A high-risk operation whole-process management and control and safety prevention and control method and system

Abstract

This invention provides a method and system for full-process control and safety prevention of high-risk operations, belonging to the field of digital management technology for high-risk operations. By forming a set of work units from work tickets, device connection relationships, and actual isolation status in the high-risk operation management platform, a spatiotemporal conflict diagram of cross-operations is constructed. This allows for a unified representation of the time overlap, equipment coupling, type coupling, and isolation deviation relationships between different operations within the same graph structure. Through graph risk propagation processing, the risk of a single operation and the propagation and coupling effects of risk among multiple operations are calculated, thereby obtaining a risk vector and risk value of operation clusters that are more consistent with the on-site cross-operation scenario. Combined with collaborative optimization, the operation execution level, isolation enhancement level, and approval upgrade identifier are jointly solved, ultimately forming a high-risk operation full-process control and safety prevention result that can be directly used for system execution.

Description

Technical Field

[0001] This invention relates to the field of digital management technology for high-risk operations, specifically to a method and system for full-process control and safety prevention of high-risk operations. Background Technology

[0002] In industrial production sites, high-risk operations such as hot work, confined space work, work at height, hoisting operations, and temporary electrical work are usually not isolated but occur simultaneously within the same equipment area, adjacent equipment areas, or the same maintenance window. While existing high-risk operation management systems can achieve online approval of work permits, qualification verification, status recording, and some alarm notifications, most systems still follow a technical approach of processing each work permit independently. This means that each work permit is approved separately, its risk is assessed separately, and control conclusions are generated separately.

[0003] In practical applications, the above methods are suitable for single-task or loosely coupled scenarios. However, when multiple high-risk tasks overlap in time, are geographically adjacent, are connected on equipment, and affect each other in terms of isolation requirements, the single-ticket independent processing method cannot accurately identify combined risks. Especially when hot work and confined space work are carried out simultaneously, multiple work tickets are executed synchronously on the same equipment link, and some isolation measures are not fully implemented while adjacent high-risk tasks are carried out in parallel, traditional methods can usually only determine whether a certain task meets the conditions of the ticket from the perspective of a single ticket, but cannot answer the questions of whether the overall risk increases after multiple tasks are superimposed, through what relationship the risk is transmitted, and how to coordinate and schedule them.

[0004] Furthermore, even when existing technologies incorporate regional or temporal conflict detection, they often rely on static rules, such as limiting simultaneous operations in the same area or restricting the number of operations within the same time period. While these solutions provide basic constraints in general scenarios, they lack sophisticated data processing capabilities for special scenarios where devices have implicit connectivity, different operation types have varying degrees of coupling, and isolation deviations amplify cross-risks. Consequently, they struggle to achieve true end-to-end safety control.

[0005] Therefore, how to perform structured modeling, risk propagation analysis, and collaborative optimization control of cross-high-risk operations based solely on a small amount of key business data already available in the high-risk operation management platform, without relying on a complex multi-source sensing system, in order to solve the problem that traditional single-ticket independent approval cannot effectively identify and control cross-combination risks, remains a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0006] The purpose of this invention is to provide a method for full-process management and safety control of high-risk operations, so as to at least solve the problem that the traditional single-ticket independent approval method is difficult to effectively identify and control combined risks in scenarios where multiple high-risk operations are carried out simultaneously.

[0007] Technical solution

[0008] To achieve the above objectives, the first aspect of the present invention provides a method for full-process management and safety control of high-risk operations, the method comprising the following steps:

[0009] Acquire work order data, device connection relationship data, and actual isolation status data from the high-risk operation management platform to construct a high-risk operation process control information group; wherein, the construction of the high-risk operation process control information group includes a set of operation units, a device connection matrix, and an isolation deviation coefficient;

[0010] Based on the high-risk operation process control information group, a cross-operation spatiotemporal conflict diagram is constructed, and the initial risk vector of each node corresponding to each operation unit is calculated.

[0011] Based on the spatiotemporal conflict graph of the cross-operation and the initial risk vector of the node, the risk propagation process of the graph is performed to obtain the propagated risk vector and the risk value of the operation cluster corresponding to each operation unit.

[0012] Based on the propagated risk vector and the job cluster risk value, a collaborative optimization solution is performed to obtain the job execution association information corresponding to each job unit; wherein, the job execution association information includes execution level, isolation enhancement level, and approval escalation identifier;

[0013] The results of full-process management and safety control for high-risk operations are generated based on the execution level, the isolation enhancement level, and the approval upgrade identifier.

[0014] To achieve the above objectives, a second aspect of the present invention provides a high-risk operation full-process control and safety prevention system, the system being used to execute the above-described high-risk operation full-process control and safety prevention method, the system comprising:

[0015] A construction unit is used to acquire work order data, device connection relationship data, and actual isolation status data from the high-risk operation management platform to construct a high-risk operation process control information group; wherein, the construction of the high-risk operation process control information group includes a set of operation units, a device connection matrix, and an isolation deviation coefficient;

[0016] The calculation unit is used to construct a spatiotemporal conflict diagram of cross-operations based on the high-risk operation process control information group, and to calculate the initial risk vector of the node corresponding to each operation unit.

[0017] The propagation unit is used to perform graph risk propagation processing based on the cross-operation spatiotemporal conflict graph and the node initial risk vector to obtain the propagated risk vector and the risk value of the operation cluster corresponding to each operation unit.

[0018] The solution unit is used to perform collaborative optimization based on the propagated risk vector and the job cluster risk value to obtain job execution association information corresponding to each job unit; wherein, the job execution association information includes execution level, isolation enhancement level and approval escalation identifier;

[0019] The generation unit is used to generate the full-process control and safety prevention results of high-risk operations based on the execution level, the isolation enhancement level, and the approval upgrade identifier.

[0020] Beneficial effects

[0021] Through the above technical solutions, this invention proposes a method and system for full-process management and safety control of high-risk operations. By forming a set of work units usable for calculation from work tickets, device connection relationships, and actual isolation states in the high-risk operation management platform, a spatiotemporal conflict graph of cross-operations is constructed. This allows for a unified representation of the temporal overlap, equipment coupling, type coupling, and isolation deviation relationships between different operations within the same graph structure. Through graph risk propagation processing, not only can the risk of a single operation be calculated, but also the propagation and coupling effects of risk across multiple operations can be calculated. This yields a risk vector and operation cluster risk value that better reflect the on-site cross-operation scenario. Combined with collaborative optimization, the operation execution level, isolation enhancement level, and approval upgrade identifier are jointly solved, ultimately forming a high-risk operation full-process management and safety control result that can be directly used for system execution. Therefore, the implementation of this invention can elevate the traditional static approval of single operations to a dynamic, collaborative, and computable safety control process oriented towards cross-operation scenarios.

