A dynamic permission control method and system for safe production operation

Abstract

A dynamic permit control method and system for safe production operations is disclosed. The method includes: responding to a special operation permit request initiated by an operation terminal; extracting the operation location coordinates and operation type identifier; identifying the associated process unit; and acquiring operational status time-series data and real-time environmental meteorological data; determining the target historical operating condition segment corresponding to the special operation permit request, and the evolution trend data of the target historical operating condition segment on the time axis; calculating abnormal evolution parameters of the associated process unit within a preset future time period based on the historical operation result tags, evolution trend data, and operational status time-series data corresponding to the target historical operating condition segment; performing spatiotemporal conflict analysis to obtain spatiotemporal conflict verification results; generating operation control instructions corresponding to the special operation permit request based on the spatiotemporal conflict verification results, and controlling the operation terminal to execute the operation control instructions. This improves the accuracy of safe production operation permits.

Description

Technical Field

[0001] This application relates to the field of safety permit control technology, specifically to a dynamic permit control method and system for safe production operations. Background Technology

[0002] In industrial production, especially in high-risk industries such as petrochemicals, natural gas, and fine chemicals, special operations (such as hot work, work at height, and confined space work) pose significant safety risks. To ensure the personal safety of workers and the stable operation of production systems, companies typically need to complete a series of risk assessments and work permit procedures before implementing special operations, ensuring that the work is carried out in a controllable and preventable environment.

[0003] Existing technologies have digitized traditional paper forms and leveraged industrial control systems and environmental monitoring systems to acquire real-time data, thus assisting in work permit approval decisions. For example, the system can set several safety thresholds based on environmental parameters of the work area (such as temperature, humidity, and toxic gas concentration) and equipment operating status parameters (such as pressure, temperature, and valve opening / closing status), and trigger an early warning mechanism when the parameters exceed the set range, thereby preventing or delaying the execution of special operations.

[0004] However, existing risk identification methods are mainly based on static data at the current point in time or single threshold rules. In actual production environments, equipment status usually has certain temporal correlations and evolutionary patterns, and the diffusion range after a hazardous medium leak is greatly affected by the real-time meteorological environment. When facing complex and cross-influenced working conditions, the risk assessment is prone to inaccurate, affecting the accuracy of safe production operation permits. Summary of the Invention

[0005] This application provides a dynamic permitting and control method and system for safe production operations, which addresses the technical problem of inaccurate risk assessment in complex and interconnected work scenarios, thereby improving the accuracy of safe production operation permits.

[0006] The first aspect of this application provides a dynamic permit control method for safe production operations, the method comprising: In response to a special operation permit request initiated by the work terminal, the work location coordinates and work type identifier in the special operation permit request are extracted. Based on a preset spatial index relationship table, the associated process units within a preset radius centered on the work location coordinates are determined, and the time-series data of the operating status of the associated process units at the current moment, as well as the real-time environmental meteorological data of the area where the work location coordinates are located are obtained. The time-series data of the operating status is matched with the pre-stored historical operation and maintenance database using a sliding window similarity matching method to obtain the target historical operating condition segment corresponding to the special operation permit request and the evolution trend data of the target historical operating condition segment on the time axis. The historical operation and maintenance database contains historical operation result tags corresponding to each historical operating condition segment. Based on the historical operation result label, the evolution trend data and the operation status time series data corresponding to the target historical operating condition segment, calculate the abnormal evolution parameters of the associated process unit within a future preset time period; Based on the real-time environmental meteorological data, the abnormal evolution parameters, the operation location coordinates, and the operation type identifier, a spatiotemporal conflict analysis is performed to obtain the spatiotemporal conflict verification results. Based on the spatiotemporal conflict verification results, the operation control instructions corresponding to the special operation permit request are generated, and the operation terminal is controlled to execute the operation control instructions.

[0007] Optionally, the time-series data of the operational status is matched with a pre-stored historical operation and maintenance database using a sliding window similarity matching method to obtain the target historical operating condition segment corresponding to the special operation permit request and the evolution trend data of the target historical operating condition segment on the time axis, specifically including: Discrete difference processing is performed on the time series data of the running state to obtain the real-time morphological change rate sequence. Based on multiple different preset time scaling factors, the real-time morphological change rate sequence is resampled along the time axis to obtain the candidate feature sequence corresponding to each preset time scaling factor. Each of the candidate feature sequences to be tested is dynamically path-normalized and matched in multiple historical operating condition records in the pre-stored historical operation and maintenance database. For the first historical operating condition record, the cumulative alignment distance between each of the candidate feature sequences to be tested and the first historical operating condition record is calculated. The first historical operating condition record is any historical operating condition record. The historical working condition record corresponding to the minimum cumulative alignment distance is determined as the target historical working condition segment; The evolution trend data is determined based on the target historical operating condition segments.

[0008] Optionally, determining the evolution trend data based on the target historical operating condition segment specifically includes: The original operating condition data within a preset prediction time window after the last cutoff time of the target historical operating condition segment are obtained from the pre-stored historical operation and maintenance database. The start time of the preset prediction time window is the same as the last cutoff time of the target historical operating condition segment. The original operating condition data is mapped inversely to a time scale based on a target preset time scaling factor to obtain the evolution trend data. The target preset time scaling factor is the preset time scaling factor used when determining the target historical operating condition segment.

[0009] Optionally, based on the historical operation result tags corresponding to the target historical operating condition segment, the evolution trend data, and the operating status time series data, the abnormal evolution parameters of the associated process unit within a preset future time period are calculated, specifically including: Based on the anomaly type in the historical operation result label, the parameter change pattern corresponding to the anomaly type is determined from the target historical operating condition segment and the evolution trend data, and an anomaly precursor sequence is generated. Based on the target preset time scaling factor, the time scale of the preceding symptom sequence is adjusted to obtain the target search template, and the target search template is used for sliding similarity matching in the running state time series data. The moment when the feature similarity is greater than the preset similarity threshold is determined as the potential abnormal starting point of the current working condition. Using each potential anomaly initiation point as a reference time, the current anomaly verification window is determined in the running state time series data, and the real-time fluctuation intensity within the current anomaly verification window is calculated. The ratio of the real-time fluctuation intensity to the historical fluctuation intensity of the abnormal precursor sequence in the corresponding historical period is calculated to obtain the intensity correction coefficient; The amplitude of the evolution trend data is corrected based on the intensity correction coefficient to obtain the abnormal evolution parameters.

[0010] Optionally, based on the real-time environmental meteorological data, the anomaly evolution parameters, the work location coordinates, and the work type identifier, a spatiotemporal conflict analysis is performed to obtain spatiotemporal conflict verification results, specifically including: Based on the real-time environmental meteorological data and the abnormal evolution parameters, the medium leakage and diffusion risk domain corresponding to each discrete time slice of the associated process unit within a preset time period is calculated. The medium leakage and diffusion risk domain is used to characterize the spatial distribution range of the hazardous medium under the discrete time slice. The spatial morphological parameters corresponding to the job type identifier are called from the preset job geometry template library, and the job influence spatial domain is constructed based on the job position coordinates. Spatiotemporal overlap analysis is performed sequentially on the spatial domain of the operation's impact and the media leakage and diffusion risk domain corresponding to each discrete time slice to obtain the spatiotemporal conflict verification results of each discrete time slice.

[0011] Optionally, based on the real-time environmental meteorological data and the anomaly evolution parameters, the media leakage and diffusion risk domain corresponding to each discrete time slice within a preset time period of the associated process unit is calculated, specifically including: The instantaneous evolution values ​​corresponding to each discrete time slice are extracted from the abnormal evolution parameters, and the medium leakage mass rate of each discrete time slice is calculated based on the instantaneous evolution values ​​and the physical properties of the associated process unit using a preset leakage source term model. The leakage mass rate of the medium and the real-time environmental meteorological data are input into a preset gas diffusion and transport model to obtain the three-dimensional spatial concentration distribution field of the hazardous medium under the discrete time slice. In the three-dimensional spatial concentration distribution field, a spatial isosurface is determined where the concentration value is equal to the preset safe concentration threshold corresponding to the hazardous medium; The closed three-dimensional region enclosed by the spatial isosurface is defined as the medium leakage and diffusion risk domain.

[0012] Optionally, based on the spatiotemporal conflict verification result, an operation control instruction corresponding to the special operation permit request is generated, specifically including: The spatiotemporal conflict verification results of each discrete time slice are traversed in time sequence. If there is a spatiotemporal conflict verification result that represents a non-empty spatiotemporal intersection, the first target discrete time slice with a non-empty spatiotemporal intersection is identified, the time point corresponding to the target discrete time slice is determined as the critical risk moment, and the operation control instruction containing the critical risk moment and the conflict type is generated. If the spatiotemporal conflict verification results of all discrete time slices indicate that the spatiotemporal intersection is empty, a work permit authorization instruction is generated, and the extreme value of the fluctuation in the abnormal evolution parameter is used as a benchmark to configure the dynamic interlock shutdown threshold of the associated process unit in the work cycle corresponding to the special work permit request, and the dynamic interlock shutdown threshold is encapsulated into the work control instruction.

[0013] Secondly, embodiments of this application provide a dynamic licensing and control system for safe production operations. The dynamic licensing and control system for safe production operations includes: one or more processors and a memory; the memory is coupled to the one or more processors, and the memory is used to store computer program code, the computer program code including computer instructions, and the one or more processors call the computer instructions to cause the dynamic licensing and control system for safe production operations to perform the method described in the first aspect and any possible implementation thereof.

[0014] Thirdly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a dynamic licensing and control system for safe production operations, cause the dynamic licensing and control system for safe production operations to perform the method described in the first aspect and any possible implementation thereof.

