
    • Y—GENERAL 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
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]

Definitions

  • the present invention relates to the technical field of workshop scheduling control, in particular to a real-time workshop scheduling method and device based on a compound agent.
  • the multi-agent manufacturing system has become a new type of intelligent manufacturing mode, which is favored by more and more people, and is regarded as an effective means to solve the current and future manufacturing difficulties.
  • multi-agent manufacturing technology has become a topic that needs to be focused on.
  • the patent application publication number CN106094759A is called a real-time control method for workshop scheduling in a complex production environment.
  • the workpiece agent and the equipment agent work together on the scheduling behavior of the production workshop.
  • the shortcomings of this design are: the workpiece agent lacks an effective computing carrier, and one or more computers are needed to provide computing power for the workpiece agent; when the number of shop orders continues to increase, the number of workpiece agents continues to increase, which will It will consume a lot of computing resources and equipment.
  • multiple artifact agents based on the same computing carrier are not conducive to system stability and maintenance operations.
  • the patent application publication number CN106527373A is a multi-agent based workshop autonomous scheduling system and method.
  • the workpiece agent group, equipment agent group and logistics agent work together to dispatch the production workshop.
  • the workpiece agent group includes a workpiece agent, a workpiece manager agent, and a job manager agent.
  • the device agent group includes a device agent, a device manager agent, and a device manager agent.
  • the shortcomings of this design the grouping of agents is based on a tree structure, which is controlled by the general manager and in charge of controlling a single agent. If an agent in charge fails, all agents under the node in charge are in a paralyzed state, if the agent in charge fails, all agents are in a paralyzed state. This has higher requirements on the stability and computing power of the computing carrier of the in-charge agent and the manager agent, and the system has higher requirements for the risk resistance of the central node.
  • the embodiment of the present invention provides a real-time scheduling method and device for a workshop based on a compound agent, which can reduce the requirements and quantity of computing carriers, and at the same time adopts a non-central node structure, which does not require a center with high stability and high computing power. node.
  • the method provided by the embodiment of the present invention includes:
  • the device agent When the device agent detects that the physical device is idle, it selects the workpiece in the workpiece buffer;
  • the generated workpiece agent updates the data in the RFID chip of the workpiece pallet where the selected workpiece is located;
  • the device provided by the embodiment of the present invention includes:
  • the device agent is used to select the workpiece in the workpiece buffer when the physical device is idle; read the data in the RFID chip of the workpiece tray where the selected workpiece is located, activate the workpiece agent and based on the data in the read RFID chip Generate a workpiece agent program corresponding to the selected workpiece;
  • the workpiece agent is used to complete the next process negotiation and update the data in the RFID chip of the workpiece pallet where the selected workpiece is located; after that, the workpiece agent program corresponding to the selected workpiece is destroyed, and the workpiece agent Switch to inactive state.
  • a composite agent is designed to realize autonomous scheduling of workshop production, which effectively reduces the number of agents in the workshop, reduces the requirements and quantity of computing carriers, and reduces the cost of workshop hardware.
  • this method can effectively control the number of concurrent workshops, reduce the cost of workshop network communication, and reduce the computing power requirements of computing equipment.
  • this method can reduce the number of workshop agents and reduce the requirements and number of computing carriers.
  • this method uses a non-central node structure and does not require high stability. And the central node of high computing power.
  • Fig. 1 is a schematic diagram of a physical device layout provided by an embodiment of the present invention.
  • Fig. 2 is a schematic diagram of the structure of a composite agent provided by an embodiment of the present invention.
  • Fig. 3 is a schematic diagram of the operation of a composite agent provided by an embodiment of the present invention.
  • Fig. 4 is a schematic flowchart of a method provided by an embodiment of the present invention.
  • FIG. 1 which includes physical equipment, a loading and unloading device, and a workpiece buffer table.
  • physical equipment can be understood as: lathes, milling machines, grinders, punching machines, planers, special processing equipment and other equipment commonly used in the industry for workpiece processing.
  • a production workshop usually has multiple physical Equipment, these physical equipment can be arranged around the loading and unloading device to facilitate the transfer of workpieces through the loading and unloading device.
  • the loading and unloading device is specifically used to put the workpiece from the workpiece buffer table into the physical equipment through the manipulator or conveyor installed on the loading and unloading device.
  • the physical device puts the workpiece from the workpiece buffer table into the physical device for processing through the loading and unloading device.
  • An RFID reader is installed on the workpiece buffer table, the workpiece transferred to the workpiece buffer table is placed in a workpiece tray, and an RFID chip is installed on the workpiece tray.
