Photovoltaic production intelligent detection and analysis system and method and storage medium

Abstract

The invention discloses a photovoltaic production intelligent detection and analysis system and method and a storage medium, and the system comprises a data collection equipment layer which is used for collecting physical signals of a silicon material, a silicon wafer, a battery and related equipment of an assembly in the whole process of photovoltaic production, and converting the physical signals into digital signals; the data machine room layer is connected with the data acquisition equipment layer and is used for receiving the digital signals and carrying out data aggregation, edge calculation, data preprocessing and local storage; the software system layer is connected with the data machine room layer, is used for receiving the processed data and realizing full functions of manufacturing and execution, and comprises multi-source data fusion processing, a production scheduling algorithm and cross-factory collaborative logic; and the cloud platform layer is connected with the software system layer and is used for realizing group-level data aggregation and intelligent analysis and global productivity resource scheduling. According to the invention, intelligent management of the whole process of photovoltaic production is realized, the detection precision and efficiency are improved, energy consumption utilization is optimized, data islands are broken, and the production safety is enhanced.

Description

Technical Field

[0001] The present invention belongs to the technical field of photovoltaic product manufacturing systems, and specifically relates to a photovoltaic production intelligent detection and analysis system, method and storage medium. Background Art

[0002] Intelligent inspection and control systems serve as intelligent quality control tools during the manufacturing of semi-finished and final products on photovoltaic production lines. Their application areas span the entire photovoltaic manufacturing process and related upstream and downstream links, including the core production stages, extended applications in supply chain and equipment management, process optimization and quality traceability standardization, product integration with other sectors, and the expansion of industry services. PV manufacturing involves multiple complex processes, including silicon material purification, wafer cutting, cell manufacturing, and module packaging. Traditional production lines face numerous challenges in quality control, energy management, and environmental monitoring. Manual inspections are inefficient, with inconsistent standards and a lack of real-time feedback. Visual inspections, for example, are prone to missed inspections, especially for microscopic defects such as microcracks and hidden cracks. Differences in standards across production lines and factories lead to poor product consistency. Furthermore, energy consumption data is opaque. Factories only monitor total power consumption, failing to accurately calibrate individual production stages. This makes it difficult to dynamically adjust equipment operating parameters based on production status for energy-saving optimization. In addition, each factory has serious information silos, making it difficult for the group's management to grasp the operating status of each base in real time. Collaboration efficiency is low, and production capacity allocation and material allocation rely on manual coordination. There are large fluctuations in quality, and the same process is executed significantly differently in different factories. Decision-making is delayed, and business analysis data needs to be manually summarized, resulting in delays in production plans. This makes it difficult to meet the needs of the photovoltaic industry's transformation from labor-intensive to intelligent and digital. Summary of the Invention

[0003] The purpose of the present invention is to provide a photovoltaic production intelligent detection and analysis system, method and storage medium, which can solve the technical problems of insufficient multi-defect joint detection capability, poor data interoperability, high energy consumption and poor adaptability in the existing technology.

[0004] In order to achieve the above-mentioned purpose, a specific embodiment of the present invention provides a photovoltaic production intelligent detection and analysis system, which includes a data acquisition equipment layer, a data room layer, a software system layer and a cloud platform layer, wherein: The data acquisition device layer is used to collect physical signals of silicon materials, silicon wafers, batteries and component-related equipment in the entire photovoltaic production process, and convert the physical signals into digital signals; The data room layer is connected to the data acquisition equipment layer, and is used to receive the digital signal and perform data aggregation, edge computing, data preprocessing and local storage; The software system layer is connected to the data room layer and is used to receive processed data and implement full manufacturing execution functions, including multi-source data fusion processing, production scheduling algorithms, and cross-factory collaboration logic; The cloud platform layer is connected to the software system layer to achieve group-level data aggregation and intelligent analysis, as well as global production capacity resource scheduling.