[0015] Fourthly, embodiments of this application provide a computer program product containing instructions that, when the computer program product is run on a dynamic licensing and control system for safe production operations, causes the dynamic licensing and control system for safe production operations to execute the method described in the first aspect and any possible implementation thereof.

[0016] In summary, one or more technical solutions provided in this application have at least the following technical effects or advantages: 1. By introducing a mechanism for extracting work location coordinates and work type identifiers into the special operation permit process, and combining it with a preset spatial index relationship table to accurately locate the associated process units around the work location, the system obtains the time-series data of the current operation status of these process units and the real-time environmental meteorological data of the work area, providing comprehensive dynamic data support for subsequent risk assessment. Furthermore, the system identifies target historical work condition segments and their time evolution trends that are highly similar to the current work condition by performing sliding window similarity matching between the time-series data of the operation status and the historical operation and maintenance database. Combined with pre-stored historical operation result tags, the system infers the possible abnormal evolution parameters of the current process unit within a preset time period in the future. On this basis, the system integrates the abnormal evolution parameters with the real-time environmental meteorological data, performs spatiotemporal conflict analysis around the specific work location and work type, effectively identifies the safety interference or environmental coupling effect that the work activity may cause within the expected execution cycle, and finally generates accurate work control instructions and issues them to the work terminal for control. This achieves dynamic and forward-looking judgment and refined response to special operation permit requests, improving the accuracy of safety production operation permits.

[0017] 2. By performing discrete difference processing on the time-series data of the operating status and constructing a real-time morphological change rate sequence, the system can characterize the dynamic change trend of the equipment's operating status. Then, by combining multiple preset time scaling factors, the system resamples this sequence along the time axis to generate multiple sets of candidate feature sequences to adapt to different time scales of operating condition evolution. Subsequently, using a dynamic path regularization matching algorithm, the system aligns each candidate feature sequence with multiple historical operating condition records in the historical maintenance database, using the historical segment with the smallest cumulative alignment distance as the target historical operating condition segment, achieving accurate matching of operating conditions at different time scales. Based on this, the system further extracts the evolved operating conditions within a preset prediction time window from the subsequent raw historical data of the target historical operating condition segment. Through inverse time scale mapping consistent with the time scaling factor used for target matching, it generates evolution trend data corresponding to the current operating condition. This achieves multi-scale similarity matching and trend prediction between the time-series data of the operating status and historical operating conditions, effectively improving the system's ability to identify complex dynamic operating condition changes and the accuracy of predicting potential risk evolution. This provides a more timely and scenario-adaptive decision-making basis for subsequent abnormal evolution parameter deduction and work permit decisions.

[0018] 3. Extract anomaly type information from the historical operation result labels corresponding to the target historical operating condition segment, and combine the parameter change patterns of this anomaly type in the historical operating condition segment and its evolution trend data to generate a representative anomaly precursor symptom sequence; use the target preset time scaling factor to adjust the time scale of the symptom sequence to construct a target search template adapted to the current operating condition rhythm, and perform sliding similarity matching in the real-time operating status time series data to accurately locate the potential anomaly starting point where the feature similarity exceeds the preset threshold; based on this, further determine the anomaly verification window in the current operating condition data, calculate the real-time fluctuation intensity within the window, and perform a ratio analysis with the fluctuation intensity of the corresponding time period of the historical symptom sequence to generate a correction coefficient reflecting the difference in anomaly intensity; based on this intensity correction coefficient, adjust the amplitude of the historical evolution trend data to obtain anomaly evolution parameters that are more in line with the actual state of the current operating condition, thereby realizing the collaborative identification and dynamic adjustment of anomaly type, time evolution characteristics and fluctuation intensity in the anomaly early warning process, effectively improving the system's response sensitivity to potential risk states and the accuracy of evolution trend prediction, and enhancing the ability to make forward-looking judgments on future anomalies in the special operation permit process. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating a dynamic permitting and control method for safe production operations in an embodiment of this application; Figure 2 This is a flowchart illustrating the process of determining target historical operating condition segments and evolution trend data in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of a dynamic permit control system for safe production operations according to an embodiment of this application.

[0020] Explanation of reference numerals in the attached drawings: 301, Central Processing Unit; 302, Read-Only Memory; 303, Random Access Memory; 304, Bus; 305, Input / Output Interface; 306, Input Section; 307, Output Section; 308, Storage Section; 309, Communication Section; 310, Driver; 311, Removable Media. Detailed Implementation

[0021] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0022] In the description of the embodiments of this application, the words "for example" or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design that is described as "for example" or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design options. Rather, the use of the words "for example" or "for instance" is intended to present the relevant concepts in a specific manner.

[0023] In the description of the embodiments of this application, the term "multiple" means two or more. For example, multiple systems means two or more systems, and multiple screen terminals means two or more screen terminals. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0024] Figure 1 This is a flowchart illustrating a dynamic permitting and control method for safe production operations in an embodiment of this application.

[0025] Please see Figure 1 This application provides a method for dynamic permit control of safe production operations, the method comprising: S101. In response to a special operation permit request initiated by the work terminal, extract the work location coordinates and work type identifier from the special operation permit request, determine the associated process units within a preset radius centered on the work location coordinates based on a preset spatial index relationship table, and obtain the time sequence data of the operating status of the associated process units at the current moment, as well as the real-time environmental meteorological data of the area where the work location coordinates are located. To achieve dynamic, safe, and accurate authorization assessment of special operation permit requests, the operation terminal first initiates the request, which includes the work location coordinates and the work type identifier. The work location coordinates identify the geographical or spatial location of the operation, typically using latitude and longitude coordinates obtained from the Global Positioning System (GPS), or two-dimensional or three-dimensional coordinate information determined by a positioning system deployed in the industrial field (such as a UWB positioning system or a Bluetooth beacon positioning system). The work type identifier indicates the specific category of the operation, such as hot work, working at height, confined space work, or temporary power supply, each type having its corresponding risk characteristics and authorization requirements.

[0026] After receiving the aforementioned permission request, the system performs a spatial query on the work location coordinates based on a preset spatial index table. The spatial index table is a spatial data structure used to quickly find information about process units around a given location coordinate. Preferably, it employs an R-tree or quad-tree structure, associating the spatial boundaries of each process unit in the industrial site with its unique unit identifier. By establishing a query range with a preset radius centered on the work location coordinates in the spatial index table, all process units within that range that are spatially adjacent to or overlap with the work location can be quickly retrieved, forming a set of associated process units.

[0027] After identifying the associated process units, the system further acquires the time-series data of their current operating status. This operating status time-series data refers to the continuous time series data of the process unit's current process parameters (such as temperature, pressure, flow rate, current, vibration, etc.), typically sourced from a DCS (Distributed Control System) or SCADA (Supervisory Control and Data Acquisition) system. The parameter data is timestamped to reflect the dynamic operating status of the process unit.

[0028] Simultaneously, the system also acquires real-time environmental meteorological data for the area where the work location is located. This data includes, but is not limited to, information on temperature, humidity, wind speed, wind direction, air pressure, flammable gas concentration, and toxic gas concentration. This data is typically collected by meteorological monitoring sensors deployed near the work area or provided by an industrial meteorological platform. Environmental meteorological data plays a crucial role in assessing the external risks of special operations; for example, certain types of work must be prohibited under conditions of high wind speeds or the presence of toxic gas leaks.

[0029] Through the above operations, the system completes the entire process of extracting key parameters from the permit request, locating relevant process units, and acquiring real-time process data and environmental information. This process lays the data foundation for subsequent similarity matching based on historical operating conditions, anomaly trend prediction, and spatiotemporal conflict analysis, ensuring that permit judgments are dynamic, real-time, and accurate. For example, when a work terminal initiates a hot work permit request, after locating the coordinates of the request at the southwest corner of the tank area, the system quickly retrieves two chemical storage tank units within a 50-meter radius using a spatial index table, and obtains their current operating parameters such as temperature, pressure, and liquid level. Simultaneously, it collects real-time meteorological data for the area, showing a wind speed of 7.2 m / s and a wind direction of southeast-east. In this case, the system can combine this data for further work permit analysis.

[0030] S102. Perform sliding window similarity matching between the running status time series data and the pre-stored historical operation and maintenance database to obtain the target historical working condition segment corresponding to the special operation permit request and the evolution trend data of the target historical working condition segment on the time axis. The historical operation and maintenance database contains historical operation result tags corresponding to each historical working condition segment. After extracting and acquiring the work location coordinates, work type identifier, associated process units, and their operational status time-series data, further assessment of the risk evolution stage of the current process operation status is needed to achieve dynamic risk determination for special operation permit requests. To this end, the system performs sliding window similarity matching between the acquired operational status time-series data and a pre-stored historical maintenance database. This identifies the target historical operational condition segment most similar to the current operational condition, and based on the evolution process of this target segment on the historical timeline, extracts its evolution trend data, providing a reference for subsequent anomaly prediction. The above matching process involves feature extraction, time-series alignment, and similarity calculation of the operational status data. Specifically, this can be achieved by performing differential processing on the operational status time-series data and combining it with multi-scale time-scaling resampling, followed by dynamic path normalization matching in the historical database to determine the optimal matching historical operational condition segment and its evolution trend. Figure 2 This is a flowchart illustrating the process of determining target historical operating condition segments and evolution trend data in an embodiment of this application. The following is a summary of the process. Figure 2Step S102 will be described in detail.

