Situation capability comprehensive evaluation system

Abstract

The invention discloses a situation capability comprehensive evaluation system, which comprises a multi-source situation awareness module, a causal diagram construction module, a multi-dimensional index coupling calculation module, a self-adaptive fusion module, a causal traceability feedback module and a situation deduction sand table module, the method comprises the following steps: acquiring sensor data, an artificial situation report and environment characteristic data through distributed edge nodes, and generating an original situation tuple marked by time and space; performing event causal relationship mining on the original situation tuple to generate a time evolution causal graph with a weight; analyzing a comprehensive situation index generation path, extracting related influence factors and conduction paths thereof, and generating a traceability map; and fusing the traceability map with the comprehensive situation index. According to the method, the problems of single data source and static model are effectively solved; and through fusion and dynamic deduction of multi-dimensional data, comprehensive analysis and evaluation of a complex situation can be realized, and the accuracy and timeliness of an evaluation result are ensured.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of situation awareness and assessment, and particularly relates to a situation capability comprehensive assessment system. BACKGROUND

[0002] With the rapid development of technology and the continuous expansion of application scenarios, the complex situation faced in modern society is increasing, especially in the fields of geological disaster warning, environmental monitoring, public safety, etc. Various sensors and monitoring devices continuously collect a large amount of spatio-temporal data, forming multi-source situation awareness data. However, due to the heterogeneity, timeliness and complexity of the data, these data are often difficult to directly use for accurate situation assessment and risk warning. The current technology mainly relies on a single data source or static model for assessment, lacks comprehensive data integration and dynamic updating mechanism, and is difficult to cope with rapidly changing complex situations.

[0003] In the prior art, although the multi-source data collection technology and data fusion method have developed to a certain extent, the problems are still very prominent. First, the traditional situation assessment method often uses a single or local data source, which fails to effectively integrate data in multiple dimensions, resulting in inaccurate or lagging assessment results. Second, the existing causal analysis and deduction methods are mostly static modeling, which cannot dynamically adjust and optimize the assessment results according to time changes, resulting in the inability to timely predict and reflect the changes in the situation in rapidly changing complex situations. SUMMARY

[0004] The purpose of the present application is to provide a situation capability comprehensive assessment system to solve the problem of single data source and static model.

[0005] Technical solution: The situation capability comprehensive assessment system according to the present application comprises a multi-source situation awareness module, a causal graph construction module, a multi-dimensional index coupling calculation module, an adaptive fusion module, a causal tracing feedback module and a situation deduction sand table module. Among them:

[0006] Multi-source situation awareness module: Collect sensor data, artificial situation reports and environmental feature data through distributed edge nodes to generate spatio-temporal labeled original situation tuples;

[0007] Causal graph construction module: used for event causal relationship mining on original situation tuples to generate a weighted time evolution causal graph;

[0008] Multi-dimensional index coupling calculation module: based on the time evolution causal graph, synchronously calculate the threat transmission coefficient, resource chain effect index and situation entropy value;

[0009] Adaptive fusion module: used for inputting the calculation results of the multi-dimensional index coupling calculation module into the fusion network with a gating mechanism according to the current situation stage characteristics, to generate a comprehensive situation index;

[0010] Causal tracing feedback module: used for analyzing the generation path of the comprehensive situation index, extracting relevant influence factors and their transmission paths, and generating a tracing atlas;

[0011] Situation deduction sand table module: used for fusing the tracing atlas and the comprehensive situation index, and generating a dynamic sand table containing multiple pre-plan evolution branches on a three-dimensional geographic base map.

[0012] Further, the multi-source situation awareness module includes a distributed data acquisition unit, a space-time labeling unit, a confidence evaluation unit, a tuple packaging unit and an edge cache unit; wherein:

[0013] Distributed data acquisition unit: deployed at the physical layer of the edge node, including the following subunits:

[0014] Sensor interface subunit: adapted to at least three kinds of heterogeneous sensors through an industrial bus protocol, used for real-time acquisition of voltage, image and motion vector data, and recording of original sampling frequency;

[0015] Artificial report analysis subunit: configured with a natural language processing engine, used for entity extraction on input text or voice situation reports, to generate a structured event description table including event type, occurrence location and reporter identity code;

[0016] Environment feature capture subunit: based on the LoRaWAN protocol to connect a meteorological station or a geological monitor, used for collecting environmental parameters including temperature and humidity, vibration amplitude and electromagnetic interference intensity;

[0017] Space-time labeling unit: used for receiving data output by the distributed data acquisition unit, and unifying the time stamps of each data to UTC standard time with an accuracy of ±1 millisecond; at the same time, according to the Beidou / GPS dual-mode positioning chip built-in the data source device, the geographic location is mapped to a three-dimensional geographic coordinate system;

[0018] Confidence evaluation unit: used for calculating confidence labels for the labeled data, including sensor confidence, artificial report confidence and environmental data confidence, respectively denoted as sensor confidence , artificial report confidence , and environmental data confidence ;

[0019] Tuple packaging unit: used for packing data with space-time labels and confidence labels into original situation tuples according to a preset format;

[0020] Edge cache unit: used for establishing a ring buffer locally at the distributed node, and temporarily storing original situation tuples in time-space marker order.

