A production line state evaluation method fusing big data analysis

Abstract

This invention discloses a production line status assessment method integrating big data analysis, relating to the fields of big data and industrial production line monitoring technology. It includes: S1, collecting multi-dimensional production line equipment data, including basic equipment operating parameters and inter-equipment interaction data; S2, performing time-series alignment and noise suppression processing on the multi-dimensional production line equipment data to generate pre-processed data. This invention ensures data quality by collecting multi-dimensional production line equipment data and performing time-series alignment and noise suppression processing. Based on the mutual information entropy algorithm, it constructs a dynamic coupling relationship network to quantify inter-equipment correlations in real time without reconstructing the network, overcoming the limitations of existing static correlations. Combined with the shortest path algorithm to locate key coupling transmission paths, and utilizing a drift transmission coefficient containing a process sensitivity coefficient and an implicit drift superposition effect quantification model incorporating a coupling attenuation factor, it accurately calculates the overall line performance degradation value and can also fit trend curves to predict the degradation rate and critical time.

Description

Technical Field

[0001] This invention relates to the field of big data and industrial production line monitoring technology, specifically a production line status assessment method that integrates big data analysis. Background Technology

[0002] In modern industrial production, the operational status of production lines directly impacts production efficiency and product quality. To monitor production line status, companies commonly employ data acquisition systems to collect equipment operating parameters in real time and utilize data analysis methods to conduct an overall assessment of production line health. These methods typically rely on statistical process control or rule-based diagnostic logic, enabling the identification of significant equipment anomalies or performance fluctuations, thus providing a degree of decision support for production management.

[0003] However, existing technologies struggle to effectively identify and quantify the latent state drift caused by the coupling effects of multiple devices. Specifically, in complex production lines, multiple equipment units interact dynamically. Minor parameter deviations in a single device can gradually propagate and accumulate through coupling links, ultimately leading to a slow decline in overall line performance, while the parameters of each individual device remain within normal threshold ranges. This drift process cannot be captured by conventional threshold alarms or single-point data analysis. Existing evaluation models are mostly based on the assumption of device independence or static correlation rules, failing to establish a real-time quantification mechanism for the dynamic coupling relationships between multiple devices. This results in the gradual degradation of production line performance going undetected and unannounced, increasing the risk of unplanned downtime and the hidden increase in maintenance costs. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a production line status assessment method that integrates big data analysis, solving the problems of existing technologies being unable to identify implicit state drift caused by multi-device coupling and unable to capture gradual performance degradation of the entire production line in a timely manner.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a production line status assessment method integrating big data analysis, comprising: S1. Collect multi-dimensional production line equipment data, which includes basic operating parameters of the equipment and data related to interactions between equipment. S2. Perform time-series alignment and noise suppression processing on multi-dimensional production line equipment data to generate preprocessed data; S3. Based on the preprocessed data, the mutual information entropy algorithm is used to calculate the coupling strength between any two devices and a dynamic coupling relationship network is constructed. The dynamic coupling relationship network uses devices as nodes and coupling strength values ​​as edge weights. The edge weights are dynamically adjusted in real time as the data is updated. The dynamic adjustment is achieved through a local topology update algorithm without reconstructing the entire network. S4. Based on the dynamic coupling relationship network, the shortest path algorithm is used to locate the key coupling transmission path of the small parameter offset of a single device. For the key coupling transmission path, the drift transmission coefficient is calculated, and a quantification model of implicit drift superposition effect is constructed to calculate the quantification value of the overall line performance degradation. The quantification model of implicit drift superposition effect introduces a coupling attenuation factor to correct the energy loss of the offset during the transmission process. S5. Based on the quantitative value of the overall line performance degradation, set multi-level early warning thresholds, trigger early warning signals, and generate corresponding maintenance suggestions based on key coupling transmission paths and core equipment. S6. Feed back the early warning decision results to the data acquisition stage, use reinforcement learning algorithms to optimize the acquisition parameters, and update the initial parameters for dynamic coupling relationship modeling.

[0006] Preferably, the collection of multi-dimensional production line equipment data includes: Based on the process dependencies of production line equipment, preset coupling and correlation dimension screening rules are used to focus on four correlation dimensions: material transfer, energy exchange, signal response, and control command interaction, and basic operating parameters of the equipment are collected. The equipment's basic operating parameters are continuously collected using a real-time sampling mode, which collects data at dynamically adjusted time intervals. The time-series slicing mode is used to collect data on the interaction between devices. The time-series slicing mode divides the time window into integer multiples of the production line process cycle time and extracts the interaction feature values ​​within each time window. The interaction feature values ​​include interaction frequency, peak intensity, and timing synchronization degree.

