WO2021258564A1 - 一种工艺节拍处理方法、系统、装置及存储介质 - Google Patents

一种工艺节拍处理方法、系统、装置及存储介质 Download PDF

Info

Publication number
WO2021258564A1
WO2021258564A1 PCT/CN2020/116016 CN2020116016W WO2021258564A1 WO 2021258564 A1 WO2021258564 A1 WO 2021258564A1 CN 2020116016 W CN2020116016 W CN 2020116016W WO 2021258564 A1 WO2021258564 A1 WO 2021258564A1
Authority
WO
WIPO (PCT)
Prior art keywords
beat
action
production
information
beat information
Prior art date
Application number
PCT/CN2020/116016
Other languages
English (en)
French (fr)
Inventor
向玉文
左志军
贺毅
姚维兵
徐华昕
张凯
Original Assignee
广州明珞装备股份有限公司
明珞汽车装备(上海)有限公司
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 广州明珞装备股份有限公司, 明珞汽车装备(上海)有限公司 filed Critical 广州明珞装备股份有限公司
Publication of WO2021258564A1 publication Critical patent/WO2021258564A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/22Matching criteria, e.g. proximity measures
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/21Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/24Classification techniques
    • G06F18/243Classification techniques relating to the number of classes
    • G06F18/24323Tree-organised classifiers
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/04Manufacturing
    • 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
    • Y02P90/30Computing systems specially adapted for manufacturing

