Industrial computer data processing method and apparatus

Abstract

The application discloses an industrial computer data processing method and device, including constructing an industrial computer control vector sequence at N continuous time stamps, determining M switching time stamps; based on the time number sequence of the control vector sequence, sorting a plurality of control parameters of the N continuous time stamps, generating a process control task chain; according to the process control task chain and the M switching time stamps, constructing an execution vector sequence; performing deviation calculation on the execution vector sequence and a pre-constructed standard execution sequence, obtaining J execution loss vectors of J sub-stage tasks; according to the J execution loss vectors, determining a to-be-corrected stage task chain of the J stage sub-tasks; the application quantifies the execution abnormal severity of each sub-stage relative to the standard process, and the constructed weighted sequence number is used as a comprehensive priority index, so that a key sub-stage with a front position and a serious deviation can be identified, and correction resources can be preferentially invested in the stage with the largest process influence.

Description

TECHNICAL FIELD

[0001] The present application relates to industrial computer data processing, in particular to an industrial computer data processing method. BACKGROUND

[0002] In the existing field of industrial process control, the analysis of computer control data is mostly focused on monitoring whether the key parameters are out of limits in real time, or performing simple statistical analysis on historical data to evaluate overall efficiency. Patent document with publication number CN113051445B discloses an industrial production data processing method and device, computer equipment and storage medium; it can continuously improve the rationality and accuracy of data filtering threshold parameters, optimize "garbage" data filtering, and reduce unnecessary resource consumption. However, in the field of similar industrial data processing as described above, when detecting abnormal control process, operation and maintenance personnel or automatic system usually process problems according to the severity of the abnormality (such as deviation amplitude) or the original execution order of the control task.

[0003] However, only according to the abnormal severity ranking, the system may prefer to handle serious problems that occur at the end of the process flow, while ignoring key problems that occur early with small deviations but have cumulative effects. And only according to the original execution order, the actual deviation state of the current process is ignored, which may lead to unnecessary checks or adjustments in the early stage of normal operation, while the response to serious abnormalities that occur later is delayed.

[0004] Therefore, the present application provides an industrial computer data processing method. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application provides an industrial computer data processing method, which solves the technical problems proposed in the background art by constructing a weighted sequence number reflecting the dual influence of position deviation as a comprehensive priority correction order index.

[0006] To achieve the above purpose, the present application is realized by the following technical scheme:

[0007] An industrial computer data processing method, characterized in that it comprises:

[0008] S1, constructing a control vector sequence of an industrial computer at N consecutive time stamps;

[0009] S2, determining M switching time stamps according to the control vector sequence;

[0010] S3, sorting a plurality of control parameters of the N consecutive time stamps based on the time number of the control vector sequence, and generating a process control task chain;

[0011] S4, constructing an execution vector sequence according to the process control task chain and the M switching time stamps;

[0012] S5, performing deviation calculation on the execution vector sequence and the pre-constructed standard execution sequence to obtain J execution loss vectors of J sub-phase tasks;

[0013] S6, determining J phase sub-tasks to be corrected according to the J execution loss vectors.

[0014] In some specific embodiments, the industrial computer constructs a control vector sequence at N consecutive time stamps, including:

[0015] S1-1, collecting a plurality of control parameters of the industrial computer at N consecutive time stamps;

[0016] S1-2, performing feature-based preprocessing on the plurality of control parameters to generate a plurality of control parameter features;

[0017] S1-3, performing feature splicing on the plurality of control parameter features to construct a control vector corresponding to each consecutive time stamp;

[0018] S1-4, traversing the N consecutive time stamps to determine the time number of the N consecutive time stamps;

[0019] S1-5, sorting the N control vectors according to the time number of the N consecutive time stamps to generate a control vector sequence.

[0020] In some specific embodiments, based on the control vector sequence, M switching time stamps are determined, including:

[0021] S2-1, calculating a composite distance in the control vector sequence;

[0022] S2-2, anchoring the M switching time stamps according to the composite distance.

