A device-oriented multi-source data fusion and information system integration control system

Abstract

The application discloses a kind of equipment-oriented multi-source data fusion and information system integration control system, belong to industrial automation and information processing technical field.It includes: data acquisition layer, for real-time acquisition of the environment and state data of physical equipment;Data transmission layer, for transmitting the environment and state data to data fusion layer;Data fusion processing layer, for pre-processing the environment and state data, to obtain basic operation data;Intelligent processing layer, for state evaluation, fault prediction and generating control instruction based on basic operation data;Control execution layer, for converting control instruction into operation action instruction, and operation action instruction is applied to physical equipment;Application service layer, for providing visual monitoring interface, and sending abnormal alarm information to user.The application adopts layered architecture design, each layer responsibility is clear, loose coupling between modules, facilitate system extension, maintenance and upgrade.

Description

Technical Field

[0001] This invention belongs to the field of industrial automation and information processing technology, specifically relating to a multi-source data fusion and information system integration control system for equipment. Background Technology

[0002] With the rapid development of IoT, edge computing, and intelligent control technologies, a large number of heterogeneous physical devices, such as sensors, actuators, programmable logic controllers (PLCs), and smart meters, are deployed in industrial settings. These devices are typically provided by different manufacturers and support multiple communication protocols (such as Modbus, OPC UA, MQTT, and PROFINET), resulting in significant differences in data formats, semantic representations, sampling frequencies, and time bases, thus creating the problem of "information silos."

[0003] Existing multi-source data fusion and control systems mostly employ centralized data processing architectures, which struggle to address issues such as spatiotemporal inconsistencies and semantic discrepancies in data from heterogeneous devices. Furthermore, traditional systems lack unified quality assessment and closed-loop adaptive mechanisms in data acquisition, transmission, fusion, and control command execution, resulting in deficiencies in real-time performance, reliability, and scalability. In addition, existing systems offer weak support for equipment status assessment, fault prediction, and control command generation, lacking digital twin-based visualization monitoring and risk warning capabilities, thus failing to meet the demands of modern industry for intelligent control and efficient operation and maintenance.

[0004] Therefore, there is an urgent need for a device-oriented multi-source data fusion and information system integration control system that can achieve unified access to heterogeneous devices, efficient data fusion, intelligent status assessment and fault prediction, as well as closed-loop adaptive control, thereby improving the overall performance and reliability of industrial automation systems. Summary of the Invention

[0005] The purpose of this invention is to address the problems in existing industrial automation systems, such as inconsistent data formats and semantics between heterogeneous devices, difficulties in spatiotemporal alignment, lack of data quality evaluation mechanisms, lack of closed-loop adaptive capability in control command execution, and poor system scalability, by providing a device-oriented multi-source data fusion and information system integration control system.

[0006] To achieve the above objectives, the present invention provides the following solution: a device-oriented multi-source data fusion and information system integration control system, comprising: a data acquisition layer, a data transmission layer, a data fusion processing layer, an intelligent processing layer, a control execution layer, and an application service layer; The data acquisition layer is connected to the device and the data transmission layer, and is used to collect environmental and status data of the physical device in real time. The data transmission layer is also connected to the data fusion processing layer, and is used to transmit the environment and state data to the data fusion layer; The data fusion processing layer is also connected to the intelligent processing layer, and is used to preprocess the environmental and state data to obtain basic operating data; The intelligent processing layer is also connected to the control execution layer, and is used to perform status assessment, fault prediction and generate control commands based on the basic operating data. The control execution layer is also connected to the data acquisition layer and is used to convert the control instructions into operation action instructions, which are applied to the physical device. The application service layer is connected to the intelligent processing layer and the data acquisition layer to provide a visual monitoring interface and send abnormal alarm information to the user.

[0007] More preferably, the data acquisition layer includes: an internal identifier generation module, a real-time data acquisition module, and a device reverse interface; The internal identifier generation module is used to access the physical device based on the protocol supported by the physical device and generate a unique internal identifier. The real-time data acquisition module is used to acquire the environmental and status data of the physical device based on the operating mode. The device reverse interface is used to perform security verification on the operation command and then transmit the operation command to the physical device.