[0031] S201. Perform discrete difference processing on the time series data of the running state to obtain the real-time morphological change rate sequence, and resample the real-time morphological change rate sequence on the time axis based on multiple different preset time scaling factors to obtain the candidate test feature sequence corresponding to each preset time scaling factor. In this embodiment, to achieve efficient similarity matching between the operational status time-series data and historical operating condition records in the historical operation and maintenance database, the currently acquired operational status time-series data first needs to undergo structured feature extraction processing. Since the operational status of industrial process units typically manifests as multi-dimensional time-series data of multiple physical quantities (such as temperature, pressure, liquid level, flow rate, etc.) dynamically changing over time, direct overall comparison is not only computationally complex but also sensitive to inconsistencies in time scale and amplitude scale, easily leading to matching distortion. Therefore, the original operational status time-series data is first subjected to discrete difference processing to extract the numerical differences between adjacent moments in the original sequence, constructing a morphological change rate sequence. The morphological change rate sequence essentially reflects the rate and trend direction of change of process parameters over time, which can enhance the key change characteristics of the operating status during dynamic evolution and effectively reduce the interference of amplitude disturbances.

[0032] The specific method of discrete difference processing is as follows: Suppose the original running state time series data is a set of time series [x1, x2, x3, ..., x...]. n If the morphological change rate sequence is [x2−x1, x3−x2, ..., x...], then the sequence is [x2−x1, x3−x2, ..., x...]. n -x n−1 This difference sequence has stronger dynamic sensitivity and can highlight the abrupt changes, fluctuation rhythms, and trend reversals of the operating conditions. It is a key intermediate representation for subsequent similarity matching.

[0033] After obtaining the real-time morphological change rate sequence, several different preset time scaling factors are introduced to further enhance the robustness and multi-scale adaptability of the matching. These preset time scaling factors are a set of representative proportional parameters discretely set after statistical analysis of typical events, based on common timescale differences in historical operating condition evolution. They refer to the proportional parameters used to compress or expand the time axis, simulating the manifestation of operating condition changes at different time scales. Different operating condition segments may exhibit the same dynamic trends but different rates of change in actual operation; therefore, the morphological change rate sequence needs to be resampled in the time dimension. The core principle of resampling is to scale the original difference sequence through interpolation or downsampling to generate multiple versions of candidate feature sequences, each corresponding to a specific time scaling factor.

[0034] For example, if the morphological change rate sequence has 100 sampling points, candidate feature sequences of length 80, 100, and 120 can be generated based on time scaling factors of 0.8, 1.0, and 1.2, respectively. In this way, the system can adapt to changes in the time scale of different historical segments when performing dynamic path regularization matching with historical working condition records, thereby improving the accuracy and flexibility of the matching.

[0035] By combining discrete difference processing with multi-timescale resampling, the expressive power and matching compatibility of operational status time-series data are significantly improved. This also provides high-quality input feature sequences for dynamic time warping matching algorithms, ultimately achieving more accurate identification of target historical operating condition segments. For example, in practical applications, if the pressure and temperature data of a certain storage tank unit show a slow upward trend, differential processing can generate a sequence of continuously positive rates of change. Combined with time scaling factor resampling, it can be found that its characteristic trend is highly similar to the early evolution stage of an overheating and pressure relief event in the historical database, providing a reliable basis for subsequent abnormal trend prediction.

[0036] S202. Dynamic path normalization and matching are performed on each of the candidate feature sequences to be tested in multiple historical operating condition records in the pre-stored historical operation and maintenance database. For the first historical operating condition record, the cumulative alignment distance between each of the candidate feature sequences to be tested and the first historical operating condition record is calculated. The first historical operating condition record is any historical operating condition record. To accurately identify historical operating condition segments most similar to the current process unit's operating state from the historical operation and maintenance database, dynamic path normalization matching needs to be performed on multiple candidate feature sequences to be tested and all historical operating condition records, and the cumulative alignment distance between them needs to be calculated. The purpose of this operation is to solve the problem of direct comparison failure caused by the nonlinear rate difference in the evolution of operating conditions on the time axis, thereby achieving dynamic similarity assessment across time scales.

[0037] Specifically, the previous processing step has constructed multiple candidate feature sequences to be tested based on different time scaling factors. Each sequence represents the dynamic change characteristics of the current operating condition at a specific time scale. The historical operation and maintenance database stores a large number of historical operating condition records. Each record consists of the operating status of a process unit within a certain historical period. The corresponding time-series structure may differ from the current feature sequence in length, time distribution, and change pattern. Therefore, a dynamic path regularization matching algorithm is used to align the candidate feature sequences to be tested with the historical operating condition records, so as to evaluate the dynamic similarity between the two under the condition of allowing nonlinear stretching and compression of the time axis.

[0038] The dynamic path warping matching is preferably implemented using the Dynamic Time Warping (DTW) algorithm. DTW is a classic method for measuring the similarity between two time series. Its basic principle is to find a path with the minimum cumulative cost in a two-dimensional distance matrix, such that the points of the two series are optimally aligned on the time axis. In specific implementation, firstly, a distance matrix is ​​constructed between the candidate feature sequence and a certain historical work condition record, where each element represents the Euclidean distance or Manhattan distance between feature values ​​at two time points. Then, the path cost is recursively calculated using dynamic programming to obtain the minimum cumulative alignment distance from the starting point to the ending point. This process is repeated for each historical work condition record, ultimately forming a set of cumulative alignment distances between each candidate feature sequence and all historical work condition records.

[0039] Through the above processing, the system can identify historical operating condition segments that best reflect the dynamic characteristics of the current operating state across different time dimensions, providing a reliable similarity basis for determining subsequent target operating condition segments and analyzing their trend evolution. This matching process not only eliminates interference caused by inconsistencies in the timeline but also preserves the overall structural characteristics of changes in operating condition morphology, significantly improving the accuracy and robustness of similarity judgment.

[0040] For example, in practical applications, if a candidate feature sequence indicates that a pressure vessel exhibits a trend of first stabilization and then gradual increase during the current period, the system uses the DTW matching algorithm to calculate the alignment distance between the sequence and all operating condition records in the historical database. Ultimately, it finds that the early state of a pressure relief event two years ago has the minimum cumulative alignment distance at a time scaling factor of 1.2. This historical segment can then be used as the target historical operating condition segment corresponding to the current request, providing a basic input for subsequent risk prediction.

[0041] S203. Determine the historical working condition record corresponding to the minimum cumulative alignment distance as the target historical working condition segment; To accurately identify the most similar historical operating condition segment corresponding to the current operating state from the historical operation and maintenance database, and to support subsequent trend prediction and risk assessment, it is necessary to determine the historical operating condition record corresponding to the minimum cumulative alignment distance as the target historical operating condition segment based on the cumulative alignment distance results obtained in the previous processing step. The core purpose of this operation is to select the historical data sample that is closest to the current operating state in terms of dynamic change pattern from multiple matching candidates, thereby maximizing the reliability of similarity matching and the reference value for subsequent analysis.

[0042] In the specific implementation process, the system has calculated the cumulative alignment distance between each candidate feature sequence to be tested and each historical working condition record in the historical operation and maintenance database using a dynamic path warping matching algorithm (preferably dynamic time warping algorithm, DTW). Each alignment distance value represents the dynamic similarity between the candidate sequence and the corresponding historical record under the condition of time axis deformability. Since multiple time scaling factors are preset, the resulting candidate feature sequences are also diverse. The system uniformly incorporates all matching results into the evaluation, performs a global comparison based on the value of the cumulative alignment distance, and selects the matching pair with the smallest cumulative alignment distance.

[0043] The minimum cumulative alignment distance indicates that the current process unit's operating status is highly consistent with the matched historical operating condition record in terms of change trend, fluctuation pattern, and dynamic cycle, and can truly reflect the evolution stage of the actual operating condition. To ensure the uniqueness and consistency of the judgment, the system adopts a minimum value selection strategy, that is, the historical operating condition record with the minimum global cumulative alignment distance is used as the target historical operating condition segment, and the corresponding time period, process parameter sequence, and operating result label of the record are extracted for subsequent trend inference and abnormal parameter calculation.

[0044] This target operating condition segment determination mechanism based on the minimum cumulative alignment distance avoids the uncertainty brought about by subjective judgment or empirical rules, and improves the data-driven capability and automation level of the entire dynamic permit control method. For example, if a reactor unit currently exhibits characteristics of continuously rising temperature and intensified pressure fluctuations within the past 30 minutes, the system performs dynamic path normalization matching between the morphological change rate sequence and hundreds of records in the historical operating condition database. Ultimately, it identifies an operating condition data recorded during a catalyst failure last year, whose DTW alignment distance is the smallest among all candidates. This record is then determined as the target historical operating condition segment to provide historical reference for subsequent risk trend analysis and permit decisions.

[0045] S204. Determine the evolution trend data based on the target historical operating condition segment.

[0046] After identifying the target historical operating condition segment that best matches the current operating state, to further achieve dynamic prediction of the future behavior of the process unit, it is necessary to extract the actual evolution process within the subsequent time period from the historical data corresponding to the target historical operating condition segment, and construct evolution trend data accordingly, as a priori reference for the calculation of subsequent abnormal evolution parameters. Since the target historical operating condition segment is obtained by matching with a specific time scaling factor, to ensure that the extracted trend data is consistent with the current state in terms of time scale, it is necessary to perform reverse time scale mapping on the original operating condition data after the target segment in combination with the preset time scaling factor to restore its time correspondence with the current time series state. Specifically, this includes the following steps: obtaining the original operating condition data within a preset prediction time window after the last deadline of the target historical operating condition segment from the pre-stored historical operation and maintenance database, wherein the start time of the preset prediction time window is the same as the last deadline of the target historical operating condition segment; performing reverse time scale mapping on the original operating condition data based on the target preset time scaling factor to obtain the evolution trend data, wherein the target preset time scaling factor is the preset time scaling factor used when determining the target historical operating condition segment.

