Digital Delivery Topology Mapping Method and System Based on Multi-Source Real-Time Data Fusion

Abstract

This invention belongs to the field of digital delivery, and particularly relates to a digital delivery topology mapping method and system based on multi-source real-time data fusion. The method acquires factory building distribution, equipment distribution and operation control logic, and preset functional block operation logic, and constructs a hierarchical clustering functional mapping space using association parsing and clustering algorithms. Responding to target functional requirements, a hierarchical response mapping path is obtained using a deep search algorithm. Based on this path, a 3D simulation model, and the performance of the display system equipment and network status, hierarchical synchronous response and distributed node anomaly monitoring are achieved. Based on the monitoring results, anomalies are traced using a hidden Markov algorithm and a forward inference model. After conflict resolution, iterative verification is performed until the function is free of anomalies and the mapping space is updated. Adjusting requirements and repeating the steps yields a completely updated mapping space, achieving precise collaboration between function and visuals under multi-source data fusion.

Description

Technical Field

[0001] This invention belongs to the field of digital delivery, and in particular relates to a digital delivery topology mapping method and system for multi-source real-time data fusion. Background Technology

[0002] With the deepening of industrial digital transformation, factory digital visualization delivery has become a core means to improve production efficiency and optimize operation management. It integrates complex functions such as system settings, dashboards, factory walkthroughs, intelligent operations, intelligent decision-making, and safety and environmental management, while also needing to consider interface layout, element display, and cross-browser and operating system compatibility. However, due to the complexity of the system structure and functions, coupled with the large amount of dynamic visual image data during real-time display, there is a mapping confusion between functional control and the screen, conflicts between control mappings and display mappings, and significant mapping lag caused by complex mapping relationships. This results in delays and display errors in the overall visualization, seriously affecting the reliability and operational efficiency of the digital system. In factory digital visualization scenarios driven by multi-source real-time data, how to establish dynamic association rules between functional control and screen display to solve the mapping confusion and conflict problems is an urgent issue to be addressed. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention proposes a digital delivery topology mapping method and system based on multi-source real-time data fusion. This method acquires the target factory building distribution, equipment distribution, and operational control logic, as well as the pre-defined functional block operational logic. It then constructs a hierarchical clustering functional mapping space using association parsing and clustering algorithms. Responding to target functional requirements, a hierarchical response mapping path is obtained using a deep search algorithm. Based on this path, a 3D simulation model of the target area, and the performance of the display system equipment and real-time network status, a hierarchical synchronous response is achieved. The functional implementation control command response and dynamic image rendering response are correlated and mapped through a correlation response delay connection relationship constructed using control command response priority, display system equipment performance, real-time network status, and image clarity, smoothness, and completeness. Anomalies are monitored using a distributed monitoring node network. Hidden Markov algorithms and forward inference models are used to trace anomalies backward. After processing with a conflict resolution strategy library, the anomalies are fed back to the 3D simulation model for iterative verification until no functional anomalies are found, at which point the path is updated to the mapping space. The response requirements are adjusted, and the steps are repeated to finally obtain a completely updated hierarchical clustering functional mapping space. This achieves precise coordination between functions and images under multi-source data fusion, improving the reliability of digital delivery.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A digital delivery topology mapping method based on multi-source real-time data fusion includes:

[0006] By responding to the target functional requirements and combining the pre-defined hierarchical clustering functional mapping space of the target factory with a deep search algorithm, the hierarchical response mapping path of the target function is obtained; the hierarchical clustering functional mapping space of the target factory is constructed by combining multi-source data of the factory with association analysis algorithms and clustering algorithms;

[0007] Based on the hierarchical response mapping path of the target function, the hierarchical synchronous response of the target function is carried out by combining the preset target area 3D simulation model and display system equipment performance and real-time network status. In addition, the hierarchical synchronous response of the target function is carried out in real time by combining the preset distributed monitoring node network to monitor the functional response mapping anomalies and corresponding dynamic screen rendering anomalies in the target area response, and obtain the real-time monitoring response information of the target function.

[0008] Based on the real-time monitoring response information of the target function, combined with the preset forward inference model, reverse anomaly tracing and localization are performed along the hierarchical response mapping path. At the same time, based on the localized anomaly information, conflict resolution is performed in combination with the conflict resolution strategy library. The hierarchical response mapping path of the target function after conflict resolution is fed back to the three-dimensional simulation model of the target area to continue simulation and anomaly monitoring until the target function is fully realized and there are no anomalies.

[0009] The hierarchical response mapping path of the target function corresponding to the target function without abnormality is fed back to the hierarchical clustering function mapping space of the target factory for real-time update, so as to obtain the hierarchical clustering function mapping space of the target factory after the single function block is updated.

[0010] Specifically, the process of constructing the hierarchical clustering functional mapping space of the target factory includes:

[0011] Acquire multi-source real-time information from the target factory, including building distribution status data in the target area, equipment distribution status and operation control logic information, and operation logic data of preset functional blocks of the digital delivery platform.

[0012] Based on the analysis of the building distribution data of the target area, the coordinates and boundary constraints of the factory boundary feature points are determined. The geometric center of the factory is calculated by least squares fitting and used as the origin of the coordinate system. The X-axis is the direction parallel to the main axis of the factory, the Y-axis is the horizontal direction perpendicular to the main axis, and the Z-axis is the vertical direction.

[0013] The data on the distribution of buildings in the target area are preprocessed, and the architectural drawings in CAD format are converted into a three-dimensional mesh model. At the same time, the geometric relationship information of the buildings is extracted and labeled according to the inclusion relationship of the target buildings. The geometric relationship information includes, but is not limited to, the vertex coordinates, thickness, height and material property parameters of walls, columns, floors, doors and windows.

[0014] Based on the geometric relationship information of the building, the three-dimensional mesh model is mapped to the factory coordinate system of the target area through coordinate transformation matrix mapping. At the same time, Boolean operation processing is performed on the mapped building model to eliminate overlapping surfaces and redundant vertices, so as to obtain the three-dimensional mapping model of the target building.

[0015] Specifically, the process of constructing the hierarchical clustering functional mapping space of the target factory also includes:

[0016] The installation position coordinates of each component in the configuration equipment distribution state are converted into coordinate values ​​in a three-dimensional coordinate system. The three-dimensional model of the corresponding equipment is retrieved according to the preset equipment three-dimensional model library. The equipment-building spatial constraint relationship and belonging relationship are established according to the building and equipment geometric information. After scaling, it is mapped to the corresponding coordinate position to obtain the initial building-equipment three-dimensional mapping model.

[0017] Based on the equipment's operation control logic information, the control causal relationship and correlation degree between various devices are obtained through correlation analysis algorithms;

[0018] Based on the causal relationship and correlation degree between various devices, the corresponding device components are taken as component nodes, and the directed relationship between the corresponding components is the first directed connection. At the same time, different colors are used to map the control relationship type of the first directed connection between different components. The length of the connection relationship and the depth of the color are used to map the correlation strength and the probability of correlation failure of the control relationship corresponding to the first directed connection between different components.

[0019] Specifically, the process of constructing the hierarchical clustering functional mapping space of the target factory also includes:

[0020] Simultaneously, based on the number of connections between corresponding devices between different buildings and the correlation strength and probability of correlation failure of the first directed connection corresponding to the control relationship, a second directed connection relationship between different building nodes is constructed. The coarseness of the second directed connection relationship is used to map the second correlation strength of different buildings. The building node is constructed from each individual building and its corresponding building location and geometric attribute parameters.

[0021] The constructed device component nodes, the first directed connection and the parameters of the first directed connection, the building nodes and the corresponding second directed connection relationship and the second association strength are combined with the graph algorithm and mapped to the initial building-equipment 3D mapping model to obtain the building-equipment hierarchical global mapping space;

[0022] A geometric deviation threshold is set, and the geometric deviation of the building and equipment is verified on the initial building-equipment 3D mapping model. When the initial building-equipment 3D mapping model is verified, the building-equipment hierarchical global mapping space is combined with the simulation algorithm and the dynamic control parameters of the corresponding equipment. Real-time data pushed by the edge gateway is received through the WebSocket protocol to verify the control logic and fault correlation causal relationship, and the verified building-equipment hierarchical global mapping space is obtained.

[0023] Specifically, the process of constructing the hierarchical clustering functional mapping space of the target factory also includes:

[0024] The pre-defined operational logic data of each functional block of the digital delivery platform is combined with a word segmentation and embedding model for textual preprocessing to obtain the pre-processed logical text information of each functional block.

[0025] Based on the preprocessed functional block logic text information and the first-level decomposition rules, the first-level main control sub-function, the first-level collaborative control sub-function, and the causal relationship between the first-level main control sub-function and the first-level collaborative control sub-function in time and space at continuous time points are obtained.

[0026] The first-level decomposition rules include control constraint rules and screen rendering constraint rules for the operation or screen changes of the corresponding device components of the target functional block at continuous time points;

[0027] The primary main control sub-function is a functional sub-module that runs through the entire implementation process of the functional block, and its implementation time is equal to the implementation time of the corresponding functional block. The primary collaborative control sub-function is a functional sub-module that is implemented within a preset time period of the functional block, and the preset time period is greater than 0 and less than the overall implementation time of the corresponding functional block.

[0028] Specifically, the process of constructing the hierarchical clustering functional mapping space of the target factory also includes:

[0029] Based on the first-level main control sub-function and the first-level cooperative control sub-function at continuous time points, combined with the analytical algorithm and the hybrid control instructions corresponding to each control sub-function, a second-level decomposition is performed to obtain the second-level main control instruction sequence corresponding to each first-level main control sub-function and the second-level cooperative control instruction sequence corresponding to each first-level cooperative control sub-function at continuous time points.

[0030] Simultaneously, based on the causal relationship between the primary main control sub-function and the primary cooperative control sub-function in time and space, the spatial constraint relationship and temporal priority relationship between the secondary main control instruction and the related secondary cooperative control instruction sequence at continuous time points are obtained.

[0031] Based on the sequence of secondary master control instructions corresponding to each functional block at consecutive time points and the sequence of secondary cooperative control instructions that are related to each secondary master control instruction, combined with the corresponding spatial constraint relationship, time priority relationship and random forest algorithm, the functional hierarchical control forest corresponding to each functional block at consecutive time points is obtained.

[0032] Specifically, the process of constructing the hierarchical clustering functional mapping space of the target factory also includes:

[0033] Based on the functional hierarchical control forest corresponding to different functional blocks, the main control instructions or cooperative control instructions with causal relationships at continuous time points are analyzed by combining the correlation analysis algorithm and the adaptive clustering algorithm. The correlation control relationship of the functional hierarchical control forest corresponding to different functional blocks at continuous time points is obtained, and the hierarchical causal correlation control forest corresponding to different functional blocks at continuous time points is obtained.

[0034] Based on the hierarchical causal control forest corresponding to different functional blocks at continuous time points, combined with the verified building-equipment hierarchical global mapping space, the secondary main control instructions or secondary collaborative control instructions are matched and aligned with the building-equipment hierarchical global mapping space through the matching algorithm. A bidirectional mapping between each secondary main control instruction or secondary collaborative control instruction and the control logic or rendering between buildings, equipment and components is established, and the hierarchical clustering functional mapping space of the target factory is obtained.

[0035] Specifically, the process of constructing the hierarchical clustering functional mapping space of the target factory also includes:

[0036] A three-dimensional simulation model of the target region is constructed based on the hierarchical clustering functional mapping space of the target factory and the simulation algorithm.

[0037] Based on the target area 3D simulation model, combined with the real-time collected target function response requirement data and screen response adjustment space, the response and delay of control logic and screen rendering at continuous time points are verified, as well as the matching and alignment anomaly verification, to obtain the verified target area 3D simulation model.

[0038] Based on the verified three-dimensional simulation model of the target area, combined with the hierarchical response mapping path of the target function, and combined with the forward inference model constructed by the hidden Markov algorithm, the sub-functional conflict detection and verification within a single target function and the multi-target functional conflict detection and verification are carried out along the connection relationship of each functional block in the hierarchical clustering functional mapping space of the target factory. The functional conflict location information and forward conflict risk transfer probability matrix corresponding to a single or cross-target function at continuous time points are obtained.