[0022] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0023] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings:

[0024] Figure 1 This is a flowchart of the steps of a high-risk operation full-process management and safety control method provided by one embodiment of the present invention.

[0025] Figure 2 This is a system structure diagram of a high-risk operation full-process control and safety prevention system provided by one embodiment of the present invention. Detailed Implementation

[0026] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0027] like Figure 1 As shown, this invention provides a method for full-process management and safety control of high-risk operations, the method comprising the following steps:

[0028] S10: Obtain work ticket data, device connection relationship data, and actual isolation status data from the high-risk operation management platform, and construct a high-risk operation process control information group; wherein, the construction of the high-risk operation process control information group includes a set of operation units, a device connection matrix, and an isolation deviation coefficient.

[0029] Specifically, the process involves acquiring work order data, device connection relationship data, and actual isolation status data from a high-risk operation management platform. Each high-risk operation record in the work order data is then analyzed. Based on the operation characteristics revealed in the analysis results, each high-risk operation record is converted into an operation unit, forming a set of operation units. These operation characteristics include operation type, operation area, associated device set, planned start time, planned end time, basic risk level, required isolation item set, and execution status information. A device connection matrix is ​​constructed based on the device connection relationship data. The actual isolation item set corresponding to each operation unit in the operation unit set is determined based on the actual isolation status data, and the isolation deviation coefficient corresponding to each operation unit is calculated. Finally, a high-risk operation process control information group is constructed based on the operation unit set, the device connection matrix, and the isolation deviation coefficient.

[0030] In this embodiment of the invention, step S10 is used to establish a unified data foundation shared by all subsequent processing procedures. It should be noted that in scenarios involving overlapping high-risk operations, if each work order is still considered merely a business form within the approval flow, it becomes difficult to directly perform unified calculations on the cross-relationships between multiple operations. Therefore, in this embodiment of the invention, the key business data in the high-risk operation management platform is first converted into a standardized set of work units, enabling subsequent relationship mapping, risk propagation, and collaborative problem-solving to all be performed under the same data representation.

[0031] In essence, the acquired data falls into three main categories. The first is work order data, reflecting the work type, work area, associated equipment, planned start and end times, basic risk level, required isolation measures, and current execution status for each high-risk operation. The second is equipment connection relationship data, reflecting whether there are process connectivity, functional relationships, or media influence relationships between different devices. The third is actual isolation status data, reflecting the isolation measures already implemented for each operation at the current moment.

[0032] Furthermore, in one executable implementation, each high-risk work record in the work ticket data can be parsed, sequentially extracting the work type, work area code, associated equipment list, planned start time, planned end time, basic risk level, required isolation item list, and execution status identifier, and converting it into a work unit. By repeating this conversion process on all work ticket records, a set of work units can be formed. Each work unit in the set corresponds to a high-risk work object to be analyzed, and each node in the subsequently constructed cross-work spatiotemporal conflict graph is directly mapped from a work unit.

[0033] In actual operation, the work type can include hot work, confined space work, work at height, hoisting work, temporary electrical work, etc.; the work area can use a unit area code, a cell area code, or a space identifier used within the enterprise; the list of associated equipment can be a set of equipment numbers for pumps, tanks, towers, containers, pipelines, valves, etc.; the basic risk level can be directly read from the enterprise's existing ticket management rules; and the execution status identifier can be used to indicate whether the work is currently pending, in progress, suspended, or closed. By uniformly extracting the above fields, different types of work tickets can be compared in subsequent processing.

[0034] In this embodiment of the invention, the device connection relationship data also needs to be constructed into a device connection matrix. Specifically, all devices participating in this round of analysis can be numbered according to a unified rule, and then the device connection relationship data can be used to determine whether there is a direct process connection, indirect influence, or functional coupling relationship between any two devices. If such a relationship exists, a connection identifier is recorded at the corresponding position in the connection matrix; if not, a non-connection identifier is recorded. In this way, subsequent steps are no longer limited to determining whether operations affect each other based on the work area, but can further identify implicit cross relationships on the device link. That is to say, even if two operations are not completely overlapping in spatial location, as long as there is a connection relationship between their associated devices, the subsequent model can still identify potential risk propagation channels between them.

[0035] Based on this, in this embodiment of the invention, it is also necessary to determine the isolation deviation result corresponding to each work unit according to the actual isolation status data. Specifically, the list of isolation items required to be implemented in the work order for each work can be compared with the list of actual isolation items that have been implemented so far, and the proportion of unimplemented isolation items in the required isolation items can be calculated to form an isolation deviation coefficient. It should be noted that this isolation deviation coefficient is used to characterize the degree of deviation between the current work and the expected requirements in terms of the implementation of isolation measures. If a work requires a large number of isolation items to be implemented, but the actual number of implemented isolation items is significantly insufficient, its isolation deviation coefficient will increase accordingly; conversely, if a work has fully implemented all the isolation requirements in the work order, its isolation deviation coefficient will be close to zero.

[0036] In practice, items such as blind flange isolation, valve closure, energy lockout, media venting, ventilation measures, and area isolation specified in the same work order can be uniformly converted into standardized isolation item codes. Then, the implemented isolation items reported on-site are also coded in the same way, and missing items are identified by set differences. This method not only reduces the impact of textual differences caused by different work order filling methods, but also allows isolation deviation results to be directly used in subsequent model calculations.

[0037] Furthermore, in one specific implementation, five high-risk operations within a certain maintenance window can be selected as the analysis objects, including two hot work operations, one confined space operation, one high-altitude operation, and one temporary power supply operation. The system first reads the corresponding five work tickets from the high-risk operation management platform, extracts the aforementioned five types of field information, and converts them into five work units. Subsequently, it reads the connection relationships between the nine devices associated with the above five operations from the device management system, establishing a 9×9 device connection matrix. Further, it reads the actual isolation implementation status of each operation from the isolation execution module, finding that one hot work operation lacks an energy isolation confirmation, and another confined space operation lacks an interlock release confirmation. Through the above processing, unified input data required for subsequent diagram construction and risk calculation can be generated.

[0038] S20: Construct a spatiotemporal conflict diagram for cross-operations based on the high-risk operation process control information group, and calculate the initial risk vector of each node corresponding to each operation unit.

