An operation sequence matching-based actual operation process specification detection method and system

Abstract

The application relates to a real operation process specification detection method and system based on operation sequence matching, which comprises the following steps: establishing a standard operation sequence and determining steps that must be executed; generating an actual operation sequence; identifying anchor operation steps; identifying interval operation steps in a real timestamp interval corresponding to two continuous anchor operation steps, and evaluating logical sequence consistency; reversely deducing an actual operation context of the interval operation steps and performing comparison verification; generating a real operation specification judgment conclusion based on the identification results of the anchor operation steps and the interval operation steps, the evaluation results of the logical sequence consistency and the comparison verification results; and in summary, the application can effectively solve the problems of high false positive rate, high false negative rate, neglect of operation context and incapability of adapting to complex processes by establishing a standard operation sequence, identifying anchor operation steps, evaluating logical sequence consistency and comparing operation contexts.

Description

Technical Field

[0001] This application relates to the technical field of process standardization testing, and in particular to a practical process standardization testing method and system based on operation sequence matching. Background Technology

[0002] In fields such as medical surgery, industrial assembly, equipment maintenance, and skills training, where strict adherence to standardized procedures is required, accurate testing and evaluation of the standardization of practical procedures are crucial for ensuring operational safety and improving work quality and efficiency.

[0003] Existing automated compliance testing methods primarily rely on pre-defined step sequences and timelines, directly comparing them with records generated by operators during actual execution. However, this method has significant limitations: First, the comparison between the pre-defined step sequences and operators is usually based on simple matching of the presence or absence of steps or time points, lacking effective modeling of the complex logical dependencies between actual steps and verification steps. It is difficult to distinguish between critical, necessary, and auxiliary steps, ultimately leading to high false positive and false negative rates. Second, existing methods generally lack consideration of the operational context, i.e., the state of the object being operated on and the system when the steps are executed. The same operation performed under different system states may have drastically different safety and effectiveness, ultimately making it difficult to guarantee safety and effectiveness. Therefore, based on these limitations, existing automated testing technologies struggle to perform state- and logic-based compliance testing on complex operational processes, severely restricting their effective application in complex processes and high-risk operational scenarios. Summary of the Invention

[0004] To address the aforementioned shortcomings, this application provides a method for detecting the standardization of practical procedures based on operation sequence matching.

[0005] The above-mentioned objective of this application is achieved through the following technical solution: A practical procedure standardization detection method based on operation sequence matching includes the following steps: Determine the target operational process, establish the standard operation sequence corresponding to the operational process, and determine the steps that must be executed. The standard operation sequence is a sequence that includes several key operation steps and their expected execution parameters. The key operation steps are associated with a standard timestamp that allows for flexible operation intervals and an operation context used to indicate the operation status. Real-time collection of operation records and generation of actual operation sequences, wherein the actual operation sequence is a sequence including several actual operation steps and their corresponding actual execution parameters, and the actual operation steps are associated with real timestamps; Identify the step type of the actual operation steps, and based on the combination characteristics of the step type and the actual execution parameters, identify the actual operation steps corresponding to each key operation step as anchor operation steps. Within the real timestamp interval corresponding to two consecutive anchoring operation steps, identify whether the actual operation sequence contains the actual operation steps corresponding to the required execution steps of the standard operation sequence within the real timestamp interval, and use them as interval operation steps, and evaluate the consistency of the logical order of all actual operation steps within the real timestamp interval with the standard operation sequence. The actual operational context is derived from the actual execution parameters of the interval operation steps within the real timestamp interval, and the derived actual operational context is compared and verified with the operational context associated with the corresponding key operation steps in the standard operation sequence. Based on the identification results of anchored operation steps and interval operation steps, the evaluation results of logical sequence consistency, and the comparison and verification results, a conclusion on the standardization of practical operation is generated.

[0006] The second objective of this invention is achieved through the following technical solution: A practical process standardization detection system based on operation sequence matching includes: The standard operation sequence establishment module is used to determine the target practical process, establish the standard operation sequence corresponding to the practical process, and determine the steps that must be executed. The standard operation sequence is a sequence including several key operation steps and their expected execution parameters. The key operation steps are associated with a standard timestamp that allows for flexible operation intervals and an operation context used to indicate the operation status. The actual operation sequence generation module is used to collect operation records in real time and generate actual operation sequences. The actual operation sequence is a sequence including several actual operation steps and their corresponding actual execution parameters. The actual operation steps are associated with real timestamps. The anchoring operation step determination module is used to identify the step type of the actual operation step, and based on the combination characteristics of the step type and the actual execution parameters, identify the actual operation steps corresponding to each key operation step as anchoring operation steps. The interval operation step determination module is used to identify whether the actual operation sequence contains the actual operation steps corresponding to the mandatory steps of the standard operation sequence within the real timestamp interval corresponding to two consecutive anchor operation steps, and to evaluate the consistency of the logical order of all actual operation steps within the real timestamp interval with the standard operation sequence. The comparison and verification module is used to reverse-engineer the actual operation context from the actual execution parameters of the interval operation steps within the real timestamp interval, and compare and verify the reverse-engineered actual operation context with the operation context associated with the corresponding key operation steps in the standard operation sequence. The standardization judgment module is used to generate a practical standardization judgment conclusion based on the identification results of anchored operation steps and interval operation steps, the evaluation results of logical sequence consistency, and the comparison and verification results.

[0007] In summary, the practical process standardization detection method and system provided in this application, based on operation sequence matching, can effectively solve the problems of high false alarm rate, high false alarm rate, neglect of operation context, and inability to adapt to complex processes in the prior art by establishing standard operation sequences, identifying anchored operation steps, evaluating the consistency of logical order, and comparing operation context. It has the effects of improving detection accuracy, reducing false alarm and false alarm rates, and effectively adapting to the nonlinear structure in complex processes, thereby improving the reliability and security of practical process standardization detection. Attached Figure Description

[0008] Figure 1 This is a flowchart of an embodiment of a practical process standardization detection method based on operation sequence matching according to this application; Figure 2 This is a flowchart of step S10 in an embodiment of a practical process standardization detection method based on operation sequence matching in this application; Figure 3 This is a flowchart of step S30 in an embodiment of a practical process standardization detection method based on operation sequence matching in this application. Detailed Implementation

[0009] The following is in conjunction with the appendix Figures 1-3 This application will be described in further detail.

[0010] In fields such as medical surgery, industrial assembly, equipment maintenance, and skills training, the standardization of practical procedures must ensure operational safety and improve work quality. Existing automated testing methods largely rely on direct comparison of preset step sequences and timelines, failing to effectively model the complex logical dependencies between actual and verification steps. This makes it difficult to distinguish between critical, necessary, and auxiliary steps. The lack of logical dependencies leads to the inability to correctly identify necessary constraints between steps, such as the decisive impact of critical steps on subsequent operational states. Furthermore, ignoring the operational context means that the state of the object being operated on and the system should be in during step execution is not included in the evaluation, and the same operation performed under different system states may produce drastically different levels of safety and effectiveness. Therefore, existing technologies cannot achieve state- and logic-based standardization testing of complex practical procedures, affecting the accuracy and reliability of testing results and thus limiting their applicability in high-risk operational scenarios.

[0011] For example, in coronary artery bypass grafting in cardiac surgery, standard procedures require aseptic disinfection and confirmation of a sterile skin surface before opening the thoracic cavity. Existing detection systems only record the timing of disinfection and incision, but do not monitor the actual microbial load and environmental parameters of the skin after disinfection. When operators perform disinfection but do not do so thoroughly, the procedure may be deemed standard due to its presence, even though the actual skin condition does not meet sterility requirements, leading to postoperative infection risks after opening the thoracic cavity. In this scenario, the logical dependency (disinfection must lead to a sterile state) is ignored, and the operational context (skin condition and operating room environment) is not modeled, making it impossible for standard detection to distinguish between the superficial presence and actual effect of the disinfection procedure, resulting in distorted detection results.

[0012] If the above problems are not addressed, the standardization testing of practical procedures will be unable to identify violations based on the operational context, increasing safety hazards in high-risk operations: in medical surgery, it may lead to irreversible complications; in industrial assembly, it may trigger a chain of equipment failures or safety accidents; at the same time, the low reliability of standardization test results will also reduce the standardization of process execution, affecting the improvement of overall operation quality and efficiency.

[0013] In one embodiment, such as Figure 1 As shown, this application discloses a method for detecting the standardization of practical procedures based on operation sequence matching, which specifically includes the following steps: S10: Determine the target practical process, establish the standard operation sequence corresponding to the practical process, and determine the steps that must be executed. The standard operation sequence is a sequence that includes several key operation steps and their expected execution parameters. The key operation steps are associated with a standard timestamp that allows for flexible operation intervals and an operation context used to indicate the operation status. In this embodiment, the target practical process refers to a specific operational process that requires standardized testing, such as equipment maintenance, production line operation, or medical surgery; the standard operation sequence refers to the predefined standardized operation steps for the target practical process, used as a reference benchmark for actual operation; the critical operation step refers to the operation in the standard operation sequence that has a decisive impact on the process status or result, and the correctness of its execution is particularly important for the standardization of the entire process; the expected execution parameter refers to the specific value or state that the operating object or system environment should have when the critical operation step is executed in the standard operation sequence; the standard timestamp refers to the time point at which the critical operation step is expected to be executed in the standard operation sequence; the flexible operation interval refers to the time range within which the critical operation step is allowed to be actually executed before and after the standard timestamp, in order to accommodate reasonable deviations in actual operation; the operation context refers to the set of states of the operating object or system environment when a specific operation step is executed, which reflects the prerequisites of the operation and the state of the result after execution; the mandatory execution step refers to the operation step in the standard operation sequence that must be executed under any circumstances to ensure the safety, effectiveness, or compliance with specific specifications of the process.

