Product test data storage and management system based on MES system

Abstract

The invention relates to the technical field of product test data management, and discloses a product test data storage and management system based on an MES system. The system comprises a dynamic configuration module which can use a self-adaptive parameter adjustment algorithm to configure data acquisition parameters according to experiment types; the closed-loop feedback module is used for binding experimental data with production batches and process parameters based on a real-time association algorithm; the two-stage verification module is used for detecting data exception through a double-layer cascade verification algorithm; and the intelligent storage module is used for generating hierarchical storage labels by using a multi-dimensional structured coding algorithm. In addition, the system integration module realizes data interaction with ERP and SCADA systems based on an interface adaptation algorithm. The system solves the problems that in a traditional data management mode, collection parameters are fixed, data relevance is poor, a verification mechanism is imperfect, storage and retrieval are inconvenient and the like, data can be efficiently collected, accurately associated, strictly verified and intelligently stored, and the enterprise data management level and production competitiveness are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of product test data management, and specifically to a product test data storage and management system based on the MES system. Background Art

[0002] In the current booming modern manufacturing industry, the quality and performance of products have become the key factors for enterprises to stand out in the fierce market competition. As the core link to ensure product quality, product testing generates a large amount of complex data, which contains key information on all aspects from product design and R & D to production and manufacturing, and is crucial for the continuous innovation, production process optimization and quality control of enterprises.

[0003] There are many drawbacks in the traditional product test data management mode. In terms of data collection, there is a lack of dynamic adaptability. Previous systems often used fixed data collection parameters. Whether it is the sampling frequency, trigger condition or sensor selection, it is difficult to flexibly adjust according to different experiment types and actual production situations. For example, when testing products with different models and different process requirements, using a fixed sampling frequency may lead to data redundancy or omission of key information. For some test scenarios with extremely high requirements for time accuracy, the fixed sampling frequency cannot meet the needs, resulting in the data collected being unable to accurately reflect the performance of the product at a specific moment; in terms of trigger condition setting, it is not intelligent enough to capture the key nodes of product state changes in a timely manner, affecting the effectiveness of data collection.

[0004] Poor data correlation is another prominent problem. The experimental data is isolated from the production batches and process parameters in the MES system, and no effective mapping relationship can be established. This makes it difficult for enterprises to quickly locate the root cause of problems when analyzing product quality issues. When product quality defects occur, it is difficult to determine which link and which batch of production in the production process have problems because there is no direct association with the production batches and corresponding process parameters, and no strong data support can be provided for quality improvement. Moreover, in terms of production process optimization, due to the lack of data correlation, the potential relationship between process parameters and product performance cannot be mined from historical data, restricting the continuous improvement and optimization of the production process by enterprises.

[0005] The imperfect data verification mechanism also brings great troubles to data management. Existing systems usually only perform simple single-dimensional verification, such as only checking whether the data exceeds a certain fixed threshold, and cannot comprehensively detect the accuracy and rationality of data from multiple levels. This single verification method cannot identify some complex abnormal situations. For example, the data is within the threshold range but does not conform to the production logic. When the production process changes or the equipment has minor faults, the data may seemingly meet the fixed threshold requirements on the surface, but in fact, it has deviated from the normal production logic, and the traditional verification mechanism is powerless, resulting in incorrect data entering the subsequent analysis and decision-making links, affecting the enterprise's judgment and decision-making.

[0006] In terms of data storage, the traditional method lacks an efficient coding and indexing mechanism. The data storage is chaotic, and it is difficult to quickly and accurately retrieve the required information. As the enterprise's data volume continues to grow, the drawbacks of this storage method become more and more obvious. It becomes extremely difficult to find the test data of a specific batch, specific time, or specific equipment, seriously affecting the data utilization efficiency. When the enterprise needs to review historical data for product quality analysis and production process optimization, it spends a lot of time on data retrieval, reducing work efficiency and increasing costs.

[0007] With the accelerating advancement of the digital transformation of the manufacturing industry, enterprises have higher and higher requirements for the management of product test data, and the traditional management mode can no longer meet the needs of enterprise development. How to achieve efficient collection, accurate association, strict verification, and intelligent storage of product test data has become an urgent problem to be solved. This is not only related to the improvement of enterprise product quality and cost control, but also related to the enterprise's competitiveness and sustainable development ability in the digital age. Summary of the Invention

[0008] The purpose of the present invention is to provide a product test data storage and management system based on the MES system to solve the problems raised in the above background technology.