[0039] Specifically, for any two work units in the set of work units, a time overlap coefficient is calculated based on the planned start time and planned end time of each work unit; a device coupling coefficient is calculated based on the device connection matrix and the associated device set corresponding to any two work units; a type coupling factor is determined based on the work type corresponding to any two work units, and the time overlap coefficient, the device coupling coefficient, the type coupling factor, and the corresponding isolation deviation coefficient are fused to obtain the edge weight; and a spatiotemporal conflict graph of cross-operation is constructed based on the edge weight.

[0040] Following this, based on the planned start and end times of each work unit in the work unit set, the planned duration of each work unit is calculated and the duration factor is determined; based on the basic risk level, isolation deviation coefficient, duration factor, and execution status factor of each work unit, the initial risk value of each work unit is calculated; the initial risk values ​​of each work unit are summarized to form the node initial risk vector.

[0041] In this embodiment of the invention, step S20 is used to complete two key processes: first, to construct the relationship structure between multiple work units; and second, to calculate the risk starting point of each work unit. It should be noted that a set of work units alone is insufficient to identify cross-risks, because cross-risks are essentially the superposition and propagation of relationships between nodes. Similarly, if only the relationships between nodes are known without considering the risk strength of the nodes themselves, it is impossible to further determine which work units are more likely to become risk sources. Therefore, this step calculates both the relationships and the risk starting point.

[0042] Firstly, regarding the construction of the spatiotemporal conflict graph for cross-tasks, in this embodiment of the invention, for any two task units, it is prioritized to determine whether they overlap in the time dimension. Specifically, based on the planned start and end times of the two task units, the proportion of time they are jointly implemented within the same time window can be identified. If the time windows of the two tasks overlap significantly, it indicates that they are more likely to coexist in actual execution, thus making it easier to form combined risks; if the two tasks are located in the same area but their planned times are significantly different, the degree of direct overlap between them will decrease. Through this time overlap processing, the key relationship of whether tasks occur simultaneously can be quantified as a component in the subsequent edge weight calculation.

[0043] Secondly, regarding the equipment relationship dimension, in this embodiment of the invention, the equipment connection matrix obtained in step S10 is compared with the set of associated equipment corresponding to each work unit. If there are many connection edges between the equipment objects associated with two work units, it indicates that even if these two work units are not located on the same plane, they may still influence each other through equipment links, media flow direction, or functional coupling relationships. Conversely, if there are almost no connection relationships between the associated equipment of two work units, their equipment coupling degree will be low. Through this processing, implicit process association risks that are difficult to detect in traditional management methods can be explicitly incorporated into the graph model.

[0044] Furthermore, in terms of work type, this embodiment of the invention also determines a type coupling factor based on the actual coupling strength between different work types. Specifically, a type relationship table can be pre-established to reflect the degree of coupling between different combinations of work types. For example, hot work and confined space work typically have a strong risk coupling because fluctuations in flammable gas and oxygen content or changes in ventilation within a confined space can significantly amplify the consequences of hot work; while high-altitude work and temporary electrical work also have overlapping effects, their coupling methods and degree of influence are usually different. By introducing a type coupling factor, the edges in the graph can reflect not only whether two work types are adjacent, but also whether these two types of work are naturally more likely to form high-consequence combination risks in the current industrial scenario.

[0045] Based on this, in this embodiment of the invention, the time overlap result, device coupling result, type coupling result, and isolation deviation result are fused to generate a comprehensive edge weight between any two work units. Furthermore, in an executable implementation, the edge weight can be represented as follows:

[0046] ;

[0047] in, This represents the edge weight between the i-th and j-th work units. Indicates the degree of time overlap. This indicates the degree of coupling between the two devices. This represents the type coupling factor between the two. and This represents the isolation deviation coefficient between the two. , , and This represents the weight parameters of each component. It's important to note that the fundamental principle of this formula lies in compressing the originally dispersed four types of intersecting factors onto the same comparable weight scale. The larger the resulting edge weight, the stronger the overall overlap between the two tasks.

[0048] Subsequently, when the edge weights meet the preset edge-building conditions, graph edges are established between the two work units, ultimately forming a spatiotemporal conflict graph for cross-operations. The nodes in this graph originate from the work units in step S10, and the edges originate from the comprehensive cross-relationships calculated in step S20. In this way, multiple work tickets that were originally scattered can be transformed into a computable cross-operation network.

[0049] On the other hand, regarding the calculation of the initial risk vector for a node, in this embodiment of the invention, the duration of the operation is first calculated based on the planned time window corresponding to each work unit, and the duration is normalized into a duration factor. It should be noted that the longer the operation duration, the longer the period of exposure to the risk environment, and the more opportunities there are for it to intersect with other operations. Therefore, the duration should be included in the calculation as part of the node's own risk.

[0050] Subsequently, based on the basic risk level, isolation deviation coefficient, duration factor, and execution status factor, the initial risk value for each work unit is calculated. Furthermore, in one executable implementation, the initial risk value can be expressed as:

[0051] ;

[0052] in, This represents the initial risk value of the i-th work unit. Indicates the basic risk level. Indicates the isolation deviation coefficient. Indicates duration factor, Indicates the execution status factor. , , and This indicates the corresponding weight. Using this formula, the isolation deviation results obtained in step S10, the basic risk level read from the work ticket, and the duration factor generated in this step can be uniformly written into the node's own risk expression.

[0053] In actual operation, the execution status factor can reflect information such as whether a task has started, is paused, or is nearing completion. Since tasks in progress have a more direct impact on the actual site, their status factor is usually higher than that of tasks that have not yet started; paused tasks fall somewhere in between. This design allows the model to more closely reflect the dynamic state of the site, rather than relying solely on static ticket information for analysis. After obtaining the initial risk value for each task unit, they are summarized in node number order to form the node initial risk vector.

[0054] In one specific implementation, it is assumed that among the five operations, two hot work operations have a high basic risk level, with one of them having an initial risk value higher than the other nodes due to being in progress and having a large isolation deviation; confined space operations, although having a similar basic risk level to hot work operations, have a higher duration factor due to their longer planned duration; high-altitude operations have a lower basic risk, but due to significant time overlap with hoisting operations, they may still become a critical node in subsequent graph propagation. After this step, the system obtains both a spatiotemporal conflict graph of cross-operations among the five nodes and the initial risk vectors corresponding to the five nodes.

[0055] S30: Based on the spatiotemporal conflict diagram of the cross-operation and the initial risk vector of the node, perform graph risk propagation processing to obtain the propagated risk vector and the risk value of the operation cluster corresponding to each operation unit.