[0008] More preferably, the data fusion processing layer includes: a spatiotemporal alignment module, a semantic alignment and mapping module, a multi-source data distribution module, a multimodal data fusion module, and a fusion quality assessment and repair module; The environment and state data are sequentially processed by the spatiotemporal alignment module and the semantic alignment and mapping module for spatiotemporal window alignment and semantic alignment to obtain mapped serialized data. The multi-source data distribution module is connected to the multimodal data fusion module and is used to distribute the mapped serialized data to different multimodal data fusion modules based on the internal identifier; The multimodal data fusion module is also connected to the fusion quality assessment and repair module, and is used to perform data fusion on the mapped serialized data using different fusion modes to obtain fused data; The fusion quality assessment and repair module is used to generate data quality tags for the fusion data, give repair instructions based on the results of the data instruction tags, and obtain the basic operating data.

[0009] More preferably, the intelligent processing layer includes: a rule processing module, a workflow maintenance module, a control command generation and issuance module, and a closed-loop control and adaptive optimization module; The rule processing module is used to continuously query and match patterns on the basic operational data based on a predefined rule set to obtain a list of actions to be executed. The workflow maintenance module is used to model the list of actions to be executed as a directed graph; The control command generation and distribution module is used to generate control commands based on the directed graph and distribute the control commands to the control execution layer; The closed-loop control and adaptive optimization module is used to perform status assessment and fault prediction based on the basic operating data, and to adjust the control commands.

[0010] More preferably, the application service layer includes: a visualization dashboard, a log management module, and an analysis and alarm module; The visualization dashboard is based on digital twin technology to construct a virtual image of the physical device, thus obtaining a physical device model; The log management module is used to store the control commands; The control commands are input into the physical device model for virtual operation. The analysis and alarm module is used to assess the risks during the virtual operation process, issue alarms for the risk results, and send abnormal alarm information.

[0011] More preferably, in the real-time data acquisition module, each acquired environmental and status data is accompanied by a quality tag, and the quality tag is used for the generation of the data quality tag in the fusion quality assessment and repair module; The quality label includes: ; In the formula, Q Indicates the data quality score; Indicates the data acquisition delay; Indicates the maximum allowed delay; Indicates the weight of the event dimension; Indicates the currently collected value; This represents the standard deviation within the historical data window; Indicates the historical median; This represents the weight of the numerical dimension.

[0012] More preferably, the data quality label includes: ; in, ; In the formula, , , These represent the average quality of the source data, the temporal consistency score, and the cross-sensor consistency score, respectively. , , These represent the weighting coefficients for the average quality of the source data, the weighting coefficients for the temporal consistency score, and the weighting coefficients for the cross-sensor consistency score, respectively. Represents the median function; W Indicates the size of the history window; Indicates local standard deviation; This represents the estimated value after fusion.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention, by setting up a spatiotemporal alignment module, a semantic alignment and mapping module, and a multimodal data fusion module, enables unified processing of device data with different protocols, frequencies, and semantics, improving data consistency and availability. By introducing data quality tags and a fusion quality assessment mechanism, the collected and fused data are scored based on multi-dimensional indicators such as latency, standard deviation, and median. The system automatically selects output, smoothing, repair, or discard strategies based on the scoring results, ensuring the accuracy and reliability of basic operational data. The intelligent processing layer supports rule matching, workflow modeling, control command generation and issuance, and adaptively adjusts control commands based on state assessment and fault prediction results, achieving closed-loop optimized control of the system and improving its intelligence and responsiveness. Through an internal identifier generation module and a weighted identifier stability algorithm, devices with different protocols can be dynamically accessed while maintaining identifier stability, avoiding control logic interruptions due to device network changes and enhancing system robustness. The device reverse interface performs identity verification, legality verification, and device status verification on operational commands, employing idempotency design and retry mechanisms to ensure the security and reliability of command execution. The application service layer constructs a virtual image of physical devices based on digital twin technology, supports the virtual operation of control commands and risk pre-assessment, and improves the intuitiveness and proactivity of system operation and maintenance by combining a visual dashboard and alarm module. Attached Figure Description