[0007] Preferably, the step of calculating the coupling strength between any two devices using the mutual information entropy algorithm includes: For any two devices and Calculate the mutual information entropy between its operating parameters and interactive data. Mutual information entropy is a value representing the coupling strength. Calculated using the following formula: ; in, Indicates equipment Value And equipment Value The joint probability density function, Indicates equipment Value The marginal probability density function, Indicates equipment Value The marginal probability density function; the coupling strength value is dynamically adjusted in real time as the preprocessed data is updated.

[0008] Preferably, after the dynamic coupling network is constructed, the method further includes: Set the coupling relationship update trigger conditions, which include the fluctuation range of device parameters reaching a preset threshold or the change rate of interactive data exceeding a set standard. When the conditions for updating coupling relationships are met, the coupling strength is automatically recalculated and the network topology is adjusted. When the coupling relationship update trigger condition is not met, an interpolation algorithm is used to smoothly update the coupling strength value.

[0009] Preferably, the calculation of the drift transfer coefficient for the critical coupling transfer path includes: For each critical coupling transmission path, a weighted product algorithm is used to calculate the drift transmission coefficient based on the coupling strength value of each node on the path and the equipment process sensitivity coefficient. Drift transfer coefficient Calculated using the following formula: ; in, Indicates the first path The coupling strength value of the strip edge, Indicates the first path The process sensitivity coefficient of each piece of equipment This indicates the number of nodes on the critical coupling transmission path; the process sensitivity coefficient is preset to an initial value based on the functional attributes of the equipment in the production line and is iteratively optimized through historical data.

[0010] Preferably, the step of constructing a quantitative model for the implicit drift superposition effect and calculating the quantitative value of the overall line performance degradation includes: Parameter offset of a single device As input; The offset values ​​transmitted across each associated device are calculated using the drift transmission coefficient of the key coupled transmission path. ; The weighted summation algorithm is used to calculate the superposition effect of all propagation offset values ​​to obtain the quantified value of overall line performance degradation. Quantitative value of overall line performance degradation Calculated using the following formula: ; in, Indicates the first The transmission offset value on each associated device Indicates the first The coupling attenuation factor of each associated device Indicates the number of associated devices.

[0011] Preferably, after calculating the quantification value of overall line performance degradation, the method further includes: Based on the historical quantitative values ​​of overall line performance degradation, a polynomial fitting algorithm is used to construct an overall line performance degradation trend curve. By analyzing the slope and curvature of the trend curve, we can predict the subsequent rate of decline and the critical warning time.

[0012] Preferably, the overall performance degradation trend curve is fitted by the following formula: ; in, Indicates time The quantification value of the overall line performance degradation. Represents the polynomial coefficients. This represents the order of the polynomial.

[0013] Preferably, the step of setting multi-level warning thresholds and triggering warning signals includes: Preset warning thresholds for four levels: safety, attention, early warning, and emergency; Based on the implicit drift quantification value and the decline trend prediction results, when the quantification value reaches the corresponding level threshold, a corresponding early warning signal is triggered. The early warning signal includes the content of the early warning information, the target audience, and the response time limit.

[0014] Preferably, the step of feeding back the early warning decision result to the data acquisition stage and optimizing the acquisition parameters using a reinforcement learning algorithm includes: For devices on critical coupling transmission paths, increase the sampling frequency of their operating parameters and the slicing density of interactive data; For weakly coupled devices, reduce the sampling frequency of their operating parameters and the slicing density of interactive data; The feedback results synchronously update the initial parameters of the dynamic coupling model and optimize the weight allocation for coupling strength calculation.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention ensures data quality by collecting multi-dimensional production line equipment data and performing time-series alignment and noise suppression. It constructs a dynamic coupling relationship network based on the mutual information entropy algorithm to quantify the correlation between equipment in real time without requiring network reconstruction, overcoming the limitations of existing static correlations. Combining the shortest path algorithm to locate key coupling transmission paths, and leveraging a drift transmission coefficient with process sensitivity coefficients and a implicit drift superposition effect quantification model incorporating coupling attenuation factors, it accurately calculates the overall line performance degradation value and can even fit trend curves to predict degradation speed and critical time. This solves the problems of existing technologies struggling to identify implicit state drift caused by multi-equipment coupling and failing to capture the superposition of minute parameter shifts in individual equipment. Furthermore, it triggers signals in a timely manner through multi-level early warning and generates maintenance suggestions, using reinforcement learning to optimize collected parameters and initial modeling parameters. Ultimately, this reduces unplanned downtime, lowers maintenance costs, and significantly improves production line stability and efficiency. Attached Figure Description

