CN107944491A - Mass property symbolism maps control figure construction method - Google Patents
Mass property symbolism maps control figure construction method Download PDFInfo
- Publication number
- CN107944491A CN107944491A CN201711188848.1A CN201711188848A CN107944491A CN 107944491 A CN107944491 A CN 107944491A CN 201711188848 A CN201711188848 A CN 201711188848A CN 107944491 A CN107944491 A CN 107944491A
- Authority
- CN
- China
- Prior art keywords
- msubsup
- sequence
- mass property
- influence factor
- quality characteristic
- Prior art date
- Legal status (The legal status 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 status listed.)
- Granted
Links
- 238000010276 construction Methods 0.000 title claims abstract description 9
- 238000013507 mapping Methods 0.000 claims abstract description 48
- 108090000623 proteins and genes Proteins 0.000 claims description 34
- 238000005070 sampling Methods 0.000 claims description 13
- 238000004458 analytical method Methods 0.000 claims description 4
- 230000002265 prevention Effects 0.000 abstract description 3
- 230000002159 abnormal effect Effects 0.000 description 18
- 238000000034 method Methods 0.000 description 18
- 238000010586 diagram Methods 0.000 description 11
- 238000003745 diagnosis Methods 0.000 description 5
- 238000004519 manufacturing process Methods 0.000 description 5
- 238000000513 principal component analysis Methods 0.000 description 4
- 239000000047 product Substances 0.000 description 4
- 208000024891 symptom Diseases 0.000 description 3
- 230000005856 abnormality Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000002301 combined effect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000003908 quality control method Methods 0.000 description 2
- 108010014172 Factor V Proteins 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000010835 comparative analysis Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
- G06F16/9038—Presentation of query results
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- General Factory Administration (AREA)
- Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)
Abstract
Description
技术领域technical field
本发明属于质量诊断与预防技术领域，尤其涉及一种质量特性符号化映射控制图构建方法。The invention belongs to the technical field of quality diagnosis and prevention, and in particular relates to a method for constructing a quality characteristic symbolic mapping control chart.
背景技术Background technique
从数学角度看，质量诊断和预控的过程实际上就是把症状空间的向量映射到故障源空间，即实现空间X(症状空间)到空间Y(故障源)的映射F(映射关系)。映射关系F是未知的，质量诊断的实质也就是综合各种知识和方法，找出这种映射关系F，进而应用这种关系。在以后发生质量问题时能快速找出问题根源，以便实时控制；From a mathematical point of view, the process of quality diagnosis and pre-control is actually to map the vector of symptom space to the fault source space, that is, to realize the mapping F (mapping relationship) from space X (symptom space) to space Y (fault source). The mapping relationship F is unknown, and the essence of quality diagnosis is to synthesize various knowledge and methods to find out this mapping relationship F, and then apply this relationship. When quality problems occur in the future, the root cause of the problem can be quickly found for real-time control;
传统的控制图只能通过质量特性数据随时间的变化的情况，来判断生成过程是否处于稳态，但是无法完成对生产过程出现异常波动的诊断与溯源；传统控制图只能描述质量特性数据的波动规律，无法描述波动规律与波动源的关联关系。在实际应用过程中，只能靠工程技术人员的经验知识与主观判断出现异常波动的故障源及原因。The traditional control chart can only judge whether the production process is in a steady state through the change of quality characteristic data over time, but it cannot complete the diagnosis and traceability of abnormal fluctuations in the production process; the traditional control chart can only describe the quality characteristic data. Fluctuation law, unable to describe the relationship between fluctuation law and fluctuation source. In the actual application process, the source and cause of abnormal fluctuations can only be judged by the experience knowledge and subjective judgment of engineers and technicians.
发明内容Contents of the invention
本发明的发明目的是：为了解决现有技术中存在的以上问题，本发明提出了一种质量特性符号化映射控制图构建方法，避免人为主观判断造成的误差，实现对质量特性影响因素的预测。The purpose of the present invention is: in order to solve the above problems in the prior art, the present invention proposes a method for constructing a quality characteristic symbolic mapping control chart, avoiding errors caused by human subjective judgments, and realizing the prediction of factors affecting quality characteristics .
本发明的技术方案是：一种质量特性符号化映射控制图构建方法，包括以下步骤：The technical solution of the present invention is: a method for constructing a quality characteristic symbolic mapping control chart, comprising the following steps:
S1、将休哈特稳态区域及上下控制限外区域等分为多个区域，对每一个等分区域采用多元组符号序列进行表示，构建质量特性值符号化序列，形成基因序列单链；S1. Divide the Shewhart steady-state region and the region outside the upper and lower control limits into multiple regions, and use a multi-group symbol sequence to represent each equal region, construct a symbolic sequence of quality characteristic values, and form a single chain of gene sequence;
S2、对每个时间域内质量特性影响因素在每个采样时刻下的观测值进行主元分析，并建立质量特性值与其影响因素的映射关系，根据步骤S1中构建的质量特性值符号化序列得到质量特性影响因素序列，再对每个质量特性影响因素进行归一化处理，构建质量特性影响因素符号化序列，形成基因序列单链；S2. Perform principal component analysis on the observed values of the quality characteristic influencing factors in each time domain at each sampling moment, and establish the mapping relationship between the quality characteristic value and its influencing factors, and obtain according to the quality characteristic value symbolic sequence constructed in step S1 Sequence of influencing factors of quality characteristics, and then normalize each influencing factor of quality characteristics, construct a symbolic sequence of influencing factors of quality characteristics, and form a single chain of gene sequence;
S3、将质量特性值符号化序列与质量特性影响因素符号化序列采用关联双链基因模式进行组合，构建质量特性符号化映射控制图。S3. Combining the symbolized sequence of quality characteristic values and the symbolized sequence of quality characteristic influencing factors using the associated double-stranded gene pattern to construct a quality characteristic symbolized mapping control chart.