[0014] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 This is a schematic diagram of the control system framework for multi-source data fusion and information system integration for devices according to an embodiment of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] Example 1: like Figure 1 As shown, this embodiment provides a device-oriented multi-source data fusion and information system integration control system, including: a data acquisition layer, a data transmission layer, a data fusion processing layer, an intelligent processing layer, a control execution layer, and an application service layer. The data acquisition layer is connected to the device and the data transmission layer to collect environmental and status data of the physical device in real time. The data transmission layer is also connected to the data fusion processing layer to transmit environmental and status data to the data fusion layer. The data fusion processing layer is also connected to the intelligent processing layer to preprocess the environmental and status data to obtain basic operating data. The intelligent processing layer is also connected to the control execution layer to perform status assessment, fault prediction, and generate control commands based on the basic operating data. The control execution layer is also connected to the data acquisition layer to convert control commands into operating action commands, which are then applied to the physical device. The application service layer is connected to the intelligent processing layer and the data acquisition layer to provide a visual monitoring interface and send abnormal alarm information to the user.

[0019] Further implementation involves a data acquisition layer comprising: an internal identifier generation module, a real-time data acquisition module, and a device reverse interface; the internal identifier generation module is used to access the physical device based on the protocols supported by the physical device and generate a unique internal identifier; the real-time data acquisition module is used to acquire the environmental and status data of the physical device based on the operating mode; and the device reverse interface is used to transmit the operating action commands to the physical device after performing security verification on the operating action commands.

[0020] Specifically, the internal identifier generation module first maintains a protocol adapter pool, dynamically selects an adapter for handshake based on the connection parameters reported by the physical device (such as protocol type, IP address, port, device ID, etc.), and calls the internal identifier generation algorithm after a successful handshake. ; In the formula, Indicates an internal identifier; The SHA-256 hash function outputs a 128-bit unique identifier; || represents string concatenation; Indicates the protocol type encoding; Indicates the vendor ID; Indicates the device serial number or MAC address; This represents the Unix timestamp at the time of access.

[0021] In addition, to prevent hash collisions, the internal identifier generation module also maintains an identifier mapping table. M : ;in, Indicates external identifier.

[0022] Meanwhile, in dynamic environments, physical devices may change network locations or protocol versions; therefore, the internal identifier generation module supports a weighted identifier stability algorithm. ; In the formula, Indicates stability score; , , These represent the protocol matching degree, MAC address matching degree, and device name matching degree, respectively. , , These represent the protocol matching weight coefficient, MAC address matching weight coefficient, and device name matching weight coefficient, respectively.

[0023] When the stability score exceeds the threshold T s When necessary, the original identifier can be reused to avoid workflow interruptions caused by identifier changes in upper-layer applications.

[0024] The real-time data acquisition module acquires environmental and status data from physical devices according to the set operating mode (such as periodic acquisition, event triggering, change threshold, mixed mode, etc.), including temperature, pressure, vibration, current, switch status, running time, etc.

[0025] The periodic acquisition method includes: setting a sampling interval. At any moment Read data; the collected values ​​are represented as: In the formula, Indicates the first n The time of the next sampling; Indicates the initial sampling time; n Indicates the sampling sequence number; Indicates in Data values ​​collected at all times; Indicates the read function; Indicates the target device identifier; Indicates the register / data address.

[0026] The change threshold acquisition method includes: only when the data change exceeds a set threshold. Uploaded in time: ;in, Indicates the amount of data change; This represents the sampled value at the time of the last upload; This indicates that an action is triggered.

[0027] The event-triggered data acquisition method involves the physical device actively pushing an interrupt or event signal and reading the associated data group. ;in, Indicates an event signal; Indicates a related data group; This indicates a batch read function; Represents a list of register addresses; Indicates the time when the event occurred.

[0028] In addition, a quality label is attached to each collected environmental and status data point: ; In the formula, Q Indicates the data quality score; Indicates the data acquisition delay; Indicates the maximum allowed delay; Indicates the weight of the event dimension; Indicates the currently collected value; This represents the standard deviation within the historical data window; Indicates the historical median; This represents the weight of the numerical dimension.

[0029] To enable collaborative sampling across multiple physical devices, this module also includes adaptive sampling rate adjustment based on network status, dynamically adjusting the sampling rate: ; ; In the formula, Indicates the new sampling frequency; Indicates the reference sampling frequency; Indicates the network load factor; Indicates the adjustment factor; Indicates current network traffic; This indicates network bandwidth.