[0016] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a flowchart illustrating the construction process of the dynamic coupling relationship network of the present invention. Figure 3 This is a flowchart of the closed-loop optimization feedback process of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1-3 This embodiment provides a production line status assessment method integrating big data analytics, using an automotive parts assembly line as an application scenario. This line comprises 20 core production devices, covering multiple functional modules such as material conveying, processing and forming, and testing and verification. The devices form complex coupling relationships through material transfer, energy exchange, signal response, and control command interaction. Traditional assessment methods struggle to capture the overall line performance degradation caused by the accumulation of hidden drift. The following detailed description of the production line status assessment method integrating big data analytics of this invention is provided in conjunction with specific technical details.

[0019] First, a multi-dimensional production line equipment data acquisition process is executed. Based on the process dependencies of the production line equipment, preset coupling and correlation dimension filtering rules are established, focusing on collecting basic equipment operating parameters across four correlation dimensions: material transfer, energy exchange, signal response, and control command interaction. These basic operating parameters include key indicators such as equipment speed, operating temperature, output pressure, energy consumption data, and control voltage. Real-time sampling is used for continuous acquisition of these basic operating parameters. The initial time interval for real-time sampling is set to 1 second, and this interval is dynamically adjusted according to the real-time operating load of the production line. When the production line load exceeds 80% of the rated load, the time interval is shortened to 0.5 seconds; when the load is below 30%, it is extended to 2 seconds. The data related to the interaction between devices is collected using a time-series slicing mode. The time-series slicing mode divides the time window into integer multiples of the production line process cycle time. In this embodiment, the production line process cycle time is 10 seconds, so the time window is set to 10 seconds, that is, the interaction feature value is extracted once every 10 seconds. The interaction feature value includes interaction frequency, peak intensity, and time synchronization degree. The interaction frequency counts the number of data interactions between devices per unit time, the peak intensity records the maximum amplitude of the interaction data, and the time synchronization degree is obtained by calculating the normalized result of the device action response time difference.

[0020] After data acquisition, time alignment and noise suppression are performed on multi-dimensional production line equipment data to generate preprocessed data. Time alignment uses linear interpolation, with the system time of the production line main controller as the benchmark, to calibrate the timestamps of data collected from different devices, ensuring that all data are consistent in the time dimension. Noise suppression uses wavelet threshold denoising, selecting a db4 wavelet basis to perform three-level decomposition of the data. The noise threshold is set to 1.2 times the root mean square value of the high-frequency coefficients after decomposition. After processing the high-frequency noise coefficients through the threshold function, the data is reconstructed, effectively filtering out noise signals caused by electromagnetic interference and sensor errors, ensuring data accuracy.

[0021] Based on preprocessed data, the mutual information entropy algorithm is used to calculate the coupling strength between any two devices and construct a dynamic coupling relationship network. For any two devices X and Y, the mutual information entropy I(X;Y) between their operating parameters and interactive data is calculated as the coupling strength value. The formula for calculating the mutual information entropy I(X;Y) is as follows: ; Where p(x,y) represents the joint probability density function of device X taking the value x and device Y taking the value y, p(x) represents the marginal probability density function of device X taking the value x, and p(y) represents the marginal probability density function of device Y taking the value y. Both the joint probability density function and the marginal probability density function are calculated using the kernel density estimation method. The dynamic coupling relationship network uses devices as nodes and coupling strength values ​​as edge weights. The edge weights are dynamically adjusted in real time as the preprocessed data is updated. The dynamic adjustment is achieved through a local topology update algorithm without reconstructing the entire network. After the dynamic coupling relationship network is constructed, coupling relationship update trigger conditions are set. An update is triggered when the fluctuation amplitude of device parameters reaches a preset threshold or the change rate of interactive data exceeds a set standard. In this embodiment, the parameter fluctuation amplitude threshold is set to ±5%, and the change rate standard of interactive data is set to 20%. When the update trigger conditions are met, the coupling strength is automatically recalculated and the network topology is adjusted; when not met, a linear interpolation algorithm is used to smoothly update the coupling strength value, ensuring that the network state matches the actual operating state of the devices in real time, solving the problem that traditional static association models cannot adapt to dynamic device interactions.

