Intelligent urban management information visualization comprehensive management method and system

Abstract

The invention discloses an intelligent city management information visualization comprehensive management method and system, particularly relates to the technical field of city management, and is used for solving the problem of misalignment of emergency response caused by false association and causal confusion in the prior art. Precise management is realized through multi-modal time sequence data feature alignment, causal chain network dynamic construction and a visual decision mapping technology, firstly, multi-dimensional event features of video monitoring, sensors and text work orders are extracted, a cross-modal association chain is constructed, interference chains are screened out through causal strength evaluation and anti-fact verification, and a visual decision mapping algorithm is established. According to the method, dynamic correction parameters are generated in combination with space-time weights and feedback priorities, finally, a visual guidance scheme is output by fusing a causal chain network and a geographic information system, decision optimization is driven through causal logic, and city event disposal accuracy and cross-department cooperation efficiency are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of urban management. More specifically, the present invention relates to an intelligent urban management information visualization integrated management method and system. Background Art

[0002] In the field of urban management, event recognition and decision support usually rely on the correlation analysis of multi-modal data (such as video surveillance, sensor monitoring, text work orders, etc.). Event features are mainly extracted through rule matching or deep learning models based on a single modality, and statistical correlation metrics (such as co-occurrence frequency, temporal synchronization) are used to establish cross-modal event associations. For example, after identifying a traffic accident through video frame analysis, the environmental sensor data fluctuations in the same area are directly associated as causal evidence. After mapping different modality data to a unified feature space, this technical path uses a shallow association network for relationship modeling, without fully considering the potential confounding variable interference between multi-source data.

[0003] The prior art fails to effectively distinguish spurious correlation from real causal chains during the mining of multi-modal implicit event associations, resulting in decision-making suggestions deviating from actual requirements. For example, when multiple events (such as traffic congestion, air quality deterioration) are caused by common confounding factors (such as road construction), the surface data synchronization is easily misjudged as a direct causal relationship, ignoring the implicit interference of confounding variables on the relationship of multi-modal data, which will lead to resource scheduling errors (such as wrongly shutting down associated facilities) or risk prediction failures (such as failing to identify the real risk source), seriously affecting the emergency response efficiency of urban management. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide an intelligent urban management information visualization integrated management method and system to solve the problems raised in the above background art.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] An intelligent urban management information visualization integrated management method, comprising the following steps:

[0007] S1. Obtain the multi-modal time-series data stream of the urban management system and extract event features respectively, and map the event features to a multi-dimensional feature space;

[0008] Among them, the multi-modal time-series data stream includes video surveillance data, sensor monitoring data, and text work order data;

[0009] S2. Construct an initial event association chain based on the cross-modal temporal synchronization and co-occurrence frequency of event features in the multi-dimensional feature space;

[0010] S3. Screen out the interfering event chains that are synchronized but lack sufficient causal relationships in the initial event association chain to generate a purified association chain;

[0011] S4. Conduct causal consistency verification on the purified association chain;

[0012] S5. Adjust the spatio-temporal weight parameters and coverage density association factors of the purified association chain, and generate a dynamic correction weight by combining the feedback priorities of the text work order data;

[0013] S6. Integrate the dynamic correction weight and the purified association chain that passes the causal consistency verification to construct a cross-modal causal chain network and generate a visualization guidance plan.

[0014] In a preferred embodiment, obtain the multi-modal time-series data stream of the urban management system and extract event features respectively, and map the event features to a multi-dimensional feature space, including:

[0015] Obtain video surveillance data and extract the spatial distribution features in video frames through a pre-trained convolutional neural network;

[0016] Obtain sensor monitoring data and extract the time-series fluctuation features of sensor values through a recurrent neural network;

[0017] Obtain text work order data and extract the semantic event label features in the work order text through a natural language processing model;

[0018] Perform dimensionality reduction processing on the spatial distribution features, time-series fluctuation features, and semantic event label features respectively, map them to a multi-dimensional feature space with the same dimension, and adjust the directions of each feature vector so that the calculation results of the cosine similarity between different modal feature vectors are comparable.