[0005] In one or more embodiments of the present invention, the data acquisition device layer includes a sensor network, a machine vision system, a PLC controller, and an RFID reader / writer, wherein: The sensor network is used to collect physical signals such as temperature, pressure, and gas and convert them into analog electrical signals and then into digital signals; The machine vision system is used to perform optical imaging of photovoltaic products and combine it with AI defect recognition; The PLC controller is used to collect device status information and issue control instructions; The RFID reader is used for radio frequency identification of material tracking tags.

[0006] In one or more embodiments of the present invention, the data room layer includes: a data aggregation unit, an edge computing unit, a data pre-processing unit, and a local storage unit, wherein: The data aggregation unit receives data transmitted from different types of devices at the data acquisition device layer through the industrial protocol gateway and aggregates and integrates the scattered data; at the same time, the data aggregation unit buffers and forwards the data to ensure the integrity and continuity of the data; The edge computing unit is used to quickly analyze and process the real-time aggregated data. The edge computing unit performs real-time feature extraction and preliminary screening on the image data collected by the machine vision system, quickly identifying product images with obvious defects and providing valuable information for subsequent in-depth analysis. The data preprocessing unit is used to remove noise data and outliers in the collected data using a data cleaning algorithm; the data preprocessing unit uses an interpolation algorithm to fill in missing data, and the data preprocessing unit performs standardized conversion on data in different formats; The local storage unit is used to construct an all-flash storage array to cache the pre-processed high-frequency collected data, and to use a time series database to efficiently store and manage the time series data in the production process.

[0007] In one or more embodiments of the present invention, the software system layer includes a data stream processing unit, a production scheduling unit, and a quality closed-loop control unit, wherein: The data stream processing unit is used to perform multi-source data fusion processing on the received data, and adopt different fusion strategies according to the type and source of the data; The production scheduling unit builds a production scheduling model based on a reinforcement learning algorithm, taking into account multiple factors such as equipment load, material inventory, order delivery date, and process priority; The quality closed-loop control unit receives quality inspection data collected by the machine vision system, EL inspection equipment, etc. in real time, and performs matching analysis with the production process parameter database.

[0008] In one or more embodiments of the present invention, the cloud platform layer includes a data center, a microservice architecture, and an application layer module, wherein: The data center platform pulls data from the data center of each factory in real time through a distributed data acquisition engine, supporting high-frequency data acquisition at the millisecond level. The data center platform deeply cleans, converts, and stores the collected raw data, and uses big data processing tools to convert massive amounts of equipment status data, process parameter data, production capacity data, etc. into standardized data sets, forming a unified data warehouse to provide data support for upper-level applications. The microservice architecture is used to modularize the various business functions of photovoltaic production and disassemble them into independent microservices. Each microservice interacts through a lightweight communication protocol, enabling flexible calling and combination of services. When business needs change, the corresponding microservices can be easily updated, expanded or replaced. The application layer modules at least include production scheduling applications, process monitoring applications, quality management applications and energy management applications.

[0009] In one or more embodiments of the present invention, the production scheduling application is used to formulate a scientific and reasonable production schedule based on cloud-based algorithms, combining multi-dimensional information such as market demand, equipment capacity, and material supply, and to achieve real-time monitoring and dynamic adjustment of the production process; The process monitoring application is used to display the production line operation status to management personnel in real time through the web and app interface, including key data such as equipment OEE, product yield, energy consumption indicators, etc., and supports remote viewing of production video monitoring images; The quality management application is used to establish a cloud-based quality database, conduct in-depth analysis and mining of quality inspection data, generate quality reports in real time, and promptly trigger early warnings and conduct traceability analysis for situations such as defective products exceeding thresholds; The energy management application is used to collect various energy consumption data in the production process, build an energy consumption model, and achieve reasonable allocation and precise control of energy through intelligent optimization algorithms, thereby reducing energy consumption per unit of production capacity.

[0010] In one or more embodiments of the present invention, the data acquisition equipment layer is connected to the data room layer by wired and / or wireless means to transmit the collected data or receive real-time control signals transmitted by the data room layer; the data room layer is connected to the software system layer through a data pipeline to transmit preprocessed data streams to the software system layer and / or receive control instructions transmitted from the software system layer, and the software system layer is communicatively connected to the cloud platform layer to transmit single-level factory data to the cloud platform layer and / or receive instructions from the cloud platform layer.