[0039] Based on the functional conflict location information and forward conflict risk transfer probability matrix corresponding to single or cross-target functions at continuous time points, starting from the component nodes or building nodes where conflicts exist, and combining the forward conflict risk transfer probability matrix with the preset conflict resolution strategy library, conflict resolution and connection mapping updates are performed along the nodes and connection relationships in the hierarchical clustering functional mapping space of the target factory, and the three-dimensional simulation model of the target area after real-time update of conflict resolution and connection mapping is obtained.

[0040] Specifically, the screen response adjustment space is used to adjust the screen resolution of different areas of the real-time display screen based on the real-time acquisition of the display system equipment performance, network status, importance of the real-time display screen area, and the requirements for screen clarity, completeness, and smoothness of display at continuous time points. This ensures that the real-time display screen meets the corresponding threshold requirements for screen clarity, completeness, and smoothness. The importance of the real-time display screen area is determined by the correlation between the buildings, equipment, and equipment operating status, operating results, and background images in different areas of the corresponding screen and the preset target functional requirements. The real-time display screen area includes a central area and an edge area. The edge area includes edge areas in the direction of screen movement and edge areas in the non-direction of screen movement. The edge areas in the direction of screen movement are obtained by splicing the average grayscale values ​​of pixels extracted from adjacent edges of images corresponding to different frames at continuous time points with the intrinsic entropy across time points.

[0041] A digital delivery topology mapping system that integrates multi-source real-time data fusion includes: a response module, a simulation module, an anomaly reasoning and resolution module, and a feedback update module;

[0042] The response module is used to respond to the target functional requirements by combining the preset target factory hierarchical clustering functional mapping space with a deep search algorithm to obtain the hierarchical response mapping path of the target function.

[0043] The simulation module, based on the hierarchical response mapping path of the target function, combines a preset 3D simulation model of the target area with the performance of the display system equipment and the real-time network status to perform hierarchical synchronous response of the target function. It also combines a preset distributed monitoring node network to monitor in real time the functional response mapping anomalies and corresponding dynamic image rendering anomalies in the target area response, and obtain real-time monitoring response information of the target function. The hierarchical response includes the function implementation control command response and the dynamic image rendering response. The function implementation control command response and the dynamic image rendering response are correlated and mapped through the correlation response delay connection relationship constructed by the control command response priority, the performance of the display system equipment, the real-time network status, and the image clarity, smoothness, and completeness.

[0044] The anomaly reasoning and resolution module, based on the real-time monitoring response information of the target function, combines the Hidden Markov algorithm and the constructed forward reasoning model to perform reverse anomaly tracing and localization along the hierarchical response mapping path. At the same time, it resolves conflicts based on the located anomaly information and the conflict resolution strategy library, and feeds back the hierarchical response mapping path of the target function after conflict resolution to the three-dimensional simulation model of the target area to continue simulation and anomaly monitoring until the target function is fully realized and there are no anomalies.

[0045] The feedback update module is used to feed back the hierarchical response mapping path of the target function corresponding to the target function without anomalies to the hierarchical clustering function mapping space of the target factory for real-time update, so as to obtain the hierarchical clustering function mapping space of the target factory after the single function block is updated.

[0046] Compared with the prior art, the beneficial effects of the present invention are:

[0047] This invention addresses the shortcomings of existing technologies by constructing a dynamic hierarchical functional mapping space through multi-source data fusion. It precisely integrates building geometric constraints, causal relationships in equipment control, and hierarchical instruction sequences for functional logic, providing a high-precision topological mapping relationship for functional responses. When responding to demands, it generates multi-dimensional paths, combining 3D simulation models with real-time equipment / network status to achieve coordinated execution of functional control instructions and visual rendering. A dynamic latency model adaptively allocates resources, prioritizing the smoothness of critical instructions and core visuals. Distributed monitoring captures anomalies, and a Hidden Markov Model is used to trace back along the path to locate the root cause node. A conflict resolution strategy is used for closed-loop correction and iterative verification. Simultaneously, anomaly-free paths are fed back to update the mapping space, driving the continuous evolution of equipment topology and control logic. A forward conflict risk matrix predicts cross-functional competition, proactively resolving spatiotemporal conflicts. Dynamic optimization of visual rendering resource allocation ensures the clarity of core information based on regional importance hierarchy, and edge region cross-frame stitching technology maintains visual continuity. This significantly improves the reliability of functional execution, system self-healing ability, and long-term adaptability of digital delivery in complex factory environments. Attached Figure Description

[0048] Figure 1 This is a flowchart of the digital delivery topology mapping method for multi-source real-time data fusion according to the present invention;

[0049] Figure 2 This is a block diagram of the digital delivery topology mapping system for multi-source real-time data fusion according to the present invention. Detailed Implementation

[0050] Example 1

[0051] Please see Figure 1 The present invention provides an embodiment of a digital delivery topology mapping method for multi-source real-time data fusion, comprising the following steps:

[0052] S1. By combining the target functional requirements with the preset target factory hierarchical clustering functional mapping space and deep search algorithm, the hierarchical response mapping path of the target function is obtained.

[0053] It should be further explained that the target factory hierarchical clustering function mapping space in this embodiment is constructed by combining multi-source data of the factory with association analysis algorithms and clustering algorithms; for example, the response target function requirements in this application include, but are not limited to, setting functions, cockpit, factory roaming, intelligent operation, intelligent decision-making, safety and environmental management, interface display rendering, etc.; wherein the setting functions include switching between internal and external scenes, multilingual switching, and voice broadcast; factory roaming includes, but is not limited to, roaming path, equipment list, and monitoring points; intelligent operation includes, but is not limited to, one-click start / stop, intelligent incineration, energy wave ash cleaning, sound wave temperature measurement, and motor brain; intelligent decision-making includes, but is not limited to, accident self-healing, intelligent inspection, predictive maintenance, alarm management, MIS / SIS system; safety and environmental management includes, but is not limited to, infrared monitoring, digital security, carbon emission monitoring, and intelligent tagging, etc.; the response target function requirements in this application also include verifying the accuracy of the answers to questions by various functions and the correctness of command interaction; interface display rendering includes, but is not limited to, main interface layout, element display, color matching, and font style, etc.; compatibility aspects include operation under different browsers and operating systems.

[0054] It should be further explained that the process of obtaining the hierarchical response mapping path of the target function in this embodiment includes:

[0055] Based on the initial association information of functional blocks in the hierarchical clustering functional mapping space of the target functional requirements and the target factory, the starting node of the deep search is determined and the initial search node is obtained. The initial association information includes the first-level main control sub-function corresponding to the target function and the associated functional block identifier. It should be further noted that the initial search node is bound to the unique code of the first-level main control sub-function and the corresponding function start timestamp to ensure that it corresponds one-to-one with the functional block identifier in the mapping space.

[0056] Based on the time priority relationship, spatial constraint relationship and association rules between the primary main control sub-function and the primary collaborative control sub-function in the hierarchical clustering function mapping space of the target factory, the path constraint conditions for deep search are set to obtain the search constraint set; the constraint conditions include the execution sequence of sub-functions, the spatial association range of equipment and the mapping rules between control commands and rendering commands.

[0057] Based on the node connection relationship between the initial search node and the target factory hierarchical clustering functional mapping space, the child nodes are expanded layer by layer according to the depth-first strategy to obtain the candidate child node set of the current level; the node connection relationship reflects the control causal relationship between equipment components and buildings, as well as the calling and control relationship of functional sub-modules; the node connection relationship includes the first directed connection and the second directed connection.

[0058] The candidate child node set is validated based on the search constraint set, and child nodes that do not meet the time priority, spatial constraints and sub-function association rules are removed to obtain the valid child node set. The validation content includes whether the secondary control instructions corresponding to the child nodes meet the spatiotemporal constraints and functional logic dependencies of the master-cooperative instructions.

[0059] Based on the set of valid child nodes, the node expansion and verification steps are executed recursively until a termination node that meets the conditions for the complete implementation of the target function is found, thus obtaining a complete path segment from the start node to the termination node; the termination node corresponds to the final state of the target function implementation, including the completion markers of all necessary control instructions and rendering instructions.

[0060] Based on the hierarchical structure of the target factory hierarchical clustering functional mapping space, the complete path segments are hierarchically divided and associated to obtain the hierarchical response mapping path of the target function. The path contains the connection relationship of each level node, the sequence of control instructions and the corresponding rendering instruction association information, and is consistent with the hierarchical logic of the mapping space. The hierarchical structure includes building layer, equipment layer, functional block layer and functional sub-module layer.

[0061] For example, this embodiment uses the intelligent incineration functional block in the digital delivery platform as an example to illustrate the process of obtaining the corresponding hierarchical response mapping path, including:

[0062] Based on the text description of the "intelligent incineration" functional requirements, extract functional keywords such as "incinerator control," "combustion parameter adjustment," and "furnace temperature monitoring," as well as "furnace temperature 850±50℃" and "air volume 1200±50m³ / h." 3 The target parameters, such as " / h", are matched with the functional block identifiers and parameter ranges stored in the hierarchical clustering function mapping space of the target factory. The functional block with the highest relevance is selected. This functional block corresponds to the first-level main control sub-function "Incineration Full Process Control", which covers the entire life cycle of the smart incineration function. This is used as the starting node for deep search to obtain the initial search node. The functional block identifier, for example, is "FB - Smart Incineration - 001".

[0063] Based on the time priority relationships, spatial constraints, and association rules between the primary main control sub-function "Full-Process Control of Incineration" and the primary collaborative control sub-functions "Dynamic Adjustment of Furnace Temperature" and "Real-Time Adaptation of Air Volume" recorded in the hierarchical clustering function mapping space of the target factory, and based on the principle that air volume adjustment must be based on furnace temperature monitoring values, the path constraints for deep search are set as follows: in terms of time, the timestamp offset between the main control command and the collaborative control command does not exceed the preset delay threshold length; in terms of space, the spatial distance between the associated devices does not exceed the boundary of the incineration workshop; and logically, the collaborative sub-function must be triggered by the operating status of the main control sub-function to obtain the search constraint set. It should be further noted that the relevant time priority relationships in this embodiment include, but are not limited to, "the incinerator start command must be executed before the feeding device start command"; and the spatial constraints include, but are not limited to, "the incinerator, primary air fan, and feeding device must be located in the same incineration workshop area".

[0064] Based on the initial search node "Incineration Process Control", the node connection relationships stored in the hierarchical clustering function mapping space of the target factory are invoked, and the sub-nodes are expanded layer by layer according to the depth-first strategy: the equipment layer expands nodes such as the incinerator body, primary air fan, feeding device, and acoustic thermometer; the functional sub-module layer expands first-level collaborative control sub-functional nodes such as "Dynamic Furnace Temperature Adjustment", "Real-time Air Volume Adaptation", and "Feeding Speed ​​Control"; and the building layer expands incineration workshop nodes. The above expanded nodes are summarized to obtain the candidate sub-node set of the current level. It should be further noted that the node connection relationships stored in the hierarchical clustering function mapping space of the target factory in this embodiment include, but are not limited to, the first directed connection between the incinerator and the primary air fan, the second directed connection between the incineration workshop and the auxiliary workshop, and the function call relationship between "Incineration Process Control" and "Dynamic Furnace Temperature Adjustment".

[0065] The candidate child node set is validated based on the search constraint set. Specifically, the time validation removes nodes whose "feeding device start" timestamp is earlier than the "incinerator start" timestamp; the spatial validation removes "standby fan" nodes located outside the incineration workshop; and the logical validation removes "airflow adjustment" nodes that are not associated with "furnace temperature monitoring value". The valid child node set is obtained by retaining the nodes that pass the validation, such as the incinerator body, primary air fan, and "furnace temperature dynamic adjustment".

[0066] Based on the effective set of child nodes, with "dynamic furnace temperature adjustment" as the new current node, the node expansion and verification steps are recursively executed. Specifically, the second-level child nodes such as "furnace temperature sampling", "furnace temperature deviation calculation" and "air adjustment command generation" are expanded, and their time sequence (sampling → calculation → command generation), spatial association (specifically, including the corresponding incinerator furnace area) and logical dependency are verified until a termination node that meets the conditions for the complete realization of the "intelligent incineration" function is found. The termination node includes the "incineration process completed" flag and the execution completion flags of all necessary control commands and rendering commands such as "furnace temperature stabilized at 850℃" and "flue gas emissions meet standards" to obtain a complete path segment from the initial search node to the termination node.