[0047] After determining the target historical operating condition segment, to support the prediction of the future trend of the current operating status, it is necessary to further extract subsequent raw operating condition data that are continuous with the target historical operating condition segment from the historical operation and maintenance database. That is, historical time series data within a preset prediction time window after the final deadline of the target historical operating condition segment. The preset prediction time window is not an arbitrarily selected period on the timeline, but refers to a continuous duration in the historical database whose start time is strictly adjacent to or coincides with the final deadline of the target historical operating condition segment. Its purpose is to capture the immediate subsequent development state of the historical operating condition after its completion, ensuring that the extracted data is a natural continuation of the target historical operating condition segment in terms of temporal causality. Thus, by using the real consequences that immediately follow in the historical records, a seamless and accurate inference can be made about the future evolution trend of the current operation. It is a time length set according to system requirements for predictive analysis, such as 10 minutes, 30 minutes, or 1 hour, to reflect the state change trend of the target historical operating condition in its subsequent natural evolution process. The purpose of this step is to obtain the "future part" of the target's historical operating conditions as a reference for subsequent abnormal evolution inference, thereby establishing a prediction mechanism based on historical evolution and improving the accuracy of judging the risk trend of the current operating conditions.

[0048] To ensure consistency between the extracted raw operating condition data and the current operating state's time dynamic structure, a reverse time-scale mapping process is required. Since the target historical operating condition segment is obtained by matching a predetermined time scaling factor with the current real-time morphological change rate sequence, this time scaling factor reflects the compression or expansion ratio of the target historical operating condition segment on the time axis. Therefore, to ensure the temporal rhythm of the trend data remains consistent with the current state, subsequent raw operating condition data needs to be inversely transformed according to this predetermined time scaling factor to align its time scale with the current state. The principle of reverse time-scale mapping is to perform an inverse scaling operation on the data time axis. For example, if the predetermined time scaling factor is 1.2, indicating that the target historical segment was compressed by 20% during the matching process, the corresponding trend data should expand its time axis back to the original ratio, i.e., the interval between each sampling point is shortened to 1 / 1.2 of the original, thus restoring its true dynamic rhythm.

[0049] Through the two operations described above, the system ultimately obtains a segment of evolution trend data that is highly consistent with the current operating condition in terms of time structure and reflects the historical natural evolution trend. This data is used for subsequent calculation of abnormal evolution parameters and risk prediction. For example, if the current operating state of a high-pressure reactor highly matches a segment of operating conditions within a year where a slow temperature rise caused overpressure, and this segment was obtained through a time scaling factor of 0.8, the system will extract the temperature and pressure change data for the next 30 minutes at the beginning and end of this segment, and perform time axis expansion processing based on a time scaling factor of 0.8 to generate trend data. This data serves as a historical mapping of the possible evolution path of the current operating state in the next 30 minutes, thereby providing data support for the dynamic permission judgment of special operations.

[0050] S103. Calculate the abnormal evolution parameters of the associated process unit within a preset time period in the future based on the historical operation result label corresponding to the target historical operating condition segment, the evolution trend data, and the operation status time series data. Step S103 aims to, based on the determination of the target historical operating condition segment and trend extraction, and combined with the current actual operating status data, to extrapolate and calculate the possible abnormal development paths of related process units in the future, thereby forming a quantitative prediction of risk. This step is a key analytical link for achieving proactive risk identification. Its essence lies in identifying possible abnormal triggering signals and their evolution trends by comparing the current operating conditions with historical abnormal samples with similar operating patterns. Specifically, it may include the following steps: Based on the anomaly type in the historical operation result label, the parameter change pattern corresponding to the anomaly type is determined from the target historical operating condition segment and the evolution trend data, and an anomaly precursor sequence is generated. Based on the target preset time scaling factor, the time scale of the preceding symptom sequence is adjusted to obtain the target search template, and the target search template is used for sliding similarity matching in the running state time series data. The moment when the feature similarity is greater than the preset similarity threshold is determined as the potential abnormal starting point of the current working condition. Using each potential anomaly initiation point as a reference time, the current anomaly verification window is determined in the running state time series data, and the real-time fluctuation intensity within the current anomaly verification window is calculated. The ratio of the real-time fluctuation intensity to the historical fluctuation intensity of the abnormal precursor sequence in the corresponding historical period is calculated to obtain the intensity correction coefficient; The amplitude of the evolution trend data is corrected based on the intensity correction coefficient to obtain the abnormal evolution parameters.

[0051] Historical operation result tags are structured information tags attached to historical operation and maintenance data. They are used to identify the operational status after the completion of a historical operating condition segment, including whether an anomaly occurred, the type of anomaly (such as excessively high temperature, sudden pressure changes, abnormal equipment vibration, etc.), the time of the anomaly, and the parameters that may have been affected. These tags are derived from historical event records, alarm logs, or automatically labeled models trained based on expert experience, and can provide a basis for anomaly tracing in subsequent analysis.

[0052] To identify potential trends similar to historical anomalies in the current operating condition, the system first extracts a sequence of process parameters highly correlated with the anomaly type from the data of the target historical operating condition segment and its evolution trend. It then further analyzes the typical variation patterns of these parameters before the anomaly occurs. For example, if the anomaly type is a sudden pressure surge, it is necessary to extract continuous variation data of pressure-related parameters in that segment from 5 to 15 minutes before the anomaly occurs. A time series template reflecting the precursory characteristics of this anomaly is constructed using indicators such as difference, slope, and volatility. This template, known as the "anomaly precursor sequence," is essentially a dynamic signal pattern that is repeatable in the preceding stages of multiple historical anomalies and can be used to identify precursors to anomalies in the current operating condition.

[0053] The target preset time scaling factor refers to the time scale transformation coefficient introduced in the aforementioned similarity matching step to align the current running time series data with historical data in terms of dynamic rhythm. Its value is usually a floating-point number, such as 0.8 or 1.2, representing the proportion of time that historical data is compressed or stretched relative to current data.

[0054] Since the anomaly precursor sequence is extracted from historical data, its original timescale may deviate from the current operating state. To ensure accurate identification in the current data, the anomaly precursor sequence needs to be "timescale adjusted" based on this time scaling factor. This adjustment process includes resampling or interpolating the original sequence to ensure its time coordinates are consistent with the current operating state, generating a "target search template." For example, if the scaling factor is 1.25, the time interval of 1 second in the original sequence needs to be shortened to 0.8 seconds to simulate the current system's operating rhythm.

[0055] Subsequently, the system uses a sliding window approach to match the target search template with the data within each window of the current running time series data, employing similarity calculation methods such as Dynamic Time Warping (DTW) and cosine similarity. When the similarity value at a certain moment exceeds a set threshold (e.g., similarity greater than 0.85), it indicates that the current running state exhibits a behavioral pattern similar to historical anomaly precursors. This moment is marked as a "potential anomaly initiation point," representing that the system may have entered the early stages of anomaly evolution.

[0056] After detecting potential anomaly initiation points, the system needs to further verify whether the anomaly has a tendency to continue developing. To this end, a fixed-length "current anomaly verification window" is defined forward from each potential anomaly initiation point. The window length can be set according to the historical anomaly evolution cycle, such as 5 minutes or 10 minutes. Within this verification window, the system performs volatility analysis on key process parameters (such as temperature, pressure, and flow rate) and calculates their "real-time volatility intensity." Volatility intensity is an indicator that measures the severity of parameter changes, and can be calculated using statistical methods such as standard deviation, variance, maximum-minimum difference, and moving average slope. For example, for a pressure sequence, its standard deviation can be calculated within the window as the volatility intensity value. If this value is significantly higher than the normal operating value, it indicates that the system is currently in an unstable state. This step can be used to screen out critical moments with actual evolutionary risks from multiple potential anomaly initiation points, avoiding false alarms and excessive intervention.

[0057] To further quantify the difference in evolution intensity between the current operating conditions and historical anomalies, the system needs to calculate the ratio of the "real-time fluctuation intensity" within the current anomaly verification window to the "historical fluctuation intensity" within the same time period before the historical anomaly occurred. The historical fluctuation intensity can be obtained by applying the same calculation method to the corresponding time period in the target historical operating condition segment and its anomaly precursor sequence. The ratio of the two is defined as the intensity correction coefficient, reflecting the degree of evolution of the current system state relative to historical anomalies. For example, if the current pressure fluctuation intensity is 0.15 and the historical intensity is 0.10, then the intensity correction coefficient is 1.5. This coefficient will be used subsequently to dynamically adjust the amplitude of trend prediction results, making the prediction results closer to the actual evolution potential of the current system.

[0058] After obtaining the intensity correction coefficient, the system needs to adjust the amplitude of the extracted evolution trend data to generate the final abnormal evolution parameters. The evolution trend data is the predicted evolution path of the target historical operating conditions over a future period, and the original amplitude reflects the degree of response of the historical system to abnormal evolution.

[0059] The system multiplies each numerical point on the trend curve by an intensity correction factor to obtain a new trend with a magnified or reduced amplitude. This process maintains the trend shape but dynamically adjusts its evolution amplitude, thereby improving the adaptability of the prediction results to the current state. For example, if the historical trend shows that the temperature will rise from 80°C to 95°C within 10 minutes, and the current intensity correction factor is 1.2, the prediction result will be adjusted to rise to 99°C within 10 minutes. The final abnormal evolution parameters include predicted extreme values, rate of increase, and fluctuation range, serving as key inputs for subsequent permissible judgments and risk warnings.

[0060] If the current system identifies an abnormal precursor symptom in a high-pressure device that is similar to a historical "sudden pressure surge before deflagration," with a similarity of 0.91, the system will use that moment as the potential anomaly starting point and analyze pressure fluctuations over the next 5 minutes. If the fluctuation intensity is 1.6 times that of historical events, the historical trend will be amplified, ultimately predicting that the current device may exceed its design pressure limit within the next 10 minutes. This will trigger a work permit blocking mechanism and generate dynamic control instructions. This implementation process embodies a dynamic risk identification and trend prediction mechanism based on historical-current collaborative modeling.