[0023] In some specific embodiments, the composite distance in the control vector sequence is calculated, including:

[0024] S2-1-1, slidingly extracting a ternary control vector composed of three consecutive control vectors in the control vector sequence; wherein the ternary control vector includes a front vector, a central vector and a rear vector;

[0025] S2-1-2, calculating a composite distance between the front vector and the rear vector in the ternary control vector.

[0026] In some specific embodiments, the M switching time stamps are anchored according to the composite distance, including:

[0027] S2-2-1. Compare the composite distance between the front vector and the back vector with a preset threshold.

[0028] S2-2-2 If the composite distance is greater than the preset threshold, then the continuous timestamps corresponding to the center vector are determined to be the switching timestamps;

[0029] S2-2-3, Traverse the control vector sequence, and slide to execute S2-2-1 and S2-2-2 until M switching timestamps are obtained.

[0030] In some specific embodiments, an execution vector sequence is constructed based on the process control task chain and M switching timestamps, including:

[0031] S4-1. Anchor M switching timestamps in the process control task chain, and divide the process control task chain based on the M switching timestamps to obtain J sub-stage tasks;

[0032] S4-2. Based on the continuous timestamps occupied by the J sub-stage tasks in the process control task chain and their several control parameters, determine the task duration and stage control quality of the J sub-stage tasks.

[0033] S4-3. Traverse the J sub-stage tasks, characterize the task duration and stage control quality of each sub-stage task, and concatenate them into a binary execution vector for each sub-stage task until the binary execution vector of the J sub-stage tasks is constructed.

[0034] S4-4. Determine the actual stage number based on the time sequence of the J sub-stage tasks in the process control task chain.

[0035] S4-5. Based on the actual stage number, sort the binary execution vectors of the J sub-stage tasks and generate an execution vector sequence.

[0036] In some specific embodiments, based on J execution loss vectors, a chain of J stage subtasks to be corrected is determined, including:

[0037] S6-1. Determine the J stage weights based on the J execution loss vectors of the J sub-stage tasks;

[0038] S6-2. Calculate the J weighted indices of the standard stage number and the corresponding stage weight;

[0039] S6-3. Sort the J stage subtasks based on the J weighted sequence numbers to generate the stage task chain to be corrected.

[0040] In some specific embodiments, J stage weights are determined based on J execution loss vectors of J sub-stage tasks, including:

[0041] S6-1-1. Calculate the vector magnitude of the J execution loss vectors and define them as the J execution losses;

[0042] S6-1-2. Determine the J standard stage numbers of the J execution losses in the standard execution sequence;

[0043] S6-1-3. Normalize the J execution losses to obtain the J stage weights of the J standard stage numbers.

[0044] This invention provides an industrial computer data processing method, which has the following beneficial effects:

[0045] This invention effectively quantifies the severity of execution anomalies of each sub-stage relative to the standard process by generating stage weights corresponding to the sequence numbers of each standard stage. Furthermore, this invention quantifies the execution loss of each sub-stage relative to the standard process and generates stage weights from this loss, which are then weighted and fused with the position sequence number of the sub-stage in the standard process to construct a weighted sequence number as a comprehensive priority index. This comprehensive priority index can identify critical sub-stages that are located earlier and have serious deviations, ensuring that correction resources are prioritized for the stage that has the greatest impact on the overall process.

[0046] In a second aspect, the present invention provides an industrial computer data processing device, comprising:

[0047] The control sequence module is used to construct the control vector sequence of the industrial computer over N consecutive timestamps;

[0048] The time anchoring module is used to determine M switching timestamps based on the control vector sequence;

[0049] The task chain generation module is used to generate a process control task chain by sorting several control parameters of N consecutive timestamps based on the time number of the control vector sequence.

[0050] The execution sequence module is used to construct an execution vector sequence based on the process control task chain and M switching timestamps;

[0051] The loss calculation module is used to calculate the deviation between the execution vector sequence and the pre-built standard execution sequence to obtain J execution loss vectors for J sub-stage tasks;

[0052] The module to be corrected is used to determine the task chain of J stage subtasks to be corrected based on J execution loss vectors.