[0067] Based on the hierarchical structure of the target factory's hierarchical clustering functional mapping space, the complete path segment is hierarchically divided: the building layer extracts the incineration workshop nodes and the second directed connection with the auxiliary workshop; the equipment layer extracts the incinerator, primary air fan, and other nodes and the first directed connection between the equipment; the functional sub-module layer extracts sub-functions such as "incineration full-process control" and "furnace temperature dynamic adjustment" and the corresponding secondary control command sequence; based on the association relationship of each level in the mapping space, the paths of each level are associated and integrated to obtain the hierarchical response mapping path of the "intelligent incineration" function. This path includes the connection relationship of each level node, the control command sequence such as "incinerator start-up → air volume adjustment → feed control", and the corresponding furnace temperature rendering, equipment status display, and other rendering command association information; the association relationship of each level in the mapping space includes, but is not limited to, the temperature sensor of the incinerator and the control module of the primary air fan corresponding to "furnace temperature dynamic adjustment";

[0068] S2. Based on the hierarchical response mapping path of the target function, the hierarchical synchronous response of the target function is performed by combining the preset target area 3D simulation model and the performance of the display system equipment and the real-time network status. In addition, the hierarchical synchronous response of the target function is performed in real time by combining the preset distributed monitoring node network to monitor the functional response mapping anomalies and corresponding dynamic screen rendering anomalies in the target area response, and obtain the real-time monitoring response information of the target function. It should be further noted that the performance of the display system equipment in this embodiment refers to the operating performance of the hardware device configured for the digital delivery platform in this example.

[0069] It should be further explained that the layered response in this application includes function implementation control command response and dynamic screen rendering response; the function implementation control command response and dynamic screen rendering response are associated and mapped through the associated response delay connection relationship constructed by the function implementation control command response priority, display system equipment performance, real-time network status and screen clarity, smoothness and completeness;

[0070] It should be further explained that the implementation process of the associated response delayed connection in this embodiment is as follows:

[0071] Based on the hierarchical division of the target function in the hierarchical response mapping path, including the building layer, equipment layer, and functional sub-module layer, the function implementation control command response is executed sequentially according to the hierarchical priority: first, the area control command corresponding to the building layer is triggered, such as workshop start-up and shutdown control; then, the equipment operation command of the equipment layer is executed, such as incinerator adjustment; and finally, the sub-function command of the functional sub-module layer is completed, such as temperature monitoring, to obtain the hierarchical function control command execution sequence.

[0072] Based on the hierarchical function control command execution sequence and the preset target area 3D simulation model, dynamic screen rendering response is triggered in the same hierarchical order. Specifically, the building layer renders the spatial state of the corresponding area, such as highlighting the workshop boundary; the equipment layer renders the equipment operation status, such as valve opening and closing animation; and the functional submodule layer renders the parameter visualization results, such as temperature curves, to obtain the hierarchical screen rendering sequence.

[0073] Based on the performance of the display system equipment and the real-time network status, the execution load of the layered function control commands and screen rendering sequences is monitored to obtain data on the current system's response processing capacity. The performance of the display system equipment includes processor load, memory usage, and graphics card rendering capability. The real-time network status includes, but is not limited to, bandwidth usage, transmission latency, and connection stability.

[0074] Based on the priority of the function implementation control instruction response, a processing priority is assigned to the hierarchical function control instruction execution sequence. Higher priority instructions occupy system resources first, resulting in a priority-ordered sequence of function instructions.

[0075] Based on the requirements for image clarity, smoothness, and integrity, and combined with the system's response processing capability data, the rendering parameters of the layered image rendering sequence are dynamically adjusted to ensure that the rendering effect meets the preset standards and obtain the adapted image rendering sequence. The rendering parameters include detail level, frame rate, number of image elements, and pixel resolution of the corresponding image area.

[0076] Based on the priority-sorted function instruction sequence, the adapted screen rendering sequence, and the system response processing capability data, an associated response delay connection relationship is constructed. Specifically, the time difference between the execution of the function instruction and the corresponding screen rendering is obtained, and a delay threshold is set to ensure the consistency of the timing between the function action and the screen display. The function instruction ID is bound to the rendering element ID to realize the association mapping between the two.

[0077] Based on a pre-set distributed monitoring node network, the execution status of hierarchical function control commands and the display status of corresponding screen rendering are collected in real time, such as whether they are synchronized or distorted. By comparing the threshold in the associated response delay connection relationship, abnormal function response mapping and abnormal dynamic screen rendering are identified, and real-time monitoring response information of the target function is obtained. The execution status includes whether it is completed or abnormal.

[0078] S3. Based on the real-time monitoring response information of the target function, combined with the preset forward inference model, reverse anomaly tracing and localization are performed along the hierarchical response mapping path. At the same time, according to the localized anomaly information, conflict resolution is performed in combination with the conflict resolution strategy library. The hierarchical response mapping path of the target function after conflict resolution is fed back to the three-dimensional simulation model of the target area to continue simulation and anomaly monitoring until the target function is fully realized and there are no anomalies.

[0079] S4. Feed back the hierarchical response mapping path of the target function corresponding to the target function without abnormality to the hierarchical clustering function mapping space of the target factory for real-time update, and obtain the hierarchical clustering function mapping space of the target factory after the single function block is updated.

[0080] S5. Adjust the response target function requirements and repeat the above steps until the target function is fully implemented and there are no anomalies, and obtain a fully updated target factory hierarchical clustering function mapping space.

[0081] It should be further explained that the construction process of the target factory hierarchical clustering functional mapping space in this embodiment includes:

[0082] The system acquires multi-source real-time data from the target factory, including data on the building distribution status of the target area, the distribution status and operation control logic of the configured equipment, and the operation logic data of the preset function blocks of the digital delivery platform. For example, the building distribution status data in this embodiment includes, but is not limited to, the three-dimensional layout coordinates of the workshop, the spatial positions of walls and columns, the height and zoning boundaries of floors, and the direction and width of fire escape routes. The distribution status and operation control logic information of the configured equipment includes the installation and positioning coordinates of the incinerator, the real-time speed and current of the blower, the logic rule of "triggering shutdown if temperature exceeds the standard" in the equipment control loop, and the connection relationship between sensors and actuators. The operation logic data of the preset function blocks of the digital delivery platform includes the triggering condition of "detecting furnace pressure before startup" in the "intelligent incineration" function block, the execution steps of the "one-click start / stop" function block (e.g., starting the blower before feeding materials), and the equipment vibration and temperature parameter thresholds associated with the "predictive maintenance" function block.

[0083] Based on the analysis of the building distribution data of the target area, the coordinates and boundary constraints of the factory boundary feature points are determined. The geometric center of the factory is calculated by least squares fitting and used as the origin of the coordinate system. The X-axis is the direction parallel to the main axis of the factory, the Y-axis is the horizontal direction perpendicular to the main axis, and the Z-axis is the vertical direction.

[0084] The data on the distribution of buildings in the target area are preprocessed, and the architectural drawings in CAD format are converted into a three-dimensional mesh model. At the same time, the geometric relationship information of the buildings is extracted and labeled according to the inclusion relationship of the target buildings. The geometric relationship information includes, but is not limited to, the vertex coordinates, thickness, height and material property parameters of walls, columns, floors, doors and windows.

[0085] Based on the geometric relationship information of the building, the 3D mesh model is mapped to the factory coordinate system of the target area through coordinate transformation matrix mapping. At the same time, Boolean operation processing is performed on the mapped building model to eliminate overlapping surfaces and redundant vertices, thus obtaining the target building 3D mapping model. It should be further explained that the Boolean operation processing in this embodiment is as follows: the face elements of each component are quickly retrieved through spatial hash index, and the normal vectors of the face are compared with the vertex coordinates to identify overlapping surfaces. The criteria for determining overlapping surfaces are: the overlap of face vertex coordinates is greater than or equal to a preset threshold and the normal vectors are in the same direction. Boolean difference operation is performed on the overlapping surfaces to retain the face elements of the main component, and Boolean union operation is performed on the connecting surfaces of adjacent components to eliminate gaps. At the same time, a greedy algorithm is used to traverse the model vertices, calculate the spatial distance between vertices, merge vertices with a distance ≤ a preset minimum threshold as a single vertex, delete duplicate vertex coordinate records, and finally obtain the target building 3D mapping model without overlapping surfaces and redundant vertices.

[0086] The installation location coordinates of each component in the configuration equipment distribution state are converted into coordinate values ​​in a three-dimensional coordinate system. The three-dimensional model of the corresponding equipment is retrieved according to the preset equipment three-dimensional model library. The equipment-building spatial constraint relationship and belonging relationship are established according to the building and equipment geometric information. After scaling, it is mapped to the corresponding coordinate position to obtain the initial building-equipment three-dimensional mapping model. It should be further noted that the preset equipment three-dimensional model library in this embodiment is constructed by those skilled in the art based on the specific equipment or component information configured in each building of the corresponding factory and combined with three-dimensional modeling software. For example, the belonging relationship is: equipment B is installed in the C position area of ​​building A.

[0087] Based on the equipment's operation control logic information, the control causal relationship and correlation degree between various devices are obtained through correlation analysis algorithms. For example, the operation control logic of the incinerator and the primary air fan is that before the incinerator ignition command is triggered, the primary air fan must have been started and reached the preset air volume. Through correlation analysis algorithms, it is found that the co-occurrence frequency of the incinerator ignition event and the primary air fan start event is a%, and the incinerator ignition command is blocked when the primary air fan is not started. Therefore, there is a control causal relationship between the two: primary air fan start → incinerator ignition, and the correlation degree is a.

[0088] Based on the causal relationship and degree of correlation between various devices, the corresponding device components are taken as component nodes, and the directed relationship between the corresponding components is taken as the first directed connection. At the same time, different colors are used to map the control relationship type of the first directed connection between different components. The length of the connection relationship and the depth of the color are used to map the correlation strength and the probability of correlation failure of the control relationship corresponding to the first directed connection between different components, respectively.

[0089] For example, based on the operation control logic information of the equipment and the control causal relationship and correlation degree between various devices obtained by the correlation analysis algorithm, the unique identifier and three-dimensional coordinates of the equipment components are extracted, and each equipment component is mapped to a component node in three-dimensional space to obtain a set of component nodes; based on the direction of the control causal relationship, such as the control command issuing end pointing to the receiving end, directed line segments are established between the corresponding component nodes as the first directed connection to obtain an initial set of directed connections;

[0090] Based on the preset color coding rules of the control relationship type, each connection in the initial set of directed connections is matched with the corresponding color according to its type to obtain directed connections with type identifiers; it should be further noted that the control relationship types include but are not limited to start control, parameter adjustment, and fault interlock; the preset color coding rules include but are not limited to green for start control, blue for parameter adjustment, and red for fault interlock;

[0091] Based on the correlation strength value, a length mapping function is preset, where the higher the correlation strength, the shorter the connection length. The length parameter of each directed connection with a type identifier is calculated and the line segment size is adjusted to obtain a directed connection with a length attribute.

[0092] Based on the associated fault probability values, a color depth gradient rule is preset, where the higher the fault probability, the higher the color saturation. The color depth of each directed connection with a length attribute is adjusted to obtain directed connections with fault probability identifiers. The set of component nodes and directed connections with type, length, and color depth attributes are spatially integrated using a graph algorithm to generate a device component association space containing visualization information on control relationship type, association strength, and fault probability.

[0093] Simultaneously, based on the number of connections between corresponding devices between different buildings and the correlation strength and probability of correlation failure of the first directed connection corresponding to the control relationship, a second directed connection relationship between different building nodes is constructed. The coarseness of the second directed connection relationship is used to map the second correlation strength of different buildings. The building node is constructed from each individual building and its corresponding building location and geometric attribute parameters.

