Bearing production quality detection system based on production data analysis

Abstract

The invention discloses a bearing production quality detection system based on production data analysis, and relates to the technical field of bearing production, and the system comprises a production data collection module which is used for collecting production data in the machining process of a bearing, and adding a process identification field and a time identification field for the production data in the collection process; and the data binding module is used for acquiring the unique identification information of the bearing individuals, and binding the production data which are acquired in different production processes and are provided with process identification fields and time identification fields to the corresponding bearing individuals based on the unique identification information. According to the invention, unified acquisition, binding, analysis and analysis of production data formed in a plurality of production processes of the bearing are realized; a process identification field and a time identification field are added to production data, and the production data in different production processes are collected based on unique identification information of a bearing individual, so that the production data can reflect a complete manufacturing process by taking the bearing individual as an index.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of bearing production, in particular to a bearing production quality detection system based on production data analysis. BACKGROUND

[0002] With the development of bearing manufacturing towards high precision and high reliability, the manufacturing process control of bearings in grinding, superfinishing and assembly and other machining processes has become an important factor affecting the quality of finished products. In actual production process, production data reflecting the state of processing or assembly process, such as processing parameters, assembly parameters, etc., are usually collected in each machining process for equipment operation monitoring, process stability analysis or production process recording, so as to ensure the continuity and controllability of the production process. In the existing bearing production process, although production data can be collected in different production processes and basically distinguished by process identification or time information, production data is usually stored or analyzed independently in units of processes, mainly for state monitoring or statistical analysis within a single process, and a cross-process production data collection method based on bearing individual indexing has not been formed. Due to the lack of unified binding and organization of production data of the same bearing individual in multiple production processes, it is difficult for production data to reflect the complete manufacturing process evolution state from the bearing individual level. In addition, in the prior art, the detection of bearing quality is mainly concentrated on the detection of geometric size, appearance or performance in the finished product stage, or only based on abnormal data in a single process. Since the processing state between different production processes has sequentiality and correlation, if the cross-process production data is not analyzed systematically, it is difficult to identify the potential quality risk of bearings caused by abnormal process state evolution in the manufacturing process. Therefore, the present application provides a bearing production quality detection system based on production data analysis. SUMMARY

[0003] The purpose of the present application is to provide a bearing production quality detection system based on production data analysis to solve the problems mentioned in the background.

[0004] The present application can be realized by the following technical scheme: a bearing production quality detection system based on production data analysis, comprising a production data acquisition module, a data binding module, a data analysis module, a data identification module and a bearing judgment module. The production data acquisition module is used to collect production data in the grinding process, superfinishing process and assembly process of the bearing, and to attach a process identification field and a time identification field to the production data during the collection process. The data binding module is used to obtain unique identification information of the bearing individual, and based on the unique identification information, production data collected in different production processes and carrying process identification fields and time identification fields are bound to the corresponding bearing individual to form a production data set indexed by the bearing individual; The data analysis module is used to group the production data set according to the production process, and extract process characteristic parameters based on the grouped production data to construct a process characteristic data set containing a process identification field, a time sequence field and a process characteristic parameter field; The data recognition module is used to perform cross-process correlation analysis on process characteristic parameters of different production processes based on a plurality of process characteristic data sets corresponding to the same bearing individual, and generate abnormal data identification information when the corresponding relationship of the process characteristic parameters in the process sequence or the time sequence deviates. The bearing determination module is used to output the quality detection result of the corresponding bearing individual based on the existence state or distribution state of the abnormal data identification information.

[0005] The further technical improvement of the application is that the production data acquisition module acquires production data, including: When the bearing enters the grinding process, the superfinishing process and the assembly process, the target process in which the bearing is located is determined based on the current production process state, and a process identification field corresponding only to the target process is generated accordingly; During the execution of the target process, the production data reflecting the state of bearing machining or assembly process are continuously acquired according to the process rhythm of the target process to form a production data sequence that can reflect the change of process execution process; At the same time of acquiring the production data sequence, a corresponding time identification field is generated for each piece of production data based on a unified time reference to maintain the time sequence consistency of the production data within the same process; The process identification field, the time identification field and the corresponding production data sequence are structurally combined to form a production data basic unit for subsequent cross-process correlation analysis.

[0006] The further technical improvement of the application is that the data binding module binds the production data and the bearing individual, including: Before the production data corresponding to the bearing individual is written into the production data set, the unique identification information of the bearing individual is obtained and a bearing index key is generated; The production data output by each production process and carrying process identification fields and time identification fields are received, and the production data are classified by process according to the process identification fields; Based on the process classification, the production data under the same process classification are arranged in time sequence according to the time identification fields to form an ordered production data sequence corresponding to the production process; The ordered production data sequence of each production procedure is bound with the bearing index key and written into the production data set, so that the production data set realizes the collection of cross-procedure production data of the same bearing individual with the bearing index key as the index.