[0011] In another aspect of the present invention, a photovoltaic production intelligent detection and analysis method is provided, the photovoltaic production intelligent detection and analysis method comprising the following steps: Collect photovoltaic production data and convert it into digital signals; Aggregate, edge compute, pre-process and locally store digital signals; Receive processed data and perform multi-source data fusion processing, production scheduling, and cross-factory collaborative decision-making; Aggregate data from various factories and conduct intelligent analysis and capacity scheduling.

[0012] In another aspect of the present invention, a photovoltaic production intelligent detection and analysis method is provided, the photovoltaic production intelligent detection and analysis method comprising the following steps: The cloud platform layer performs intelligent analysis and capacity scheduling based on the collected data, generates corresponding control instructions, and sends the control instructions to the software system layer; The software system layer receives the control instructions, parses them and sends them to the data room layer; The data room layer receives control instructions and converts them into signals suitable for execution by the data acquisition device layer through industrial protocol gateways and communication lines; The PLC controller, sensor network or other control units in the data acquisition device layer receive signals and perform corresponding operations.

[0013] In another aspect of the present invention, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the photovoltaic production intelligent detection and analysis system are implemented.

[0014] Beneficial Effects: Compared to existing technologies, the photovoltaic production intelligent detection and analysis system of this invention, from the data acquisition device layer to the cloud platform layer, ensures the timely and accurate collection and transmission of photovoltaic production data. Various sensors and acquisition units at the data acquisition device layer capture first-hand data from every production process. Preliminary processing and standardization at the data center layer ensures data quality, providing a precise data foundation for subsequent analytical decisions and reducing production misjudgments caused by data delays or errors.

[0015] Collaborative processing at the data center and software system levels enables efficient integration and in-depth analysis of massive amounts of production data. Edge computing pre-screens and processes key data at the data center level, reducing the burden on the cloud. Software system functions such as multi-source data fusion, production scheduling algorithms, and closed-loop quality control further unlock the value of data, enabling rapid insight into production process issues and trends, and improving the sophistication and responsiveness of production management.

[0016] The cloud platform aggregates data from each factory and conducts intelligent analysis and capacity scheduling, optimizing the allocation of photovoltaic production resources from a global perspective. Leveraging big data analysis and AI technology, it deeply mines production data to predict market demand, optimize production schedules, and provide early warning of equipment failures. This enables intelligent decision-making during the production process, improving production efficiency, reducing operating costs, and enhancing market competitiveness.

[0017] Stable communication protocols and data transmission mechanisms between each layer ensure the reliable operation of the entire system. Industrial protocols ensure the stability of data collection and transmission, and APIs / message queues enable reliable communication between the software system layer and the cloud platform layer, allowing data to flow accurately between each link. Even in complex production environments, the system can maintain uninterrupted operation and reduce the risk of data loss or communication interruptions.

[0018] This process breaks down data silos across various aspects of photovoltaic production, enabling vertical data integration from the equipment layer to the cloud, as well as horizontal data integration for cross-factory collaboration. Data from different factories and workshops is aggregated and shared on the cloud platform, promoting collaboration across departments within the enterprise, improving the overall coherence and consistency of production processes, and facilitating unified management and resource allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments described in the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort.

[0020] Figure 1 This is a schematic diagram of the structure of a photovoltaic production intelligent detection and analysis system in one embodiment of the present invention; Figure 2 This is a structural diagram of a data acquisition device layer in one embodiment of the present invention; Figure 3 This is a structural diagram of a data room layer in one embodiment of the present invention; Figure 4A schematic diagram of the structure of the software system layer in one embodiment of the present invention; Figure 5 This is a schematic diagram of the structure of the cloud platform layer in one embodiment of the present invention; Figure 6 This is a flow chart of a photovoltaic production intelligent detection and analysis method according to one embodiment of the present invention; Figure 7 This is a flow chart of a photovoltaic production intelligent detection and analysis method according to another embodiment of the present invention; Figure 8 This is the architectural intent of the photovoltaic production intelligent detection and analysis system in one embodiment of the present invention; Figure 9 This is the architectural intent of a photovoltaic production intelligent detection and analysis system in another embodiment of the present invention; Figure 10 This is a demonstration diagram (1) of the photovoltaic production intelligent detection and analysis system in one embodiment of the present invention; Figure 11 This is a demonstration diagram (2) of the photovoltaic production intelligent detection and analysis system in one embodiment of the present invention; Figure 12 This is a detection demonstration diagram (3) of the photovoltaic production intelligent detection and analysis system in one embodiment of the present invention. DETAILED DESCRIPTION