[0056] It should be noted that before performing risk propagation processing based on the cross-operation spatiotemporal conflict graph and the node initial risk vector, the process further includes: constructing a scenario enhancement propagation coefficient based on the time overlap coefficient, type coupling factor, and isolation deviation coefficient between any two operation units; constructing a normalized propagation weight based on the edge weights in the cross-operation spatiotemporal conflict graph and the scenario enhancement propagation coefficient; and writing the normalized propagation weight into the cross-operation spatiotemporal conflict graph.

[0057] Subsequently, using the initial risk vector of the node as the initial input, iterative risk propagation is performed based on the normalized propagation weight. The iteration stops when the results of two adjacent iterations meet the convergence condition, and the converged node risk value is recorded as the propagated risk value. The propagated risk values ​​corresponding to each job unit are summarized to form the propagated risk vector. The job cluster risk value is calculated based on the propagated risk value and the intra-cluster normalized propagation weight.

[0058] In this embodiment of the invention, step S30 aims to address the problem that traditional high-risk operation management methods can only identify individual operation risks and struggle to identify risks arising from cross-combinations. It should be noted that in scenarios where multiple high-risk operations are carried out simultaneously, the actual risk of a particular operation is not entirely equal to its inherent risk, but is influenced by the risk input from adjacent operations. For example, although a confined space operation may have been approved, if a hot work operation is being carried out concurrently in an adjacent equipment area, and the equipment connectivity and isolation gaps provide conditions for risk propagation, then the overall risk of the confined space operation in actual execution will be amplified. Traditional single-ticket methods cannot reflect this phenomenon of risk amplification at nodes due to surrounding operations, while graph risk propagation processing is precisely designed to solve this problem.

[0059] Using the spatiotemporal conflict graph of cross-operations and the initial risk vector as input, the scenario enhancement propagation coefficient is further constructed based on the initial edge weights, and normalized propagation weights are generated. Then, using the initial risk vector as the zeroth round propagation input, multiple rounds of risk propagation are performed on each operation node according to the iterative propagation formula until the convergence condition is met, so as to obtain the propagated risk vector.

[0060] In this embodiment of the invention, enhanced risk propagation capability is first constructed based on the graph edge weights obtained in step S20. Specifically, it considers not only the comprehensive edge weights between two work units, but also their propagation enhancement factors in the current scenario, particularly the degree of temporal overlap, the degree of work type coupling, and the degree of isolation deviation. In other words, if two work units not only have general conflict edges, but also have long temporal overlap, strong type coupling, and large isolation gaps, the risk propagation capability between these two work units will be further enhanced. Through this processing, the graph propagation model can become a targeted risk propagation model adapted to high-risk work intersection scenarios.

[0061] Furthermore, in one executable implementation, the scenario enhancement propagation coefficient can be expressed as:

[0062] ;

[0063] in, This represents the scene enhancement propagation coefficient between the i-th job unit and the j-th job unit. Indicates the degree of time overlap. Indicates type coupling factor, and This represents the corresponding isolation deviation coefficient. , and This indicates an enhancement parameter. Based on the same structural edge weights, this coefficient grants a higher propagation capability to combinations of operations that are more likely to experience risk linkages in reality.

[0064] Subsequently, the edge weights from step S20 are combined with the scene enhancement propagation coefficients, and the input weights of each node are normalized. This ensures that the propagation influence from different neighboring nodes can participate in the iteration on a uniform scale. This normalization process does not change the pattern of risk magnitude but is used to prevent propagation instability caused by excessively high total input when a node has many adjacencies. Through this process, subsequent graph risk propagation can converge more smoothly to a result reflecting the equilibrium state of cross-risks.

[0065] After preparing the propagation weights, in this embodiment of the invention, the initial risk vector of the node obtained in step S20 is used as the propagation starting point to perform multiple rounds of iterative propagation on the spatiotemporal conflict graph of cross-operation tasks. The basic meaning is that in each iteration, the new risk of a node consists of two parts: one is the node's original initial risk, and the other is the external risk input by all neighboring nodes according to the propagation weights. Thus, in the first round of propagation, nodes with high initial risk will release their risk impact to neighboring nodes; in the second round and subsequent rounds of propagation, nodes whose risk has been increased will continue to transmit their impact to their neighboring nodes. After multiple iterations, the risk in the entire cross-operation network will gradually stabilize.

[0066] Furthermore, in one executable implementation, graph risk propagation can be represented as:

[0067] ;

[0068] in, This represents the risk value of the i-th job unit in the (k+1)-th iteration. This represents the initial risk value of the i-th work unit. Let N(i) represent the propagation attenuation coefficient, and let N(i) represent the set of adjacent job units that conflict with the i-th job unit. This represents the normalized propagation weight from the j-th job unit to the i-th job unit. This represents the risk value of the j-th work unit in the k-th iteration. It should be noted that the first part of this formula is used to retain the node's own risk, while the second part is used to introduce the input risk of adjacent nodes. Therefore, it can reflect both the inherent danger of the node and the amplification effect brought about by surrounding cross-operations.

[0069] In practical applications, when the change between the results of two adjacent rounds of propagation is lower than a preset threshold, the propagation can be considered to have converged. At this point, the risk value corresponding to each node is the post-propagation risk value. By summarizing all post-propagation risk values ​​in node order, the post-propagation risk vector can be obtained. Compared with the initial node risk vector, the post-propagation risk vector better reflects the true risk pattern in cross-operation scenarios.

[0070] Furthermore, in this embodiment of the invention, not only is the post-propagation risk of a single node calculated, but also the connected substructures or high-weight dense substructures in the graph are aggregated for analysis to form a cluster risk value. It should be noted that in on-site management, it is often necessary to simultaneously determine whether a group of overlapping operations requires coordinated handling as a whole, rather than processing individual nodes separately. Therefore, this embodiment of the invention further comprehensively calculates the node risk and edge coupling strength within the cluster to obtain a cluster-level risk value. The higher the cluster-level risk value, the more likely the group of operations as a whole may require synchronous peak-shifting, isolation enhancement, or approval escalation.

[0071] Furthermore, in one executable implementation, the job cluster risk value can be expressed as:

[0072] ;

[0073] in, This represents the risk value of job cluster C. This represents the risk value after propagation in the i-th work unit. Indicates the intra-cluster coupling enhancement parameter. Let represent the set of edges within a cluster. The first part of the formula represents the sum of the risks of nodes within the cluster, while the second part represents the coupling amplification effect brought about by the edge relationships within the cluster. Therefore, the cluster risk is no longer a simple summation, but rather reflects the characteristic that the overall risk is further amplified when high-risk nodes are adjacent to each other.

