CN-122007987-A - Polishing process optimization method and system combining big data analysis

Abstract

The embodiment of the application provides a polishing process optimization method and a polishing process optimization system combining big data analysis, which relate to the technical field of big data, and are characterized by firstly integrating basic workpiece attributes, operation of polishing equipment, process parameter execution and polishing effect feedback data, establishing a polishing process multisource data linkage set, mining dynamic coupling relations of different process parameters based on the polishing process multisource data linkage set to form a coupling rule set, constructing a prediction model taking the basic workpiece attributes and initial process parameters as input and predicting polishing effects as output according to the process parameter coupling rule set and historical polishing effect feedback data, inputting basic workpiece attribute data to be polished, iteratively adjusting the process parameters by combining real-time equipment operation data, generating a dynamic optimization parameter sequence, and finally outputting a complete optimization scheme integrating equipment operation, process time sequence and quality control nodes. The application comprehensively utilizes data, accurately optimizes the polishing process and improves the quality and efficiency.

Inventors

• Hong Wanxian
• ZHANG XIHU
• LU XIAOWU

Assignees

• 广东粤轻卫浴科技有限公司

Dates

Publication Date

20260512

Application Date

20260123

Claims (10)

1. 1. A polishing process optimization method in combination with big data analysis, the method comprising: Combining basic attribute data of a workpiece, operation data of polishing equipment, process parameter execution data and polishing effect feedback data to establish a polishing process multi-source data linkage set, wherein the polishing process multi-source data linkage set comprises real-time synchronous links and historical association indexes of all data types; Based on the polishing process multisource data linkage set, dynamic coupling relations among different process parameters are mined to form a process parameter coupling rule set, and the process parameter coupling rule set comprises parameter variation collaborative logic and parameter adaptation constraint conditions; According to the technological parameter coupling rule set and the historical polishing effect feedback data, a polishing effect prediction model is constructed, and the polishing effect prediction model takes workpiece basic attribute data and initial technological parameters as input and outputs predicted polishing effect data; Inputting basic attribute data of a workpiece to be polished into the polishing effect prediction model, and iteratively adjusting process parameters by combining operation data of real-time polishing equipment to generate a dynamic optimization parameter sequence; based on the dynamic optimization parameter sequence, integrating the equipment operation adaptation requirement, the process execution time sequence and the quality control node, and outputting a complete optimization scheme of the polishing process.
2. 2. The polishing process optimization method in combination with big data analysis according to claim 1, wherein the mining the dynamic coupling relation between different process parameters based on the polishing process multi-source data linkage set to form a process parameter coupling rule set comprises: Extracting all historical process parameter execution data from the polishing process multisource data linkage set, dividing parameter categories according to polishing process stages, wherein each parameter category corresponds to a core execution link in the polishing process; extracting specific process parameter items under each parameter category, and recording complete change tracks of each process parameter item under different polishing scenes, wherein the change tracks comprise parameter initial values, intermediate adjustment values and final stable values; Associating technological parameter item change tracks of different parameter categories under the same polishing scene, and establishing a parameter change time axis to enable the change nodes of all the parameter items to realize synchronous alignment in the time dimension; analyzing the change association between different process parameter items in a single parameter category, tracking the influence path of the numerical adjustment of one process parameter item on other process parameter items in the same category, and recording the time delay and change amplitude association generated by influence; Analyzing the cross influence of process parameter items among different parameter categories, extracting cross-category parameter adjustment association cases for adjusting process parameter items in other parameter categories caused by the change of the process parameter items in one parameter category, generating association rules for describing the influence relation among the parameter categories based on the cross-category parameter adjustment association cases; Aiming at polishing time related parameter items and polishing liquid type related parameter items, comparing the adapting process cases under different workpiece shapes, determining parameter adapting constraint conditions, and recording application ranges and limiting boundaries of the combined use of the polishing time related parameter items and the polishing liquid type related parameter items; Integrating parameter association logic, association rules, parameter variation cooperative logic and parameter adaptation constraint conditions in a single category to form a preliminary parameter coupling rule; traversing historical polishing process cases in the polishing process multisource data linkage set, and verifying the applicability of the preliminary parameter coupling rules by using actual parameter change and effect feedback data; optimizing a preliminary parameter coupling rule according to a verification result, supplementing a parameter coupling special adaptation item under a special working condition, and correcting rule contents which are inconsistent with an actual polishing process case; classifying and sorting the optimized parameter coupling rules according to parameter types and coupling types to form a structured process parameter coupling rule set, adding scene identifications for the process parameter coupling rule set, marking the applicable ranges of workpiece materials, shapes and equipment types corresponding to each parameter coupling rule, wherein the process parameter coupling rule set comprises parameter variation collaborative logic and parameter adaptation constraint conditions.
3. 3. The polishing process optimization method in combination with big data analysis according to claim 1, wherein the coupling rule set and the historical polishing effect feedback data according to the process parameters construct a polishing effect prediction model, the polishing effect prediction model takes workpiece basic attribute data and initial process parameters as inputs, outputs predicted polishing effect data, and comprises: Extracting historical polishing effect feedback data from the polishing process multisource data linkage set, splitting data content according to polishing quality evaluation dimensions, wherein each polishing quality evaluation dimension corresponds to a core polishing effect index; Correlating the historical polishing effect feedback data with corresponding process parameter execution data and workpiece basic attribute data to form a correlation data set corresponding to each polishing process case; extracting parameter coupling rules corresponding to each associated data set from the technological parameter coupling rule set, and labeling collaborative logic and constraint conditions followed by parameter changes in the associated data sets; Constructing a model basic structure, wherein the model basic structure comprises an attribute characteristic input layer, a parameter coupling rule embedding layer, an effect index prediction layer and a result output layer, and information circulation is realized among the layers through a data transmission channel; converting material characteristics and shape characteristics in the basic attribute data of the workpiece into characteristic vectors which can be identified by the model, and taking the characteristic vectors as a standard input format of an attribute characteristic input layer; Converting the process parameter execution data into parameter vectors according to parameter types, fusing the parameter vectors with corresponding parameter coupling rule codes to form parameter-rule fusion vectors, and inputting the parameter coupling rule embedding layers; in the parameter coupling rule embedding layer, processing attribute feature vectors and parameter-rule fusion vectors through a feature interaction algorithm, and mining nonlinear association between the attribute feature vectors and the parameter-rule fusion vectors to generate an association feature matrix; Inputting the associated feature matrix into an effect index prediction layer, wherein the effect index prediction layer independently predicts each polishing effect index through a multidimensional regression algorithm to generate a single index prediction value, integrates all single index prediction values to form complete predicted polishing effect data, and outputs the complete predicted polishing effect data according to a preset format through a result output layer; Collecting a plurality of associated data sets as training samples, inputting the training samples into a model basic structure for iterative training, and updating algorithm parameters and characteristic interaction weights of each layer through the iterative training; Setting a cross verification process, dividing a training sample into a plurality of subsets, selecting a hierarchical structure and an algorithm type of a model foundation structure based on training and verification results of different subsets, combining the trained model foundation structure, updated algorithm parameters and verification selected structural configuration, constructing a polishing effect prediction model, and outputting predicted polishing effect data by taking workpiece foundation attribute data and initial process parameters as inputs.
4. 4. The polishing process optimization method combined with big data analysis according to claim 3, wherein in the parameter coupling rule embedding layer, the attribute feature vector and the parameter-rule fusion vector are processed through a feature interaction algorithm, nonlinear association between the attribute feature vector and the parameter-rule fusion vector is mined, and an association feature matrix is generated, and the method comprises the following steps: performing dimension expansion processing on the attribute feature vector of the input parameter coupling rule embedding layer to enable the dimension of the attribute feature vector to be consistent with the dimension of the parameter-rule fusion vector, mapping the attribute feature vector subjected to dimension expansion to a high-dimensional feature space by adopting a feature mapping algorithm, and converting the attribute feature vector into a high-dimensional attribute feature representation; weighting the characteristic components corresponding to the parameter coupling rule codes in the parameter-rule fusion vector through weighting operation so as to increase the weight of the rule features in the feature interaction; constructing a two-way feature interaction channel, wherein one feature interaction channel is used for processing direct interaction of attribute features and parameter features, and the other feature interaction channel is used for processing indirect interaction of attribute features and rule features; In the feature interaction channel of the indirect interaction, the matrix level interaction of the high-dimensional attribute feature representation and the regular feature part in the weighted parameter-rule fusion vector is realized through matrix multiplication operation, so as to generate an indirect interaction feature matrix; performing weight distribution on the direct interaction feature vector and the indirect interaction feature matrix by adopting a feature weighting algorithm, wherein the weight value is set based on the influence degree of attribute features, parameter features and rule features in the historical data on the polishing effect; Performing dimension splicing on the weighted direct interaction feature vector and the indirect interaction feature matrix to form a preliminary interaction feature set, wherein the preliminary interaction feature set contains feature information in different interaction modes; The method comprises the steps of selecting features with the correlation with a polishing effect prediction target higher than a preset threshold value from a preliminary interaction feature set through a feature selection algorithm as key features, grouping the key features according to the correlation degree of the key features with different polishing effect indexes to form grouped key features; the key features after grouping are converted into a structured associated feature matrix by adopting a matrix reconstruction algorithm, and an associated feature matrix containing all key interaction features is generated, wherein the associated feature matrix is used for reflecting nonlinear association among attribute features, parameter features and rule features, the row dimension of the associated feature matrix corresponds to feature categories, the column dimension corresponds to sample identifications, and matrix elements are feature values.
5. 5. The polishing process optimization method in combination with big data analysis according to claim 1, wherein inputting the basic attribute data of the workpiece to be polished into the polishing effect prediction model, iteratively adjusting the process parameters in combination with the real-time polishing equipment operation data, generating a dynamic optimization parameter sequence, comprises: collecting basic attribute data of a workpiece to be polished, wherein the basic attribute data comprises workpiece material composition information, shape structure information and initial surface state information, and the dimension of the basic attribute data is consistent with that of the workpiece basic attribute data in the polishing process multisource data linkage set; Converting basic attribute data of a workpiece to be polished into standard feature vectors, and preprocessing the data according to the input format requirement of the polishing effect prediction model; Setting an initial process parameter set of a workpiece to be polished, wherein the initial process parameter set comprises polishing pressure related parameters, polishing speed related parameters, polishing time related parameters and polishing liquid type related parameters, and the initial parameter value is set based on conventional polishing process experience; Inputting a standard feature vector and an initial technological parameter set into the polishing effect prediction model, wherein the polishing effect prediction model outputs corresponding predicted polishing effect data, and the predicted polishing effect data comprises predicted values of various polishing effect indexes; Collecting real-time operation data after the polishing equipment is started, wherein the real-time operation data comprises equipment operation state information, parameter actual execution value information and equipment load information, and the dimension of the real-time operation data is consistent with the dimension of the equipment operation data in the polishing process multisource data linkage set; Comparing the actual execution value of the parameters in the real-time operation data with the parameter values in the initial process parameter set, extracting deviation information between the actual execution value of the parameters and the parameter values, matching the deviation information based on the process parameter coupling rule set, identifying the deviation type and deducing the influence trend of the deviation type on the polishing effect; determining technological parameter items to be adjusted according to the deviation type, the influence trend, the predicted polishing effect data and the technological parameter coupling rule set, and calculating an adjustment direction according to parameter change cooperative logic and an adaptation constraint condition; Calculating the adjustment amplitude of each technological parameter item to be adjusted, wherein the adjustment amplitude is set based on the deviation degree, the difference between the predicted polishing effect and the expected effect and the constraint requirement in the parameter coupling rule; Parameter updating is carried out on the initial technological parameter set according to the adjusting direction and the adjusting amplitude, an adjusted technological parameter set is generated, and the adjusted technological parameter set is input into the polishing effect prediction model to obtain new predicted polishing effect data; Continuously collecting running data of the real-time polishing equipment, and repeating the steps of parameter comparison, deviation analysis, adjustment direction determination, adjustment amplitude calculation and parameter updating to form a process parameter iteration adjustment loop; Setting iteration termination conditions, stopping iteration adjustment when the predicted polishing effect data meets a preset effect threshold value and the parameter deviation in the real-time equipment operation data is in a preset range, integrating all adjusted process parameter sets in the iteration process, and sequencing according to the adjustment time sequence and the polishing procedure stage to form a dynamic optimization parameter sequence comprising a parameter change track, an adjustment triggering condition and a corresponding equipment operation state.
6. 6. The method according to claim 5, wherein calculating the adjustment range of each item of the process parameters to be adjusted, the adjustment range being set based on the deviation degree, the difference between the predicted polishing effect and the expected effect, and the constraint requirements in the parameter coupling rule, comprises: Constructing an adjustment amplitude calculation flow, wherein the adjustment amplitude calculation flow comprises a deviation influence quantization step, an effect difference quantization step, a rule constraint quantization step and a comprehensive calculation step, and the adjustment amplitude calculation is completed by cooperation of the steps; in the deviation influence quantization step, converting the deviation degree of the actual execution value of the parameter in the real-time operation data and the initial process parameter into a quantization index, wherein the quantization index is used for reflecting the influence of the deviation on the process execution; In the effect gap quantization step, calculating the gap between each index predicted value in the predicted polishing effect data and a preset expected effect index value, converting the gap into a standardized effect gap quantized value, wherein the size of the quantized value corresponds to the obvious degree of the effect gap; in the rule constraint quantization step, extracting an adaptive constraint condition related to a process parameter item to be adjusted in the process parameter coupling rule set, and converting the constraint condition into a quantization constraint value, wherein the quantization constraint value is used for reflecting the allowable range of parameter adjustment; Carrying out standardization processing on the deviation influence quantization index, the effect gap quantization value and the rule constraint quantization value based on the unified evaluation scale to obtain a standardized deviation influence value, an effect gap value and a rule constraint value; Setting weight coefficients of a deviation influence value, an effect gap value and a rule constraint value, wherein the weight coefficients are set based on the importance degree of each factor on parameter adjustment and history adjustment experience; multiplying the standardized deviation influence value, the effect difference value and the rule constraint value with the corresponding weight coefficients respectively to obtain weighted evaluation results of all the factors, and carrying out summation operation on the weighted evaluation results of all the factors in a comprehensive calculation step to obtain a comprehensive evaluation value of parameter adjustment, wherein the comprehensive evaluation value is used for reflecting the necessity and adjustment space of parameter adjustment; constructing an adjustment amplitude mapping function, wherein the adjustment amplitude mapping function takes the comprehensive evaluation value as input, takes the parameter adjustment amplitude as output, and sets the corresponding relation between the comprehensive evaluation value and the actual adjustment amplitude in the process case based on a large number of historical parameter adjustment; inputting the comprehensive evaluation value of parameter adjustment into an adjustment amplitude mapping function to obtain a preliminary adjustment amplitude value, wherein the preliminary adjustment amplitude value is the reference amplitude of parameter adjustment; Performing boundary verification on the primary adjustment amplitude value by combining the current numerical range of the technological parameter item to be adjusted, so that the adjusted parameter value is between the maximum range and the minimum range allowed by the technology; and correcting the primary adjustment amplitude value according to the boundary verification result, if the primary adjustment amplitude value leads to the parameter exceeding the boundary, correcting the adjustment amplitude value into a difference value or a sum value of the boundary value and the current parameter value, referring to historical parameter adjustment amplitude data of the same type of workpieces, fine-tuning the corrected adjustment amplitude value, enabling the adjustment amplitude to be more fit with the actual process execution requirement, and determining the final adjustment amplitude of each process parameter item to be adjusted.
7. 7. The polishing process optimization method in combination with big data analysis according to claim 1, wherein the integrating equipment operation adaptation requirements, process execution timing and quality control nodes based on the dynamic optimization parameter sequence outputs a complete polishing process optimization scheme, comprising: extracting parameter configuration information of each process stage from the dynamic optimization parameter sequence, dividing process execution units according to the sequence of the polishing procedure, wherein each process execution unit corresponds to a group of parameter configuration and one process execution stage; Analyzing the parameter configuration information of each process execution unit, combining the equipment operation data in the polishing process multisource data linkage set, retrieving equipment operation records matched with each parameter configuration from the historical equipment operation data, and extracting a standard operation action sequence and parameter set values from the equipment operation records; generating a device control instruction set of each process execution unit based on the standard operation action sequence and the parameter set value, wherein the device control instruction set comprises a designated device running mode, the standard operation action sequence and a parameter calibration target value; Planning a process execution time sequence according to the time node of parameter adjustment in the dynamic optimization parameter sequence and the sequence of the process execution units, and marking the starting time, the duration time and the connection time of each process execution unit; Extracting historical polishing quality control data from the polishing process multisource data linkage set, determining quality control nodes of each process execution unit by combining parameter change nodes in a dynamic optimization parameter sequence, and labeling control objects and control modes; Configuring a corresponding quality detection program and a detection triggering condition for each quality control node; Integrating parameter configuration information, equipment operation adaptation requirements, process execution time sequence and quality control nodes of each process execution unit to form a process execution unit scheme, wherein each process execution unit scheme corresponds to a complete process execution link; Extracting equipment maintenance data corresponding to a dynamic optimization parameter sequence from the polishing process multisource data linkage set, and combining the process execution main flow to formulate equipment preposed maintenance suggestions and process maintenance plans; According to the main process of process execution, equipment preposed maintenance proposal and process maintenance plan, estimating human resources, time resources and material resources required by the polishing process execution to form a resource demand list; Establishing an emergency treatment flow in the process of executing the process, and defining triggering conditions, executing steps and an executing main body of a countermeasure aiming at possible equipment faults, parameter deviations and quality anomalies; Integrating the process execution main flow, the equipment operation adaptation requirement, the quality control node, the equipment maintenance suggestion, the resource demand list and the emergency treatment flow to form a complete optimization scheme of the polishing process, and outputting the complete optimization scheme according to a standardized format.
8. 8. The polishing process optimization method in combination with big data analysis according to claim 7, wherein the emergency treatment flow in the execution process of the construction process, for possible equipment failure, parameter deviation, quality abnormality, explicit countermeasure and responsibility division, comprises: Extracting fault case data in the historical polishing process from the polishing process multisource data linkage set, and counting possible equipment fault types, parameter deviation types and quality abnormality types to form a fault abnormality type list; For each equipment fault type, analyzing common reasons, performance characteristics and influence ranges of the faults on process execution, and refining effective fault processing steps based on historical processing experience; aiming at each parameter deviation type, analyzing triggering conditions, variation trends and influence degrees on polishing effects generated by deviation, and combining the technological parameter coupling rule set to formulate deviation correction measures; Aiming at each quality anomaly type, analyzing key nodes, propagation paths and influence results on the final polishing quality, and making anomaly containment measures and quality remedy schemes; setting response priority corresponding to each fault abnormality type, wherein the response priority is set based on the influence degree of fault abnormality on process progress, quality and cost; dividing emergency treatment responsibility departments and responsibility personnel, and marking fault exception handling ranges and responsibility authorities corresponding to each responsibility main body; Defining an emergency treatment process, wherein the emergency treatment process comprises an automatic logic chain for fault exception triggering, measure matching and execution and result feedback processing; Setting an emergency treatment time threshold, and labeling the longest allowable time of treatment aiming at each fault exception type; The method comprises the steps of performing fault processing, deviation correction measures, exception restraining measures, response priority, execution main body definition, emergency processing flow and time threshold value, generating a preliminary emergency processing rule base, updating the preliminary emergency processing rule base based on historical processing data, adding special scene rules, correcting execution logic, and generating an emergency processing rule base containing all fault exception type processing schemes.
9. 9. The polishing process optimization method in combination with big data analysis according to claim 2, wherein analyzing the cross influence of process parameter items between different parameter categories, extracting cross-category parameter adjustment association cases in which the change of process parameter items in one parameter category causes the adjustment of process parameter items in other parameter categories, and generating association rules describing influence relationships between parameter categories based on the cross-category parameter adjustment association cases, comprises: Screening a complete polishing process case containing a plurality of parameter type process parameter item change records from the polishing process multisource data linkage set, wherein the complete polishing process case contains a definite parameter change time node and a corresponding adjustment record; Classifying and marking the technological parameter items in each complete polishing technological case according to the parameter types, and marking the parameter types to which each technological parameter item belongs and the functional positioning in the process execution; Constructing a parameter change tracking map, marking the numerical value change of each process parameter item at different time nodes by taking time as a horizontal axis and taking a parameter category as a vertical axis, and intuitively presenting a parameter change track; Identifying a first change node of a process parameter item in a parameter category from a parameter change tracking map, and taking the first change node as a starting point of cross influence analysis; tracking the numerical value change of the process parameter item in other parameter categories after the starting point, and recording the category, the adjustment time and the adjustment amplitude of the first process parameter item with numerical value adjustment; Analyzing the association relation between the initial parameter change and the subsequent parameter adjustment, extracting cross-category parameter adjustment association cases directly caused by the initial parameter change, and recording parameter category combinations, initial parameter change characteristics, subsequent parameter adjustment characteristics and corresponding polishing scene information in the cross-category parameter adjustment association cases; classifying cross-class parameter adjustment associated cases, and dividing case types according to a combination form of a starting parameter class and an affected parameter class, wherein each case type comprises a plurality of similar cross-class parameter adjustment associated cases; In each case type, counting the corresponding relation between the initial parameter variation amplitude and the subsequent parameter adjustment amplitude, and calculating and recording the numerical proportion relation between the initial parameter variation amplitude and the subsequent parameter adjustment amplitude and the variation mode thereof; analyzing differences of cross-category parameter influences under different polishing scenes in the same case type, extracting the adjusting effect of scene characteristics on the influence degree, and marking the influence weights of scene factors; Counting the common data modes of cross-category parameter influence in each case type, wherein the common data modes comprise influence delay time, adjustment amplitude proportion relation and parameter change direction relation, and generating a preliminary influence association rule based on the common data modes; And integrating the preliminary influence association rules of all case types, supplementing influence special adaptation items in special scenes, and generating a cross-category parameter influence association rule set, wherein the cross-category parameter influence association rule set is used for describing the cross-category influence relation of technological parameter items among different parameter categories.
10. 10. A polishing process optimization system incorporating big data analysis, comprising a processor and a readable storage medium storing a program which, when executed by the processor, implements the polishing process optimization method incorporating big data analysis of any one of claims 1-9.

Description

Polishing process optimization method and system combining big data analysis Technical Field The application relates to the technical field of big data, in particular to a polishing process optimization method and system combining big data analysis. Background In the field of polishing processes, with the continuous increase of requirements of industrial production on product quality and efficiency, the conventional polishing process optimization method gradually exposes various limitations. Currently, many polishing process optimizations rely primarily on experienced operators who rely on long-term personal accumulation of experience to adjust process parameters in order to achieve the desired polishing effect. However, the above-described manner of relying on human experience has significant drawbacks. On the one hand, the manual experience has subjectivity and limitation, and different operators can have different understanding and adjusting modes of process parameters, so that the stability and consistency of polishing effect are difficult to ensure. On the other hand, with the continuous increase of the types and the complexity of products, the influence of various factors on the polishing effect is difficult to comprehensively consider only by artificial experience, and the fine process optimization is difficult to realize. In addition, some existing polishing process optimization methods collect partial data, but the data source is single, and only the process parameters are usually focused, so that comprehensive utilization of various information such as basic properties of workpieces, running states of polishing equipment and feedback of polishing effects is omitted. The optimization mode of the single data source cannot comprehensively and accurately reflect the complex relation in the polishing process, and potential dynamic coupling rules among process parameters are difficult to mine, so that improvement of the optimization effect of the polishing process is limited, and the requirements of modern industrial production on the high-quality and high-efficiency polishing process cannot be met. Disclosure of Invention In view of the above, the present application aims to provide a polishing process optimization method and system combining big data analysis. According to a first aspect of the present application, there is provided a polishing process optimization method in combination with big data analysis, the method comprising: Combining basic attribute data of a workpiece, operation data of polishing equipment, process parameter execution data and polishing effect feedback data to establish a polishing process multi-source data linkage set, wherein the polishing process multi-source data linkage set comprises real-time synchronous links and historical association indexes of all data types; Based on the polishing process multisource data linkage set, dynamic coupling relations among different process parameters are mined to form a process parameter coupling rule set, and the process parameter coupling rule set comprises parameter variation collaborative logic and parameter adaptation constraint conditions; According to the technological parameter coupling rule set and the historical polishing effect feedback data, a polishing effect prediction model is constructed, and the polishing effect prediction model takes workpiece basic attribute data and initial technological parameters as input and outputs predicted polishing effect data; Inputting basic attribute data of a workpiece to be polished into the polishing effect prediction model, and iteratively adjusting process parameters by combining operation data of real-time polishing equipment to generate a dynamic optimization parameter sequence; based on the dynamic optimization parameter sequence, integrating the equipment operation adaptation requirement, the process execution time sequence and the quality control node, and outputting a complete optimization scheme of the polishing process. According to a second aspect of the present application, there is provided a polishing process optimization system incorporating big data analysis, the polishing process optimization system incorporating big data analysis including a processor and a readable storage medium storing a program which, when executed by the processor, implements the foregoing polishing process optimization method incorporating big data analysis. According to any one of the aspects, the polishing process multisource data linkage set is constructed by integrating workpiece basic attribute data, polishing equipment operation data, process parameter execution data and polishing effect feedback data, dynamic coupling relations among different process parameters are mined based on the polishing process multisource data linkage set, a process parameter coupling rule set is formed, cooperative logic and adaptation constraint conditions among the parameters are revealed, a polishing effect pr