Definitions

  • the present invention relates to the technical field of production processes, in particular to a process beat processing method, system, device and storage medium.
  • process beat In the production process, all the production actions that complete a process is usually called a beat, also called a Cycle (that is, process beat). Each process beat may be composed of one action group or multiple action groups, and each action group is composed of many production actions.
  • the collection of process beat data is essential for subsequent statistical analysis to improve production efficiency and production quality.
  • the abnormal process beats caused by the above two situations are not considered, which leads to inaccurate data such as the sequence of actions, the number of actions, and the duration of the beat, which affects the subsequent statistical analysis.
  • the development of technology it is more and more common for the same production line to adapt to different process production. If the original process beat processing method is continued, it will definitely greatly increase the abnormal ratio of process beat data, which will lead to subsequent statistics on the production process.
  • the analysis is seriously inconsistent with the actual production situation, which prevents the manufacturer from making process adjustments and optimizations based on the actual production situation, which is not conducive to reducing production costs and restricts the improvement of production efficiency and production quality.
  • the purpose of the present invention is to provide a process beat processing method, system, device and storage medium, which can adapt to the process beat statistics under various conditions and greatly improve the accuracy of the process beat statistical results. It meets the needs of manufacturers to improve production quality and production efficiency, and reduce production costs.
  • a process beat processing method including the following steps:
  • the abnormal beat information is matched with the feature decision tree obtained by pre-training to determine the second beat information.
  • step of obtaining the action data of each production action includes:
  • the message queue of the Internet of Things is analyzed and processed to obtain the action data of each production action.
  • the action data includes station ID, action ID, action duration, and flag bit parameters.
  • the flag bit parameters include 1 and 0.
  • the production action corresponding to the flag bit parameter of 1 is the start of a process cycle or The end action, the production action corresponding to the flag parameter being 0 is an intermediate action of a process beat.
  • the step of obtaining the first beat information of each process beat according to the motion data includes:
  • the Key of the key-value pair includes the station ID
  • the Value of the key-value pair includes the action ID and the action duration
  • the first beat information includes a first action sequence, a first action quantity, and a first beat duration.
  • the step of verifying the first beat information and determining abnormal beat information includes at least one of the following:
  • the first threshold value range is a preset threshold value range of the number of process cycle actions
  • the second threshold value range is a preset threshold value range of process cycle time length.
  • the second beat information includes a second action sequence, a second action number, and a second beat duration
  • the abnormal beat information is matched with a feature decision tree obtained by pre-training to determine the second beat information.
  • the steps include:
  • the second action sequence, the second action quantity, and the second beat duration are stored.
  • the process beat processing method further includes the step of training a feature decision tree, which is specifically:
  • the action sequence of the process beat in the preset time period is acquired as a training data set, and the feature decision tree is obtained through machine learning training.
  • a process beat processing system including:
  • the action data acquisition module is used to acquire the action data of each production action
  • the first beat information determining module is used to obtain the first beat information of each process beat according to the motion data
  • a verification module configured to verify the first beat information and determine abnormal beat information
  • the matching module is used to match the abnormal beat information with the feature decision tree obtained by pre-training to determine the second beat information.
  • a process beat processing device including:
  • At least one processor At least one processor
  • At least one memory for storing at least one program
  • the at least one processor When the at least one program is executed by the at least one processor, the at least one processor implements the process beat processing method.
  • a computer-readable storage medium stores therein instructions executable by a processor, and the instructions executable by the processor are used to execute the process beat processing method when executed by the processor.
  • the beneficial effects of the present invention are: the process beat processing method, system, device and storage medium of the present invention, by matching the abnormal beat information with the feature decision tree, the determined second beat information can accurately represent the process beat in the production process It can adapt to the process beat statistics under various conditions, and greatly improves the accuracy of the process beat statistical results, which is convenient for subsequent process adjustment and optimization of the actual production situation based on the process beat statistical results, ensuring production efficiency and production quality. Improve, reduce production risks and production costs.
  • FIG. 1 is a flow chart of the steps of a process beat processing method provided by an embodiment of the present invention
  • FIG. 2 is a structural block diagram of a process beat processing system provided by an embodiment of the present invention.
  • FIG. 3 is a structural block diagram of a process beat processing device provided by an embodiment of the present invention.
  • Figure 4 is a schematic diagram of a feature decision tree provided by an embodiment of the present invention.
  • Fig. 5 is a schematic flow chart of a process tick processing method provided by a preferred embodiment of the present invention.
  • an embodiment of the present invention provides a process beat processing method, including the following steps:
  • the action data includes information such as the station performing the production action, the duration of the production action, and the sequence of the production action in a process beat.
  • the action data needs to be acquired in real time.
  • the above step S101 specifically includes the following steps:
  • S1011 collect production data in real time and upload it to the message queue of the Internet of Things
  • the production data in the production process can be collected in real time through the collector, and the production data can be uploaded to the IoT message queue at a preset time.
  • the production data in the Internet of Things message queue can be distributed and consumed in real time through a data analysis program.
  • the program reads the production data, it integrates, analyzes, converts, and cleans the production data, and performs processing on important attribute values.
  • the verification process obtains the action data corresponding to each production action. It should be understood that the production data in the above step S1011 is only the real-time status data of the current production process of a certain machine and equipment, and does not completely represent a process beat or a production action, nor does it include the execution of a certain production action.
  • the distributed processing capability of the big data platform is utilized, the action data can be obtained quickly and accurately, and the processing efficiency of the process beat is improved.
  • the action data includes station ID, action ID, action duration, and flag bit parameters.
  • the flag bit parameters include 1 and 0.
  • the production action corresponding to the flag bit parameter of 1 is the start of a process cycle or End action, the production action corresponding to the flag bit parameter being 0 is an intermediate action of a process beat.
  • the action data includes information such as station ID, action ID, action time length, and flag position parameters.
  • the station ID indicates the station performing the production action. Since the production data may be collected at the same time, multiple stations may be collected. Need to use the station ID to distinguish, the station ID of all production actions in a process cycle is the same; the action ID represents the label of the production action, and the label corresponding to each production action can be set in advance; the action duration represents the time consumed to execute the production action , The action time needs to be obtained in real time; the flag bit parameter is used to determine the start and end of a process beat, the flag bit parameter can be 1 or 0, the flag bit parameter 1 indicates the beginning of a process beat, and the production action corresponding to the flag bit parameter 0 is It is an intermediate action of a craft beat, until the next flag bit parameter 1, which means that a craft beat ends, and this is used as the start of the next craft beat. For example, if the sequence of flag bit parameters is 1000000100010000000001, the sequence corresponds to three process beats, which
  • the flag bit parameter can accurately and concisely indicate the production actions included in a process beat, which is convenient for subsequent statistics such as the duration and action sequence of the process beat.
  • the action data may also include the action group ID and the line body ID.
  • a process beat may include one or more action groups, and each action group contains multiple production actions.
  • the production actions are performed in the unit of the action group, and the action group ID represents the action group.
  • Label; line body value ID is the dimension element of the front-end BI report.
  • the first beat information includes the action sequence of the production action in the process beat (ie the first action sequence), the number of production actions in the process beat (ie the first action number), and the total duration of the production action in the process beat (ie, the first action sequence).
  • One beat duration) the first beat information can be obtained based on the motion data of all production actions in the statistical process beat.
  • the Key of the key-value pair includes the station ID
  • the Value of the key-value pair includes the action ID and the action duration
  • the key-value pair is cached in the Redis cluster
  • the station ID can be used as the Key value, and the action duration and the action ID can be connected by a string as the Value value, so as to obtain the key-value pair corresponding to each production action and cache it in Redis In the cluster.
  • the station ID of a certain production action is 1001
  • the action duration is 20s
  • the action ID is A1
  • the Key value of the corresponding key-value pair can be expressed as "1001”
  • the Value value can be expressed as "20, A1”. Since the Key values of all production actions under the same workstation are the same, the corresponding Value values can be used as a set, and the key-value pairs obtained in this way are cached in the Redis cluster.
  • flag bit parameter of a certain action data when the flag bit parameter of a certain action data is received is 0, it means that the current production action is an intermediate action of the process beat, and only buffers are not read at this time; when the flag bit parameter of a certain action data is received is 1 , It means that the current process beat has ended.
  • the flag bit parameter of a certain action data is received is 1 .
  • the current process beat After completing the cache of the key-value pair corresponding to the action data, read all the Value values cached in the Redis cluster, count the action duration to get the first beat duration, and count the number of actions to get the first For the number of actions, sort the action IDs in order to obtain the first action sequence (such as A1-A2-A4-A7-A8).
  • the first beat information includes the first action sequence, the first action number, and the first beat duration.
  • the station ID and the action group ID can be connected by character strings to obtain the Key value
  • the action duration, the line body ID and the action ID can be connected by character strings as the Value value to generate key-value pairs corresponding to each production action.
  • the Key value can be expressed as "1001, A003"
  • the action duration is 20s
  • the line ID is 405
  • the action ID is A00301 (the previous part is consistent with the action group IDA003)
  • the Value can be expressed as "20,405,A00301". This is suitable for the situation where a process beat contains multiple action groups.
  • the action group ID and the line body ID will be omitted. It should be understood that when a process beat contains multiple action groups, the implementation can be further refined and modified in various ways. The following implementations I won't repeat it in the example.
  • the first beat information also includes the station ID. Since the production actions in the same process beat are all executed by the same station, the station IDs are all consistent, and the station ID in the first beat information is also consistent with each production action. The station ID is the same. The station ID can be used to subsequently select the corresponding feature decision tree for matching.