[0021] Further, the causal graph construction module includes an event causal relationship mining unit, a causal weight calculation unit, and a time evolution unit; wherein:

[0022] Event causal relationship mining unit: used for mining event causal relationships of original situation tuples; according to the time stamp, spatial coordinates and confidence label in the original data, the order and relevance of event occurrence are identified, and a Bayesian network-based inference method is used to calculate the conditional dependence relationship between events, and then a graph structure with causal relationships is generated;

[0023] Causal weight calculation unit: used for assigning a weight to each edge in the event causal relationship graph, and the calculation formula of the weight is: , wherein, is the causal relationship weight between event and event , is the conditional probability from event to event , and is the confidence of event and event , is a weighting coefficient;

[0024] Time evolution unit: used for dynamically adjusting the correlation between events in the causal graph according to the time sequence of event occurrence, and generating a causal graph with time evolution.

[0025] Further, the event causal relationship mining unit includes:

[0026] Data preprocessing: standardizing the sensor data, manual reports and environmental feature data to ensure data consistency;

[0027] Event identification and ordering: by analyzing the time-space marker data, relevant events are identified and ordered according to time sequence;

[0028] Inference of causal relationship between events: on the basis of event ordering, Bayesian network algorithm is used to infer the causal relationship between events. Specifically, for each pair of events , the conditional probability is calculated, which represents the probability of event occurring after event occurs ; if the conditional probability exceeds a preset threshold , it is considered that event Regarding the event It has causal influence, that is, it establishes edges ;

[0029] Constructing a causal graph structure: Based on the established edges, construct a causal graph structure according to the causal influence between events; specifically, each event... Represented as a node in a graph, an event and causal relationship between If an edge is represented as a directed edge in a graph, then the weight of the edge is... .

[0030] Furthermore, the time evolution unit includes:

[0031] Event time series sorting: Sort the timestamps of events to obtain an event sequence. Each event Corresponding to a timestamp And satisfy: ;

[0032] Dynamic updating of temporal relationships: dynamically adjusting the strength of causal relationships between events in the cause-effect graph based on the temporal order of events;

[0033] Time adjustment of causal relationship weights: Based on the time difference between events and the dynamically adjusted influence, the weights between events in the causal graph are updated. The updated weights are calculated using the following formula: ,in, For the original weights, The causal weights are adjusted for time. The coefficient for the attenuation of influence due to time difference;

[0034] Generate a causal graph with time evolution: Based on the adjusted causal weights above, reconstruct a causal graph with time evolution.

[0035] Furthermore, the multi-dimensional index coupling calculation module includes a threat transmission coefficient calculation unit, a resource chain effect index calculation unit, and a situation entropy value calculation unit; wherein:

[0036] Threat Transmission Coefficient Calculation Unit: Used to calculate the degree of threat transmission between nodes in the system, i.e., the threat transmission coefficient, based on the event relationships in the time evolution causal graph. The formula is: ,in, For the threat from the incident Transmission to events The coefficients, the denominator is the event. To all related events The sum of weights;

[0037] Resource chain effect index calculation unit: for calculating the mutual dependence and chain effect index of resources between events, the formula is: , wherein, is the resource chain effect index, indicating the effect of event under the influence of event ; and is the resource weight of event and ;

[0038] Situation entropy value calculation unit: for calculating the situation entropy value of the system, indicating the uncertainty of the system state, first, based on all event nodes in the time evolution causal diagram and the causal relationship weight between events , the state probability of each event is calculated ; then the situation entropy value is calculated according to the state probability of all events, the formula is: , wherein, is the situation entropy value, indicating the uncertainty of the event state in the system.

[0039] Further, the adaptive fusion module includes a current situation phase identification unit, a multi-dimensional index input unit, a gating mechanism unit, and a comprehensive situation index calculation unit; wherein:

[0040] Current situation phase identification unit: for identifying and extracting the current situation phase characteristics according to the current situation state of the system, including normal, alert, early warning and emergency;

[0041] Multi-dimensional index input unit: for receiving the calculation results from the multi-dimensional index coupling calculation module, including threat transmission coefficient, resource chain effect index and situation entropy value, and standardizing them to convert them to a unified scale, respectively denoted as threat transmission coefficient , resource chain effect index and situation entropy value ;

[0042] Gating mechanism unit: for adjusting the fusion degree of multi-dimensional indexes according to the current situation phase characteristics, and dynamically adjusting the fusion proportion of each index through the gating mechanism, specifically, according to the current situation phase characteristics, using the gating network to calculate the fusion coefficient ;

[0043] Comprehensive situation index calculation unit: for calculating the final comprehensive situation index by weighted sum of each weighted multi-dimensional index according to the output of the gating mechanism .