[0007] The further technical improvement of the present application is that the step of constructing the procedure characteristic data set by the data analysis module comprises: The production data set indexed by the bearing individual is read, and the production data is divided into the grinding procedure production data subset, the superfinishing procedure production data subset and the assembly procedure production data subset according to the procedure identification field; In each production data subset, the production data is arranged in time sequence according to the time identification field to form the time sequence field of the corresponding production procedure; For the time sequence field, the change amount between adjacent sampling points is calculated and the distribution range of the change amount is counted to generate the procedure characteristic parameter representing the stability of the production procedure; The procedure identification field, the time sequence field and the procedure characteristic parameter field are structurally combined to form the procedure characteristic data set of the corresponding bearing individual.

[0008] The further technical improvement of the present application is that the procedure characteristic parameter acquisition method comprises: The production data subset corresponding to the same bearing individual is read and the time sequence field is formed according to the time identification field; In the time sequence field, the procedure execution process is divided into at least two continuous segments according to a preset segment division rule, and the continuous segments correspond to different stages of the procedure execution respectively; The change amount between adjacent sampling points is calculated in each continuous segment, and the segment stability parameter is generated based on the continuous change characteristics of the change amount in the continuous segment; The segment stability parameters corresponding to each continuous segment are summarized to obtain the procedure characteristic parameter representing the stability of the production procedure, and the procedure identification field, the time sequence field and the procedure characteristic parameter field are structurally combined to form the procedure characteristic data set of the corresponding bearing individual.

[0009] The further technical improvement of the present application is that when the data recognition module performs cross-procedure correlation analysis on the procedure characteristic parameters of different production procedures, it comprises: According to the predetermined order of the production procedures, the procedure characteristic data sets formed in the grinding procedure, the superfinishing procedure and the assembly procedure of the same bearing individual are read in sequence; The procedure characteristic parameters corresponding to each production procedure are sequentially connected in the production procedure order to form the procedure characteristic parameter evolution path describing the manufacturing process evolution state of the bearing individual in the multiple production procedures; The process characteristic parameters between adjacent production processes are processed in segments along the process characteristic parameter evolution path, so that the cross-process correlation analysis is limited to the process characteristic parameter evolution path, thereby avoiding correlation analysis of non-adjacent production processes or process characteristic parameters without manufacturing sequence relationship.

[0010] Further technical improvements of the present application are that the segment-by-segment correlation processing includes: According to the predetermined sequence of production processes, the process characteristic parameter evolution path corresponding to the same bearing individual is divided into a plurality of correlation segments composed of adjacent production processes, and each correlation segment contains the process characteristic parameters of the previous production process and the process characteristic parameters of the subsequent production process. For each correlation segment, based on the process characteristic parameters of the previous production process and the process characteristic parameters of the subsequent production process, the cross-process corresponding relationship features corresponding to the correlation segment are formed respectively. The cross-process corresponding relationship features formed by each correlation segment are sequentially recorded according to the production process sequence to obtain a segment-by-segment correlation result set. Based on the segment-by-segment correlation result set, it is judged whether the corresponding relationship of the process characteristic parameters deviates in the process sequence or the time sequence.

[0011] Further technical improvements of the present application are that the judgment basis of the data recognition module when judging whether the process characteristic parameters deviate includes: Along the process characteristic parameter evolution path corresponding to the same bearing individual, the process characteristic parameter pairs corresponding to adjacent production processes are selected in sequence according to the predetermined sequence of production processes. For each group of process characteristic parameter pairs, the corresponding relationship features of the process characteristic parameters of the previous production process and the process characteristic parameters of the subsequent production process in terms of numerical change direction, change continuity and stage stability are determined respectively, and the corresponding relationship features are sequentially recorded according to the production process sequence to form a structure description set of the process characteristic parameter evolution path. The corresponding relationship features in the structure description set are used as the corresponding relationship judgment basis representing the cross-process evolution behavior of the process characteristic parameters of the bearing individual under normal manufacturing state. When the corresponding relationship features of the subsequently obtained process characteristic parameters fail to maintain consistency with the corresponding relationship judgment basis, it is determined that the corresponding relationship of the process characteristic parameters deviates, and corresponding abnormal data identification information is generated.

[0012] Further technical improvements of the present application are that the quality detection results output by the bearing determination module include: When the read abnormal data identification information represents that at least one production process associated section deviates abnormally, output a quality detection result representing that the bearing individual has a quality risk, and synchronously output the production process associated section position information corresponding to the abnormal deviation; When no abnormal data identification information representing that the production process associated section deviates abnormally is read, output a quality detection result representing that the bearing individual has normal quality.