[0061] S104. Based on the real-time environmental meteorological data, the abnormal evolution parameters, the operation location coordinates, and the operation type identifier, perform spatiotemporal conflict analysis to obtain spatiotemporal conflict verification results. After predicting the current operational risk trend, the system needs to further consider the location characteristics of the work space, the evolution trend of potentially released high-risk media, and the impact of current environmental meteorological conditions on the diffusion of hazardous substances. This involves constructing a dynamic assessment scenario that conforms to spatiotemporal semantics to determine whether there is a real-world conflict risk where the operational activity overlaps with the risk area. Specifically, the core of step S104 is to analyze the dynamic interaction between the work space domain and the media leakage risk domain. Simultaneously, it combines real-time wind speed, wind direction, atmospheric stability, and other meteorological elements to extrapolate the spatial expansion trend of the risk domain formed by abnormal evolution parameters. Based on this, it systematically identifies and quantifies the potential risk intersections that special operation permits may constitute. To achieve this, step S104 specifically includes steps S1041-S1043, which are described in detail below.

[0062] S1041. Based on the real-time environmental meteorological data and the abnormal evolution parameters, calculate the medium leakage and diffusion risk domain corresponding to each discrete time slice of the associated process unit within a preset time period. The medium leakage and diffusion risk domain is used to characterize the spatial distribution range of the hazardous medium under the discrete time slice. During the spatiotemporal conflict verification process, the system first needs to construct a dynamic distribution model of the potential spatial diffusion of hazardous media to reflect the potential risk areas formed at each moment within the prediction period, providing a physical basis for the superposition analysis of the impact domain of subsequent operations. Therefore, the key to step S1041 lies in simulating and reconstructing the time-varying spatial concentration field of hazardous media in each discrete time slice based on the hazardous media release status described by the current abnormal evolution parameters and the actual impact of real-time meteorological conditions (such as wind speed, wind direction, atmospheric stability, etc.) on diffusion behavior, relying on a professional diffusion and transport model, and identifying risk areas above the safety threshold in each time slice, thereby defining the "media leakage diffusion risk domain". This spatial domain can be regarded as the temporal projection of the impact range of the hazardous source in space, accurately characterizing the combined effect of leakage evolution and meteorological response. Specifically, it may include the following steps: extracting the instantaneous evolution values ​​corresponding to each discrete time slice from the abnormal evolution parameters, and calculating the media leakage mass rate of each discrete time slice based on the instantaneous evolution values ​​and the physical properties of the associated process unit using a preset leakage source term model; The leakage mass rate of the medium and the real-time environmental meteorological data are input into a preset gas diffusion and transport model to obtain the three-dimensional spatial concentration distribution field of the hazardous medium under the discrete time slice. In the three-dimensional spatial concentration distribution field, a spatial isosurface is determined where the concentration value is equal to the preset safe concentration threshold corresponding to the hazardous medium; The closed three-dimensional region enclosed by the spatial isosurface is defined as the medium leakage and diffusion risk domain.

[0063] In conducting spatiotemporal conflict analysis for dynamic permit control, predicting the diffusion behavior of potentially hazardous media is a fundamental prerequisite for assessing regional conflict risks. Therefore, it is essential to derive the leakage evolution characteristics for each time slice based on the currently predicted abnormal situation. The key to this process is extracting time-related "instantaneous evolution values" from "abnormal evolution parameters." Abnormal evolution parameters are trend predictions of key media (such as flammable gases and corrosive liquids) over a predetermined future timeframe, generated by the model based on historical operating condition segments and real-time time-series data. These predictions are typically numerical curves showing changes over time. The data points can represent the evolution of a specific process unit, such as pipeline end pressure, temperature within a closed tank, or gas volumetric flow rate.

[0064] For discrete-time slices, the core lies in sampling each specific moment from this dynamic evolution trend to form "instantaneous evolution values." In this embodiment, the instantaneous evolution values ​​are generally the instantaneous changes in the target parameters (such as instantaneous pressure increases, temperature gradient changes, etc.). Alternatively, depending on the requirements of the specific leakage model, they can be extracted as differential forms such as the cumulative rate of volume change, the rate of temperature rise, and the slope of system pressure drop. The significance of extracting these values ​​is that they can be input as driving factors into subsequent leakage source term models, thereby enabling the risk simulation results to dynamically respond to the current evolution trend.

[0065] After obtaining the instantaneous evolution values ​​of the slice state, the corresponding media leakage mass rate needs to be calculated based on this state and the physical attributes of the identified associated process units. Associated process units generally include valve segments, storage tank units, transmission branches, etc., and refer to the lowest physical unit where the equipment system and the work location space have spatial and process associations. In actual operation, the system retrieves the structural attributes (such as pipe diameter, wall thickness, pressure rating), media physicochemical parameters (such as critical pressure, gas-liquid distribution coefficient), and leakage logic state (such as safety valve opening status, whether the fluid state exceeds limits) of the unit from the process diagram database or material management model, and matches them with the instantaneous evolution values ​​to construct a leakage behavior input set that conforms to the actual boundary conditions.

[0066] The leakage source model used in this method refers to a mathematical model constructed based on the physical mechanism of sudden leakage of hazardous chemicals. Its implementation can take the form of an orifice flow model, a fissure flow model, or a liquid-gas two-phase injection model. Common simulation formulas include mass flux calculation formulas based on Bernoulli's equation or leakage power estimation methods recommended by the DOE. For example, if the associated process unit is a high-pressure reactor pipeline, its leakage source model will use the critical flow model for compressible gases. Knowing that the current projection slope increases sharply to a pressure jump of 30%, the predicted results of the current cross-sectional area, internal pressure, and external pressure can be substituted to calculate the medium mass leakage rate (in kg / s) over a short period. This data directly determines the source strength accuracy of the downstream gas diffusion concentration model and will have a decisive impact on the risk space assessment results.

[0067] For example, in a certain work permit scenario, if a rapid 15% pressure increase in a reactor within a related process unit is predicted within 10 minutes, and the reactor wall thickness is insufficient to withstand this magnitude, the source term model, combined with equipment leakage history and design data, determines that a stable Class C gas leak will occur after 8 minutes. Based on the cross-sectional area and medium properties, the mass flow rate at this time is calculated to be approximately 0.85 kg / s. This rate will be used as the source strength input for the gas diffusion model to initiate subsequent three-dimensional diffusion field simulations, and will be directly used in the risk domain construction during the spatiotemporal conflict verification process. This approach not only reflects the responsiveness to real-time operating conditions but also the predictive and dynamic characteristics of permit assessment.

[0068] After calculating the media leakage mass rate for each discrete time slice, the system must further simulate the real-time diffusion behavior of the medium in space to construct a dynamic spatial risk model that accurately reflects the consequences of leakage. To this end, this embodiment uses the "media leakage mass rate" and the "real-time environmental meteorological data" corresponding to the work area as joint driving factors input into the "gas diffusion and transport model" (hereinafter referred to as the "diffusion model"), thereby calculating and outputting the three-dimensional spatial concentration distribution field at each time point. This concentration distribution field is the core physical basis for constructing the risk domain, and its accuracy directly affects the reliability of subsequent spatiotemporal conflict analysis.

[0069] The media leakage mass rate is estimated from the previous processing step using instantaneous evolution values ​​and equipment physical structural parameters. It is typically expressed in kg / s and represents the instantaneous output capacity of the leakage source strength within a certain time slice. These leakage source parameters essentially determine the energy or mass input variables of the initial boundary conditions in the diffusion field, while the diffusion model is used to transform this source strength information into a spatially distributed hazard through physical transport mechanisms.

[0070] Gas diffusion and transport models are numerical simulation systems based on fluid mechanics, gas transport equations, and environmental response mechanisms. Their principles are typically based on Gaussian diffusion, Lagrange particle transport, or CFD (Computational Fluid Dynamics) methods to predict and model the transport behavior of leaked substances in the air under a given wind field. In implementation, the system calls pre-defined models implemented by embedded or integrated external modules (such as the AERMOD model recommended by the US EPA, the GIDAPS model from Germany, etc.) and sets the model input parameters, including: leak source location (operation coordinates), source strength (mass rate), release height (given by equipment structure), real-time meteorological elements (wind speed, wind direction, atmospheric stability, temperature stratification), and terrain parameters (such as ground roughness, surface thermal reflectivity). Real-time meteorological data is generally collected through the interface of regional meteorological stations. If there is insufficient sampling frequency, it can be interpolated and processed to form a dynamic input stream on a 5-10 minute scale.

[0071] In practical operation, the model uses the proposed discrete-time slices as the simulation step size, initiating the calculation process slice by slice. Initial boundary conditions are constructed for each time point, and then the distribution concentration values ​​of the hazardous medium in the three-dimensional spatial range (x, y, z axes) are iteratively calculated based on the control equations of the diffusion simulation. The output result is denoted as the "spatial concentration distribution field." Its representation can be a three-dimensional raster data matrix (volume grid), a vector cloud, or isopleths of point clouds, etc.

[0072] For example, in a certain time slice during the pressure vessel leak prediction process, if the leak source strength, calculated in the previous step, is 1.25 kg / s, and the meteorological conditions are moderate wind speed (3.5 m / s), stability category D, and wind direction north-northeast, the system will use this to construct the source term and wind field conditions as input into the AERMOD model. Within a simulation height of 10 meters, it will generate an asymmetric concentration cloud cluster extending southeastward from the leak source. The output concentration distribution will be expressed as a unit volume distribution (mg / m³) representing the hazardous gas concentration at any point in the three-dimensional region.