  • S103 Verify the first beat information, and determine the abnormal beat information.
  • step S103 can be implemented in at least two schemes, and the details are as follows.
  • step S103 it is specifically:
  • the first threshold value range is a preset threshold value range of the number of actions of the process cycle.
  • C min , C max are calculated by the average value algorithm, which is the value of continuous calculation and accumulation of historical data of production. With continuous statistical calculation of data, their values become more and more accurate. No detailed explanation.
  • step S103 it is specifically:
  • the second threshold value range is the preset threshold value range of the process beat duration, when the beat duration of a certain process beat does not meet the preset threshold range [T min , T max ], that is, the beat duration is too short or too long , It means that the first beat information is abnormal beat information.
  • T min and T max are also calculated by an average value algorithm.
  • the above two verification conditions can also be combined for verification, that is, if any one of the above two threshold ranges is not met, it means that the first beat information is abnormal beat information.
  • the first beat information is verified by the two verification conditions of the number of actions and the beat duration, which ensures that abnormal beat information can be filtered out, thereby facilitating subsequent processing of the abnormal beat information to obtain accurate beats. information.
  • a dictionary table of "characteristic abnormal actions” can be added as the verification condition.
  • the dictionary table includes several impossible adjacent actions. For example, the two production actions of A1 and A5 cannot be adjacent to each other.
  • A1-A5" is added to the dictionary table as a characteristic abnormal action. If the sequence segment "A1-A5" appears in the first action sequence, it means that the corresponding first beat information is abnormal beat information.
  • the embodiment of the present invention can improve the verification conditions of the abnormal beat information, and further improves the accuracy of the statistical results of the process beat.
  • S104 Match the abnormal beat information with the feature decision tree obtained by pre-training, and determine the second beat information.
  • FIG. 4 is a schematic diagram of a feature decision tree under a certain station in an embodiment of the present invention.
  • the root node of the feature decision tree is used to indicate the start action of the process beat
  • the intermediate node is used to indicate the intermediate action of the process beat.
  • the leaf node is used to indicate the end action of the process beat, and it can be known that each leaf node corresponds to a preset action sequence.
  • the action sequence in the abnormal beat information can be matched with the feature decision tree, and the preset action sequence with higher similarity can be found, so that accurate second beat information can be obtained to replace the abnormal first beat information, and the abnormal process beat data Converted into normal process beat data, push display and storage.
  • the second beat information includes the second action sequence, the second action number, and the second beat duration.
  • S1042 select the second action sequence according to the number of appearances of the third action sequence in the production process and the most recent appearance time;
  • the action sequence of the abnormal beat information is "A1-A2-A4-A7-A8”
  • the corresponding feature decision tree can be found according to its station ID, and then the feature decision tree can be matched to "A1-A2 according to the similarity" -A4-A7-A8-A9-A10-A12-14" and "A1-A2-A4-A7-A8-A10-A12-A13" these two preset action sequences with higher similarity, and then pass these two
  • the number of occurrences of a preset action sequence in the historical production process is selected as the second action sequence; if the number of occurrences is the same, you can also select the preset with the appearance time closer to the current time according to its recent appearance time
  • the action sequence serves as the second action sequence.
  • S1043 Obtain the corresponding second action quantity according to the second action sequence, and obtain the second beat duration according to the second action sequence and the preset action duration;
  • the second action quantity can be obtained by counting the number of production actions included in the second action sequence, and the second beat duration can be obtained according to the preset action length of each production action in the second action sequence.
  • the determined second beat information can accurately represent the craft beat in the production process, can adapt to the statistics of the craft beat in various situations, and greatly improve the craftsmanship.
  • the accuracy of the beat statistical results is convenient for subsequent process adjustment and optimization of the actual production situation based on the statistical results of the process beats, ensuring the improvement of production efficiency and production quality, and reducing production risks and production costs.
  • the process beat processing method further includes a step of training a feature decision tree, which is specifically:
  • the action sequence obtained from the production data of the normal production process within a preset time period can be used as the training data set, and the feature decision tree can be obtained through machine learning training. It can also be used in real-time and verified to be the first beat that is not abnormal beat information.
  • the action sequence of the information is trained as a training data set.
  • the obtained feature decision tree can be divided according to the station ID, and each station corresponds to a feature decision tree, which is convenient for subsequent matching of abnormal beat information.
  • the first beat information that filters out abnormal beat information can be used as a training data set to train the feature decision tree in real time, and continuously improve the prediction of the feature decision tree through machine learning. Set the accuracy and reliability of the action sequence.
  • the number of occurrences and the most recent occurrence time of each action sequence can be counted in real time, and the leaf nodes of the feature decision tree can be labeled to facilitate the selection of the most accurate second beat information.
  • FIG. 5 is a schematic flow diagram of a process beat processing method provided by a preferred embodiment of the present invention.
  • the production data is collected in real time by the collector and uploaded to the message queue of the Internet of Things, using the distributed processing of the big data platform Ability to obtain the action data of each production action, and then combine the distributed Redis cluster to complete the statistics of the process beat.
  • the normal beat information is directly output, and it is also used as the training data to train the feature decision tree, and the abnormal beat information is passed
  • the feature decision tree is matched and revised to obtain accurate beat information.
  • This embodiment can further optimize the process tick processing process and improve the utilization of production data resources.
  • the feature decision tree changes in real time, it can meet the process tick processing in the real-time production process to the greatest extent, and further ensure the production efficiency. And the improvement of production quality reduces production risks and production costs.
  • an embodiment of the present invention also provides a process beat processing system, including:
  • the action data acquisition module is used to acquire the action data of each production action
  • the first beat information determining module is used to obtain the first beat information of each process beat according to the motion data
  • the verification module is used to verify the first beat information and determine the abnormal beat information
  • the matching module is used to match the abnormal beat information with the feature decision tree obtained by pre-training, determine the second beat information, and save the second beat information.
  • a process tick processing system can execute a process tick processing method provided in a method embodiment of the present invention, and can execute any combination of implementation steps of the method embodiments, and has corresponding functions and beneficial effects of the method.
  • an embodiment of the present invention also provides a process beat processing device, including:
  • At least one processor At least one processor
  • At least one memory for storing at least one program
  • the at least one processor When the at least one program is executed by the at least one processor, the at least one processor implements the process beat processing method.
  • a process tick processing device can execute a process tick processing method provided in a method embodiment of the present invention, can execute any combination of implementation steps of the method embodiments, and has corresponding functions and beneficial effects of the method.
  • the embodiment of the present invention also provides a computer-readable storage medium, in which instructions executable by the processor are stored, and the instructions executable by the processor are used to execute the process beat processing method described above when executed by the processor.
  • a computer-readable storage medium can execute a process tick processing method provided by the method embodiment of the present invention, can execute any combination of implementation steps of the method embodiment, and has the corresponding functions and beneficial effects of the method .
  • the embodiments of the present invention can be realized or implemented by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer-readable memory.
  • the method can be implemented in a computer program using standard programming techniques—including a non-transitory computer-readable storage medium configured with a computer program, where the storage medium so configured enables the computer to operate in a specific and predefined manner—according to the specific The methods and drawings described in the examples.
  • Each program can be implemented in a high-level process or object-oriented programming language to communicate with the computer system. However, if necessary, the program can be implemented in assembly or machine language. In any case, the language can be a compiled or interpreted language. In addition, the program can be run on a programmed application specific integrated circuit for this purpose.
  • the processes (or variants and/or combinations thereof) described herein can be executed under the control of one or more computer systems configured with executable instructions, and can be used as codes that are executed collectively on one or more processors (e.g., , Executable instructions, one or more computer programs, or one or more applications), implemented by hardware or a combination thereof.
  • the computer program includes a plurality of instructions executable by one or more processors.
  • the method can be implemented in any type of computing platform that is operably connected to a suitable computing platform, including but not limited to a personal computer, a mini computer, a main frame, a workstation, a network or a distributed computing environment, a separate or integrated computer Platform, or communication with charged particle tools or other imaging devices, etc.
  • a suitable computing platform including but not limited to a personal computer, a mini computer, a main frame, a workstation, a network or a distributed computing environment, a separate or integrated computer Platform, or communication with charged particle tools or other imaging devices, etc.
  • Aspects of the present invention can be implemented by machine-readable codes stored on non-transitory storage media or devices, whether removable or integrated into computing platforms, such as hard disks, optical reading and/or writing storage media, RAM, ROM, etc., so that they can be read by a programmable computer, and when the storage medium or device is read by the computer, it can be used to configure and operate the computer to perform the processes described herein.
  • machine-readable code or part of it, can be transmitted over a wired or wireless network.
  • a medium includes instructions or programs that implement the steps described above in combination with a microprocessor or other data processor
  • the invention described herein includes these and other different types of non-transitory computer-readable storage media.
  • the present invention also includes the computer itself.
  • a computer program can be applied to input data to perform the functions described herein, thereby converting the input data to generate output data that is stored in non-volatile memory.
  • the output information can also be applied to one or more output devices such as displays.
  • the converted data represents physical and tangible objects, including specific visual depictions of physical and tangible objects produced on the display.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Business, Economics & Management (AREA)
  • Manufacturing & Machinery (AREA)
  • Health & Medical Sciences (AREA)
  • Economics (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Marketing (AREA)
  • Primary Health Care (AREA)
  • Strategic Management (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种工艺节拍处理方法、系统、装置和存储介质,其中方法包括:获取各个生产动作的动作数据(S101);根据动作数据得到各个工艺节拍的第一节拍信息(S102);对第一节拍信息进行验证,确定异常节拍信息(S103);将异常节拍信息与预先训练得到的特征决策树进行匹配,确定第二节拍信息(S104)。该方法通过将异常节拍信息与特征决策树进行匹配,确定的第二节拍信息可以准确表示生产工艺中的工艺节拍,可适应各种情况下的工艺节拍统计,且大大提高了工艺节拍统计结果的准确性,便于后续根据工艺节拍统计结果对实际生产情况对出工艺调整和优化,确保了生产效率和生产质量的提高,降低了生产风险及生产成本。本方法可广泛应用于生产工艺技术领域。