[0074] In one specific implementation, assuming that among the aforementioned five tasks, a confined space task, although initially having a slightly lower risk value than a hot work task in progress, may have its risk value increased to the highest level among all nodes after graph propagation due to its strong overlap with two hot work tasks and a temporary power supply task. Furthermore, a task cluster consisting of a confined space task and two hot work tasks will have a significantly higher risk value than a single high-altitude task cluster due to its higher edge weights and mutually reinforcing node risks. Therefore, the system can identify task combinations that truly require priority collaborative processing, rather than simply intuitively judging which task is more dangerous based on its type.

[0075] In another possible implementation, after propagation, nodes with risk values ​​exceeding a preset threshold can be marked as high-priority nodes, and clusters with job cluster risk values ​​exceeding a preset threshold can be marked as key collaborative clusters. These high-priority nodes and key collaborative clusters will participate in the optimization solution or receive higher control weights in step S40. This approach can further improve the algorithm's solution efficiency in large-scale job set scenarios without changing the overall technical approach.

[0076] Through step S30, the embodiment of the present invention combines the spatiotemporal conflict diagram of cross-operations with the initial risk vector of nodes, forming a processing leap from a static relationship diagram to a dynamic propagation risk diagram, so that the combined risks that could not be identified by single-ticket logic can be quantitatively calculated.

[0077] S40: Perform collaborative optimization based on the propagated risk vector and the job cluster risk value to obtain job execution association information corresponding to each job unit; wherein, the job execution association information includes execution level, isolation enhancement level and approval upgrade identifier.

[0078] Specifically, the execution level variables, isolation enhancement levels, and approval escalation identifiers corresponding to each work unit are defined, and the execution level number corresponding to each work unit is determined based on the execution level variables. A residual conflict attenuation coefficient is constructed based on the execution level number, isolation enhancement level, and approval escalation identifier of each work unit. A collaborative optimization objective function is constructed based on the post-propagation risk value, the normalized propagation weight, the residual conflict attenuation coefficient, and the cost of each control action. The collaborative optimization objective function is constrained and solved based on unique level constraints, hard conflict non-synchronous-level constraints, overall residual risk threshold constraints, isolation enhancement resource constraints, and approval escalation capacity constraints to obtain the work execution association information corresponding to each work unit, including the execution level, isolation enhancement level, and approval escalation identifier.

[0079] Furthermore, a risk-guided initial solution is generated based on the propagated risk value of each job unit and the job cluster risk value of each job cluster. A candidate solution evaluation function is constructed based on the risk-guided initial solution. An adaptive tabu length is constructed based on the maximum job cluster risk value and the average job cluster risk value among all current job clusters. In the case of locally excessive residual conflicts in the candidate solution, the local conflict cost of the job unit to be repaired at each execution level is calculated. The repair execution level of the job unit to be repaired is determined according to the minimum local conflict cost, and the optimal collaborative solution result is output based on the tabu search iteration result.

[0080] In this embodiment of the invention, step S40 is used to address the problem of how to form an optimal control scheme after identifying cross-combination risks. It should be noted that traditional high-risk operation management systems, even if they can detect some conflicts, usually only provide a prompt, leaving managers to decide based on experience whether to stagger peak times, increase isolation, or escalate approvals. This approach makes it difficult to simultaneously minimize both global risk and execution cost when multiple cross-operations exist concurrently. This embodiment of the invention further uses the propagated risk vector and operation cluster risk value obtained in step S30 directly as optimization inputs to construct a collaborative control solution process for cross-combination high-risk operation scenarios.

[0081] In this embodiment of the invention, three types of control variables are first defined for each work unit. The first type is the execution level, which indicates which execution batch or time level the work is scheduled to be implemented in. The second type is the isolation enhancement level, which indicates whether the work needs to be further isolated on top of the original isolation measures. The third type is the approval escalation indicator, which indicates whether the work needs to be upgraded in terms of approval level or add joint signature nodes. These three types of control variables can cover the three most common, direct, and executable control actions in the management of cross-risk operations.

[0082] The introduction of execution levels addresses the issue of staggered execution between multiple tasks. If there is a strong overlap between two tasks, they can be assigned to different execution levels to reduce the probability of them coexisting in time. The enhanced isolation level addresses the inadequacy of existing isolation measures by adding isolation items to reduce the transmission channels between tasks. The upgraded approval flag addresses the lack of additional review for high-risk tasks by adding approval nodes to ensure that high-risk tasks receive more rigorous review before implementation.

[0083] To characterize the mitigating effect of control actions on residual conflicts between tasks, this embodiment of the invention further constructs a residual conflict attenuation coefficient. Specifically, if two tasks are assigned to different execution levels, their time overlap will decrease in execution; if isolation measures are added to either task, the available risk propagation channels between them will shrink; if either task is upgraded to an approval level, it generally means that additional human verification and measure confirmation will further reduce its uncertainty. Therefore, this embodiment of the invention incorporates the level difference, isolation enhancement level, and approval upgrade status into the residual conflict expression.

[0084] Furthermore, in one executable implementation, the residual conflict attenuation coefficient can be expressed as:

[0085] ;

[0086] in, This represents the residual conflict attenuation coefficient between the i-th and j-th work units. and This indicates the execution level number corresponding to the two. and This indicates the corresponding level of isolation enhancement. and This indicates the corresponding approval upgrade indicator. , and This indicates the parameters that control actions affect. The principle behind this formula is that the greater the hierarchical difference, the more enhanced the isolation, and the stricter the approval process, the smaller the residual conflict that will ultimately remain between the two tasks.

[0087] Based on this, in this embodiment of the invention, the costs of adjusting execution levels, enhancing isolation, upgrading approval processes, and the residual coupling risks after control are all incorporated into the collaborative optimization objective. In other words, the system does not simply aim to postpone all tasks or upgrade all approval processes, but rather to achieve a balance between sufficiently low risk and acceptable on-site execution costs. Therefore, the optimization objective of this embodiment is a comprehensive objective, which includes the schedule disruption costs caused by task staggering, the management costs of increasing isolation and upgrading approval processes, and the combined risk costs remaining after control.

[0088] Furthermore, in one executable implementation, the collaborative optimization objective function can be expressed as:

[0089] ;

[0090] Where F represents the collaborative optimization objective value, This represents the execution adjustment cost corresponding to the i-th job unit being assigned to a certain execution level. This represents the isolation enhancement cost coefficient for the i-th job unit. This represents the approval upgrade cost coefficient for the i-th work unit. Indicates the normalized propagation weight. and This represents the risk value after propagation at the corresponding node. This represents the residual conflict attenuation coefficient after the control action has been performed. , , and This represents the objective weight. It should be noted that the last term in this objective function represents the combined risk remaining after control measures are implemented; it is the core risk term in the entire optimization process.