[0073] This three-dimensional concentration distribution field is used not only to extract spatial isosurfaces for safety thresholds in subsequent processing, but also to provide quantitative support for dynamic spatiotemporal overlap analysis. Furthermore, the resolution of the concentration distribution can be dynamically adjusted; higher accuracy allows for more refined calculation of risk domain boundaries, thereby improving the accuracy of special operation overlap judgments and the robustness of safety boundary control. Through the above modeling process, the system can constructively predict the spatiotemporal projection characteristics of accident consequences under different weather and operational conditions, and help identify potential operational conflict areas in advance, thus strongly supporting the scientific and intelligent nature of the dynamic permit control mechanism.

[0074] After successfully constructing a spatial isosurface in the three-dimensional spatial concentration distribution field where the concentration value is equal to the preset safe concentration threshold corresponding to the hazardous medium, the system needs to further analyze and process the spatial geometric characteristics of the isosurface, determine the closed three-dimensional region enclosed by it as the "medium leakage and diffusion risk domain", and register it as the risk space boundary under the current discrete time slice.

[0075] The determination of this "closed three-dimensional region" goes beyond the abstract extraction of the concentration critical surface. Instead, it uses isosurfaces as a shell to systematically encompass all spatial cells with concentrations greater than or equal to the threshold. Logically, this is approximates using the isosurface as a hazardous boundary crust, dividing the entire three-dimensional space into "hazardous zones" and "non-hazardous zones." Based on this boundary, the system creates a three-dimensional volumetric mesh data model. Its basic structure can be an octree voxel (also suitable for rapid spatial relationship determination), or it can combine boundary mesh Boolean operations (such as constructing B-Rep or CSG forms) to generate a rigorous geometric description. The core purpose of the above processing operations is to clearly define the minimum enclosing space of the hazard source's energy influence at each slice moment and assign it computable, visual, and comparable geometric entity features.

[0076] The generation of this "medium leakage and diffusion risk domain" is a key technical task supporting the spatiotemporal conflict judgment of special operation permits, and a core bridge in the transition from "simulated concentration data" to "spatiotemporal decision constraint rules." Its significance lies in the fact that without further abstracting and merging the hazardous "point values" displayed by the concentration value distribution into a unified regional entity, it is impossible to complete the geometric intersection calculation with the "operation impact spatial domain," nor can dynamic spatial conflict tracking under multiple time slices be performed. Therefore, materializing the three-dimensional structure enclosed by the isosurface into a "risk domain" is one of the key points in the logical closed loop of the special permit analysis process.

[0077] In its implementation, the system uses spatial point set judgment and in vivo envelope detection algorithms to filter all cells in the concentration grid whose attribute values ​​are greater than or equal to a threshold. It then constructs a closed-structure node relationship using isosurfaces as the shell and internal concentration points as the filling. This process is largely based on a grid spatial discretization strategy, utilizing numerical connectivity analysis (such as 26-neighbor topological connectivity) to connect the target region into a continuous three-dimensional group. It can also choose to perform multi-threshold enclosing clustering based on density field co-topology fractions to achieve comprehensive modeling of irregularly shaped regions. Simultaneously, this risk domain can be assigned meta-attribute information, such as medium type, current time, leakage rate distribution, and wind direction, for subsequent multi-source tracing analysis and evolution trajectory visualization in the risk management system.

[0078] For a practical example, if a pressure relief event occurs in a chlorine tank area, the system detects low wind speeds and a high-temperature inversion layer as the meteorological conditions. The diffusion model predicts that at the 6-minute mark, an isosurface with a concentration of 5 ppm will form an ellipsoidal three-dimensional structure extending southeastward and curving upwards. Using this isosurface boundary, the system calculates a "chlorine diffusion risk domain." The spatial morphology of this domain differs significantly from the standard spherical region under the calm wind assumption, helping users more scientifically identify which areas pose a risk to special operations at the current moment. Ultimately, this spatial region participates in the subsequent conflict domain comparison logic, immediately triggering the operation control unit to issue a "temporary suspension of operations" command to the affected area and automatically generating coordinated scheduling suggestions.

[0079] In summary, explicitly defining the volume enclosed by the isosurface as the "medium leakage and diffusion risk domain" not only forms a dynamic risk space boundary but also provides a quantifiable and algorithmically analyzable spatial basis for special operation permits based on physical simulation. This enables permit management to have the characteristics of high timeliness and high decision-making accuracy, and is a necessary step to realize a dynamic intelligent safety control system.

[0080] S1042. Call the spatial morphological parameters corresponding to the operation type identifier from the preset operation geometry template library, and construct the operation influence spatial domain based on the operation position coordinates based on the spatial morphological parameters; To achieve accurate conflict determination of permit requests, the system also needs to define the spatial range that special operations can reach or influence on at the same spatiotemporal scale. Therefore, step S1042 needs to be executed, which involves calling the spatial morphological parameters corresponding to the operation type identifier from the preset operation geometry template library, and constructing an operation influence spatial domain based on the operation location coordinates. The core purpose of this step is to map the "operation object" from an abstract event to a spatial entity so that it can participate in subsequent spatial relationship calculations.

[0081] The operational geometry template library mentioned here is a pre-established structured engineering parameter library that records the space required for different types of special operations during actual operation, such as high-altitude welding, hot work, confined space operations, and pressurized operations. Each type of operation is assigned a standard spatial geometry model. These models typically include three-dimensional shape definitions (such as rectangular prisms, prisms, hemispheres, etc.), dimensional parameters (such as radius, height, operating radius zone, and construction operation field of view cone, etc.), and behavioral extension factors (such as personnel activity radius and heat radiation propagation angle).

[0082] Whenever a work terminal submits a work permit request, the attached work type identifier field (e.g., "Hot Work at Heights") is directly used for indexing and matching in the template library to select the corresponding set of morphological parameters. Based on the selected parameter set and the work location coordinates in the request, the system instantiates the geometry to construct the work's influence space domain. This process typically utilizes a 3D modeling engine to project the morphological parameters onto the coordinate system of the scene containing the work coordinates, generating a volumetric structure centered on that point, and determining its bounding box or constructing a multi-faceted mesh structure to support volume intersection analysis.

[0083] A key characteristic of this construction process is the instability and dynamic nature of the workspace. For example, high-altitude maintenance work may extend laterally due to the use of climbing equipment, while confined space access work involves vertical passageways for the body and equipment. Therefore, the workspace not only reflects physical occupancy but also needs to incorporate estimates of the activity domain during the work process, such as using dynamic buffer coefficients to extend boundaries. To address this need, the system embeds adaptation rules into the template library, such as adjusting the height based on different work platforms or setting a difference model between the "horizontal influence radius" and "vertical risk height" based on the work type. This information is typically generated from historical samples of on-site safety parameters and process specifications, ultimately forming standardized, reusable templates.

[0084] For example, for a request for "hot work on top of container," with the work type being "hot work," the system retrieves a corresponding template defined as "top hemisphere + safety buffer zone," where the hemisphere diameter is set to 3 meters and the outward safety factor is 1.5 meters. If the work coordinates are (102.5, 46.1, 25.0) meters, a three-dimensional work spherical cap is automatically generated relative to the 3D model of the process unit, with that coordinate as the center, a radius of 4.5 meters, and a height of 2.25 meters, defined as the "Work Influence Spatial Domain." This domain then participates in the geometric intersection calculation with the "Media Leakage and Diffusion Risk Domain" on a time-slice basis to determine whether there is an intersection between the work area and the leakage concentration exceeding the limit area at any given time.

[0085] In summary, step S1042 expresses the scope of operational activities using a spatial structure. Its implementation not only accurately reflects the ontological spatial impact of special operations but also introduces spatial computing infrastructure for dynamic permit determination. By introducing a standard operation template system and a geometric abstract modeling mechanism, it effectively bridges the physical / logical gap between traditional static text-based approval and actual operations, making operation permits more standardized, automated, and dynamically responsive.

[0086] S1043. Perform spatiotemporal overlap analysis on the spatial domain of the operation impact and the medium leakage and diffusion risk domain corresponding to each discrete time slice in sequence to obtain the spatiotemporal conflict verification results of each discrete time slice.

[0087] To achieve dynamic safety assessment of work permits, step S1043 needs to be performed. This involves sequentially conducting spatiotemporal overlap analysis on the work impact spatial domain and the media leakage and diffusion risk domain corresponding to each discrete time slice, obtaining the spatiotemporal conflict verification results for each discrete time slice. This step performs fine-grained modeling and temporal analysis of potential interaction scenarios between leakage risk and the work space, and is the core of the decision-making for intelligent control of dynamic work permits.

[0088] To accurately perform spatiotemporal conflict analysis, the system first needs to implement the aforementioned "sequential" analysis mechanism. Because the diffusion of leaked gas exhibits strong temporal evolution, the risk space expands, shifts, or contracts over time. While the location of special operations generally remains stable in the short term, its impact domain persists over time. Therefore, the system divides the entire prediction period (e.g., 30 minutes) into several discrete time slices (which can be cut every 30 seconds or 1 minute). Within each time slice, the system performs geometric relationship reasoning between the fixed "operation impact domain" and the "leakage diffusion risk domain" at that moment.

[0089] Overlap analysis is essentially a three-dimensional Boolean operation problem. In practice, the system represents two spatial domains by constructing polyhedra or voxel structures and uses spatial topology relationship judgment algorithms (such as ST_INTERSECTS or Axis-Aligned Bounding Box based on boundary point set judgment, AABB fast overlap test) to determine whether there is a spatial intersection between the two. If the intersection is not empty, it means that the current operation location and the leakage and diffusion risk present a spatial conflict under this time slice, and it is marked as "conflict"; otherwise, it is marked as "safe".