Description

一种工艺节拍处理方法、系统、装置及存储介质 技术领域
本发明涉及生产工艺技术领域,尤其涉及一种工艺节拍处理方法、系统、装置及存储介质。
背景技术
当前,工业生产竞争越来越激烈,各厂家在提高生产效率与降低生产成本等方面进行技术及体制改革,提出了产线的非标需求,即从原来大规模地建厂建生产线,转变为在原有的生产线上进行生产模型修改,使同一个产线可以适应不同工艺生产的需求。同时,生产质量也是当前客户厂家追求的目标,当前市场价格竞争透明度高,使得厂家对产品的质量要求更高。
在生产过程中,通常把加工完成一道工序工艺的所有生产动作叫一个节拍,也称为一个Cycle(即工艺节拍)。每个工艺节拍可能由一个动作组或多个动作组组成,而且每个动作组由许多生产动作组成,工艺节拍数据的采集对后续进行统计分析从而提高生产效率和生产质量至关重要。在采集器上传生产数据的时候,存在网络中断的情况,那样采集器发送的生产数据不连续或不完整,使得对应的工艺节拍数据异常;此外,在生产过程中,配置的工艺工序也会存在着“转台”的情况,即在同一个工位下同一台机器设备,生产的A的某个工艺节拍的所有生产动作还没有完成,产线就转为生产B,该机器设备则需要立即执行生产B的工艺节拍,从而导致生产A的工艺节拍数据异常。在上述两种情况下,如果不对异常的节工艺进行检测并校正,就会得到错误的动作序列、动作数量和节拍时长。然而在现有技术中,并没有考虑到上述两种情况所带来的异常工艺节拍,从而导致得到的动作序列、动作数量、节拍时长等数据不准确,影响了后续的统计分析,而且,随着工艺发展,同一个产线适应不同工艺生产的情况越来越常见,若继续采用原有的工艺节拍处理方法,必定会大大地增加工艺节拍数据的异常比例,从而导致后续对生产工艺的统计分析与实际生产情况严重不符,使得厂家无法针对实际生产情况作出工艺调整和优化,不利于降低生产成本,限制了生产效率和生产质量的提高。
发明内容
为了解决上述技术问题,本发明的目的在于:提供一种工艺节拍处理方法、系统、装置及存储介质,可适应各种情况下的工艺节拍统计,且大大提高了工艺节拍统计结果的准确性, 满足了厂家提高生产质量和生产效率、以及降低生产成本的需求。
本发明所采用的第一技术方案是:
一种工艺节拍处理方法,包括以下步骤:
获取各个生产动作的动作数据;
根据所述动作数据得到各个工艺节拍的第一节拍信息;
对所述第一节拍信息进行验证,确定异常节拍信息;
将所述异常节拍信息与预先训练得到的特征决策树进行匹配,确定第二节拍信息。
进一步,所述获取各个生产动作的动作数据这一步骤,其包括:
实时采集生产数据并上传至物联网消息队列;
对所述物联网消息队列进行分析处理,得到各个生产动作的动作数据。
进一步,所述动作数据包括工位ID、动作ID、动作时长以及标志位参数,所述标志位参数包括1和0,所述标志位参数为1对应的生产动作是一个工艺节拍的开始动作或结束动作,所述标志位参数为0对应的生产动作是一个工艺节拍的中间动作。
进一步,所述根据所述动作数据得到各个工艺节拍的第一节拍信息这一步骤,其包括:
根据各所述生产动作创建若干个键值对,所述键值对的Key包括所述工位ID,所述键值对的Value包括所述动作ID和所述动作时长,将所述键值对缓存到Redis集群中;
确定所述生产动作的标志位参数为1,读取所述Redis集群中缓存的键值对,根据所述键值对的Value统计得到当前工艺节拍的第一节拍信息;
清空所述Redis集群;
其中,所述第一节拍信息包括第一动作序列、第一动作数量以及第一节拍时长。
进一步,所述对所述第一节拍信息进行验证,确定异常节拍信息这一步骤,包括以下至少之一:
确定所述第一动作数量超出第一阈值范围,获取所述第一动作数量对应的第一节拍信息作为异常节拍信息;
确定所述第一节拍时长超出第二阈值范围,获取所述第一节拍时长对应的第一节拍信息作为异常节拍信息;
其中,所述第一阈值范围为预设的工艺节拍动作数量的阈值范围,所述第二阈值范围为预设的工艺节拍节拍时长的阈值范围。
进一步,所述第二节拍信息包括第二动作序列、第二动作数量以及第二节拍时长,所述将所述异常节拍信息与预先训练得到的特征决策树进行匹配,确定第二节拍信息这一步骤, 其包括:
将所述异常节拍信息对应的第一动作序列与所述特征决策树中的预设动作序列进行相似度匹配,选取相似度高于预设阈值的预设动作序列作为第三动作序列;
根据所述第三动作序列在生产过程中的出现次数和最近出现时间,选取第二动作序列;
根据所述第二动作序列得到对应的第二动作数量,根据所述第二动作序列和预设动作时长得到第二节拍时长;
保存所述第二动作序列、所述第二动作数量以及所述第二节拍时长。
进一步,所述工艺节拍处理方法还包括训练特征决策树的步骤,其具体为:
获取预设时段内工艺节拍的动作序列作为训练数据集,通过机器学习训练得到所述特征决策树。
本发明所采用的第二技术方案是:
一种工艺节拍处理系统,包括:
动作数据获取模块,用于获取各个生产动作的动作数据;
第一节拍信息确定模块,用于根据所述动作数据得到各个工艺节拍的第一节拍信息;
验证模块,用于对所述第一节拍信息进行验证,确定异常节拍信息;
匹配模块,用于将所述异常节拍信息与预先训练得到的特征决策树进行匹配,确定第二节拍信息。
本发明所采用的第三技术方案是:
一种工艺节拍处理装置,包括:
至少一个处理器;
至少一个存储器,用于存储至少一个程序;
当所述至少一个程序被所述至少一个处理器执行,使得所述至少一个处理器实现所述工艺节拍处理方法。
本发明所采用的第四技术方案是:
一种计算机可读存储介质,其中存储有处理器可执行的指令,所述处理器可执行的指令在由处理器执行时用于执行所述工艺节拍处理方法。
本发明的有益效果是:本发明一种工艺节拍处理方法、系统、装置及存储介质,通过将异常节拍信息与特征决策树进行匹配,确定的第二节拍信息可以准确表示生产工艺中的工艺节拍,可适应各种情况下的工艺节拍统计,且大大提高了工艺节拍统计结果的准确性,便于后续根据工艺节拍统计结果对实际生产情况对出工艺调整和优化,确保了生产效率和生产质 量的提高,降低了生产风险及生产成本。
附图说明
图1为本发明实施例提供的工艺节拍处理方法的步骤流程图;
图2为本发明实施例提供的工艺节拍处理系统的结构框图;
图3为本发明实施例提供的工艺节拍处理装置的结构框图;
图4为本发明实施例提供的特征决策树示意图;
图5为本发明一较佳实施例提供的工艺节拍处理方法的流程示意图。
具体实施方式
下面结合附图和具体实施例对本发明做进一步的详细说明。对于以下实施例中的步骤编号,其仅为了便于阐述说明而设置,对步骤之间的顺序不做任何限定,实施例中的各步骤的执行顺序均可根据本领域技术人员的理解来进行适应性调整。
在本发明的描述中,多个的含义是两个以上,如果有描述到第一、第二只是用于区分技术特征为目的,而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量或者隐含指明所指示的技术特征的先后关系。此外,除非另有定义,本文所使用的所有的技术和科学术语与本技术领域的技术人员通常理解的含义相同。本文说明书中所使用的术语只是为了描述具体的实施例,而不是为了限制本发明。
参照图1,本发明实施例提供了一种工艺节拍处理方法,包括以下步骤:
S101、获取各个生产动作的动作数据。
具体地,动作数据包括该执行该生产动作的工位、该生产动作的时长以及该生产动作在一个工艺节拍中的顺序等信息,在工艺节拍的处理中,动作数据需要实时获取。上述步骤S101具体包括以下步骤:
S1011、实时采集生产数据并上传至物联网消息队列;
具体地,可通过采集器来实时采集生产过程中的生产数据,并按预设的时间上传生产数据到物联网消息队列中。
S1012、对物联网消息队列进行分析处理,得到各个生产动作的动作数据;