  • the workpiece buffer table is equipped with an RFID reader; the workpiece is placed in the workpiece tray; the workpiece tray is equipped with an RFID chip for the RFID reader to read and write the information; the workpiece tray RFID chip stores the workpiece index number and the workpiece geometry Dimensions, process status parameters, current status parameters, process flow information.
  • the composite agent runs on a computing carrier, which is connected to the workpiece buffer RFID reader and the physical device, and the composite agent includes a device agent and a workpiece agent. That is, the composite agent is deployed on every physical device in the workshop.
  • the workpiece agent described in this embodiment can refer to a pre-defined program according to the common understanding in the industry, which is started by the device agent, and the parameters required by the program are determined by the device.
  • the agent reads from the RFID chip of the workpiece pallet.
  • the device agent monitors the state of the physical device through the physical communication interface; and, the device agent controls the workpiece buffer RFID reader and writer through the physical communication interface, from the workpiece pallet Read the information data in the RFID chip.
  • the composite agent exists in the computing carrier (the computing carrier can be an embedded computing device, industrial computer, etc.), and the computing carrier is connected through a physical communication interface (Ethernet, serial port, CAN bus, USB, etc.) RFID reader to physical equipment and workpiece buffer.
  • the embodiment of the present invention provides a real-time scheduling method for a workshop based on a compound agent, as shown in FIG. 4, including:
  • the device agent When the device agent detects that the physical device is idle, it selects the workpiece in the workpiece buffer.
  • the generated workpiece agent updates the data in the RFID chip of the workpiece pallet where the selected workpiece is located.
  • the carrier of the device agent is a physical device in the workshop; the device agent matches the physical device where it is currently located, and drives the physical device where it is currently located to other agents in the workshop. Provide processing services.
  • the method further includes: selecting the next processed part from the workpiece buffer by the device agent according to the workshop environment and the task status.
  • the data carrier of the artifact agent is located on the artifact carrier; the state of the artifact agent includes: an inactive state and an activated state, wherein, in the inactive state, the artifact agent does not occupy computing resources; In the activated state, the carrier of the device agent of the artifact agent is activated, and computing resources are provided to the artifact agent.
  • the workpiece carrier is a container for holding the workpiece with an RFID chip.
  • the workpiece carrier is a workpiece tray for holding the workpiece, and the pallet is attached with an RFID chip for recording the data of the current workpiece state.
  • the device agent in the composite agent controls the physical device and monitors the state of the physical device through the physical communication interface, and the device agent controls the workpiece buffer RFID reader through the physical communication interface to read and modify the information data in the RFID of the workpiece tray.
  • the device agent selects a workpiece in the workpiece buffer according to the corresponding decision algorithm, reads the information data in the pallet RFID chip, and generates the corresponding workpiece agent program.
  • the workpiece agent enters Active state (program segment in operation).
  • the activated workpiece agent completes the next process processing negotiation, it updates the RFID chip information data in the workpiece tray through the physical communication bus, and then destroys its own running program and becomes inactive (static state data).
  • the generating a workpiece agent program corresponding to the selected workpiece according to the read data in the RFID chip includes:
  • a workpiece agent program is generated and transmitted to the physical device through the physical communication interface.
  • the internal equipment agent of the compound agent selects the corresponding workpiece from the workpiece buffer for processing according to the corresponding decision-making algorithm (machine learning, genetic algorithm, dynamic programming algorithm), and the geometric size and process state required for processing
  • the information and current status information are stored in the RFID chip of the workpiece pallet.
  • the device agent obtains this information through the physical communication interface, generates an NC program and transmits it to the physical device.
  • the completion of the next process processing negotiation includes:
  • the active workpiece agent initiates a processing negotiation request to the physical equipment in the workshop according to the current status information and process status information of the selected workpiece.
  • the physical equipment in the workshop is selected for the next process. For example: when the workpiece being processed completes the current process, the equipment agent generates the corresponding workpiece agent program according to the state information of the workpiece, that is, the workpiece agent turns into the active state.
  • the active workpiece agent initiates a processing negotiation request to other physical equipment in the workshop according to the current status information of the workpiece and the process flow information, and selects the appropriate physical equipment according to the corresponding decision-making algorithm (reinforcement learning, contract network mechanism, cellular immune algorithm, etc.) Carry out the next process.
  • the workpiece agent After completing the selection, the workpiece agent updates the information data in the RFID chip of the workpiece tray, and after initiating a movement request to the transportation equipment (AGV, conveyor belt), destroys its own running program, that is, the workpiece agent turns into an inactive state.
  • AGV transportation equipment
  • the embodiment of the present invention also provides a real-time scheduling device for a workshop based on a compound agent as shown in Figs. 2 and 3, wherein:
  • the device agent is used to select the workpiece in the workpiece buffer when the physical device is idle; read the data in the RFID chip of the workpiece tray where the selected workpiece is located, activate the workpiece agent and based on the data in the read RFID chip Generate a workpiece agent program corresponding to the selected workpiece.
  • the workpiece agent is used to complete the next process negotiation and update the data in the RFID chip of the workpiece pallet where the selected workpiece is located; after that, the workpiece agent program corresponding to the selected workpiece is destroyed, and the workpiece agent Switch to inactive state.
  • the physical equipment puts the workpiece from the workpiece buffer table into the physical device for processing through the loading and unloading device; an RFID reader is installed on the workpiece buffer table, and the workpiece transferred to the workpiece buffer table is placed on the workpiece buffer table.
  • an RFID chip is installed on the workpiece pallet; a composite agent runs on a computing carrier, the computing carrier is connected to the workpiece buffer RFID reader and the physical device, and the composite agent includes equipment Agent and artifact agent.
  • the carrier of the device agent is a physical device in the workshop; the device agent matches the physical device where it is currently located, and drives the physical device where it is currently located to provide processing services to other agents in the workshop ;
  • the equipment agent is also used to select the next processing part from the workpiece buffer according to the workshop environment and task status.
  • the data carrier of the artifact agent is located on the artifact carrier;
  • the state of the artifact agent includes: an inactive state and an activated state, wherein, in the inactive state, the artifact agent does not occupy computing Resource; when in the activated state, activate the carrier of the device agent of the artifact agent, and provide computing resources to the artifact agent.
  • the device agent is specifically used to extract geometric size information, process state information, and current state information of the selected workpiece from the data in the RFID chip; and generate workpiece intelligence based on the extracted information Body program and transmitted to the physical device through the physical communication interface.
  • the workpiece agent is specifically used to initiate a processing negotiation request to the physical equipment in the workshop according to the current status information and process status information of the selected workpiece when it is in the active state; select the The next process is processed in the physical equipment in the workshop.
  • Patent Application Publication No. CN106094759A called a real-time control method for workshop scheduling in a complex production environment.
  • the workpiece agent and the equipment agent work together on the scheduling behavior of the production workshop.
  • the shortcomings of this design are: the workpiece agent lacks an effective computing carrier, and one or more computers are needed to provide computing power for the workpiece agent; when the number of shop orders continues to increase, the number of workpiece agents continues to increase, which will It will consume a lot of computing resources and equipment.
  • multiple artifact agents based on the same computing carrier are not conducive to system stability and maintenance operations.
  • Patent Application Publication No. CN106527373A named as a multi-agent based workshop autonomous scheduling system and method.
  • the workpiece agent group, equipment agent group and logistics agent work together to dispatch the production workshop.
  • the workpiece agent group includes a workpiece agent, a workpiece manager agent, and a job manager agent.
  • the device agent group includes a device agent, a device manager agent, and a device manager agent.
  • the grouping of agents is based on a tree structure, which is controlled by the general manager and in charge of controlling a single agent. If an agent in charge fails, all agents under the node in charge are in a paralyzed state, and if the agent in charge of the master fails, all agents are in a paralyzed state. This has higher stability and computing power requirements for the computing carriers of the in-charge agent and the manager agent.
  • the composite agent includes a device agent and a workpiece agent.
  • the composite agent exists on every physical device in the workshop. among them:
  • the agent carrier is the physical equipment in the workshop, which matches its own physical equipment to provide this type of processing service to other agents in the workshop; drives the physical equipment to perform processing services; has the ability to make independent decisions based on the workshop environment and The task status selects the next part to be processed in its own workpiece buffer.
  • the feature of the work piece agent is that the agent has no specific calculation carrier, and its data carrier is located on the work piece carrier.
  • the agent has two states, one is inactive state, and the other is active state. When in the inactive state, the agent has no computing capabilities and only data is stored on the workpiece carrier; when in the activated state, the agent’s computing carrier is the carrier of the device agent that activates it. At this time, the agent is based on the current Processing information, as well as workshop environment information, make independent decisions about processing equipment for the next process.
  • a composite agent is designed to realize autonomous scheduling of workshop production, which effectively reduces the number of agents in the workshop, reduces the requirements and quantity of computing carriers, and reduces the cost of workshop hardware.
  • this method can effectively control the number of concurrent workshops, reduce the cost of workshop network communication, and reduce the computing power requirements of computing equipment.
  • this method can reduce the number of workshop agents and reduce the requirements and number of computing carriers.
  • this method uses a non-central node structure and does not require high stability. And the central node of high computing power.