[0053] Compared with the prior art, the beneficial effects of the industrial computer data processing device of the present invention are the same as those of the industrial computer data processing method described above, so they will not be repeated here. Attached Figure Description

[0054] Figure 1 This is a flowchart illustrating an industrial computer data processing method according to the present invention.

[0055] Figure 2 This is a schematic diagram of the execution vector sequence generation process described in this invention;

[0056] Figure 3 This is a schematic diagram of the process for generating the task chain of the stage to be corrected as described in this invention;

[0057] Figure 4 This is a structural block diagram of an industrial computer data processing device according to the present invention. Detailed Implementation

[0058] 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.

[0059] Example 1: Please refer to Figures 1 to 3 This invention provides an industrial computer data processing method, comprising the following steps:

[0060] S1. Construct a control vector sequence for an industrial computer at N consecutive timestamps;

[0061] S2. Determine M switching timestamps based on the control vector sequence;

[0062] S3. Based on the time number of the control vector sequence, sort several control parameters of N consecutive timestamps to generate a process control task chain;

[0063] Specifically, the process control task chain represents a time sequence of several control parameters organized in N time number order, covering the entire process control task.

[0064] S4. Construct an execution vector sequence based on the process control task chain and M switching timestamps;

[0065] S5. Calculate the deviation between the execution vector sequence and the pre-built standard execution sequence to obtain J execution loss vectors for the J sub-stage tasks;

[0066] For example, the standard execution sequence represents a binary feature vector sequence containing J standard sub-stages, pre-constructed using historical best process data or expert experience.

[0067] Specifically, in this embodiment, the execution vector sequence contains J sub-stage tasks, the number of which is consistent with the number of sub-stages in the standard execution sequence, indicating that the actual process and the standard process are aligned in terms of operation stage division. The application of this deviation calculation method is based on the premise that the actual process and the standard process have the same stage structure. If the number of identified sub-stage tasks does not match the standard process, it indicates that there is a structural deviation or major anomaly in the actual process flow, and an alarm needs to be triggered and the current deviation analysis should be paused.

[0068] S6. Based on the J execution loss vectors, determine the task chain of the J stage subtasks to be corrected.

[0069] In this embodiment, the process control task is divided into stages by constructing a control vector sequence representing the control state and anchoring the switching timestamp. An execution vector sequence is constructed based on the task duration and control quality of each sub-stage, and the deviation is calculated with the standard execution sequence to quantify the execution loss of each stage. Finally, based on the severity of the execution loss and its time position in the standard process, a task chain of stages to be corrected is generated, which integrates the degree of deviation and the time factor. This task chain provides decision support for the industrial computer control system to perform process correction in the form of optimized processing priority, ensuring that the sub-stages with the greatest impact on the overall process are processed first.

[0070] In this embodiment, step S1 specifically includes:

[0071] S1-1: Collect several control parameters from the industrial computer at N consecutive timestamps;

[0072] For example, several control parameters represent the operating state variables and control command signals of an industrial computer in a specific process control technology. The operating state variables can be temperature, pressure, flow rate, liquid level, etc., while the control command signals can be valve opening degree, motor set speed, heating power set value, etc.

[0073] S1-2. Perform feature preprocessing on several control parameters to generate several control parameter features;

[0074] The feature preprocessing includes performing standardization, normalization, or dimensionless operations on the control parameters to eliminate the differences in dimensions and magnitudes between different parameters, forming parameter features with uniform dimensions and format.

[0075] S1-3. Perform feature concatenation on several control parameter features to construct a control vector corresponding to each consecutive timestamp;

[0076] Feature concatenation involves linking all control parameter features at the same timestamp in a predetermined order to generate a high-dimensional control vector that comprehensively represents the multidimensional control state of the industrial computer at that timestamp.

