Digital factory monitoring data processing method

Abstract

The invention provides a digital factory monitoring data processing method, which relates to the technical field of industrial process monitoring, and comprises the following steps of: acquiring monitoring data of factory monitoring points, establishing a topological association graph, calculating data transmission characteristics among the monitoring points, establishing a three-layer data transmission matrix, and calculating abnormal propagation credibility; the method is used for generating a multi-level anomaly propagation topological graph and finally outputting comprehensive evaluation analysis. According to the method, the comprehensiveness and accuracy of abnormal propagation feature extraction are remarkably improved, the reliability of propagation path recognition is enhanced, potential risks are found in time through an early warning mechanism, more effective technical support is provided for abnormal propagation analysis of a complex industrial system, and the method has important practical value.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial process monitoring, and more specifically, to a digital factory monitoring data processing method. Background Art

[0002] Anomaly propagation analysis in industrial production processes is a key link to ensure production stability and product quality. Traditional anomaly analysis methods mainly rely on single-point monitoring and static threshold judgment, which makes it difficult to effectively characterize the dynamic propagation characteristics of anomalies in complex process systems. With the improvement of industrial automation and informatization, a large amount of equipment operation parameters, process parameters and quality inspection data have been accumulated in the production process. Anomaly propagation analysis based on these multi-source heterogeneous data has become a research hotspot. Existing studies mainly use correlation analysis, causal inference and graph theory to construct anomaly propagation models, and track the diffusion path of anomalies by extracting the correlation between parameters. However, these methods often regard anomaly propagation as a static process, ignoring the dynamic evolution characteristics of parameter correlation in industrial production, resulting in insufficient accuracy and timeliness in identifying anomaly propagation paths.

[0003] The current anomaly propagation analysis technology has the following main problems: first, there is a lack of systematic characterization methods for multi-scale data transmission characteristics, which makes it difficult to accurately capture the anomaly propagation laws at different time scales such as rapid response, steady-state process and long-term evolution; second, most of the existing propagation feature extraction methods are based on single-dimensional statistical indicators, and fail to fully integrate multi-dimensional feature information such as time lag, amplitude correlation and fluctuation consistency, which affects the reliability of propagation path identification; third, the traditional expression of anomaly propagation topological structure is too simple, which cannot effectively distinguish propagation paths with different credibility levels, and lacks early warning mechanisms for dynamic conversion and intensity mutation of propagation paths. These technical problems seriously restrict the actual application effect of anomaly propagation analysis in complex industrial systems. Summary of the invention

[0004] In order to solve the above technical problems, the present invention is proposed. The present invention provides a digital factory monitoring data processing method, which can solve the problems of incomplete feature extraction, unreliable propagation path identification and insufficient early warning capability in industrial abnormal propagation analysis to a certain extent.

[0005] According to one aspect of the present invention, a method for processing digital factory monitoring data is provided, which comprises: Obtain monitoring data from factory monitoring points, establish a topological association diagram containing previous, current and subsequent association points based on the process flow, and evaluate and eliminate abnormal data; Based on the topological association graph, an adaptive time window method is used to calculate data transmission characteristics; A three-layer data transfer matrix is ​​established based on the transfer feature, and when an abnormal monitoring point is detected, the abnormal propagation credibility is calculated based on the data transfer matrix; A multi-level anomaly propagation topology diagram is generated based on the anomaly propagation credibility, and a comprehensive evaluation analysis is output in combination with the data transmission characteristics.

[0006] Furthermore, the monitoring data includes: equipment operation parameters, process parameters and quality inspection data, and a monitoring point topology association diagram including preceding, current and subsequent association points is established based on the process flow.

[0007] Furthermore, the data transmission characteristics include: time lag, amplitude correlation and fluctuation consistency; The time lag is obtained by calculating the time difference between the peak values ​​of the data at adjacent monitoring points; The amplitude correlation is obtained by calculating the ratio of the data changes of adjacent monitoring points; The fluctuation consistency is obtained by calculating the overlap of the fluctuation directions of the data of adjacent monitoring points.

[0008] Furthermore, data analysis is performed based on the adaptive time window method, the parameter change cycle is identified through fast Fourier transform and wavelet analysis, the local steady-state characteristics and dynamic change characteristics of the data are captured in combination with the piecewise autocorrelation analysis and sliding entropy method, and the analysis window length is dynamically adjusted.

[0009] Furthermore, the determination of the window length adopts a multi-constraint adaptive mechanism. By setting a hard constraint range based on the process response time, the window length and sliding step are dynamically adjusted for different situations such as periodic fluctuations, sudden abnormalities and operating condition conversions, and a differentiated parameter adjustment strategy is established.

[0010] Furthermore, the data transmission characteristics between adjacent monitoring points are calculated based on the determined time window, including: Calculate time lag based on peak identification; Analyze the amplitude correlation by changing the ratio; Assessing fluctuation consistency based on directional coincidence; Meanwhile, the nonlinear feature processing and the coupling effect of multiple transfer paths are considered, and the piecewise linearization and path decomposition methods are used to improve the calculation accuracy.

[0011] Furthermore, based on the data transfer characteristics, the sliding window method is used to divide the original data into a fast response layer, a steady-state process layer and a long-term evolution layer according to the sampling period, and a three-layer data transfer matrix is ​​constructed, including: An asymmetric time lag matrix is ​​constructed through cross-correlation analysis of high-frequency data; A symmetric amplitude correlation matrix is ​​constructed using the dynamic correlation coefficients of the aligned data; A symmetric volatility consistency matrix is ​​constructed through a dynamic time warping algorithm for long-term data.

[0012] Furthermore, a dynamic linkage mechanism is established based on the three-layer data transmission matrix. The time lag change of the rapid response layer triggers the adjustment of the calculation parameters of the steady-state process layer. The change of the correlation intensity of the steady-state process layer triggers the re-evaluation of the long-term evolution layer. The new collaborative mode of the long-term evolution layer feedback optimizes the feature extraction of the first two layers. The credibility of the abnormal propagation is calculated by comprehensively evaluating the time dimension, intensity dimension and collaborative dimension.

[0013] Furthermore, a three-layer propagation topology diagram is constructed based on the abnormal propagation credibility, and the propagation intensity and timing characteristics are expressed through graphic features to generate a comprehensive report including abnormal source location, impact range analysis and propagation feature evaluation.

[0014] Furthermore, based on the comprehensive report, an early warning mechanism is established for abnormal propagation characteristics, including: Monitoring time-lag fluctuations to warn of transmission path conversion risks; Track the intensity attenuation changes to warn of mutation spread trends; Analyze the possible problems of coupling strength enhancement warning path merging; Assess key node anomalies and warn of the formation of new transmission hubs.

[0015] Compared with the prior art, the present invention realizes the multi-dimensional characterization of abnormal propagation characteristics by constructing a three-layer data transfer matrix, which significantly improves the comprehensiveness and accuracy of abnormal propagation analysis. At the same time, a multi-level propagation topology structure based on credibility is proposed, which intuitively displays the importance and reliability of the propagation path through different graphic features, and establishes an early warning mechanism for abnormal propagation characteristics, which can timely discover potential risks such as propagation path conversion, intensity mutation, and path merging. And through the dynamic linkage mechanism between levels, the adaptive adjustment of feature updates is realized, so that the abnormal propagation analysis can better adapt to the dynamic characteristics of complex industrial systems. It significantly improves the practicality and reliability of industrial abnormal propagation analysis, and provides more effective technical support for industrial process monitoring and fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order 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 prior art descriptions. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings: Figure 1 is a flow chart of a digital factory monitoring data processing method according to an embodiment of the present invention; Figure 24 is a flow chart of calculating the anomaly propagation credibility according to an embodiment of the present invention. DETAILED DESCRIPTION

[0017] Below, the exemplary embodiments according to the present invention will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments of the present invention, and it should be understood that the present invention is not limited to the exemplary embodiments described here.

[0018] As mentioned in the above background technology, there are three main outstanding problems in the existing industrial abnormal propagation analysis technology: First, there is a lack of systematic characterization methods for multi-scale data transmission characteristics, which makes it difficult to accurately capture the abnormal propagation laws on different time scales such as rapid response, steady-state process and long-term evolution; second, the existing propagation feature extraction methods are mostly based on single-dimensional statistical indicators, and fail to fully integrate multi-dimensional feature information such as time lag, amplitude correlation and fluctuation consistency, which affects the reliability of propagation path identification; third, the traditional abnormal propagation topological structure expression method is too simple to effectively distinguish propagation paths with different credibility levels, and lacks an early warning mechanism for dynamic conversion and intensity mutation of propagation paths. Our invention targets these technical pain points and proposes a digital factory monitoring data processing method. By constructing a three-layer data transfer matrix to characterize the multi-scale transmission relationship between parameters, integrating the time dimension, intensity dimension and collaborative dimension features to calculate the abnormal propagation credibility, and establishing a multi-level propagation topological graph structure and early warning mechanism, it effectively solves the problems of incomplete feature extraction, unreliable propagation path identification and weak early warning capability in industrial abnormal propagation analysis, and belongs to the field of industrial process monitoring technology.

[0019] Figure 1 FIG. 1 is a flow chart of a method for processing digital factory monitoring data according to an embodiment of the present invention. Figure 1 As shown, in the digital factory monitoring data processing method, it includes: S1: Obtain monitoring data from factory monitoring points, establish a topological association diagram containing preceding, current, and subsequent association points based on the process flow, and evaluate and eliminate abnormal data.

[0020] Acquire data from monitoring points distributed in various production links. The collected real-time monitoring data covers three dimensions: equipment operation parameters, process parameters, and quality inspection data.

[0021] Among them, the equipment operating parameters reflect the real-time operating status of the equipment, including the operating status data of the speed parameters, temperature parameters, pressure parameters, and vibration parameters of the production equipment, which can directly reflect the health status of the equipment; The process parameters reflect the operation status of the production process, including the process control data of flow parameters, concentration parameters, liquid level parameters, and output parameters in the process, which are directly related to the production process control of the product; The quality inspection data reflects the quality characteristics of the product, including the product's size parameters, weight parameters, density parameters, and composition parameters, which determine whether the final product meets the quality requirements.

[0022] On the basis of obtaining the above monitoring data, grasp the correlation between each monitoring point. By analyzing the factory's production process, sorting out the material flow path and process connection sequence, establish a topological correlation diagram of the monitoring points.

[0023] In this topological association diagram, the association relationship between monitoring points is described at three levels: the previous association point, the current monitoring point, and the subsequent association point. The previous association point is the related monitoring point located upstream of the monitoring point to be analyzed in the process flow, and data changes often transmit and affect the status of the current monitoring point; the current monitoring point is the target point for data analysis and is the core object of abnormal analysis; the subsequent association point refers to the related monitoring point located downstream of the monitoring point to be analyzed in the process flow, and the status may be affected by data changes at the current monitoring point.

[0024] In order to ensure the accuracy of abnormal propagation analysis, it is necessary to conduct a comprehensive quality assessment of the collected real-time monitoring data. First, pay attention to the integrity of the data, check whether there are missing values ​​or abnormal discontinuities in the data sequence, and ensure the continuity of the data; secondly, evaluate the consistency of the data, check whether the data at different monitoring points are unified in terms of dimensional units, and avoid analysis deviations caused by inconsistent units; at the same time, verify the timeliness of the data, ensure that the data collection time meets the requirements of real-time analysis, and ensure the timeliness of the analysis results; and verify the validity of the data, determine whether the data value is within a reasonable range, and eliminate abnormal values ​​that are obviously beyond the normal range.

[0025] The results of the above data quality assessment are quantitatively judged by the preset quality threshold. When the data quality assessment result of a monitoring point is lower than the preset quality threshold, it indicates that the data point may have problems such as collection anomaly, sensor failure or communication interruption, and is removed from the data set.

[0026] Furthermore, a topological association diagram of monitoring points is established, including: Obtain a complete process flow chart and process pipeline instrumentation diagram, and extract the material transfer network structure therein. Based on the principle of material balance, identify the material transfer relationship between each process unit, including: material input node, material conversion node and material output node. For each material transfer path, determine the layout of its monitoring points, mainly including: process parameter measurement points such as flow, pressure, liquid level, and process parameter measurement points such as temperature and composition. In particular, for material diversion nodes and material confluence nodes, establish a multi-path transfer structure in the topological association diagram, use directed edges to represent the material flow direction, and mark the flow distribution ratio. For the material circulation loop, a closed-loop structure is used to represent it, and characteristic parameters such as circulation ratio and material change cycle are marked.

[0027] Furthermore, for direct mechanical coupling systems, identify the corresponding relationship between the driving device and the driven device, and use solid lines to connect them in the topological association diagram, and annotate parameters such as transmission ratio and efficiency. For indirect energy coupling systems, analyze the energy transfer link between the heat source device, heat transfer medium and heat receiving device, and use dotted lines to connect them in the topological association diagram, and annotate characteristics such as heat transfer coefficient and delay time. For process control systems, extract the components of the control loop, including: measuring elements, controllers, actuators and controlled objects, use special symbols to represent the control loop structure in the topological association diagram, and annotate information such as control methods and control parameters.

[0028] Furthermore, based on the process mechanism and process dynamics principles, the causal relationship between key process parameters is identified, for example, changes in reactor temperature and pressure caused by changes in feed flow, and chain changes in subsequent product quality parameters. By analyzing historical data, the characteristic quantities of parameter transfer are extracted: including transfer time, transfer intensity and transfer direction. In the topological association diagram, connecting lines of different thicknesses are used to represent the transfer intensity, arrows are used to mark the transfer direction, and characteristic parameters such as transfer time are marked on the connecting lines.

[0029] Among them, the calculation of the transfer time adopts the cross-correlation analysis method to pre-process the time series data of the two monitoring points, including removing outliers, data standardization and eliminating trend items. When the sliding time window is used to calculate the cross-correlation function, the window length is set according to the process characteristics and is 3-5 times the parameter change cycle. During the calculation process, the cross-correlation function is obtained by evaluating the correlation between the two parameter sequences under different time delays. The corresponding time delay when the cross-correlation function reaches the maximum value is the parameter transfer time. If multiple local maxima are detected, a reasonable transfer time must be selected in combination with the process characteristics. When the parameter change has periodic characteristics, the transfer time should be determined within a complete cycle.

[0030] The calculation of transfer strength adopts the partial correlation analysis method. When calculating the transfer strength between two parameters, first identify and eliminate other parameter sets that may have an impact, and calculate the partial correlation coefficient between the target parameter pairs. This coefficient reflects the true degree of correlation between the two parameters after excluding the influence of other factors. When there are multiple influencing parameters, a recursive method is used to gradually calculate the higher-order partial correlation coefficients. If the absolute value of the partial correlation coefficient is greater than the preset threshold, it is considered that there is a significant transfer relationship between the parameters. When the working conditions change significantly, the transfer strength needs to be recalculated to adapt to the new working conditions.

[0031] The transfer direction is determined by the Granger causality test method, and two prediction models are constructed with and without the parameter to be tested. The first model uses the historical data of the two parameters for prediction, and the second model only uses the historical data of the predicted parameter. By comparing the prediction results of the two models, if the prediction effect of the model containing the parameter to be tested is significantly better than that of the model containing only a single parameter, it can be determined that the parameter to be tested has a significant causal effect on the target parameter. When performing a two-way causal test, it is necessary to test the influence of the two parameters on each other at the same time. If a two-way causal relationship is detected, the main transfer direction should be determined in combination with the transfer intensity.

[0032] The topological association diagram finally established is a multi-level association network, in which each monitoring point is marked with its position, functional attributes and association characteristics in the process flow. The connection relationship between the monitoring points not only reflects the material flow path, but also includes the energy transfer channel and information transfer mechanism.

[0033] S2: Calculate the data transmission characteristics between monitoring points based on the topological association graph, and use the adaptive time window method to calculate the transmission characteristics.

[0034] Based on the real-time data of the monitoring points in the topological association diagram, the data transmission characteristics between adjacent monitoring points are calculated, including time lag, amplitude correlation and fluctuation consistency. The time lag is obtained by calculating the time difference between the peak values ​​of the data of adjacent monitoring points, the amplitude correlation is obtained by calculating the ratio of the changes in the data of adjacent monitoring points, and the fluctuation consistency is obtained by calculating the overlap of the fluctuation directions of the data of adjacent monitoring points.

[0035] In order to adapt to the dynamic characteristics of parameter changes in industrial production processes, the adaptive time window method is used for data analysis. When determining the length of the adaptive time window, the main periodic components of parameter changes are identified by performing spectrum analysis and autocorrelation analysis on historical data. When the process parameters show obvious periodic fluctuations, the length of the time window at least includes the complete change cycle; if the parameter changes lack regularity, the statistical characteristics of the parameters are calculated to determine the appropriate window length.

[0036] Furthermore, when performing spectrum analysis on historical data, a hybrid method of fast Fourier transform combined with wavelet analysis is used. While performing fast Fourier transform to identify the main frequency components, wavelet transform is used to capture the mutation characteristics and local characteristics in the data. When the data contains multiple frequency components, not only spectrum peak detection is performed to determine the main period and secondary period, but also the energy proportion of each frequency component needs to be evaluated to screen out the frequency components that have actual physical significance for parameter changes. When the parameters show non-stationary characteristics, the wavelet packet analysis method is used to achieve a fine characterization of the parameters in the time-frequency domain, especially for the high-frequency transient characteristics caused by sudden anomalies.

[0037] When analyzing the autocorrelation characteristics, the method of segmented autocorrelation analysis combined with sliding entropy is used. When calculating the autocorrelation function of the data sequence, the local steady-state characteristics of the sequence are identified through segmented processing. For each segment, its autocorrelation function and sample entropy are calculated. The sudden change of sample entropy usually indicates a change in the dynamic characteristics of the system. When the autocorrelation coefficient shows obvious periodic fluctuations, the period can be used as a characteristic period of parameter changes, but the stability of the period needs to be verified by conditional autocorrelation analysis. When the autocorrelation coefficient decays below the preset threshold, the corresponding time delay can be used to determine the correlation length of the data. If multiple significant autocorrelation peaks are found, they need to be screened and confirmed in combination with typical periodic characteristics.

[0038] The time window length is determined using a multi-constraint adaptive mechanism, including: The hard constraint range of the window length is set, with the lower limit not less than 3 times the fastest process response time to ensure that the complete dynamic response process can be captured; the upper limit is not more than 1.5 times the longest process cycle to avoid introducing irrelevant long-term fluctuations. Within the hard constraint range, the dynamic adjustment of the window length follows the following rules: For periodic fluctuations under normal operating conditions, the window length is an integer multiple of the period; When a sudden anomaly is detected, the dual time window mechanism is activated: the original long time window is maintained for steady-state feature analysis, while the short time window is opened for transient feature extraction. The analysis results of the two time windows are weighted fused; During the working condition conversion process, the window length is dynamically adjusted with the parameter change rate: the faster the change rate, the shorter the window length, but it must meet the minimum length constraint; when the change tends to be stable, the window length gradually returns to the standard length; Considering the computing resource constraints, the adaptive relationship between the window sliding step size and the window length is set: when the window is short, the step size can be 1 / 10 of the window length; when the window is long, the step size can be appropriately increased, but not more than 1 / 5 of the window length.

[0039] For different types of process parameters, differentiated window length adjustment strategies are established.

[0040] In a certain time window, the time lag between adjacent monitoring points is calculated. The time lag reflects the time required for the process parameters to be transferred from the upstream monitoring point to the downstream monitoring point. By identifying the peak points in the monitoring data, the time difference between the peaks of the adjacent monitoring points is calculated to obtain the lag time of parameter transfer. To improve the calculation accuracy, not only the time difference of the main peak value is calculated, but also the time difference of the secondary peak value is calculated, and a more reliable time lag is obtained through weighted average. In the calculation process, the influence of false peaks is eliminated through filtering.

[0041] The calculation of amplitude correlation mainly observes the proportional relationship between the data changes of adjacent monitoring points. By calculating the change of monitoring data within the time window and obtaining the ratio of the change of adjacent monitoring points, the attenuation or amplification characteristics in the parameter transfer process are reflected. To ensure the stability of the calculation results, the data changes at multiple time points are considered, and the calculated ratios are statistically analyzed. When there are nonlinear characteristics, the piecewise linearization method is used to process the relationship between the data changes to reflect the amplitude correlation characteristics under different working conditions.

[0042] Fluctuation consistency reflects the degree of coordination of the change direction of data from adjacent monitoring points. By comparing the change direction of data from adjacent monitoring points in the same time window and calculating the proportion of time when the change direction overlaps, the fluctuation consistency index is obtained. When calculating specifically, it is necessary to determine the direction of data change, including: rising, falling, or flat, and count the time periods when the change direction of adjacent monitoring points is consistent, and calculate the proportion of the consistent time period to the total analysis time. In order to avoid interference from small fluctuations, a corresponding change threshold is set, and only changes that exceed the threshold are considered valid fluctuations.

[0043] In the actual calculation process, it is necessary to make appropriate adjustments based on the specific process characteristics. For example, when the process exhibits strong nonlinear characteristics, the piecewise linearization method is used for data preprocessing, and the parameter variation range is divided into multiple intervals based on the process mechanism. When the parameter values ​​are in different intervals, the corresponding linear transformation function is used for data conversion. If the parameter changes span multiple intervals, smoothing is required at the interval boundaries to ensure the continuity of the converted data. When the process has saturation characteristics, a compression nonlinear function should be used for data conversion near the saturation region. If an exponential or power function relationship is found between the parameters, the nonlinear relationship is converted into a linear relationship through logarithmic transformation.

[0044] When there are multiple transfer paths, the path decomposition and path integration methods are used to calculate the eigenvalues. Based on the process flow, all possible transfer paths are identified, including direct transfer paths and indirect transfer paths. When calculating the eigenvalues ​​of the direct transfer path, the influence of other paths needs to be temporarily ignored. If multiple transfer paths exist in parallel, the eigenvalues ​​on each path are calculated separately, and the weighting coefficients are set according to the importance of the paths. When paths cross, the mutual influence between the paths needs to be considered, and a path coupling correction mechanism needs to be established.

[0045] S3: A three-layer data transfer matrix is ​​established based on the data transfer characteristics, and when an abnormal monitoring point is detected, the abnormal propagation credibility is calculated based on the data transfer matrix.

[0046] Based on the calculated time lag, amplitude correlation and fluctuation consistency data, the sliding window method is used to divide the original data into a rapid response layer, a steady-state process layer and a long-term evolution layer according to the sampling period. When dividing the data, ensure that there are overlapping areas between the data windows of each level to achieve data connection between levels. If data is missing at a certain level, interpolation is performed based on the data characteristics of the adjacent levels.

[0047] In the fast response layer, an n×n time lag matrix T is constructed, where n is the number of monitoring points. For each element Tij in the matrix, the time lag characteristics from monitoring point i to monitoring point j are determined by cross-correlation analysis of high-frequency data. Specifically, the data sequence of monitoring point i is used as the input signal, and the data sequence of monitoring point j is used as the output signal. The time delay τ that maximizes the function is found by calculating the cross-correlation function Rij(τ) = E[xi(t)·xj(t+τ)], which is the value of the matrix element Tij. The time lag matrix is ​​not symmetric because the parameter transfer is directional and is continuously updated through a sliding time window, which can dynamically reflect changes in the response relationship between parameters.

[0048] Based on the time lag characteristics of the fast response layer, the steady-state process layer constructs the amplitude correlation matrix R. During the construction process, the hourly data is first time-aligned according to the corresponding element values ​​in the time lag matrix T to ensure that the truly corresponding data segments are compared. By calculating the dynamic correlation coefficient of the aligned data sequence, the element value Rij in the amplitude correlation matrix R is obtained, which reflects the correlation strength between parameters i and j. Unlike the time lag matrix, the amplitude correlation matrix R is symmetrical because the correlation strength between parameters is mutual. A sliding time window is used to continuously calculate the correlation coefficient, and the calculation window is dynamically adjusted according to the time lag changes detected by the fast response layer.

[0049] Based on the time lag and amplitude correlation features, the long-term evolution layer constructs the fluctuation consistency matrix W. It is also n×n dimensional, and its elements Wij characterize the long-term synergistic relationship between parameters i and j. During the construction process, the daily or weekly data are aligned and preprocessed in combination with the features of the first two layers, and then the long-term similarity of the parameter sequence is evaluated by the dynamic time warping algorithm. The fluctuation consistency matrix W is symmetrical, and its element values ​​reflect the degree of coordination of the fluctuation patterns between parameters. The feature calculation at this level continuously tracks the long-term change trend of the parameters. When the synergy pattern changes, it will feedback and adjust the feature calculation parameters of the first two layers to form a complete feature update mechanism.

[0050] Furthermore, in the dynamic linkage mechanism of the three-layer matrix, when the fast response layer detects a change in the time lag feature, it is manifested as a significant change in the value of an element Tij in the time lag matrix T. First, the reliability of this change is evaluated, and the ratio of the change amplitude to the historical fluctuation range is calculated to determine whether it is necessary to trigger the adjustment of the steady-state process layer. If it is confirmed that adjustment is required, the steady-state process layer will change the calculation method of the amplitude correlation accordingly, including adjusting the time offset of the data alignment to match the new time lag feature; at the same time, the calculation window length of the correlation coefficient is changed. When the time lag becomes shorter, the calculation window is reduced to improve sensitivity, and when the time lag becomes longer, the window is appropriately expanded to enhance stability.

[0051] When the steady-state process layer finds a significant change in the correlation strength, that is, the element values ​​in the amplitude correlation matrix R show a continuous deviation, the potential impact of this change on the long-term fluctuation characteristics is evaluated. First, the duration and trend of the change in correlation strength are analyzed. If the duration of the change exceeds the preset threshold or shows an obvious unidirectional evolution trend, the re-evaluation mechanism of the long-term evolution layer is triggered. The long-term evolution layer will adjust the calculation parameters of the fluctuation consistency, including expanding the time span of trend analysis, increasing the weight coefficient for emerging patterns, and updating the baseline value of synergy evaluation. At the same time, re-examine the synergistic features in the historical data to evaluate whether the new correlation pattern represents a fundamental change in the behavior of the system.

[0052] If the long-term evolution layer identifies a new collaborative pattern, it will be manifested as a new stable feature in the fluctuation consistency matrix W, and this change will be fed back to the feature extraction process of the first two layers. For the fast response layer, the system will optimize the calculation parameters of time lag, including adjusting the time window size of the cross-correlation analysis, updating the dynamic threshold of the lag judgment, and changing the signal preprocessing method to better adapt to the new collaborative pattern. For the steady-state process layer, the criteria for amplitude association are adjusted, including updating the calculation method of the correlation coefficient, changing the criteria for the significance test, and optimizing the data alignment strategy.

[0053] Based on the three-layer data transfer matrix, the credibility of anomaly propagation is calculated by comprehensively evaluating the time lag characteristics, amplitude correlation characteristics and fluctuation consistency characteristics, including: The detected abnormal signal is feature extracted to identify the initial position and time point of the abnormality. Based on the time lag matrix of the fast response layer, the theoretical propagation time of the abnormal signal from the initial position to each relevant monitoring point is calculated, and these times are used as the benchmark features of abnormal propagation. When the actual observed abnormal propagation time is consistent with the theoretical time, it is considered that the propagation path has a high time dimension credibility.

[0054] On the basis of time feature verification, the amplitude correlation matrix of the steady-state process layer is further used to evaluate the intensity characteristics of anomaly propagation. By analyzing the change in correlation strength between parameters before and after the anomaly occurs, the amplitude attenuation law of the anomaly signal during the propagation process is calculated. If the observed amplitude change trend is consistent with the historical correlation pattern and the change amplitude is within a reasonable range, the credibility of the intensity dimension of the propagation path is improved.

[0055] The matching degree between abnormal propagation and long-term synergy characteristics is further evaluated by combining the fluctuation consistency matrix of the long-term evolution layer. By comparing the current abnormal propagation mode with the historically accumulated synergy mode, it is determined whether this propagation conforms to the long-term coupling law between parameters. If it is found that the abnormal propagation path is highly consistent with the known synergy relationship and the fluctuation characteristics during the propagation process remain consistent, the synergy credibility of the propagation path is further confirmed.

[0056] Among them, the time dimension refers to the time series characteristics of the propagation of anomalies in the system, which is reflected in the response delay and transmission order between parameters. It is described by the time lag matrix, reflecting the causal relationship and dynamic response characteristics between different monitoring points; The intensity dimension is used to describe the amplitude variation characteristics of the anomaly during the propagation process. It is represented by the amplitude correlation matrix, which reflects the coupling strength and signal attenuation law between parameters, including the amplitude attenuation, amplification or conversion characteristics of the abnormal signal during the propagation process. The synergy dimension is used to characterize the long-term synergistic relationship between parameters and the consistency of fluctuation patterns. It is described by the fluctuation consistency matrix, which reflects the stable characteristics and coupling laws between parameters on a longer time scale, including the synchronization, periodicity and trend consistency of parameter changes.

[0057] The credibility of the time dimension, intensity dimension and coordination dimension is combined into an abnormal propagation credibility index by using a weighted fusion method. The weight distribution takes into account the reliability and importance of the features of different dimensions, and is dynamically adjusted according to the actual working conditions.

[0058] As the anomaly spreads, the credibility evaluation results are adjusted in real time. When a significant change is found in the credibility of a certain propagation path, the credibility distribution of the entire propagation network is re-evaluated.

[0059] S4: Generate a multi-level anomaly propagation topology map based on the anomaly propagation credibility, combine the transmission characteristics, and output a comprehensive evaluation analysis.

[0060] Based on the calculated anomaly propagation credibility, a multi-level anomaly propagation topology structure is constructed. Specifically: a grading threshold for anomaly propagation credibility is set, and the propagation path with the highest credibility is classified into the main propagation path layer, the path with the second highest credibility is classified into the secondary propagation path layer, and the path with relatively low credibility but still has reference value constitutes the potential propagation path layer.

[0061] In the main propagation path layer, the propagation links with the highest credibility are marked. By analyzing the time lag matrix of the rapid response layer, the propagation timing of the anomaly on the main propagation path is calculated, including the propagation delay time of the anomaly from the source to each affected node. At the same time, based on the amplitude correlation matrix of the steady-state process layer, the intensity change of the abnormal signal during the propagation process is quantified, and the attenuation or amplification effect of each propagation link is calculated. Combined with the fluctuation consistency matrix of the long-term evolution layer, the stability of the main propagation path is evaluated to determine whether these propagation paths are consistent with the long-term operation characteristics of the system.

[0062] Similarly, in the secondary transmission path layer, identify transmission paths that are slightly less credible but still important. These paths may have certain uncertainties or be interfered with by other factors, but they still need to be focused on. Analyze the timing characteristics, intensity characteristics, and stability characteristics on these paths, but use different expressions from the main paths when visualizing and assessing risks, so as to quickly identify different levels of transmission risks.

[0063] For the potential transmission path layer, focus on transmission paths that are currently less credible but may be activated under certain conditions. They usually show weaker transmission characteristics or are not highly matched with historical data, so they are included in the monitoring scope. Continue to track the evolution trend of these potential paths and discover new transmission patterns in a timely manner.

[0064] Specifically, for the main propagation path layer, the line width of the propagation path is proportional to the propagation strength. The stronger the propagation relationship, the thicker the line. At the same time, solid arrows are used to mark the propagation direction at the node connection. The size of the arrow reflects the time lag of the propagation. Shorter time lags are represented by larger arrows.

[0065] The secondary propagation path layer uses dotted lines to represent propagation links. The density of dotted lines is used to represent the credibility level of propagation. The higher the credibility, the denser the dotted lines. The connection between nodes uses hollow arrows. The transparency of the arrows varies with the strength of propagation. The stronger the propagation relationship, the lower the transparency of the arrow.

[0066] For the potential propagation path layer, dotted lines are used to represent propagation links. The spacing between dotted lines reflects the uncertainty of propagation. The larger the spacing, the higher the uncertainty. Nodes are connected by dotted arrows. The length of the arrow is proportional to the degree of time lag. Longer time lags are represented by longer arrows. In addition, the activity of the propagation path is represented by the flashing frequency of the arrow. The higher the frequency, the greater the possibility that the path will be transformed into a higher level.

[0067] After constructing a complete multi-level topology map, a comprehensive analysis report is generated. By analyzing the monitoring point where the abnormal characteristics first appear, the source location of the abnormality is pointed out in combination with the time lag characteristics and the credibility of the propagation path. The propagation of the abnormality at different levels is described in detail, including the number of monitoring points involved in each level, the scope of the process units covered, and the correlation between the levels.

[0068] The report focuses on analyzing the impact range of the anomaly, including not only the areas that have been affected, but also predicting the areas that may be affected. For the analysis of the propagation sequence, a detailed timeline is provided, marking the occurrence time and expected arrival time of the anomaly at each key node.

[0069] In terms of propagation intensity, the intensity variation law of abnormal signals on various propagation paths is quantitatively described, and the impact of cumulative effect and coupling effect is evaluated.

[0070] It also further marks the weak links and potential risk points that need special attention, and provides targeted suggestions for subsequent monitoring and control.

[0071] Provide special warning information for areas where the transmission path may change or the intensity may suddenly change, including: When a fluctuation in the time lag characteristics of a certain propagation node is detected, for example, the time lag that was originally stable at 30 seconds is suddenly shortened to 15 seconds, and at the same time, the node begins to show enhanced amplitude correlation characteristics with the nodes on other potential propagation paths, an early warning of a possible switch in the propagation path is issued, indicating a downward trend in the stability of the current main propagation path, and possible activation of new propagation paths, and an estimated time window for the switch to occur is given.

[0072] If a change is observed in the intensity attenuation law of a certain transmission link, for example, the originally stable linear attenuation suddenly turns into exponential attenuation, and this change is accompanied by an increase in the consistency of fluctuations of adjacent nodes, an intensity mutation warning is generated, including a quantitative analysis of the intensity change, the boundary conditions for the mutation, and an assessment of the possible impact range. When the intensity mutation may lead to a significant increase in the rate of propagation and diffusion, the warning level will be raised to the highest level.

[0073] When it is found that the coupling strength of multiple propagation paths is enhanced at the same time, the focus is on analyzing the possible merging of propagation paths. For example, two originally relatively independent secondary propagation paths may merge into one main propagation path due to changes in certain process conditions. This warns of the risk of reorganization of the topological structure and evaluates the possible propagation intensity amplification effect after the merger.

[0074] If the fluctuation consistency index of a certain propagation node is abnormal, and the node has a key position on multiple potential propagation paths, it will be warned that the node may become a new propagation hub, including the evolution trend of node characteristics, possible activated propagation path combinations, and the analysis of the cascade effect that may be caused by this.

[0075] In summary, the digital factory monitoring data processing system based on the embodiment of the present invention is explained, which realizes the multi-dimensional characterization of abnormal propagation characteristics by constructing a three-layer data transfer matrix, significantly improving the comprehensiveness and accuracy of abnormal propagation analysis. At the same time, a multi-level propagation topology structure based on credibility is proposed, which intuitively displays the importance and reliability of the propagation path through different graphic features, and establishes an early warning mechanism for abnormal propagation characteristics, which can timely discover potential risks such as propagation path conversion, intensity mutation, and path merging. And through the dynamic linkage mechanism between levels, the adaptive adjustment of feature updates is realized, so that the abnormal propagation analysis can better adapt to the dynamic characteristics of complex industrial systems. The practicality and reliability of industrial abnormal propagation analysis are significantly improved, providing more effective technical support for industrial process monitoring and fault diagnosis.

Claims

1. A digital factory monitoring data processing method, characterized in that: include: Obtain monitoring data from factory monitoring points, establish a topological association diagram containing previous, current and subsequent association points based on the process flow, and evaluate and eliminate abnormal data; Based on the topological association graph, an adaptive time window method is used to calculate data transmission characteristics; A three-layer data transfer matrix is ​​established based on the transfer feature, and when an abnormal monitoring point is detected, the abnormal propagation credibility is calculated based on the data transfer matrix; A multi-level anomaly propagation topology diagram is generated based on the anomaly propagation credibility, and a comprehensive evaluation analysis is output in combination with the data transmission characteristics.

2. The digital factory monitoring data processing method according to claim 1 is characterized in that: The monitoring data includes: equipment operation parameters, process parameters and quality inspection data, and a monitoring point topology association diagram including preceding, current and subsequent association points is established based on the process flow.

3. The digital factory monitoring data processing method according to claim 2 is characterized in that: The data transmission characteristics include: time lag, amplitude correlation and fluctuation consistency; The time lag is obtained by calculating the time difference between the peak values ​​of the data at adjacent monitoring points; The amplitude correlation is obtained by calculating the ratio of the data changes of adjacent monitoring points; The fluctuation consistency is obtained by calculating the overlap of the fluctuation directions of the data of adjacent monitoring points.

4. The digital factory monitoring data processing method according to claim 3 is characterized in that: Data analysis is performed based on the adaptive time window method, parameter change cycles are identified through fast Fourier transform and wavelet analysis, and the local steady-state characteristics and dynamic change characteristics of the data are captured in combination with the piecewise autocorrelation analysis and sliding entropy method, and the analysis window length is dynamically adjusted.

5. The digital factory monitoring data processing method according to claim 2 is characterized in that: The determination of the window length adopts a multi-constraint adaptive mechanism. By setting a hard constraint range based on the process response time, the window length and sliding step are dynamically adjusted for different situations such as periodic fluctuations, sudden abnormalities and operating condition conversions, and a differentiated parameter adjustment strategy is established.

6. The digital factory monitoring data processing method according to claim 5 is characterized in that: Calculate the data transmission characteristics between adjacent monitoring points based on the determined time window, including: Calculate time lag based on peak identification; Analyze the amplitude correlation by changing the ratio; Assessing fluctuation consistency based on directional coincidence; Meanwhile, the nonlinear feature processing and the coupling effect of multiple transfer paths are considered, and the piecewise linearization and path decomposition methods are used to improve the calculation accuracy.

7. The digital factory monitoring data processing method according to claim 6 is characterized in that: Based on the data transfer characteristics, the sliding window method is used to divide the original data into a rapid response layer, a steady-state process layer and a long-term evolution layer according to the sampling period, and a three-layer data transfer matrix is ​​constructed, including: An asymmetric time lag matrix is ​​constructed through cross-correlation analysis of high-frequency data; A symmetric amplitude correlation matrix is ​​constructed using the dynamic correlation coefficients of the aligned data; A symmetric volatility consistency matrix is ​​constructed through a dynamic time warping algorithm for long-term data.

8. The digital factory monitoring data processing method according to claim 7, characterized in that: A dynamic linkage mechanism is established based on the three-layer data transmission matrix. The time lag change of the rapid response layer triggers the adjustment of the calculation parameters of the steady-state process layer. The change of the correlation intensity of the steady-state process layer triggers the re-evaluation of the long-term evolution layer. The new collaborative mode of the long-term evolution layer feedback optimizes the feature extraction of the first two layers; and the credibility of anomaly propagation calculated in the time dimension, intensity dimension and collaborative dimension is comprehensively evaluated.

9. The digital factory monitoring data processing method according to claim 8, characterized in that: A three-layer propagation topology diagram is constructed based on the abnormal propagation credibility, the propagation intensity and timing characteristics are expressed through graphic features, and a comprehensive report including abnormal source location, impact range analysis and propagation feature evaluation is generated.

10. The digital factory monitoring data processing method according to claim 9, characterized in that: Based on the comprehensive report, an early warning mechanism is established for abnormal propagation characteristics, including: Monitoring time-lag fluctuations to warn of transmission path conversion risks; Track the intensity attenuation changes to warn of mutation spread trends; Analyze the possible problems of coupling strength enhancement warning path merging; Assess key node anomalies and warn of the formation of new transmission hubs.

Images