[0094] Exemplarily, based on the three-dimensional building model of the target area factory, the unique identifier and spatial boundary coordinates of each building are extracted, including the vertex coordinates of the minimum bounding cuboid, and the correspondence between the building identity and the spatial range is established to obtain the building basic information set;

[0095] Based on the cross-building equipment connection list, a cross-building equipment connection matrix is constructed, and the matrix elements record the association strength and association failure probability of a single connection to obtain the equipment connection information between buildings. It should be further noted that the connection list includes, but is not limited to, the building identifiers, equipment IDs, and connection types of the equipment at both ends of the connection, such as electrical connection, pipeline connection, and data communication connection;

[0096] Based on the equipment connection information between buildings, the connection quantities are summarized by building pairs, and the weighted association strength of the building pair is calculated: taking the association strength of a single connection as the weight, the connection quantities are weighted and summed, and then multiplied by (1 - average association failure probability) for correction, where the average association failure probability is the mean of the failure probabilities of all connections of the building pair, to obtain the second association strength between buildings;

[0097] Analyze the control instruction flow direction of the equipment connections in the building pair. For example, in the equipment connection, the instruction is sent from the equipment in workshop A to the equipment in control room B for reception, and determine the direction of the second directed connection, such as from the building where the instruction is sent to the building where the instruction is received, to obtain the connection direction parameter;

[0098] Based on the second association strength, a piecewise non-linear mapping rule is used to set the connection thickness: when the association strength is in the interval [0, 0.3), the thickness increases linearly with the strength (growth rate k1); when it is in the interval [0.3, 0.7), the growth rate is increased to k2 (k2 > k1); when it is in the interval [0.7, 1], the growth rate is decreased to k3 (k3 < k1), and at the same time, a minimum thickness threshold is set to calculate the connection thickness parameter corresponding to the building pair;

[0099] Associate and integrate the building identifiers, connection direction parameters, thickness parameters corresponding to the second association strength, and connection type tags in the building basic information set to generate the second directed connection relationship between different buildings, where each connection includes a direction arrow, thickness, and type tag, and the connection thickness directly maps the second association strength of the building pair; the connection type tag, for example, "E" is marked for electrical connection;

[0100] Combine the constructed equipment component nodes, the first directed connection and its parameters, the building nodes, and the corresponding second directed connection relationship and the second association strength, and map them into the initial building-equipment three-dimensional mapping model using graph algorithms to obtain the building-equipment hierarchical global mapping space;

[0101] Set a geometric deviation threshold and verify the geometric deviation of the building and equipment in the initial building-equipment 3D mapping model. When the initial building-equipment 3D mapping model is verified, combine the simulation algorithm and the dynamic control parameters of the corresponding equipment in the building-equipment hierarchical global mapping space. Receive real-time data pushed by the edge gateway through the WebSocket protocol to verify the control logic and fault correlation causal relationship, and obtain the verified building-equipment hierarchical global mapping space.

[0102] Based on the design accuracy level of buildings and equipment, differentiated geometric deviation thresholds are set to obtain a hierarchical geometric deviation threshold set; it should be further noted that in this embodiment, buildings and equipment correspond to linear deviation thresholds and volume deviation thresholds, respectively.

[0103] Based on the initial building-equipment 3D mapping model, the surface point coordinates of the building and equipment are sampled according to the grid density. The sampled coordinates are compared with the design coordinates to calculate the linear deviation value. At the same time, the deviation value between the actual volume and the design volume of the building-equipment 3D mapping model is calculated through the grid integration algorithm to obtain the geometric deviation dataset.

[0104] The geometric deviation dataset is compared with the hierarchical geometric deviation threshold set. If all deviations do not exceed the threshold, a geometric verification pass flag is obtained.

[0105] Based on the control logic rules and fault correlation causal relationships recorded in the building-equipment hierarchical global mapping space, typical control scenarios and high-frequency fault modes are extracted to construct a verification scenario library; typical control scenarios include, but are not limited to, normal start-up and shutdown and parameter adjustment scenarios; high-frequency fault modes include, but are not limited to, equipment-level faults and cross-equipment cascading faults.

[0106] Based on the dynamic control parameters of the corresponding equipment, the parameter threshold range for control logic verification and the correlation strength threshold for fault causation verification are set to obtain the verification judgment criteria.

[0107] A long-lived connection is established with the edge gateway via the WebSocket protocol to receive real-time data at fixed intervals. A sliding window is used to cache the data to obtain a time-seriesd real-time data stream. The window duration is adapted to the control scenario cycle. The real-time data includes, but is not limited to, device operating parameters, control command execution status, and fault alarm signals.

[0108] The real-time data stream is time-aligned with the control scenarios in the verification scenario library. The matching degree between the control instruction execution sequence and the preset logic is calculated by the dynamic time warping algorithm. At the same time, the deviation between the actual value and the target value during the parameter adjustment process is tracked. If the deviation meets the preset standard, it is considered as passing, and the control logic verification result is obtained.

[0109] For fault signals in real-time data streams, a dynamic Bayesian network model is used to infer potential triggering sources and calculate the causal correlation strength between fault signals and each candidate triggering source. If the strength meets the correlation strength standard, it is considered to pass. At the same time, it is verified whether the fault chain reaction conforms to the causal path recorded in the mapping space. If the path conformity meets the preset standard, it is considered to pass, and the fault correlation causal relationship verification result is obtained.

[0110] When both the control logic verification result and the fault correlation causal relationship verification result meet the judgment criteria, the building-equipment hierarchical global mapping space is marked as having passed the verification. For items that fail, the logic rule parameters and causal relationship strength values ​​in the mapping space are corrected by combining real-time data. The verification is iterated until it passes, and the verified building-equipment hierarchical global mapping space is obtained.

[0111] This construction process, through multi-dimensional data fusion and refined modeling, achieved a comprehensive digital mapping of the target factory from physical space to functional logic, yielding significant technical results. Firstly, by integrating multi-source information such as building distribution, equipment status, and functional block logic, combined with the establishment of a factory coordinate system and the transformation to a 3D mesh model, and eliminating overlapping surfaces and redundant vertices through Boolean operations, the accuracy of the spatial relationship representation between buildings and equipment was greatly improved. Geometric deviations were strictly controlled within design thresholds, providing a high-fidelity 3D carrier for subsequent spatial correlation analysis. Secondly, based on correlation analysis algorithms, the causal relationships between equipment control were mined and visualized using a first directed connection with type, intensity, and failure probability attributes. Simultaneously, a second directed connection was constructed through a cross-building connection matrix, with coarse-grained mapping of correlation strength, clearly depicting the hierarchical control logic at the equipment and building levels. This transformed the factory operation rules from implicit to explicit, improving the interpretability of logical correlations. Third, through multiple iterations of geometric deviation verification and control logic and fault causal verification, combined with real-time data to dynamically correct model parameters, the mapping space is ensured to be reliable in both geometric accuracy and logical consistency. It can accurately reproduce the physical state of the factory and accurately simulate the control process and fault propagation path, providing a precise digital foundation for applications such as target function response and conflict detection, and effectively supporting the efficient realization of functions such as intelligent factory operation and decision optimization.

[0112] The pre-defined operational logic data of each functional block of the digital delivery platform is combined with a word segmentation and embedding model for textual preprocessing to obtain the pre-processed logical text information of each functional block.

[0113] For example, based on the pre-set operation logic data of each functional block in the digital delivery platform, the content related to the core logic of the function operation is filtered out, redundant comments, duplicate descriptions and format symbols are removed, and text fragments containing equipment operation instructions, trigger conditions and constraint relationships are identified and retained to obtain the initial logic text set; the operation logic data of each functional block includes, but is not limited to, natural language descriptions, flowchart text descriptions, rule entries, etc.

[0114] Based on a domain-specific terminology database of the functional block's domain, a domain-specific word segmentation dictionary is constructed. A bidirectional longest matching method is used to segment text fragments in the initial logical text set, prioritizing matching of specialized terms in the dictionary to avoid term fragmentation, resulting in a segmented logical text sequence. The domain includes, but is not limited to, incineration control and equipment maintenance. The specialized terminology database includes, but is not limited to, equipment names, control command terms, and logical relational terms, constructed from a distributed library combined with a deep search algorithm. The domain-specific word segmentation dictionary includes, but is not limited to, tagging terms' parts of speech and logical roles, such as "start" as an action word and "if" as a conditional conjunction.

[0115] Logical unit boundaries are identified based on logical connectors. The segmented logical text sequence is divided into trigger condition units, execution action units, and constraint units. For example, "before the incinerator starts" is a trigger condition unit, and "the primary air fan needs to run" is a constraint unit. Logical role labels are added to each unit to obtain a set of logical units with role labels. Logical connectors include "when...", "then...", "must...", "subsequently...".

[0116] Record the correspondence between synonyms and standard terms, such as mapping "open" and "start" to "start". Construct a functional block logical semantic mapping table, perform semantic standardization on the text in the logical unit set with role labels, replace synonyms with standard terms, correct inconsistencies in expression, and obtain a standardized logical unit set.

[0117] Based on the word segmentation and embedding model, each term in the standardized logical unit set is converted into a vector representation. At the same time, the role label information of the logical unit is integrated during the embedding process, so that the vectors of the trigger condition unit and the execution action unit present a differentiated distribution in the semantic space. All term vectors are integrated to form the preprocessed logical text information of each functional block, including but not limited to term vectors and logical unit role association information.

[0118] Based on the preprocessed functional block logic text information and the first-level decomposition rules, the first-level main control sub-function, the first-level collaborative control sub-function, and the causal relationship between the first-level main control sub-function and the first-level collaborative control sub-function in time and space at continuous time points are obtained; the first-level decomposition rules include the control constraint rules and screen rendering constraint rules for the operation or screen transformation of the corresponding device components of the target functional block at continuous time points.

[0119] For example, based on the terminology vectors and logical unit role association information contained in the preprocessed functional block logic text information, a first-level decomposition rule is constructed, specifically: Control constraint rules include timing trigger rules, which define the priority of trigger conditions for device component operation at consecutive time points, such as pre-triggered, synchronous, and post-triggered relationships based on the main control command. Pre-triggered triggers require state confirmation to be completed before the main control command is executed; synchronous triggers require time alignment with the main control command; and post-triggered triggers require startup according to preset logic after the main control command is executed. State dependency rules specify the pre-conditions that device component operation depends on, such as the execution of an action unit requiring the state of the trigger condition unit to be satisfied, and triggering a blocking mechanism when the state is not satisfied. Operation scope rules include, but... This is not limited to limiting the operational boundaries of device components, such as ensuring that device operations based on spatial relationships do not exceed the control permissions corresponding to the building area to which they belong; the screen rendering constraint rules include element mapping rules, which establish the correspondence between the operating state of device components and screen rendering elements, such as the highlighting of rendering elements corresponding to the device startup state, and the dynamic value update of rendering elements corresponding to the parameter adjustment state; timing synchronization rules, which stipulate that screen rendering changes must match the timing of the execution of device control commands, such as the rendering changes must start in logical order after the control command is issued, and must not be earlier than the command execution feedback; and linkage constraint rules, which limit the screen rendering linkage relationship corresponding to the operating states of multiple devices, such as the rendering elements must be displayed hierarchically according to the association strength when the states of the main device and the collaborative devices change.

[0120] Based on the constructed first-level decomposition rules, the trigger condition units, execution action units, and constraint restriction units in the preprocessed functional block logic text information are matched to identify the core execution action units that dominate the implementation of the function as first-level main control sub-functions, and the auxiliary execution action units that cooperate with the core actions as first-level collaborative control sub-functions.

[0121] By analyzing the timing triggering rules and state dependency rules, the execution order, start interval and state transmission relationship of the first-level main control sub-function and the first-level cooperative control sub-function at continuous time points are extracted to obtain the temporal causal relationship.

[0122] By analyzing the operation range rules and linkage constraint rules, the spatial distribution association of the device components corresponding to the primary main control sub-function and the primary collaborative control sub-function is extracted, and the spatial causal relationship of the rendered elements is obtained. The temporal and spatial causal relationships are integrated to obtain the primary main control sub-function, the primary collaborative control sub-function and the causal relationship between the two in time and space at continuous time points.