[0019] In a preferred embodiment, construct an initial event association chain based on the cross-modal time-series synchronization and co-occurrence frequency of event features in the multi-dimensional feature space, including:

[0020] Based on the time-series alignment relationship of the spatial distribution feature sequence of video surveillance data, the time-series fluctuation features of sensor monitoring data, and the semantic event label features in the multi-dimensional feature space, calculate the cosine similarity of different modal feature vectors within the same time window as the cross-modal time-series synchronization index;

[0021] Statistically calculate the co-occurrence frequency of different modal feature vectors with an occurrence probability exceeding a preset threshold within the same geographical area. The co-occurrence frequency is calculated based on the ratio of the number of times multi-modal events occur simultaneously to the total number of events in historical data;

[0022] Fuse the cross-modal temporal synchronization index and the co-occurrence frequency with weights to generate an initial event association chain. The initial event association chain contains the association strength values and association directions between event nodes. The association direction is determined by the order of feature vectors within the time window.

[0023] In a preferred embodiment, filter out the interference event chains that are synchronized but lack sufficient causal relationships in the initial event association chain to generate a purified association chain, including:

[0024] Calculate the causal strength values between event nodes in the initial event association chain through a causal strength evaluation method based on temporal prediction error. The causal strength value is a quantization index of the influence degree of the precursor event node on the successor event node;

[0025] Verify the causal relationship between event nodes based on counterfactual analysis. If the decrease amplitude of the occurrence probability of the precursor event node after the successor event node is intervened and eliminated is lower than the preset threshold, then determine the corresponding association chain as an interference event chain;

[0026] Generate a filtering rule based on the causal strength value and the counterfactual analysis result, filter out the interference event chains with a causal strength value lower than the preset causal strength threshold or that do not meet the conditions in the counterfactual analysis, and retain the remaining association chains as the purified association chain.

[0027] In a preferred embodiment, perform causal consistency verification on the purified association chain, including:

[0028] Verify whether the logical trigger order of sub-events in the purified association chain conforms to the preset urban management business rules. The preset urban management business rules include the logical causal constraint conditions between event types and the trigger time interval range;

[0029] Generate counterfactual temporal data corresponding to the interference event chain, simulate the causal relationship failure scenario through the temporal data reconstruction method, and extract the cross-modal abnormal window caused by the causal relationship failure;

[0030] Calculate the temporal dependence strength attenuation rate between the temporal fluctuation characteristics of sensor monitoring data and the spatial distribution characteristic sequence of video surveillance data within the cross-modal abnormal window;

[0031] If the logical trigger order conforms to the preset urban management business rules and the temporal dependence strength attenuation rate is within the preset threshold tolerance range, then determine that the causal consistency verification passes.

[0032] In a preferred embodiment, adjust the spatio-temporal weight parameters and coverage density association factors of the purified association chain, and generate dynamic correction weights in combination with the feedback priority of text work order data, including:

[0033] Calculate the time decay factor and spatial proximity based on the occurrence time and geographical location encoding of event nodes in the purification association chain, and generate spatio-temporal weight parameters;

[0034] Statistically analyze the distribution density of event types in the same geographical area in historical data to generate a coverage density correlation factor. The event type distribution density is the ratio of the number of occurrences of an event type within a preset time period to the area of the region;

[0035] Extract the feedback priority from the semantic event label features of text work order data. The feedback priority is calculated based on the weighted average of the classification score of the urgency marked in the work order text and the historical processing time;

[0036] Fuse the spatio-temporal weight parameters, coverage density correlation factor, and feedback priority through conditional rules to generate a dynamically corrected weight.

[0037] In a preferred embodiment, the time decay factor is calculated by exponential decay according to the interval between the event occurrence time and the current time, and the spatial proximity is calculated based on the geographical location encoding distance between event nodes.

[0038] In a preferred embodiment, the conditional rules for conditional rule fusion are as follows: when the feedback priority reaches the preset emergency threshold, the dynamically corrected weight is determined by the product of the spatio-temporal weight parameters and the coverage density correlation factor; when the feedback priority does not reach the emergency threshold, the dynamically corrected weight is determined by the ratio of the coverage density correlation factor to the feedback priority.