[0091] Regarding constraint settings, this embodiment of the invention also introduces several business-necessary restrictions. Specifically, each task must be assigned to one and only one execution level; tasks with hard conflicts cannot be located at the same execution level; the overall residual risk of all tasks after taking control actions must not exceed a preset threshold; the total amount of isolation enhancements must not exceed the current available resource capacity; and the number of approval escalations must not exceed the approval capacity that the administrator can bear. Through these constraints, it can be ensured that the output control scheme is not a theoretically optimal value, but a feasible solution that can be truly executed in real business processes.

[0092] Regarding the solution method, this embodiment of the invention employs an improved tabu search strategy to solve the aforementioned collaborative optimization problem. It should be noted that while traditional tabu search is suitable for solving discrete combinatorial optimization problems, its direct application to this scenario often results in problems such as high randomness in the search starting point, oscillations near high-risk job clusters, and a large number of locally conflicting solutions. Therefore, this embodiment of the invention has been adapted to the specific scenario.

[0093] Specifically, the algorithm first constructs a risk-guided initial solution based on the post-propagation risk vector and job cluster risk values ​​obtained in step S30. This means prioritizing jobs with high propagation risk and jobs within high-cluster risk job clusters to receive safer hierarchical arrangements or higher-level control actions during the initial allocation phase, thus preventing the algorithm from starting the search from an excessively low-quality initial state. Subsequently, in each round of the search, candidate solutions are evaluated, considering not only the optimization objective itself but also the penalties for constraint violations. This objective value plus penalty approach allows the algorithm to retain some exploration space in the early stages of the search while gradually approaching feasible optimal solutions in subsequent stages.

[0094] To enhance the search's adaptability to high-risk scenarios, this embodiment of the invention also uses the job cluster risk value to adjust the tabu length. Furthermore, when a particularly prominent high-risk job cluster exists in the current scenario, the algorithm automatically increases the tabu length, thereby reducing repeated bounces near that high-risk area; if the overall risk distribution is relatively uniform, the tabu length can be relatively shortened to improve search flexibility. Through this design, a direct linkage is formed between the optimization algorithm and the aforementioned graph propagation model, meaning that the risk cluster structure identified in the previous stage is not merely for display but directly participates in the solution process in the next stage.

[0095] Furthermore, in this embodiment of the invention, when a candidate solution has a relatively good overall objective value but still has significantly excessive residual conflicts in some areas, local backtracking repair can be triggered. Specifically, the system will sequentially test the local costs of the key operations that cause local conflicts at different execution levels, and select the level with the minimum combined local conflicts and execution disturbances for adjustment. Through this local repair mechanism, the entire candidate solution can be avoided from being completely discarded due to the unreasonable arrangement of a few key nodes, thereby improving the solution efficiency.

[0096] In one specific implementation, if step S30 identifies the highest risk value for a cluster of operations consisting of two hot work operations and one confined space operation, then in the risk-guided initial solution, one hot work operation can be prioritized for the first execution level, the confined space operation for the second execution level, and the other hot work operation marked as requiring an additional level of isolation and approval escalation. During subsequent searches, the system may continue to attempt to stagger high-altitude operations from confined space operations, or to escalate temporary power supply operations to a more stringent approval status. Ultimately, when the overall residual risk is below a threshold and the cost of delay is acceptable, the output optimal solution includes the execution level, isolation enhancement level, and approval escalation flag for each operation. This result will serve as the direct input for step S50.

[0097] In another possible implementation, jobs whose risk value after propagation is significantly lower than the threshold and whose risk value within their job cluster is also low can be given a lower search priority during the optimization process, or even have their original hierarchical arrangement maintained. This does not change the basic operating principle of the invention, but it can reduce the solution overhead and improve overall processing efficiency when the job scale is large.

[0098] Through step S40, the embodiment of the present invention achieves a key transformation from identifying combined risks to generating the optimal control scheme, so that the management and control of cross-high-risk operations no longer relies on human experience, but has clear data input, calculation process and output results.

[0099] S50: Generate the full-process control and safety prevention results for high-risk operations based on the execution level, the isolation enhancement level, and the approval upgrade identifier.

[0100] Specifically, staggered execution results are generated based on the execution level corresponding to each work unit, isolation enhancement results are generated based on the isolation enhancement level corresponding to each work unit, and approval upgrade results are generated based on the approval upgrade identifier corresponding to each work unit. The staggered execution results, isolation enhancement results, and approval upgrade results are correlated and summarized to form the full-process control and safety prevention results for high-risk operations. When the execution status information or actual isolation status information corresponding to any work unit changes, the affected work unit and its adjacent work units are extracted to form an affected subgraph, and local risk recalculation is performed based on the affected subgraph. If the local risk recalculation results show that the risk value of the corresponding work cluster exceeds a preset threshold, the collaborative optimization solution is re-executed and the full-process control and safety prevention results for high-risk operations are updated.

[0101] Once the optimal collaborative control solution is obtained, the execution level, isolation enhancement level, and approval escalation identifier can be mapped to specific digital management and control instructions. Corresponding suggestions for staggered execution, isolation enhancement, and approval escalation can then be displayed on the approval interface. Furthermore, by combining the post-propagation risk value and the risk value of the job cluster, combined risk warning information can be added to high-risk jobs and high-risk job clusters. This allows approvers to directly view the conflict analysis results and perform actions such as approval, approval after adjustment, rejection, or triggering recalculation during the approval process.

[0102] In this embodiment of the invention, step S50 is used to transform the optimal control solution obtained in step S40 into a system-executable business result, and to perform local updates based on state changes during job execution. It should be noted that the optimization solution itself still provides variable results in a computational sense; only by mapping these results to actual control actions such as staggered execution, enhanced isolation, and escalated approval processes can a truly comprehensive, end-to-end security and control collaborative process be formed.

[0103] In this embodiment of the invention, the staggered execution results are first generated based on the execution level corresponding to each work unit. Specifically, jobs with the same execution level can be grouped into the same execution batch, while jobs with different execution levels can be sequentially executed according to their hierarchical order. In this way, high-risk job combinations that would otherwise overlap in time can be separated in the execution plan through hierarchical division, thereby reducing the cross-risk caused by parallel implementation.

[0104] Secondly, isolation enhancement results are generated based on the isolation enhancement level corresponding to each work unit. Specifically, if the isolation enhancement level of a certain work unit is greater than zero, additional isolation control requirements are added to it in addition to the original required isolation items. For example, an additional valve status verification, an additional energy isolation confirmation, an additional ventilation confirmation, or an additional adjacent equipment sealing verification may be required. It should be noted that the embodiments of the present invention do not limit the specific content of isolation enhancement; any form that can reduce the probability of the existence of risk propagation channels in actual business operations can be used as the implementation form of isolation enhancement results.