[0009] To achieve the above purpose, the present invention provides the following technical solution: A product test data storage and management system based on the MES system, the system includes: Dynamic configuration module: used to dynamically configure data acquisition parameters according to the experiment type by using an adaptive parameter adjustment algorithm, and the data acquisition parameters include sampling frequency, trigger condition, and sensor selection; Closed-loop feedback module: based on a real-time association algorithm, bind the experimental data with the production batch and process parameters in the MES system to generate a data-production mapping relationship chain; Two-level verification module: perform anomaly detection on the collected data through a double-layer cascaded verification algorithm. The first level performs basic threshold verification and data integrity verification, and the second level performs logical rationality verification in combination with the production context; Intelligent storage module: Generates hierarchical storage tags for experimental data according to batch number, timestamp, device identifier, and processing stage by using a multi-dimensional structured coding algorithm.

[0010] Preferably, the adaptive parameter adjustment algorithm includes: Construct a mapping rule library for experiment types and acquisition parameters, where the rule library contains parameter templates for different experimental scenarios; Match the corresponding template according to the MES production task instruction received in real time, and dynamically adjust the smooth transition of the sampling frequency through a sliding window mechanism; Use a fuzzy inference system to prioritize the triggering conditions, and dynamically activate or block specific sensors based on the signal-to-noise ratio of the sensor signals.

[0011] Preferably, the real-time association algorithm includes: Extract the production batch code and process parameter sequence in the MES system to generate a time synchronization index; Construct a data stream alignment model, and perform sliding matching between the timestamp of the experimental data and the time index of the production batch; Store the matching results through an incremental hash table to form a traceable data-production mapping relationship chain.

[0012] Preferably, the double-level cascade verification algorithm includes: At the first level, set a dynamic threshold interval, calculate the upper and lower limit thresholds according to the historical data distribution of the device, and verify the integrity of the data fields; At the second level, load the production context rule library, which contains process constraints, device status association relationships, and an exception pattern library; Analyze the rationality of the data sequence through a time series inference engine, and mark the data as abnormal if a violation of the context logic is detected.

[0013] Preferably, the multi-dimensional structured coding algorithm includes: Encode the batch number and device identifier into a fixed-length binary sequence, and generate a unique index through a hash function; Perform segmented coding on the timestamp to distinguish between year-month-day and millisecond-level precision; Construct a tree-like tag structure based on the processing stage type, supporting data retrieval according to the experimental preparation, execution, and post-processing stages.

[0014] Preferably, the system further includes: System integration module: Realize data interaction with ERP and SCADA systems based on an interface adaptation algorithm, supporting modular expansion and protocol conversion; The interface adaptation algorithm includes: Define a unified data exchange protocol to convert the experimental data format into the standard input format of the target system; Adopt a protocol proxy mechanism to analyze the material coding rules of the ERP system and the real-time data interface of the SCADA system; Implement cross-system data buffering through an asynchronous message queue.

[0015] Preferably, the fuzzy inference system includes: Design a three-input single-output fuzzy controller with the input variables being the experimental complexity, environmental noise level, and equipment load rate; Construct triangular membership functions to partition the fuzzy sets of the input variables, with the output variable being the sensor activation weight; Generate sensor selection instructions through defuzzification by the centroid method.

[0016] Preferably, the data stream alignment model includes: Divide the experimental data stream into data blocks of fixed length, and attach a check code and a timestamp to each data block; Adopt the dynamic time warping algorithm to align the production batch time axis and the experimental data time axis; Fill the data missing segments caused by transmission delay through an interpolation compensation mechanism.

[0017] Preferably, the time series inference engine includes: Construct a hidden Markov model to describe the transition probability of the normal data sequence; Extract the feature vectors of the abnormal data segments, including the statistical deviation degree, the density of mutation points, and the status of associated devices; Calculate the abnormal probability through a Bayesian network, and trigger the MES warning signal if it exceeds the preset threshold.

[0018] Preferably, the protocol proxy mechanism includes: Analyze the bill of materials structure of the ERP system and extract the material attribute fields related to the experimental data; Construct a state machine model for the communication channel of the SCADA system to manage the connection establishment, data encapsulation, and error retransmission processes; Maintain the stability of the interface through a heartbeat detection and timeout reconnection mechanism.