[0014] S20: Real-time acquisition of operation records and generation of actual operation sequence, wherein the actual operation sequence is a sequence including several actual operation steps and their corresponding actual execution parameters, and the actual operation steps are associated with real timestamps; In this embodiment, the actual operation sequence refers to the operation steps formed by real-time collection of operation records, reflecting the actual execution process of the operator; the actual operation step refers to each specific operation behavior in the actual operation sequence; the actual execution parameter refers to the actual value or state of the operation object or system environment when the actual operation step is executed; the real timestamp refers to the actual time point when the actual operation step is executed in the actual operation sequence.

[0015] S30: Identify the step type of the actual operation steps, and based on the combination characteristics of the step type and the actual execution parameters, identify the actual operation steps corresponding to each key operation step as anchor operation steps. In this embodiment, the anchoring operation step refers to the actual operation step that is identified in the actual operation sequence as corresponding to the key operation step in the standard operation sequence, and serves as a key reference point for process matching.

[0016] S40: Within the real timestamp interval corresponding to two consecutive anchored operation steps, identify whether the actual operation sequence contains the actual operation steps corresponding to the required steps of the standard operation sequence within the real timestamp interval, and use them as interval operation steps, and evaluate the consistency of the logical order of all actual operation steps within the real timestamp interval with the standard operation sequence. In this embodiment, the interval operation step refers to the actual operation step included in the actual operation sequence between two consecutive anchor operation steps, which corresponds to the required execution steps of the standard operation sequence within the real time interval; logical order consistency refers to the degree to which the execution order of the actual operation steps conforms to the logical dependencies defined in the standard operation sequence.

[0017] S50: Derive the actual operational context from the actual execution parameters of the interval operation steps within the real timestamp interval, and compare and verify the actual operational context derived from the reverse derivation with the operational context associated with the corresponding key operation steps in the standard operation sequence. In this embodiment, the actual operating context refers to the set of states that are actually in the operation object or system environment when the step is executed, which is derived from the actual execution parameters of the interval operation steps.

[0018] S60: Based on the identification results of anchored operation steps and interval operation steps, the evaluation results of logical sequence consistency, and the comparison and verification results, generate a conclusion on the standardization of practical operation.

[0019] In this embodiment, the conclusion on the standardization of practical operation refers to the final judgment on whether the practical operation conforms to the standard, based on the comparison and evaluation results of the actual operation sequence and the standard operation sequence.

[0020] For example, suppose there is a device maintenance procedure with the goal of "replacing the filter of device A". Existing methods may not be able to effectively detect whether critical safety steps are missed during the filter replacement process or whether the operation is performed in an incorrect device state.

[0021] First, the target practical process is determined to be "replacing the filter element of device A". To this end, a corresponding standard operating sequence is established, which includes the following key operating steps: 1. Disconnect the power supply to device A (key operating step, expected execution parameters: power off, standard timestamp: T1, operating context: device is in a safe power-off state, this step must be performed); 2. Open the casing of device A (key operating step, expected execution parameters: casing is open, standard timestamp: T2, operating context: casing is open); 3. Remove the old filter element (key operating step, expected execution parameters: old filter element has been removed, standard timestamp: T3, operating context: casing is open); 4. Install a new filter element (critical operation step, expected execution parameters: new filter element installed, standard timestamp: T4, operation context: new filter element is in place, mandatory step); 5. Close the casing of device A (critical operation step, expected execution parameters: casing is closed, standard timestamp: T5, operation context: casing is closed); 6. Turn on the power to device A (critical operation step, expected execution parameters: power status is on, standard timestamp: T6, operation context: device is in normal power-on state). Each critical operation step is given a flexible operating range that allows for reasonable deviations.

[0022] Furthermore, real-time data is collected on user A's operation of "replacing the filter of device A" at location A, generating an actual operation sequence. For example, the actual operation sequence might be: Actual operation step 1: Open the casing of device A (real timestamp: t1, actual execution parameter: casing status: open); Actual operation step 2: Remove the old filter (real timestamp: t2, actual execution parameter: old filter status: removed); Actual operation step 3: Disconnect the power to device A (real timestamp: t3, actual execution parameter: power status: off); Actual operation step 4: Install the new filter (real timestamp: t4, actual execution parameter: new filter status: installed); Actual operation step 5: Close the casing of device A (real timestamp: t5, actual execution parameter: casing status: closed); Actual operation step 6: Turn on the power to device A (real timestamp: t6, actual execution parameter: power status: on).

[0023] Furthermore, the step types of the actual operation steps are identified, and based on the combination characteristics of the step type and the actual execution parameters, the actual operation steps corresponding to each key operation step are identified as anchor operation steps. For example, actual operation step 1 "open the casing of device A" is identified as corresponding to the key operation step "open the casing of device A" in the standard operation sequence, thus becoming an anchor operation step; similarly, actual operation steps 3 "disconnect the power supply of device A", 4 "install a new filter", 5 "close the casing of device A", and 6 "connect the power supply of device A" are also identified as anchor operation steps.

[0024] Furthermore, within the real timestamp interval corresponding to two consecutive anchoring operation steps, it is identified whether the actual operation sequence contains the actual operation steps corresponding to the mandatory execution steps of the standard operation sequence within the real timestamp interval, and the consistency of the logical order is evaluated; for example, consider the real timestamp interval from the anchoring operation step "disconnect the power supply of device A" (actual operation step 3) to the anchoring operation step "install a new filter" (actual operation step 4), i.e., t3 to t4; within this real timestamp interval, the mandatory execution steps of the standard operation sequence are "disconnect the power supply of device A" and "install a new filter", and the actual operation sequence contains actual operation steps 3 and 4 within this real timestamp interval.

[0025] Furthermore, the consistency of the logical sequence was evaluated. The standard operating sequence requires first disconnecting the power to device A, then opening the casing of device A, then removing the old filter, and finally installing the new filter. However, in the actual operating sequence, opening the casing of device A and removing the old filter occur before disconnecting the power to device A. This identifies an inconsistency in the logical sequence. Simultaneously, the actual operating context is derived by reversing the actual execution parameters of the interval operation steps within the real timestamp interval. This derived operating context is then compared and verified with the operating context associated with the corresponding key operation steps in the standard operating sequence. For example, when actual operation step 1, "open the casing of device A," is executed (real timestamp t1), its actual execution parameters show "casing status: open." The derived operating context is "casing is open." However, the expected operating context for the "open the casing of device A" step in the standard operating sequence is "device is in a safe power-off state" and "casing is open." Comparison reveals that at time t1, the device power is not disconnected, therefore the actual operating context differs from the expected operating context.

[0026] Finally, based on the identification results of the anchoring operation steps and the interval operation steps, the evaluation results of the consistency of the logical order, and the comparison and verification results, a conclusion on the standardization of practical operation is generated. In the above example, due to the inconsistency of the logical order (opening the shell and taking out the filter first, and then cutting off the power) and the inconsistency of the operation context (opening the shell while the equipment is not powered off), the final conclusion is "non-standard operation", and it can be further pointed out that "the power disconnection step is executed too late, which poses a safety risk".

[0027] Compared to existing automated standardization detection methods, as described in the background section, which mainly rely on simple comparisons between preset step sequences and timelines, it is difficult to effectively model complex logical dependencies and lacks consideration of operational context. In the example above, if existing technology is used, it may only be able to detect whether the necessary step of "disconnecting the power to device A" has been executed, or whether its execution time is within a preset time window. However, existing technology is unable to detect the logical sequence error of opening the casing of device A and removing the old filter before disconnecting the power to device A. This embodiment, by evaluating the logical sequence consistency of all actual operation steps with the standard operation sequence between consecutive anchored operation steps, can accurately identify such sequence conflicts, thereby making up for the shortcomings of existing technology in modeling logical dependencies.

[0028] Furthermore, existing technologies generally lack consideration of the operational context, failing to determine whether an operation was performed under the correct system conditions. In the example above, user A performed the operation of "opening the casing of device A" while the device was not powered off. If only the presence or absence of steps or the time point are compared, this potential safety risk will not be detected. This embodiment, by reverse-engineering the actual operational context from the actual execution parameters of the interval operation steps and comparing and verifying it with the operational context associated with the corresponding key operation steps in the standard operation sequence, can effectively detect situations where the operational context is inconsistent. Thus, this embodiment can identify potential operational risks or normative defects caused by context deviations, improving the comprehensiveness and accuracy of standardized process detection.

[0029] In summary, the technical solution of this embodiment, by introducing features such as key operation steps, mandatory execution steps, operation context, and logical sequence consistency assessment, can achieve standardization detection of practical operation processes based on state and logic. This enables a deeper understanding of the inherent logic and environmental dependencies of the operation, providing more accurate and comprehensive standardization judgments. It can effectively solve the problem of limited application of existing technologies in complex processes and high-risk operation scenarios.

[0030] In one embodiment, such as Figure 2 As shown, step S10 includes: S11: Based on the standard operation sequence, determine several key operation step pairs, and according to the operation context associated with each key operation step pair, determine at least one intermediate state that must be satisfied in the operation context transition from the previous key operation step to the next key operation step in the key operation step pair. In this embodiment, a critical operation step pair refers to two adjacent or logically related critical operation steps in a standard operation sequence. Determining intermediate states is to refine the state transition paths between critical steps, ensuring that the system state evolution between two critical operations is as expected, rather than a direct jump or omission of necessary intermediate steps. For example, for the critical step pair of "starting the device" and "setting device operating parameters," if the context after "starting the device" is "the device is in standby mode," and the context before "setting device operating parameters" is "the device is in running mode," then the intermediate state may be "the device switches from standby mode to running mode." These intermediate states may be achieved through self-checks, initialization, or other operations, or they may be defined by analyzing domain knowledge bases, device manuals, or expert experience to define the intermediate states that may exist and must be passed between each critical operation step pair.