[0021] In order to enable those skilled in the art to better understand the technical solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0022] As mentioned in the background technology, there are issues such as quality control, energy consumption optimization, data silos, and production safety in the photovoltaic manufacturing process.

[0023] In response to the above technical issues, such as Figure 1As shown, the present invention introduces a photovoltaic production intelligent detection and analysis system, including a data acquisition device layer, a data computer room layer, a software system layer and a cloud platform layer, wherein the data acquisition device layer is equivalent to the edge as the "nerve endings" of the industrial site, and is set to collect data on the production line to realize the conversion of physical signals into digital signals; the data computer room layer is often set near the data acquisition device layer and connected to it by wired and / or wireless means to realize local calculation and storage of raw data; the software system layer serves as a business logic carrier to realize the decision engine of full-function manufacturing execution and multi-factory collaboration, and the cloud platform layer is used for level data aggregation and serves as the center for intelligent analysis of production capacity resource scheduling. In this embodiment, the data acquisition equipment layer is used to collect physical signals from silicon materials, silicon wafers, batteries and component-related equipment in the entire photovoltaic production process, and convert the physical signals into digital signals; the data room layer is connected to the data acquisition equipment layer, and is used to receive digital signals and perform data aggregation, edge computing, data preprocessing and local storage; the software system layer is connected to the data room layer, and is used to receive processed data and realize the full functions of manufacturing execution, including multi-source data fusion processing, production scheduling algorithms and cross-factory collaborative logic; the cloud platform layer is connected to the software system layer, and is used to realize group-level data aggregation and intelligent analysis, as well as global production capacity resource scheduling.

[0024] In this embodiment, the wired connection between the data acquisition device layer and the data room layer can be industrial Ethernet and / or RS485 bus, and the wireless connection can be 5G private network or WIFI6; the connection between the data room layer and the software system layer can be through a data channel established by Python. The code for implementing the data channel and multi-source data preprocessing can be: Python # Typical data flow pseudocode class DataPipeline: def __init__(self): self.input = KafkaConsumer("plant1-raw-data") self.output = gRPCCannel(software_system) def process(self): while True: raw = self.input.poll() cleaned = DataCleaner.apply_rules(raw)# Data cleaning enriched = FeatureEngineer.extract(cleaned)# Feature extraction self.output.push(enriched) Real-time control interface: Protocol: gRPC (Protobuf encoding) Batch data interface: Protocol: RESTful API (JSON format); In this embodiment, the connection between the software system layer and the cloud platform layer adopts a hybrid cloud connection mode.