具体地,可通过数据解析程序分布式实时消费物联网消息队列中的生产数据,程序在读取到生产数据后,对生产数据进行整合、解析、转换、清洗等处理,以及对重要属性值进行验证处理,得到对应各个生产动作的动作数据。应当理解的是,上述S1011步骤中的生产数据仅为当前某个机器设备进行生产工艺的实时状态数据,并不能完整表示一个工艺节拍或是一个生产动作,也并不包含其执行某个生产动作的所需的时间、该生产动作在一个工艺节拍 中的执行顺序等信息,因此需要通过数据解析程序对一段时间内的所有生产数据进行整合、解析、转换、清洗等处理,从而得到用于表示该段时间内各个生产动作的时长和顺序的动作数据。
本发明实施例中,通过实时采集生产数据并上传至物联网消息队列进行分析处理,利用了大数据平台的分布式处理能力,能够快速、准确地得到动作数据,提高了工艺节拍的处理效率。
进一步作为可选的实施方式,动作数据包括工位ID、动作ID、动作时长以及标志位参数,标志位参数包括1和0,标志位参数为1对应的生产动作是一个工艺节拍的开始动作或结束动作,标志位参数为0对应的生产动作是一个工艺节拍的中间动作。
具体地,动作数据包括工位ID、动作ID、动作时长以及标志位参数等信息,工位ID表示执行该生产动作的工位,由于采集生产数据时可能同时对多个工位进行采集,因此需要用工位ID进行区分,一个工艺节拍中所有生产动作的工位ID一样;动作ID表示该生产动作的标号,可以预先设置对应每个生产动作的标号;动作时长表示执行该生产动作的耗时,动作时长需要实时获取;标志位参数用于判断一个工艺节拍的开始与结束,标志位参数可以是1或者0,标志位参数1表示一个工艺节拍的开始,标志位参数0对应的生产动作即为一个工艺节拍的中间动作,直到下一个标志位参数1,则表示一个工艺节拍结束,并且以此作为下一个工艺节拍的开始。例如,标志位参数组成的序列为1000000100010000000001,则该序列对应三个工艺节拍,分别可以用10000001、10001和10000000001表示。
本发明实施例中,通过标志位参数可以准确、简洁地表示一个工艺节拍包含的生产动作,便于后续统计工艺节拍的时长和动作序列等信息。
可选地,动作数据还可以包括动作组ID和线体ID。
具体地,一个工艺节拍可能包括一个或多个动作组,而每个动作组则包含多个生产动作,实际应用中,生产动作都是以动作组为单位进行,动作组ID则表示动作组的标号;线体值ID是前端BI报表的维度要素。
S102、根据动作数据得到各个工艺节拍的第一节拍信息。
具体地,第一节拍信息包括工艺节拍中生产动作的动作序列(即第一动作序列)、工艺节拍中生产动作的数量(即第一动作数量)以及工艺节拍中生产动作的总时长(即第一节拍时长),第一节拍信息可以根据统计工艺节拍中所有生产动作的动作数据得到。上述步骤S102具体包括以下步骤:
S1021、根据各生产动作创建若干个键值对,键值对的Key包括工位ID,键值对的Value 包括动作ID和动作时长,将键值对缓存到Redis集群中;
具体地,对于每一个生产动作,可将其工位ID作为Key值,将其动作时长和动作ID进行字符串相连作为Value值,从而得到对应每个生产动作的键值对,并缓存在Redis集群中。例如,某个生产动作的工位ID为1001,动作时长为20s,动作ID为A1,则对应的键值对的Key值可以表示为“1001”,Value值可以表示为“20,A1”。由于同一个工位下的所有生产动作的Key值一样,对应Value值可以作为一个集合,这样得到的键值对缓存在Redis集群中。
S1022、确定生产动作的标志位参数为1,读取Redis集群中缓存的键值对,根据键值对的Value统计得到当前工艺节拍的第一节拍信息;
具体地,当接收到某个动作数据的标志位参数为0,表示当前的生产动作为工艺节拍的中间动作,此时只缓存不读取;当接收到某个动作数据的标志位参数为1,则表示当前工艺节拍已经结束,在完成该动作数据对应的键值对的缓存后,读取Redis集群中缓存的所有的Value值,统计动作时长得到第一节拍时长,统计动作的数量得到第一动作数量,按照顺序将动作ID排序得到第一动作序列(如A1-A2-A4-A7-A8)。
S1023、清空Redis集群;
其中,第一节拍信息包括第一动作序列、第一动作数量以及第一节拍时长。
具体地,完成当前工艺节拍的统计后,需要清空Redis集群的缓存,并重复步骤S1021和步骤S1023完成下一个工艺节拍的统计,同时由于当前工艺节拍的结束动作对应于下一个工艺节拍的开始动作,因此实际应用中需要把该结束动作对应的键值对重新缓存。
本发明实施例中,结合分布式Redis集群实现工艺节拍的统计,可完成实时海量数据与硬盘IO的交换,提高了工艺节拍处理的效率以及系统的稳定性。
可选地,也可将工位ID和动作组ID进行字符串相连得到Key值,将动作时长、线体ID和动作ID进行字符串相连作为Value值,生成对应各个生产动作的键值对。例如,工位ID为1001,动作组ID为A003,那么Key值可以表示为“1001,A003”;动作时长为20s,线体ID为405,动作ID为A00301(前一部分与动作组IDA003一致),则Value值可以表示为“20,405,A00301”。这样适用于一个工艺节拍包含多个动作组的情况。
在下面实施例的描述中,将略去动作组ID和线体ID,应当理解的是,在一个工艺节拍包含多个动作组时,可对实施方式进一步细化作出种种变形,在下面的实施例中不再赘述。
可选地,第一节拍信息还包括工位ID,由于同一工艺节拍中生产动作都由同一工位执行,故工位ID均一致,第一节拍信息中的工位ID也与各个生产动作的工位ID一致。工位ID可 用于后续选取对应的特征决策树进行匹配。
S103、对第一节拍信息进行验证,确定异常节拍信息。
具体地,当采集器上传生产数据时如果网络发生波动甚至中断,或者生产过程中更换产线的生产产品,都会导致得到的第一节拍信息中的动作序列不完整,以及动作数量和节拍时长发生错误。因此需要根据预先设定的阈值范围对第一节拍信息进行验证,确定其是否为异常节拍信息。上述步骤S103可至少采用两种方案来实现,具体如下所示。
对于步骤S103的第一实施例,其具体为:
S1031、确定第一动作数量超出第一阈值范围,获取第一动作数量对应的第一节拍信息作为异常节拍信息;
具体地,第一阈值范围为预设的工艺节拍动作数量的阈值范围,当某一工艺节拍包含的动作数量不满足预设的阈值范围[C min,C max],即动作数量过少或过多,则表示该第一节拍信息为异常节拍信息。本发明实施例中,C min和C max是通过平均值算法计算出来,是生产的历史数据不断的计算积累完善的值,随着不断的数据统计计算,其取值越来越精确,在此不作详细说明。
对于步骤S103的另一实施例,其具体为:
S1032、确定第一节拍时长超出第二阈值范围,获取第一节拍时长对应的第一节拍信息作为异常节拍信息;
具体地,第二阈值范围为预设的工艺节拍节拍时长的阈值范围,当某一工艺节拍的节拍时长不满足预设的阈值范围[T min,T max],即节拍时长过短或过长,则表示该第一节拍信息为异常节拍信息。本发明实施例中,T min和T max同样是通过平均值算法计算出来的。
可选地,也可以综合上述两个验证条件进行验证,即如果上述两个阈值范围中的任何一个不满足,则该表示第一节拍信息为异常节拍信息。
本发明实施例中,通过动作数量和节拍时长的两个验证条件来对第一节拍信息进行验证,确保了可以筛选出异常的节拍信息,从而便于后续对该异常节拍信息进行处理得到准确的节拍信息。
可选地,还可增加“特征异常动作”字典表作为验证条件,该字典表中包括若干个不可能出现的相邻动作,如A1和A5这两个生产动作不可能相邻,则将“A1-A5”作为特征异常动作加入该字典表,若第一动作序列中出现了“A1-A5”这个序列片段,则表示对应的第一节拍信息为异常节拍信息。本发明实施例可以对异常节拍信息的验证条件进行完善,进一步提高了工艺节拍统计结果的准确性。