进一步地，所述步骤S1将休哈特稳态区域及上下控制限外区域等分为多个区域，对每一个等分区域采用多元组符号序列进行表示，构建质量特性值符号化序列，形成基因序列单链，具体包括以下分步骤：Further, the step S1 divides the Shewhart steady-state region and the region outside the upper and lower control limits into multiple regions equally, and uses a multi-group symbol sequence to represent each equally divided region, and constructs a symbolic sequence of quality characteristic values, forming Single-stranded gene sequence, specifically including the following sub-steps:
S11、将休哈特稳态区域及上下控制限外区域，以中心线μ线为起点，等分为八个区域，表示为{(-∞,μ-3σ),[μ-3σ,μ-2σ],(μ-2σ,μ-σ],(μ-σ,μ],(μ,μ+σ],(μ+σ,μ+2σ]，(μ+2σ,μ+3σ],(μ+3σ,+∞)}，其中休哈特稳态区域为μ±3σ；S11. Divide the Shewhart steady-state area and the area outside the upper and lower control limits into eight equal areas starting from the centerline μ line, expressed as {(-∞, μ-3σ), [μ-3σ, μ - 2σ],(μ-2σ,μ-σ],(μ-σ,μ],(μ,μ+σ],(μ+σ,μ+2σ],(μ+2σ,μ+3σ],( μ+3σ,+∞)}, where the Shewhart steady-state region is μ±3σ;
S12、将步骤S11中的八个等分区域分别采用符号{D,C,B,A,a,b,c,d}八元组进行表示；S12. The eight equally divided areas in step S11 are respectively represented by symbols {D, C, B, A, a, b, c, d} octet;
S13、将各个时间域内不同采样时刻的质量特性数据点表示为符号序列QD＝{xi}＝{…,…,A,B,C,a,b,c,D,d,…,…}，得到质量特性值符号化序列，形成基因序列单链。S13. Express the quality characteristic data points at different sampling moments in each time domain as a symbol sequence Q D ={ xi }={...,...,A,B,C,a,b,c,D,d,...,... }, to obtain the symbolized sequence of quality characteristic values, and form a single-chain gene sequence.
进一步地，所述步骤S2对每个时间域内质量特性影响因素在每个采样时刻下的观测值进行主元分析，并建立质量特性值与其影响因素的映射关系，根据步骤S1中构建的质量特性值符号化序列得到质量特性影响因素序列，再对每个质量特性影响因素进行归一化处理，构建质量特性影响因素符号化序列，形成基因序列单链，具体包括以下分步骤：Further, the step S2 performs principal component analysis on the observed values of the quality characteristic influencing factors in each time domain at each sampling moment, and establishes a mapping relationship between the quality characteristic value and its influencing factors. According to the quality characteristic constructed in step S1 Value symbolized sequence to obtain the sequence of quality characteristic influencing factors, and then normalize each quality characteristic influencing factor, construct the symbolized sequence of quality characteristic influencing factors, and form a gene sequence single chain, which specifically includes the following sub-steps:
S21、根据每个时间域内质量特性影响因素在每个采样时刻下的观测值，提取影响因素主元信息，将质量特性影响因素主元序列表示为{V1,V2,V3,…VB}，其中质量特性影响因素表示为[V1,V2,V3,…VN]，观测值表示为[x1,…,xN]；S21. According to the observed value of the influencing factors of quality characteristics in each time domain at each sampling moment, extract the pivotal component information of influencing factors, and express the pivotal element sequence of influencing factors of quality characteristics as {V 1 , V 2 , V 3 ,…V B }, where the factors affecting quality characteristics are expressed as [V 1 ,V 2 ,V 3 ,…V N ], and the observed values are expressed as [x 1 ,…,x N ];
S22、对一个质量特性值与其影响因素建立一个映射再对一个质量特性值与其影响因素主元建立一个映射根据步骤S1中构建的质量特性值符号化序列得到质量特性影响因素序列：S22. Establish a mapping between a quality characteristic value and its influencing factors Then establish a mapping between a quality characteristic value and its influencing factor pivot According to the symbolized sequence of quality characteristic values constructed in step S1, the sequence of influencing factors of quality characteristics is obtained:
得到质量特性影响因素主元序列：Get the pivot sequence of factors affecting quality characteristics:
S23、将不同质量特性影响因素在不同时刻所造成的影响划分等级，对每个质量特性影响因素进行归一化处理并将归一化处理后的分值划分为10个等级[0-9]，通过设定稳定状态等级阈值，将分值处于稳定区域的质量特性影响因素状态表示为0，将分值处于不稳定区域的质量特性影响因素状态表示为1，从而将质量特性影响因素状态空间SNF表示为：S23. Classify the impact of different quality characteristic influencing factors at different times, perform normalization processing on each quality characteristic influencing factor, and divide the normalized scores into 10 grades [0-9] , by setting the stable state level threshold, the state of the quality characteristic influencing factors whose score is in the stable region is expressed as 0, and the state of the quality characteristic influencing factors whose score is in the unstable region is expressed as 1, so that the quality characteristic influencing factor state space SNF is expressed as:
将质量特性影响因素状态空间SKF表示为：The state space S KF of quality characteristic influencing factors is expressed as:
得到质量特性影响因素符号化序列，形成基因序列单链。Obtain the symbolized sequence of factors affecting quality characteristics, and form a single strand of gene sequence.