[0077] S1-4. Traverse N consecutive timestamps and determine the time number of the N consecutive timestamps;

[0078] S1-5. Based on the timestamps of N consecutive timestamps, sort the N control vectors and generate a control vector sequence.

[0079] Among them, the control vector sequence is an ordered sequence of vectors organized in chronological order, used to characterize the evolution of the control state of an industrial computer at N consecutive time stamps.

[0080] In this embodiment, by collecting multidimensional control parameters of an industrial computer at N consecutive time stamps, and performing feature preprocessing and splicing on them, a control vector representing its dynamic control state is constructed. Then, these control vectors are arranged in time sequence according to the time number to generate an ordered control vector sequence, which describes the continuous trajectory of the industrial computer's control state evolving over time in the process control task.

[0081] In this embodiment, step S1 further includes:

[0082] S2-1. Calculate the composite distance in the control vector sequence;

[0083] Furthermore, calculating the composite distance in the control vector sequence includes:

[0084] S2-1-1. In the control vector sequence, slide to extract a ternary control vector consisting of three consecutive control vectors; wherein, the ternary control vector includes: the first vector, the middle vector, and the last vector;

[0085] S2-1-2. In the ternary control vector, calculate the composite distance between the front vector and the back vector.

[0086] For example, the composite distance employs a composite metric that combines Euclidean distance and cosine distance to simultaneously capture the combined abrupt changes in the numerical and directional aspects of the front and back vectors.

[0087] Specifically, single distance metrics have limitations. Euclidean distance primarily reflects the difference in absolute values ​​between vectors but is insensitive to changes in direction; cosine distance only measures the difference in the angle between vectors, completely ignoring changes in their magnitudes. In industrial process control, a critical operation mode switch (such as switching from "heating" to "cooling") is often accompanied by changes in the magnitude of control parameters and a reconstruction of the orientation between parameters. Therefore, using a composite distance metric can more comprehensively identify complex state transitions.

[0088] In this embodiment, the formula for calculating the composite distance is:

[0089] ;

[0090] in, This represents the composite distance; the larger the value, the greater the difference between the front and back vectors. This represents the front vector in the ternary control vector. This represents the rear vector in the ternary control vector; Let be the Euclidean distance between the front vector and the back vector. and Let these represent the L2 norms of the front and back vectors, respectively. For a very small positive number, such as This is used to prevent the denominator from being zero and to ensure the stability of the first term in the calculation. This can represent the normalized Euclidean distance, used to measure the relative numerical difference between the preceding and following vectors; furthermore, For cosine similarity, therefore... This represents the cosine distance, used to measure the difference in direction between the preceding and following vectors; These are the preset coefficients for normalized Euclidean distance. These are preset coefficients for the cosine distance, satisfying α≥0, β≥0, and α+β=1. They are used to adjust the relative importance of numerical and directional differences in the composite distance. This composite distance integrates the normalized numerical and directional differences.

[0091] In this embodiment, by sliding to extract the triplet structure on the control vector sequence, the composite distance between the preceding and following vectors is calculated. This composite distance integrates the normalized Euclidean distance (which measures the relative difference in values) and the cosine distance (which measures the difference in direction), and can effectively represent the overall degree of deviation between the preceding and following vectors in the control state, thereby identifying the comprehensive abrupt changes in the industrial computer control state in terms of values ​​and parameters.

[0092] In this embodiment, step S1 further includes:

[0093] S2-2. Based on the composite distance, anchor M switching timestamps;

[0094] Further, anchoring M switching timestamps based on the composite distance includes:

[0095] S2-2-1. Compare the composite distance between the front vector and the back vector with a preset threshold.

[0096] S2-2-2 If the composite distance is greater than the preset threshold, then the continuous timestamps corresponding to the center vector are determined to be the switching timestamps;

[0097] S2-2-3, Traverse the control vector sequence, and slide to execute S2-2-1 and S2-2-2 until M switching timestamps are obtained.