[0019] Compared with the prior art, the beneficial effects of the present invention are: In the data acquisition stage, the dynamic configuration module of the system plays a crucial role. Through the adaptive parameter adjustment algorithm, the data acquisition parameters are flexibly configured according to the experiment type. The established mapping rule library between experiment types and acquisition parameters enables the system to quickly match parameter templates under different experimental scenarios. In the performance testing of electronic products and the durability testing of mechanical products, the most suitable sampling frequency, triggering conditions, and sensors can be selected according to their respective characteristics. The sliding window mechanism realizes the smooth transition of the sampling frequency, avoiding data errors caused by sudden frequency changes. The fuzzy inference system is used to prioritize the triggering conditions, and specific sensors are dynamically activated or blocked based on the signal-to-noise ratio of the sensor signals, ensuring that the collected data is both comprehensive and accurate, greatly improving the quality and efficiency of data acquisition.

[0020] The closed-loop feedback module successfully establishes a close connection between the experimental data and the production batches and process parameters in the MES system. The real-time association algorithm extracts the production batch code and the process parameter sequence to generate a time synchronization index, slides and matches the experimental data timestamp with it by means of a data stream alignment model, and stores the results through an incremental hash table, forming a traceable data-production mapping relationship chain. This function has inestimable value in product quality traceability and production process optimization. When product quality problems occur, enterprises can quickly locate the specific production batches and related process parameters, analyze the causes of the problems, and take targeted improvement measures. Through the correlation analysis of a large amount of historical data, enterprises can also discover the potential laws between process parameters and product performance, providing a strong basis for the optimization of production processes, thereby continuously improving product quality and reducing production costs.

[0021] The two-level verification module provides a solid guarantee for the accuracy and reliability of the data. The double-layer cascaded verification algorithm performs anomaly detection on the collected data in two levels. The first level sets a dynamic threshold range, calculates the upper and lower limit thresholds according to the historical data distribution of the device, and verifies the integrity of the data fields, effectively identifying data that significantly deviates from the normal range and incomplete data. The second level loads the production context rule library and uses a temporal reasoning engine to analyze the rationality of the data sequence, capable of detecting abnormal data that, although within the threshold range, does not conform to the production logic. During the production process, when the device has minor faults or the process parameters change slightly, this module can promptly detect data anomalies, preventing incorrect data from entering subsequent links, ensuring the reliability of the data, and providing accurate data support for enterprise decision-making.

[0022] The multi-dimensional structured coding algorithm adopted by the intelligent storage module greatly improves the efficiency of data storage and retrieval. The batch number and equipment identification are encoded as a fixed-length binary sequence and a unique index is generated, the timestamp is segmented and encoded, and a tree-like label structure is constructed based on the processing stage type, making data storage more orderly. When enterprises perform data retrieval, they can quickly locate the required data based on multiple dimensions such as batch number, timestamp, equipment identification or processing stage. When analyzing the test data of a specific device within a certain period of time, or viewing the data of a batch of products at different processing stages, it can be quickly obtained, which greatly improves the efficiency of data utilization and saves time and labor costs.

[0023] The system integration module realizes data interaction with ERP and SCADA systems based on the interface adaptation algorithm, further expanding the functions and application scope of the system. Define a unified data exchange protocol, convert the experimental data format into the standard input format of the target system, and ensure the smooth transmission of data between different systems. Use the protocol proxy mechanism to parse the material coding rules of the ERP system and the real-time data interface of the SCADA system, and realize the deep integration and sharing of data. Cross-system data buffering is achieved through asynchronous message queues, which improves the stability and response speed of the system. This enables enterprises to integrate data resources from multiple systems, conduct more comprehensive and in-depth data analysis and decision-making, and improve the overall operational efficiency and management level of the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is a working principle diagram of a product test data storage and management system based on the MES system described in the present invention; Figure 2 It is a workflow diagram of the real-time association algorithm; Figure 3 It is a workflow diagram of the multi-dimensional structured coding algorithm; Figure 4 Workflow diagram for data flow alignment model. DETAILED DESCRIPTION

[0025] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions 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 creative work are within the scope of protection of the present invention.