[0025] In a further embodiment, Figure 2 As shown, the data acquisition equipment layer includes sensor network units: installed at key locations throughout photovoltaic production, they collect real-time physical signals such as temperature, pressure, gas composition, and concentration. Signal conversion circuits convert these physical signals into analog electrical signals, which are then converted to digital signals via analog-to-digital converters. For example, during the silicon wafer cutting process, temperature sensors are tightly attached to the cutting equipment to monitor the temperature in the cutting area in real time. During the cell sintering process, gas sensors monitor the concentrations of gases such as oxygen and hydrogen in the environment, providing essential data for process optimization and safety assurance. Machine vision system units: deployed at various workstations in the production of silicon wafers, cells, and modules, they consist of high-resolution industrial cameras, optical lenses, and corresponding lighting. The industrial cameras use optical lenses to image photovoltaic products, and lighting assists in obtaining clear image information. The system uses a deep learning-based image recognition algorithm to analyze the collected images in real time, accurately identifying defects such as microcracks on the silicon wafer surface, hidden cracks in the cell, and poor solder joints in the module package and bubbles in the EVA film, with recognition accuracy reaching a preset pixel level. PLC controller units: These connect to various production equipment on the production line. They collect operational status parameters, such as power on / off status, motor speed, and operating time. They also issue control commands, enabling precise control of the equipment. For example, system feedback on equipment failures can be used to promptly shut down equipment or adjust operating parameters based on optimized process parameters. RFID reader / writer units: These are installed at key points in the material flow and work in conjunction with RFID tags attached to materials. During the transfer of silicon materials, wafer storage, and material handovers during the cell and module production processes, RFID readers rapidly read or write relevant material information, enabling precise tracking and management, ensuring visibility and traceability of material flows. For example, during the module packaging process, RFID tags on auxiliary materials such as glass and backsheets are read to verify that the material specifications and batches meet production requirements.

[0026] In a further embodiment, Figure 3As shown, the data center layer includes a data aggregation unit: This unit receives data transmitted from various types of devices in the data acquisition device layer via industrial protocol gateways. This includes data from PLC controllers using the OPC UA protocol, sensor network data using the Modbus protocol, and machine vision system data using custom protocols, thereby aggregating and integrating these various types of data. Furthermore, the MQTT messaging middleware is used to buffer and forward data to ensure data integrity and continuity. The edge computing unit: This unit, equipped with high-performance edge computing servers, runs a stream processing framework to rapidly analyze and process real-time aggregated data. For example, it monitors temperature, pressure, and other data collected by the sensor network in real time. When data exceeds the preset process range, an alarm is immediately triggered and preliminary fault diagnosis and analysis is performed. Image data collected by the machine vision system is subjected to real-time feature extraction and preliminary screening to quickly identify product images with obvious defects, providing valuable information for subsequent in-depth analysis. The data preprocessing unit: This unit uses data cleaning algorithms to remove noise and outliers from the collected data, such as using a sliding average to smooth temperature sensor data. Interpolation algorithms are used to fill in missing data, such as providing a reasonable estimate of occasionally missing gas concentration data. At the same time, data in different formats is standardized and converted, with unified units and dimensions to ensure data consistency and comparability. Local Storage Unit: Build an all-flash storage array to cache pre-processed, high-frequency data, ensuring data is not lost in the event of anomalies such as network outages. The storage period is [X] days. Furthermore, a time series database is used to efficiently store and manage time series data during the production process. For example, process parameters for each workstation in the silicon wafer production process are stored in chronological order to facilitate subsequent data query and analysis.

[0027] In a further embodiment, Figure 4As shown in Figure 1, the software system layer includes a data stream processing module that performs multi-source data fusion on received data, employing different fusion strategies based on data type and source. For example, it correlates and fuses equipment operating status data collected by the PLC controller with product image data collected by the machine vision system to achieve a comprehensive assessment of equipment status and product quality. It also uses a time series database to align process parameter data, ensuring that data from each wafer passing through each workstation accurately corresponds to the same point in time, providing accurate data support for process optimization. The production scheduling module, based on a reinforcement learning algorithm, constructs a production scheduling model that comprehensively considers factors such as equipment load, material inventory, order delivery dates, and process priorities. By evaluating and analyzing the real-time status of each factory, it dynamically generates an optimal production schedule and monitors the production process in real time. In the event of unexpected situations such as equipment failures and order changes, it can quickly adjust the scheduling plan to ensure efficient production. For example, in the event of a sudden equipment failure at a factory, some orders were promptly reassigned to other factories with sufficient production capacity, reducing order delays by 72%. The closed-loop quality control module receives real-time quality inspection data from machine vision systems, EL inspection equipment, and other systems, and compares and analyzes it against a database of production process parameters. When quality defects are detected, adjustments are automatically triggered. For example, to address issues like broken grid lines in printed solar cells, the module automatically adjusts the pressure and speed parameters of the printing equipment. This achieves closed-loop quality control during the production process and improves product quality stability.