进一步地，所述步骤S3将质量特性值符号化序列与质量特性影响因素符号化序列采用关联双链基因模式进行组合，构建质量特性符号化映射控制图，具体为：根据步骤S2中建立的质量特性值与其影响因素的映射关系，将步骤S1中形成的基因序列单链与步骤S2中形成的基因序列单链，采用关联双链基因模式组合为双链基因序列，从而得到质量特性符号化映射控制图。Further, the step S3 combines the symbolized sequence of quality characteristic values and the symbolized sequence of quality characteristic influencing factors using the associated double-stranded gene pattern to construct a quality characteristic symbolized mapping control chart, specifically: according to the quality characteristic established in step S2 The mapping relationship between the characteristic value and its influencing factors is to combine the single-stranded gene sequence formed in step S1 and the single-stranded gene sequence formed in step S2 into a double-stranded gene sequence using the associated double-stranded gene pattern, so as to obtain the symbolic mapping of quality characteristics Control Charts.
本发明的有益效果是：本发明的质量特性符号化映射控制图通过建立质量特性值与其影响因素的映射关系，即建立了质量特性数据波动规律与引起波动的影响因素之间的映射关系，一方面能反映质量特性数据的波动规律，另一方面能通过映射关系提前发现故障源，从而人为主观判断造成的误差，实现对质量特性的影响因素进行预测和调整，使得质量特性处于有效稳态，达到预防控制的目的。The beneficial effects of the present invention are: the quality characteristic symbolic mapping control chart of the present invention establishes the mapping relationship between the quality characteristic value and its influencing factors, that is, establishes the mapping relationship between the fluctuation law of the quality characteristic data and the influencing factors causing the fluctuation. On the one hand, it can reflect the fluctuation law of quality characteristic data. On the other hand, it can find the fault source in advance through the mapping relationship, so that the error caused by human subjective judgment can realize the prediction and adjustment of the influencing factors of quality characteristic, so that the quality characteristic is in an effective steady state. To achieve the purpose of prevention and control.
附图说明Description of drawings
图1是本发明的质量特性符号化映射控制图构建方法流程示意图。Fig. 1 is a schematic flow chart of the method for constructing a quality characteristic symbolic mapping control chart according to the present invention.
图2是本发明的质量特性值符号化序列构建示意图。Fig. 2 is a schematic diagram of the construction of the symbolized sequence of quality characteristic values in the present invention.
图3是本发明的质量特性影响因素主元提取与映射结构基础构建示意图。Fig. 3 is a schematic diagram of principal component extraction and mapping structure foundation construction of quality characteristic influencing factors in the present invention.
图4是本发明的质量特性影响因素符号化序列构建示意图。Fig. 4 is a schematic diagram of constructing a symbolized sequence of quality characteristic influencing factors in the present invention.
图5是本发明的质量特性符号化控制图结构示意图。Fig. 5 is a schematic structural diagram of the quality characteristic symbolized control diagram of the present invention.
具体实施方式Detailed ways
为了使本发明的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本发明进行进一步详细说明。应当理解，此处所描述的具体实施例仅用以解释本发明，并不用于限定本发明。In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.
如图1所示，为本发明的质量特性符号化映射控制图构建方法流程示意图。一种质量特性符号化映射控制图构建方法，包括以下步骤：As shown in FIG. 1 , it is a schematic flow chart of the method for constructing the quality characteristic symbolic mapping control chart of the present invention. A method for constructing a quality characteristic symbolic mapping control chart, comprising the following steps:
S1、将休哈特稳态区域及上下控制限外区域等分为多个区域，对每一个等分区域采用多元组符号序列进行表示，构建质量特性值符号化序列，形成基因序列单链；S1. Divide the Shewhart steady-state region and the region outside the upper and lower control limits into multiple regions, and use a multi-group symbol sequence to represent each equal region, construct a symbolic sequence of quality characteristic values, and form a single chain of gene sequence;
S2、对每个时间域内质量特性影响因素在每个采样时刻下的观测值进行主元分析，并建立质量特性值与其影响因素的映射关系，根据步骤S1中构建的质量特性值符号化序列得到质量特性影响因素序列，再对每个质量特性影响因素进行归一化处理，构建质量特性影响因素符号化序列，形成基因序列单链；S2. Perform principal component analysis on the observed values of the quality characteristic influencing factors in each time domain at each sampling moment, and establish the mapping relationship between the quality characteristic value and its influencing factors, and obtain according to the quality characteristic value symbolic sequence constructed in step S1 Sequence of influencing factors of quality characteristics, and then normalize each influencing factor of quality characteristics, construct a symbolic sequence of influencing factors of quality characteristics, and form a single chain of gene sequence;
S3、将质量特性值符号化序列与质量特性影响因素符号化序列采用关联双链基因模式进行组合，构建质量特性符号化映射控制图。S3. Combining the symbolized sequence of quality characteristic values and the symbolized sequence of quality characteristic influencing factors using the associated double-stranded gene pattern to construct a quality characteristic symbolized mapping control chart.