[0090] To improve analysis efficiency and accuracy, the system often introduces a multi-layer spatial indexing strategy. For example, it constructs a spatial indexing framework based on R-tree or octree structures in a 3D coordinate scene to quickly prune the domain to be analyzed, significantly reducing the Boolean intersection logarithm required for computation. Furthermore, to capture the potential perturbation expansion of the risk domain due to meteorological changes, the system also performs a "virtual boundary expansion" calculation on the risk domain boundary. This involves adding a weak expansion layer (e.g., 3-5%) outside the isosurface of known safe concentrations to prevent missed detection errors caused by model accuracy.

[0091] The system executes the above analysis operations sequentially according to time slices, and records the analysis results of the relationship between the two domains at each time point in real time. Finally, these results are integrated into a "spatial-temporal conflict verification result sequence", where each element represents whether the operation location intersects with the risk domain at a certain discrete time point, thereby further forming a temporal risk map.

[0092] In practical applications, for example, if a temporary hot work application is submitted to start at the southwest corner of an equipment pipe rack, and preliminary anomaly trend analysis predicts a low-speed micro-leakage may occur in a nearby gas branch within the next 15 minutes, the system runs a diffusion model and obtains the risk concentration isosurface at discrete minute intervals. Then, it performs a geometric intersection calculation between the encapsulated volume domain and the hot work spatial domain at each time point. The analysis shows partial overlap between the two during the 7th to 10th minute, and the system generates a "conflict exists in some time periods" flag, suggesting either "delay the start of the work by 10 minutes" or "enhance area ventilation" based on safety regulations.

[0093] Therefore, step S1043 not only provides a static assessment of the relationship between the risk domain and the workspace, but also implements a dynamic, sequential safety risk measurement mechanism. Its role is to make the final determination of whether an actual operation meets safety requirements, and it is also the critical decision point for the entire intelligent permitting system to transform from data perception to risk avoidance. Through this mechanism, the system can make complex risks explicit and sequential, providing safety managers with a clear, quantifiable, and traceable risk control path, and achieving automated permit processing.

[0094] S105. Based on the spatiotemporal conflict verification result, generate the operation control instruction corresponding to the special operation permit request, and control the operation terminal to execute the operation control instruction.

[0095] After completing the time-slice spatiotemporal overlap analysis of the operational impact space domain and the media leakage and diffusion risk domain and obtaining complete spatiotemporal conflict verification results, the system needs to further process these results at the decision level to achieve dynamic control response to special operation permits. The core objective of step S105 is to generate precise operation control instructions based on the spatiotemporal conflict verification results and implement response control through the operation terminal to ensure that the operation process is always within a controllable safety threshold. At this point, if any non-empty intersection time slices are detected, it indicates that there are potential risk events in the future operation period. The system will identify the critical time point when the risk occurs and generate dynamic instructions with safety response content (such as delay, pause, partial release, etc.) accordingly. Conversely, if the analysis results are entirely safe, the system will issue operation permit instructions and automatically configure dynamic interlocking control thresholds based on the operating condition fluctuation trend to enhance the anomaly prevention and control capabilities during operation. Specifically, it may include the following steps: performing a time-series traversal of the spatiotemporal conflict verification results of each discrete time slice; if there is a spatiotemporal conflict verification result that represents a non-empty spatiotemporal intersection, identifying the first target discrete time slice with a non-empty spatiotemporal intersection, determining the time point corresponding to the target discrete time slice as the critical risk moment, and generating the operation control instruction that includes the critical risk moment and the conflict type. If the spatiotemporal conflict verification results of all discrete time slices indicate that the spatiotemporal intersection is empty, a work permit authorization instruction is generated, and the extreme value of the fluctuation in the abnormal evolution parameter is used as a benchmark to configure the dynamic interlock shutdown threshold of the associated process unit in the work cycle corresponding to the special work permit request, and the dynamic interlock shutdown threshold is encapsulated into the work control instruction.

[0096] After completing the overlap analysis of the spatial domain of operational impact and the risk domain of media leakage and diffusion under each discrete time slice and obtaining a set of spatiotemporal conflict verification results, the system enters the decision feedback stage for work permits. This stage requires further extraction of key information from the analysis results and generation of accurate and effective work control instructions. The implementation of this step relies on a time-series traversal processing mechanism for the spatiotemporal conflict verification result sequence. Its underlying logic is to identify the turning point from a safe state to an unsafe state from the dynamic evolution of risk, thereby providing a constraint basis for the automated generation of response strategies.

[0097] During implementation, the system reads and judges the spatiotemporal conflict verification results of each discrete time slice in chronological order. The spatiotemporal conflict verification result is a discrimination value obtained based on Boolean spatial relationships in each time slice, typically represented by Boolean states or multi-level label encoding (such as SAFE, WARNING, DANGER) to indicate the risk overlap judgment result. If a non-empty spatial relationship such as linear volume intersection or boundary tangency is detected between two spatial domains in a certain time slice, the system records that time slice as a candidate risk point.

[0098] When a non-empty intersection is encountered for the first time during the traversal, the system explicitly defines that time slice as the critical risk moment—that is, the moment when the work permit conditions are first violated. This moment is the upper limit of the time during which the work can continue without triggering a safety conflict. The determination of this moment serves as an effective watershed in the work permit judgment model to block the safety window, and its existence ensures that control instructions are immediate and forward-looking.

[0099] Upon identifying a critical risk moment, the system will simultaneously extract the corresponding conflict type information. Conflict type is a risk classification dimension defined when the system constructs the risk domain. For example, the intersection of a heat source's influence area and a volatile gas risk zone falls under the category of thermal ignition-flammable gas conflict; similarly, the ground operation radius of the equipment crossing a combustible cloud area falls under the category of mechanical activity-fire gas cloud conflict. These types are defined in a dictionary format within the operation template library, serving to assist in generating detailed, categorized response strategies.

[0100] Ultimately, the system integrates critical risk moments with corresponding conflict types to construct a structured operation control instruction object. This object includes not only whether operation is permitted, the allowed delay time, or a prohibition notice, but may also include safety measure recommendations (such as temporarily setting up ventilation or shielding measures, or configuring online combustible gas monitoring). This instruction is pushed to the operation terminal, enabling dispatchers and on-site personnel to respond to the system's risk assessment in real time.

[0101] For example, in a case where a hot work permit request was submitted, the system detected that the hemispherical thermal impact space created by the operation would overlap with the risk domain of butane gas released from the underground tank area starting at the 8th minute, according to the predicted diffusion model. After time-series traversal analysis, the system determined the 8th minute as the critical risk moment and identified the conflict type as thermal operation-combustible gas phase overlap. The final automatically generated operation control instruction was: hot work is prohibited for 8 minutes; it is recommended to set up an upwind ventilation system and deploy a portable LEL gas detection device; the re-application period for the permit should be after the 12th minute of the predicted timeline.

[0102] Therefore, through the systematic analysis of the full-time series risk situation and the extraction of key information, this step realizes the closed-loop response from analysis to control, ensuring that the work permit not only has reasonable results, but also has time accuracy and risk classification basis, greatly improving the intelligent response level of safety work and the industrial collaborative execution efficiency.

[0103] After the spatio-temporal overlap analysis under each discrete time slice is completed, the system will form a complete sequence of spatio-temporal conflict verification results, where each time slice corresponds to a clear judgment result of the spatial relationship. At this time, if it is found through sequential traversal that there is no spatial intersection between the operation influence spatial domain and the medium leakage diffusion risk domain in all time slices, it means that the current operation request location does not involve any potential medium leakage influence range within the entire preset prediction window (such as the next 30 minutes), and there are sufficient safety working conditions. Therefore, the system automatically generates a work permit authorization instruction, allowing the special operation to be carried out at this location and time period.

[0104] This process does not take non-conflict as the sole sufficient condition for work permit, but on the basis of global risk-free, adds a dynamic process control mechanism to further enhance the working condition perception ability and linkage response ability during operation execution. To achieve this goal, the system will introduce a dynamic interlock shutdown threshold mechanism. The core of this mechanism is to use the abnormal evolution parameters obtained in the aforementioned modeling process to pre-predict potential fluctuations during the operation process and configure automatic response thresholds accordingly.

[0105] The abnormal evolution parameter is a numerical expression of the future state change obtained by the system through historical working condition pattern matching and trend extrapolation. It contains the trend information on how the equipment fluctuation amplitude evolves. The maximum amplitude index is the fluctuation extreme value, representing the upper limit of the most likely process disturbance during the prediction period. Considering this value as the linkage reference benchmark for shutdown interlock has two aspects: First, it comes from the credible data window formed by historical evolution rules, with good practical feasibility and reference; second, it has both time sensitivity and adaptability and can be adjusted according to the changes in prediction behavior within the permit window. Therefore, it is more flexible and responsive than traditional static threshold libraries (such as fixed alarm values in PLC).

[0106] In practice, the system will set a set of linkage monitoring rules for the workspace in the data structure of the authorization instruction, including but not limited to: configuring sensor monitoring parameters (temperature, pressure, VOC concentration), real-time alarm levels, and shutdown linkage control logic; while the dynamic interlock shutdown threshold is injected into the linkage rule template as a dynamic parameter input, ensuring that once any sensed value exceeds the threshold in future actual working conditions, the corresponding area control, exhaust linkage, air supply lock-up, or equipment shutdown action will be triggered immediately. This combined authorization mechanism of "safety authorization + dynamic monitoring + real-time intervention" ensures that the operation not only starts safely, but is also under closed-loop monitoring throughout the entire process.

[0107] For example, in a temporary confined space cleaning operation permit request submitted by a factory, system analysis revealed that the operation location was completely outside the leakage risk zone for the next 30 minutes, and no spatial conflicts were found during the simulation. Therefore, the system immediately generated an operation permit instruction. Simultaneously, based on the historical failure modes of the adjacent variable pressure scrubbing tower, the system predicted a potential upward trend in pressure fluctuations, with the peak pressure fluctuation expected to occur at 0.35 MPa between the 13th and 20th minutes of the evolution curve. Simultaneously with the permit authorization, the system configured this 0.35 MPa value as a dynamic interlock shutdown threshold and set up continuous real-time pressure monitoring of the pipeline section of the cleaning subunit during the operation period. If a sudden event exceeding this value occurs, the system automatically triggers a shutdown interlock and alarm, ensuring steady-state operation of the permitted scenario under continuously verified safety conditions.