[0123] The primary main control sub-function is a functional sub-module that runs throughout the entire implementation process of the functional block, and its implementation time is equal to the overall implementation time of the corresponding functional block. The primary collaborative control sub-function is a functional sub-module that is implemented within a preset time period of the functional block, and the preset time period is greater than 0 and less than the overall implementation time of the corresponding functional block. For example, in the "intelligent incineration" functional block, the primary main control sub-function is "full-process control of incineration". This functional block runs continuously throughout the entire process from incinerator start-up preparation, ignition and heating, stable combustion to shutdown and cooling, and its implementation time is consistent with the overall implementation time of the "intelligent incineration" functional block. The primary collaborative control sub-functions include "dynamic adjustment of furnace temperature" (which only runs from the incinerator ignition and heating to stable combustion stage, and its implementation time is less than the overall implementation time of the "intelligent incineration" functional block) and "feed speed adaptation control" (which only adjusts the feed rate in real time according to the furnace temperature during the stable combustion stage, and its implementation time is less than the overall implementation time of the "intelligent incineration" functional block).

[0124] Based on the first-level main control sub-function and the first-level cooperative control sub-function at continuous time points, combined with the analytical algorithm and the hybrid control instructions corresponding to each control sub-function, a second-level decomposition is performed to obtain the second-level main control instruction sequence corresponding to each first-level main control sub-function and the second-level cooperative control instruction sequence corresponding to each first-level cooperative control sub-function at continuous time points.

[0125] For example, based on the primary main control sub-function and the primary collaborative control sub-function at continuous time points, the hybrid control instructions corresponding to each control sub-function are extracted. The hybrid control instructions include hardware execution instructions, software logic instructions, rendering trigger instructions and instruction association identifiers. The hybrid control instructions are structured and parsed using a parsing algorithm.

[0126] Instructions of different dimensions are separated by instruction type tags, which include hardware, software, and rendering. A mapping relationship between cross-type instructions is established based on instruction association identifiers to obtain a classified instruction set.

[0127] Based on the continuous timeline of the primary main control sub-function, the hardware execution instructions and software logic instructions in the classification instruction set are split according to the time node. The split instructions are arranged in the order of execution and a rendering trigger instruction is embedded. The rendering trigger instruction is synchronized with the hardware execution instructions to form the preliminary secondary instruction sequence of the primary main control sub-function.

[0128] Based on the preset time period of the first-level collaborative control sub-function, hardware execution instructions, software logic instructions and rendering trigger instructions corresponding to the time period in the classified instruction set are extracted, sorted according to the instruction dependency relationship, to form the preliminary second-level instruction sequence of the first-level collaborative control sub-function. Among them, the software logic instructions need to be executed based on hardware and obtain feedback results.

[0129] Record the triggering conditions of preceding and subsequent instructions, construct an instruction dependency matrix, perform timing verification on the preliminary second-level instruction sequence, and adjust the order of conflicting instructions. Specifically, when one of two dependent instructions has a planned execution time earlier than the planned execution time of the preceding instruction it depends on, insert a waiting flag bound to the completion time of the preceding instruction into the execution sequence of the subsequent instruction. This forces the subsequent instruction to be delayed until the preceding instruction is completed before it starts, thereby eliminating timing conflicts and obtaining the verified second-level instruction sequence.

[0130] A timestamp and execution status bit are added to the verified secondary instruction sequence. The timestamp is associated with each consecutive time point, precisely corresponding to the planned execution time of the instruction. The execution status bit includes three states: not executed, executing, and completed, used to track the execution progress of the instruction in real time. This ultimately yields the secondary main control instruction sequence corresponding to each primary main control sub-function and the secondary collaborative control instruction sequence corresponding to each primary collaborative control sub-function at consecutive time points. The time coverage of the secondary main control instruction sequence is completely consistent with the operating cycle of its corresponding primary main control sub-function, covering the complete time cycle from the function block's initiation to its final completion (including all stages such as preparation, execution, and termination). The time coverage of the secondary collaborative control instruction sequence is strictly limited to the preset operating time period of its corresponding primary collaborative control sub-function. Furthermore, based on the causal relationship between the primary main control sub-function and the primary collaborative control sub-function in time and space, the spatial constraints and temporal priority relationships between the secondary main control instructions and the associated secondary collaborative control instruction sequences at consecutive time points are obtained.

[0131] It should be further explained that this embodiment extracts the spatial correlation elements based on the causal relationship between the primary main control sub-function and the primary collaborative control sub-function in time and space. The spatial correlation elements include the building area to which the associated equipment component belongs, the equipment installation location level, and the spatial boundary corresponding to the operation permission. The temporal correlation elements include the order of execution of the sub-functions, the time dependency of the triggering conditions, and the time interval of state transmission, thus obtaining an initial set of spatiotemporal correlation elements.

[0132] Based on the spatial association elements in the initial spatiotemporal association element set, the three-dimensional coordinates and region identifiers of the equipment components corresponding to the secondary main control commands and associated secondary collaborative control commands in the building-equipment 3D mapping model are called. The spatial inclusion algorithm is used to determine whether the equipment components are in the same building partition. If they are in the same partition, they are marked as strong spatial association; otherwise, they are marked as weak spatial association. The effective spatial range of command execution is determined by combining the operation permission boundary. If the range is exceeded, it is marked as spatially restricted. A spatial association matrix is ​​constructed, which records the spatial association type and restricted state between commands. The spatial constraint relationship corresponding to the sequence of secondary main control commands and associated secondary collaborative control commands at continuous time points is obtained. The spatial constraint relationship includes the strong association spatial range, the weak association spatial range, and the spatial restricted condition.

[0133] Based on the temporal correlation elements in the initial spatiotemporal correlation element set, the temporal triggering rules in the causal relationship of the first-level sub-functions are analyzed. These temporal triggering rules include pre-triggered, synchronous, and post-triggered events. Mapped to the second-level instruction level, the basic temporal order between instructions is determined. Pre-triggered events correspond to the second-level collaborative control instructions needing to be executed before the second-level main control instructions. Synchronous triggering corresponds to the two maintaining time alignment. Post-triggered events correspond to the second-level collaborative control instructions needing to lag behind the second-level main control instructions. A conditional dependency chain between instructions is established by tracing the causal relationship chain. This conditional dependency chain requires the execution of the second-level collaborative control instructions to be based on the specific state feedback of the second-level main control instructions. A time conflict resolution rule is set. When multiple instructions overlap in time, the execution window of the collaborative instructions is adjusted based on the priority of the second-level main control instructions. A time correlation matrix is ​​constructed. This time correlation matrix records the temporal order, dependency conditions, and conflict resolution strategies between instructions. The temporal priority relationship between the second-level main control instructions and the associated second-level collaborative control instruction sequences at consecutive time points is obtained. This temporal priority relationship includes the basic execution order, conditional dependency relationship, and conflict resolution method.

[0134] Based on the sequence of secondary master control instructions corresponding to each functional block at consecutive time points and the sequence of secondary cooperative control instructions that are related to each secondary master control instruction, combined with the corresponding spatial constraint relationship, time priority relationship and random forest algorithm, the functional hierarchical control forest corresponding to each functional block at consecutive time points is obtained.

[0135] Based on the functional hierarchical control forest corresponding to different functional blocks, the main control instructions or cooperative control instructions with causal relationships at continuous time points are analyzed by combining the correlation analysis algorithm and the adaptive clustering algorithm. The correlation control relationship of the functional hierarchical control forest corresponding to different functional blocks at continuous time points is obtained, and the hierarchical causal correlation control forest corresponding to different functional blocks at continuous time points is obtained.

[0136] Based on the hierarchical causal control forest corresponding to different functional blocks at continuous time points, combined with the verified building-equipment hierarchical global mapping space, the secondary main control instructions or secondary collaborative control instructions are matched and aligned with the building-equipment hierarchical global mapping space through the matching algorithm. A bidirectional mapping between each secondary main control instruction or secondary collaborative control instruction and the control logic or rendering between buildings, equipment and components is established to obtain the hierarchical clustering functional mapping space of the target factory.

[0137] It should be further explained that this embodiment is based on a hierarchical causal control forest corresponding to different functional blocks at continuous time points, extracting the instruction identifier, operation object type, execution condition description, and rendering element requirements of the secondary main control instructions and secondary cooperative control instructions contained therein, and constructing an instruction feature set;

[0138] Based on the verified building-equipment hierarchical global mapping space, the regional identifiers and spatial boundaries of buildings, the unique codes and functional attributes of equipment, the installation locations and executable operations of components, and the rendering types and status parameters of visual elements are extracted to construct a spatial element index library. The index items of this spatial element index library include entity identifiers, functional descriptions, responsive command types, and visual attributes.

[0139] A bidirectional matching algorithm is used to match the instruction feature set with the spatial element index. Specifically, for control logic instructions, which include hardware execution instructions and software logic instructions, the operation object type in the instruction is accurately matched with the unique code of the device and component in the spatial element index. Logical verification is performed by combining the execution condition description with the functional attributes of the device and component. This logical verification verifies whether the device and component have the functional basis to execute the instruction, determines the one-way mapping of the control logic from the instruction to the building, device, and component, and associates the status feedback interface of the device and component with the execution status bit of the instruction, so that changes in the operating status of the device and component can trigger the instruction status update in reverse, forming a bidirectional mapping of control logic.

[0140] For rendering commands, which are rendering trigger commands, the rendering element requirements in the command are matched with the rendering type of the visual element in the spatial element index library. The command association identifier is combined with the unique code of the device and component to bind them, ensuring that the rendering element is associated with the corresponding device and component, and determining the one-way rendering mapping from the command to the visual element. At the same time, the state change interface of the visual element is associated with the trigger identifier of the command, so that the display state change of the visual element can be associated with the trigger state of the command, forming a two-way rendering mapping.

[0141] Based on the hierarchical structure of the hierarchical causal control forest, which consists of functional blocks, first-level sub-functions, and second-level instructions, the bidirectional mapping of control logic and rendering is clustered hierarchically. Mappings of the same functional block are grouped into one category, and mappings of the same level of sub-function are grouped into one sub-category. The spatial hierarchy of the building-equipment hierarchical global mapping space is also associated, which consists of buildings, equipment, and components. Finally, the hierarchical clustering functional mapping space of the target factory is obtained, which contains hierarchical bidirectional mappings of control logic and rendering.

[0142] A three-dimensional simulation model of the target region is constructed based on the hierarchical clustering functional mapping space of the target factory and the simulation algorithm.

[0143] Based on the target area 3D simulation model, combined with the real-time collected target function response requirement data and screen response adjustment space, the response and delay of control logic and screen rendering at continuous time points are verified, as well as the matching and alignment anomaly verification, to obtain the verified target area 3D simulation model.

[0144] It should be further explained that the screen response adjustment space in this embodiment is used to adjust the screen resolution of different areas of the real-time display screen based on the real-time acquisition of the display system equipment performance, network status, importance of the real-time display screen area, and the requirements for screen clarity, completeness, and smoothness of display at continuous time points. This ensures that the real-time display screen meets the corresponding threshold requirements for screen clarity, completeness, and smoothness. The importance of the real-time display screen area is determined by the correlation between the buildings, equipment, and equipment operating status, operating results, and background screen in different areas of the corresponding screen and the preset target functional requirements. The real-time display screen area includes the central area and the edge area. The edge area includes the edge area in the direction of screen movement and the edge area in the non-direction of screen movement. The edge area in the direction of screen movement is obtained by splicing the average gray value of pixels extracted from adjacent edges of the corresponding images of different frames at continuous time points with the intrinsic entropy across time points.

[0145] It should be further explained that, in this embodiment, based on each frame of the real-time display screen, an image segmentation algorithm is used to identify and extract independent regions in the screen. The independent regions include building display area, equipment display area, operation status display area, and background display area. The operation status display area contains parameter values ​​and animation identifiers. Each independent region is assigned spatial boundary coordinates and element type labels. The element type labels include building type, equipment type, status type, and background type, thus obtaining a set of regional elements.

[0146] Based on the type of response target functional requirements, a function-element association weight table is constructed. The types of response target functional requirements include intelligent operation, safety and environmental management, and intelligent decision-making. The function-element association weight table presets the basic association weights between different element types and functional requirements. Among them, in the intelligent operation requirement, the weight of equipment elements is higher than that of building elements, and in the safety and environmental management requirement, the weight of status elements is higher than that of background elements. In addition, a weighting coefficient is set for the abnormal status indicators in the operation status display area. The abnormal status indicators include alarm signals and parameters exceeding the standard, thus obtaining the association weight parameter set.