在步骤S1中，休哈特控制图把过程稳定受控状态质量特性数据点集中在μ±3σ区域，若99.73％的数据点落入此区域则说明过程受控。根据数据点是否落入本区域判断有效控制的准则是模糊的，但是控制图能够通过数据点位置的变化反映过程中的任何变化。控制图中数据点的位置分布状况所出现的趋势、链状、周期和超界等，能够说明过程中是否出现异常。In step S1, the Shewhart control diagram concentrates the data points of the quality characteristics in the stable and controlled state of the process in the μ±3σ region, and if 99.73% of the data points fall into this region, it means that the process is under control. The criterion for judging effective control based on whether a data point falls into this region is ambiguous, but the control chart can reflect any change in the process through the change of the position of the data point. The trends, chains, cycles, and out-of-bounds of the position distribution of data points in the control chart can indicate whether there is an abnormality in the process.
将休哈特稳态区域及上下控制限外区域等分为多个区域，对每一个等分区域采用多元组符号序列进行表示，构建质量特性值符号化序列，形成基因序列单链，具体包括以下分步骤：Divide the Shewhart steady-state region and the region outside the upper and lower control limits into multiple regions, and use a multigroup symbol sequence to represent each equal region, construct a symbolic sequence of quality characteristic values, and form a single chain of gene sequence, specifically including The following sub-steps:
S11、将休哈特稳态区域及上下控制限外区域，以中心线μ线为起点，等分为八个区域，表示为{(-∞,μ-3σ),[μ-3σ,μ-2σ],(μ-2σ,μ-σ],(μ-σ,μ],(μ,μ+σ],(μ+σ,μ+2σ]，(μ+2σ,μ+3σ],(μ+3σ,+∞)}，其中休哈特稳态区域为μ±3σ，μ为质量特性值的数学期望，σ为质量特性值的方差；S11. Divide the Shewhart steady-state area and the area outside the upper and lower control limits into eight equal areas starting from the centerline μ line, expressed as {(-∞, μ-3σ), [μ-3σ, μ- 2σ],(μ-2σ,μ-σ],(μ-σ,μ],(μ,μ+σ],(μ+σ,μ+2σ],(μ+2σ,μ+3σ],( μ+3σ,+∞)}, where the Shewhart steady-state region is μ±3σ, μ is the mathematical expectation of the quality characteristic value, and σ is the variance of the quality characteristic value;
S12、将步骤S11中的八个等分区域分别采用符号{D,C,B,A,a,b,c,d}八元组进行表示；S12. The eight equally divided areas in step S11 are respectively represented by symbols {D, C, B, A, a, b, c, d} octet;
S12、将各个时间域内不同采样时刻的质量特性数据点表示为符号序列QD＝{xi}＝{…,…,A,B,C,a,b,c,D,d,…,…}，得到质量特性值符号化序列，形成基因序列单链。S12. Express the quality characteristic data points at different sampling moments in each time domain as a symbol sequence Q D ={ xi }={...,...,A,B,C,a,b,c,D,d,...,... }, to obtain the symbolized sequence of quality characteristic values, and form a single-chain gene sequence.
符号序列QD反映各个时间域内不同采样时刻的质量特性数据点的变化过程。例如，当连续出现字符序列{…,c,b,a,A,B,C,…}时，则表示出现上升趋势异常；当出现字符序列{…,A,A,B,B,C,C,…}时，则表示发生中心一侧上升趋势异常；当出现字符序列{…,C,B,A,a,b,c,…}时，则表示下降趋势异常；当出现字符序列{…,a,a,b,b,c,c,…}时，则表示发生中心一侧下降趋势异常；当字符序列中出现{…,D,…}或{…,d,…}时，则表示超界异常；当连续出现字符序列The symbol sequence Q D reflects the change process of the quality characteristic data points at different sampling moments in each time domain. For example, when the character sequence {…,c,b,a,A,B,C,…} appears continuously, it means that there is an abnormal upward trend; when the character sequence {…,A,A,B,B,C, C,…}, it means that the upward trend of the center side is abnormal; when the character sequence {…,C,B,A,a,b,c,…} appears, it means that the downward trend is abnormal; when the character sequence { …,a,a,b,b,c,c,…}, it means that the downward trend on the center side is abnormal; when {…,D,…} or {…,d,…} appear in the character sequence, It means an out-of-bounds exception; when a sequence of characters appears consecutively
等在一定时间内间隔相同的上升或下降的重复字符序列时，则表示出现周期症状；当连续出现超过12个字符序列{…,A,a,A,a,A,a,…}在中心线集中时，则表示紧中心异常。When repeated character sequences with the same rising or falling intervals within a certain period of time, it indicates periodic symptoms; when more than 12 character sequences {…,A,a,A,a,A,a,…} appear in the center When the lines are concentrated, it indicates a tight center anomaly.