[0098] Specifically, the switching timestamp represents the time point at which the industrial computer's operating mode undergoes a significant change during the control process, and is used to anchor the stage boundary of the process control task.

[0099] In this embodiment, by dynamically comparing the composite distance generated by the sliding window with a preset threshold, the moment when the control process changes significantly can be identified; by traversing the entire control vector sequence, M switching timestamps are extracted, thus realizing the anchoring of the boundaries of each operation stage in the industrial computer process control task.

[0100] In this embodiment, step S4 specifically includes:

[0101] S4-1. Anchor M switching timestamps in the process control task chain, and divide the process control task chain based on the M switching timestamps to obtain J sub-stage tasks;

[0102] Furthermore, the sub-stage task represents a continuous control parameter segment between two adjacent switching timestamps (or between the start and end timestamps and the switching timestamp) in the process control task, corresponding to a relatively stable operation stage.

[0103] S4-2. Based on the continuous timestamps occupied by the J sub-stage tasks in the process control task chain and their several control parameters, determine the task duration and stage control quality of the J sub-stage tasks.

[0104] For example, the task duration can be calculated based on the time difference corresponding to the start and end timestamps of the sub-stage task;

[0105] The stage control quality can be defined as a composite index of multiple key control parameters.

[0106] For example, key control parameters need to be maintained within a preset target range or near a set value during sub-stage tasks; therefore, the calculation of stage control quality includes:

[0107] First, for each key control parameter, its stability, deviation, and fluctuation range are calculated and fused into a single control score for that parameter. Then, the single scores of all key control parameters are normalized and weighted and fused to finally generate the stage control quality of the sub-stage task.

[0108] S4-3. Traverse the J sub-stage tasks, characterize the task duration and stage control quality of each sub-stage task, and concatenate them into a binary execution vector for each sub-stage task until the binary execution vector of the J sub-stage tasks is constructed.

[0109] Specifically, the binary execution vector is a two-dimensional vector formed by concatenating the normalized task duration and the normalized stage control quality, and is used to quantify the overall level of execution efficiency and control performance of the corresponding sub-stage task.

[0110] S4-4. Determine the actual stage number based on the time sequence of the J sub-stage tasks in the process control task chain.

[0111] S4-5. Based on the actual stage number, sort the binary execution vectors of the J sub-stage tasks and generate an execution vector sequence.

[0112] Furthermore, the execution vector sequence represents a set of vectors representing the execution performance (execution efficiency and control performance) of each sub-stage task arranged according to the actual stage number. Compared with the control vector sequence mentioned above, which detects instantaneous state changes, its time granularity is coarser, and it can explore the evolution pattern of operation mode from the dimension of task execution performance.

[0113] In this embodiment, step S6 specifically includes:

[0114] S6-1. Determine the J stage weights based on the J execution loss vectors of the J sub-stage tasks;

[0115] S6-2. Calculate the J weighted indices of the standard stage number and the corresponding stage weight;

[0116] It should be noted that the weighted sequence number is the product of the standard stage sequence number and the corresponding stage weight.

[0117] In the practice of process correction task chains, sorting based solely on the severity of deviations (stage weight) may lead to late-stage critical issues being addressed before early-stage critical issues, and the cumulative effect of early issues may have a greater impact on subsequent stages. Conversely, sorting based solely on time position (standard stage number) fails to reflect the degree of anomaly in the current execution. Therefore, to balance the dual impact of time position and deviation severity on correction priority, this embodiment uses the product of the two as the priority indicator.

[0118] This product term combines the time position of the sub-stage in the standard process with the severity of the deviation currently being executed, and is used to quantitatively assess the overall priority of the stage in the overall process correction: the earlier the stage is (the smaller the sequence number) and the more severe the deviation (the higher the weight), the higher its overall priority and should be corrected first.