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)
  • General Factory Administration (AREA)

Abstract

一种基于复合智能体的车间实时调度方法及装置，能够降低对运算载体的要求和数量，同时采用的是无中心节点结构，不需要高稳定性和高运算力的中心节点。本发明包括：设备智能体检测到物理设备空闲时，选择工件缓冲区内的工件；读取所选择工件所在工件托盘的RFID芯片内的数据，激活工件智能体并根据读取的RFID芯片内的数据生成对应所选择工件的工件智能体程序；在完成下一道工序加工协商后，所生成的工件智能体更新所选择工件所在工件托盘的RFID芯片内的数据；销毁对应所选择工件的工件智能体程序，并将工件智能体切换为非激活状态。

Description

一种基于复合智能体的车间实时调度方法及装置 技术领域
本发明涉及车间调度控制技术领域，尤其涉及一种基于复合智能体的车间实时调度方法及装置。
背景技术
随着计算机和网络技术的发展，多智能体制造系统已成为一种新型智能制造模式，受到越来越多人的青睐，被认为是解决当前及以后制造业困境的有效手段。在中国发展智能制造的过程中，多智能体制造技术成为一个需要重点关注的课题。
目前也已有一些多智能体制造被研发并应用，例如：
专利申请公布号CN106094759A，名为一种复杂生产环境下车间调度实时控制方法。以工件智能体和设备智能体共同作用于生产车间调度行为。这种设计的不足之处是：工件智能体缺少有效运算载体，需要额外一台或多台计算机为工件智能体提供运算能力；当车间订单数量不断增加，工件智能体的数量不断增加，这将会消耗大量运算资源和设备，与此同时，多个工件智能体基于同一个运算载体上不利于系统稳定以及维护操作等。
专利申请公布号CN106527373A，为名基于多智能体的车间自主调度系统和方法。以工件智能体组、设备智能体组和物流智能体共同作用于生产车间调度。工件智能体组内包含工件智能体、工件分管智能体和工件总管智能体，设备智能体组内包含设备智能体、设备分管智能体和设备总管智能体。这种设计的不足之处：智能体分组是基于树形结构，由总管控制分管，分管控制单个智能体。若某个分管智能体故障后，该分管节点下的所有智能体均处于瘫痪状态，若总 管智能体故障后，所有智能体处于瘫痪状态。这对分管智能体和总管智能体的运算载体有较高的稳定性和运算能力要求，系统对于中心节点的抗风险性要求较高。
发明内容
本发明的实施例提供一种基于复合智能体的车间实时调度方法及装置，能够降低对运算载体的要求和数量，同时采用的是无中心节点结构，不需要高稳定性和高运算力的中心节点。
为达到上述目的，本发明的实施例采用如下技术方案：
第一方面，本发明的实施例提供的方法，包括：
设备智能体检测到物理设备空闲时，选择工件缓冲区内的工件；
读取所选择工件所在工件托盘的RFID芯片内的数据，激活工件智能体并根据读取的RFID芯片内的数据生成对应所选择工件的工件智能体程序；
在完成下一道工序加工协商后，所生成的工件智能体更新所选择工件所在工件托盘的RFID芯片内的数据；
销毁所述对应所选择工件的工件智能体程序，并将工件智能体切换为非激活状态。
第一方面，本发明的实施例提供的装置，包括：
设备智能体，用于检测到物理设备空闲时，选择工件缓冲区内的工件；读取所选择工件所在工件托盘的RFID芯片内的数据，激活工件智能体并根据读取的RFID芯片内的数据生成对应所选择工件的工件智能体程序；
所述工件智能体，用于完成下一道工序加工协商，并更新所选择工件所在工件托盘的RFID芯片内的数据；之后，销毁所述对应所选择工件的工件智能体程序，并将工件智能体切换为非激活状态。
本实施例通过设计一种复合智能体，来实现车间生产自主调度，有效的降低了车间内智能体数量，降低了运算载体的要求和数量，降低车间硬件成本。通过赋予工件智能体两种状态，有效的控制车间并发数量，降低车间网络通讯成本，降低运算设备的运算能力要求。本实施例中，相比其他多智能体的车间调度方法，该方法可以降低车间智能体数量，降低对运算载体的要求和数量，同时该方法采用的是无中心节点结构，不需要高稳定性和高运算力的中心节点。
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简单地介绍，显而易见地，下面描述中的附图仅仅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它的附图。
图1为本发明实施例提供的物理设备布局示意图。
图2为本发明实施例提供的复合智能体结构示意图。
图3为本发明实施例提供的复合智能体运行示意图。
图4为本发明实施例提供的方法流程示意图。
具体实施方式