本发明的字符序列特征变化能够反映数据的过程变化。不同局部特征的字符序列说明不同的异常模式的存在，这个远远大于传统八大异常模式特征，可以有利于通过信息识别技术来扩展更实际更具体的异常模式。传统的八大异常模式只给出了基本的异常判别准则，但是更多的异常规律需要人为判断。本发明以传统异常模式为基础，通过字符序列特征规律并结合故障信息，运用字符化的信息识别技术可以完善和补充更多的异常模式库，更有利于质量特性的诊断和预控。The characteristic change of the character sequence in the present invention can reflect the process change of the data. The character sequences of different local features indicate the existence of different abnormal patterns, which are far greater than the traditional eight abnormal pattern characteristics, and can be beneficial to expand more practical and specific abnormal patterns through information recognition technology. The traditional eight anomaly models only give basic anomaly discrimination criteria, but more abnormal laws require human judgment. Based on the traditional abnormal pattern, the present invention can improve and supplement more abnormal pattern databases by using character sequence feature rules combined with fault information and using characterized information recognition technology, which is more conducive to the diagnosis and pre-control of quality characteristics.
如图2所示，为本发明的质量特性值符号化序列构建示意图。其中，1,2...I分别表示时间域，第一序列为质量特性值序列，进行质量特性值序列符号化形成基因序列单链。转换的质量特性字符序列蕴含了已知和未知的质量特性数据异常模式，随着质量特性的演化操作，集合整个时间域内的质量特性数据字符序列构成质量特性海量数据信息库。As shown in FIG. 2 , a schematic diagram is constructed for the quality characteristic value symbolization sequence of the present invention. Among them, 1, 2...I represent the time domain respectively, and the first sequence is the quality characteristic value sequence, and the quality characteristic value sequence is symbolized to form a single chain of gene sequence. The converted character sequence of quality characteristics contains known and unknown abnormal patterns of quality characteristic data. With the evolution operation of quality characteristics, the quality characteristic data character sequence in the entire time domain is collected to form a mass characteristic data database.
在步骤S2中，影响质量特性数据波动规律的要素就是制造过程中的质量因素，这些因素构成了质量特性波动源信息集合。质量特性变异的结果与质量影响因素的表现形式之间具有一定的规律性。如果制造过程中质量影响因素相似(即相同级别的操作人员、检验人员；相同类型及经济精度的设备系统；相同加工难度的工件；相同经济精度的加工方法及同类型的质量特征值；相同检测误差的检测方法；相近的加工环境等)，则加工后的质量特性数据也应该满足同类型的分布规律。这奠定了数据波动到要素状态变化映射关系的基础。In step S2, the elements that affect the fluctuation law of the quality characteristic data are the quality factors in the manufacturing process, and these factors constitute the information set of the quality characteristic fluctuation source. There is a certain regularity between the results of quality characteristic variation and the manifestations of quality influencing factors. If the quality influencing factors in the manufacturing process are similar (that is, operators and inspectors of the same level; equipment systems of the same type and economic precision; workpieces of the same processing difficulty; processing methods of the same economic precision and quality characteristic values of the same type; Error detection method; similar processing environment, etc.), the quality characteristic data after processing should also meet the same type of distribution law. This lays the foundation for the mapping relationship between data fluctuations and feature state changes.
所述步骤S2对每个时间域内质量特性影响因素在每个采样时刻下的观测值进行主元分析，并建立质量特性值与其影响因素的映射关系，根据步骤S1中构建的质量特性值符号化序列得到质量特性影响因素序列，再对每个质量特性影响因素进行归一化处理，构建质量特性影响因素符号化序列，形成基因序列单链，具体包括以下分步骤：The step S2 performs principal component analysis on the observed values of the quality characteristic influencing factors in each time domain at each sampling moment, and establishes a mapping relationship between the quality characteristic value and its influencing factors, and symbolizes the quality characteristic value according to the quality characteristic value constructed in step S1 Sequence to obtain the sequence of influencing factors of quality characteristics, and then normalize each influencing factor of quality characteristics, construct a symbolic sequence of influencing factors of quality characteristics, and form a single chain of gene sequence, which specifically includes the following sub-steps:
S21、根据每个时间域内质量特性影响因素在每个采样时刻下的观测值，提取主元信息，将质量特性影响因素主元序列表示为{V1,V2,V3,…VB}，其中质量特性影响因素表示为[V1,V2,V3,…VN]，观测值表示为[x1,…,xN]，VB为第B个质量特性影响因素主元，VN为第N个质量特性因素，xN为第N个影响因素的观察值；S21. According to the observed value of quality characteristic influencing factors in each time domain at each sampling moment, extract the pivotal component information, and express the pivotal component sequence of quality characteristic influencing factors as {V 1 , V 2 , V 3 ,...V B } , where the influencing factors of quality characteristics are expressed as [V 1 , V 2 , V 3 ,…V N ], the observed values are expressed as [x 1 ,…,x N ], and V B is the Bth quality characteristic influencing factor pivot, V N is the Nth quality characteristic factor, x N is the observed value of the Nth influencing factor;