[0030] The device's reverse interface receives operational commands (such as start / stop, parameter setting, mode switching, etc.) from the control execution layer. After rigorous security verification, it converts these commands into protocol action commands recognizable by the physical device and sends them to the target device. The security verification process includes: identity verification, command validity verification, and device status verification. Identity verification verifies whether the command originates from an authorized control execution layer module, based on JWT or a device certificate. Command validity verification is based on a set of valid commands and parameter range constraints; the verification function is as follows: ;in, Indicates an indicator function; Indicates the execution of the action command; Represents the set of legal instructions; Indicates command parameters; This refers to the valid value range of the parameter. Device status verification involves temporarily reading the current device status before the command is issued. Check if the prerequisites are met: ;in, This indicates a precondition function, such as a "stop command" which requires the device to be in a "running" state before it can be executed.

[0031] Run action instructions in a unified format that pass security verification. Convert to device raw protocol frame ;in, Indicates the opcode; Indicates the value of the instruction parameter; Indicates the original protocol frame; Represents the protocol encoder function; Indicates the register address; Represents a byte stream.

[0032] Meanwhile, the execution action command queue for each physical device adopts an idempotent design, with each execution action command carrying a unique transaction ID. If a physical device does not respond, a retry strategy is triggered. ;in, Indicates the first k Waiting time for each retry; k Indicates the sequence number of the retry attempt; Indicates the base retry time; This indicates the maximum number of retries. If the maximum number of retries is exceeded, an execution failure event will be reported and logged in the log management module.

[0033] The data fusion processing layer includes: a spatiotemporal alignment module, a semantic alignment and mapping module, a multi-source data distribution module, a multimodal data fusion module, and a fusion quality assessment and repair module. Environmental and state data are sequentially processed by the spatiotemporal alignment module and the semantic alignment and mapping module for spatiotemporal window alignment and semantic alignment, resulting in mapped serialized data. The multi-source data distribution module is connected to the multimodal data fusion module and is used to distribute the mapped serialized data to different multimodal data fusion modules based on internal identifiers. The multimodal data fusion module is also connected to the fusion quality assessment and repair module, which is used to perform data fusion on the mapped serialized data using different fusion modes to obtain fused data. The fusion quality assessment and repair module generates data quality labels for the fused data and provides repair instructions based on the results of the data instruction labels, resulting in basic operational data.

[0034] Specifically, the spatiotemporal alignment module adopts a timestamp-based interpolation alignment strategy and sets a uniform time window size. and alignment period For any data stream i At the moment of target alignment t Its alignment value The value is obtained by interpolation between adjacent original sampling points: ; In the formula, , , The timestamps are adjacent original sampling timestamps; , They represent Time and The original sampled value at time.

[0035] For discrete state variables, the nearest neighbor value is used: ; In the formula, Indicates the variable to be traversed; This represents the set of original sampling time points.

[0036] Define standard semantic space , where each standard semantic tag Includes: standard name, standard unit, value range, and data type. For those with accompanying raw semantic tags... Tag raw raw data points x The semantic mapping function is: ; in, ; In the formula, This represents the data points after semantic mapping; Represents a semantic mapping function; Represents standard semantic tags; Represents a normalized value; Indicates the original data offset; Indicates the scaling factor of the original data; , These represent the standard data offset and the standard data scaling factor, respectively.

[0037] After spatiotemporal alignment and semantic mapping, the mapped serialized data is obtained. ;in, Represents standardized numerical values.

[0038] The multi-source data distribution module distributes mapped serialized data to different multimodal data fusion modules based on internal identifiers, thereby achieving parallel processing and load balancing of data streams.

[0039] The mapping functions from physical devices to the multimodal data fusion module include: ; In the formula, Indicates the instance number of the multimodal data fusion module; Represents a consistent hash function; This indicates the number of instances of the multimodal data fusion module.

[0040] The distributed data blocks are: ; In the formula, d Represents a single data point; m This indicates the number of the multimodal data fusion module.

[0041] The multimodal data fusion module employs an adaptive weighted fusion method for fusing data of the same modality, dynamically allocating weights based on the historical accuracy of each data source. , ; in, ; In the formula, This represents the estimated value after fusion; n Indicates the number of data acquisition devices; Indicates the first i The weight of each data acquisition device; Indicates the first i Measurement values ​​from a data acquisition device; Indicates the first i Historical variance of each data acquisition device.

[0042] For heterogeneous data, feature-level concatenation is used. ,in, This represents the fused feature vector; This represents a vector concatenation function; These represent feature vectors of different modalities. They are then input into a fusion network (such as LSTM or GRU) to obtain fused data.