[0147] For each independent region in the regional element set, extract the functional attributes corresponding to its element type label. The functional attributes include the specific device ID associated with the device element and the parameter type corresponding to the status element. By matching the function-element association weight table, calculate the basic association degree between the region and the current target functional requirement, and then add the real-time coefficient of the operating status. The real-time alarm status coefficient is higher than the normal status coefficient, and the dynamic parameter display coefficient is higher than the static parameter coefficient to obtain the regional association degree score.

[0148] Based on the regional correlation score, combined with the spatial location attributes of the region in the image, including the region where the core equipment is located and the preset key monitoring area corresponding to the functional requirements, the weighted summation method is used to calculate the comprehensive importance of each independent region, where the correlation score accounts for a% and the spatial location weight accounts for b%. The comprehensive importance is divided into three levels: high, medium and low, to obtain the regional importance level set.

[0149] Based on the regional importance level set, a preset rendering parameter gradient rule is established: high-level regions correspond to the highest resolution and highest clarity, where the highest resolution is the highest pixel density and the highest clarity is the highest anti-aliasing level and the highest detail texture sampling rate; medium-level regions correspond to medium resolution and clarity; low-level regions correspond to basic resolution and basic clarity, where the basic resolution is the lowest pixel density and the basic clarity is simplified detail texture. At the same time, a transition coefficient is set for adjacent regions to avoid visual discontinuities caused by sudden resolution changes, thus obtaining the initial rendering parameter configuration.

[0150] The system monitors changes in the operating status display area in real time. These changes include parameter anomaly triggers and device status switching. When a status change is detected, the system updates the real-time coefficient of the corresponding area's operating status, recalculates the overall importance and adjusts the level, and synchronously updates the rendering parameters of that area. When an alarm area is temporarily upgraded from a low level to a high level, the resolution and clarity are immediately adjusted to obtain dynamically adapted rendering parameter configurations.

[0151] Dynamically adaptable rendering parameter configurations are applied to the rendering process of each frame of the image, so that high-importance areas maintain high detail display, while low-importance areas are simplified in rendering as needed, thereby achieving optimized allocation of screen resources and prominent display of information related to the target function.

[0152] It should be further explained that the process of splicing across time points in this embodiment includes:

[0153] Based on the real-time display of multiple frames acquired at continuous time points, inter-frame difference operations are performed on adjacent frames. The motion vector of each image block is calculated by a block matching algorithm. The motion vector consists of a direction vector and a displacement amplitude. Principal component analysis is performed on the motion vectors of all image blocks. The direction vector with the largest variance contribution is determined as the image movement direction. Along this direction, the image edge is divided into a front edge region and a back edge region. Perpendicular to this direction, it is divided into two non-moving direction edge regions on both sides. Each region is assigned a spatial boundary identifier and a direction attribute label.

[0154] For the front and rear edge regions along the direction of image movement, edge bands are extracted according to a preset pixel width. The edge bands include the effective display area inside the edge and the transition area outside the edge. Within the edge bands of the current frame and consecutive previous frames, a set of pixels is extracted at a uniform sampling interval. A weighted average operation is performed on the set of pixels, and the weighting coefficient increases with the distance of the pixel from the inside of the edge. The intrinsic entropy is calculated based on the probability distribution of pixel gray values, forming a gray mean sequence and an intrinsic entropy sequence for each frame, which are then summarized into an edge feature parameter set.

[0155] A cross-time stitching process is performed on the edge feature parameter set at consecutive time points: the edge pixels of the previous frame and the current frame are aligned along the time axis, and the gray-level mean sequence is linearly fitted using the least squares method to generate an inter-frame transition gray-level value sequence; a sliding window matching algorithm is applied to the intrinsic entropy sequence, the window size of which is adapted to the image update cycle, and the entropy change rate is kept continuous by adjusting the randomness of the pixel gray-level distribution within the window; the transition gray-level value sequence and the adjusted entropy distribution are integrated to form an initial cross-time stitching edge, which includes the connecting pixel band between the historical frame and the current frame and the feature parameter deviation value of the pixel band, the deviation value being the difference between the actual sequence and the ideal continuous sequence.

[0156] The structured elements of the current frame image are extracted, including the three-dimensional boundary coordinates of the building area, the geometric contour parameters of the equipment, the position coordinates of the operation status display area, and the identifiers of the associated rendering elements. The hierarchical clustering function of the target factory is invoked to map the control and rendering instructions recorded in the space for the next moment. The control instructions include the equipment motion trajectory parameters and spatial range definition values, and the rendering instructions include the update frequency of state parameters and the element style conversion rules. An association mapping table between the current image elements and the instructions for the next moment is constructed. The mapping table records the coordinate offset, state conversion identifier, and rendering priority of each image element.

[0157] Based on the association mapping table, the next moment's display screen is predicted and generated. Specifically, the spatial coordinates of the current frame image elements are transformed and deduced according to the offset in the mapping table. Combined with the element update rules in the rendering instructions, the device position and parameter display style are adaptively adjusted to generate a prediction frame containing the three-dimensional coordinates and pixel attributes of each element. The pixel attributes include gray level, texture features and transparency parameters.

[0158] Based on the spatial range of the predicted frame and the direction of image movement, the intersection of the edge region of the current frame and the edge region of the predicted frame is calculated to determine the overlapping stitching area. The overlapping stitching area is the intersection pixel band of the outer edge of the current frame and the inner edge of the predicted frame. Based on the feature parameter deviation range of the initial cross-time stitching edge, the pixel transition parameters of the overlapping stitching area are set, including the gray value mean difference threshold and the intrinsic entropy matching degree threshold. The threshold is determined by historical stitching error statistics.

[0159] Based on the feature parameter deviation values ​​of the initial cross-time stitching edge and the image elements of the predicted frame, the predicted pixel compensation value is calculated: for regions where the mean difference in grayscale values ​​exceeds a threshold, a linear gradient grayscale sequence is generated according to the grayscale trend of the corresponding region in the predicted frame, and the gradient value of the sequence increases with the deviation amount; for regions where the intrinsic entropy matching degree is lower than a threshold, a local pixel entropy adjustment algorithm is adopted to make the entropy value distribution continuous by increasing or decreasing the pixel grayscale fluctuation amount, and the adjustment amount is positively correlated with the deviation value; the above processing results are summarized to form the predicted pixel compensation value.

[0160] The predicted pixel compensation values ​​are stored in a frame buffer queue, which manages the compensation value data in timestamp order. When the next frame is generated, the corresponding compensation value is retrieved from the queue and a weighted average fusion operation is performed with the connecting pixel band of the initial cross-time stitching edge. The weighting coefficient changes linearly with the distance of the pixel from the center of the overlapping area. The edge region pixel sequence output after fusion satisfies the constraints of grayscale continuity and entropy stability, realizing seamless stitching across time points and ensuring that there is no tearing or discontinuity at the edge during the frame movement.

[0161] It should be further explained that the process of obtaining the screen response adjustment space in this embodiment includes:

[0162] Based on the hierarchical response mapping path of the real-time display screen, an image segmentation algorithm is used to identify and extract independent regions in the screen. These independent regions include building display area, equipment display area, operation status display area, and background display area. Each independent region is assigned spatial boundary coordinates and feature type labels, including building, equipment, status, and background categories. Combined with a region importance level set, which includes high, medium, and low levels, a region attribute matrix is ​​constructed. The region attribute matrix records the region identifier, feature type, importance level, and initial resolution parameters.

[0163] The system collects and displays system device performance data and network status data in real time. The system device performance data includes processor load rate, memory usage rate, and graphics card rendering frame rate. The network status data includes bandwidth utilization rate, data transmission latency, and packet loss rate. The device performance data is converted into device load factor through normalization processing. The value range of the device load factor corresponds to the increasing load level. The network status data is converted into network quality factor. The value range of the network quality factor corresponds to the decreasing quality level.

[0164] Based on the requirements for image clarity, integrity, and smoothness, a requirement weight coefficient is set. The clarity requirement corresponds to the importance level of the area, with higher levels having higher weights than medium and low levels. The smoothness requirement is related to the device load factor and network quality factor. The higher the load and the worse the quality, the higher the smoothness weight. The integrity requirement is fixed as the baseline value, and a requirement weight vector is constructed.

[0165] A screen resolution adjustment model is constructed. The initial resolution parameters in the regional attribute matrix are used as the benchmark. The product of the device load factor and the network quality factor is used as the system constraint coefficient. The product of the demand weight vector and the regional importance level is used as the regional priority coefficient. The screen resolution adjustment coefficient is calculated by the ratio of the system constraint coefficient and the regional priority coefficient. When the coefficient is greater than 1, the screen resolution is increased. When the coefficient is less than 1, the screen resolution is decreased.

[0166] Differential adjustments are made for the central and edge regions. The edge regions include the edge regions along the image movement direction and the edge regions without movement direction. The central region directly corrects the resolution parameters according to the image resolution adjustment coefficient. The lower limit of the coefficient for high importance regions is the baseline value. The edge regions along the image movement direction have a cross-time splicing compensation coefficient superimposed in the coefficient calculation. The cross-time splicing compensation coefficient is set based on the continuity deviation between the mean gray value and the intrinsic entropy. The edge regions without movement direction are multiplied by the basic adjustment coefficient according to the region importance level. The basic adjustment coefficient is lower than the adjustment range of the central region.

[0167] A screen resolution transition mechanism is established. When the difference in screen resolution adjustment coefficients between adjacent frames exceeds the transition threshold, a step transition sequence is generated. The step transition sequence gradually approaches the target screen resolution at frame intervals. At the same time, the transition coefficients in the rendering parameter gradient rules are associated to ensure that there are no visual gaps in the resolution change.

[0168] The system monitors the actual clarity, smoothness, and completeness of the adjusted image in real time. The actual clarity is calculated using the structural similarity index, the smoothness is calculated using the frame interval fluctuation rate, and the completeness is calculated using the region missing rate. The monitored values ​​are compared with the preset requirement thresholds. If they are not met, the system constraint coefficient and the region priority coefficient are recalculated, and the adjustment steps are executed iteratively.

[0169] By integrating the regional attribute matrix, screen resolution adjustment model, differentiated adjustment rules and transition mechanism, a screen response adjustment space is formed. This space contains a dynamically updated screen resolution parameter table and corresponding rendering parameter adaptation rules, which can automatically output the optimal screen resolution configuration for each region based on real-time data.

[0170] Based on the verified three-dimensional simulation model of the target area, combined with the hierarchical response mapping path of the target function, and combined with the forward inference model constructed by the hidden Markov algorithm, the sub-functional conflict detection and verification within a single target function and the multi-target functional conflict detection and verification are carried out along the connection relationship of each functional block in the hierarchical clustering functional mapping space of the target factory. The functional conflict location information and forward conflict risk transfer probability matrix corresponding to a single or cross-target function at continuous time points are obtained.

[0171] It should be further explained that the steps for obtaining functional conflict location information and the forward conflict risk transfer probability matrix in this embodiment include:

[0172] All nodes in the building layer, equipment layer, and functional block layer of the hierarchical response mapping path are extracted as hidden state nodes. Five types of conflict state attributes are defined for each node: no conflict, resource conflict, timing conflict, spatial conflict, and logical conflict. The mapping relationship between the node and real-time monitoring parameters such as the feedback signal of the device actuator, spatial coordinate offset, instruction execution delay, and rendering element distortion identifier in the 3D simulation model of the target area is established as the observed state variable to obtain the state space of the hidden Markov model.

[0173] Based on the first directed connection of the causal relationship between nodes in the hierarchical response mapping path and the second directed connection of the spatial dependency relationship, the frequency of conflict state transitions between adjacent nodes is statistically analyzed based on historical verification data. When the source node is in a conflict-free state and the target node meets the spatiotemporal constraint rules and resource load requirements, the probability of transitioning to a conflict-free state is set as the baseline value. When the source node is in a resource conflict state and the resources of the target node overlap with the conflicting resources of the source node, the probability of transitioning to a resource conflict state is set to increase. When the source node is in a temporal conflict state and the instruction execution windows of the target node overlap and there is no buffering mechanism, the probability of transitioning to a temporal conflict state is set to increase. The transition rules of all node pairs are integrated to generate a state transition probability matrix.