[0026] See also Figures 1-4 The present invention provides a product test data storage and management system based on the MES system, and its overall implementation scheme is as follows: The whole realizes the efficient acquisition, accurate association, strict verification, and orderly storage and interaction of product test data through the collaboration of multiple functional modules. Before the product test process starts, the system first activates the dynamic configuration module, which, based on different experiment types, dynamically configures the data acquisition parameters by means of an adaptive parameter adjustment algorithm. These acquisition parameters include sampling frequency, trigger conditions, and sensor selection, and their reasonable configuration is crucial for obtaining accurate and effective test data. For example, when conducting performance tests on electronic components, depending on whether the test focuses on the transient response or steady-state characteristics of the components, the dynamic configuration module will select different sensors and adjust the sampling frequency and trigger conditions accordingly.

[0027] After the parameter configuration is completed, the data acquisition stage begins. During the acquisition process, the closed-loop feedback module starts to function. It binds the acquired experimental data with the production batch and process parameters in the MES system based on a real-time association algorithm, generating a data-production mapping relationship chain. The establishment of this relationship chain enables each set of test data to be accurately corresponding to a specific production batch and process conditions, providing strong support for subsequent data analysis and quality traceability.

[0028] The acquired data then enters the two-level verification module, where anomaly detection is carried out through a double cascaded verification algorithm. First, at the first level, basic threshold verification and data integrity verification are performed to ensure that the data meets the requirements in terms of basic numerical range and field integrity; then, at the second level, logical rationality verification is carried out in combination with the production context, comprehensively judging the rationality of the data from multiple aspects such as production process and equipment status, effectively excluding the interference of abnormal data on subsequent analysis.

[0029] The verified data is stored and managed by the intelligent storage module. The intelligent storage module adopts a multi-dimensional structured coding algorithm to generate hierarchical storage labels according to batch number, timestamp, equipment identification, and processing stage, realizing the efficient storage and convenient retrieval of data. For example, when it is necessary to query the test data of a specific batch of products at a certain processing stage during the production process, the required data can be quickly located and obtained through these hierarchical storage labels.

[0030] The following further elaborates on the present invention through 5 embodiments.

[0031] Embodiment 1: In this embodiment, the specific implementation of the adaptive parameter adjustment algorithm is emphasized. The core of the adaptive parameter adjustment algorithm lies in constructing a mapping rule library for experimental types and acquisition parameters, which pre-stores parameter templates for different experimental scenarios. For example, in the experimental type of automotive engine performance testing, for the scenario of detecting engine torque changes, specific sampling frequencies, trigger conditions, and applicable sensor types are set, and these parameter combinations form a parameter template and are stored in the mapping rule library. When the system receives the MES production task instruction in real time, it will quickly match the corresponding template in the mapping rule library according to the experimental type information in the instruction.

[0032] After matching the corresponding template, the sliding window mechanism is used to dynamically adjust the smooth transition of the sampling frequency. Suppose the current experiment requires switching from a lower sampling frequency to a higher sampling frequency. The sliding window mechanism will gradually adjust the sampling frequency at a certain time interval to avoid the impact of frequency mutation on the stability of data acquisition. Specifically, let the current sampling frequency be , the target sampling frequency be , the sliding window size be , the frequency increment for each adjustment be , then after adjustments, the sampling frequency , where . During the adjustment process, through the real-time monitoring of the collected data, it is ensured that the sampling frequency after each adjustment can stably obtain valid data.

[0033] A fuzzy inference system is used to prioritize the trigger conditions and dynamically activate or block specific sensors based on the signal-to-noise ratio of the sensor signals. The fuzzy inference system designs a three-input single-output fuzzy controller, and the input variables are the experimental complexity , the environmental noise level , and the equipment load rate . The experimental complexity is used to measure the complexity of the experiment itself. For example, a complex multi-physical field coupling experiment has a higher complexity; the environmental noise level reflects the noise interference degree of the environment where the experiment is located; the equipment load rate represents the working load of the equipment when collecting data. The output variable is the sensor activation weight .