[0013] Compared with the prior art, the present application has the following beneficial effects: The present application realizes unified collection, binding, analysis and analysis of production data formed in multiple production processes of bearings; by attaching a process identification field and a time identification field to the production data, and collecting the production data in different production processes based on the unique identification information of the bearing individual, the production data can reflect the complete manufacturing process of the bearing individual as an index, providing a structured data basis for subsequent cross-process analysis; Moreover, the present application analyzes the production data set indexed by the bearing individual, constructs a process characteristic data set containing a process identification field, a time sequence field and a process characteristic parameter field, and performs cross-process correlation analysis based on the process characteristic data sets formed in multiple production processes of the same bearing individual; by correlating the process characteristic parameters between adjacent production processes along the process characteristic parameter evolution path, and combining the corresponding relationship changes in process order or time order, the abnormal evolution behavior of the bearing in the manufacturing process can be more accurately identified, thereby improving the pertinence and reliability of abnormal identification; On the other hand, the present application outputs the quality detection result of the bearing individual based on the existence state or distribution state of the abnormal data identification information in the bearing determination module, and synchronously outputs the corresponding production process associated section position information when there is abnormal deviation, so that the quality detection result can not only reflect whether the bearing has a quality risk, but also indicate the specific manufacturing link where the abnormality occurs, which is beneficial to traceability analysis and process optimization of the production process, thereby fully utilizing the production data resources without changing the existing production process, and improving the comprehensiveness and refinement level of bearing production quality detection. BRIEF DESCRIPTION OF DRAWINGS

[0014] In order to facilitate understanding of those skilled in the art, the present application will be further described below with reference to the accompanying drawings.

[0015] Figure 1 The system logic diagram of the present application. DETAILED DESCRIPTION

[0016] To further illustrate the technical means and effects taken by the present application to achieve the predetermined inventive purpose, the specific embodiments, structures, features and effects according to the present application are described in detail below in combination with the drawings and preferred embodiments.

[0017] Referring to Figure 1 The present application provides a bearing production quality detection system based on production data analysis, which comprises a production data acquisition module, a data binding module, a data analysis module, a data recognition module and a bearing judgment module. The production data acquisition module is used to acquire production data in the grinding process, superfinishing process and assembly process of the bearing, and to attach a process identification field and a time identification field to the production data during the acquisition process. The production data acquisition module continuously records the processing or assembly process state of the bearing in the grinding process, superfinishing process and assembly process in the form of production data, and clearly indicates the production process to which each piece of production data belongs through the process identification field, and clearly indicates the time sequence of each piece of production data in the same production process through the time identification field, so that the subsequent steps can organize and analyze the production data without confusing the process source and the time sequence.

[0018] The step of acquiring production data by the production data acquisition module comprises: When the bearing enters the grinding process, superfinishing process and assembly process, the target process in which the bearing is located is determined based on the current production process state, and a process identification field corresponding only to the target process is generated accordingly; During the execution of the target process, the production data reflecting the processing or assembly process state of the bearing is continuously acquired according to the process rhythm of the target process, so as to form a production data sequence that can reflect the changes in the execution process of the process; At the same time of acquiring the production data sequence, a corresponding time identification field is generated for each piece of production data based on a unified time reference, so as to maintain the time sequence consistency of the production data in the same process; The process identification field, time identification field and corresponding production data sequence are structurally combined to form a production data basic unit for subsequent cross-process correlation analysis.

[0019] Specifically, when the bearing enters the grinding process, superfinishing process and assembly process, the production data acquisition module determines the target process in which the bearing is located based on the current production process state, and the production process state includes the station occupancy state, the process start state and the tooling fixture state. By comprehensively judging the production process state, the target process corresponding to the bearing at present is determined, and a process identification field corresponding only to the target process is generated accordingly; In this embodiment, when the station occupancy state corresponds to a grinding station, the process start state is started, and the tooling fixture state is locked, it is determined that the target process is a grinding process, and a process identifier field uniquely corresponding to the grinding process is generated.

[0020] In the execution process of the target process, production data reflecting the state of bearing processing or assembly is continuously collected according to the process cycle of the target process to form a production data sequence that can reflect changes in the process execution process. In this embodiment, the spindle current, spindle speed and feed rate are continuously collected according to a process cycle of two hundred milliseconds in the grinding process, the sand belt tension, vibration amplitude and processing time are continuously collected according to a process cycle of three hundred milliseconds in the superfinishing process, and the assembly force, press-fit displacement and assembly time are continuously collected according to a process cycle of one hundred milliseconds in the assembly process, thereby forming a production data sequence corresponding to the target process.

[0021] At the same time of collecting the production data sequence, a corresponding time identifier field is generated for each piece of production data based on a unified time reference, and the unified time reference is a timestamp corresponding to the production line unified clock. By generating a corresponding time identifier field for each piece of production data in the same process, the consistency of the production data in the same process in time sequence is maintained. In this embodiment, continuously increasing time identifier fields are generated for the spindle current data continuously collected in the grinding process.

[0022] Finally, the process identifier field, the time identifier field and the corresponding production data sequence are structurally combined, the process identifier field, the time identifier field and the production data sequence are field-aligned and packaged to form a production data basic unit for subsequent cross-process correlation analysis, so that the production data basic units formed by different production processes are called and executed for cross-process correlation analysis in a unified field structure.

[0023] The data binding module is used to obtain unique identification information of a bearing individual, and based on the unique identification information, production data collected in different production processes and having a process identifier field and a time identifier field are bound to the corresponding bearing individual to form a production data set indexed by the bearing individual. The data binding module changes the production data dispersed in different production processes from "dispersed records in the process dimension" to "cross-process collection in the bearing individual dimension", so that the production data formed by the same bearing individual in the grinding process, the superfinishing process and the assembly process can be uniformly written into the production data set indexed by the bearing individual, thereby providing a consistent data entry and index basis for subsequent cross-process correlation analysis according to the bearing individual.