[0174] The conflict state attributes of the hidden state nodes are associated with the observed state variables. Resource conflict states are associated with resource occupancy exceeding the limit and texture loading failure events. Temporal conflict states are associated with instruction execution delay exceeding the threshold and spatial coordinate offset mutation events. Logical conflict states are associated with instruction sequence reversal and device status feedback not matching expectations events. The occurrence probability of observed events under each conflict state is statistically analyzed to form a matching probability set between observed state variables and hidden conflict states, thus constructing an observation probability matrix.

[0175] The complete hierarchical response mapping path of the input target function is extracted as a hidden state sequence. Based on the real-time acquired observation state variable sequence, the Viterbi decoding algorithm is used to start from the previous node and calculate the probability of each conflict state transitioning to the current node by combining the state transition probability matrix. The current node's observed state variables are associated with the matching probability calculated by the observation probability matrix. The conflict probability of all path nodes is iteratively calculated and the global optimal conflict state sequence is obtained by backtracking. The conflict location information and conflict type identifier of each sub-function node within a single target function are output.

[0176] Extract the hierarchical response mapping path of parallel target functions, identify path intersection nodes, construct the device resource competition fusion state and space occupation conflict fusion state at the intersection nodes, expand the hidden Markov model state space and add fusion conflict states and their transition probabilities, based on the multi-path observation state variable sequence, use the forward-backward algorithm to calculate the state probability from the forward starting node to the intersection node and the state probability from the backward ending node to the intersection node, fuse the forward and backward probabilities to obtain the fusion conflict state probability of the intersection node, and output the conflict location information and conflict correlation matrix of cross-target functions.

[0177] Based on the conflict location information, the hierarchical attributes and conflict types of the conflict nodes are extracted. Starting from the current conflict node, the subsequent nodes are traversed along the directed connection of the hierarchical response mapping path. The probability of the conflict state transitioning to the subsequent node is calculated according to the state transition probability matrix. The transition probabilities and conflict type inheritance relationships of all conflict nodes to the subsequent nodes are integrated to generate a forward conflict risk transition probability matrix with clear downstream conflict triggering probability and type mapping rules.

[0178] Based on the functional conflict location information and forward conflict risk transfer probability matrix corresponding to single or cross-target functions at continuous time points, starting from the component nodes or building nodes where conflicts exist, and combining the forward conflict risk transfer probability matrix with the preset conflict resolution strategy library, conflict resolution and connection mapping updates are performed along the nodes and connection relationships in the hierarchical clustering functional mapping space of the target factory, and the three-dimensional simulation model of the target area after real-time update of conflict resolution and connection mapping is obtained.

[0179] It should be further explained that the construction process of the preset conflict resolution strategy library in this embodiment includes:

[0180] Based on historical conflict case data and corresponding resolution records of the target factory, the historical conflict case data includes the conflict occurrence timestamp, unique identifier of the equipment involved and its building area, and conflict type, including but not limited to resource preemption / time sequence offset / spatial overlap / logical contradiction, and associated functional block ID. The corresponding resolution records include the executed control command sequence, rendering parameter adjustment values, and conflict resolution success rate. Conflict feature keywords, such as "incinerator-01", "resource preemption", and "10:00:00", are extracted using a bidirectional longest matching word segmentation algorithm in natural language processing. Entity linking technology is used to map these keywords to a preset equipment encoding table, functional block dictionary, and time format specification. Simultaneously, a rule engine parses the command parameters and effect evaluation values ​​in the resolution actions to obtain standardized historical strategy entries containing conflict feature vectors, resolution action sequences (including command IDs and parameters), and effect evaluation matrices (including success rate and resource consumption values). Command parameters include, for example, "delay setting value" and "priority identifier"; effect evaluation values ​​include, for example, "success rate index"; and the conflict feature vector consists of equipment ID, conflict type code, and normalized time parameter value.

[0181] Based on standardized historical strategy entries, a strategy graph structure is constructed using graph database modeling technology. The attributes of conflict type nodes include feature vectors and occurrence frequency (number of occurrences per unit time). Strategy node attributes include applicable scenario labels, such as "device-level resource conflict" or "building-level spatial conflict," execution action sequences (stored as an array of instruction IDs and corresponding parameters), historical success rate (recent average), and resource consumption coefficient (representing the degree of resource consumption; lower values ​​indicate lower consumption). Edge attributes between nodes include matching degree (calculated using cosine similarity, ranging from 0 to 1; higher values ​​indicate better matching) and substitution coefficient (0 to 1, representing the feasibility of strategy A replacing strategy B). Conflict type nodes and strategy nodes are connected via "applicability" edges, and strategy nodes are connected via "substitution" edges, thus obtaining the initial strategy graph database. The feature vectors of conflict type nodes include feature dimensions such as device level weight, time offset, and spatial coordinate deviation.

[0182] Based on an initial policy graph database, a deep reasoning model is constructed using a hybrid algorithm of case-based reasoning and rule-based reasoning. The case-based reasoning module uses a weighted Euclidean distance algorithm, assigning weights based on device ID matching degree, conflict type encoding, time parameters, and spatial parameters to calculate the similarity between the feature vector of a new conflict and the feature vector of a conflict type node in the policy graph, returning the Top-N similar conflict nodes and their associated policy nodes. The rule-based reasoning module has a built-in conflict-policy adaptation rule library, such as "stop instructions for security-level devices have higher priority than start instructions" and "spatial overlap conflicts preferentially adopt the region division strategy." When the policy node returned by case-based reasoning does not match the device model of a new conflict, the execution action parameters in the policy are corrected through parameter mapping rules (such as adjusting instruction delay parameters according to device specifications). For new conflicts without similar cases, an initial resolution strategy is generated using a decision tree algorithm, with conflict type, security level of the involved devices, and functional block priority as splitting features, resulting in a deep reasoning model that includes case retrieval, parameter adaptation, and new policy generation functions.

[0183] Based on the core parameters of policy nodes in the policy graph database, including security weight (e.g., policies associated with security-level devices have higher security weights than policies associated with ordinary devices), historical success rate (the probability that a policy has successfully resolved conflicts in the past), and resource consumption coefficient (the degree of system resource consumption during policy execution), a weighted summation algorithm with preset weight allocation is used to comprehensively calculate the above three parameters according to a set proportion (e.g., security weight has the highest proportion, followed by historical success rate, and finally resource consumption coefficient), to obtain the priority score of each policy; then the policies are sorted in descending order of score, finally forming a priority ranking rule that can be directly used for policy selection when resolving conflicts.

[0184] Based on the need to verify the effectiveness of the new conflict resolution strategy, two types of core data were collected: First, 3D simulation model verification data, including the simulated conflict resolution success rate (the percentage of scenarios where the strategy successfully eliminates conflicts), instruction execution delay error (the deviation between actual execution time and planned time), timing conflict elimination rate (the proportion of timing contradictions corrected by the strategy to the total initial conflicts), and peak resource consumption (the maximum consumption of system resources during strategy execution); Second, field test data, including actual device response time (the interval between receiving an instruction and starting execution), screen rendering synchronization (the time difference between the execution of a function instruction and the rendering of the corresponding screen), device status feedback accuracy (the degree of matching between the actual device status and the system's recorded status), and anomaly recovery time (the time it takes for the system to recover to a normal state after the strategy is triggered). Based on the two types of data and the correlation analysis algorithm, the consistency and reliability of the strategy in virtual simulation and real-world scenarios were verified.

[0185] When the simulation verification success rate meets the preset standard and the measured effect meets the preset threshold (including latency error and rendering synchronization index), the new strategy entry is added as a node to the strategy graph database through the incremental update algorithm. At the same time, the matching degree between the new strategy and the existing conflict type nodes is calculated and "applicable" edges are established. The "substitution" edge attributes between related strategy nodes are updated. The historical success rate and priority score of the strategy node are recalculated at a preset period using the sliding window algorithm (the window size is the preset duration) to obtain a dynamically updated conflict resolution strategy library.

[0186] This embodiment of the technical solution uses a hierarchical clustering functional mapping space of the target factory as its core carrier to achieve efficient and reliable control of the entire process of factory functions from response to operation, with significant technical effects. Firstly, multi-source data fusion modeling and refined processing construct a hierarchical space containing a two-way mapping of physical space and functional logic, accurately associating buildings, equipment, and control commands, providing a high-fidelity digital foundation for functional response and ensuring the accuracy of functional path planning. Secondly, the hierarchical response mechanism, combined with distributed monitoring, synchronously executes control and rendering commands hierarchically, capturing anomalies in real time, ensuring the synchronization and accuracy of functional execution and visual display. Thirdly, a forward inference model based on Hidden Markov Models accurately detects intra-functional and cross-functional conflicts, and, combined with a conflict resolution strategy library, quickly locates and resolves conflicts, significantly improving the stability of system operation and reducing the risk of functional execution interruption. Fourthly, the visual response adjustment space dynamically adapts to the screen resolution, combined with cross-time-point edge stitching technology, optimizing rendering resources based on device performance, network status, and regional importance, while ensuring continuous, unbroken screen edges, significantly improving the clarity, smoothness, and visual integrity of the display. Ultimately, by continuously optimizing the mapping space through a feedback update mechanism, a closed loop of functional response from planning, simulation, anomaly handling to iterative optimization was achieved, effectively supporting the efficient implementation of functions such as intelligent factory operation and decision-making.

[0187] Example 2

[0188] Please see Figure 2 Another embodiment of the present invention provides a digital delivery topology mapping system for multi-source real-time data fusion, comprising: a response module, a simulation module, an anomaly reasoning and resolution module, and a feedback update module;

[0189] The response module is used to respond to the target functional requirements by combining the preset target factory hierarchical clustering functional mapping space with a deep search algorithm to obtain the hierarchical response mapping path of the target function.

[0190] The simulation module, based on the hierarchical response mapping path of the target function, combines a preset 3D simulation model of the target area with the performance of the display system equipment and the real-time network status to perform hierarchical synchronous response of the target function. It also combines a preset distributed monitoring node network to monitor in real time the functional response mapping anomalies and corresponding dynamic image rendering anomalies in the target area response, and obtain real-time monitoring response information of the target function. The hierarchical response includes the function implementation control command response and the dynamic image rendering response. The function implementation control command response and the dynamic image rendering response are correlated and mapped through the correlation response delay connection relationship constructed by the control command response priority, the performance of the display system equipment, the real-time network status, and the image clarity, smoothness, and completeness.

[0191] The anomaly reasoning and resolution module, based on the real-time monitoring response information of the target function, combines the Hidden Markov algorithm and the constructed forward reasoning model to perform reverse anomaly tracing and localization along the hierarchical response mapping path. At the same time, it resolves conflicts based on the located anomaly information and the conflict resolution strategy library, and feeds back the hierarchical response mapping path of the target function after conflict resolution to the three-dimensional simulation model of the target area to continue simulation and anomaly monitoring until the target function is fully realized and there are no anomalies.

[0192] The feedback update module is used to feed back the hierarchical response mapping path of the target function corresponding to the target function without anomalies to the hierarchical clustering function mapping space of the target factory for real-time update, so as to obtain the hierarchical clustering function mapping space of the target factory after the single function block is updated; at the same time, the response target function requirements are adjusted, and the above steps are repeated until the target function is fully implemented and there are no anomalies, so as to obtain the fully updated hierarchical clustering function mapping space of the target factory.

[0193] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art, under the guidance of the present invention, can make changes, modifications, substitutions and variations to the above embodiments without departing from the spirit and scope of the claims. All of these variations are within the protection scope of the present invention.

[0194] If the technical solution disclosed herein involves personal information, the product using this technical solution has clearly informed the user of the personal information processing rules and obtained the user's voluntary consent before processing the personal information. If the technical solution disclosed herein involves sensitive personal information, the product using this technical solution has obtained the user's separate consent before processing the sensitive personal information, and also meets the requirement of "express consent". For example, at personal information collection devices such as cameras, clear and prominent signs are set up to inform users that they have entered the scope of personal information collection and that personal information will be collected. If an individual voluntarily enters the collection scope, it is deemed that they have agreed to the collection of their personal information; or on the personal information processing device, with clear signs / information informing users of the personal information processing rules, authorization is obtained from the individual through pop-up information or by asking the individual to upload their personal information; wherein, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the types of personal information processed.