[0034] Triangular membership functions are constructed to divide the fuzzy sets of the input variables. Taking the experimental complexity as an example, it is divided into three fuzzy subsets: low, medium, and high, and the membership degree of each input value in different fuzzy subsets is determined through triangular membership functions. Similarly, for the environmental noise level and the equipment load rate Perform a similar fuzzy division. Fuzzy inference is performed on the input variables through a fuzzy rule base, which contains a large number of rules determined based on experience and experiments, such as "if the experimental complexity is high, the environmental noise level is high, the equipment load rate is medium, then the sensor activation weight is low" and other rules. Finally, the center-of-gravity method is used to defuzzify and generate sensor selection instructions. According to the magnitude of the sensor activation weight , specific sensors are activated or blocked, thereby optimizing the data acquisition process.

[0035] Example 2: This example details the implementation process of the real-time association algorithm. The real-time association algorithm first extracts the production batch code and process parameter sequence in the MES system to generate a time synchronization index. The production batch code is the unique identifier of each production batch. For example, "20230801-001" represents the first batch of products produced on August 1, 2023; the process parameter sequence contains key parameters in the production process, such as temperature, pressure, etc. By extracting this information and combining time information, a time synchronization index is generated to provide a time reference for subsequent data matching.

[0036] Construct a data stream alignment model to perform sliding matching between the timestamps of experimental data and the time index of production batches. The experimental data stream is divided into data blocks of fixed length, and each data block is attached with a check code and a timestamp. Let the data block length be , the check code be , the timestamp be , and the data block be represented as . The dynamic time warping algorithm is used to align the production batch time axis and the experimental data time axis. The dynamic time warping algorithm can find the best time alignment path between two time series with different lengths and speed changes. Assume that the production batch time series is , and the experimental data time series is . Through the dynamic time warping algorithm, a warping path is calculated, where represents the th time point in the production batch time series aligned with the th time point in the experimental data time series .

[0037] Fill the data missing segments caused by transmission delays through an interpolation compensation mechanism. When it is found that the experimental data time series When there are large data missing segments in the time interval, according to the changing trends of the data before and after, linear interpolation or other appropriate interpolation methods are used to fill the data. Let the missing data point be , and its adjacent data points before and after are and , and the timestamps are , and , respectively. Then, through the linear interpolation formula

[0038]

[0039] calculate the value of the missing data point to ensure the integrity and continuity of the data, so as to accurately establish the data-production mapping relationship chain.

[0040] Embodiment 3: This embodiment deeply introduces the specific operations of the double-level cascade verification algorithm. In the first level of the double-level cascade verification algorithm, a dynamic threshold interval is set. Calculate the upper and lower limit thresholds according to the historical data distribution of the device. Assume that the historical data set of the device is , and calculate the mean value and the standard deviation of the data through statistical analysis. Then, the lower limit of the dynamic threshold , and the upper limit of the dynamic threshold , where and are coefficients determined according to the experimental requirements and data characteristics. During the data acquisition process, compare the collected data with the dynamic threshold interval. If or , it is marked as suspected abnormal data. At the same time, verify the integrity of the data fields, check whether the data contains all necessary field information, such as timestamps, device identifiers, measurement values, etc. If there are missing fields, determine that the data is incomplete and perform corresponding processing.

[0041] In the second level, load the production context rule library, which includes process constraints, device status association relationships, and abnormal pattern libraries. For example, in the production test of a certain chemical product, the process constraints stipulate that the reaction temperature is within a specific interval, the device status association relationship indicates that the operating states of two certain devices should maintain a certain synchronization, and the abnormal pattern library stores common abnormal patterns such as sudden temperature rise and abnormal pressure fluctuation. Analyze the rationality of the data sequence through the time series inference engine. The time series inference engine constructs a hidden Markov model to describe the transition probability of the normal data sequence. Let the state set of the hidden Markov model be , and the state transition probability matrix be , where represents the transition from state to state Probability.

[0042] Extract the feature vectors of the abnormal data segment, including the statistic deviation degree, the density of mutation points, and the status of associated devices. The statistic deviation degree is used to measure the difference between the data and the normal statistical characteristics. The density of mutation points represents the density of mutation points in the data sequence. The status of associated devices reflects the operating status of related devices when the data is abnormal. Calculate the abnormal probability through the Bayesian network. Let the node set of the Bayesian network be , and the edge set be . Given the observed data , calculate the abnormal probability according to the Bayesian formula , where represents the abnormal event, and represents all possible events. If the abnormal probability exceeds the preset threshold, it is marked as abnormal data and triggers the MES warning signal to notify relevant personnel for timely processing.