[0022] Based on the dynamic coupling relationship network, the shortest path algorithm is used to locate the critical coupling transmission path of a single device with a small parameter offset. In this embodiment, the shortest path algorithm selected is Dijkstra's algorithm. First, the coupling strength weight of each edge in the dynamic coupling relationship network is converted into its reciprocal. Then, Dijkstra's algorithm is used to calculate the shortest path with the minimum weight between devices. This path is the critical coupling transmission path with the maximum total coupling strength in the original network. By calculating the path length after weighting the coupling strength between devices, the three paths with the maximum total coupling strength are selected as the critical coupling transmission paths.

[0023] For each critical coupling transmission path, calculate the drift transmission coefficient. The calculation formula is: ; in, This represents the coupling strength value of the k-th edge on the path. Let N represent the process sensitivity coefficient of the k-th device on the path, and N represent the number of nodes on the critical coupling transmission path. The process sensitivity coefficient is preset with initial values ​​based on the functional attributes of the equipment in the production line: core processing equipment is initially set to 0.9, auxiliary conveying equipment to 0.3, and detection equipment to 0.7. Then, iterative optimization using gradient descent based on historical operating data from the past three months is employed to obtain a stable process sensitivity coefficient. Subsequently, a quantification model of the implicit drift superposition effect is constructed to calculate the quantified value of the overall line performance degradation, and the parameter offset of a single device is considered. As input, the transfer offset value of the offset on each associated device is calculated using the drift transfer coefficient of the key coupled transfer path. Then, a weighted summation algorithm is used to calculate the superposition effect of all propagation offset values ​​to obtain the quantified value of overall line performance degradation. The calculation formula is as follows: ; in, This represents the transfer offset value on the j-th associated device. Let M represent the coupling attenuation factor of the j-th associated device, and M represent the number of associated devices. The coupling attenuation factor is determined based on the physical distance between devices and the signal transmission medium. The attenuation factor value is obtained by experimentally measuring the transmission loss of offset at different distances and using exponential fitting. For example, the coupling attenuation factor for adjacent devices is set to 0.95, and for devices separated by one device, it is set to 0.85, effectively correcting for energy loss during offset transmission. This model can accurately capture the transmission and superposition effects of minute parameter offsets of a single device through the coupling path. Even if the parameters of a single device are within the normal threshold range, it can quantify the gradual degradation of the overall line performance, solving the technical problem of unidentifiable implicit drift in the background technology.

[0024] After calculating the quantified value of overall line performance degradation, a polynomial fitting algorithm is used to construct the overall line performance degradation trend curve based on the historical quantified values. The fitting formula is as follows: ; Where P(t) represents the quantized value of the line performance degradation at time t. Here, represents the polynomial coefficients, and n represents the order of the polynomial. In this embodiment, the polynomial order n=3 was determined through experiments. The least squares method was used to fit the historical data of the past 6 months to obtain the polynomial coefficients. By analyzing the slope and curvature of the trend curve, the subsequent rate of decline and the critical warning time point can be predicted. For example, when the slope of the trend curve is greater than 0.05 / day and the curvature is greater than 0, it is determined that the rate of decline is accelerating, and an early warning is issued to allow sufficient maintenance time.

[0025] Set multi - level warning thresholds and trigger warning signals according to the quantified value of the overall line performance decline. Preset warning thresholds for four levels: safe, attention, warning, and emergency. The thresholds are determined by statistically analyzing the fault data of the production line in the past two years. When the quantified value of the overall line performance decline D ≤ 0.1, it is the safe level; when 0.1 < D ≤ 0.3, it is the attention level; when 0.3 < D ≤ 0.6, it is the warning level; when D > 0.6, it is the emergency level. According to the quantified value of the hidden drift and the prediction result of the decline trend, when the quantified value reaches the corresponding level threshold, trigger the corresponding warning signal. The warning signal includes the content of the warning message, the push object, and the response time limit. For the safe level, only record the log without active push. For the attention level, push a prompt message to the production line administrator with a response time limit of 24 hours. For the warning level, push a detailed analysis report to the maintenance engineer with a response time limit of 6 hours. For the emergency level, push an alarm message to both the administrator and the maintenance engineer and trigger on - site audible and visual alarms with a response time limit of 1 hour. At the same time, generate corresponding maintenance suggestions based on the key coupling transfer path and the core influencing equipment. For example, when a small rotational speed deviation of a certain processing machine tool is detected and transmitted through the coupling path, it is recommended to calibrate the rotational speed sensor of the machine tool first and check the transmission accuracy of the associated conveying equipment to achieve precise maintenance.