[0105] Secondly, the system generates approval upgrade results based on the approval upgrade flags corresponding to each work unit. If a work is marked as requiring approval upgrade, the system can add a layer of countersigning, an additional responsible person's review, or a special risk review process to the original approval chain. In this way, work with higher risks after dissemination or that is in a key work cluster can be subject to stricter management and control before implementation.

[0106] Subsequently, the results of staggered execution, enhanced isolation, and upgraded approval processes are correlated and summarized to form the overall management and safety control results for high-risk operations. It should be noted that these results are not merely a simple list, but a comprehensive management output that corresponds one-to-one with the risk results of each work unit, each work cluster, and the results of the preceding propagation analysis. This output can be used for display on the approval interface, for issuing on-site execution instructions, and as a baseline input for subsequent status change recalculation.

[0107] Furthermore, this embodiment of the invention also considers the dynamic changes that occur during the execution of high-risk operations. For example, an operation may start earlier, end later, be temporarily suspended, or an isolation measure may be completed during the operation. In such cases, if the original optimization results are still used, the latest on-site status may not be reflected. Therefore, this embodiment of the invention introduces a local update mechanism in step S50.

[0108] Specifically, when the execution status information or actual isolation status information corresponding to any job unit changes, the system first identifies the job unit and its adjacent job units, and constructs an affected subgraph accordingly. The reason for not directly recalculating all jobs is that changes in the field status often only affect local relationships, not the entire job network. By extracting the affected subgraph, the efficiency of dynamic updates can be improved while preserving the overall stability of the global results.

[0109] After the affected subgraph is constructed, the system will recalculate the initial local risk based on the subgraph and perform local propagation calculations to determine whether the risk of the affected nodes and their respective job clusters has changed substantially. If the risk value of the corresponding job cluster is still lower than the preset threshold after local propagation, the original control solution can continue to be used, and only the state display is refreshed; if the risk value of a job cluster rises and exceeds the threshold after local propagation, step S40 is triggered again, and collaborative optimization is performed only on the jobs within the affected cluster, replacing the old results with the new results. Through this mechanism, a dynamic update process of state change, local identification, local propagation, local re-optimization, and result update can be formed.

[0110] In one feasible implementation, local risk updates can still maintain the principle of consistency with global propagation, that is, retaining the initial local risk of a node and superimposing the propagation effects from local neighboring nodes. In other words, although the implementation of this invention reduces the computational scope in the dynamic stage, it does not change its graph propagation principle and collaborative solution logic; it only improves real-time performance at the engineering implementation level through local recalculation.

[0111] In one specific implementation, if a hot work operation originally scheduled for the second execution level is applied for commencement ahead of schedule due to completion of preparations, the system detects a change in the execution status of the operation and extracts the operation and its adjacent confined space operations, high-altitude operations, and temporary power supply operations to form an affected subgraph. Subsequently, the system recalculates the initial risk and post-propagation risk of the nodes in this subgraph and finds that the risk of the intersection cluster of the hot work operation and the confined space operation has increased significantly, exceeding the current residual risk threshold. At this point, the system calls back step S40 to resolve the operations within the cluster, ultimately adjusting the confined space operation from the second level to the third level and adding a level of approval upgrade for the hot work operation. The updated control results are then written back to the system, thereby preventing the original control scheme from failing due to the change in status.

[0112] In another possible implementation, if an existing isolation gap in a confined space operation has been filled in on-site, the system will still trigger the extraction of the affected subgraph and local recalculation. Since the isolation deviation coefficient of this operation decreases, its propagation risk and cluster risk may also decrease. If the decreased risk indicates that the previously set approval upgrade is no longer necessary, the system can cancel the approval upgrade flag after re-optimization, thereby avoiding over-control. Therefore, the dynamic update mechanism in this embodiment can strengthen control when risk increases and appropriately relax control when risk decreases, making the entire high-risk operation management process more refined.

[0113] Through step S50, the embodiment of the present invention ultimately forms a complete process from data acquisition, relationship modeling, risk propagation, collaborative optimization to dynamic updating, enabling the full-process management of cross-high-risk operations to truly possess continuity and updability, rather than being static and unchanging after a one-time approval.

[0114] like Figure 2 As shown, this invention provides a high-risk operation full-process management and safety control system. The system is used to execute the above-described high-risk operation full-process management and safety control method. The system includes:

[0115] A construction unit is used to acquire work order data, device connection relationship data, and actual isolation status data from the high-risk operation management platform to construct a high-risk operation process control information group; wherein, the construction of the high-risk operation process control information group includes a set of operation units, a device connection matrix, and an isolation deviation coefficient;

[0116] The calculation unit is used to construct a spatiotemporal conflict diagram of cross-operations based on the high-risk operation process control information group, and to calculate the initial risk vector of the node corresponding to each operation unit.

[0117] The propagation unit is used to perform graph risk propagation processing based on the cross-operation spatiotemporal conflict graph and the node initial risk vector to obtain the propagated risk vector and the risk value of the operation cluster corresponding to each operation unit.

[0118] The solution unit is used to perform collaborative optimization based on the propagated risk vector and the job cluster risk value to obtain job execution association information corresponding to each job unit; wherein, the job execution association information includes execution level, isolation enhancement level and approval escalation identifier;

[0119] The generation unit is used to generate the full-process control and safety prevention results of high-risk operations based on the execution level, the isolation enhancement level, and the approval upgrade identifier.

[0120] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a microcontroller, chip, or processor to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0121] The optional embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details described above. Within the scope of the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical solutions of the embodiments of the present invention, and these simple modifications all fall within the protection scope of the embodiments of the present invention. It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not further describe the various possible combinations.

[0122] Furthermore, various different embodiments of the present invention can be combined in any way, as long as they do not violate the spirit of the embodiments of the present invention, they should also be regarded as the content disclosed by the embodiments of the present invention.