[0043] Example 4: This example focuses on the specific implementation of the multi-dimensional structured coding algorithm. The multi-dimensional structured coding algorithm encodes the batch number and device identifier into a fixed-length binary sequence and generates a unique index through a hash function. Assume the batch number is , and the device identifier is . Combine them into a string , and use the hash function to generate a unique index . The selection of the hash function needs to ensure that different inputs can generate different index values as much as possible to reduce conflicts. For example, use the SHA-256 hash function to calculate different combinations of batch numbers and device identifiers to obtain a fixed-length hash value as the unique index.

[0044] Encode the timestamp in segments to distinguish between year-month-day and millisecond-level precision. Assume the timestamp is . Decompose it into year , month , day , hour , minute , second and millisecond . Encode the year, month, and day into one segment. For example, use binary encoding to represent the first 4 bits of the year, the 2 bits of the month, and the 5 bits of the day; encode the hour, minute, second, and millisecond into another segment. For example, use binary encoding to represent the 5 bits of the hour, the 6 bits of the minute, the 6 bits of the second, and the 10 bits of the millisecond. Such a segmented encoding method facilitates quick retrieval of data according to the time range.

[0045] Construct a tree - like tag structure based on the type of processing stage, supporting data retrieval according to the experimental preparation, execution, and post - processing stages. Taking the experimental preparation stage as an example, create a sub - node "Experimental Preparation" under the root node of the tree - like tag structure, and then further divide it into grand - children nodes such as "Equipment Calibration" and "Sample Preparation" under this sub - node. When storing data, add the corresponding tree - like tag to it according to the processing stage to which the data belongs. For example, if a piece of data is collected during equipment calibration, it is marked as "Experimental Preparation - Equipment Calibration". When retrieving data, by traversing the tree - like tag structure, the required data can be quickly located, improving the data retrieval efficiency.

[0046] Example 5:

[0047] This example elaborates in detail the working process of the system integration module and the interface adaptation algorithm. The system integration module realizes data interaction with ERP and SCADA systems based on the interface adaptation algorithm, supporting modular expansion and protocol conversion. The interface adaptation algorithm first defines a unified data exchange protocol and converts the experimental data format into the standard input format of the target system. Suppose the experimental data format is , and the standard input format of the ERP system is , and the standard input format of the SCADA system is . Through the data format conversion function , where can be or , the experimental data is format - converted. For example, the timestamp format in the experimental data is converted from one representation to the format required by the ERP system.

[0048] Adopt a protocol proxy mechanism to parse the material coding rules of the ERP system and the real - time data interface of the SCADA system. Parse the bill of materials structure of the ERP system and extract the material attribute fields related to the experimental data. Suppose the bill of materials of the ERP system is , and extract the material number , material name , material specification and other attribute fields related to the experimental data. Build a state - machine model for the communication channel of the SCADA system to manage the connection establishment, data encapsulation, and error re - transmission processes. The state - machine model includes a connection - establishment state , data - transmission state , error - handling state , etc. In the connection - establishment state, by sending a connection request instruction Attempt to establish a connection with the SCADA system; in the data transmission state, encapsulate the converted data according to the protocol of the SCADA system, such as adding a frame header, a frame tail, and a check code, etc.; in the error handling state, if a data transmission error occurs, execute the corresponding error retransmission strategy according to the error type.

[0049] Implement cross-system data buffering through an asynchronous message queue. The asynchronous message queue serves as a data buffer, receiving the data to be transmitted sent from this system and the data received from the ERP and SCADA systems. Suppose the asynchronous message queue , when there is data that needs to be sent to the target system, put the data into the message queue ; when receiving data from the target system, take out the data from the message queue . Maintain the stability of the interface through the heartbeat detection and timeout reconnection mechanism. Regularly send heartbeat detection messages , if the response from the target system is not received within the specified time , it is determined that the connection has timed out, and execute the timeout reconnection operation to ensure the stable progress of data interaction between systems.

[0050] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not expressly listed, or elements inherent to such process, method, article or device.