S22、对一个质量特性值与其影响因素建立一个映射其中XN为第N个质量特性值，为XN对应的第N个影响因素；再对一个质量特性值与其影响因素主元建立一个映射其中XK为提取主元之后第K个质量特性值，为提取主元之后XK对应的第B个影响因素；根据步骤S1中构建的质量特性值符号化序列得到质量特性影响因素序列：S22. Establish a mapping between a quality characteristic value and its influencing factors where X N is the Nth quality characteristic value, is the Nth influencing factor corresponding to X N ; then establish a mapping between a quality characteristic value and its influencing factor principal Where X K is the Kth quality property value after extracting the pivot, After extracting the principal element, X K corresponds to the Bth influencing factor; according to the quality characteristic value symbolized sequence constructed in step S1, the quality characteristic influencing factor sequence is obtained:
得到质量特性影响因素序列：Get the sequence of influencing factors of quality characteristics:
S23、将不同质量特性影响因素在不同时刻所造成的影响划分等级，对每个质量特性影响因素进行归一化处理并将归一化处理后的分值划分为10个等级[0-9]，通过设定稳定状态等级阈值，将分值处于稳定区域的质量特性影响因素状态表示为0，将分值处于不稳定区域的质量特性影响因素状态表示为1，从而将质量特性影响因素状态空间SNF表示为：S23. Classify the impact of different quality characteristic influencing factors at different times, perform normalization processing on each quality characteristic influencing factor, and divide the normalized scores into 10 grades [0-9] , by setting the stable state level threshold, the state of the quality characteristic influencing factors whose score is in the stable region is expressed as 0, and the state of the quality characteristic influencing factors whose score is in the unstable region is expressed as 1, so that the quality characteristic influencing factor state space SNF is expressed as:
将质量特性影响因素主元状态空间SKF表示为：The principal component state space S KF of quality characteristic influencing factors is expressed as:
得到质量特性影响因素符号化序列，形成基因序列单链。Obtain the symbolized sequence of factors affecting quality characteristics, and form a single strand of gene sequence.
如图3所示，为本发明的质量特性影响因素主元提取与映射结构基础构建示意图。As shown in FIG. 3 , it is a schematic diagram of principal component extraction and mapping structure foundation construction of quality characteristic influencing factors in the present invention.
在步骤S23中，由于质量特性因素涉及5M1E几个方面，有的因素可测量，有的不可测量，有的是定性描述的，有的可定量描述，所以要统一描述其状态变化，必须用一定的知识方法统一描述与分析。不同影响因素的定量数值之间没有可比性，不能进行对比分析，因此，将各个因素的定量值进行归一化处理，这样，所有的无量纲元素之间就具有了相似性，可以进行进一步的分析。产品质量特性水平是不同因素综合作用的结果，所以给各个因素赋予不同的影响因子，影响因子的大小表示该因素对产品质量的影响程度，因素综合作用的结果最终以产品质量数据的形式体现出来。In step S23, since the quality characteristic factors involve several aspects of 5M1E, some factors can be measured, some cannot be measured, some can be described qualitatively, and some can be described quantitatively, so to uniformly describe the state changes, certain knowledge must be used Methods unified description and analysis. The quantitative values of different influencing factors are not comparable, and comparative analysis cannot be carried out. Therefore, the quantitative values of each factor are normalized, so that all dimensionless elements have similarities and can be further analyzed. analyze. The level of product quality characteristics is the result of the combined effect of different factors, so each factor is given a different impact factor, the size of the impact factor indicates the degree of influence of the factor on product quality, and the result of the combined effect of factors is finally reflected in the form of product quality data .
将不同质量特性影响因素在不同时刻或时段所造成的影响划分等级，对每个质量特性影响因素VN∈[0,1]进行归一化处理并将归一化处理后的分值划分为10个等级[0-9]，通过设定稳定状态等级阈值[α,β]，将分值处于稳定区域的质量特性影响因素状态表示为0，将分值处于不稳定区域的质量特性影响因素状态表示为1，从而将质量特性影响因素状态空间SNF表示为：Divide the impact of different quality characteristic influencing factors at different times or periods into grades, normalize each quality characteristic influencing factor V N ∈ [0,1] and divide the normalized score into 10 levels [0-9], by setting the stable state level threshold [α, β], the state of the quality characteristic influencing factors whose score is in the stable region is expressed as 0, and the quality characteristic influencing factors whose score is in the unstable region The state is expressed as 1, so the state space SNF of quality characteristic influencing factors is expressed as:
将质量特性影响因素主元状态空间SKF表示为：The principal component state space S KF of quality characteristic influencing factors is expressed as:
得到质量特性影响因素符号化序列，形成基因序列单链。稳定状态等级阈值[α,β]表示稳定状态等级的最大范围，不同产品不同质量特性的影响因素的处于稳定状态的等级范围不同，可以根据经验得出。Obtain the symbolized sequence of factors affecting quality characteristics, and form a single strand of gene sequence. Steady-state grade threshold [α, β] represents the maximum range of steady-state grades, and the stable-state grade ranges of factors affecting different quality characteristics of different products are different, which can be obtained based on experience.