S104、将异常节拍信息与预先训练得到的特征决策树进行匹配,确定第二节拍信息。
具体地,如图4所示为本发明实施例某一工位下的特征决策树示意图,特征决策树的根节点用于表示工艺节拍的开始动作,中间节点用于表示工艺节拍的中间动作,叶子节点用于表示工艺节拍的结束动作,由此可知每个叶子节点都对应着一个预设动作序列。可将异常节拍信息中的动作序列与特征决策树进行匹配,找到相似度较高的预设动作序列,从而得到准确的第二节拍信息以替代异常的第一节拍信息,从而把异常工艺节拍数据转换成正常的工艺节拍数据推送显示及存储。第二节拍信息包括第二动作序列、第二动作数量以及第二节拍时长。上述步骤S104具体包括以下步骤:
S1041、将异常节拍信息对应的第一动作序列与特征决策树中的预设动作序列进行相似度匹配,选取相似度高于预设阈值的预设动作序列作为第三动作序列;
S1042、根据第三动作序列在生产过程中的出现次数和最近出现时间,选取第二动作序列;
例如,异常节拍信息的动作序列为“A1-A2-A4-A7-A8”,可根据其工位ID找到对应的特征决策树,然后在该特征决策树中按照相似度匹配到“A1-A2-A4-A7-A8-A9-A10-A12-14”和“A1-A2-A4-A7-A8-A10-A12-A13”这两个相似度较高的预设动作序列,再通过这两个预设动作序列在历史成产过程中的出现次数选择出现次数较多的做为第二动作序列;若出现次数一样多,还可以根据其最近出现时间选择出现时间较为接近当前时间的预设动作序列作为第二动作序列。
S1043、根据第二动作序列得到对应的第二动作数量,根据第二动作序列和预设动作时长得到第二节拍时长;
S1044、保存第二动作序列、第二动作数量以及第二节拍时长。
具体地,统计第二动作序列包含的生产动作的数量即可得到第二动作数量,根据第二动作序列中每个生产动作的预设动作长即可得到第二节拍时长。
本发明实施例中,通过将异常节拍信息与特征决策树进行匹配,确定的第二节拍信息可以准确表示生产工艺中的工艺节拍,可适应各种情况下的工艺节拍统计,且大大提高了工艺节拍统计结果的准确性,便于后续根据工艺节拍统计结果对实际生产情况对出工艺调整和优化,确保了生产效率和生产质量的提高,降低了生产风险及生产成本。
进一步作为可选的实施方式,工艺节拍处理方法还包括训练特征决策树的步骤,其具体为:
获取预设时段内工艺节拍的动作序列作为训练数据集,通过机器学习训练得到特征决策树。
具体地,可采用预设时段内正常生产过程的生产数据得到的动作序列作为训练数据集,通过机器学习训练得到特征决策树,还可采用实时的、经过验证不为异常节拍信息的第一节拍信息的动作序列作为训练数据集进行训练。得到的特征决策树可按照工位ID进行划分,每一个工位对应一个特征决策树,便于后续对异常节拍信息进行匹配。
应该理解的是,在实时的工艺节拍处理过程中,筛除异常节拍信息的第一节拍信息即可作为训练数据集,实时地对特征决策树进行训练,通过机器学习不断提高特征决策树中预设动作序列的准确性和可靠性。
可选地,还可以实时统计各个动作序列的出现次数和最近出现时间,并在特征决策树的叶子节点进行标注,便于选取最为准确的第二节拍信息。
如图5所示为本发明一较佳实施例提供的工艺节拍处理方法的流程示意图,本实施例中,通过采集器实时采集生产数据上传至物联网消息队列,利用大数据平台的分布式处理能力得到各个生产动作的动作数据,然后结合分布式Redis集群完成工艺节拍的统计,对工艺节拍进行验证后,正常的节拍信息直接输出,同时也作为训练数据训练特征决策树,异常的节拍信息通过特征决策树进行匹配并修正,从而得到准确的节拍信息。本实施例可进一步优化工艺节拍处理的流程,提高生产数据资源的利用率,同时由于特征决策树是实时变化的,可以最大限度的满足实时的生产过程中的工艺节拍处理,进一步确保了生产效率和生产质量的提高,降低了生产风险及生产成本。
参照图2,本发明实施例还提供了一种工艺节拍处理系统,包括:
动作数据获取模块,用于获取各个生产动作的动作数据;
第一节拍信息确定模块,用于根据动作数据得到各个工艺节拍的第一节拍信息;
验证模块,用于对第一节拍信息进行验证,确定异常节拍信息;
匹配模块,用于将异常节拍信息与预先训练得到的特征决策树进行匹配,确定第二节拍信息并保存第二节拍信息。
本发明实施例的一种工艺节拍处理系统,可执行本发明方法实施例所提供的一种工艺节拍处理方法,可执行方法实施例的任意组合实施步骤,具备该方法相应的功能和有益效果。
参照图3,本发明实施例还提供了一种工艺节拍处理装置,包括:
至少一个处理器;
至少一个存储器,用于存储至少一个程序;
当上述至少一个程序被上述至少一个处理器执行,使得上述至少一个处理器实现上述工艺节拍处理方法。
本发明实施例的一种工艺节拍处理装置,可执行本发明方法实施例所提供的一种工艺节拍处理方法,可执行方法实施例的任意组合实施步骤,具备该方法相应的功能和有益效果。
本发明实施例还提供了一种计算机可读存储介质,其中存储有处理器可执行的指令,上述处理器可执行的指令在由处理器执行时用于执行上述工艺节拍处理方法。
本发明实施例的一种计算机可读存储介质,可执行本发明方法实施例所提供的一种工艺节拍处理方法,可执行方法实施例的任意组合实施步骤,具备该方法相应的功能和有益效果。
应当认识到,本发明的实施例可以由计算机硬件、硬件和软件的组合、或者通过存储在非暂时性计算机可读存储器中的计算机指令来实现或实施。所述方法可以使用标准编程技术—包括配置有计算机程序的非暂时性计算机可读存储介质在计算机程序中实现,其中如此配置的存储介质使得计算机以特定和预定义的方式操作——根据在具体实施例中描述的方法和附图。每个程序可以以高级过程或面向对象的编程语言来实现以与计算机系统通信。然而,若需要,该程序可以以汇编或机器语言实现。在任何情况下,该语言可以是编译或解释的语言。此外,为此目的该程序能够在编程的专用集成电路上运行。
此外,可按任何合适的顺序来执行本文描述的过程的操作,除非本文另外指示或以其他方式明显地与上下文矛盾。本文描述的过程(或变型和/或其组合)可在配置有可执行指令的一个或多个计算机系统的控制下执行,并且可作为共同地在一个或多个处理器上执行的代码(例如,可执行指令、一个或多个计算机程序或一个或多个应用)、由硬件或其组合来实现。所述计算机程序包括可由一个或多个处理器执行的多个指令。
进一步,所述方法可以在可操作地连接至合适的任何类型的计算平台中实现,包括但不限于个人电脑、迷你计算机、主框架、工作站、网络或分布式计算环境、单独的或集成的计算机平台、或者与带电粒子工具或其它成像装置通信等等。本发明的各方面可以以存储在非暂时性存储介质或设备上的机器可读代码来实现,无论是可移动的还是集成至计算平台,如硬盘、光学读取和/或写入存储介质、RAM、ROM等,使得其可由可编程计算机读取,当存储介质或设备由计算机读取时可用于配置和操作计算机以执行在此所描述的过程。此外,机器可读代码,或其部分可以通过有线或无线网络传输。当此类媒体包括结合微处理器或其他数据处理器实现上文所述步骤的指令或程序时,本文所述的发明包括这些和其他不同类型的非暂时性计算机可读存储介质。当根据本发明所述的方法和技术编程时,本发明还包括计算机本身。
计算机程序能够应用于输入数据以执行本文所述的功能,从而转换输入数据以生成存储至非易失性存储器的输出数据。输出信息还可以应用于一个或多个输出设备如显示器。在本 发明优选的实施例中,转换的数据表示物理和有形的对象,包括显示器上产生的物理和有形对象的特定视觉描绘。
以上所述,只是本发明的较佳实施例而已,本发明并不局限于上述实施方式,只要其以相同的手段达到本发明的技术效果,凡在本发明的精神和原则之内,所做的任何修改、等同替换、改进等,均应包含在本发明保护的范围之内。在本发明的保护范围内其技术方案和/或实施方式可以有各种不同的修改和变化。