Claims

1. A method for full-process management and safety control of high-risk operations, characterized in that, The method includes the following steps: Acquire work order data, device connection relationship data, and actual isolation status data from the high-risk operation management platform to construct a high-risk operation process control information group; wherein, the construction of the high-risk operation process control information group includes a set of operation units, a device connection matrix, and an isolation deviation coefficient; Based on the high-risk operation process control information group, a cross-operation spatiotemporal conflict diagram is constructed, and the initial risk vector of each node corresponding to each operation unit is calculated. Specifically, the construction of a cross-operation spatiotemporal conflict diagram based on the high-risk operation process control information group includes: for any two operation units in the operation unit set, calculating a time overlap coefficient based on the planned start and end times of each operation unit; calculating an equipment coupling coefficient based on the equipment connection matrix and the associated equipment sets corresponding to any two operation units; determining a type coupling factor based on the operation types corresponding to any two operation units, and fusing the time overlap coefficient, the equipment coupling coefficient, the type coupling factor, and the corresponding isolation deviation coefficient to obtain edge weights; and constructing a cross-operation spatiotemporal conflict diagram based on the edge weights. A scenario enhancement propagation coefficient is constructed based on the time overlap coefficient, type coupling factor, and isolation deviation coefficient between any two work units; a normalized propagation weight is constructed based on the edge weights in the cross-work spatiotemporal conflict graph and the scenario enhancement propagation coefficient, and the normalized propagation weight is written into the cross-work spatiotemporal conflict graph. Based on the spatiotemporal conflict graph of the cross-operation and the initial risk vector of the node, the risk propagation process of the graph is performed to obtain the post-propagation risk vector and the risk value of the operation cluster corresponding to each operation unit. Specifically, this includes: taking the initial risk vector of the node as the initial input, performing iterative risk propagation based on the normalized propagation weight, stopping the iteration when the results of two adjacent iterations meet the convergence condition, and recording the converged node risk value as the post-propagation risk value; summarizing the post-propagation risk values ​​corresponding to each operation unit to form the post-propagation risk vector, and calculating the operation cluster risk value based on the post-propagation risk value and the normalized propagation weight within the cluster. Based on the propagated risk vector and the risk value of the job cluster, a collaborative optimization solution is performed to obtain the job execution association information corresponding to each job unit; wherein, the job execution association information includes execution level, isolation enhancement level, and approval escalation identifier; Specifically, this includes: defining execution level variables, isolation enhancement levels, and approval escalation identifiers corresponding to each work unit, and determining the execution level number corresponding to each work unit based on the execution level variables; constructing a residual conflict attenuation coefficient based on the execution level number, isolation enhancement level, and approval escalation identifier corresponding to each work unit; constructing a collaborative optimization objective function based on the post-propagation risk value, the normalized propagation weight, the residual conflict attenuation coefficient, and the cost of each control action; and constraining the collaborative optimization objective function based on unique level constraints, hard conflict non-same-level constraints, overall residual risk threshold constraints, isolation enhancement resource constraints, and approval escalation capacity constraints to obtain work execution association information corresponding to each work unit, including execution level, isolation enhancement level, and approval escalation identifier. The constraint solution for the collaborative optimization objective function includes: generating a risk-guided initial solution based on the propagated risk value of each job unit and the job cluster risk value of each job cluster; constructing a candidate solution evaluation function based on the risk-guided initial solution; constructing an adaptive tabu length based on the maximum job cluster risk value and the average job cluster risk value among all current job clusters; calculating the local conflict cost of the job unit to be repaired at each execution level when there are excessively high residual conflicts in the candidate solution; determining the repair execution level of the job unit to be repaired based on the minimum local conflict cost; and outputting the optimal collaborative solution result based on the tabu search iteration result. The results of full-process management and safety control for high-risk operations are generated based on the execution level, the isolation enhancement level, and the approval upgrade identifier.

2. The method for full-process control and safety prevention of high-risk operations according to claim 1, characterized in that, Establish a high-risk operation process control information group, including: The high-risk operation records in the work ticket data are parsed, and based on the operation characteristics in the parsing results, the high-risk operation records are converted into operation units to form a set of operation units; The operation characteristics include operation type, operation area, set of associated equipment, planned start time, planned end time, basic risk level, set of required isolation items, and execution status information; Based on the device connection relationship data, a device connection matrix is ​​constructed; based on the actual isolation status data, the actual isolation item set corresponding to each work unit in the work unit set is determined; and the isolation deviation coefficient corresponding to each work unit is calculated. Based on the set of work units, the equipment connection matrix, and the isolation deviation coefficient, a high-risk work process control information group is constructed.

3. The method for full-process control and safety prevention of high-risk operations according to claim 1, characterized in that, Calculate the initial risk vector for each node corresponding to each work unit, including: Based on the planned start time and planned end time of each work unit in the work unit set, calculate the planned duration of each work unit and determine the duration factor. Calculate the initial risk value for each work unit based on its basic risk level, isolation deviation coefficient, duration factor, and execution status factor. The initial risk values ​​corresponding to each work unit are summarized to form the node initial risk vector.

4. The method for full-process control and safety prevention of high-risk operations according to claim 1, characterized in that, Based on the execution level, the isolation enhancement level, and the approval escalation identifier, a high-risk operation full-process control and safety prevention result is generated, including: Based on the execution level corresponding to each work unit, generate the staggered execution results of the work; based on the isolation enhancement level corresponding to each work unit, generate the isolation enhancement results; based on the approval upgrade identifier corresponding to each work unit, generate the approval upgrade results. The results of staggered operation, enhanced isolation, and upgraded approval are correlated and summarized to form the full-process management and safety control results for high-risk operations. If the execution status information or actual isolation status information corresponding to any work unit changes, the affected work unit and its adjacent work units are extracted to form an affected subgraph, and a local risk recalculation is performed based on the affected subgraph. If the local risk recalculation results show that the risk value of the corresponding job cluster exceeds the preset threshold, the collaborative optimization solution is re-executed and the results of the full-process control and safety prevention of the high-risk operation are updated.

5. A high-risk operation full-process control and safety prevention system, characterized in that, The system is used to execute the high-risk operation full-process control and safety prevention method according to any one of claims 1-4, and the system includes: A construction unit is used to acquire work order data, device connection relationship data, and actual isolation status data from the high-risk operation management platform to construct a high-risk operation process control information group; wherein, the construction of the high-risk operation process control information group includes a set of operation units, a device connection matrix, and an isolation deviation coefficient; The calculation unit is used to construct a spatiotemporal conflict diagram of cross-operations based on the high-risk operation process control information group, and to calculate the initial risk vector of the node corresponding to each operation unit. The propagation unit is used to perform graph risk propagation processing based on the cross-operation spatiotemporal conflict graph and the node initial risk vector to obtain the propagated risk vector and the risk value of the operation cluster corresponding to each operation unit. The solution unit is used to perform collaborative optimization based on the propagated risk vector and the job cluster risk value to obtain job execution association information corresponding to each job unit; wherein, the job execution association information includes execution level, isolation enhancement level and approval escalation identifier; The generation unit is used to generate the full-process control and safety prevention results of high-risk operations based on the execution level, the isolation enhancement level, and the approval upgrade identifier.

Images