[0031] The key operational steps are identified by analyzing the target operational process to pinpoint all operational nodes that could cause irreversible state transitions in the corresponding operational object or system environment. This analysis, combined with historical data, filters out nodes where omission or incorrect execution would trigger pre-defined risks, forming a set of key operational steps. Irreversible state transition operational nodes are those steps that, once executed, fundamentally change the state of the operational object or system environment and cannot be easily reversed to the previous state, such as equipment startup, valve opening, and data submission. Filtering out nodes that would trigger pre-defined risks through historical data ensures the safety of the target operational process. The key operational steps identified are not only state transition points, but also crucial risk control points for process safety, efficiency, and compliance. The identification process can involve manual or semi-automatic identification of state transition points in the process through expert experience, flowchart analysis, and failure mode and effects analysis. It can also utilize statistical methods or machine learning models to identify operational nodes associated with high-risk events by analyzing historical operation logs, incident reports, and violation records. Alternatively, process mining techniques can be used to automatically discover process models from large amounts of operation logs and identify key decision points or state transition points within these models, while simultaneously combining predefined risk rule bases and historical risks. Event tagging involves risk assessment and screening of identified key decision points or state transition points to construct a set of key operational steps. Operational context is defined by extracting features from the expected execution parameters of each key operational step that characterize the specific system state formed or caused by that step, and defining these features as the operational context associated with that key operational step. Operational context is an abstract representation of the system's state after the execution of a key operational step; it is not the operation itself, but rather the environmental or object state resulting from the operation. Defining operational context by extracting features from expected execution parameters allows for the association of operational behavior with the resulting state, providing contextual comparison. For example, for a key operational step of "opening a valve," the expected execution parameters may include "valve number," "opening angle," and "opening time." We can extract "valve number X is in the open state, opening angle is Y" as the operational context. Alternatively, we can predefine a contextual description language or ontology, mapping the expected execution parameters of different key operational steps to specific state features within this language. For instance, through a rule engine or semantic parser, the expected parameters for "setting the temperature to 100 degrees Celsius" can be parsed into a set of state features such as "temperature sensor reading = 100℃" and "heater status = ON," serving as the operational context.

[0032] S12: Based on the determined intermediate state and the constraints of the operation object corresponding to the actual operation process, determine several candidate operation steps that can realize the intermediate state transition. Furthermore, constraints include physical constraints, security specifications, and logical dependencies.

[0033] In this embodiment, constraints are rules that guide operational behavior to ensure the compliance, safety, and effectiveness of operations. For example, if the intermediate state is "the equipment is in operation," the constraints may include physical constraints and safety regulations such as "power is connected," "cooling system is started," and "safety door is closed." In this case, candidate operation steps that can achieve the intermediate state transition may include "pressing the start button," "checking the power indicator light," and "confirming the coolant level." In one implementation, all operations that may lead to a specific intermediate state can be listed by analyzing process specification documents and equipment operation manuals, combined with expert knowledge. At the same time, the listed operations are matched with predefined physical (such as equipment connection method, material compatibility), safety (such as operation sequence, protective measures), and logical (such as data dependency, preconditions) constraints to filter out candidate operation steps that meet the conditions.

[0034] S13: Select the candidate operation steps that are necessary to achieve the intermediate state transition and whose execution order is uniquely limited by the constraint conditions from the candidate operation steps, and then select them as the steps that must be executed.

[0035] In this embodiment, for example, in the case of "the device switching from standby state to running state", if pressing the start button is the only operation that can change the device from standby to running, and it must be performed after checking the power connection and before setting the running parameters, then pressing the start button may be determined as a necessary step. In one implementation, a decision tree or rule engine can also be constructed, and candidate operation steps and constraints can be input. By analyzing the constraints, such as "operation A must precede operation B", "operation C and operation D must be performed simultaneously", "operation E is the only way to reach state F", etc., operations that are both irreplaceable and have a fixed order on a specific intermediate state transition path can be identified and marked as necessary steps.

[0036] For example, suppose a practical procedure is a "nuclear power plant reactor startup procedure". First, analyze the practical procedure to identify operational nodes such as "control rods raised to critical position", "main coolant pump start-up", and "turbine grid connection". These nodes will cause irreversible transitions in the reactor or power grid state. Further, combined with historical accident data, it was found that "control rods raised too quickly" or "control rods raised before the main coolant pump started" had caused risks. Based on this, operations such as "control rods raised to critical position", "main coolant pump start-up", and "turbine grid connection" were identified as critical operation steps. Further, from the expected execution parameters of "control rods raised to critical position" (such as control rod position and raising speed), "the reactor reaches the critical state and the power slowly increases" was extracted as its operational context. Based on the standard operation sequence, "main coolant pump start-up" and "control rods raised to critical position" were identified as a critical operation step pair. Based on the associated operational context, the necessary intermediate state between "main coolant pump not started" and "reactor reaches criticality" is determined to be "main coolant system operating normally, flow rate reaching preset value". Furthermore, based on this intermediate state of "main coolant system operating normally, flow rate reaching preset value", and the constraints in the nuclear power plant operating procedures (e.g., physical constraints: coolant pump must have power; safety regulations: coolant flow rate must be within safe limits; logical dependencies: flow rate can only be measured after coolant pump is started), several candidate operational steps to achieve this intermediate state transition are identified, such as "check coolant pump power supply", "start coolant pump", etc. The steps are: "start the coolant pump", "check the coolant flow meter reading", and "adjust the coolant flow rate". Finally, from the candidate operation steps, the steps necessary to achieve the intermediate state of "normal operation of the main coolant system and the flow rate reaching the preset value" and whose execution order is uniquely defined by the constraints are selected. For example, "starting the coolant pump" is necessary and must be performed before "checking the coolant flow meter reading", while "checking the coolant pump power supply" must be performed before "starting the coolant pump". Therefore, "checking the coolant pump power supply", "starting the coolant pump", and "checking the coolant flow meter reading" are determined as necessary steps and arranged in their uniquely defined order.

[0037] Through the above technical solutions, this application can more accurately establish the standard operating sequence of practical processes. Specifically, by precisely defining the necessary steps between key operation steps, it can avoid blind spots in standardization detection caused by ambiguous definitions of intermediate operations in complex processes, thereby improving the accuracy and reliability of standardization detection of practical processes. By associating operational behaviors with system state context and further considering physical, security, and logical constraints, it can ensure that the standard sequence not only covers the main operations but also includes all necessary intermediate links to ensure process safety and compliance. This allows for more effective identification of non-standard behaviors such as omissions, errors, or reversed sequences in actual operations.

[0038] In one embodiment, such as Figure 3 As shown, step S30 includes: S31: Identify the step type of the actual operation steps, and construct the comparison features of the actual operation steps based on the step type and the actual execution parameters. In this embodiment, identifying the step type of the actual operation step refers to the label or description that classifies the operation behavior, such as "open valve," "start pump," "input data," etc. Its function is to provide a classification basis for subsequent comparison, enabling the differentiation of different types of operations. The step type can be identified by matching keywords in the actual operation record using a preset rule base. For example, if the record contains words such as "open" or "start," it is identified as the "open" type. Alternatively, natural language processing technology can be used to perform semantic analysis on the operation description text, extracting operation verbs and objects to infer the step type. Based on the step type of the actual operation step and the actual operation... The actual execution parameters constitute the comparison features of the actual operation steps. These comparison features are unique or nearly unique identifiers used to describe an actual operation step. They combine the operation type and specific execution details to provide a structured data representation for effective similarity assessment with standard key operation steps. Specifically, the step type can be encoded as an enumeration value or string, and the actual execution parameters can be standardized and then concatenated into a feature vector or hash value. Alternatively, a multi-dimensional feature space can be constructed, where one dimension represents the step type and other dimensions represent various aspects of the actual execution parameters, such as the operation object, operation value, and operation result, forming a comprehensive feature descriptor.

[0039] S32: Identify the step types of key operation steps, and construct reference features for key operation steps based on the step types and expected execution parameters. In this embodiment, identifying the step type of key operation steps is similar to that of actual operation steps, and is a classification of standard operation behaviors to ensure comparability between standards and actual operations at the type level. The step type of key operation steps is usually predefined and marked when the standard operation sequence is established; or, it can be achieved by parsing the standard operation process document, extracting descriptive information of each key operation step, and mapping it to a predefined step type classification system; reference features refer to the structured description of standard key operation steps, used as a benchmark for comparison, and providing a clear standardized reference point for similarity assessment; specifically, the step type of key operation steps and expected execution parameters (such as preset values, ranges, target states, etc.) can be encoded and combined to form a structure compatible with the comparison features of actual operation steps; or, the same or similar feature construction methods as the comparison features of actual operation steps can be used to ensure the consistency of feature dimensions and semantics, for example, combining the step type and expected execution parameters into a reference feature vector.

[0040] S33: Perform similarity assessment on the reference features and the corresponding comparison features, and select key operation steps that meet the preset similarity conditions within the elastic operation range of the corresponding standard timestamps of the actual operation sequence, and use them as anchor operation steps.

[0041] In this embodiment, similarity assessment refers to the process of quantifying the degree of similarity between two features. It typically outputs a numerical value representing the level of similarity to determine the degree of matching between the actual operation steps and the standard key operation steps. Various algorithms, such as cosine similarity, the reciprocal of Euclidean distance, and Jaccard similarity, can be used to calculate the similarity between feature vectors. Alternatively, a weighted similarity function can be designed, assigning different weights based on the step type and the importance of different execution parameters to more accurately reflect the similarity in business logic. The flexible operation interval refers to the allowed upward and downward fluctuation range of the standard timestamp. The preset similarity condition is the threshold for judging a successful match, aiming to comprehensively consider both time tolerance and feature matching degree to improve the identification of anchored operation steps. To ensure accuracy, specifically, a time window can be determined based on the standard timestamps of key operation steps and a preset flexible range. Within this time window, the similarity between the comparison features of all actual operation steps and the reference features of the current key operation step is evaluated. Finally, actual operation steps with similarity scores higher than a preset threshold are marked as potential anchoring operation steps. Alternatively, a sliding time window approach can be used to search for actual operation steps that are temporally close to the key operation steps in the actual operation sequence. For each actual operation step that meets the temporal condition, its feature similarity with the key operation step is calculated and compared with a preset similarity threshold. Only actual operation steps that simultaneously meet the time constraint and the similarity threshold are confirmed as anchoring operation steps.