为使本领域技术人员更好地理解本发明的技术方案，下面结合附图和具体实施方式对本发明作进一步详细描述。下文中将详细描述本发明的实施方式，所述实施方式的示例在附图中示出，其中自始至终相同或类似的标号表示相同或类似的元件或具有相同或类似功能的元件。下面通过参考附图描述的实施方式是示例性的，仅用于解释本发明，而不能解释为对本发明的限制。本技术领域技术人员可以理解，除非特意声明，这里使用的单数形式“一”、“一个”、“所述”和“该”也可包括复数形式。应该进一步理解的是，本发明的说明书 中使用的措辞“包括”是指存在所述特征、整数、步骤、操作、元件和/或组件，但是并不排除存在或添加一个或多个其他特征、整数、步骤、操作、元件、组件和/或它们的组。应该理解，当我们称元件被“连接”或“耦接”到另一元件时，它可以直接连接或耦接到其他元件，或者也可以存在中间元件。此外，这里使用的“连接”或“耦接”可以包括无线连接或耦接。这里使用的措辞“和/或”包括一个或更多个相关联的列出项的任一单元和全部组合。本技术领域技术人员可以理解，除非另外定义，这里使用的所有术语(包括技术术语和科学术语)具有与本发明所属领域中的普通技术人员的一般理解相同的意义。还应该理解的是，诸如通用字典中定义的那些术语应该被理解为具有与现有技术的上下文中的意义一致的意义，并且除非像这里一样定义，不会用理想化或过于正式的含义来解释。
本实施例可以实现在一种如图1所示的系统上，所述系统包括了物理设备、上下料装置和工件缓冲台。其中，物理设备可以理解为：车床、铣床、磨床、冲压机、刨床、特种加工设备等目前业内常用的、用于工件加工的设备，在实际应用中，一个生产车间中通常会布置多个物理设备，这些物理设备可以布置在上下料装置周围，以便于通过上下料装置传递工件。上下料装置，具体用于通过上下料装置上安装的机械手或者传送带将工件从工件缓冲台放入物理设备。
具体的，所述物理设备通过所述上下料装置将工件从工件缓冲台放入物理设备中进行加工。所述工件缓冲台上安装RFID读写器，转移至所述工件缓冲台的工件放置在工件托盘中，所述工件托盘上安装RFID芯片。其中，工件缓冲台配备有RFID读写器；工件放置在工件托盘中；工件托盘配备有RFID芯片，可供RFID读写器读写其中的信息；工件托盘RFID芯片储存有工件索引号、工件几何尺寸、工艺状态参数、当前状态参数、工艺流程信息。
所述复合智能体在运算载体上运行，所述运算载体连接至所述工件缓冲区RFID读写器和所述物理设备，所述复合智能体包括设备智能体和工件智能体。即将复合智能体部署在车间中的每一台物理设备上。
需要说明的是，本实施例中所述的工件智能体，可以按照业内通常的理解，指的是一段预先定义好的程序，它是由设备智能体启动这段程序，程序所需参数由设备智能体从工件托盘RFID芯片读取。
进一步的，还包括：所述设备智能体通过所述物理通讯接口监测物理设备状态；和，所述设备智能体通过所述物理通讯接口控制工件缓冲区RFID读写器，从所述工件托盘上的RFID芯片内读取信息数据。例如：如图2所示，复合智能体存在于运算载体(运算载体可以是嵌入式运算设备、工控机等)，运算载体通过物理通讯接口(以太网、串口、CAN总线、USB等接口)连接至物理设备和工件缓冲区RFID读写器。
本发明实施例提供一种基于复合智能体的车间实时调度方法，如图4所示，包括：
S1、设备智能体检测到物理设备空闲时，选择工件缓冲区内的工件。
S2、读取所选择工件所在工件托盘的RFID芯片内的数据，激活工件智能体并根据读取的RFID芯片内的数据生成对应所选择工件的工件智能体程序。
S3、在完成下一道工序加工协商后，所生成的工件智能体更新所选择工件所在工件托盘的RFID芯片内的数据。
S4、销毁所述对应所选择工件的工件智能体程序，并将工件智能体切换为非激活状态。
在本实施例中，所述设备智能体的载体为车间中的物理设备；所述设备智 能体匹配当前所在的物理设备，并驱动所述当前所在的物理设备向所述车间中的其它智能体提供加工服务。
因此所述方法还包括：通过所述设备智能体根据车间环境和任务状态从所述工件缓冲区中选择下一个加工零件。
所述工件智能体的数据载体位于工件载体上；所述工件智能体的状态包括：非激活状态和激活状态，其中，处于所述非激活状态时，所述工件智能体不占用运算资源；处于所述激活状态时，激活所述工件智能体的设备智能体的载体，向所述工件智能体提供运算资源。其中，工件载体为附带有RFID芯片的用于盛放工件的容器，例如：工件载体是工件托盘，用于盛放工件，托盘上并附带有RFID芯片，用于记录当前工件状态的数据。
例如：复合智能体内的设备智能体通过物理通讯接口控制物理设备并监测物理设备状态，设备智能体通过物理通讯接口控制工件缓冲区RFID读写器，读取和修改工件托盘RFID内的信息数据。在物理设备空闲时，设备智能体根据相应的决策算法选择工件缓冲区内的某一工件，读取其托盘RFID芯片内的信息数据，生成相应的工件智能体程序，此时该工件智能体进入激活状态(运行中的程序段)。当激活的工件智能体完成下一道工序加工协商，通过物理通讯总线更新工件托盘内RFID芯片信息数据，随后销毁自身运行程序，变成非激活状态(静态状态数据)。
具体的，所述根据读取的RFID芯片内的数据生成对应所选择工件的工件智能体程序，包括：
从所述RFID芯片内的数据中，提取所选取工件的几何尺寸信息、工艺状态信息和当前状态信息。根据所提取的信息生成工件智能体程序，并通过所述物 理通讯接口传输至所述物理设备。例如：如图3所示，复合智能体内部设备智能体根据相应决策算法(机器学习、遗传算法、动态规划算法)从工件缓冲区选择相应的工件进行加工，加工所需的几何尺寸、工艺状态信息、当前状态信息都储存在工件托盘的RFID芯片内。设备智能体通过物理通讯接口获取这些信息，生成NC程序并传输至物理设备内。
可选的，所述完成下一道工序加工协商，包括：