[0119] For example, the sorting logic of the weighted sequence is as follows:

[0120] Phase 1: Standard Phase Number = 1, Phase Weight = 0.9 → Weighted Phase Number = 0.9 (Highest Priority)

[0121] Phase 2: Standard Phase Number = 2, Phase Weight = 0.8 → Weighted Phase Number = 1.6

[0122] Phase 3: Standard phase number = 3, phase weight = 0.95 → weighted phase number = 2.85 (although the deviation is the most serious, its priority is lower than that of phase 1 because of its later position).

[0123] S6-3. Sort the J stage subtasks based on the J weighted sequence numbers to generate the stage task chain to be corrected.

[0124] Ultimately, the task chain to be corrected does not represent a new execution order in time, but a processing queue arranged in descending order of comprehensive priority. More precisely, it generates an optimized processing priority list that integrates the standard stage position and the current execution deviation, which is used to guide the industrial computer control system to prioritize the adjustment of key control parameters or key sub-stage tasks that have the greatest impact on the overall process when performing process correction tasks.

[0125] In this embodiment, the execution loss vector of each sub-stage task is transformed into a stage weight representing the severity of the deviation. This weight is then innovatively multiplied by a standard stage number to construct a weighted sequence reflecting the dual impact of "position-deviation" as a comprehensive priority index. Based on this index, the sub-stage tasks are sorted, generating a task chain for process correction. This task chain, in the form of an optimized processing priority list, provides decision support for the industrial computer control system, enabling it to prioritize stages that are critical and severely deviated when performing correction tasks. This leads to more efficient restoration of process stability and improved overall control performance.

[0126] Furthermore, step S6-1 specifically includes:

[0127] S6-1-1. Calculate the vector magnitude of the J execution loss vectors and define them as the J execution losses;

[0128] In this embodiment, the vector magnitude represents the degree of comprehensive deviation of the execution loss vector in the binary parameter space consisting of execution time and stage control quality. Therefore, using the vector magnitude as a quantitative indicator of execution loss can integrate the combined effects of execution time deviation and control quality deviation to generate a scalarized comprehensive loss value.

[0129] S6-1-2. Determine the J standard stage numbers of the J execution losses in the standard execution sequence;

[0130] Furthermore, the standard stage number of the execution loss is obtained by the position index of its corresponding sub-stage task in the execution vector sequence, and this position index corresponds one-to-one with the preset number of each standard sub-stage in the standard execution sequence.

[0131] S6-1-3. Normalize the J execution losses to obtain the J stage weights of the J standard stage numbers;

[0132] Specifically, execution loss reflects the overall deviation of a sub-stage task from the standard process. The larger the value, the more severe the execution anomaly in that stage. Therefore, through normalization, the execution loss can be converted into a stage weight corresponding to the standard stage number. This weight represents the severity of the deviation in the current process.

[0133] In this embodiment, the deviations of the execution loss vector in both task duration and control quality are fused into a scalarized comprehensive loss by calculating the magnitude of the execution loss vector. Subsequently, the standard stage number is determined based on the position of the sub-stage in the execution sequence, and the stage weight corresponding to each standard stage number is generated by normalizing the comprehensive loss value. This weighting system quantifies the severity of the execution anomaly of each sub-stage relative to the standard process, providing support for determining the correction priority through the weighted sequence number (the product of the standard stage number and the stage weight).

[0134] Example 2: See Figure 4 This embodiment 2 differs from embodiment 1 in that it also discloses an industrial computer data processing device. This device is used to implement the above-described method embodiments, and details already described will not be repeated. The terms "module," "unit," and "subunit" used below refer to combinations of software and / or hardware that perform a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0135] like Figure 4 As shown, Figure 4 This is a structural block diagram of an industrial computer data processing device according to the present invention, the device comprising:

[0136] The control sequence module is used to construct the control vector sequence of the industrial computer over N consecutive timestamps;

[0137] The time anchoring module is used to determine M switching timestamps based on the control vector sequence;

[0138] The task chain generation module is used to generate a process control task chain by sorting several control parameters of N consecutive timestamps based on the time number of the control vector sequence.