[0042] For example, suppose there is a standard operating procedure, one of the key operating steps of which is "open valve A to 50%". The expected execution parameters of this key operating step include "valve A" and "open to 50%", and its standard timestamp is T1, with a flexible operating range of ±5 seconds. When the actual operation record is collected in real time, the following situations may occur: Actual operation record 1: "The operator opens valve A halfway." At this point, semantic parsing can identify the step type as "open valve." Combining the step type "open valve," the actual execution parameter "valve A," and quantifying "opening halfway" as "opening to 50%," a comparison feature is formed. The key operation step's step type is "open valve." Combining the step type "open valve," the expected execution parameters "valve A" and "opening to 50%," a reference feature is formed. At this point, the comparison feature of actual operation record 1 and the reference feature of the key operation step are highly similar. If the actual timestamp of this operation falls within the elastic operation interval of T1, and the similarity assessment result is higher than the preset threshold, then this actual operation step can be identified as an anchored operation step.

[0043] Actual operation record 2: "Operator opens valve A, opening degree is 48%", where the step type is identified as "open valve". Combining the step type "open valve" with the actual execution parameters "valve A" and "opening degree is 48%", a comparison feature is formed. At this time, although the opening degree is slightly off, if the similarity evaluation algorithm can tolerate such a small range of numerical differences, and the actual timestamp of the operation is still within the elastic operation range, the similarity score may still meet the preset threshold. Thus, this actual operation step can also be identified as the anchor operation step.

[0044] Actual operation record 3: "Operator adjusts valve B". The step type is identified as "adjust valve". The step type "adjust valve" and the actual execution parameter "valve B" form a comparison feature. At this time, both the step type and the operation object do not match the key operation steps. The similarity evaluation result will be far below the preset threshold. Therefore, this actual operation step will not be identified as the anchor operation step.

[0045] In summary, this embodiment can flexibly and accurately identify the anchored operation steps corresponding to the standard key operation steps from the actual operation sequence, even if there are certain variations or ambiguities in the actual operation.

[0046] Through the above technical solution, this application can effectively solve the problem of identification difficulties caused by subtle differences between actual operation steps and standard key operation steps. Specifically, by constructing comparison features and reference features and combining them with similarity assessment, this solution can quantify the degree of matching between actual operation and standard operation, thereby tolerating non-critical deviations in operation details to a certain extent. At the same time, by introducing a flexible operation range of standard timestamps, the time matching also has a certain degree of fault tolerance, which can improve the accuracy of anchor operation step identification and avoid negative impacts on the standardization detection results due to misjudgment or omission of anchor operation steps, thereby improving the reliability and practicality of the standardization detection of practical operation processes.

[0047] In one embodiment, step S40 includes: S41: Parse the standard operation sequence, extract the logical dependencies between the starting key operation step and the ending key operation step corresponding to the current real timestamp interval, and construct the standard sequential directed graph of the real timestamp interval based on the logical dependencies. In this embodiment, logical dependencies can indicate that certain operations must be completed before other operations, or that the execution of certain operations is a prerequisite for other operations. Constructing a standard sequential directed graph is a key means of making the logical order and dependencies between critical operation steps explicit, which can intuitively represent the order and dependency paths between operation steps. For example, logical dependencies can be obtained through expert experience analysis, rule extraction based on historical regulatory documents, or by using machine learning models to learn and infer from a large amount of compliant operation data. When constructing a directed graph, each step that must be executed can be represented as a node, and the logical dependencies between these nodes are represented as directed edges.

[0048] S42: Within the real timestamp interval, generate an execution subsequence based on the execution order of the actual operation steps, and map the actual operation steps in the execution subsequence to the corresponding nodes of the standard sequential directed graph; In this embodiment, generating an execution subsequence means sorting the actual collected operation records according to their real timestamps to form an ordered chain of actual operation steps. Then, each actual operation step of the execution subsequence is matched and mapped with the corresponding node in the standard sequential directed graph. For example, the corresponding node of the actual operation step in the standard graph can be determined by comparing the step type, actual execution parameters and expected execution parameters of key operation steps.

[0049] S43: By comparing the sequential relationship of the actual operation steps in the actual execution sequence with the directed path relationship between the nodes of the actual operation steps in the standard sequential directed graph, all existing sequence conflicts are identified. In this embodiment, a sequence conflict refers to an order relationship in the actual execution sequence that is opposite to the dependency direction defined by the standard directed graph model. The identification of sequence conflicts can employ various algorithms. For example, each pair of adjacent or non-adjacent actual operation steps in the actual execution sub-sequence can be traversed to check if a directed path exists in the standard directed graph, and to determine if the actual execution order is consistent with the direction of that directed path. If actual operation step A is executed before B in the actual execution sequence, but a directed path from B to A exists in the standard directed graph, then a sequence conflict exists. Another approach is to use topological sorting or graph traversal algorithms to detect whether the actual execution sequence conforms to the topological order of the standard directed graph.

[0050] S44: Calculate the consistency metric value of the real timestamp interval based on the identified sequence conflict, and compare the consistency metric value with the preset consistency tolerance threshold to complete the logical sequence consistency assessment.

[0051] In this embodiment, the consistency metric is a value derived by comprehensively considering all identified sequence conflicts. It reflects the degree to which the actual operation sequence deviates from the standard logical order. The calculation method may include counting the number of conflicts, the severity of the conflicts (e.g., conflicts involving critical steps have higher weights), and the impact of the conflicts on subsequent operations. For example, different weights can be assigned to each type of sequence conflict, and then the weights of all conflicts can be summed to obtain a total conflict score, which can then be converted into a consistency metric through a predefined function. By comparing the consistency metric with a preset consistency tolerance threshold, it can be determined whether the current operation sequence meets the normative requirements. The consistency tolerance threshold can be dynamically adjusted according to factors such as the risk level of the process and the complexity of the operation.

[0052] For example, as a specific implementation, suppose that in a certain equipment maintenance process, between the two anchored operation steps of "turning off the power" and "disconnecting the connection cable," the standard operation sequence specifies three mandatory steps: "checking that the power indicator light is off," "pressing the emergency stop button," and "waiting for the equipment to completely stop." These mandatory steps have the following logical dependencies: checking that the power indicator light is off is necessary to confirm that the power is off before pressing the emergency stop button for safety; and waiting for the equipment to completely stop is a prerequisite for disconnecting the connection cable. First, the above logical dependencies are analyzed, and a standard sequence directed graph is constructed. In this graph, "checking that the power indicator light is off" is the predecessor node of "pressing the emergency stop button," and "pressing the emergency stop button" is the predecessor node of "waiting for the equipment to completely stop." In actual operation, if the operator presses the emergency stop button first after turning off the power... The process involves pressing the emergency stop button, checking if the power indicator light is off, and finally waiting for the device to completely stop. This generates an execution sub-sequence of actual operation steps, which is then mapped to corresponding nodes in a standard sequential directed graph. Further, the order of the actual execution sequence is compared with the directed path relationship in the standard sequential directed graph. In this example, pressing the emergency stop button actually occurs before checking if the power indicator light is off. However, in the standard directed graph, "checking if the power indicator light is off" is a predecessor to "pressing the emergency stop button," thus identifying a sequence conflict. Subsequently, a consistency metric is calculated based on the identified sequence conflict for the actual timestamp interval. For example, different weights can be assigned based on the severity of the conflict; conflicts involving safety-critical steps have higher weights. If the calculated consistency metric is lower than a preset consistency tolerance threshold, the operation process is deemed to have a logically non-standard sequence.

[0053] Through the above technical solutions, this application can make the logical dependencies of the steps that must be executed in the standard operation sequence explicit into a quantifiable directed graph of the standard sequence, thereby providing a benchmark model for the standardization evaluation of the actual operation sequence. Furthermore, this application can identify sequence conflicts in the actual operation sequence that do not conform to the standard logical order, rather than just missing judgment steps. In addition, by calculating the consistency metric and comparing it with the tolerance threshold, this application can quantitatively evaluate the consistency of the logical sequence of the operation process, thereby effectively solving the problem of the difficulty in accurately judging operation sequence errors, improving the accuracy and precision of the standardization detection of the actual operation process, and thus reducing the safety hazards and operational risks caused by operation sequence errors.

[0054] In one embodiment, after step S41, the method further includes: S411: Identify the cyclic execution logic for the steps that must be executed. When a cyclic execution step is identified, construct a cyclic structure in the standard sequential directed graph, including the cyclic entry node, the sequence of cyclic body nodes, and the cyclic closing node, and define the range of cyclic iterations and the cyclic termination condition. In this embodiment, step S411 aims to identify the sequence of steps repeatedly executed in the standard operation process and convert it into a verifiable loop structure. For example, the loop pattern can be identified by syntactic analysis of the standard operation sequence or by pattern matching based on an expert knowledge base. The loop structure can be represented as a specific set of nodes in a graph, where the loop entry node indicates the start of the loop, the loop body node sequence indicates the repeated operations inside the loop, and the loop closing node indicates the end of the loop. The range of loop iterations can be preset to the minimum and maximum allowed number of iterations, while the loop termination condition can be a specific state of an operation object or a parameter reaching a preset value.

[0055] S412: Perform branch selection logic identification for necessary execution steps. When a branch selection execution step is identified, construct a branch structure in the standard sequential directed graph, including branch judgment nodes and several branch path nodes, and define logical preconditions for each branch path. In this embodiment, step S412 aims to identify the steps in the standard operation process that select different execution paths based on specific conditions; wherein, the branch selection logic can be identified by analyzing the condition judgment statements in the standard operation sequence or based on predefined decision rules; the branch structure in the diagram can be derived from a branch judgment node to multiple branch path nodes, each branch path node representing a possible execution path; the logical precondition is the condition that triggers a specific branch path, such as a sensor reading, user input, or system status.