[0028] In a further embodiment, Figure 5As shown, the cloud platform layer includes a data center: Serving as a data aggregation center, it pulls data from each factory's data center in real time through a distributed data collection engine, supporting high-frequency data collection in milliseconds. The collected raw data is deeply cleaned, converted, and stored. Using big data processing tools such as Hadoop and Spark, massive amounts of equipment status data, process parameter data, and production capacity data are converted into standardized datasets, forming a unified data warehouse that provides data support for upper-layer applications. A microservices architecture modularizes the various business functions of photovoltaic production into independent microservices, such as work order management, quality management, equipment management, and energy management. These microservices interact through lightweight communication protocols (such as RESTful APIs), enabling flexible invocation and combination between services. As business needs change, corresponding microservices can be easily updated, expanded, or replaced, improving the overall scalability and flexibility of the system. For example, adding a quality inspection service for new photovoltaic modules requires only developing and integrating the corresponding microservices, eliminating the need for a major system overhaul. Application layer modules encompass multiple functional modules, including production scheduling, process monitoring, quality management, and energy management. The production scheduling application is based on cloud-based algorithms and combines multi-dimensional information such as market demand, equipment capacity, and material supply to formulate scientific and reasonable production scheduling plans, and realize real-time monitoring and dynamic adjustment of the production process; the process monitoring application displays the production line operation status to managers in real time through the Web and APP interface, including key data such as equipment OEE (overall equipment efficiency), product yield, energy consumption indicators, and supports remote viewing of production video monitoring images; the quality management application establishes a cloud-based quality database, conducts in-depth analysis and mining of quality inspection data, generates quality reports in real time, and promptly triggers early warnings and conducts traceability analysis for situations such as defective products exceeding thresholds; the energy management application collects various energy consumption data in the production process, constructs an energy consumption model, and realizes reasonable allocation and precise control of energy through intelligent optimization algorithms, thereby reducing energy consumption per unit of production capacity.

[0029] In a further embodiment, Figures 6 and 7 The figure shows the process entity diagram of the photovoltaic production intelligent detection and analysis method in the present invention. Figure 1 , a photovoltaic production intelligent detection and analysis method includes: S11, collecting photovoltaic production data and converting it into digital signals; S12, aggregate, edge compute, pre-process and locally store digital signals; S13, receiving the processed data and performing multi-source data fusion processing, production scheduling, and cross-factory collaborative decision-making; S14. Aggregate data from each factory and conduct intelligent analysis and capacity scheduling.

[0030] The above method is a process of multi-source data from collection to preprocessing, fusion and scheduling, and the method also includes: S21. The cloud platform layer performs intelligent analysis and capacity scheduling based on the collected data, generates corresponding control instructions, and sends the control instructions to the software system layer; S22, the software system layer receives the control instruction, parses it and sends it to the data room layer; S23, the data room layer receives the control instructions and converts them into signals suitable for execution by the data acquisition device layer through the industrial protocol gateway and communication lines; S23. The PLC controller, sensor network or other control units in the data acquisition device layer receive the signal and perform corresponding operations.

[0031] In a further embodiment, Figures 8-12 As shown, the architecture and application scenarios of the present invention include at least: Silicon wafer inspection: During the silicon wafer cutting process, high-resolution cameras and AI algorithms are used to detect defects such as microcracks and edge collapse on the silicon wafer surface in real time to improve silicon wafer quality. Cell manufacturing monitoring: Real-time monitoring of the velvet structure and printed grid line quality during the cell manufacturing process, identifying potential defects such as hidden cracks or PID defects, and optimizing process parameters. Module packaging inspection: Detects solder joints, EVA film bubbles, backplane scratches, etc. to ensure module packaging quality and reduce scrap rates. Supply chain and equipment management: Quality inspection of raw materials to control quality fluctuations at the source; infer equipment anomalies through defect patterns to achieve predictive maintenance. Data-driven optimization: Linking inspection data with the MES system to automatically adjust process parameters, establish a defect database, and assist in ISO quality management certification.