处于激活状态的工件智能体，根据所选取工件的的当前状态信息和工艺状态信息，向车间内的物理设备发起加工协商请求。根据预设的决策算法，选择所述车间内的物理设备中进行下一道工序加工。例如：当正在加工的工件完成当前工序加工时，设备智能体根据工件的状态信息生成对应的工件智能体程序，即该工件智能体转入激活状态。处于激活状态的工件智能体根据工件当前状态信息和工艺流程信息向车间内其他物理设备发起加工协商请求，根据相应的决策算法(强化学习、合同网机制、细胞免疫算法等)选择合适的物理设备进行下一道工序加工。完成选择后，工件智能体更新工件托盘RFID芯片内信息数据，向运输设备(AGV、传送带)发起移动请求后，销毁自身运行程序，即工件智能体转入非激活状态。
本发明实施例还提供一种如图2、3所示的，基于复合智能体的车间实时调度装置，其中：
设备智能体，用于检测到物理设备空闲时，选择工件缓冲区内的工件；读取所选择工件所在工件托盘的RFID芯片内的数据，激活工件智能体并根据读取的RFID芯片内的数据生成对应所选择工件的工件智能体程序。
所述工件智能体，用于完成下一道工序加工协商，并更新所选择工件所在 工件托盘的RFID芯片内的数据；之后，销毁所述对应所选择工件的工件智能体程序，并将工件智能体切换为非激活状态。
具体的，所述物理设备通过所述上下料装置将工件从工件缓冲台放入物理设备中进行加工；所述工件缓冲台上安装RFID读写器，转移至所述工件缓冲台的工件放置在工件托盘中，所述工件托盘上安装RFID芯片；复合智能体在运算载体上运行，所述运算载体连接至所述工件缓冲区RFID读写器和所述物理设备，所述复合智能体包括设备智能体和工件智能体。
具体的，所述设备智能体的载体为车间中的物理设备；所述设备智能体匹配当前所在的物理设备，并驱动所述当前所在的物理设备向所述车间中的其它智能体提供加工服务；所述设备智能体，还用于根据车间环境和任务状态从所述工件缓冲区中选择下一个加工零件。
具体的，所述工件智能体的数据载体位于工件载体上；所述工件智能体的状态包括：非激活状态和激活状态，其中，处于所述非激活状态时，所述工件智能体不占用运算资源；处于所述激活状态时，激活所述工件智能体的设备智能体的载体，向所述工件智能体提供运算资源。
在本实施例中，所述设备智能体，具体用于从所述RFID芯片内的数据中，提取所选取工件的几何尺寸信息、工艺状态信息和当前状态信息；根据所提取的信息生成工件智能体程序，并通过所述物理通讯接口传输至所述物理设备。
所述工件智能体，具体用于在处于激活状态时，根据所选取工件的的当前状态信息和工艺状态信息，向车间内的物理设备发起加工协商请求；根据预设的决策算法，选择所述车间内的物理设备中进行下一道工序加工。
在现有方案中，存在很多问题，例如：专利申请公布号CN106094759A，名为一种复杂生产环境下车间调度实时控制方法。以工件智能体和设备智能体共同作用于生产车间调度行为。这种设计的不足之处是：工件智能体缺少有效运算载体，需要额外一台或多台计算机为工件智能体提供运算能力；当车间订单数量不断增加，工件智能体的数量不断增加，这将会消耗大量运算资源和设备，与此同时，多个工件智能体基于同一个运算载体上不利于系统稳定以及维护操作等。
再例如：专利申请公布号CN106527373A，为名基于多智能体的车间自主调度系统和方法。以工件智能体组、设备智能体组和物流智能体共同作用于生产车间调度。工件智能体组内包含工件智能体、工件分管智能体和工件总管智能体，设备智能体组内包含设备智能体、设备分管智能体和设备总管智能体。
这种设计的不足之处：智能体分组是基于树形结构，由总管控制分管，分管控制单个智能体。若某个分管智能体故障后，该分管节点下的所有智能体均处于瘫痪状态，若总管智能体故障后，所有智能体处于瘫痪状态。这对分管智能体和总管智能体的运算载体有较高的稳定性和运算能力要求。
在本实施例中，复合智能体包含设备智能体和工件智能体。复合智能体存在于车间中每一台物理设备上。其中：
设备智能体的特征在于：该智能体载体是车间中的物理设备，它匹配自身物理设备向车间其他智能体提供该类加工服务；驱动物理设备进行加工服务；拥有自主决策能力，根据车间环境以及任务状态选择自身工件缓冲区下一个加工零件。
工件智能体的特征在于：该智能体没有特定运算载体，其数据载体位于工件载体上。该智能体拥有两种状态，一是非激活状态，二是激活状态。处于非 激活状态时，该智能体没有运算能力，只有数据储存于工件载体上；处于激活状态时，该智能体的运算载体是激活它的设备智能体的载体，此时，该智能体根据当前加工信息，以及车间环境信息，自主决策下一道工序加工设备。
本实施例通过设计一种复合智能体，来实现车间生产自主调度，有效的降低了车间内智能体数量，降低了运算载体的要求和数量，降低车间硬件成本。通过赋予工件智能体两种状态，有效的控制车间并发数量，降低车间网络通讯成本，降低运算设备的运算能力要求。本实施例中，相比其他多智能体的车间调度方法，该方法可以降低车间智能体数量，降低对运算载体的要求和数量，同时该方法采用的是无中心节点结构，不需要高稳定性和高运算力的中心节点。
本说明书中的各个实施例均采用递进的方式描述，各个实施例之间相同相似的部分互相参见即可，每个实施例重点说明的都是与其他实施例的不同之处。尤其，对于设备实施例而言，由于其基本相似于方法实施例，所以描述得比较简单，相关之处参见方法实施例的部分说明即可。以上所述，仅为本发明的具体实施方式，但本发明的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本发明揭露的技术范围内，可轻易想到的变化或替换，都应涵盖在本发明的保护范围之内。因此，本发明的保护范围应该以权利要求的保护范围为准。