[0056] If the standard ordered directed graph contains a cycle structure or a branch structure, then after step S44, the following steps are also included: S4401: Perform loop compliance verification on the execution subsequence of the actual operation steps mapped into the loop structure, specifically including: identifying each complete iteration of the loop body node sequence and counting the actual number of iterations, determining whether it is within the loop iteration count range, and determining whether the loop termination condition is met based on the actual execution parameters of the relevant actual operation steps after the last iteration. In this embodiment, step S4401 aims to confirm whether the execution of the loop in actual operation conforms to the standard; specifically, it identifies the repetitive execution mode corresponding to the loop body node sequence in the actual operation sequence, counts the actual number of iterations, compares the actual number of iterations with the predefined range of loop iterations to determine whether it is within the allowed range; in addition, it also determines whether the loop ends correctly according to the preset loop termination condition based on the actual execution parameters of the relevant actual operation steps after the last iteration.

[0057] S4402: Perform path conformity verification on the execution subsequence of the actual operation steps mapped into the branch structure. Specifically, this includes: extracting the actual execution parameters of the actual operation steps corresponding to the branch judgment node, inferring the logical preconditions that are satisfied based on the actual execution parameters, thereby determining the branch path, and verifying the consistency between the node sequence associated with the execution subsequence of the actual operation steps and the branch path. In this embodiment, step S4402 aims to confirm whether the selection of the branch path in the actual operation is correct. Specifically, the actual execution parameters of the actual operation steps corresponding to the branch judgment node are extracted, such as a certain measurement value or user selection. Based on the extracted actual execution parameters, the logical preconditions satisfied by the actual operation are inferred, thereby determining the actual execution branch path. Finally, the node sequence associated with the execution sub-sequence of the actual operation steps is compared and verified with the branch path defined in the standard to determine whether the actual selected path matches the conditions.

[0058] S4403: Update the logical order consistency assessment of the real timestamp interval based on the results of the cyclic compliance verification and the path compliance verification.

[0059] In this embodiment, step S4403 aims to integrate the verification results for complex control flow into the overall logical order consistency evaluation. Specifically, if loop or branch verification finds non-compliant situations, such as loop count exceeding the range, loop not terminating correctly, or branch path selection error, these non-compliant items are treated as important conflicts or deviations and used to adjust or correct the original logical order consistency metric, thereby providing a more accurate compliance evaluation.

[0060] Furthermore, loop compliance verification refers to checking whether the actual number of iterations is within the preset range and determining whether the parameters after the last iteration meet the loop termination condition; branch compliance verification refers to determining the path to be triggered based on the actual execution parameters of the branch judgment node, comparing the actual execution node sequence with the standard branch path, and determining the branch violation if they are inconsistent.

[0061] For example, suppose there is a practical procedure for device startup, which includes a self-test loop and a fault handling branch. When constructing the standard operation sequence, the "self-test" is first identified as a loop execution step, with its loop body node sequence including "check power supply," "check connections," and "run diagnostic program." The loop iteration count is defined as 2-5 times, and the loop termination condition is "normal diagnostic result." Simultaneously, a branch decision node is identified after "run diagnostic program," with the logical preconditions being "normal diagnostic result" or "abnormal diagnostic result." If the diagnostic result is normal, the branch path "start the main system" is entered; if the diagnostic result is abnormal, the branch path "perform fault troubleshooting" is entered. When the actual operator executes this device startup procedure, their operation record is collected in real time. For example, the actual operation sequence shows "check power supply," "check connections," and "run diagnostic program." The steps "Power On", "Check Connection", and "Run Diagnostic Program" were executed six times, followed by "Start Main System". During loop compliance verification, it was found that the actual number of iterations exceeded the preset loop iteration range, thus the loop compliance verification failed. For example, after the actual operation sequence showed "Run Diagnostic Program" with an abnormal diagnosis, the operator executed "Start Main System". During path compliance verification, the actual execution parameters corresponding to "Run Diagnostic Program" (abnormal diagnosis result) were extracted, and the logical precondition of "abnormal diagnosis result" should be met. Therefore, the operator should enter the "Execute Troubleshooting" branch path. However, the actual operation selected the "Start Main System" branch path, thus the path compliance verification failed. Finally, the above verification results are used to update the overall logical sequence consistency assessment.

[0062] Through the above technical solutions, loop and branch structures are explicitly constructed in the standard sequential directed graph, and the corresponding iteration range, termination conditions, and logical preconditions are defined, thereby enabling a deeper understanding of the dynamic execution logic of the standard process. In actual operation, targeted loop compliance verification and path compliance verification can effectively identify normative defects such as non-compliance with loop count requirements, loops not terminating as expected, or incorrect branch path selection. This can improve the comprehensiveness and accuracy of logical sequence consistency assessment, avoid misjudgments or omissions caused by improper handling of complex logical structures, and thus provide a more reliable basis for determining the standardization of practical processes.

[0063] In one embodiment, step S44 includes: S441: Identify the nodes in the standard ordered directed graph corresponding to the actual operation steps involved in the sequence conflict, calculate the shortest reverse path depth of the identified node in the standard ordered directed graph, and set the basic conflict weight for the sequence conflict based on the shortest reverse path depth and the preset depth influence coefficient. In this embodiment, a sequential conflict refers to an execution order of actual operation steps that is opposite to the logical dependency direction defined by the standard sequential directed graph. Using graph traversal algorithms, such as breadth-first search or depth-first search, a reverse search can be performed in the standard sequential directed graph to identify the two nodes involved in the conflict and calculate the shortest path length from the later-executed node to the earlier-executed node in the standard graph. Alternatively, the shortest paths between all node pairs in the standard sequential directed graph can be pre-calculated and stored for direct querying when a conflict occurs. The basic conflict weight can be defined as the product of the shortest reverse path depth and the depth influence coefficient, or calculated using a preset function. For example, the weight can be set to be inversely proportional to the depth, meaning that the smaller the depth (the closer the conflict is to the critical path), the higher the conflict weight. The depth influence coefficient is a configurable parameter, and the influence of depth on the weight can be adjusted according to actual needs, for example, using a linear, exponential, or piecewise function.

[0064] Furthermore, in this embodiment, the step of setting the basic conflict weight for the sequence conflict based on the shortest reverse path depth and a preset depth influence coefficient specifically includes the following calculation logic: Let the shortest reverse path depth of the nodes corresponding to the two actual operation steps involved in a certain identified sequence conflict in the standard ordered directed graph be... ( ≥1); Since the smaller the depth of the shortest reverse path, the closer the two operation steps are related in the standard process, such as a direct pre- or post-processing relationship, the higher the risk of sequence conflict. Therefore, an inverse decay function can be used to calculate the basic conflict weight: ,in, This is a preset depth influence coefficient, for example, it can be set to 10; This is the decay control constant, used to control the rate at which the weight decreases with distance; for example, it can take a value of 1 or 2. For instance, if the operator reverses two steps with a direct dependency, then... =1, then =10 / 1 1 =10; If the steps that are far apart are reversed, then it is possible that... =3, then =10 / 3≈3.333.

[0065] S442: Based on the step type of the actual operation steps involved in the sequential conflict and the relative temporal position of the sequential conflict within the real timestamp interval, the basic conflict weight is dynamically adjusted to generate the final conflict impact value of the sequential conflict. In this embodiment, the final conflict impact value refers to a further refinement of the basic conflict weights to more accurately reflect the actual impact. Different step types (e.g., safety-related steps, critical decision steps) and relative temporal positions (e.g., early, middle, and late stages of the process) can affect the severity of the conflict. Adjustment factors can be preset for different step types. For example, the adjustment factor can be higher for conflicts involving emergency stops or critical parameter settings. At the same time, conflicts in the early stage of the process may have a greater chain reaction than conflicts in the late stage of the process. Therefore, different adjustment factors can be applied according to the relative temporal positions.

[0066] Furthermore, when generating the final conflict impact value of sequence conflicts, a step type adjustment factor can be introduced. and relative time position adjustment factor The basic conflict weights are dynamically adjusted using the following formula: In practice, different types of steps can be assigned values ​​using a pre-defined data dictionary. For example, safety-related procedures (such as disconnecting the power). =2.0; Key Decision Steps =1.5; Common auxiliary steps =1.0; Similarly, based on the relative temporal position, the actual timestamp interval is divided into early, middle, and late stages. Considering that the chain reaction caused by operational errors in the early stage is greater, the early stage can be set to 1.0. =1.2, =1.0, final stage =0.8; Adjust the factor through the above steps. and relative time position adjustment factor It can convert the severity of a conflict into a quantitative value that a computer can directly process, thereby improving the accuracy of the assessment.

[0067] S443: Integrate the final conflict impact value of all sequential conflicts within the real timestamp interval, and combine it with the actual operation steps within the real timestamp interval and the number of steps that must be executed, and calculate and generate a consistency metric value through a predefined consistency metric synthesis algorithm; In this embodiment, the consistency metric refers to a comprehensive indicator that quantifies the operational compliance across the entire real timestamp interval. The consistency metric synthesis algorithm can be a weighted summation model that accumulates all conflict impact values ​​and imposes additional penalties on omitted mandatory steps. Furthermore, the consistency metric synthesis algorithm needs to ensure that omitting mandatory steps will have a multiplicative negative impact on the metric value. This is a specific requirement for the consistency metric synthesis algorithm, emphasizing the importance of mandatory steps. Omitting mandatory steps usually indicates a serious deviation from the process, which may lead to security risks or process failure. Therefore, a stronger penalty than for general sequential conflicts is required. For example, the penalty factor for omitting mandatory steps can be set to a coefficient much larger than that of other conflict impact values, or an exponential penalty mechanism can be used.