如图4所示，为本发明的质量特性影响因素符号化序列构建示意图。如果影响因素主元数为B个，那么影响因素主元符号序列的组合状态有2B。转换的质量特性影响因素字符序列蕴含了稳态和非稳态的质量特性影响因素正常与异常信息。As shown in FIG. 4 , a schematic diagram is constructed for the symbolized sequence of quality characteristic influencing factors in the present invention. If the number of pivots of influencing factors is B, then there are 2 B combined states of the pivot symbol sequence of influencing factors. The character sequence of the converted quality characteristic influencing factors contains the normal and abnormal information of the steady-state and unsteady-state quality characteristic influencing factors.
在步骤S3中，本发明的质量特性字符序列蕴含了已知和未知的质量特性异常模式，并且根据控制图在时间轴上建立了质量特性数据字符序列与影响因素信息字符序列映射关系，通过把包含不同数据波动规律的字符序列空间与包含不同故障因素组合的字符序列空间以及两者之间的时间序列的映射关系，作为质量特性符号化映射控制图的三个组成成分，采用关联双链基因模式共同构建起质量特性符号化映射控制图的基本结构。质量特性两个序列空间中包含有正常模式与异常模式的字符序列，质量特性序列空间是预测诊断质量异常的基础，序列空间中所包含的异常模式字符序列代表质量特性波动变异的主要特征。在序列空间中数据异常波动与故障因素组合在以时间轴建立的质量特性字符信心库中存在固定对应关系。In step S3, the quality characteristic character sequence of the present invention contains known and unknown quality characteristic abnormal patterns, and establishes the mapping relationship between the quality characteristic data character sequence and the influencing factor information character sequence on the time axis according to the control chart, by putting The mapping relationship between the character sequence space containing different data fluctuation rules, the character sequence space containing different failure factor combinations, and the time series between the two, as the three components of the quality characteristic symbolic mapping control chart, adopts the associated double-chain gene The patterns jointly build the basic structure of the quality characteristic symbolic mapping control chart. The two sequence spaces of quality characteristics contain character sequences of normal mode and abnormal pattern. The quality characteristic sequence space is the basis for predicting and diagnosing quality abnormalities. The abnormal pattern character sequences contained in the sequence space represent the main characteristics of quality characteristic fluctuations and variations. In the sequence space, there is a fixed correspondence between the combination of abnormal data fluctuations and failure factors in the quality characteristic character confidence library established by the time axis.
如图5所示，为本发明的质量特性符号化控制图结构示意图。本发明的质量特性符号化控制图以生产过程质量特性数据与故障信息为分析对象，以统计方法与控制图为辅助转换工具，以数据库为载体，以计算机为工具对质量控制过程中产生的大量质量特性数据进行储存、检索、识别、处理及分析，并以质量控制学知识对结果进行解释，从而最终揭示蕴藏质量特性数据与故障信息序列中具有实践性、经验性的质量特性波动规律。As shown in FIG. 5 , it is a schematic structural diagram of the quality characteristic symbolized control diagram of the present invention. The quality characteristic symbolized control chart of the present invention takes the quality characteristic data and fault information of the production process as the analysis object, uses the statistical method and the control chart as the auxiliary conversion tool, uses the database as the carrier, and uses the computer as a tool to analyze a large number of data generated in the quality control process. The quality characteristic data is stored, retrieved, identified, processed and analyzed, and the results are interpreted with the knowledge of quality control science, so as to finally reveal the practical and empirical quality characteristic fluctuation laws contained in the quality characteristic data and fault information sequence.
本领域的普通技术人员将会意识到，这里所述的实施例是为了帮助读者理解本发明的原理，应被理解为本发明的保护范围并不局限于这样的特别陈述和实施例。本领域的普通技术人员可以根据本发明公开的这些技术启示做出各种不脱离本发明实质的其它各种具体变形和组合，这些变形和组合仍然在本发明的保护范围内。Those skilled in the art will appreciate that the embodiments described here are to help readers understand the principles of the present invention, and it should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical revelations disclosed in the present invention without departing from the essence of the present invention, and these modifications and combinations are still within the protection scope of the present invention.