[0032] Compared with the existing technology, the present invention uses high-precision sensors and machine vision systems to conduct all-round, real-time quality inspections on photovoltaic products, accurately identify various types of defects, and improve inspection efficiency and accuracy. The present invention can monitor the energy consumption of each production link in real time, dynamically adjust equipment operating parameters through intelligent optimization algorithms, reduce energy waste, and improve energy utilization efficiency. The present invention breaks the information island, realizes the real-time aggregation and sharing of multi-factory data, promotes inter-departmental collaboration, and improves the coherence and consistency of production processes. Based on big data analysis and AI technology, the present invention deeply mines production data, provides intelligent decision-making support such as production scheduling, process optimization, and equipment maintenance, and improves enterprise operational efficiency. The present invention adopts a "core cloud + edge node" layered structure, combined with 5G + cloud + network + big data technology, to achieve millisecond-level response, support PB-level data storage and dynamic expansion.

[0033] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0034] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A system that specifies the functions of a box or boxes.

[0035] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture including an instruction system that is implemented in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0036] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0037] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims, not the foregoing description, and all variations within the meaning and range of equivalents of the claims are intended to be included therein. Any reference sign in a claim should not be construed as limiting the claim to which it relates.

[0038] In addition, it should be understood that although this specification is described in terms of implementation methods, not every implementation method contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

Claims

1. A photovoltaic production intelligent detection and analysis system, characterized in that: It includes data acquisition equipment layer, data room layer, software system layer and cloud platform layer, among which, The data acquisition device layer is used to collect physical signals of silicon materials, silicon wafers, batteries and component-related equipment in the entire photovoltaic production process, and convert the physical signals into digital signals; The data room layer is connected to the data acquisition equipment layer, and is used to receive the digital signal and perform data aggregation, edge computing, data preprocessing and local storage; The software system layer is connected to the data room layer and is used to receive processed data and implement full manufacturing execution functions, including multi-source data fusion processing, production scheduling algorithms, and cross-factory collaboration logic; The cloud platform layer is connected to the software system layer to achieve group-level data aggregation and intelligent analysis, as well as global production capacity resource scheduling.

2. A photovoltaic production intelligent detection and analysis system according to claim 1, characterized in that: The data acquisition equipment layer includes a sensor network, a machine vision system, a PLC controller and an RFID reader. The sensor network is used to collect physical signals such as temperature, pressure, and gas and convert them into analog electrical signals and then into digital signals; The machine vision system is used to perform optical imaging of photovoltaic products and combine it with AI defect recognition; The PLC controller is used to collect device status information and issue control instructions; The RFID reader is used for radio frequency identification of material tracking tags.

3. The photovoltaic production intelligent detection and analysis system according to claim 1, characterized in that: The data room layer includes: a data aggregation unit, an edge computing unit, a data pre-processing unit and a local storage unit, wherein: The data aggregation unit receives data transmitted from different types of devices at the data acquisition device layer through the industrial protocol gateway and aggregates and integrates the scattered data; at the same time, the data aggregation unit buffers and forwards the data to ensure the integrity and continuity of the data; The edge computing unit is used to quickly analyze and process the real-time aggregated data. The edge computing unit performs real-time feature extraction and preliminary screening on the image data collected by the machine vision system, quickly identifying product images with obvious defects and providing valuable information for subsequent in-depth analysis. The data preprocessing unit is used to remove noise data and outliers in the collected data using a data cleaning algorithm; the data preprocessing unit uses an interpolation algorithm to fill in missing data, and the data preprocessing unit performs standardized conversion on data in different formats; The local storage unit is used to construct an all-flash storage array to cache the pre-processed high-frequency collected data, and to use a time series database to efficiently store and manage the time series data in the production process.

4. The photovoltaic production intelligent detection and analysis system according to claim 1, characterized in that: The software system layer includes a data stream processing unit, a production scheduling unit and a quality closed-loop control unit, wherein: The data stream processing unit is used to perform multi-source data fusion processing on the received data, and adopt different fusion strategies according to the type and source of the data; The production scheduling unit builds a production scheduling model based on a reinforcement learning algorithm, taking into account multiple factors such as equipment load, material inventory, order delivery date, and process priority; The quality closed-loop control unit receives quality inspection data collected by the machine vision system, EL inspection equipment, etc. in real time, and performs matching analysis with the production process parameter database.