Claims (13)

  1. 一种基于复合智能体的车间实时调度方法，其特征在于，包括：
    设备智能体检测到物理设备空闲时，选择工件缓冲区内的工件；
    读取所选择工件所在工件托盘的RFID芯片内的数据，激活工件智能体并根据读取的RFID芯片内的数据生成对应所选择工件的工件智能体程序；
    在完成下一道工序加工协商后，所生成的工件智能体更新所选择工件所在工件托盘的RFID芯片内的数据；
    销毁所述对应所选择工件的工件智能体程序，并将工件智能体切换为非激活状态。
  2. 根据权利要求1所述的方法，其特征在于，所述物理设备通过所述上下料装置将工件从工件缓冲台放入物理设备中进行加工；
    所述工件缓冲台上安装RFID读写器，转移至所述工件缓冲台的工件放置在工件托盘中，所述工件托盘上安装RFID芯片；
    所述复合智能体在运算载体上运行，所述运算载体连接至所述工件缓冲区RFID读写器和所述物理设备，所述复合智能体包括设备智能体和工件智能体。
  3. 根据权利要求2所述的方法，其特征在于，所述设备智能体的载体为车间中的物理设备；
    所述设备智能体匹配当前所在的物理设备，并驱动所述当前所在的物理设备向所述车间中的其它智能体提供加工服务；
    所述方法还包括：
    通过所述设备智能体根据车间环境和任务状态从所述工件缓冲区中选择下一个加工零件。
  4. 根据权利要求3所述的方法，其特征在于，所述工件智能体的数据载体位于工件载体上；
    所述工件智能体的状态包括：非激活状态和激活状态，其中，处于所述非激活状态时，所述工件智能体不占用运算资源；处于所述激活状态时，激活所述工件智能体的设备智能体的载体，向所述工件智能体提供运算资源。
  5. 根据权利要求2所述的方法，其特征在于，还包括：
    所述设备智能体通过所述物理通讯接口监测物理设备状态；
    和，所述设备智能体通过所述物理通讯接口控制工件缓冲区RFID读写器，从所述工件托盘上的RFID芯片内读取信息数据。
  6. 根据权利要求2所述的方法，其特征在于，所述根据读取的RFID芯片内的数据生成对应所选择工件的工件智能体程序，包括：
    从所述RFID芯片内的数据中，提取所选取工件的几何尺寸信息、工艺状态信息和当前状态信息；
    根据所提取的信息生成工件智能体程序，并通过所述物理通讯接口传输至所述物理设备。
  7. 根据权利要求6所述的方法，其特征在于，所述完成下一道工序加工协商，包括：
    处于激活状态的工件智能体，根据所选取工件的的当前状态信息和工艺状态信息，向车间内的物理设备发起加工协商请求；
    根据预设的决策算法，选择所述车间内的物理设备中进行下一道工序加工。
  8. 一种基于复合智能体的车间实时调度装置，其特征在于，包括：
    设备智能体，用于检测到物理设备空闲时，选择工件缓冲区内的工件；读 取所选择工件所在工件托盘的RFID芯片内的数据，激活工件智能体并根据读取的RFID芯片内的数据生成对应所选择工件的工件智能体程序；
    所述工件智能体，用于完成下一道工序加工协商，并更新所选择工件所在工件托盘的RFID芯片内的数据；之后，销毁所述对应所选择工件的工件智能体程序，并将工件智能体切换为非激活状态。
  9. 根据权利要求8所述的装置，其特征在于，所述物理设备通过所述上下料装置将工件从工件缓冲台放入物理设备中进行加工；
    所述工件缓冲台上安装RFID读写器，转移至所述工件缓冲台的工件放置在工件托盘中，所述工件托盘上安装RFID芯片；
    复合智能体在运算载体上运行，所述运算载体连接至所述工件缓冲区RFID读写器和所述物理设备，所述复合智能体包括设备智能体和工件智能体。
  10. 根据权利要求9所述的装置，其特征在于，所述设备智能体的载体为车间中的物理设备；
    所述设备智能体匹配当前所在的物理设备，并驱动所述当前所在的物理设备向所述车间中的其它智能体提供加工服务；
    所述设备智能体，还用于根据车间环境和任务状态从所述工件缓冲区中选择下一个加工零件。
  11. 根据权利要求10所述的装置，其特征在于，所述工件智能体的数据载体位于工件载体上；
    所述工件智能体的状态包括：非激活状态和激活状态，其中，处于所述非激活状态时，所述工件智能体不占用运算资源；处于所述激活状态时，激活所述工件智能体的设备智能体的载体，向所述工件智能体提供运算资源。
  12. 根据权利要求9所述的装置，其特征在于，所述设备智能体，具体用于 从所述RFID芯片内的数据中，提取所选取工件的几何尺寸信息、工艺状态信息和当前状态信息；根据所提取的信息生成工件智能体程序，并通过所述物理通讯接口传输至所述物理设备。
  13. 根据权利要求12所述的装置，其特征在于，所述工件智能体，具体用于在处于激活状态时，根据所选取工件的的当前状态信息和工艺状态信息，向车间内的物理设备发起加工协商请求；根据预设的决策算法，选择所述车间内的物理设备中进行下一道工序加工。
PCT/CN2019/109886 2019-09-25 2019-10-08 一种基于复合智能体的车间实时调度方法及装置 WO2021056605A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910911650.4 2019-09-25
CN201910911650.4A CN110764472B (zh) 2019-09-25 2019-09-25 一种基于复合智能体的车间实时调度方法及装置