Claims

1. A digital delivery topology mapping method based on multi-source real-time data fusion, characterized by the following steps: include: By combining the preset hierarchical clustering function mapping space of the target factory with the deep search algorithm in response to the target functional requirements, the hierarchical response mapping path of the target function is obtained. The hierarchical clustering function mapping space of the target factory is constructed by combining multi-source data of the factory with association analysis algorithms and clustering algorithms; The process of constructing the hierarchical clustering functional mapping space of the target factory includes: Based on the equipment's operation control logic information, the control causal relationship and correlation degree between various devices are obtained through correlation analysis algorithms; Based on the causal relationship and degree of correlation between various devices, the corresponding device components are taken as component nodes, and the directed relationship between the corresponding components is taken as the first directed connection. At the same time, different colors are used to map the control relationship type of the first directed connection between different components. The length of the connection relationship and the depth of the color are used to map the correlation strength and the probability of correlation failure of the control relationship corresponding to the first directed connection between different components, respectively. Simultaneously, based on the number of connections between corresponding devices between different buildings and the correlation strength and probability of correlation failure of the first directed connection corresponding to the control relationship, a second directed connection relationship between different building nodes is constructed. The coarseness of the second directed connection relationship is used to map the second correlation strength of different buildings. The building node is constructed from each individual building and its corresponding building location and geometric attribute parameters. The constructed device component nodes, the first directed connection and the parameters of the first directed connection, the building nodes and the corresponding second directed connection relationship and the second association strength are combined with the graph algorithm and mapped to the initial building-equipment 3D mapping model to obtain the building-equipment hierarchical global mapping space; The digital delivery platform pre-processes the pre-defined operational logic data of each functional block by combining it with a word segmentation and embedding model to obtain the pre-processed logical text information of each functional block. Based on the preprocessed functional block logic text information and the first-level decomposition rules, the first-level main control sub-function, the first-level collaborative control sub-function, and the causal relationship between the first-level main control sub-function and the first-level collaborative control sub-function in time and space at continuous time points are obtained. The first-level decomposition rules include control constraint rules and screen rendering constraint rules for the operation or screen transformation of the device components corresponding to the target functional block at continuous time points. The first-level main control sub-function is a functional sub-module that runs through the entire functional block implementation process, and its corresponding implementation time is equal to the implementation time of the corresponding functional block; the first-level collaborative control sub-function is a functional sub-module that is implemented within a preset time period of the functional block, and the preset time period is greater than 0 and less than the overall implementation time of the corresponding functional block. Based on the first-level main control sub-function and the first-level cooperative control sub-function at continuous time points, combined with the analytical algorithm and the hybrid control instructions corresponding to each control sub-function, a second-level decomposition is performed to obtain the second-level main control instruction sequence corresponding to each first-level main control sub-function and the second-level cooperative control instruction sequence corresponding to each first-level cooperative control sub-function at continuous time points. Simultaneously, based on the causal relationship between the primary main control sub-function and the primary cooperative control sub-function in time and space, the spatial constraint relationship and temporal priority relationship between the secondary main control instruction and the related secondary cooperative control instruction sequence at continuous time points are obtained. Based on the sequence of secondary master control instructions corresponding to each functional block at consecutive time points and the sequence of secondary cooperative control instructions that are related to each secondary master control instruction, combined with the corresponding spatial constraint relationship, time priority relationship and random forest algorithm, the functional hierarchical control forest corresponding to each functional block at consecutive time points is obtained. Based on the functional hierarchical control forest corresponding to different functional blocks, the main control instructions or cooperative control instructions with causal relationships at continuous time points are analyzed by combining the correlation analysis algorithm and the adaptive clustering algorithm. The correlation control relationship of the functional hierarchical control forest corresponding to different functional blocks at continuous time points is obtained, and the hierarchical causal correlation control forest corresponding to different functional blocks at continuous time points is obtained. Based on the hierarchical causal control forest corresponding to different functional blocks at continuous time points, combined with the verified building-equipment hierarchical global mapping space, the secondary main control instructions or secondary collaborative control instructions are matched and aligned with the building-equipment hierarchical global mapping space through the matching algorithm. A bidirectional mapping between each secondary main control instruction or secondary collaborative control instruction and the control logic or rendering between buildings, equipment and components is established to obtain the hierarchical clustering functional mapping space of the target factory. Based on the hierarchical response mapping path of the target function, the hierarchical synchronous response of the target function is carried out by combining the preset target area 3D simulation model and display system equipment performance and real-time network status. In addition, the hierarchical synchronous response of the target function is carried out in real time by combining the preset distributed monitoring node network to monitor the functional response mapping anomalies and corresponding dynamic screen rendering anomalies in the target area response, and obtain the real-time monitoring response information of the target function. Based on the real-time monitoring response information of the target function, combined with the preset forward inference model, reverse anomaly tracing and localization are performed along the hierarchical response mapping path. At the same time, based on the localized anomaly information, conflict resolution is performed in combination with the conflict resolution strategy library. The hierarchical response mapping path of the target function after conflict resolution is fed back to the three-dimensional simulation model of the target area to continue simulation and anomaly monitoring until the target function is fully realized and there are no anomalies. The hierarchical response mapping path of the target function corresponding to the target function without abnormality is fed back to the hierarchical clustering function mapping space of the target factory for real-time update, so as to obtain the hierarchical clustering function mapping space of the target factory after the single function block is updated.

2. The digital delivery topology mapping method for multi-source real-time data fusion as described in claim 1, characterized in that, The process of constructing the hierarchical clustering functional mapping space of the target factory also includes: Acquire multi-source real-time information of the target factory, including building distribution status data of the target area, configuration equipment distribution status and operation control logic information, and operation logic data of preset function blocks of the digital delivery platform; Based on the analysis of the building distribution data of the target area, the coordinates and boundary constraints of the factory boundary feature points are determined. The geometric center of the factory is calculated by least squares fitting and used as the origin of the coordinate system. The X-axis is the direction parallel to the main axis of the factory, the Y-axis is the horizontal direction perpendicular to the main axis, and the Z-axis is the vertical direction. The data on the distribution of buildings in the target area are preprocessed, and the architectural drawings in CAD format are converted into a three-dimensional mesh model. At the same time, the geometric relationship information of the buildings is extracted and labeled according to the inclusion relationship of the target buildings. Based on the geometric relationship information of the building, the three-dimensional mesh model is mapped to the factory coordinate system of the target area through coordinate transformation matrix mapping. At the same time, Boolean operation processing is performed on the mapped building model to eliminate overlapping surfaces and redundant vertices, so as to obtain the three-dimensional mapping model of the target building.

3. The digital delivery topology mapping method for multi-source real-time data fusion as described in claim 2, characterized in that, The process of constructing the hierarchical clustering functional mapping space of the target factory also includes: The installation position coordinates of each component in the configuration equipment distribution state are converted into coordinate values ​​in a three-dimensional coordinate system. The three-dimensional model of the corresponding equipment is retrieved according to the preset equipment three-dimensional model library. The equipment-building spatial constraint relationship and belonging relationship are established according to the building and equipment geometric information. After scaling, it is mapped to the corresponding coordinate position to obtain the initial building-equipment three-dimensional mapping model.

4. The digital delivery topology mapping method for multi-source real-time data fusion as described in claim 3, characterized in that, The process of constructing the hierarchical clustering functional mapping space of the target factory also includes: A geometric deviation threshold is set, and the geometric deviation of the building and equipment is verified on the initial building-equipment 3D mapping model. When the initial building-equipment 3D mapping model is verified, the building-equipment hierarchical global mapping space is combined with the simulation algorithm and the dynamic control parameters of the corresponding equipment. Real-time data pushed by the edge gateway is received through the WebSocket protocol to verify the control logic and fault correlation causal relationship, and the verified building-equipment hierarchical global mapping space is obtained.

5. The digital delivery topology mapping method for multi-source real-time data fusion as described in claim 4, characterized in that, The process of constructing the hierarchical clustering functional mapping space of the target factory also includes: A three-dimensional simulation model of the target region is constructed based on the hierarchical clustering functional mapping space of the target factory and the simulation algorithm. Based on the target area 3D simulation model, combined with the real-time collected target function response requirement data and screen response adjustment space, the response and delay of control logic and screen rendering at continuous time points are verified, as well as the matching and alignment anomaly verification, to obtain the verified target area 3D simulation model. Based on the verified three-dimensional simulation model of the target area, combined with the hierarchical response mapping path of the target function, and combined with the forward inference model constructed by the hidden Markov algorithm, the sub-functional conflict detection and verification within a single target function and the multi-target functional conflict detection and verification are carried out along the connection relationship of each functional block in the hierarchical clustering functional mapping space of the target factory. The functional conflict location information and forward conflict risk transfer probability matrix corresponding to a single or cross-target function at continuous time points are obtained. Based on the functional conflict location information and forward conflict risk transfer probability matrix corresponding to single or cross-target functions at continuous time points, starting from the component nodes or building nodes where conflicts exist, and combining the forward conflict risk transfer probability matrix with the preset conflict resolution strategy library, conflict resolution and connection mapping updates are performed along the nodes and connection relationships in the hierarchical clustering functional mapping space of the target factory, so as to obtain the three-dimensional simulation model of the target area after real-time update of conflict resolution and connection mapping.

6. The digital delivery topology mapping method for multi-source real-time data fusion as described in claim 5, characterized in that, The screen response adjustment space is used to adjust the screen resolution of different areas of the real-time display screen based on the real-time acquisition of the display system equipment performance, network status, importance of the real-time display screen area, and the requirements for screen clarity, completeness, and smoothness at continuous time points. This ensures that the real-time display screen meets the corresponding threshold requirements for screen clarity, completeness, and smoothness. The importance of the real-time display screen area is determined by the correlation between the buildings, equipment, and equipment operating status and results corresponding to different areas in the screen and the background screen with the preset target functional requirements. The real-time display screen area includes the central area and the edge area. The edge region includes the edge region along the image movement direction and the edge region outside the image movement direction; the edge region along the image movement direction is obtained by splicing the average gray value of the pixels at the adjacent edges of the corresponding images of different frames at continuous time points with the intrinsic entropy across time points.

7. A digital delivery topology mapping system based on multi-source real-time data fusion, implemented according to any one of claims 1-6, characterized in that, include: The module consists of a response module, a simulation module, an anomaly reasoning and resolution module, and a feedback update module. The response module is used to respond to the target functional requirements by combining the preset target factory hierarchical clustering functional mapping space and deep search algorithm to obtain the hierarchical response mapping path of the target function. The simulation module performs hierarchical synchronous response of the target function based on the hierarchical response mapping path of the target function, combined with the preset three-dimensional simulation model of the target area and the display system equipment performance and real-time network status. It also combines the preset distributed monitoring node network to monitor the functional response mapping anomalies and corresponding dynamic screen rendering anomalies in the target area response in real time, thereby obtaining real-time monitoring response information of the target function. The layered response includes function implementation control command response and dynamic screen rendering response; the function implementation control command response and dynamic screen rendering response are associated and mapped through the associated response delay connection relationship constructed by the control command response priority, display system device performance, real-time network status and screen clarity, smoothness and completeness. The anomaly reasoning and resolution module, based on the real-time monitoring response information of the target function, combines the Hidden Markov algorithm and the constructed forward reasoning model to perform reverse anomaly tracing and localization along the hierarchical response mapping path. At the same time, it resolves conflicts based on the located anomaly information and the conflict resolution strategy library, and feeds back the hierarchical response mapping path of the target function after conflict resolution to the three-dimensional simulation model of the target area to continue simulation and anomaly monitoring until the target function is fully realized and there are no anomalies. The feedback update module is used to feed back the hierarchical response mapping path of the target function corresponding to the target function without abnormality to the hierarchical clustering function mapping space of the target factory for real-time update, so as to obtain the hierarchical clustering function mapping space of the target factory after the single function block is updated.

Images