[0043] The methods for obtaining data quality labels for fused data by the fusion quality assessment and repair module include: ; in, ; In the formula, , , These represent the average quality of the source data (as indicated by the quality labels). Q The results include timing consistency score and cross-sensor consistency score. , , These represent the weighting coefficients for the average quality of the source data, the weighting coefficients for the temporal consistency score, and the weighting coefficients for the cross-sensor consistency score, respectively. Represents the median function; W Indicates the size of the history window; This represents the local standard deviation.

[0044] when When the value is ≥0.9, the fused data is directly output as the basic operating data; when 0.7≤ When the value is less than 0.9, the fused data is lightly smoothed, and the output is the base running data; when 0.5 ≤ When the value is less than 0.7, the fused data is interpolated and repaired before being output as the base running data; when... When the value is less than 0.5, the fused data is discarded and marked as missing.

[0045] The intelligent processing layer includes: a rule processing module, a workflow maintenance module, a control command generation and distribution module, and a closed-loop control and adaptive optimization module. The rule processing module is used to continuously query and match patterns on basic operating data based on predefined rule sets to obtain a list of actions to be executed. The workflow maintenance module is used to model the list of actions to be executed as a directed graph. The control command generation and distribution module is used to generate control commands based on the directed graph and distribute the control commands to the control execution layer. The closed-loop control and adaptive optimization module is used to perform state evaluation and fault prediction based on basic operating data, and adjust the control commands based on the state evaluation and fault prediction results.

[0046] First, define the rule set. Each of the rules Represented as: ;in, Representation rules j The conditional expression; Representation rules j Action list; rule processing module in time window Internal moment t Arrival of basic operational data stream Perform continuous queries: ;in, This indicates a rule matching indicator function; a value of 1 indicates a rule. r j At any moment t 0 indicates that the event has been triggered.

[0047] Summarize all actions triggered by the rules to obtain a list of actions to be executed. : .

[0048] Construct a directed workflow graph based on the list of actions to be performed. ,in, V Represents a set of vertices, each v This indicates an action to be performed; E Represents the set of edges, which are the actions to be performed. v i With the action to be performed v j Directed edges are defined. Simultaneously, a dependency matrix is ​​constructed, where each element represents an action to be executed. v i With the action to be performed v j Dependency relationship, a value of 1 indicates an action to be performed. v i Actions to be performed v j Execute before the specified condition, otherwise return 0. Based on the directed workflow graph and dependency matrix, obtain the topological sorting result of the directed workflow graph: ;in, This represents a topological sorting function, which can be a sorting algorithm based on Kahn's algorithm or depth-first search.

[0049] Based on the topological sorting result of the directed workflow graph, specific control instructions are generated and issued to the control execution layer. The control instruction generation function is as follows: ;in, This indicates an instruction generation function, used to convert the action to be executed into a standard instruction format by combining context information; This indicates the context information. Afterwards, control commands are sequentially distributed to the control execution layer.

[0050] In this embodiment, the state is evaluated based on the device state evaluation function: ; In the formula, Represents physical devices o At any moment t ; health status; The state evaluation function can be represented by an evaluation algorithm based on thresholds, statistical models, or machine learning models. Represents physical devices o The basic operational data.

[0051] Fault prediction includes: ; In the formula, The fault prediction function can be represented by regression models, time series models (LSTM, Transformer), etc. Indicates model parameters.

[0052] After adjusting the control commands, the model parameters are updated using closed-loop feedback: ;in, Indicates the learning rate; Represents the gradient operator; Indicates the actual implementation effect; Indicates the expected goal.

[0053] The application service layer includes: a visualization dashboard, a log management module, and an analysis and alarm module. The visualization dashboard uses digital twin technology to build a virtual image of the physical device, resulting in a physical device model. The log management module stores control commands. The control commands are input into the physical device model for virtual operation. The analysis and alarm module assesses the risks during the virtual operation process, issues alarms for risk results, and sends abnormal alarm information.