[0139] The execution sequence module is used to construct an execution vector sequence based on the process control task chain and M switching timestamps;

[0140] The loss calculation module is used to calculate the deviation between the execution vector sequence and the pre-built standard execution sequence to obtain J execution loss vectors for J sub-stage tasks;

[0141] The module to be corrected is used to determine the task chain of J stage subtasks to be corrected based on J execution loss vectors.

[0142] In the aforementioned device, a control vector sequence is constructed through the control sequence module; M switching timestamps are anchored through the time anchoring module; a process control task chain is generated through the task chain generation module; an execution vector sequence is constructed through the execution sequence module; J execution loss vectors for J sub-stage tasks are obtained through the loss calculation module; and the task chain to be rectified for J stage sub-tasks is determined through the correction module. This solves the problem of a single priority ranking index for stage sub-tasks.

[0143] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means.

[0144] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. An industrial computer data processing method, characterized in that, include: S1. Construct a control vector sequence for an industrial computer at N consecutive timestamps; S2. Determine M switching timestamps based on the control vector sequence; S3. Based on the time number of the control vector sequence, sort several control parameters of N consecutive timestamps to generate a process control task chain; S4. Construct an execution vector sequence based on the process control task chain and M switching timestamps; S5. Calculate the deviation between the execution vector sequence and the pre-built standard execution sequence to obtain J execution loss vectors for the J sub-stage tasks; S6. Based on the J execution loss vectors, determine the task chain of the J stage subtasks to be corrected. The step of constructing an execution vector sequence based on the process control task chain and M switching timestamps includes: S4-1. Anchor M switching timestamps in the process control task chain, and divide the process control task chain based on the M switching timestamps to obtain J sub-stage tasks; S4-2. Based on the continuous timestamps occupied by the J sub-stage tasks in the process control task chain and their several control parameters, determine the task duration and stage control quality of the J sub-stage tasks. S4-3. Traverse the J sub-stage tasks, characterize the task duration and stage control quality of each sub-stage task, and concatenate them into a binary execution vector for each sub-stage task until the binary execution vector of the J sub-stage tasks is constructed. S4-4. Determine the actual stage number based on the time sequence of the J sub-stage tasks in the process control task chain. S4-5. Based on the actual stage number, sort the binary execution vectors of the J sub-stage tasks and generate an execution vector sequence. The step of determining the chain of stage tasks to be corrected for J stage subtasks based on J execution loss vectors includes: S6-1. Determine the J stage weights based on the J execution loss vectors of the J sub-stage tasks; S6-2. Calculate the J weighted indices of the standard stage number and the corresponding stage weight; The weighted sequence number is the product of the standard stage sequence number and the corresponding stage weight. The product term integrates the time position of the sub-stage task in the standard process and the severity of the deviation currently being executed, and is used to quantitatively evaluate the comprehensive priority of the sub-stage task in the overall process correction. The comprehensive priority indicates that the sub-stage task that is earlier in the process and has a more severe deviation has a higher comprehensive priority and should be corrected first. S6-3. Sort the J stage subtasks based on the J weighted indexes and generate the stage task chain to be corrected. The step of determining the J stage weights based on the J execution loss vectors of the J sub-stage tasks includes: S6-1-1. Calculate the vector magnitude of the J execution loss vectors and define them as the J execution losses; S6-1-2. Determine the J standard stage numbers of the J execution losses in the standard execution sequence; S6-1-3. Normalize the J execution losses to obtain the J stage weights of the J standard stage numbers.

2. The industrial computer data processing method according to claim 1, characterized in that, Constructing a control vector sequence for an industrial computer at N consecutive time stamps includes: S1-1: Collect several control parameters from the industrial computer at N consecutive timestamps; S1-2. Perform feature preprocessing on several control parameters to generate several control parameter features; S1-3. Perform feature concatenation on several control parameter features to construct a control vector corresponding to each consecutive timestamp; S1-4. Traverse N consecutive timestamps and determine the time number of the N consecutive timestamps; S1-5. Based on the timestamps of N consecutive timestamps, sort the N control vectors and generate a control vector sequence.