[0026] Finally, feedback the warning decision result to the data acquisition link, and use the reinforcement learning algorithm to optimize the acquisition parameters. The reinforcement learning algorithm uses the data acquisition efficiency and evaluation accuracy as the reward function. For the equipment on the key coupling transfer path, increase the sampling frequency of its operating parameters and the slicing density of the interaction data. In this embodiment, the sampling frequency is increased to twice the original, and the slicing density is adjusted to 5 seconds / window. For the weakly - coupled equipment, reduce the sampling frequency of its operating parameters and the slicing density of the interaction data. The sampling frequency is reduced to 50% of the original, and the slicing density is adjusted to 20 seconds / window. The feedback result synchronously updates the initial parameters of the dynamic coupling modeling and optimizes the weight distribution of the coupling strength calculation. For example, increase the coupling strength calculation weight of the core equipment by 15% to make the subsequent construction of the coupling relationship network more in line with the key node state of the production line, forming a closed - loop mechanism of "acquisition - processing - evaluation - feedback - optimization", reducing the computing power consumption of data acquisition and processing while ensuring the evaluation accuracy.

[0027] In summary, this invention ensures data quality by collecting multi-dimensional production line equipment data and performing time-series alignment and noise suppression. It constructs a dynamic coupling relationship network based on the mutual information entropy algorithm to quantify the correlation between equipment in real time without requiring network reconstruction, overcoming the limitations of existing static correlations. Combining the shortest path algorithm to locate key coupling transmission paths, and leveraging a drift transmission coefficient with process sensitivity coefficients and a implicit drift superposition effect quantification model incorporating coupling attenuation factors, it accurately calculates the overall line performance degradation value and can even fit trend curves to predict degradation speed and critical time. This solves the problems of existing technologies struggling to identify implicit state drift caused by multi-equipment coupling and failing to capture the superposition of minute parameter shifts in individual equipment. Furthermore, it triggers signals in a timely manner through multi-level early warning and generates maintenance suggestions, using reinforcement learning to optimize collected parameters and initial modeling parameters. Ultimately, this reduces unplanned downtime, lowers maintenance costs, and significantly improves production line stability and efficiency.

[0028] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0029] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A production line status assessment method integrating big data analysis, characterized in that, include: S1. Collect multi-dimensional production line equipment data, which includes basic operating parameters of the equipment and data related to interactions between equipment. S2. Perform time-series alignment and noise suppression processing on multi-dimensional production line equipment data to generate preprocessed data; S3. Based on the preprocessed data, the mutual information entropy algorithm is used to calculate the coupling strength between any two devices and a dynamic coupling relationship network is constructed. The dynamic coupling relationship network uses devices as nodes and coupling strength values ​​as edge weights. The edge weights are dynamically adjusted in real time as the data is updated. The dynamic adjustment is achieved through a local topology update algorithm without reconstructing the entire network. S4. Based on the dynamic coupling relationship network, the shortest path algorithm is used to locate the key coupling transmission path of the small parameter offset of a single device. For the key coupling transmission path, the drift transmission coefficient is calculated, and a quantification model of implicit drift superposition effect is constructed to calculate the quantification value of the overall line performance degradation. The quantification model of implicit drift superposition effect introduces a coupling attenuation factor to correct the energy loss of the offset during the transmission process. S5. Based on the quantitative value of the overall line performance degradation, set multi-level early warning thresholds, trigger early warning signals, and generate corresponding maintenance suggestions based on key coupling transmission paths and core equipment. S6. Feed back the early warning decision results to the data acquisition stage, use reinforcement learning algorithms to optimize the acquisition parameters, and update the initial parameters for dynamic coupling relationship modeling.

2. The production line status assessment method integrating big data analysis according to claim 1, characterized in that, The collected multi-dimensional production line equipment data includes: Based on the process dependencies of production line equipment, preset coupling and correlation dimension screening rules are used to focus on four correlation dimensions: material transfer, energy exchange, signal response, and control command interaction, and basic operating parameters of the equipment are collected. The equipment's basic operating parameters are continuously collected using a real-time sampling mode, which collects data at dynamically adjusted time intervals. The time-series slicing mode is used to collect data on the interaction between devices. The time-series slicing mode divides the time window into integer multiples of the production line process cycle time and extracts the interaction feature values ​​within each time window. The interaction feature values ​​include interaction frequency, peak intensity, and timing synchronization degree.

3. The production line status assessment method integrating big data analysis according to claim 1, characterized in that, The calculation of the coupling strength between any two devices using the mutual information entropy algorithm includes: For any two devices and Calculate the mutual information entropy between its operating parameters and interactive data. Mutual information entropy is a value representing the coupling strength. Calculated using the following formula: ； in, Indicates equipment Value And equipment Value The joint probability density function, Indicates equipment Value The marginal probability density function, Indicates equipment Value The marginal probability density function; the coupling strength value is dynamically adjusted in real time as the preprocessed data is updated.

4. The production line status assessment method integrating big data analysis according to claim 1, characterized in that, After the dynamic coupling relationship network is constructed, the method further includes: Set the coupling relationship update trigger conditions, which include the fluctuation range of device parameters reaching a preset threshold or the change rate of interactive data exceeding a set standard. When the conditions for updating coupling relationships are met, the coupling strength is automatically recalculated and the network topology is adjusted. When the coupling relationship update trigger condition is not met, an interpolation algorithm is used to smoothly update the coupling strength value.

5. The production line status assessment method integrating big data analysis according to claim 1, characterized in that, The calculation of drift transfer coefficients for key coupled transmission paths includes: For each critical coupling transmission path, a weighted product algorithm is used to calculate the drift transmission coefficient based on the coupling strength value of each node on the path and the equipment process sensitivity coefficient. Drift transfer coefficient Calculated using the following formula: ； in, Indicates the first path The coupling strength value of the strip edge, Indicates the first path The process sensitivity coefficient of each piece of equipment This indicates the number of nodes on the critical coupling transmission path; the process sensitivity coefficient is preset to an initial value based on the functional attributes of the equipment in the production line and is iteratively optimized through historical data.

6. The production line status assessment method integrating big data analysis according to claim 1, characterized in that, The construction of the implicit drift superposition effect quantification model and the calculation of the overall line performance degradation quantification value include: Parameter offset of a single device As input; The offset values ​​transmitted across each associated device are calculated using the drift transmission coefficient of the key coupled transmission path. ; The weighted summation algorithm is used to calculate the superposition effect of all propagation offset values ​​to obtain the quantified value of overall line performance degradation. Quantitative value of overall line performance degradation Calculated using the following formula: ； in, Indicates the first The transmission offset value on each associated device Indicates the first The coupling attenuation factor of each associated device Indicates the number of associated devices.

7. The production line status assessment method integrating big data analysis according to claim 1, characterized in that, After calculating the quantification value of overall line performance degradation, the method further includes: Based on the historical quantitative values ​​of overall line performance degradation, a polynomial fitting algorithm is used to construct an overall line performance degradation trend curve. By analyzing the slope and curvature of the trend curve, we can predict the subsequent rate of decline and the critical warning time.

8. The production line status assessment method integrating big data analysis according to claim 7, characterized in that, The overall performance degradation trend curve is fitted by the following formula: ； in, Indicates time The quantification value of the overall line performance degradation. Represents the polynomial coefficients. This represents the order of the polynomial.

9. The production line status assessment method integrating big data analysis according to claim 1, characterized in that, The setting of multi-level early warning thresholds to trigger early warning signals includes: Preset warning thresholds for four levels: safety, attention, early warning, and emergency; Based on the implicit drift quantification value and the decline trend prediction results, when the quantification value reaches the corresponding level threshold, a corresponding early warning signal is triggered. The early warning signal includes the content of the early warning information, the target audience, and the response time limit.

10. The production line status assessment method integrating big data analysis according to claim 1, characterized in that, The step of feeding back the early warning decision results to the data acquisition stage and optimizing the acquisition parameters using a reinforcement learning algorithm includes: For devices on critical coupling transmission paths, increase the sampling frequency of their operating parameters and the slicing density of interactive data; For weakly coupled devices, reduce the sampling frequency of their operating parameters and the slicing density of interactive data; The feedback results synchronously update the initial parameters of the dynamic coupling model and optimize the weight allocation for coupling strength calculation.

Images