[0024] The steps of the data binding module binding the production data and the bearing individual include: Before the production data corresponding to the bearing individual is written into the production data set, the unique identification information of the bearing individual is obtained and the bearing index key is generated; The production data output by each production process is received, with a process identification field and a time identification field, and the production data is classified by process according to the process identification field; On the basis of process classification, the production data under the same process classification is arranged in time sequence according to the time identification field, forming an ordered production data sequence corresponding to the production process; The ordered production data sequence of each production process is bound with the bearing index key and written into the production data set, so that the production data set realizes the collection of cross-process production data of the same bearing individual with the bearing index key as the index.

[0025] Specifically, before the production data corresponding to the bearing individual is written into the production data set, the data binding module obtains the unique identification information of the bearing individual and generates the bearing index key, and the unique identification information is the identification information that can be stably read in the production circulation process; In this embodiment, the unique identification information is "B202512170001", the data binding module encapsulates the unique identification information and generates the bearing index key as "IDX_B202512170001", and temporarily stores the bearing index key in the station circulation record corresponding to the bearing individual, so that the production data received subsequently can be accurately pointed to the same bearing individual.

[0026] The data binding module receives the production data output by each production process, with a process identification field and a time identification field, and classifies the production data by process according to the process identification field. In this embodiment, the production data output by the grinding process is received, the process identification field is the grinding process and contains spindle current data with time identification fields of 08:00:00.200 and 08:00:00.400, the production data output by the superfinishing process is received, the process identification field is the superfinishing process and contains vibration amplitude data with time identification fields of 08:10:00.300 and 08:10:00.600, the production data output by the assembly process is received, the process identification field is the assembly process and contains assembly force data with time identification fields of 08:20:00.100 and 08:20:00.200, and the production data is written into the corresponding process classification buffer area.

[0027] On the basis of the process classification, the data binding module arranges the production data under the same process classification in time sequence according to the time identifier field to form an ordered production data sequence corresponding to the production process. In this embodiment, the spindle current data with the time identifier field of 08:00:00.200 is arranged before the spindle current data with the time identifier field of 08:00:00.400 under the grinding process classification, so as to form the ordered production data sequence of the grinding process. The vibration amplitude data with the time identifier field of 08:10:00.300 is arranged before the vibration amplitude data with the time identifier field of 08:10:00.600 under the superfinishing process classification, so as to form the ordered production data sequence of the superfinishing process. The assembly force data with the time identifier field of 08:20:00.100 is arranged before the assembly force data with the time identifier field of 08:20:00.200 under the assembly process classification, so as to form the ordered production data sequence of the assembly process.

[0028] Finally, the data binding module binds the ordered production data sequence of each production process with the bearing index key and writes it into the production data set, so that the production data set realizes the collection of cross-process production data of the same bearing individual with the bearing index key as the index; In this embodiment, the "IDX_B202512170001" is used as the index key, and the ordered production data sequence of the grinding process, the ordered production data sequence of the superfinishing process and the ordered production data sequence of the assembly process are written into the production data set, so as to form the cross-process production data collection record indexed by the bearing index key in the production data set, for reading and performing cross-process correlation analysis in the subsequent steps.

[0029] The data analysis module is used for grouping processing the production data set according to the production process, and extracting the process characteristic parameters based on the grouped production data to construct a process characteristic data set containing a process identifier field, a time sequence field and a process characteristic parameter field; The data analysis module groups the production data according to the production process based on the production data set indexed by the bearing individual and forms a time sequence field corresponding to the production process, so as to maintain the time sequence consistency of the production data within the same production process, and on this basis, extracts the process characteristic parameters capable of representing the stability of the production process from the time sequence field, so that the subsequent steps can carry out cross-process comparison and correlation analysis based on the structured process characteristic data set without directly relying on a large amount of original production data.

[0030] The step of constructing the process characteristic data set by the data analysis module includes: The production data set indexed by the bearing individual is read, and the production data is divided into a grinding process production data subset, a superfinishing process production data subset and an assembly process production data subset according to the process identifier field. In each production data subset, the production data is arranged in time sequence according to the time identifier field to form a time sequence field corresponding to the production process; For the time sequence field, the change amount between adjacent sampling points is calculated and the distribution range of the change amount is counted to generate a process characteristic parameter representing the stability of the production process; The process identifier field, the time sequence field and the process characteristic parameter field are structurally combined to form a process characteristic data set corresponding to the bearing individual.

[0031] The process characteristic parameter acquisition method comprises: Reading the production data subset corresponding to the same bearing individual and forming a time sequence field according to the time identifier field; In the time sequence field, the process execution process is divided into at least two continuous segments according to a preset segment division rule, and the continuous segments correspond to different stages of process execution respectively; In each continuous segment, the change amount between adjacent sampling points is calculated, and a segment stability parameter is generated based on the continuous change characteristics of the change amount in the continuous segment; The segment stability parameters corresponding to each continuous segment are summarized to obtain a process characteristic parameter representing the stability of the production process, and the process identifier field, the time sequence field and the process characteristic parameter field are structurally combined to form a process characteristic data set corresponding to the bearing individual.