[0068] Furthermore, a predefined consistency metric synthesis algorithm is used to calculate the degree of non-standardization of operations across the entire real timestamp interval, and the metric value is... A higher value indicates a more severe inconsistency. This consistency measure synthesis algorithm is implemented through weighted summation combined with an exponential penalty mechanism for missed required steps. The specific formula is as follows: ,in, This indicates all data within the given timestamp interval. The sum of the final conflict impact values ​​of each sequential conflict; This indicates the number of required steps that were missed within the given timestamp interval; The base for the fundamental penalty can be much larger than the conventional one. The value; To multiply the penalty coefficient, and ≥2, the above calculation formula can ensure at the algorithm level the accurate consideration of the exponential negative impact of missing steps on the overall evaluation.

[0069] S444: Compare the consistency metric value with a preset consistency tolerance threshold, which is adaptively determined based on the complexity of the operational context transformation between the starting critical operation step and the ending critical operation step corresponding to the current real timestamp interval.

[0070] In this embodiment, by comparing the calculated consistency metric with an acceptable consistency tolerance threshold, a conclusion of whether the operation is compliant or non-compliant can be drawn. Furthermore, multiple thresholds can be set, corresponding to different levels of compliance, thus providing more granular evaluation results. The consistency tolerance threshold is adaptively determined based on the complexity of the operational context transition between the starting and ending critical operation steps corresponding to the current real timestamp interval. This adaptability is to improve the accuracy and flexibility of the evaluation. Since different operation stages or context transitions may have different complexities and risks, different tolerance levels are required. For example, the tolerance threshold should be more stringent for intervals involving high-risk or complex equipment operations. A context transition complexity evaluation model can be pre-constructed. This model can consider factors such as the number of state variables in the operational context of the starting and ending critical operation steps, the dependencies between state variables, and the range of changes these state variables may undergo during the transition process. It calculates a complexity score and maps it to the corresponding tolerance threshold.

[0071] Furthermore, to avoid rigid, one-size-fits-all judgments, the consistency tolerance threshold is adaptively determined based on the complexity of operational context transformation. The steps can be implemented using the following algorithm: First, calculate the complexity score of the operation context transformation. : ,in, This refers to the number of state variables contained in the operational context associated with the initial critical operation step and the final critical operation step. This refers to the score of the correlation and dependency between these state variables; This refers to the range of changes that may occur during a state transition; , , The corresponding preset weights are used; further, an adaptive tolerance threshold is calculated based on the complexity score: ,in, This refers to the basic maximum tolerance threshold, which can be set to, for example, 20. This is a sensitivity coefficient, specifically a score for the complexity of contextual shifts. The higher the threshold, the greater the operational risk, and the more consistent the tolerance threshold for sequence conflicts or deviations. It should be as low as possible.

[0072] It should be noted that the preset weights, basic maximum tolerance thresholds, and sensitivity coefficients in the above examples can all be set by those skilled in the art based on actual needs and historical experience, and need not be elaborated again.

[0073] For example, suppose in a practical process of equipment maintenance, within a certain real timestamp interval, the standard operation sequence requires first turning off the main power supply, which is a mandatory step, then disconnecting the connection cable, and finally checking the equipment status. Therefore, the corresponding standard sequence directed graph shows that "turning off the main power supply" is a prerequisite for "disconnecting the connection cable," and "disconnecting the connection cable" is a prerequisite for "checking the equipment status." In actual operation, if the operator mistakenly disconnects the connection cable first and then turns off the main power supply, a sequence conflict will occur. In this case, the corresponding steps of disconnecting the connection cable and turning off the main power supply in the standard sequence directed graph are identified. For a given node, assuming there's a directed edge from "Power Off" to "Disconnect Line" in a standard sequential directed graph, but in actual operation, disconnecting the connection line is executed before power off, the reverse path is from "Disconnect Line" to "Power Off". If the depth of this reverse path is 1, meaning there's a direct dependency, then the shortest reverse path depth is calculated to be 1. Based on a preset depth influence coefficient, a basic conflict weight is set for this conflict. Furthermore, considering that power off is a safety-related step type, and this conflict occurs early in the process, a higher adjustment factor can be set for this step type and its timing position. The final conflict impact value will be dynamically increased. Additionally, if other sequential conflicts exist within the actual timestamp interval (e.g., checking device status before disconnecting the cable), their final conflict impact values ​​will be calculated similarly. Furthermore, if the operator completely omits the mandatory step of shutting down the main power, the consistency metric synthesis algorithm will impose a multiplicative negative impact. For example, if the penalty factor for omitting a mandatory step is 100, the final consistency metric value will increase significantly. Finally, the final impact values ​​of all conflicts are integrated, combined with the actual number of operational steps and the number of mandatory steps, to determine the final consistency metric value. A predefined consistency metric synthesis algorithm calculates the consistency metric value for the real timestamp interval. Simultaneously, based on the contextual transformation complexity of the starting and ending critical operation steps corresponding to the real timestamp interval, an adaptive consistency tolerance threshold is determined. For example, if the real timestamp interval involves high-voltage circuit operations, the contextual transformation complexity is high, and the tolerance threshold may be set to a lower value; conversely, if it is simply a parameter check, the contextual transformation complexity is low, and the tolerance threshold may be higher. Finally, the calculated consistency metric value is compared with the adaptive tolerance threshold to determine whether the operation interval conforms to the specification.

[0074] Through the above technical solution, this application can overcome the limitations of traditional methods that simply identify conflicts but cannot distinguish their severity. It can dynamically adjust the influence weight of conflicts based on various factors such as the depth of sequential conflicts, the type of steps involved, and the timing of occurrence, making the evaluation results more objective and accurate. In particular, by ensuring that omitting a necessary step will have a multiplicative negative impact on the metric, this application can improve the sensitivity and penalty for omitting key operations, thereby effectively avoiding insufficient normative evaluation due to the absence of key steps. In addition, by adaptively determining the consistency tolerance threshold, the evaluation criteria can be adjusted according to the complexity of contextual shifts and risk levels at different operational stages, avoiding rigid judgments that are generalized, thereby improving the practicality and reliability of normative detection.

[0075] In one embodiment, step S50 includes: S51: Based on the preset context mapping rule base, the actual execution parameters of the interval operation steps are parsed and matched to determine the state set of the operation object corresponding to the interval operation steps, and the state set is defined as the actual operation context derived in reverse. In this embodiment, the context mapping rule base refers to a pre-established knowledge base containing rules that map the execution parameters of various actual operation steps to specific states of the operation object. These rules can be simple key-value pair mappings, conditional logic expressions, state machine transition rules, or even mapping relationships trained based on machine learning models. For example, the rule base can define "when the valve opening is 100% and the pressure sensor reading exceeds the threshold, the valve is fully open and the system pressure is normal." Its function is to provide an interpretable mechanism, transforming discrete or continuous actual execution parameters into semantically meaningful operational context descriptions. Parsing and matching the actual execution parameters of interval operation steps refers to obtaining detailed execution data of the actual operation steps occurring within a specific time interval. For example, operation commands, equipment feedback, and sensor data are parsed and matched. The parsing process may involve data cleaning, format conversion, and feature extraction. Matching compares the parsed parameters with rules in the context mapping rule base to find the context definition that best matches the current parameter state. Through the above parsing and matching process, the specific state of the relevant operation object when performing the operation steps in this range can be identified. At this time, the state set may include multiple dimensions, such as the on / off state of the equipment, parameter settings, ambient temperature, pressure, etc. Finally, the identified state set of operation objects is integrated to form a description reflecting the current actual operating environment and object state, which is the reverse-derived actual operating context. This actual operating context is derived from actual observation data rather than preset standards.

[0076] Furthermore, the operational context is a multi-factor state set including equipment operation status factors, material status factors, human-machine interaction status factors, and safety lock status factors. The context mapping rule base pre-stores the mapping rules and coupling constraints between execution parameters and each status factor. The coupling constraints are logical conditions that must be satisfied simultaneously between different factors. During reverse derivation, the actual execution parameters are parsed into the actual status values ​​of each factor according to the context mapping rule base, and then the coupling relationships between each factor are checked to form a complete actual operational context.

[0077] S52: Based on the standard operation sequence, obtain the expected operation context associated with the key operation steps corresponding to the operation steps in the current interval; In this embodiment, the standard operation sequence refers to a predefined standardized operation process, which includes the expected execution parameters of each key operation step and its associated expected operation context. After identifying the interval operation steps in the actual operation sequence, the expected operation context that the key operation step should have when it is executed is retrieved from the standard operation sequence according to its correspondence with the key operation steps in the standard operation sequence. The expected operation context represents the ideal state that the operation object should be in when the operation step is executed under the standard process. For example, if the interval operation step corresponds to "start the cooling pump" in the standard sequence, then the expected operation context may include "coolant level is normal" and "power is connected".

[0078] S53: Compare and verify the actual operation context with the expected operation context. The comparison and verification includes state completeness comparison, state value conformity comparison, and logical relationship consistency comparison.

[0079] In this embodiment, state completeness comparison refers to comparing whether the state elements contained in the actual operating context are consistent with all the state elements required in the expected operating context, that is, whether the actual context contains all the key state information required by the expected context; for example, if the expected context requires normal liquid level and stable temperature, but the actual context only detects normal liquid level, then the state completeness is insufficient; state value conformity comparison refers to comparing whether the specific values ​​of the state elements that coexist in the actual operating context and the expected operating context are within the allowable range or whether they conform to the expected values. For example, if the expected context requires normal liquid level and stable temperature, but the actual context only detects normal liquid level, then the state completeness is insufficient; The context requires "pressure between 1.0-1.2 MPa," but the actual context shows a pressure of 1.3 MPa, which is inconsistent with the expected value. This may involve comparison of numerical ranges, matching of enumerated values, etc. Logical relationship consistency comparison refers to checking whether the logical relationship between each state element in the actual operation context is consistent with the logical relationship defined in the expected operation context. For example, the expected context may define "when state A is true, state B must be false." If A is true and B is also true in the actual context, the logical relationship is inconsistent. This helps to discover more complex contextual errors, rather than just deviations in a single state.