[0108] Therefore, this step not only concludes the assessment of the safe and workable state but also marks the beginning of the full-lifecycle control phase of the work permit. Extending the prediction results into execution control conditions is the core guarantee for this dynamic permit management method to achieve an intelligent safety closed loop.

[0109] Please see Figure 3 This is a schematic diagram of the structure of a dynamic permit control system for safe production operations in an embodiment of this application.

[0110] It should be noted that, Figure 3 The structure of the dynamic permit control system for safe production operations shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0111] like Figure 3As shown, a dynamic permit control system for safe production operations includes a central processing unit 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory 302 or a program loaded from a storage section 308 into a random access memory 303, such as executing the methods described in the above embodiments. The random access memory 303 also stores various programs and data required for system operation. The central processing unit 301, the read-only memory 302, and the random access memory 303 are interconnected via a bus 304. An input / output interface 305 is also connected to the bus 304.

[0112] The following components are connected to the input / output interface 305: an input section 306 including audio input devices, push-button switches, etc.; an output section 307 including an LCD display, audio output devices, indicator lights, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card, modem, etc. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the input / output interface 305 as needed. A removable medium 311, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 310 as needed so that computer programs read from it can be installed into the storage section 308 as needed.

[0113] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing computer programs for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311. When the computer program is executed by central processing unit 301, it performs the various functions defined in the present invention.

[0114] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory, flash memory, optical fiber, portable compact disk read-only memory, optical storage devices, magnetic storage devices, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0115] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those shown in the drawings.

[0116] Specifically, the dynamic licensing and control system for safe production operations in this embodiment includes a processor and a memory. The memory stores a computer program, and when the computer program is executed by the processor, it implements the dynamic licensing and control method for safe production operations provided in the above embodiment.

[0117] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in a dynamic licensing and control system for safe production operations described in the above embodiments; or it may exist independently and not assembled into the dynamic licensing and control system for safe production operations. The storage medium carries one or more computer programs, which, when executed by a processor of the dynamic licensing and control system for safe production operations, cause the system to implement the dynamic licensing and control method for safe production operations provided in the above embodiments.

Claims

1. A dynamic permitting and control method for safe production operations, characterized in that, The method includes: In response to a special operation permit request initiated by the work terminal, the work location coordinates and work type identifier in the special operation permit request are extracted. Based on a preset spatial index relationship table, the associated process units within a preset radius centered on the work location coordinates are determined, and the time-series data of the operating status of the associated process units at the current moment, as well as the real-time environmental meteorological data of the area where the work location coordinates are located are obtained. The time-series data of the operating status is matched with the pre-stored historical operation and maintenance database using a sliding window similarity matching method to obtain the target historical operating condition segment corresponding to the special operation permit request and the evolution trend data of the target historical operating condition segment on the time axis. The historical operation and maintenance database contains historical operation result tags corresponding to each historical operating condition segment. Based on the historical operation result label, the evolution trend data and the operation status time series data corresponding to the target historical operating condition segment, calculate the abnormal evolution parameters of the associated process unit within a future preset time period; Based on the real-time environmental meteorological data, the abnormal evolution parameters, the operation location coordinates, and the operation type identifier, a spatiotemporal conflict analysis is performed to obtain the spatiotemporal conflict verification results. Based on the spatiotemporal conflict verification results, the operation control instruction corresponding to the special operation permit request is generated, and the operation terminal is controlled to execute the operation control instruction.

2. The method according to claim 1, characterized in that, The step of performing sliding window similarity matching between the operational status time-series data and the pre-stored historical operation and maintenance database to obtain the target historical operating condition segment corresponding to the special operation permit request and the evolution trend data of the target historical operating condition segment on the time axis specifically includes: Discrete difference processing is performed on the time series data of the running state to obtain the real-time morphological change rate sequence. Based on multiple different preset time scaling factors, the real-time morphological change rate sequence is resampled along the time axis to obtain the candidate feature sequence corresponding to each preset time scaling factor. Each of the candidate feature sequences to be tested is dynamically path-normalized and matched in multiple historical operating condition records in the pre-stored historical operation and maintenance database. For the first historical operating condition record, the cumulative alignment distance between each of the candidate feature sequences to be tested and the first historical operating condition record is calculated. The first historical operating condition record is any historical operating condition record. The historical working condition record corresponding to the minimum cumulative alignment distance is determined as the target historical working condition segment; The evolution trend data is determined based on the target historical operating condition segments.

3. The method according to claim 2, characterized in that, The determination of the evolution trend data based on the target historical operating condition segment specifically includes: From the above The original operating condition data within a preset prediction time window after the last cutoff time of the target historical operating condition segment is obtained from the pre-stored historical operation and maintenance database. The start time of the preset prediction time window is the same as the last cutoff time of the target historical operating condition segment. The original operating condition data is mapped inversely to a time scale based on a target preset time scaling factor to obtain the evolution trend data. The target preset time scaling factor is the preset time scaling factor used when determining the target historical operating condition segment.

4. The method according to claim 3, characterized in that, The calculation of abnormal evolution parameters of the associated process unit within a preset future time period based on the historical operation result tags corresponding to the target historical operating condition segment, the evolution trend data, and the operating status time series data specifically includes: Based on the anomaly type in the historical operation result label, the parameter change pattern corresponding to the anomaly type is determined from the target historical operating condition segment and the evolution trend data, and an anomaly precursor sequence is generated. Based on the target preset time scaling factor, the time scale of the preceding symptom sequence is adjusted to obtain the target search template, and the target search template is used for sliding similarity matching in the running state time series data. The moment when the feature similarity is greater than the preset similarity threshold is determined as the potential abnormal starting point of the current working condition. Using each potential anomaly initiation point as a reference time, the current anomaly verification window is determined in the running state time series data, and the real-time fluctuation intensity within the current anomaly verification window is calculated. The ratio of the real-time fluctuation intensity to the historical fluctuation intensity of the abnormal precursor sequence in the corresponding historical period is calculated to obtain the intensity correction coefficient; The amplitude of the evolution trend data is corrected based on the intensity correction coefficient to obtain the abnormal evolution parameters.

5. The method according to claim 1, characterized in that, Based on the real-time environmental meteorological data, the anomaly evolution parameters, the work location coordinates, and the work type identifier, a spatiotemporal conflict analysis is performed to obtain spatiotemporal conflict verification results, specifically including: Based on the real-time environmental meteorological data and the abnormal evolution parameters, the medium leakage and diffusion risk domain corresponding to each discrete time slice of the associated process unit within a preset time period is calculated. The medium leakage and diffusion risk domain is used to characterize the spatial distribution range of the hazardous medium under the discrete time slice. The spatial morphological parameters corresponding to the job type identifier are called from the preset job geometry template library, and the job influence spatial domain is constructed based on the job position coordinates. Spatiotemporal overlap analysis is performed sequentially on the spatial domain of the operation's impact and the media leakage and diffusion risk domain corresponding to each discrete time slice to obtain the spatiotemporal conflict verification results of each discrete time slice.

6. The method according to claim 5, characterized in that, The step of calculating the medium leakage and diffusion risk domain corresponding to each discrete time slice within a preset time period for the associated process unit based on the real-time environmental meteorological data and the abnormal evolution parameters specifically includes: The instantaneous evolution values ​​corresponding to each discrete time slice are extracted from the abnormal evolution parameters, and the medium leakage mass rate of each discrete time slice is calculated based on the instantaneous evolution values ​​and the physical properties of the associated process unit using a preset leakage source term model. The leakage mass rate of the medium and the real-time environmental meteorological data are input into a preset gas diffusion and transport model to obtain the three-dimensional spatial concentration distribution field of the hazardous medium under the discrete time slice. In the three-dimensional spatial concentration distribution field, a spatial isosurface is determined where the concentration value is equal to the preset safe concentration threshold corresponding to the hazardous medium; The closed three-dimensional region enclosed by the spatial isosurface is defined as the medium leakage and diffusion risk domain.

7. The method according to claim 5, characterized in that, The step of generating the operation control instruction corresponding to the special operation permit request based on the spatiotemporal conflict verification result specifically includes: The spatiotemporal conflict verification results of each discrete time slice are traversed in time sequence. If there is a spatiotemporal conflict verification result that represents a non-empty spatiotemporal intersection, the first target discrete time slice with a non-empty spatiotemporal intersection is identified, the time point corresponding to the target discrete time slice is determined as the critical risk moment, and the operation control instruction containing the critical risk moment and the conflict type is generated. If the spatiotemporal conflict verification results of all discrete time slices indicate that the spatiotemporal intersection is empty, a work permit authorization instruction is generated, and the extreme value of the fluctuation in the abnormal evolution parameter is used as a benchmark to configure the dynamic interlock shutdown threshold of the associated process unit in the work cycle corresponding to the special work permit request, and the dynamic interlock shutdown threshold is encapsulated into the work control instruction.

8. A dynamic permitting and control system for safe production operations, characterized in that, The dynamic permit control system for safe production operations includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to cause the dynamic permit control system for safe production operations to perform the method as described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that, When the instruction is executed on the dynamic permit control system for safe production operations, the dynamic permit control system for safe production operations performs the method as described in any one of claims 1-7.

10. A computer program product, characterized in that, When the computer program product is run on the dynamic licensing and control system for safe production operations, the dynamic licensing and control system for safe production operations performs the method as described in any one of claims 1-7.

Images