Publications (1)

Publication Number Publication Date
WO2021056605A1 true WO2021056605A1 (zh) 2021-04-01

Family

ID=69330223

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/109886 WO2021056605A1 (zh) 2019-09-25 2019-10-08 一种基于复合智能体的车间实时调度方法及装置

Country Status (2)

Country Link
CN (1) CN110764472B (zh)
WO (1) WO2021056605A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115796387A (zh) * 2023-01-29 2023-03-14 深圳市普拉托科技有限公司 托盘调度寻优方法、装置、设备及存储介质

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113377073B (zh) * 2021-06-28 2022-09-09 西南交通大学 一种基于双层多智能体系统的柔性作业车间调度优化方法

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106527373A (zh) * 2016-12-05 2017-03-22 中国科学院自动化研究所 基于多智能体的车间自主调度系统和方法
CN106873564A (zh) * 2017-04-26 2017-06-20 南京航空航天大学 基于智能车间的流动式多智能体实时调度系统及方法
CN109164762A (zh) * 2018-07-12 2019-01-08 南京航空航天大学 一种制造系统的智能模块构建方法、制造系统以及制造方法

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9875120B2 (en) * 2013-06-24 2018-01-23 Microsoft Technology Licensing, Llc Virtualized components in computing systems
FR3033655B1 (fr) * 2015-03-10 2017-04-21 Spring Tech Procede de generation d'un programme d'usinage

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106527373A (zh) * 2016-12-05 2017-03-22 中国科学院自动化研究所 基于多智能体的车间自主调度系统和方法
CN106873564A (zh) * 2017-04-26 2017-06-20 南京航空航天大学 基于智能车间的流动式多智能体实时调度系统及方法
CN109164762A (zh) * 2018-07-12 2019-01-08 南京航空航天大学 一种制造系统的智能模块构建方法、制造系统以及制造方法

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115796387A (zh) * 2023-01-29 2023-03-14 深圳市普拉托科技有限公司 托盘调度寻优方法、装置、设备及存储介质

Also Published As

Publication number Publication date
CN110764472A (zh) 2020-02-07
CN110764472B (zh) 2022-04-05

Similar Documents

Publication Publication Date Title
WO2021056605A1 (zh) 一种基于复合智能体的车间实时调度方法及装置
CN106873564B (zh) 基于智能车间的流动式多智能体实时调度方法
US8527080B2 (en) Method and system for managing process jobs in a semiconductor fabrication facility
US7089076B2 (en) Scheduling multi-robot processing systems
CN111882215A (zh) 一种含有agv的个性化定制柔性作业车间调度方法
CN108229853B (zh) 一种基于静态调度预测的智慧工厂动态协同调度方法
CN116976653A (zh) 一种基于mes系统的多工厂产能协调方法
WO2020103810A1 (zh) Agv基点式通讯系统、通讯方法以及计算机可读存储介质
CN103885393B (zh) 集群控制系统以及其运行方法
CN106598483A (zh) 一种面向异构内存的可重构内存管理系统
CN203224756U (zh) 集群控制系统
CN101414170B (zh) 用于集成电路生产线的amhs系统及其实现方法
Wang et al. A smart factory solution to hybrid production of multi-type products with reduced intelligence
Hasan et al. Performance modelling of dispatching strategies under resource failure scenario in reconfigurable manufacturing system
Wang et al. Probing an Easy-to-Deploy Multi-Agent Manufacturing System Based on Agent Computing Node: Architecture, Implementation, and Case Study
Wu et al. Schedulability and scheduling of dual-arm cluster tools with residency time constraints based on Petri net
JPS59214910A (ja) フレキシブル複合加工セルの運転制御方式
Ray et al. A production management information model for discrete manufacturing
Hrúz et al. Composition of conflict-free Petri net models for control of flexible manufacturing systems
Halevi Integration of PM into CIM
Ng et al. Fast self-organizing holonic based multi-equipment control
CN106776427A (zh) 通信方法及装置
JPH03105632A (ja) 相互スタンバイシステム
Zhu Real-time adaptive task schedule for a flexible manufacturing system
KR100221101B1 (ko) 유연 생산 시스템용 절삭 데이터 베이스 시스템

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: 19946754

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19946754

Country of ref document: EP

Kind code of ref document: A1