5. The photovoltaic production intelligent detection and analysis system according to claim 1, characterized in that: The cloud platform layer includes: data center, microservice architecture and application layer modules, among which, The data center platform pulls data from the data center of each factory in real time through a distributed data acquisition engine, supporting high-frequency data acquisition at the millisecond level. The data center platform deeply cleans, converts, and stores the collected raw data, and uses big data processing tools to convert massive amounts of equipment status data, process parameter data, production capacity data, etc. into standardized data sets, forming a unified data warehouse to provide data support for upper-level applications. The microservice architecture is used to modularize the various business functions of photovoltaic production and disassemble them into independent microservices. Each microservice interacts through a lightweight communication protocol, enabling flexible calling and combination of services. When business needs change, the corresponding microservices can be easily updated, expanded or replaced. The application layer modules at least include production scheduling applications, process monitoring applications, quality management applications and energy management applications.

6. The photovoltaic production intelligent detection and analysis system according to claim 5, characterized in that: The production scheduling application is used to formulate scientific and reasonable production scheduling plans based on cloud-based algorithms, combining multi-dimensional information such as market demand, equipment capacity, and material supply, and to achieve real-time monitoring and dynamic adjustment of the production process; The process monitoring application is used to display the production line operation status to management personnel in real time through the web and app interface, including key data such as equipment OEE, product yield, energy consumption indicators, etc., and supports remote viewing of production video monitoring images; The quality management application is used to establish a cloud-based quality database, conduct in-depth analysis and mining of quality inspection data, generate quality reports in real time, and promptly trigger early warnings and conduct traceability analysis for situations such as defective products exceeding thresholds; The energy management application is used to collect various energy consumption data in the production process, build an energy consumption model, and achieve reasonable allocation and precise control of energy through intelligent optimization algorithms, thereby reducing energy consumption per unit of production capacity.

7. The photovoltaic production intelligent detection and analysis system according to claim 1, characterized in that: The data acquisition equipment layer is connected to the data room layer by wired and / or wireless means to transmit the collected data or receive real-time control signals transmitted by the data room layer; the data room layer is connected to the software system layer by a data pipeline to transmit preprocessed data streams to the software system layer and / or receive control instructions transmitted from the software system layer, and the software system layer is communicatively connected to the cloud platform layer to transmit single-level factory data to the cloud platform layer and / or receive instructions from the cloud platform layer.

8. A photovoltaic production intelligent detection and analysis method, wherein the traffic analysis method is applied to the photovoltaic production intelligent detection and analysis system according to any one of claims 1 to 7, characterized in that: The photovoltaic production intelligent detection and analysis method comprises the following steps: Collect photovoltaic production data and convert it into digital signals; Aggregate, edge compute, pre-process and locally store digital signals; Receive processed data and perform multi-source data fusion processing, production scheduling, and cross-factory collaborative decision-making; Aggregate data from various factories and conduct intelligent analysis and capacity scheduling.

9. A photovoltaic production intelligent detection and analysis method, wherein the traffic analysis method is applied to the photovoltaic production intelligent detection and analysis system according to any one of claims 1 to 7, characterized in that: The photovoltaic production intelligent detection and analysis method comprises the following steps: The cloud platform layer performs intelligent analysis and capacity scheduling based on the collected data, generates corresponding control instructions, and sends the control instructions to the software system layer; The software system layer receives the control instructions, parses them and sends them to the data room layer; The data room layer receives control instructions and converts them into signals suitable for execution by the data acquisition device layer through industrial protocol gateways and communication lines; The PLC controller, sensor network or other control units in the data acquisition device layer receive signals and perform corresponding operations.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the photovoltaic production intelligent detection and analysis method according to any one of claims 8 to 9 or executes the photovoltaic production intelligent detection and analysis system according to any one of claims 1 to 7.

Images