[0044] Further, the causal tracing feedback module comprises a comprehensive situation index analysis unit, an influence factor extraction unit, a conduction path tracking unit, and a tracing map construction unit; wherein:

[0045] The comprehensive situation index analysis unit is configured to analyze the comprehensive situation index, and identify relevant events in the generation path of the comprehensive situation index and the correlation therebetween;

[0046] The influence factor extraction unit is configured to extract relevant influence factors from the generation path of the comprehensive situation index, and determine the contribution of the influence factors to the situation assessment result;

[0047] The conduction path tracking unit is configured to track the conduction path of the influence factors, and identify the transmission chain of the influence factors in the causal graph; according to the event causal relationship mining result, the conduction path of the influence factors in the causal graph is reversely tracked, and the influence source is gradually determined;

[0048] The tracing map construction unit is configured to construct a tracing map according to the extraction result of the conduction path and the influence factors.

[0049] Further, the situation deduction sand table module comprises a tracing map fusion unit, a dynamic sand table generation unit, and an evolution path display unit; wherein:

[0050] The tracing map fusion unit is configured to fuse the tracing map generated by the causal tracing feedback module and the comprehensive situation index calculated by the adaptive fusion module; specifically, each event node in the tracing map is aligned with the relevant event in the comprehensive situation index; then, by using a weighted average method, the influence of each event in the tracing map on the situation is combined with the calculation result of the comprehensive situation index, to obtain fused situation data ;

[0051] The dynamic sand table generation unit is configured to generate a dynamic sand table comprising multiple pre-plan evolution branches on a three-dimensional geographic base map according to the fused situation data .

[0052] The evolution path display unit is configured to display the multiple pre-plan evolution branches in a dynamic manner on the three-dimensional sand table, to intuitively display the situation evolution under different scenarios.

[0053] Further, the dynamic sand table generation unit comprises:

[0054] Event space position mapping: mapping the space position of each event to the three-dimensional geographic map, and mapping the space coordinates of each event to the three-dimensional geographic map; ;

[0055] Evolution path initialization: initializing the evolution path based on the fused situation data This generates multiple evolutionary branches, each representing a contingency plan evolution path under different scenarios.

[0056] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: By introducing a combination of multi-source situational awareness, causal graph construction and time evolution analysis, the present invention effectively solves the problems of single data source and static model; through the fusion of multi-dimensional data and dynamic inference, it can achieve comprehensive analysis and evaluation of complex situations, ensuring the accuracy and timeliness of the evaluation results; especially in rapidly changing environments, the present invention can adjust the situational assessment in real time, providing decision-makers with dynamic analysis based on causal relationships. Attached Figure Description

[0057] Figure 1 This is a schematic diagram of a comprehensive situational capability assessment system.

[0058] Figure 2 A schematic diagram of the module for constructing a cause-effect graph. Detailed Implementation

[0059] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0060] like Figure 1 As shown, the situational awareness comprehensive assessment system of the present invention includes a multi-source situational awareness module, a causal graph construction module, a multi-dimensional index coupling calculation module, an adaptive fusion module, a causal source tracing feedback module, and a situational simulation sand table module; wherein:

[0061] Multi-source situational awareness module: Collects sensor data, manual situational reports and environmental feature data through distributed edge nodes to generate spatiotemporally labeled raw situational tuples;

[0062] Causal graph construction module: used to mine the causal relationships of events from the original situation tuples and generate a weighted time evolution causal graph;

[0063] Multidimensional index coupling calculation module: Based on the time evolution causal graph, it simultaneously calculates the threat transmission coefficient, resource chain effect index and situation entropy value;

[0064] Adaptive fusion module: Based on the characteristics of the current situation stage, it inputs the calculation results of the multi-dimensional index coupling calculation module into a fusion network with a gating mechanism to generate a comprehensive situation index.

[0065] Causal source tracing feedback module: used to analyze the generation path of the comprehensive situation index, extract relevant influencing factors and their transmission paths, and generate a source tracing map;

[0066] Situation simulation sand table module: used to integrate source map and comprehensive situation index, and generate a dynamic sand table containing multiple contingency plan evolution branches on a three-dimensional geographic base map.

[0067] The multi-source situation awareness module comprises a distributed data acquisition unit, a space-time marking unit, a confidence evaluation unit, a tuple packaging unit and an edge cache unit; wherein:

[0068] The distributed data acquisition unit is deployed at the physical layer of the edge node and comprises the following sub-units:

[0069] The sensor interface sub-unit is adapted to at least three kinds of heterogeneous sensors through an industrial bus protocol, is used for acquiring voltage, image and motion vector data in real time, and records the original sampling frequency;

[0070] The artificial report analysis sub-unit is configured with a natural language processing engine, is used for entity extraction on input text or voice situation reports, and generates a structured event description table including event type, occurrence location and reporter identity code;

[0071] The environmental feature capture sub-unit is connected with a meteorological station or a geological monitor based on the LoRaWAN protocol, is used for collecting environmental parameters including temperature and humidity, vibration amplitude and electromagnetic interference intensity;

[0072] The space-time marking unit is used for receiving data output by the distributed data acquisition unit, and unifying the time stamps of the data to the UTC standard time with an accuracy of ±1 millisecond; meanwhile, according to the Beidou / GPS dual-mode positioning chip built in the data source device, the geographic location is mapped to a three-dimensional geographic coordinate system (longitude X, latitude Y, altitude Z);

[0073] The confidence evaluation unit is used for calculating the confidence labels of the marked data, including sensor confidence, artificial report confidence and environmental data confidence, which are respectively denoted as sensor confidence , artificial report confidence and environmental data confidence ; specifically,

[0074] The sensor confidence is linearly weighted according to signal strength, historical failure rate and current environmental interference degree, and the calculation formula is: wherein, is the normalized signal strength, is the device failure probability, is the environmental interference coefficient, is the corresponding weighting coefficient;

[0075] The artificial report confidence is calculated based on the reporting personnel's authority level and the report logic integrity , and the calculation formula is: ;

[0076] The environmental data confidence is calculated according to the precision level direct assignment A;

[0077] Tupel packaging unit: for packing data with spatiotemporal markers and confidence labels into original situation tuples in a preset format, the specific format being: header field, timestamp, spatial coordinates, data payload, and confidence label;

[0078] Edge cache unit: for locally establishing a ring buffer at the distributed node, and temporarily storing original situation tuples in spatiotemporal marker order, which ensures the continuity of evaluation when data transmission is interrupted.

[0079] As Figure 2 shown, the causal graph construction module includes an event causal relationship mining unit, a causal weight calculation unit, and a time evolution unit; wherein:

[0080] Event causal relationship mining unit: for mining event causal relationships of original situation tuples; according to the timestamp, spatial coordinates, and confidence label in the original data, the order and relevance of event occurrence are identified, and a reasoning method based on Bayesian network is used to calculate the conditional dependence relationship between events, and then a graph structure with causal relationships is generated;

[0081] Causal weight calculation unit: for assigning weights to each edge (i.e., causal relationship) in the event causal relationship graph, the specific operation being: according to the confidence of each event, the spatiotemporal relevance between events, the reliability of historical data, and the current environmental factors, a weighted average method is used to generate the weight of the edge in combination with the conditional probability calculated in the Bayesian network, and the calculation formula of the weight is: wherein, is the weight of the causal relationship between event and event , is the conditional probability from event to event , and are the confidences of event and event , and is the weighted coefficient;

[0082] Time evolution unit: for dynamically adjusting the correlation between events in the causal graph according to the time sequence of event occurrence, and generating a causal graph with time evolution; the above unit can accurately mine the causal relationships between events and assign appropriate weights to each relationship by comprehensively considering the spatiotemporal markers, confidence labels, and time evolution relationships between events of the original situation tuples, thereby providing an accurate basis for subsequent situation assessment and deduction.

[0083] The event causal relationship mining unit includes:

[0084] Data preprocessing: Standardize sensor data, manual reports, and environmental characteristic data to ensure data consistency;

[0085] Event identification and sorting: By analyzing spatiotemporal stamped data, relevant events are identified and sorted in chronological order. Event sorting is achieved by comparing the timestamps of the events.

[0086] Reasoning for causal relationships between events: Based on event ranking, Bayesian network algorithms are used to reason about the causal relationships between events. Specifically, for each pair of events... Calculate its conditional probability , indicating an event Post-event probability of occurrence The formula for calculating conditional probability is: ,in, For the event and events The joint probability of simultaneous occurrence For the event The marginal probability of occurrence; if the conditional probability Exceeding the preset threshold Then the event is considered Regarding the event It has causal influence, that is, it establishes edges ;

[0087] Constructing a causal graph structure: Based on the established edges, construct a causal graph structure according to the causal influence between events; specifically, each event... Represented as a node in a graph, an event and causal relationship between If an edge is represented as a directed edge in a graph, then the weight of the edge is... Through the above steps, the event causal relationship mining unit can extract the causal relationships between events from the original situation tuple and generate a weighted time evolution causal graph, providing a foundation for subsequent situation analysis and inference.

[0088] The time evolution unit includes:

[0089] Event time series sorting: Sort the timestamps of events to obtain an event sequence. Each event Corresponding to a timestamp And satisfy: The purpose of event sequencing is to ensure the temporality of causal relationships, with earlier events influencing later events;

[0090] Temporal relationship dynamic update: dynamically adjust the causal relationship strength between events in the causal diagram based on the time sequence of events, specifically according to the time difference of event occurrence to calculate the influence of event on event , the influence is dynamically adjusted by a decay function: , where is the decay coefficient, which determines the decay speed of the time delay on the influence, is the time difference between event and event ; as time goes on, the influence of earlier events on subsequent events gradually decreases;

[0091] Temporal adjustment of causal relationship weight: based on the time difference of events and the dynamically adjusted influence, update the weight between events in the causal diagram; for each pair of events and , the causal relationship weight is adjusted over time, and the updated weight is calculated by the following formula: , where is the original weight, is the causal relationship weight after time adjustment, is the influence decay coefficient caused by the time difference; in this way, the weight gradually decreases over time, reflecting the changes in causal relationship over time;

[0092] Generate a causal diagram with time evolution: according to the adjusted causal relationship weight above, reconstruct a causal diagram with time evolution; through the above steps, the time evolution unit can dynamically adjust the relationship between events in the causal diagram according to the time sequence of events, generate a causal diagram with time evolution, and accurately reflect the time sequence changes and causal influence of events, providing strong support for subsequent situation deduction and decision-making.

[0093] The multi-dimensional index coupling calculation module includes a threat transmission coefficient calculation unit, a resource chain effect index calculation unit, and a situation entropy value calculation unit; wherein:

[0094] Threat transmission coefficient calculation unit: used to calculate the transmission degree of threats between nodes in the system according to the event relationship in the time evolution causal diagram, i.e. the threat transmission coefficient, the formula is: , where is the coefficient of threat transmission from event to event , the denominator is the event to all associated events the sum of the weights of the events, ensuring the conduction coefficient normalization, so that the sum of the conduction coefficients is 1; the coefficient reflects the degree of influence of a certain event on other events;

[0095] Resource chain effect index calculation unit: for calculating the mutual dependence and chain effect index of resources between events, the formula is: , wherein, is the resource chain effect index, indicating the influence of event on the resources of event ; and is the resource weight of event and , reflecting the resource influence of each event, and the index reflects the resource influence of an event on other events on the resource chain;

[0096] Situation entropy value calculation unit: for calculating the situation entropy value of the system, indicating the uncertainty of the system state. First, based on all event nodes in the time evolution causal graph and the causal relationship weight between events, the state probability of each event is calculated, the formula is: , wherein, is the state probability of event , indicating the importance of event in the causal graph; then the situation entropy value is calculated according to the state probability of all events, the formula is: , wherein, is the situation entropy value, indicating the uncertainty of the event state in the system; a higher situation entropy value indicates a higher uncertainty of the system state, and a lower situation entropy value indicates a more certain system state; through the above calculation, the threat conduction coefficient, resource chain effect index and situation entropy value can be accurately calculated according to the time evolution causal graph, providing an important basis for the comprehensive situation assessment of the system.

[0097] The adaptive fusion module includes a current situation phase identification unit, a multi-dimensional index input unit, a gating mechanism unit and a comprehensive situation index calculation unit; wherein:

[0098] Current situation phase identification unit: for identifying and extracting the current situation phase characteristics according to the current situation state of the system, including normal, alert, early warning and emergency;

[0099] Multi-dimensional index input unit: for receiving the calculation results from the multi-dimensional index coupling calculation module, including the threat conduction coefficient, the resource chain effect index and the situation entropy value, and standardizing them to convert them into a unified scale, respectively denoted as threat conduction coefficient , resource chain effect index and situation entropy value ;

[0100] Gating mechanism unit: used to adjust the fusion degree of multi-dimensional indicators according to the current situation stage characteristics, and dynamically adjust the fusion proportion of each indicator through the gating mechanism, specifically, according to the current situation stage characteristics, the fusion coefficient is calculated using the gating network , the formula is: , wherein, current situation stage characteristics, is the fusion coefficient, indicating the fusion weight of each indicator in the current situation stage; represents the Sigmoid activation function, which limits the fusion coefficient between 0 and 1, ensuring that the contribution of each indicator in the comprehensive situation index is adaptively adjusted according to the situation stage;

[0101] Comprehensive situation index calculation unit: used to weight and sum each weighted multi-dimensional indicator according to the output of the gating mechanism to generate the final comprehensive situation index ; the specific formula is: , wherein, is the comprehensive situation index, is the fusion coefficient, indicating the contribution of each indicator in the current situation stage; through the above processing, the comprehensive situation index can dynamically adjust the weight of each multi-dimensional indicator according to the current situation stage, thereby generating a comprehensive evaluation result that conforms to the actual situation.

[0102] The causal tracing feedback module includes a comprehensive situation index analysis unit, an influence factor extraction unit, a conduction path tracking unit, and a tracing atlas construction unit; wherein:

[0103] Comprehensive situation index analysis unit: used to analyze the comprehensive situation index to identify relevant events in its generation path and their associated relationships; specifically, based on the weighted time evolution causal graph generated by the causal graph construction module, the event nodes related to the calculation of the comprehensive situation index are extracted; and the causal graph is analyzed using a graph traversal algorithm (such as depth-first search DFS or breadth-first search BFS) to determine the relevant events that affect the comprehensive situation index; then by tracking the time evolution order of these event nodes, the time order and causal relationship in its generation path are identified;

[0104] Influence factor extraction unit: used to extract relevant influence factors from the generation path of the comprehensive situation index and determine their contribution to the situation assessment result, specifically, from the key events obtained by analysis, the corresponding multi-dimensional indicators (such as threat conduction coefficient, resource chain effect index, situation entropy value, etc.) are identified, and the contribution of each indicator to the comprehensive situation index is calculated; then using a weighted calculation method, the influence of each indicator is quantified as an influence factor , the formula is: wherein, is the weight of the th event on the comprehensive situation index, is the multi-dimensional index value corresponding to the event, is the total number of impact factors;

[0105] the conduction path tracking unit is used to track the conduction path of the impact factor and identify the transmission chain of the impact factor in the causal graph; according to the event causal relationship mining result, the conduction path of the impact factor in the causal graph is tracked in reverse, and the impact source is gradually determined; the cumulative weight of the path is calculated , reflecting the importance of each path in the conduction process of the impact factor, and the formula is: wherein, is the weight of the th event node in the path, is the number of event nodes in the path;

[0106] the traceability graph construction unit is used to construct a traceability graph according to the extraction result of the conduction path and the impact factor; specifically, according to the conduction path of the impact factor, the nodes and edges in the graph are constructed, wherein the nodes represent the impact factors or key events, and the edges represent the transmission direction and strength of the causal relationship; the finally generated traceability graph contains each impact factor and its mutual relationship, and shows the entire influence process from the source event to the comprehensive situation index; by tracking the conduction chain of each impact factor, the system can clearly reveal the source event of the situation assessment and its transmission process, and provide more transparent and traceable basis for subsequent decision-making and emergency response.

[0107] The situation deduction sand table module includes a traceability graph fusion unit, a dynamic sand table generation unit, and an evolution path display unit; wherein:

[0108] the traceability graph fusion unit is used to fuse the traceability graph generated by the causal traceability feedback module and the comprehensive situation index calculated by the adaptive fusion module, to ensure the correlation and consistency of the two, and specifically, each event node in the traceability graph is aligned with the related events in the comprehensive situation index, to ensure that the information of each event in the traceability graph and the comprehensive situation index is consistent; then, by using the weighted average method, the influence of each event in the traceability graph on the situation is combined with the calculation result of the comprehensive situation index, to obtain the fused situation data , and the fusion technology formula is: wherein, is the fused situation data, is the weight of the th event, is the total number of events, is the impact factor of the th event in the traceability graph, for events contribution to the comprehensive situation index; the fusion result reflects the comprehensive influence of the traceability map and the comprehensive situation index;

[0109] Dynamic sand table generation unit: for generating a dynamic sand table containing multiple pre-plan evolution branches on a three-dimensional geographic base map according to the fused situation data , a dynamic sand table containing multiple pre-plan evolution branches is generated on a three-dimensional geographic base map;

[0110] Evolution path display unit: for displaying multiple pre-plan evolution branches on a three-dimensional sand table in a dynamic manner, intuitively displaying the evolution of the situation under different scenarios; specifically, according to the priority, weight and current situation of each evolution path, the display level of each branch in the sand table is dynamically adjusted to ensure that the key path is displayed first; at the same time, color coding, transparency adjustment and other visual means are used to display the relative importance and possibility of each evolution path; for high-probability or high-impact evolution paths, more obvious markers are used for highlighting; through the above steps, the situation deduction sand table module can generate a dynamic sand table containing multiple pre-plan evolution branches on a three-dimensional geographic base map, reflecting the situation changes in real time and providing multi-angle and dynamic evolution prediction for decision-making.

[0111] The dynamic sand table generation unit includes:

[0112] Event space position mapping: mapping the spatial position of each event on a three-dimensional geographic map, and mapping the spatial coordinates of each event to a three-dimensional geographic map;

[0113] Evolution path initialization: based on the fused situation data , multiple evolution branches are generated, each branch representing a pre-plan evolution path under different scenarios; each path is dynamically adjusted according to different situation changes, time evolution and influence factor transmission process; through the above steps, the dynamic sand table generation unit can generate a dynamic sand table containing multiple pre-plan evolution branches in real time, displaying the evolution process of the situation under different scenarios.

Claims

1. A comprehensive situational capability assessment system, characterized in that, It includes a multi-source situational awareness module, a cause-effect graph construction module, a multi-dimensional indicator coupling calculation module, an adaptive fusion module, a cause-effect tracing feedback module, and a situation simulation sandbox module; among which: Multi-source situational awareness module: Collects sensor data, manual situational reports and environmental feature data through distributed edge nodes to generate spatiotemporally labeled raw situational tuples; Causal graph construction module: used to mine the causal relationships of events from the original situation tuples and generate a weighted time evolution causal graph; Multidimensional index coupling calculation module: Based on the time evolution causal graph, it simultaneously calculates the threat transmission coefficient, resource chain effect index and situation entropy value; Adaptive fusion module: Based on the characteristics of the current situation stage, it inputs the calculation results of the multi-dimensional index coupling calculation module into a fusion network with a gating mechanism to generate a comprehensive situation index. Causal source tracing feedback module: used to analyze the generation path of the comprehensive situation index, extract relevant influencing factors and their transmission paths, and generate a source tracing map; Situation simulation sand table module: used to integrate source map and comprehensive situation index, and generate a dynamic sand table containing multiple contingency plan evolution branches on a three-dimensional geographic base map.

2. The situational capability comprehensive assessment system according to claim 1, characterized in that, The multi-source situational awareness module includes a distributed data acquisition unit, a spatiotemporal marking unit, a confidence evaluation unit, a tuple encapsulation unit, and an edge caching unit; wherein: Distributed data acquisition unit: The physical layer deployed at edge nodes, including the following sub-units: Sensor interface subunit: Adapts to at least three heterogeneous sensors via industrial bus protocol to acquire voltage, image and motion vector data in real time and record the raw sampling frequency; The manual report parsing subunit is configured with a natural language processing engine to extract entities from the input text or voice situation reports and generate a structured event description table, including the event type, location of occurrence, and the identity code of the reporting personnel. Environmental feature capture subunit: Based on the LoRaWAN protocol, it connects to a weather station or geological monitoring instrument to collect environmental parameters, including temperature, humidity, vibration amplitude and electromagnetic interference intensity; Spatiotemporal marking unit: used to receive data output from the distributed data acquisition unit and unify the timestamps of each data to UTC standard time with an accuracy of ±1 millisecond; at the same time, based on the Beidou / GPS dual-mode positioning chip built into the data source device, the geographical location is mapped to a three-dimensional geographic coordinate system; Confidence assessment unit: Used to calculate confidence labels for labeled data, including sensor confidence, human report confidence, and environmental data confidence, denoted as sensor confidence. Confidence level of manual reports Environmental data confidence level ; Tuple Encapsulation Unit: Used to package data with spatiotemporal markers and confidence labels into raw situation tuples according to a preset format; Edge caching unit: used to establish a circular buffer on the local distributed node to temporarily store the original situation tuples in spatiotemporal tag order.

3. The situational capability comprehensive assessment system according to claim 1, characterized in that, The causal graph construction module includes an event causal relationship mining unit, a causal weight calculation unit, and a time evolution unit; wherein: Event Causality Mining Unit: Used to mine event causal relationships from the original situation tuples; based on the timestamps, spatial coordinates and confidence labels in the original data, it identifies the order and correlation of events, and uses a Bayesian network-based inference method to calculate the conditional dependencies between events, thereby generating a graph structure with causal relationships; Causal weight calculation unit: used to assign weights to each edge in the event causal relationship graph. The formula for calculating the weights is: ,in, For the event and events The weight of the causal relationship between them For the event To the event The conditional probability, and For the event and events confidence level These are weighting coefficients; Temporal Evolution Unit: Used to dynamically adjust the relationships between events in a causal graph based on the chronological order of events, generating a causal graph with temporal evolution.

4. The situational capability comprehensive assessment system according to claim 3, characterized in that, The event causality mining unit includes: Data preprocessing: Standardize sensor data, manual reports, and environmental characteristic data to ensure data consistency; Event identification and sorting: By analyzing spatiotemporal labeled data, relevant events are identified and sorted in chronological order; Reasoning for causal relationships between events: Based on event ranking, Bayesian network algorithms are used to reason about the causal relationships between events. Specifically, for each pair of events... Calculate its conditional probability , indicating an event Post-event probability of occurrence If conditional probability Exceeding the preset threshold Then the event is considered Regarding the event It has causal influence, that is, it establishes edges ; Constructing a causal graph structure: Based on the established edges, construct a causal graph structure according to the causal influence between events; specifically, each event... Represented as a node in a graph, an event and causal relationship between If an edge is represented as a directed edge in a graph, then the weight of the edge is... .

5. The situational capability comprehensive assessment system according to claim 4, characterized in that, The time evolution unit includes: Event time series sorting: Sort the timestamps of events to obtain an event sequence. Each event Corresponding to a timestamp And satisfy: ; Dynamic updating of temporal relationships: dynamically adjusting the strength of causal relationships between events in the cause-effect graph based on the temporal order of events; Time adjustment of causal relationship weights: Based on the time difference between events and the dynamically adjusted influence, the weights between events in the causal graph are updated. The updated weights are calculated using the following formula: ,in, For the original weights, The causal weights are adjusted for time. The coefficient for the attenuation of influence due to time difference; Generate a causal graph with time evolution: Based on the adjusted causal weights above, reconstruct a causal graph with time evolution.

6. The situational capability comprehensive assessment system according to claim 5, characterized in that, The multidimensional index coupling calculation module includes a threat transmission coefficient calculation unit, a resource chain effect index calculation unit, and a situation entropy value calculation unit; wherein: Threat Transmission Coefficient Calculation Unit: Used to calculate the degree of threat transmission between nodes in the system, i.e., the threat transmission coefficient, based on the event relationships in the time evolution causal graph. The formula is: ,in, For the threat from the incident Transmission to events The coefficients, the denominator is the event. To all related events The sum of weights; Resource Chain Effect Index Calculation Unit: Used to calculate the interdependence and chain effect index of resources between events. The formula is: ,in, The resource chain effect index represents the event. In the event Resource effects under influence and For the event and Resource weights; The situation entropy calculation unit is used to calculate the situation entropy value of the system, representing the uncertainty of the system state. First, it calculates the entropy value based on all event nodes and the weights of causal relationships between events in the time-evolution causal graph. Calculate the state probability for each event. Then, the situation entropy value is calculated based on the state probabilities of all events, using the following formula: ,in, The state entropy value represents the uncertainty of the event state in the system.

7. The situational capability comprehensive assessment system according to claim 1, characterized in that, The adaptive fusion module includes a current situation stage identification unit, a multi-dimensional indicator input unit, a gating mechanism unit, and a comprehensive situation index calculation unit; wherein: Current Situation Stage Identification Unit: Used to identify and extract current situation stage characteristics based on the current situation status of the system, including normal, alert, warning, and emergency. Multidimensional Indicator Input Unit: This unit receives calculation results from the multidimensional indicator coupling calculation module, including the threat transmission coefficient, resource chain effect index, and situational entropy value. It then standardizes these results, converting them into a unified scale, and denoted as the threat transmission coefficient. Resource chain effect index and situational entropy ; Gating mechanism unit: Used to adjust the degree of fusion of multi-dimensional indicators according to the characteristics of the current situation stage, and dynamically adjust the fusion weight of each indicator through the gating mechanism. Specifically, the fusion coefficient is calculated using a gating network based on the characteristics of the current situation stage. ; The comprehensive situation index calculation unit is used to sum the weighted multidimensional indicators based on the output of the gating mechanism to generate the final comprehensive situation index. .

8. The situational capability comprehensive assessment system according to claim 1, characterized in that, The causal source tracing feedback module includes a comprehensive situation index analysis unit, an impact factor extraction unit, a transmission path tracing unit, and a source tracing map construction unit; wherein: Comprehensive Situation Index Analysis Unit: Used to analyze the comprehensive situation index and identify relevant events and their relationships in its generation path; Impact Factor Extraction Unit: Used to extract relevant impact factors from the generation path of the comprehensive situation index and determine their contribution to the situation assessment results; Transmission Path Tracking Unit: Used to track the transmission path of influencing factors and identify the transmission chain of influencing factors in the causal graph; based on the results of event causal relationship mining, it traces back along the transmission path of influencing factors in the causal graph to gradually determine the source of influence; Source Mapping Unit: Used to construct source maps based on the extraction results of transmission paths and influencing factors.

9. The situational capability comprehensive assessment system according to claim 1, characterized in that, The situation simulation sand table module includes a source map fusion unit, a dynamic sand table generation unit, and an evolution path display unit; wherein: The source map fusion unit is used to fuse the source map generated by the causal source feedback module with the comprehensive situation index calculated by the adaptive fusion module. Specifically, it first aligns each event node in the source map with the relevant events in the comprehensive situation index; then, it combines the impact of each event in the source map on the situation with the calculation result of the comprehensive situation index using a weighted average method to obtain the fused situation data. ; Dynamic sand table generation unit: used to generate the sand table based on the fused situational data. A dynamic sand table containing multiple contingency plan evolution branches is generated on a three-dimensional geographic base map; Evolution path display unit: used to dynamically display multiple contingency plan evolution branches on a 3D sand table, intuitively showing the situational evolution under different scenarios.

10. The situational capability comprehensive assessment system according to claim 9, characterized in that, The dynamic sand table generation unit includes: Event Spatial Location Mapping: Plotting the spatial location of each event on a 3D geographic map, and assigning spatial coordinates to each event. Mapped onto a 3D geographic map; Evolutionary path initialization: based on fused situational data This generates multiple evolutionary branches, each representing a contingency plan evolution path under different scenarios.

Images