[0054] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A device-oriented multi-source data fusion and information system integration control system, characterized in that, include: The data acquisition layer, data transmission layer, data fusion and processing layer, intelligent processing layer, control and execution layer, and application service layer are all included. The data acquisition layer is connected to the device and the data transmission layer, and is used to collect environmental and status data of the physical device in real time. The data transmission layer is also connected to the data fusion processing layer, and is used to transmit the environment and state data to the data fusion layer; The data fusion processing layer is also connected to the intelligent processing layer, and is used to preprocess the environmental and state data to obtain basic operating data; The intelligent processing layer is also connected to the control execution layer, and is used to perform status assessment, fault prediction and generate control commands based on the basic operating data. The control execution layer is also connected to the data acquisition layer and is used to convert the control instructions into operation action instructions, which are applied to the physical device. The application service layer is connected to the intelligent processing layer and the data acquisition layer to provide a visual monitoring interface and send abnormal alarm information to the user.

2. The device-oriented multi-source data fusion and information system integration control system according to claim 1, characterized in that, The data acquisition layer includes: an internal identifier generation module, a real-time data acquisition module, and a device reverse interface; The internal identifier generation module is used to access the physical device based on the protocol supported by the physical device and generate a unique internal identifier. The real-time data acquisition module is used to acquire the environmental and status data of the physical device based on the operating mode. The device reverse interface is used to perform security verification on the operation command and then transmit the operation command to the physical device.

3. The device-oriented multi-source data fusion and information system integration control system according to claim 2, characterized in that, The data fusion processing layer includes: a spatiotemporal alignment module, a semantic alignment and mapping module, a multi-source data distribution module, a multimodal data fusion module, and a fusion quality assessment and repair module; The environment and state data are sequentially processed by the spatiotemporal alignment module and the semantic alignment and mapping module for spatiotemporal window alignment and semantic alignment to obtain mapped serialized data. The multi-source data distribution module is connected to the multimodal data fusion module and is used to distribute the mapped serialized data to different multimodal data fusion modules based on the internal identifier; The multimodal data fusion module is also connected to the fusion quality assessment and repair module, and is used to perform data fusion on the mapped serialized data using different fusion modes to obtain fused data; The fusion quality assessment and repair module is used to generate data quality tags for the fusion data, give repair instructions based on the results of the data instruction tags, and obtain the basic operating data.

4. The device-oriented multi-source data fusion and information system integration control system according to claim 1, characterized in that, The intelligent processing layer includes: a rule processing module, a workflow maintenance module, a control command generation and issuance module, and a closed-loop control and adaptive optimization module; The rule processing module is used to continuously query and match patterns on the basic operational data based on a predefined rule set to obtain a list of actions to be executed. The workflow maintenance module is used to model the list of actions to be executed as a directed graph; The control command generation and distribution module is used to generate control commands based on the directed graph and distribute the control commands to the control execution layer; The closed-loop control and adaptive optimization module is used to perform status assessment and fault prediction based on the basic operating data, and to adjust the control commands.

5. A device-oriented multi-source data fusion and information system integration control system according to claim 1, characterized in that, The application service layer includes: a visual dashboard, a log management module, and an analysis and alerting module; The visualization dashboard is based on digital twin technology to construct a virtual image of the physical device, thus obtaining a physical device model; The log management module is used to store the control commands; The control commands are input into the physical device model for virtual operation. The analysis and alarm module is used to assess the risks during the virtual operation process, issue alarms for the risk results, and send abnormal alarm information.

6. A device-oriented multi-source data fusion and information system integration control system according to claim 3, characterized in that, In the real-time data acquisition module, each acquired environmental and state data is accompanied by a quality tag, which is used for the generation of data quality tags in the fusion quality assessment and repair module. The quality label includes: ； In the formula, Q Indicates the data quality score; Indicates the data acquisition delay; Indicates the maximum allowed delay; Indicates the weight of the event dimension; Indicates the currently collected value; This represents the standard deviation within the historical data window; Indicates the historical median; This represents the weight of the numerical dimension.

7. A device-oriented multi-source data fusion and information system integration control system according to claim 6, characterized in that, The data quality labels include: ； in, ； In the formula, , , These represent the average quality of the source data, the temporal consistency score, and the cross-sensor consistency score, respectively. , , These represent the weighting coefficients for the average quality of the source data, the weighting coefficients for the temporal consistency score, and the weighting coefficients for the cross-sensor consistency score, respectively. Represents the median function; W Indicates the size of the history window; Indicates local standard deviation; This represents the estimated value after fusion.

Images