[0080] For example, suppose that in the operation process of a chemical reactor, a certain operational step involves adding a catalyst. The actual execution parameters for this operational step might include the temperature sensor readings inside the reactor, the agitator speed, the status of the feed valve, and the catalyst batch information input by the operator through the human-machine interface. A preset context mapping rule base might contain rules such as: "If the temperature is between 40-50℃, the status is suitable reaction temperature; if the agitator speed is between 100-120 rpm, the status is suitable stirring speed; if the feed valve is open, the status is feeding." Through these rules, the actual execution parameters can be used to deduce the actual operational context, such as "Temperature: 45℃, Stirring speed: 110 rpm, Feed valve: Open." Simultaneously, in the standard operation sequence, this step involves adding a catalyst. The expected operating context for the catalyst addition step might be defined as "Temperature: 40-50℃, stirring speed: 100-120rpm, feed valve: closed". During the comparative verification, the state completeness comparison specifically checks whether the actual context includes all state items in the expected context; the state value conformity comparison specifically checks whether the actual temperature of 45℃ is within the expected range of 40-50℃, and whether the actual stirring speed of 110rpm is within the expected range of 100-120rpm; while the logical consistency comparison specifically identifies logical conflicts between "feed valve: open" in the actual context and "feed valve: closed" in the expected context. Through the above comparisons, even if the operator adds the catalyst on time, the potential risk of operating with the feed valve not closed can be accurately identified.

[0081] Through the above technical solution, this application can perform contextual assessment on each operational step in the practical process, thereby improving the accuracy and comprehensiveness of practical standardization detection. Specifically, this solution can not only detect deviations in the operation sequence and logical order, but also reveal subtle discrepancies in the environment and state during operation, effectively identifying potential operational risks or standardization defects caused by contextual deviations, thus making the conclusions of practical standardization judgment more convincing.

[0082] In one embodiment, after step S60, the method further includes: S70: Based on the identification results of anchored operation steps and interval operation steps, the evaluation results of logical sequence consistency, and the comparison and verification results, locate the source of normative defects. In this embodiment, locating the source of the non-standard operation defect refers to finding the specific link or operation point that leads to the non-standard operation. Its function is to concretize the non-standard judgment to one or more specific steps or states in the operation sequence. This can be achieved by backtracking analysis on the identification results of the anchor operation steps and interval operation steps on which the practical standard operation judgment conclusion depends, the evaluation results of the logical sequence consistency, and the comparison and verification results. For example, a decision tree or rule set can be constructed, and the results can be traced down layer by layer according to the type of judgment conclusion and the specific values ​​or states of each evaluation indicator until the smallest unit that leads to the non-standard operation is located, such as the absence of a certain anchor operation step, the execution error of a certain interval operation step, the logical sequence conflict within a certain time period, or the deviation of a certain operation context.

[0083] S80: Based on the source of the normative defect, identify its corresponding violation type and violation context. The violation type includes missing key operation steps, missing mandatory steps, logical sequence errors, and inconsistencies in operation context. In this embodiment, identifying the corresponding violation type and its violation context refers to classifying and contextualizing the identified defect sources. This serves to standardize complex defect information and provide sufficient information for subsequent rectification. Violation type refers to the categorization of the nature of the defect. For example, "missing critical operation steps" indicates that an important anchoring operation step was not executed; "missing mandatory steps" indicates that a necessary operation step within a specific timeframe was omitted; "incorrect logical order" indicates that the execution order of operation steps does not conform to the standard; and "inconsistent operation context" indicates that the environment or state during operation is inconsistent with expectations. Violation context provides specific details of when the defect occurred, such as which critical operation step was missing, which mandatory step was executed in the wrong order, or which operation context was deviated from.

[0084] S90: Based on a pre-built violation guidance mapping knowledge base, it matches corresponding standard rectification operation guidelines for violation types and their violation contexts.

[0085] In this embodiment, based on a pre-built violation guidance mapping knowledge base, corresponding standard rectification operation guidelines are matched for violation types and their violation contexts. This aims to provide specific corrective measures for identified violations, transforming diagnostic results into actionable plans to form a closed-loop feedback loop. The pre-built violation guidance mapping knowledge base can be a database containing a large number of rules or cases, storing the mapping relationships between various violation types and violation context combinations and corresponding standard rectification operation guidelines. For example, when a "missing key operation step" is identified and the context is "starting the device," the violation guidance mapping knowledge base can match the guidance "Please check the device power connection and re-perform the startup operation." The violation guidance mapping knowledge base can be built based on domain expert experience or can be learned and optimized through historical operation data and rectification records.

[0086] For example, suppose that during a practical equipment maintenance operation, a compliance judgment conclusion is generated according to the method of the aforementioned embodiment, indicating that the operation is not in compliance. As a specific implementation method, the source of the compliance defect is first located. For example, by analyzing the evaluation results of logical sequence consistency, it is found that within a certain real timestamp interval, that is, between the anchor operation step "turn off the main power" and the anchor operation step "disconnect the connection line", the execution order of the step "check that the power indicator light is off" in the actual operation sequence does not conform to the standard. At this time, the corresponding violation type and its violation context are identified. For example, the violation type is identified as a logical sequence error, and its violation context is... The description states: "After turning off the main power, you should first check that the power indicator light is off before disconnecting the connection cable; however, in actual operation, users disconnect the connection cable before checking that the power indicator light is off." Based on a pre-built violation guidance mapping knowledge base, corresponding standard rectification operation guidelines are matched for violation types and their violation contexts. For example, there may be a rule in the violation guidance mapping knowledge base: "If the violation type is a logical sequence error and the context involves disconnecting the connection cable before checking the power indicator light, it is recommended to strictly follow the operating procedures and ensure that the device is completely powered off by checking the power indicator light before disconnecting any connection cable to avoid the risk of electric shock." This guideline is then provided to operators.

[0087] Through the above technical solution, this application can transform the conclusions of practical operation standardization judgment into actionable feedback information, enabling operators not only to understand whether the operation is standardized, but also to clearly understand the specific reasons for non-standard operation, the location of the non-standard operation, and the corresponding corrective measures. Through this diagnostic and guidance capability, the practical value of practical operation standardization testing can be enhanced, helping operators to quickly locate and correct errors, thereby effectively reducing operational risks, improving operational efficiency and quality, and promoting continuous improvement of operational skills and processes.

[0088] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0089] In one embodiment, a practical process standardization detection system based on operation sequence matching is provided. This system corresponds one-to-one with the practical process standardization detection method based on operation sequence matching described in the previous embodiment. The practical process standardization detection system based on operation sequence matching includes: The standard operation sequence establishment module is used to determine the target practical process, establish the standard operation sequence corresponding to the practical process, and determine the steps that must be executed. The standard operation sequence is a sequence including several key operation steps and their expected execution parameters. The key operation steps are associated with a standard timestamp that allows for flexible operation intervals and an operation context used to indicate the operation status. The actual operation sequence generation module is used to collect operation records in real time and generate actual operation sequences. The actual operation sequence is a sequence including several actual operation steps and their corresponding actual execution parameters. The actual operation steps are associated with real timestamps. The anchoring operation step determination module is used to identify the step type of the actual operation step, and based on the combination characteristics of the step type and the actual execution parameters, identify the actual operation steps corresponding to each key operation step as anchoring operation steps. The interval operation step determination module is used to identify whether the actual operation sequence contains the actual operation steps corresponding to the mandatory steps of the standard operation sequence within the real timestamp interval corresponding to two consecutive anchor operation steps, and to evaluate the consistency of the logical order of all actual operation steps within the real timestamp interval with the standard operation sequence. The comparison and verification module is used to reverse-engineer the actual operation context from the actual execution parameters of the interval operation steps within the real timestamp interval, and compare and verify the reverse-engineered actual operation context with the operation context associated with the corresponding key operation steps in the standard operation sequence. The standardization judgment module is used to generate a practical standardization judgment conclusion based on the identification results of anchored operation steps and interval operation steps, the evaluation results of logical sequence consistency, and the comparison and verification results.

[0090] Optional, a standard operation sequence establishment module, including: The intermediate state determination submodule is used to determine several key operation step pairs based on the standard operation sequence, and to determine at least one intermediate state that must be satisfied in the operation context of the key operation step pair when the operation context of the previous key operation step is transformed into the operation context of the next key operation step, according to the operation context associated with each key operation step pair. The candidate step determination submodule is used to determine several candidate operation steps that can achieve the intermediate state transition based on the determined intermediate state and the constraints of the operation object corresponding to the actual operation process. The execution step filtering submodule is used to filter out candidate operation steps from the candidate operation steps that are necessary to achieve intermediate state transitions and whose execution order is uniquely limited by constraints, and these steps are the mandatory execution steps.

[0091] Specific limitations regarding the operational sequence matching-based practical process standardization detection system can be found in the above-described limitations of the operational sequence matching-based practical process standardization detection method, and will not be repeated here. Each module in the aforementioned operational sequence matching-based practical process standardization detection system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the computer device's memory as software, so that the processor can call and execute the operations corresponding to each module.

[0092] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for detecting the standardization of practical procedures based on operation sequence matching, characterized in that, Including the following steps: Determine the target operational process, establish the standard operation sequence corresponding to the operational process, and determine the steps that must be executed. The standard operation sequence is a sequence that includes several key operation steps and their expected execution parameters. The key operation steps are associated with a standard timestamp that allows for flexible operation intervals and an operation context used to indicate the operation status. Real-time collection of operation records and generation of actual operation sequences, wherein the actual operation sequence is a sequence including several actual operation steps and their corresponding actual execution parameters, and the actual operation steps are associated with real timestamps; Identify the step type of the actual operation steps, and based on the combination characteristics of the step type and the actual execution parameters, identify the actual operation steps corresponding to each key operation step as anchor operation steps. Within the real timestamp interval corresponding to two consecutive anchoring operation steps, identify whether the actual operation sequence contains the actual operation steps corresponding to the required execution steps of the standard operation sequence within the real timestamp interval, and use them as interval operation steps, and evaluate the consistency of the logical order of all actual operation steps within the real timestamp interval with the standard operation sequence. The actual operational context is derived from the actual execution parameters of the interval operation steps within the real timestamp interval, and the derived actual operational context is compared and verified with the operational context associated with the corresponding key operation steps in the standard operation sequence. Based on the identification results of anchored operation steps and interval operation steps, the evaluation results of logical sequence consistency, and the comparison and verification results, a conclusion on the standardization of practical operation is generated.