Claims (4)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711188848.1A CN107944491B (en) | 2017-11-24 | 2017-11-24 | Quality characteristic symbolization mapping control chart construction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711188848.1A CN107944491B (en) | 2017-11-24 | 2017-11-24 | Quality characteristic symbolization mapping control chart construction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107944491A true CN107944491A (en) | 2018-04-20 |
| CN107944491B CN107944491B (en) | 2021-01-05 |
Family
ID=61949567
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711188848.1A Expired - Fee Related CN107944491B (en) | 2017-11-24 | 2017-11-24 | Quality characteristic symbolization mapping control chart construction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107944491B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114077228A (en) * | 2020-08-14 | 2022-02-22 | 欧姆龙(上海)有限公司 | System and method for quality control of products |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6811706B1 (en) * | 1999-05-05 | 2004-11-02 | Eric J. Wahlberg | Activated sludge process optimization |
| JP2005275766A (en) * | 2004-03-24 | 2005-10-06 | Seiko Epson Corp | Control chart data processing apparatus, control chart data processing method and program |
| KR20150034482A (en) * | 2013-09-26 | 2015-04-03 | 성균관대학교산학협력단 | Controling method of very small fraction of nonconforming |
| CN105653871A (en) * | 2016-01-06 | 2016-06-08 | 电子科技大学 | Symbolic multi-quality characteristic analysis method based on relative entropy and gray relational degree |
| CN105678078A (en) * | 2016-01-06 | 2016-06-15 | 电子科技大学 | Symbolized quality characteristic grey prediction method of complicated electromechanical system |
| CN107122907A (en) * | 2017-04-28 | 2017-09-01 | 电子科技大学 | A kind of analysis of electronic product symbolism mass property and failure cause retroactive method |
| CN107122609A (en) * | 2017-04-28 | 2017-09-01 | 电子科技大学 | A kind of electromechanical product quality evaluation method based on mass property gene theory |
-
2017
- 2017-11-24 CN CN201711188848.1A patent/CN107944491B/en not_active Expired - Fee Related
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6811706B1 (en) * | 1999-05-05 | 2004-11-02 | Eric J. Wahlberg | Activated sludge process optimization |
| JP2005275766A (en) * | 2004-03-24 | 2005-10-06 | Seiko Epson Corp | Control chart data processing apparatus, control chart data processing method and program |
| KR20150034482A (en) * | 2013-09-26 | 2015-04-03 | 성균관대학교산학협력단 | Controling method of very small fraction of nonconforming |
| CN105653871A (en) * | 2016-01-06 | 2016-06-08 | 电子科技大学 | Symbolic multi-quality characteristic analysis method based on relative entropy and gray relational degree |
| CN105678078A (en) * | 2016-01-06 | 2016-06-15 | 电子科技大学 | Symbolized quality characteristic grey prediction method of complicated electromechanical system |
| CN107122907A (en) * | 2017-04-28 | 2017-09-01 | 电子科技大学 | A kind of analysis of electronic product symbolism mass property and failure cause retroactive method |
| CN107122609A (en) * | 2017-04-28 | 2017-09-01 | 电子科技大学 | A kind of electromechanical product quality evaluation method based on mass property gene theory |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114077228A (en) * | 2020-08-14 | 2022-02-22 | 欧姆龙(上海)有限公司 | System and method for quality control of products |
| CN114077228B (en) * | 2020-08-14 | 2025-04-29 | 欧姆龙(上海)有限公司 | System and method for controlling product quality |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107944491B (en) | 2021-01-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109459993B (en) | An online adaptive fault monitoring and diagnosis method for process industry process | |
| CN109308063B (en) | Stage division method of fermentation process based on score matrix | |
| CN108803520B (en) | A Dynamic Process Monitoring Method Based on Variable Nonlinear Autocorrelation Elimination | |
| CN111638707B (en) | Intermittent process fault monitoring method based on SOM clustering and MPCA | |
| CN107436597A (en) | A kind of chemical process fault detection method based on sparse filtering and logistic regression | |
| CN105574587B (en) | A method for online working condition process monitoring of plastic injection molding process | |
| CN108776831A (en) | A kind of complex industrial process Data Modeling Method based on dynamic convolutional neural networks | |
| CN110245460A (en) | A Fault Monitoring Method for Batch Process Based on Multi-Stage OICA | |
| WO2017088208A1 (en) | Data-difference-driven self-learning dynamic optimization method for batch process | |
| CN112199829A (en) | Multi-condition and multi-stage batch process monitoring method based on density peak clustering and just-in-time learning | |
| CN114662406A (en) | Method for establishing rock joint peak shear strength prediction model with assistance of machine learning | |
| CN107403196A (en) | Instant learning modeling method based on spectral clustering analysis | |
| CN111079856A (en) | CSJITL-RVM-based multi-period intermittent process soft measurement modeling method | |
| CN110032799A (en) | A kind of the angle similarity divided stages and monitoring method of microbiological pharmacy process | |
| CN106845825B (en) | Strip steel cold rolling quality problem tracing and control method based on improved PCA | |
| CN112381145A (en) | Gaussian process regression multi-model fusion modeling method based on nearest correlation spectral clustering | |
| CN107944491A (en) | Mass property symbolism maps control figure construction method | |
| Wu et al. | Application of fuzzy pushdown automaton on prediction of quality control for spinning yarn: Application of fpda on quality prediction | |
| CN117314202B (en) | Product carbon efficiency monitoring and optimizing method and system based on industrial Internet | |
| Song et al. | Fault diagnosis and process monitoring using a statistical pattern framework based on a self-organizing map | |
| CN110517724B (en) | Method for deducing gene regulation network by using single cell transcription and gene knockout data | |
| CN114368768A (en) | Aluminum hydroxide crystal seed granularity refinement explosion prediction model and method based on LSTM | |
| CN112183569A (en) | FDA and SOM based intermittent industrial process reaction phase clustering and fault classification visualization | |
| TWI639908B (en) | Method for detecting and diagnosing an abnormal process | |
| Barkia et al. | Semiconductor yield loss' causes identification: A data mining approach |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20210105 Termination date: 20211124 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |