System for monitoring and early warning based on big data and automatic root cause analysis in photovoltaic cell production
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-05-21
- Publication Date
- 2026-08-11
AI Technical Summary
当生产出现异常，如良率下降、电性能参数偏移或出现特定不良类型时，传统监控方法面临巨大挑战：
Smart Images
Figure CN122549922A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of photovoltaic cell production technology, and in particular to a system for photovoltaic cell production based on big data monitoring, early warning, and automatic root cause analysis. Background Technology
[0002] The photovoltaic cell manufacturing process is a precision manufacturing process involving multiple complex steps such as texturing, diffusion, etching, deposition, and printing. This process is characterized by: complex processes (e.g., tubular, trough, chain, etc.), numerous parameters (including process parameters such as temperature, pressure, and flow rate, as well as testing parameters such as film thickness, sheet resistance, and efficiency), and a massive amount of data (derived from high-frequency acquired equipment data and material traceability data throughout the entire process). When production anomalies occur, such as decreased yield, deviations in electrical performance parameters, or the emergence of specific defect types, traditional monitoring methods face significant challenges. 1) Fragmented multi-source data: Process parameter data (such as furnace tube temperature), detection data (such as film thickness, sheet resistance, etc.), and equipment status data (such as alarm information) may be scattered in different systems, making data correlation analysis very difficult.
[0003] 2) Low analysis efficiency: Traditional analysis methods are based on post-incident problems, requiring manual data extraction from various reporting systems (such as MES and SPC), followed by data cleaning, integration, and visualization analysis. This process is time-consuming and labor-intensive, often taking hours or even days, and cannot meet the real-time and timely requirements of the production site, thus missing the best opportunity for handling the problem.
[0004] 3) Difficulty in problem localization: The root cause of the anomaly may originate from equipment or process parameter drift in the preceding process, or from a problem with the raw materials in the current batch. Due to the long production chain, data from each link is isolated, and there is a lack of effective correlation analysis methods, making the troubleshooting process like "finding a needle in a haystack," heavily reliant on the personal experience of engineers.
[0005] 4) Lack of insight: Existing quality control is mostly limited to monitoring single-point indicators (such as SPC control charts) or simple threshold alarms. It lacks in-depth correlation and mining of multi-dimensional and cross-dimensional data, and cannot automatically form root cause analysis conclusions with guiding significance. It is also difficult to effectively precipitate and replicate expert experience. Summary of the Invention
[0006] To address at least one technical problem in the prior art, embodiments of the present invention provide a system for photovoltaic cell production based on big data monitoring, early warning, and automatic root cause analysis. This system integrates multi-dimensional production data to achieve real-time monitoring, automatic early warning, and root cause analysis functions, and can automatically generate root cause analysis reports. To achieve the above technical objectives, the technical solution adopted by embodiments of the present invention is as follows: This invention provides a system for photovoltaic cell production based on big data monitoring, early warning, and automatic root cause analysis, comprising: The data access module is used to obtain multi-dimensional production data of photovoltaic cells from the production management system; after cleaning and integrating the multi-dimensional production data, it is stored to form a battery production data table. Each record in the battery production data table is a multi-dimensional battery production record of a photovoltaic cell. The multi-dimensional early warning module is used to configure and parse early warning rules. It executes early warning rules on multi-dimensional cell production records of photovoltaic cells and triggers early warning actions when the triggering conditions configured in the early warning rules are met. Early warning actions include generating early warning events and calling root cause analysis services. The root cause analysis and report generation module is used to configure analysis templates and provide root cause analysis services. When an early warning action is triggered, the root cause analysis service loads the corresponding analysis template according to the type of the early warning event, performs multi-dimensional root cause analysis on the multi-dimensional battery production records corresponding to the early warning event, and generates a root cause analysis report. The visualization application module is used to send notification information based on early warning events, provide a visual interface of production status, and output root cause analysis reports.
[0007] Furthermore, the data access module includes: The data acquisition unit is used to acquire multi-dimensional production data of photovoltaic cells from multiple production management systems; The data cleaning unit is used to perform deduplication, noise reduction, unified format conversion, and missing value processing on the acquired photovoltaic cell production data. The data integration unit is used to associate, splice, and aggregate photovoltaic cell production data based on a unique key, resulting in a complete multi-dimensional cell production record for each photovoltaic cell. The unique key for a photovoltaic cell is one or a combination of at least two of the following: cell identifier, batch number, and equipment identifier. The data storage unit is used to generate a battery production data table based on the multi-dimensional battery production records of each photovoltaic cell and store it in the database.
[0008] Furthermore, the multi-dimensional early warning module includes: The early warning rule configuration unit is used to provide an early warning rule configuration interface, and to configure and save early warning rules in the early warning rule library. The early warning rule includes the polling cycle, applicable objects and corresponding triggering conditions. The triggering conditions are based on multi-dimensional production data in the multi-dimensional battery production record and corresponding judgment logic settings. The early warning rule parsing unit is used to load early warning rules and generate executable rule instances through the rule parser. The early warning rule execution unit is used to obtain multi-dimensional battery production records of photovoltaic cells according to a preset polling cycle, and execute the judgment logic of multi-dimensional production data in the early warning rule for the multi-dimensional battery production records of photovoltaic cells; when the triggering condition is met, the early warning action is triggered.
[0009] Furthermore, the warning rule also includes embedded first-type expert experience; the first-type expert experience includes possible reasons when the triggering condition is met, and preset actions that can be performed; when the triggering condition is met, the warning action also includes performing preset actions.
[0010] Furthermore, the root cause analysis and report generation module includes: The analysis template configuration unit provides an analysis template configuration interface for users to select analysis components from a pre-set analysis component library; it also allows users to arrange the selected analysis components according to the type of early warning event and configure them into analysis links in sequence; and it saves the configured analysis links as analysis templates. The root cause analysis unit is used to provide root cause analysis services. When an early warning action is triggered, it matches and loads the corresponding analysis template according to the type of the early warning event. According to the order of each analysis component in the analysis template, it calls each analysis component in sequence. Each analysis component performs multi-dimensional root cause analysis on the multi-dimensional battery production records corresponding to the early warning event to obtain the root cause analysis results. The report generation unit is used to assemble the root cause analysis results of each analysis component as analysis fragments to generate a root cause analysis report.
[0011] Furthermore, each analysis component performs multi-dimensional root cause analysis, including data query and judgment analysis and / or algorithm analysis; Data query and analysis includes: performing multi-dimensional correlation queries and analysis on multi-dimensional production data in multi-dimensional battery production records corresponding to the early warning event in order to obtain the root cause of the suspected problem; The algorithm analysis includes: using chi-square analysis on multi-dimensional production data in multi-dimensional battery production records to identify dimensions suspected of having problems; and using residual analysis to determine the root causes of the suspected problems for the identified dimensions.
[0012] Furthermore, the chi-square analysis employs the chi-square test formula, as shown in formula (1): (1) in, The chi-square value, Let i be the actual number of defects in the i-th dimension sample. Let be the expected number of defects for the i-th dimension sample; assume that the expected number of defects is the same for all dimensions. If the chi-square value exceeds the preset threshold, the corresponding dimension is suspected of having a problem.
[0013] Furthermore, the residual analysis employs a standardized residual calculation formula, as shown in formula (2): (2) in, To standardize the residual values, Let i be the actual number of defects in the i-th dimension sample. Let be the expected number of defects for the i-th dimension sample; assume that the expected number of defects is the same for all dimensions. If the absolute value of the standardized residual is greater than the preset residual threshold, then the root cause of the suspected problem in that dimension is determined.
[0014] Furthermore, when generating the root cause analysis report, the report generation unit also embeds a second type of expert experience, which includes possible causes and conclusive descriptions when the triggering conditions are met.
[0015] Furthermore, the visualization application module includes: The real-time early warning unit is used to send notification information based on early warning events; An interactive analysis dashboard unit is used to provide a visual production status interface, which is based on multi-dimensional root cause analysis; the production status interface allows users to perform operations on multi-dimensional production data. The report output unit is used to output and present the root cause analysis report.
[0016] The beneficial effects of the technical solution provided by the embodiments of the present invention are as follows: 1) It can integrate multi-dimensional production data from various production management systems to build a complete multi-dimensional battery production record, eliminating the problem of data isolation from multiple sources.
[0017] 2) Improved problem response timeliness: By executing pre-configured early warning rules, multi-dimensional production data can be monitored in real time, triggering early warning actions promptly. This transforms "passive post-event analysis" into "proactive real-time early warning," achieving automated monitoring and early warning based on big data. After triggering an early warning action, the analysis chain in the pre-configured analysis template is automatically executed, and a root cause analysis report is automatically generated, greatly shortening the processing cycle for quality anomalies.
[0018] 3) Automatically locate the root cause of the problem. It can automatically perform multi-dimensional correlation query judgment analysis or algorithm analysis on multi-dimensional battery production records, which automates the problem investigation and reduces the reliance on engineers' experience.
[0019] 4) Deepen the analysis and insight: Through algorithms such as chi-square test and residual analysis, we have achieved correlation mining and analysis of multi-dimensional production data, and provided in-depth analysis conclusions that far exceed simple threshold alarms.
[0020] 5) Improved system analysis flexibility: Through configurable early warning rules and analysis links, it can flexibly adapt to the monitoring and analysis needs of different factories, processes, and products, and has strong scalability and adaptability.
[0021] 6) The analysis threshold has been lowered by embedding expert experience into early warning rules and root cause analysis reports. Ordinary process personnel can also use root cause analysis reports to understand the possible causes of problems in the photovoltaic cell production process, locate the root cause of the problem, and realize the standardization and large-scale reuse of expert experience. Attached Figure Description
[0022] Figure 1 This is a schematic diagram of a system based on big data monitoring and early warning and automatic root cause analysis in photovoltaic cell production according to an embodiment of the present invention.
[0023] Figure 2 This is a schematic diagram of the data access module in an embodiment of the present invention.
[0024] Figure 3 This is a schematic diagram of the multi-dimensional early warning module in an embodiment of the present invention.
[0025] Figure 4 This is a schematic diagram of the root cause analysis and report generation module in an embodiment of the present invention.
[0026] Figure 5 This is a schematic diagram of the visualization application module in an embodiment of the present invention. Detailed Implementation
[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.
[0028] In the description of the embodiments of the present invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the present invention. In addition, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.
[0029] In the description of the embodiments of the present invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can also refer to the internal connection of two components; and they can refer to a wireless connection or a wired connection. Those skilled in the art can understand the specific meaning of the above terms in the present invention based on the specific circumstances.
[0030] Furthermore, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.
[0031] This invention proposes a system for photovoltaic cell production based on big data monitoring, early warning, and automatic root cause analysis, such as... Figure 1 As shown, it includes: The data access module is used to obtain multi-dimensional production data of photovoltaic cells from the production management system; after cleaning and integrating the multi-dimensional production data, it is stored to form a battery production data table. Each record in the battery production data table is a multi-dimensional battery production record of a photovoltaic cell. The multi-dimensional early warning module is used to configure and parse early warning rules. It executes early warning rules on multi-dimensional cell production records of photovoltaic cells and triggers early warning actions when the triggering conditions configured in the early warning rules are met. Early warning actions include generating early warning events and calling root cause analysis services. The root cause analysis and report generation module is used to configure analysis templates and provide root cause analysis services. When an early warning action is triggered, the root cause analysis service loads the corresponding analysis template according to the type of the early warning event, performs multi-dimensional root cause analysis on the multi-dimensional battery production records corresponding to the early warning event, and generates a root cause analysis report. The visualization application module is used to send notification information based on early warning events, provide a visual interface of production status, and output root cause analysis reports.
[0032] This application integrates multi-dimensional production data from various production management systems to construct a complete multi-dimensional battery production record. By executing pre-configured early warning rules, it can monitor multi-dimensional production data in real time, triggering early warning actions promptly. After triggering an early warning action, it automatically executes the analysis links in the pre-configured analysis template, improving the timeliness of problem response. Performing multi-dimensional root cause analysis can automatically conduct multi-dimensional correlation query judgment analysis or algorithm analysis, thus automating problem troubleshooting.
[0033] (a) Data Access Module Data access module such as Figure 2 As shown, it includes: (1.1) Data acquisition unit, used to acquire multi-dimensional production data of photovoltaic cells from multiple production management systems; In this embodiment, the production management system includes a data acquisition system, an EAP (Equipment Automation Program) system, a MES (Manufacturing Execution System) system, and a traceability system. The data acquisition system can directly collect electrical performance data (such as voltage, current, etc.) and process parameter data (such as temperature, pressure, humidity, flow rate, etc.) from equipment sensors, PLCs, and SCADA systems. The EAP system includes pre-processed process parameter data, detection data (such as film thickness, sheet resistance, etc.), and equipment status data (such as alarm information). The MES system includes production modeling data (such as process, equipment, furnace tube, temperature zone, boat, etc. data), raw material data, etc. The traceability system includes flow information data during the photovoltaic cell production process.
[0034] (1.2) Data cleaning unit, used to perform deduplication, noise reduction, unified format conversion and missing value processing on the acquired photovoltaic cell production data; In this embodiment, deduplication of production data involves removing duplicate received production data; denoising of production data involves filtering obviously abnormal or erroneous production data (e.g., data exceeding physical limits); unified format conversion of production data is used to convert production data of different formats (JSON, XML, CSV, time-series data streams) into a unified structure; missing value handling of production data is used to handle missing data values according to a preset strategy, such as filling in null values.
[0035] (1.3) Data integration unit, used to associate, splice and aggregate the production data of photovoltaic cells according to unique keys (such as battery ID, batch number, equipment number, etc.) to obtain a complete multi-dimensional battery production record for each photovoltaic cell.
[0036] (1.4) Data storage unit, used to form a battery production data table based on the multi-dimensional battery production records of each photovoltaic cell and store it in the database.
[0037] In this embodiment, the battery production data table can be stored in databases such as ClickHouse, Redis, and MongoDB to support different application scenarios.
[0038] The data access module can integrate multi-source heterogeneous production data to construct a complete multi-dimensional battery production record, eliminating the problem of multi-source data isolation and facilitating subsequent data correlation analysis.
[0039] (II) Multi-dimensional early warning module Multi-dimensional early warning module such as Figure 3 As shown, it includes: (2.1) Early warning rule configuration unit, used to provide an early warning rule configuration interface, and to configure early warning rules and save early warning rules in the early warning rule library; the early warning rule includes the polling cycle, applicable objects and corresponding triggering conditions; the triggering conditions are based on multi-dimensional production data in multi-dimensional battery production records and corresponding judgment logic settings; In some embodiments, the applicable objects may be production output, yield, etc.; if the applicable object is yield, the corresponding triggering conditions are some production data and judgment logic related to yield, such as raw material batch, certain process parameter data, and corresponding judgment logic.
[0040] (2.2) Early warning rule parsing unit, used to load early warning rules and generate executable rule instances through rule parser; The rule parser can "translate" and "compile" warning rules into executable rule instances that can be executed at high speed; this process transforms flexible business configurations into executable technical instructions, which is the key to achieving both flexibility and high performance.
[0041] (2.3) The early warning rule execution unit is used to obtain multi-dimensional battery production records of photovoltaic cells according to a preset polling cycle, and execute the judgment logic of multi-dimensional production data in the early warning rule for the multi-dimensional battery production records of photovoltaic cells; when the triggering conditions are met, the early warning action is triggered, including generating an early warning event and calling the root cause analysis service; The execution of the early warning rules forms an automated closed loop of "polling - acquiring production data - judgment - early warning action", transforming "post-event passive analysis" into "proactive real-time early warning", and realizing automated monitoring and early warning based on big data.
[0042] In some embodiments, the warning rule also includes an embedded first type of expert experience; the first type of expert experience includes possible reasons when the triggering condition is met, and preset actions that can be performed; when the triggering condition is met, the warning action also includes performing the preset actions.
[0043] The possible reasons for the fulfillment of triggering conditions can be embedded into the root cause analysis report. Ordinary process engineers can also use this report to understand the possible causes of problems in photovoltaic cell production, pinpoint the root cause, and achieve standardized and large-scale reuse of expert experience. The early warning rule execution unit also executes preset actions to address problems promptly and prevent them from escalating.
[0044] (III) Root Cause Analysis and Report Generation Module Root cause analysis and report generation module, such as Figure 4 As shown, it includes: (3.1) Analysis template configuration unit, used to provide an analysis template configuration interface for users to select analysis components from a preset analysis component library; and for users to arrange the selected analysis components according to the type of early warning event and configure them into analysis links in sequence; and to save the configured analysis links as analysis templates; Users can customize the root cause analysis process by dragging and dropping. The analysis template configuration unit allows users to form personalized analysis processes based on specific warning event types, covering multiple dimensions such as equipment, furnace tubes, temperature zones, and raw materials.
[0045] (3.2) Root cause analysis unit, used to provide root cause analysis service. When an early warning action is triggered, the corresponding analysis template is matched and loaded according to the type of the early warning event; according to the order of each analysis component in the analysis template, each analysis component is called in sequence. Each analysis component performs multi-dimensional root cause analysis on the multi-dimensional battery production records corresponding to the early warning event to obtain the root cause analysis results. Each analysis component performs multi-dimensional root cause analysis, including data query and judgment analysis and / or algorithm analysis; Data query and analysis includes: performing multi-dimensional correlation queries and analysis on multi-dimensional production data in multi-dimensional battery production records corresponding to the early warning event in order to obtain the root cause of the suspected problem; For example, at the equipment level, it automatically queries whether there are other alarms or maintenance records for the equipment during the same period; at the raw material level, it automatically correlates and analyzes the source of raw materials used in the current batch and the raw material quality test results; at the process parameter level, it automatically performs process parameter traceability queries and compares them with the process parameter values in the standard formula.
[0046] Algorithm analysis includes: using chi-square analysis on multi-dimensional production data in multi-dimensional battery production records to identify dimensions suspected of having problems; Chi-square analysis uses the chi-square test formula, as shown in formula (1): (1) in, The chi-square value, This represents the actual number of defects in the i-th dimension sample (e.g., the actual number of defects in the i-th device). Let be the expected number of defects for the i-th dimension sample; assume that the expected number of defects is the same for all dimensions. If the chi-square value exceeds a preset threshold (e.g., 0.05), the corresponding dimension is suspected of having a problem. For example, if the dimension is a device, it indicates that at least one device is defective and is suspected of having a problem.
[0047] For dimensions suspected of having problems, residual analysis is used to determine the root cause of the suspected problems; The residual analysis uses the standardized residual calculation formula, as shown in formula (2): (2) in, To standardize the residual values, Let i be the actual number of defects in the i-th dimension sample. Let be the expected number of defects for the i-th dimension sample; assume that the expected number of defects is the same for all dimensions. If the absolute value of the standardized residual is greater than the preset residual threshold (e.g., 2), then the root cause of the suspected problem in that dimension is determined. For example, if the absolute value of the standardized residual of device A is 2.5, which is greater than 2, then device A is suspected to be the root cause of the problem at the device level. Similarly, by analyzing the furnace tube level and the temperature zone level, we can locate the specific root cause of the problem, which can be represented as the specific device, furnace tube, or temperature zone.
[0048] This application can automatically perform multi-dimensional correlation query judgment analysis or algorithm analysis on multi-dimensional battery production records, and can automatically locate the root cause of the problem; through algorithms such as chi-square test and residual analysis, it realizes correlation mining analysis of multi-dimensional production data, and provides in-depth analysis conclusions far beyond simple threshold alarms.
[0049] (3.3) Report generation unit, used to assemble the root cause analysis results of each analysis component as analysis fragments to generate a root cause analysis report.
[0050] In some embodiments, a second type of expert experience is also embedded when generating the root cause analysis report. This second type of expert experience includes possible causes and conclusive descriptions when the triggering conditions are met.
[0051] This results in a well-structured, multi-dimensional, and conclusive root cause analysis report that is presented directly to the user, rather than the raw data.
[0052] (iv) Visualization Application Module Visual application modules such as Figure 5 As shown, it includes: The real-time early warning unit is used to send notification information based on early warning events; for example, it can send notification messages to users in the form of lists or messages to remind them of the current anomalies and ensure timely response.
[0053] The interactive analysis dashboard unit provides a visual interface for production status, which is based on multi-dimensional root cause analysis. The production status interface allows users to perform operations on multi-dimensional production data, such as downloading and filtering, so that users can independently explore and analyze the data.
[0054] The report output unit is used to output and present the root cause analysis report. In addition to the root cause analysis results, the root cause analysis report also includes possible causes and conclusive descriptions when the triggering conditions are met, so that users can read and review it.
[0055] Finally, it should be noted that the above specific embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.
Claims
1. A system for monitoring, early warning, and automatic root cause analysis based on big data in photovoltaic cell production, characterized in that, include: The data access module is used to obtain multi-dimensional production data of photovoltaic cells from the production management system; After cleaning and integrating the multi-dimensional production data, it is stored to form a battery production data table. Each record in the battery production data table is a multi-dimensional battery production record of a photovoltaic cell. The multi-dimensional early warning module is used to configure and parse early warning rules. It executes early warning rules on the multi-dimensional cell production records of photovoltaic cells and triggers early warning actions when the triggering conditions configured in the early warning rules are met. Early warning actions include generating early warning events and invoking root cause analysis services; The root cause analysis and report generation module is used to configure analysis templates and provide root cause analysis services; When an early warning action is triggered, the root cause analysis service loads the corresponding analysis template according to the type of the early warning event, performs multi-dimensional root cause analysis on the multi-dimensional battery production records corresponding to the early warning event, and generates a root cause analysis report. The visualization application module is used to send notification information based on early warning events, provide a visual interface of production status, and output root cause analysis reports.
2. The system for photovoltaic cell production based on big data monitoring, early warning, and automatic root cause analysis as described in claim 1, characterized in that, The data access module includes: The data acquisition unit is used to acquire multi-dimensional production data of photovoltaic cells from multiple production management systems; The data cleaning unit is used to perform deduplication, noise reduction, unified format conversion, and missing value processing on the acquired photovoltaic cell production data. The data integration unit is used to associate, splice, and aggregate photovoltaic cell production data based on a unique key, resulting in a complete multi-dimensional cell production record for each photovoltaic cell. The unique key for a photovoltaic cell is one or a combination of at least two of the following: cell identifier, batch number, and equipment identifier. The data storage unit is used to generate a battery production data table based on the multi-dimensional battery production records of each photovoltaic cell and store it in the database.
3. The system for photovoltaic cell production based on big data monitoring, early warning, and automatic root cause analysis as described in claim 1, characterized in that, The multi-dimensional early warning module includes: The early warning rule configuration unit is used to provide an early warning rule configuration interface, and to configure and save early warning rules in the early warning rule library. The early warning rule includes the polling cycle, applicable objects and corresponding triggering conditions. The triggering conditions are based on multi-dimensional production data in the multi-dimensional battery production record and corresponding judgment logic settings. The early warning rule parsing unit is used to load early warning rules and generate executable rule instances through the rule parser. The early warning rule execution unit is used to obtain multi-dimensional battery production records of photovoltaic cells according to a preset polling cycle, and execute the judgment logic of multi-dimensional production data in the early warning rule for the multi-dimensional battery production records of photovoltaic cells; when the triggering condition is met, the early warning action is triggered.
4. The system for photovoltaic cell production based on big data monitoring, early warning, and automatic root cause analysis as described in claim 3, characterized in that, The warning rules also include embedded first-class expert experience; the first-class expert experience includes possible reasons when the triggering conditions are met, and preset actions that can be performed; when the triggering conditions are met, the warning action also includes performing preset actions.
5. The system for photovoltaic cell production based on big data monitoring, early warning, and automatic root cause analysis as described in claim 3 or 4, characterized in that, The root cause analysis and report generation module includes: The analysis template configuration unit provides an analysis template configuration interface for users to select analysis components from a pre-set analysis component library; it also allows users to arrange the selected analysis components according to the type of early warning event and configure them into analysis links in sequence; and it saves the configured analysis links as analysis templates. The root cause analysis unit is used to provide root cause analysis services. When an early warning action is triggered, it matches and loads the corresponding analysis template according to the type of the early warning event. According to the order of each analysis component in the analysis template, it calls each analysis component in sequence. Each analysis component performs multi-dimensional root cause analysis on the multi-dimensional battery production records corresponding to the early warning event to obtain the root cause analysis results. The report generation unit is used to assemble the root cause analysis results of each analysis component as analysis fragments to generate a root cause analysis report.
6. The system for photovoltaic cell production based on big data monitoring, early warning, and automatic root cause analysis as described in claim 5, characterized in that, Each analysis component performs multi-dimensional root cause analysis, including data query and judgment analysis and / or algorithm analysis; Data query and analysis includes: performing multi-dimensional correlation queries and analysis on multi-dimensional production data in multi-dimensional battery production records corresponding to the early warning event in order to obtain the root cause of the suspected problem; The algorithm analysis includes: using chi-square analysis on multi-dimensional production data in multi-dimensional battery production records to identify dimensions suspected of having problems; and using residual analysis to determine the root causes of the suspected problems for the identified dimensions.
7. The system for photovoltaic cell production based on big data monitoring, early warning, and automatic root cause analysis as described in claim 6, characterized in that, The chi-square analysis uses the chi-square test formula, as shown in formula (1): (1) in, The chi-square value, Let i be the actual number of defects in the i-th dimension sample. Let be the expected number of defects for the i-th dimension sample; assume that the expected number of defects is the same for all dimensions. If the chi-square value exceeds the preset threshold, the corresponding dimension is suspected of having a problem.
8. The system for photovoltaic cell production based on big data monitoring, early warning, and automatic root cause analysis as described in claim 6, characterized in that, The residual analysis uses the standardized residual calculation formula, as shown in formula (2): (2) in, To standardize the residual values, Let i be the actual number of defects in the i-th dimension sample. Let be the expected number of defects for the i-th dimension sample; assume that the expected number of defects is the same for all dimensions. If the absolute value of the standardized residual is greater than the preset residual threshold, then the root cause of the suspected problem in that dimension is determined.
9. The system for photovoltaic cell production based on big data monitoring, early warning, and automatic root cause analysis as described in claim 5, characterized in that, When generating a root cause analysis report, the report generation unit also embeds a second type of expert experience, which includes possible causes and conclusive descriptions when the triggering conditions are met.
10. The system for photovoltaic cell production based on big data monitoring, early warning, and automatic root cause analysis as described in claim 1, characterized in that, The visualization application module includes: The real-time early warning unit is used to send notification information based on early warning events; An interactive analysis dashboard unit is used to provide a visual production status interface, which is based on multi-dimensional root cause analysis; the production status interface allows users to perform operations on multi-dimensional production data. The report output unit is used to output and present the root cause analysis report.