Claims (10)

  1. 一种工艺节拍处理方法,其特征在于,包括以下步骤:
    获取各个生产动作的动作数据;
    根据所述动作数据得到各个工艺节拍的第一节拍信息;
    对所述第一节拍信息进行验证,确定异常节拍信息;
    将所述异常节拍信息与预先训练得到的特征决策树进行匹配,确定第二节拍信息。
  2. 根据权利要求1所述的一种工艺节拍处理方法,其特征在于,所述获取各个生产动作的动作数据这一步骤,其包括:
    实时采集生产数据并上传至物联网消息队列;
    对所述物联网消息队列进行分析处理,得到各个生产动作的动作数据。
  3. 根据权利要求1所述的一种工艺节拍处理方法,其特征在于,所述动作数据包括工位ID、动作ID、动作时长以及标志位参数,所述标志位参数包括1和0,所述标志位参数为1对应的生产动作是一个工艺节拍的开始动作或结束动作,所述标志位参数为0对应的生产动作是一个工艺节拍的中间动作。
  4. 根据权利要求3所述的一种工艺节拍处理方法,其特征在于,所述根据所述动作数据得到各个工艺节拍的第一节拍信息这一步骤,其包括:
    根据各所述生产动作创建若干个键值对,所述键值对的Key包括所述工位ID,所述键值对的Value包括所述动作ID和所述动作时长,将所述键值对缓存到Redis集群中;
    确定所述生产动作的标志位参数为1,读取所述Redis集群中缓存的键值对,根据所述键值对的Value统计得到当前工艺节拍的第一节拍信息;
    清空所述Redis集群;
    其中,所述第一节拍信息包括第一动作序列、第一动作数量以及第一节拍时长。
  5. 根据权利要求4所述的一种工艺节拍处理方法,其特征在于,所述对所述第一节拍信息进行验证,确定异常节拍信息这一步骤,包括以下至少之一:
    确定所述第一动作数量超出第一阈值范围,获取所述第一动作数量对应的第一节拍信息作为异常节拍信息;
    确定所述第一节拍时长超出第二阈值范围,获取所述第一节拍时长对应的第一节拍信息作为异常节拍信息;
    其中,所述第一阈值范围为预设的工艺节拍动作数量的阈值范围,所述第二阈值范围为预设的工艺节拍节拍时长的阈值范围。
  6. 根据权利要求4所述的一种工艺节拍处理方法,其特征在于,所述第二节拍信息包括第 二动作序列、第二动作数量以及第二节拍时长,所述将所述异常节拍信息与预先训练得到的特征决策树进行匹配,确定第二节拍信息这一步骤,其包括:
    将所述异常节拍信息对应的第一动作序列与所述特征决策树中的预设动作序列进行相似度匹配,选取相似度高于预设阈值的预设动作序列作为第三动作序列;
    根据所述第三动作序列在生产过程中的出现次数和最近出现时间,选取第二动作序列;
    根据所述第二动作序列得到对应的第二动作数量,根据所述第二动作序列和预设动作时长得到第二节拍时长;
    保存所述第二动作序列、所述第二动作数量以及所述第二节拍时长。
  7. 根据权利要求1所述的一种工艺节拍处理方法,其特征在于,所述工艺节拍处理方法还包括训练特征决策树的步骤,其具体为:
    获取预设时段内工艺节拍的动作序列作为训练数据集,通过机器学习训练得到所述特征决策树。
  8. 一种工艺节拍处理系统,其特征在于,包括:
    动作数据获取模块,用于获取各个生产动作的动作数据;
    第一节拍信息确定模块,用于根据所述动作数据得到各个工艺节拍的第一节拍信息;
    验证模块,用于对所述第一节拍信息进行验证,确定异常节拍信息;
    匹配模块,用于将所述异常节拍信息与预先训练得到的特征决策树进行匹配,确定第二节拍信息。
  9. 一种工艺节拍处理装置,其特征在于,包括:
    至少一个处理器;
    至少一个存储器,用于存储至少一个程序;
    当所述至少一个程序被所述至少一个处理器执行,使得所述至少一个处理器实现权利要求1-7任一项所述的一种工艺节拍处理方法。
  10. 一种计算机可读存储介质,其中存储有处理器可执行的指令,其特征在于,所述处理器可执行的指令在由处理器执行时用于执行如权利要求1-7任一项所述的一种工艺节拍处理方法。
PCT/CN2020/116016 2020-06-24 2020-09-17 一种工艺节拍处理方法、系统、装置及存储介质 WO2021258564A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010591800.0A CN111860605B (zh) 2020-06-24 2020-06-24 一种工艺节拍处理方法、系统、装置及存储介质
CN202010591800.0 2020-06-24

Publications (1)

Publication Number Publication Date
WO2021258564A1 true WO2021258564A1 (zh) 2021-12-30