[0051] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A product test data storage and management system based on MES system, characterized in that: include: Dynamic configuration module: used to dynamically configure data acquisition parameters using an adaptive parameter adjustment algorithm according to the experiment type. The data acquisition parameters include sampling frequency, trigger conditions and sensor selection; Closed-loop feedback module: Based on the real-time association algorithm, the experimental data is bound to the production batches and process parameters in the MES system to generate a data-production mapping relationship chain; Two-level verification module: Anomaly detection is performed on collected data through a two-layer cascade verification algorithm. The first layer performs basic threshold verification and data integrity verification, and the second layer performs logical rationality verification based on production context. Intelligent storage module: uses a multi-dimensional structured coding algorithm to generate hierarchical storage tags for experimental data according to batch number, timestamp, equipment identification and processing stage.

2. The product test data storage and management system based on the MES system according to claim 1, characterized in that: The adaptive parameter adjustment algorithm includes: Build a mapping rule base between experiment types and acquisition parameters. The rule base contains parameter templates for different experiment scenarios. Match the corresponding template according to the MES production task instructions received in real time, and dynamically adjust the smooth transition of the sampling frequency through the sliding window mechanism; A fuzzy inference system is used to prioritize the trigger conditions and dynamically activate or shield specific sensors based on the signal-to-noise ratio of the sensor signal.

3. The product test data storage and management system based on the MES system according to claim 1, characterized in that: The real-time association algorithm includes: Extract the production batch code and process parameter sequence in the MES system and generate a time synchronization index; Build a data stream alignment model to perform sliding matching between the timestamp of experimental data and the time index of production batches; The matching results are stored in an incremental hash table to form a traceable data-production mapping relationship chain.

4. The product test data storage and management system based on the MES system according to claim 1, characterized in that: The two-layer cascade verification algorithm includes: At the first level, dynamic threshold intervals are set, upper and lower thresholds are calculated based on the historical data distribution of the device, and the integrity of the data field is verified; At the second level, a production context rule base is loaded, wherein the rule base includes process constraints, equipment status associations, and an abnormal pattern base; The rationality of the data sequence is analyzed through the time series reasoning engine, and if a violation of the context logic is detected, it is marked as abnormal data.

5. The product test data storage and management system based on the MES system according to claim 1, characterized in that: The multi-dimensional structured coding algorithm comprises: Encode the batch number and equipment identification into a fixed-length binary sequence and generate a unique index through a hash function; Encode timestamps in segments, distinguishing year-month-day and millisecond accuracy; A tree-like label structure is constructed based on the processing stage type, supporting data retrieval by experimental preparation, execution, and post-processing stages.

6. The product test data storage and management system based on the MES system according to claim 1, characterized in that: The system further comprises: System integration module: realizes data interaction with ERP and SCADA systems based on interface adaptation algorithm, and supports modular expansion and protocol conversion; The interface adaptation algorithm includes: Define a unified data exchange protocol to convert the experimental data format into the standard input format of the target system; Use protocol proxy mechanism to parse material coding rules of ERP system and real-time data interface of SCADA system; Cross-system data buffering is achieved through asynchronous message queues.

7. The product test data storage and management system based on the MES system according to claim 2, characterized in that: The fuzzy inference system comprises: Design a three-input single-output fuzzy controller, the input variables are the experimental complexity, environmental noise level and equipment load rate; Construct a triangular membership function to partition the fuzzy set of input variables, and the output variable is the sensor activation weight; The sensor selection instructions are generated by defuzzification using the center of gravity method.

8. The product test data storage and management system based on the MES system according to claim 3, characterized in that: The data stream alignment model includes: Divide the experimental data stream into data blocks of fixed length, and attach a checksum and timestamp to each data block; Use dynamic time warping algorithm to align production batch timeline with experimental data timeline; The data missing segments caused by transmission delay are filled through interpolation compensation mechanism.

9. The product test data storage and management system based on the MES system according to claim 4, characterized in that: The temporal reasoning engine comprises: Construct a hidden Markov model to describe the transition probability of normal data sequences; Extract the feature vector of the abnormal data segment, including the statistical deviation, mutation point density and related equipment status; The abnormal probability is calculated through the Bayesian network, and if it exceeds the preset threshold, the MES warning signal is triggered.

10. The product test data storage and management system based on the MES system according to claim 6, characterized in that: The protocol proxy mechanism includes: Parse the bill of materials structure of the ERP system and extract the material attribute fields related to the experimental data; Build a state machine model of the SCADA system communication channel to manage the connection establishment, data encapsulation and error retransmission processes; The stability of the interface is maintained through heartbeat detection and timeout reconnection mechanism.

Images