3. The industrial computer data processing method according to claim 1, characterized in that, Based on the control vector sequence, M switching timestamps are determined, including: S2-1. Calculate the composite distance in the control vector sequence; S2-2. Based on the composite distance, anchor M switching timestamps.

4. The industrial computer data processing method according to claim 3, characterized in that, Calculate the composite distance in the control vector sequence, including: S2-1-1. In the control vector sequence, slide to extract a ternary control vector consisting of three consecutive control vectors; wherein, the ternary control vector includes: the first vector, the middle vector, and the last vector; S2-1-2. In the ternary control vector, calculate the composite distance between the front vector and the back vector.

5. The industrial computer data processing method according to claim 4, characterized in that, Based on the composite distance, M switching timestamps are anchored, including: S2-2-1. Compare the composite distance between the front vector and the back vector with a preset threshold. S2-2-2 If the composite distance is greater than the preset threshold, then the continuous timestamps corresponding to the center vector are determined to be the switching timestamps; S2-2-3, Traverse the control vector sequence, and slide to execute S2-2-1 and S2-2-2 until M switching timestamps are obtained.

6. An industrial computer data processing apparatus for executing the industrial computer data processing method according to any one of claims 1 to 5, characterized in that, include: The control sequence module is used to construct the control vector sequence of the industrial computer over N consecutive timestamps; The time anchoring module is used to determine M switching timestamps based on the control vector sequence; The task chain generation module is used to generate a process control task chain by sorting several control parameters of N consecutive timestamps based on the time number of the control vector sequence. The execution sequence module is used to construct an execution vector sequence based on the process control task chain and M switching timestamps; The loss calculation module is used to calculate the deviation between the execution vector sequence and the pre-built standard execution sequence to obtain J execution loss vectors for J sub-stage tasks; The module to be corrected is used to determine the task chain of the J stage subtasks to be corrected based on the J execution loss vectors. The step of constructing an execution vector sequence based on the process control task chain and M switching timestamps includes: S4-1. Anchor M switching timestamps in the process control task chain, and divide the process control task chain based on the M switching timestamps to obtain J sub-stage tasks; S4-2. Based on the continuous timestamps occupied by the J sub-stage tasks in the process control task chain and their several control parameters, determine the task duration and stage control quality of the J sub-stage tasks. S4-3. Traverse the J sub-stage tasks, characterize the task duration and stage control quality of each sub-stage task, and concatenate them into a binary execution vector for each sub-stage task until the binary execution vector of the J sub-stage tasks is constructed. S4-4. Determine the actual stage number based on the time sequence of the J sub-stage tasks in the process control task chain. S4-5. Based on the actual stage number, sort the binary execution vectors of the J sub-stage tasks and generate an execution vector sequence. The step of determining the chain of stage tasks to be corrected for J stage subtasks based on J execution loss vectors includes: S6-1. Determine the J stage weights based on the J execution loss vectors of the J sub-stage tasks; S6-2. Calculate the J weighted indices of the standard stage number and the corresponding stage weight; The weighted sequence number is the product of the standard stage sequence number and the corresponding stage weight. The product term integrates the time position of the sub-stage task in the standard process and the severity of the deviation currently being executed, and is used to quantitatively evaluate the comprehensive priority of the sub-stage task in the overall process correction. The comprehensive priority indicates that the sub-stage task that is earlier in the process and has a more severe deviation has a higher comprehensive priority and should be corrected first. S6-3. Sort the J stage subtasks based on the J weighted indexes and generate the stage task chain to be corrected. The step of determining the J stage weights based on the J execution loss vectors of the J sub-stage tasks includes: S6-1-1. Calculate the vector magnitude of the J execution loss vectors and define them as the J execution losses; S6-1-2. Determine the J standard stage numbers of the J execution losses in the standard execution sequence; S6-1-3. Normalize the J execution losses to obtain the J stage weights of the J standard stage numbers.

Images