Family

ID=72988180

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/116016 WO2021258564A1 (zh) 2020-06-24 2020-09-17 一种工艺节拍处理方法、系统、装置及存储介质

Country Status (2)

Country Link
CN (1) CN111860605B (zh)
WO (1) WO2021258564A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114637264A (zh) * 2022-03-15 2022-06-17 广州明珞装备股份有限公司 动作稳定性分析方法、系统、设备及存储介质
WO2023125123A1 (zh) * 2021-12-31 2023-07-06 广州明珞装备股份有限公司 一种工艺节拍计算方法、装置、设备及存储介质
CN117991741A (zh) * 2024-04-07 2024-05-07 深圳和润达科技有限公司 基于Redis的工艺流程控制方法及装置

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112288175B (zh) * 2020-11-02 2024-06-18 联通(浙江)产业互联网有限公司 产线实时优化方法及装置
CN112352801B (zh) * 2020-11-06 2022-04-29 广东智源机器人科技有限公司 多工位扒炉上料控制方法、装置、控制设备和存储介质
CN113189935B (zh) * 2021-04-29 2022-09-30 浙江陀曼云计算有限公司 基于时序功率数据的非侵入式生产节拍预测方法及系统
CN113450125A (zh) * 2021-07-06 2021-09-28 北京市商汤科技开发有限公司 可溯源生产数据的生成方法、装置、电子设备及存储介质
CN114227701B (zh) * 2022-02-25 2022-05-10 科大智能物联技术股份有限公司 一种基于生产数据的机器人故障预测方法
CN114594745A (zh) * 2022-03-15 2022-06-07 广州明珞装备股份有限公司 瓶颈工位分析方法、分析设备及计算机存储介质
CN114625088B (zh) * 2022-03-15 2024-02-02 广州明珞装备股份有限公司 自动化生产线的效能优化方法、系统、设备及存储介质
CN114625087A (zh) * 2022-03-15 2022-06-14 广州明珞装备股份有限公司 工位异常动作分析方法、系统、设备及存储介质
CN114706361B (zh) * 2022-06-07 2022-09-02 工业富联(杭州)数据科技有限公司 离散制造生产系统修复方法、计算机装置及存储介质
CN115600930B (zh) * 2022-11-29 2023-03-28 天津佰利金电力有限公司 用于电力金具生产数据的智能管理方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110288667A1 (en) * 2009-02-12 2011-11-24 Kyoto University Industrial robot system
CN106990733A (zh) * 2017-04-24 2017-07-28 华南理工大学 一种支持工业大数据分析的装备控制器及运行方法
CN107301499A (zh) * 2017-05-27 2017-10-27 天津大学 一种基于ami数据的配电馈线统计线损率数据清洗方法
CN109657945A (zh) * 2018-12-06 2019-04-19 华中科技大学 一种基于数据驱动的工业生产过程故障诊断方法
CN110245374A (zh) * 2019-04-26 2019-09-17 广州明珞汽车装备有限公司 一种工艺时序间隔分析方法、系统、装置和存储介质

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2001282344A (ja) * 2000-03-31 2001-10-12 Nec Corp タクト生産の生産計画立案方法及びその生産計画立案システム
US6631305B2 (en) * 2000-05-03 2003-10-07 General Electric Company Capability analysis of assembly line production
CN106295840A (zh) * 2015-06-09 2017-01-04 北京福田康明斯发动机有限公司 一种优化生产流水线作业的方法、装置及生产流水线
CN111178660A (zh) * 2019-10-31 2020-05-19 青岛海尔工业智能研究院有限公司 混合产品生产线中生产节拍确定方法、装置及电子设备
CN111008905A (zh) * 2019-12-25 2020-04-14 浙江力太工业互联网有限公司 一种生产设备节拍高精度测量的方法及装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110288667A1 (en) * 2009-02-12 2011-11-24 Kyoto University Industrial robot system
CN106990733A (zh) * 2017-04-24 2017-07-28 华南理工大学 一种支持工业大数据分析的装备控制器及运行方法
CN107301499A (zh) * 2017-05-27 2017-10-27 天津大学 一种基于ami数据的配电馈线统计线损率数据清洗方法
CN109657945A (zh) * 2018-12-06 2019-04-19 华中科技大学 一种基于数据驱动的工业生产过程故障诊断方法
CN110245374A (zh) * 2019-04-26 2019-09-17 广州明珞汽车装备有限公司 一种工艺时序间隔分析方法、系统、装置和存储介质

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2023125123A1 (zh) * 2021-12-31 2023-07-06 广州明珞装备股份有限公司 一种工艺节拍计算方法、装置、设备及存储介质
CN114637264A (zh) * 2022-03-15 2022-06-17 广州明珞装备股份有限公司 动作稳定性分析方法、系统、设备及存储介质
CN114637264B (zh) * 2022-03-15 2024-05-17 广州明珞装备股份有限公司 动作稳定性分析方法、系统、设备及存储介质
CN117991741A (zh) * 2024-04-07 2024-05-07 深圳和润达科技有限公司 基于Redis的工艺流程控制方法及装置
CN117991741B (zh) * 2024-04-07 2024-06-07 深圳和润达科技有限公司 基于Redis的工艺流程控制方法及装置

Also Published As

Publication number Publication date
CN111860605B (zh) 2022-12-13
CN111860605A (zh) 2020-10-30

Similar Documents

Publication Publication Date Title
WO2021258564A1 (zh) 一种工艺节拍处理方法、系统、装置及存储介质
WO2021179572A1 (zh) 运维系统异常指标检测模型优化方法、装置及存储介质
CN112131383B (zh) 特定目标的情感极性分类方法
CN108648765B (zh) 一种语音异常检测的方法、装置及终端
CN107070940B (zh) 一种从流式登录日志中判断恶意登录ip地址的方法及装置
CN108960269B (zh) 数据集的特征获取方法、装置及计算设备
CN106709509B (zh) 一种基于时间序列特殊点的卫星遥测数据聚类方法
WO2024119746A1 (zh) 一种数据存储的方法、装置、电子设备以及存储介质
WO2024087428A1 (zh) 一种存储产品参数配置推荐方法、装置、设备及介质
CN110059126B (zh) 基于lkj异常值数据的复杂关联网络分析方法及系统
CN113448761B (zh) 一种根因定位方法及装置
CN111353860A (zh) 产品信息推送方法及系统
CN115345600B (zh) 一种rpa流程的生成方法和装置
CN108763289B (zh) 一种海量异构传感器格式数据的解析方法
CN115860141A (zh) 一种自动化机器学习交互式黑箱可视建模方法及系统
CN112732549B (zh) 基于聚类分析的测试程序分类方法
CN108491328A (zh) 一种基线测试报告自动化处理系统及方法
CN112257215B (zh) 一种产品寿命分布参数的极大似然估计求解方法及系统
TW202312042A (zh) 診斷模型的自動優化方法及自動優化系統
CN112861128A (zh) 一种批量识别机器账号的方法及系统
CN112463643A (zh) 一种软件质量预测方法
WO2020232899A1 (zh) 数据分析系统会诊方法及相关装置
CN107437001B (zh) 信息序列转换为向量化数据的方法及系统
CN109669932A (zh) 一种基于次序依赖的电力数据集修复方法
CN112244855B (zh) 一种面向心电数据的数据处理系统

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: 20941688

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: 20941688

Country of ref document: EP

Kind code of ref document: A1