[0032] Specifically, the production data set indexed by the bearing individual is first read, and the production data is divided into grinding process production data subset, superfinishing process production data subset and assembly process production data subset according to the process identifier field, wherein the process identifier field is determined by the production process state, and the production process state is updated when the bearing enters the corresponding production process and remains until the end of the production process. In the production data collection process, each piece of production data is recorded with the process identifier field corresponding to the current production process state when generated, and when the production process state changes, the production data collected subsequently is automatically associated with the updated process identifier field, so that the production data formed before and after the process switching is clearly distinguished in the process identifier field. In the data analysis stage, the production data records with the same process identifier field are divided into the grinding process production data subset, the superfinishing process production data subset or the assembly process production data subset by reading the process identifier field.

[0033] In each production data subset, the production data is arranged in time sequence according to the time identifier field, forming a time sequence field corresponding to the production process, wherein the earliest time identifier field in the production data subset of the same production process is taken as the starting point of the time sequence field, and the last time identifier field in the production data subset of the same production process is taken as the termination point of the time sequence field; the production data records with the same process identifier field are sorted according to the time identifier field in the order of time, forming a monotonically increasing time sequence field; when there is a time interval discontinuity between adjacent production data records, the original time identifier field is retained without interpolation processing, so as to ensure that the time sequence field truly reflects the collection process of the production data.

[0034] After forming the time sequence field, adjacent production data records in the time sequence field are selected in turn, the difference between the value of the latter production data record and the value of the former production data record is calculated, and the change amount sequence between adjacent sampling points is obtained; after obtaining the change amount sequence, the maximum change amount value and the minimum change amount value in the change amount sequence are determined, and the numerical interval between the maximum change amount value and the minimum change amount value is taken as the distribution range of the change amount; the distribution range of the change amount is taken as the process characteristic parameter for representing the stability degree of the production process in the entire execution process; Subsequently, on the basis of the time sequence field, the execution process of the production process is divided into at least two continuous sections according to a preset section division rule, and the continuous sections correspond to different stages of the execution of the production process; in each continuous section, the change amount between adjacent sampling points is calculated, and the section stability parameter is generated based on the continuous change characteristics of the change amount in the continuous section; the section stability parameters corresponding to each continuous section are summarized to obtain the process characteristic parameter representing the stability of the production process.

[0035] Finally, the process identifier field, the time sequence field and the process characteristic parameter field are structurally combined to form a process characteristic data set corresponding to the bearing individual, so that the process characteristic data set reflects the overall stability of the production process and the stability characteristics of the production process in different execution stages at the same time.

[0036] The data recognition module is used for cross-process correlation analysis of the process characteristic parameters of different production processes based on the multiple process characteristic data sets corresponding to the same bearing individual, and generates abnormal data identification information when the corresponding relationship of the process characteristic parameters in the process sequence or the time sequence deviates; The data recognition module organizes the process feature data sets formed in the grinding process, the superfinishing process and the assembling process of the same bearing individual along the predetermined order of the production processes, so that the cross-process correlation analysis can establish a clear corresponding relationship in the process order or time order, and the bearing individual experiencing an abnormal manufacturing process evolution state is recognized by judging whether the corresponding relationship deviates, and further generates abnormal data identification information indicating the existence of abnormal deviation, thereby providing a direct basis for the output of the quality detection result of the bearing judgment module.

[0037] The data recognition module includes the following steps when performing cross-process correlation analysis on the process feature parameters of different production processes: According to the predetermined order of the production processes, the process feature data sets formed in the grinding process, the superfinishing process and the assembling process of the same bearing individual are sequentially read; The process feature parameters corresponding to each production process are sequentially connected in the production process order to form a process feature parameter evolution path describing the manufacturing process evolution state of the bearing individual in multiple production processes; Along the process feature parameter evolution path, the process feature parameters between adjacent production processes are processed in segments to limit the cross-process correlation analysis along the process feature parameter evolution path, thereby avoiding the correlation analysis of non-adjacent production processes or process feature parameters without manufacturing sequence relationship.

[0038] The segmented correlation processing includes the following steps: According to the predetermined order of the production processes, the process feature parameter evolution path corresponding to the same bearing individual is divided into multiple correlation segments composed of adjacent production processes, and each correlation segment contains the process feature parameters of the previous production process and the process feature parameters of the subsequent production process; For each correlation segment, based on the process feature parameters of the previous production process and the process feature parameters of the subsequent production process, the cross-process corresponding relationship features corresponding to the correlation segment are formed respectively; The cross-process corresponding relationship features formed by each correlation segment are sequentially recorded according to the production process order to obtain a segmented correlation result set; Based on the segmented correlation result set, it is judged whether the corresponding relationship of the process feature parameters deviates in the process order or time order.