2. The method for detecting the standardization of practical procedures based on operation sequence matching according to claim 1, characterized in that: The process involves determining the target operational flow, establishing a standard operation sequence corresponding to the operational flow, and identifying the steps that must be executed. The standard operation sequence is a sequence including several key operational steps and their expected execution parameters. Each key operational step is associated with a standard timestamp that allows for flexible operational intervals and an operational context used to represent the operational state. This includes: Based on the standard operation sequence, several key operation step pairs are identified, and according to the operation context associated with each key operation step pair, at least one intermediate state must be satisfied in the operation context transition from the previous key operation step to the next key operation step in the key operation step pair. Based on the determined intermediate state and the constraints of the operation objects corresponding to the actual operation process, several candidate operation steps that can realize the intermediate state transition are determined. From the candidate operation steps, select the candidate operation steps that are necessary to achieve the intermediate state transition and whose execution order is uniquely determined by the constraints, and then select them as the mandatory operation steps.

3. The method for detecting the standardization of practical procedures based on operation sequence matching according to claim 1, characterized in that: The step type for identifying actual operation steps, and based on the combination characteristics of step type and actual execution parameters, identifying the actual operation steps corresponding to each key operation step as anchor operation steps, includes: Identify the step type of the actual operation steps, and construct the comparison features of the actual operation steps based on the step type and the actual execution parameters. Identify the step types of key operation steps, and construct reference features for key operation steps based on the step types and expected execution parameters. The similarity between the reference features and the corresponding comparison features is evaluated. Within the flexible operation range of the corresponding standard timestamp of the actual operation sequence, the key operation steps that meet the preset similarity conditions are selected as anchor operation steps.

4. The method for detecting the standardization of practical procedures based on operation sequence matching according to claim 1, characterized in that: The step of identifying whether the actual operation sequence contains the actual operation steps corresponding to the mandatory execution steps of the standard operation sequence within the real timestamp interval corresponding to two consecutive anchoring operation steps, and evaluating the logical order consistency between all actual operation steps within the real timestamp interval and the standard operation sequence, includes: The standard operation sequence is parsed to extract the logical dependencies between the starting critical operation step and the ending critical operation step corresponding to the current real timestamp interval, and a standard sequential directed graph of the real timestamp interval is constructed based on the logical dependencies. Within the real timestamp interval, an execution subsequence is generated according to the execution order of the actual operation steps, and the actual operation steps in the execution subsequence are mapped to the corresponding nodes of the standard sequential directed graph; By comparing the sequential relationship of the actual operation steps in the actual execution sequence with the directed path relationship between the nodes of the actual operation steps in the standard sequential directed graph, all existing sequence conflicts are identified. The consistency metric of the real timestamp interval is calculated based on the identified sequence conflicts, and the consistency metric is compared with the preset consistency tolerance threshold to complete the logical sequence consistency assessment.

5. The method for detecting the standardization of practical procedures based on operation sequence matching according to claim 4, characterized in that: After parsing the standard operation sequence, extracting the logical dependencies between the starting critical operation step and the ending critical operation step corresponding to the current real timestamp interval, and constructing a standard sequential directed graph of the real timestamp interval based on the logical dependencies, the method further includes: The loop execution logic is identified for the steps that must be executed. When a loop execution step is identified, a loop structure including the loop entry node, the loop body node sequence and the loop closing node is constructed in the standard sequential directed graph, and the loop iteration number range and the loop termination condition are defined. For steps that must be executed, branch selection logic is identified. When a branch selection execution step is identified, a branch structure including a branch judgment node and several branch path nodes is constructed in the standard sequential directed graph, and logical preconditions are defined for each branch path. If a cyclic or branching structure exists in the standard ordered directed graph, after the step of calculating the consistency metric value of the real timestamp interval based on the identified order conflicts and comparing the consistency metric value with a preset consistency tolerance threshold to complete the logical order consistency assessment, the method further includes: The execution subsequence of the actual operation steps mapped into the loop structure is verified for loop compliance. Specifically, this includes: identifying each complete iteration of the loop body node sequence and counting the actual number of iterations, determining whether it is within the loop iteration count range, and determining whether the loop termination condition is met based on the actual execution parameters of the relevant actual operation steps after the last iteration. The path conformity verification of the execution subsequence of the actual operation steps mapped into the branch structure is specifically included: extracting the actual execution parameters of the actual operation steps corresponding to the branch judgment node, inferring the logical preconditions satisfied based on the actual execution parameters, thereby determining the branch path, and verifying the consistency between the node sequence associated with the execution subsequence of the actual operation steps and the branch path. Based on the results of the cyclic compliance verification and the path compliance verification, update the logical order consistency assessment of the real timestamp interval.

6. The method for detecting the standardization of practical procedures based on operation sequence matching according to claim 4, characterized in that: The step of calculating the consistency metric value of the real timestamp interval based on the identified sequence conflicts, and comparing the consistency metric value with a preset consistency tolerance threshold to complete the logical sequence consistency assessment includes: The actual operational steps involved in identifying sequence conflicts correspond to the nodes in the standard ordered directed graph. The shortest reverse path depth of the identified node in the standard ordered directed graph is calculated, and the basic conflict weight is set for the sequence conflict based on the shortest reverse path depth and the preset depth influence coefficient. Based on the step type of the actual operation steps involved in the sequential conflict and the relative temporal position of the sequential conflict within the real timestamp interval, the basic conflict weight is dynamically adjusted to generate the final conflict impact value of the sequential conflict. The final conflict impact value of all sequential conflicts within the real timestamp interval is integrated, and combined with the actual operation steps and the number of steps that must be executed within the real timestamp interval, a consistency metric value is generated by calculating it through a predefined consistency metric synthesis algorithm. The consistency metric is compared with a preset consistency tolerance threshold, which is adaptively determined based on the complexity of the context transformation between the starting critical operation step and the ending critical operation step corresponding to the current real timestamp interval.

7. The method for detecting the standardization of practical procedures based on operation sequence matching according to claim 1, characterized in that: The step of reverse-engineering the actual operational context from the actual execution parameters of the interval operation steps within the real timestamp interval, and comparing and verifying the reverse-engineered actual operational context with the operational context associated with the corresponding key operational steps in the standard operation sequence, includes: Based on a pre-defined context mapping rule base, the actual execution parameters of the interval operation steps are parsed and matched to determine the state set of the operation object corresponding to the interval operation steps, and the state set is defined as the actual operation context derived in reverse. Based on the standard operation sequence, obtain the expected operation context associated with the key operation steps corresponding to the operation steps in the current interval; The actual operational context is compared and verified with the expected operational context. The comparison and verification includes state completeness comparison, state value conformity comparison, and logical relationship consistency comparison.

8. The method for detecting the standardization of practical procedures based on operation sequence matching according to claim 1, characterized in that: After the step of generating a practical standardization judgment conclusion based on the identification results of anchored operation steps and interval operation steps, the evaluation results of logical sequence consistency, and the comparison and verification results, it also includes: Based on the identification results of anchored operation steps and interval operation steps, the evaluation results of logical sequence consistency, and the comparison and verification results, the source of normative defects is located. Based on the source of the normative defect, identify its corresponding violation type and violation context. The violation type includes missing key operation steps, missing mandatory steps, logical sequence errors, and inconsistencies in operation context. Based on a pre-built knowledge base of violation guidance mapping, corresponding standard rectification operation guidelines are matched for violation types and their violation contexts.

9. A practical process standardization detection system based on operation sequence matching, characterized in that, include: The standard operation sequence establishment module is used to determine the target practical process, establish the standard operation sequence corresponding to the practical process, and determine the steps that must be executed. The standard operation sequence is a sequence including several key operation steps and their expected execution parameters. The key operation steps are associated with a standard timestamp that allows for flexible operation intervals and an operation context used to indicate the operation status. The actual operation sequence generation module is used to collect operation records in real time and generate actual operation sequences. The actual operation sequence is a sequence including several actual operation steps and their corresponding actual execution parameters. The actual operation steps are associated with real timestamps. The anchoring operation step determination module is used to identify the step type of the actual operation step, and based on the combination characteristics of the step type and the actual execution parameters, identify the actual operation steps corresponding to each key operation step as anchoring operation steps. The interval operation step determination module is used to identify whether the actual operation sequence contains the actual operation steps corresponding to the mandatory steps of the standard operation sequence within the real timestamp interval corresponding to two consecutive anchor operation steps, and to evaluate the consistency of the logical order of all actual operation steps within the real timestamp interval with the standard operation sequence. The comparison and verification module is used to reverse-engineer the actual operation context from the actual execution parameters of the interval operation steps within the real timestamp interval, and compare and verify the reverse-engineered actual operation context with the operation context associated with the corresponding key operation steps in the standard operation sequence. The standardization judgment module is used to generate a practical standardization judgment conclusion based on the identification results of anchored operation steps and interval operation steps, the evaluation results of logical sequence consistency, and the comparison and verification results.

10. The operational procedure standardization detection system based on operation sequence matching according to claim 9, characterized in that: The standard operation sequence establishment module includes: The intermediate state determination submodule is used to determine several key operation step pairs based on the standard operation sequence, and to determine at least one intermediate state that must be satisfied in the operation context of the key operation step pair when the operation context of the previous key operation step is transformed into the operation context of the next key operation step, according to the operation context associated with each key operation step pair. The candidate step determination submodule is used to determine several candidate operation steps that can achieve the intermediate state transition based on the determined intermediate state and the constraints of the operation object corresponding to the actual operation process. The execution step filtering submodule is used to filter out candidate operation steps from the candidate operation steps that are necessary to achieve intermediate state transitions and whose execution order is uniquely limited by constraints, and these steps are the mandatory execution steps.

Images