[0039] Further technical improvements of the present application are that the judgment basis of the data recognition module when judging whether the process feature parameters deviate includes: Along the process feature parameter evolution path corresponding to the same bearing individual, the process feature parameter pairs corresponding to adjacent production processes are sequentially selected according to the predetermined order of the production processes; For each set of process characteristic parameter pairs, the corresponding relationship characteristics of the process characteristic parameters of the previous production process and the process characteristic parameters of the subsequent production process in the numerical change direction, change continuity and stage stability are determined respectively, and the corresponding relationship characteristics are sequentially recorded according to the production process order to form a structural description set of the process characteristic parameter evolution path; The corresponding relationship characteristics in the structural description set are taken as the corresponding relationship judgment basis representing the cross-process evolution behavior of the process characteristic parameters of the bearing individual under the normal manufacturing state; When the corresponding relationship characteristics of the subsequently obtained process characteristic parameters in the process order or time order fail to maintain consistency with the corresponding relationship judgment basis, it is determined that the corresponding relationship of the process characteristic parameters deviates, and the corresponding abnormal data identification information is generated.

[0040] Specifically, the data recognition module is used to realize bearing production quality detection based on production data analysis, and the implementation process includes a forming stage of the corresponding relationship judgment basis and a deviation judgment stage of the bearing to be detected, and the two stages are executed in time sequence and are independent in logic.

[0041] In the forming stage of the corresponding relationship judgment basis, a plurality of bearing individuals confirmed to be in a normal manufacturing state are selected, and process characteristic data sets formed in the grinding process, the superfinishing process and the assembly process of the bearing individuals are obtained respectively, and the process characteristic data sets all contain a process identification field, a time sequence field and a process characteristic parameter field; in this embodiment, the process characteristic parameter field of the grinding process is the distribution range of the change amount of 0.1 to 0.5, the process characteristic parameter field of the superfinishing process is the distribution range of the change amount of 0.1 to 0.3, and the process characteristic parameter field of the assembly process is the distribution range of the change amount of 2 to 17.

[0042] The process characteristic parameters corresponding to each production process are sequentially connected according to the predetermined order of the production process to form a process characteristic parameter evolution path describing the manufacturing process evolution state of the bearing individual in a plurality of production processes, so that the process characteristic parameter evolution path can represent the continuous manufacturing state change of the bearing individual in the grinding process, the superfinishing process and the assembly process.

[0043] Subsequently, a section-by-section association process is performed, including: according to the predetermined order of the production process, the process characteristic parameter evolution path is divided into a plurality of association sections composed of adjacent production processes, each association section containing the process characteristic parameters of the previous production process and the process characteristic parameters of the subsequent production process; for each association section, the cross-process corresponding relationship characteristics corresponding to the association section are formed based on the process characteristic parameters of the previous production process and the process characteristic parameters of the subsequent production process; the cross-process corresponding relationship characteristics formed by each association section are sequentially recorded according to the production process order to obtain a section-by-section association result set.

[0044] According to the established order of the production processes, a corresponding pair of process feature parameters of adjacent production processes is selected in sequence along the process feature parameter evolution path corresponding to the same bearing individual, and for each group of process feature parameter pairs, corresponding relationship features of the process feature parameters of the previous production process and the process feature parameters of the subsequent production process in terms of the numerical change direction, change continuity and stage stability are determined respectively, and the corresponding relationship features are sequentially recorded according to the production process order to form a structural description set of the process feature parameter evolution path; and the corresponding relationship features in the structural description set are summarized as a corresponding relationship judgment criterion representing the cross-process evolution behavior of the process feature parameters of the bearing individual under the normal manufacturing state.

[0045] In the present embodiment, the process feature parameters corresponding to the grinding process and the superfinishing process present convergent change characteristics from 0.1 to 0.5 to 0.1 to 0.3 in the numerical change direction, present continuous change characteristics in the change continuity, and present stable change characteristics in the stage stability; the process feature parameters corresponding to the superfinishing process and the assembly process present divergent change characteristics from 0.1 to 0.3 to 2 to 17 in the numerical change direction, present step change characteristics in the change continuity, and present unstable change characteristics in the stage stability, and the above corresponding relationship features are recorded and used to constitute the corresponding relationship judgment criterion.

[0046] In the deviation determination stage of the bearing to be detected, for the bearing individual to be detected, the process feature data set formed in the grinding process, the superfinishing process and the assembly process is obtained, and the process feature parameter evolution path corresponding to the bearing to be detected is constructed in the same manner as in the corresponding relationship judgment criterion formation stage, and the step-by-step correlation processing is performed along the process feature parameter evolution path to form a corresponding step-by-step correlation result set.

[0047] The step-by-step correlation result set of the bearing to be detected is compared with the corresponding relationship judgment criterion formed in advance, and when the cross-process corresponding relationship features formed in any correlation segment of the bearing to be detected fail to maintain consistency with the corresponding relationship judgment criterion in terms of the numerical change direction, the change continuity or the stage stability, it is determined that the corresponding relationship of the process feature parameters of the bearing to be detected deviates, and corresponding abnormal data identification information is generated.

[0048] The bearing determination module is used for outputting the quality detection result of the corresponding bearing individual based on the existence state or distribution state of the abnormal data identification information, and the abnormal data identification information is converted into the quality detection result output which can be directly used for quality control. When the existence state of the abnormal data identification information indicates that at least one production process associated section deviates abnormally, the quality detection result indicating that the bearing individual has a quality risk is outputted, and the position information of the production process associated section corresponding to the abnormal deviation is synchronously outputted. When the abnormal data identification information indicating that the production process associated section deviates abnormally is not read, the quality detection result indicating that the quality of the bearing individual is normal is outputted, so as to realize the deterministic output and traceable positioning of the quality detection result of the bearing individual.

[0049] The quality detection result outputted by the bearing determination module comprises: When the read abnormal data identification information indicates that at least one production process associated section deviates abnormally, the quality detection result indicating that the bearing individual has a quality risk is outputted, and the position information of the production process associated section corresponding to the abnormal deviation is synchronously outputted. When the abnormal data identification information indicating that the production process associated section deviates abnormally is not read, the quality detection result indicating that the quality of the bearing individual is normal is outputted.

[0050] Specifically, after the abnormal identification of each production process associated section of the same bearing individual is completed, the abnormal data identification information is read, and the abnormal data identification information at least comprises the abnormal production process associated section identification field and the abnormal deviation type field. In this embodiment, for a certain bearing individual, one abnormal data identification information is generated in the production process associated section composed of the grinding process and the superfinishing process, and the abnormal deviation type field indicates that the associated section deviates abnormally. Based on the read abnormal data identification information, it is judged whether the abnormal data identification information indicates that at least one production process associated section deviates abnormally. When there is at least one abnormal data identification information corresponding to different or same production process associated sections, it is determined that the bearing individual has an abnormal deviation state. In the case where it is determined that the bearing individual has an abnormal deviation state, the position information of the production process associated section corresponding to the abnormal deviation is extracted according to the production process associated section identification field in the abnormal data identification information, and when the read abnormal data identification information indicates that at least one production process associated section deviates abnormally, the quality detection result indicating that the bearing individual has a quality risk is outputted, and the position information of the production process associated section corresponding to the abnormal deviation is synchronously outputted. In this embodiment, the quality detection result of "having a quality risk" is outputted, and the position information of "grinding process-superfinishing process associated section" is synchronously outputted. In the case that no abnormal data identification information representing abnormal deviation of the production process associated section is detected, it is determined that the bearing individual is in a normal manufacturing state in each production process associated section, and when no abnormal data identification information representing abnormal deviation of the production process associated section is read, a quality detection result representing normal quality of the bearing individual is output, thereby completing quality detection determination of the bearing individual.

[0051] The above formulas are dimensionless values calculated, the formulas are obtained by collecting a large amount of data to simulate a formula of the nearest real situation, and the preset parameters and threshold values in the formula are set by a person skilled in the art according to the actual situation.

[0052] The above is only a preferred embodiment of the present application, and does not limit the present application in any form. Although the present application has been disclosed as above with a preferred embodiment, it is not intended to limit the present application. Any person skilled in the art can make slight changes or modifications to the above disclosed technical content to obtain equivalent embodiments with equivalent changes without departing from the technical solution of the present application. Any simple modification, equivalent change and modification of the above embodiments made according to the technical essence of the present application still belong to the scope of the technical solution of the present application.

Claims

1. A bearing production quality inspection system based on production data analysis, characterized in that, include: The production data acquisition module is used to collect production data during the bearing processing and to add process identifier fields and time identifier fields to the production data during the acquisition process. The data binding module is used to obtain the unique identification information of each bearing and, based on the unique identification information, bind the production data collected in different production processes, which have process identification fields and time identification fields, to the corresponding bearings to form a production data set indexed by the bearings. The data parsing module is used to group the production data set according to the production process, extract process feature parameters based on the grouped production data, and construct a process feature data set containing process identifier field, time series field and process feature parameter field. The data identification module performs cross-process correlation analysis on the process characteristic parameters of different production processes based on multiple process characteristic data sets corresponding to the same bearing individual, and generates abnormal data identification information when the correspondence of process characteristic parameters in process sequence or time sequence deviates. The bearing determination module outputs the quality inspection results for the corresponding individual bearing based on the presence or distribution of abnormal data identification information.

2. The bearing production quality inspection system based on production data analysis according to claim 1, characterized in that, The bearing manufacturing process includes grinding, ultra-precision machining, and assembly.

3. The bearing production quality inspection system based on production data analysis according to claim 2, characterized in that, The steps for the production data acquisition module to acquire production data include: When a bearing enters the grinding, ultra-precision and assembly processes, the target process of the bearing is determined based on the current production process status, and a process identifier field that uniquely corresponds to the target process is generated accordingly. During the execution of the target process, production data reflecting the status of bearing processing or assembly is continuously collected according to the process rhythm of the target process, so as to form a production data sequence that can reflect the changes in the process execution. While collecting production data sequences, a corresponding time identifier field is generated for each piece of production data based on a unified time benchmark to maintain the consistency of the time sequence of production data within the same process. The process identifier field, time identifier field, and corresponding production data sequence are structurally combined to form the basic unit of production data for subsequent cross-process correlation analysis.

4. The bearing production quality inspection system based on production data analysis according to claim 3, characterized in that, The steps for the data binding module to bind production data to individual bearings include: Before the production data corresponding to an individual bearing is written into the production data set, obtain the unique identification information of the individual bearing and generate the bearing index key; Receive production data output from each production process, which includes process identifier and time identifier fields, and categorize the production data according to the process identifier field; Based on the process classification, the production data under the same process classification are arranged in chronological order according to the time identifier field, forming an ordered production data sequence for the corresponding production process; The ordered production data sequence of each production process is bound to the bearing index key and written into the production data set, so that the production data set can use the bearing index key as an index to realize the cross-process production data collection of the same bearing individual.

5. The bearing production quality inspection system based on production data analysis according to claim 1, characterized in that, The steps for the data parsing module to construct a set of process feature data include: Read the production data set indexed by individual bearings, and divide the production data into a subset of grinding process production data, a subset of ultra-precision process production data, and a subset of assembly process production data based on the process identifier field; In each subset of production data, the production data is arranged in chronological order according to the time identifier field to form the time sequence field of the corresponding production process. For time series fields, the changes between adjacent sampling points are calculated and the distribution range of the changes is statistically analyzed to generate process characteristic parameters that characterize the stability of the production process; The process identifier field, time series field, and process feature parameter field are structured and combined to form a process feature data set for the corresponding bearing individual.

6. The bearing production quality inspection system based on production data analysis according to claim 5, characterized in that, Methods for obtaining process characteristic parameters include: Read the subset of production data corresponding to the same bearing and form a time series field based on the time identifier field; In the time series field, the process execution is divided into at least two consecutive segments according to the preset segmentation rules. The consecutive segments correspond to different stages of the process execution. Within each continuous segment, the change between adjacent sampling points is calculated, and segment stability parameters are generated based on the continuous change characteristics of the change within the continuous segment. The stability parameters of each continuous segment are summarized to obtain the process characteristic parameters that characterize the stability of the production process. The process identification field, time series field and process characteristic parameter field are then combined in a structured manner to form the process characteristic data set of the corresponding bearing individual.

7. The bearing production quality inspection system based on production data analysis according to claim 1, characterized in that, When performing cross-process correlation analysis on process characteristic parameters of different production processes, the data identification module includes: According to the established sequence of production processes, the process feature data set formed by the same bearing in the grinding process, ultra-precision process and assembly process is read in sequence. The process characteristic parameters corresponding to each production process are connected sequentially according to the production process order to form the process characteristic parameter evolution path describing the evolution state of the individual bearing in multiple production processes. Along the evolution path of process characteristic parameters, the process characteristic parameters between adjacent production processes are correlated segment by segment, so that the cross-process correlation analysis is limited to the evolution path of process characteristic parameters, thereby avoiding correlation analysis of process characteristic parameters of non-adjacent production processes or processes without manufacturing sequence relationship.

8. The bearing production quality inspection system based on production data analysis according to claim 7, characterized in that, Segment-by-segment association processing, including: According to the established sequence of production processes, the evolution path of the process characteristic parameters corresponding to the same bearing individual is divided into multiple associated segments consisting of adjacent production processes. Each associated segment contains the process characteristic parameters of the preceding production process and the process characteristic parameters of the following production process. For each associated segment, based on the process characteristic parameters of the preceding production process and the process characteristic parameters of the following production process, cross-process correspondence features corresponding to that associated segment are formed respectively. The cross-process correspondence features formed by each associated segment are recorded in an orderly manner according to the production process sequence to obtain a set of segment-by-segment association results. Based on the result set of segment-by-segment association, determine whether the correspondence between process feature parameters and process sequence or time sequence has deviated.

9. The bearing production quality inspection system based on production data analysis according to claim 8, characterized in that, The criteria used by the data recognition module to determine whether process characteristic parameters have deviated include: Along the evolution path of the process characteristic parameters corresponding to the same bearing individual, and in accordance with the predetermined order of the production process, the process characteristic parameter pairs corresponding to adjacent production processes are selected in sequence; For each pair of process characteristic parameters, the correspondence between the process characteristic parameters of the preceding production process and the process characteristic parameters of the subsequent production process in terms of the direction of numerical change, continuity of change, and stage stability is determined. The correspondence characteristics are recorded in an orderly manner according to the production process sequence to form a set of structural descriptions of the evolution path of process characteristic parameters. The correspondence features in the structural description set are used as the correspondence judgment benchmark to characterize the cross-process evolution behavior of process feature parameters of individual bearings under normal manufacturing conditions; When the correspondence between the subsequently acquired process feature parameters and the process sequence or time sequence fails to match the correspondence judgment benchmark, it is determined that the correspondence between the process feature parameters has deviated, and corresponding abnormal data identification information is generated.

10. The bearing production quality inspection system based on production data analysis according to claim 1, characterized in that, The quality inspection results output by the bearing determination module include: When the abnormal data identification information read indicates that at least one production process-related segment has deviated abnormally, the quality inspection result indicating that the bearing has a quality risk is output, and the location information of the production process-related segment corresponding to the abnormal deviation is output simultaneously. When no abnormal data identifier information indicating abnormal deviation of the production process-related segment is read, the quality inspection result indicating that the quality of the individual bearing is normal is output.

Images