[0039] In a preferred embodiment, integrate the dynamically corrected weight and the purification association chain verified by causal consistency, and construct a cross-modal causal chain network and generate a visualization guidance scheme, including:

[0040] Perform weight mapping of nodes and edges on the dynamically corrected weight and the verified purification association chain to construct a cross-modal causal chain network with a directed weighted graph structure. The nodes represent event types, the edges represent causal relationships, and the weight values are the dynamically corrected weights;

[0041] Extract the critical path based on the cross-modal causal chain network. The screening conditions for the critical path are that the total path weight exceeds the preset display threshold and the path length meets the requirement of the minimum number of event chain nodes;

[0042] Overlay and render the critical path on the urban geographic information system map to generate a multi-dimensional interactive visualization guidance scheme;

[0043] Output the visualization guidance scheme and synchronously generate a cross-departmental task priority list. The task priority is sorted in descending order according to the weight values of event nodes in the critical path. Event nodes with the same weight value determine the execution order according to topological sorting.

[0044] On the other hand, the present invention provides an intelligent urban management information visualization integrated management system, including the following modules:

[0045] Data acquisition module: used to obtain the multi-modal time-series data stream of the urban management system, extract event features respectively, and map the event features to a multi-dimensional feature space;

[0046] Association construction module: used to construct an initial event association chain based on the cross-modal time-series synchronization and co-occurrence frequency of event features in the multi-dimensional feature space;

[0047] Interference screening module: used to screen out the interference event chains with sufficient synchronization but insufficient causal relationship in the initial event association chain, and generate a purified association chain;

[0048] Causal verification module: used to verify the causal consistency of the purified association chain;

[0049] Weight adjustment module: used to adjust the spatio-temporal weight parameters and coverage density correlation factors of the purified association chain, and generate a dynamic correction weight by combining the feedback priority of text work order data;

[0050] Solution generation module: used to integrate the dynamic correction weight and the purified association chain passed through the causal consistency verification, construct a cross-modal causal chain network and generate a visualization guidance solution.

[0051] Compared with the prior art, the present invention has the following beneficial effects:

[0052] 1. Through the deep causal association modeling and dynamic weight fusion mechanism of multi-modal data, the recognition accuracy and decision reliability of urban management events are significantly improved; by using cross-modal time-series synchronization analysis and counterfactual causal verification technology, it is possible to effectively identify and screen out false associations caused by confounding variables (such as road construction simultaneously causing traffic congestion and air quality fluctuations), ensuring the authenticity and interpretability of the causal chain, thereby avoiding misjudgment-induced incorrect shutdown of facilities or missed detection of risk sources; through the dynamic adjustment of spatio-temporal weight parameters and coverage density correlation factors, combined with the feedback priority of text work orders, the adaptive optimization of resource scheduling strategies is realized, solving the problem of imbalance between local event handling and global resource allocation caused by static weights in traditional methods.

[0053] 2. Through multi-modal causal modeling and dynamic parameter fusion, a closed-loop iterative causal decision-making chain is formed; it can not only dynamically correct causal biases caused by data noise or environmental changes, but also intuitively present the event association strength and handling priority through the topological structure and weight heat map of the visualization chain network, providing an operable decision-making guide for cross-departmental collaboration, and solving the defects of traditional methods that rely on manual experience for parameter adjustment, have a lagged response, and lack interpretability. Description of the Drawings

[0054] Figure 1This is the flowchart of the intelligent urban management information visualization comprehensive management method of the present invention;

[0055] Figure 2 This is the structural schematic diagram of the intelligent urban management information visualization comprehensive management system of the present invention. Specific implementation manners

[0056] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0057] Embodiment 1: Figure 1 The intelligent urban management information visualization comprehensive management method of the present invention is given, including the following steps:

[0058] S1. Obtain the multi-modal time-series data stream of the urban management system and extract event features respectively, and map the event features to a multi-dimensional feature space;

[0059] Among them, the multi-modal time-series data stream includes video surveillance data, sensor monitoring data, and text work order data;

[0060] S2. Construct an initial event association chain based on the cross-modal time-series synchronization and co-occurrence frequency of event features in the multi-dimensional feature space;

[0061] S3. Screen out the interference event chains with sufficient synchronization but insufficient causal relationship in the initial event association chain to generate a purified association chain;

[0062] S4. Perform causal consistency verification on the purified association chain;

[0063] S5. Adjust the spatio-temporal weight parameters and coverage density association factors of the purified association chain, and generate a dynamic correction weight in combination with the feedback priority of the text work order data;

[0064] S6. Integrate the dynamic correction weight and the purified association chain that passes the causal consistency verification, construct a cross-modal causal chain network and generate a visualization guidance plan.

[0065] S1. Obtain the multi-modal time-series data stream of the urban management system and extract event features respectively, and map the event features to a multi-dimensional feature space. The specific implementation is as follows:

[0066] The process of obtaining video surveillance data and extracting the spatial distribution features in video frames through a pre-trained convolutional neural network includes: accessing real-time video stream data from the urban management system, where the video stream data comes from camera devices deployed in various regions of the city and is transmitted to the data processing server at a rate of 25 frames per second; splitting the video stream data into video segments according to a fixed time window, with each video segment containing consecutive multiple frames of images; using the pre-trained convolutional neural network to extract features from each video segment, where the training data of the convolutional neural network comes from the publicly available urban traffic surveillance dataset and the training task is the localization and classification of vehicles, pedestrians, and road facilities in the video; the convolutional neural network outputs the spatial distribution features of the target objects in each video segment, and the spatial distribution features include the position coordinates, size ratios, and class probability distributions of the target objects in the video frame; splicing the spatial distribution features of multiple frames of images in the same video segment in chronological order to form a spatial distribution feature sequence of the video surveillance data.

[0067] The process of obtaining sensor monitoring data and extracting the temporal fluctuation features of sensor values through a recurrent neural network includes: accessing sensor devices deployed in urban drainage pipe networks, traffic intersections, and environmental monitoring stations, where the sensor devices include flow meters, temperature and humidity sensors, and air quality monitors; the sensor devices upload monitoring data at a fixed sampling frequency, and the sampling frequency is set according to the sensor type, with the flow meter collecting data once per second and the temperature and humidity sensor collecting data once every 10 seconds; aligning the sensor monitoring data according to the device number and timestamp to generate the temporal data stream of each sensor; using the recurrent neural network to extract features from the temporal data stream, where the input layer of the recurrent neural network receives the sequence of sensor values sorted by time, the hidden layer adopts a long short-term memory unit structure, and the output layer generates the temporal fluctuation features of the sensor values; the temporal fluctuation features include the mean, variance, peak frequency, and differential change amount between adjacent sampling points of the sensor values within a preset time window; for the temporal fluctuation features of multiple sensors in the same geographical area, weighted fusion is performed according to the spatial positions of the sensors to generate the regional-level sensor temporal fluctuation features, where when performing the weighted fusion of the sensor spatial positions, the weight values are dynamically calculated according to the deployment density of the sensor devices in the urban drainage pipe network and traffic intersections.

[0068] The process of obtaining text work order data and extracting semantic event label features from the work order text through a natural language processing model includes: exporting historical work order data and real-time reported work order data from the urban management work order system, where the work order data includes a text description field, a reporting timestamp, and a geographic location code; preprocessing the work order text, which includes removing stop words, word segmentation, part-of-speech tagging, and named entity recognition; inputting the preprocessed work order text into a natural language processing model, which adopts a sequence encoding structure based on the attention mechanism, and the model training task is to identify event types, impact scopes, and urgency level labels from the work order text; the natural language processing model outputs the semantic event label features of the work order text, and the semantic event label features include a probability distribution vector of the event type, a keyword weight list of the impact scope, and a classification score of the urgency level; aggregating the semantic event label features of multiple work orders within the same geographic region, and the aggregation method is to take the weighted average of the probability distribution vectors of the same event type, and the weight is determined by the interval between the work order reporting time and the current time, and the shorter the interval, the higher the weight.

[0069] The process of separately performing dimensionality reduction on the spatial distribution features, temporal fluctuation features, and semantic event label features, mapping them to a multi-dimensional feature space with the same dimension, and adjusting the directions of each feature vector so that the cosine similarity calculation results between different modality feature vectors are comparable includes: performing dimensionality reduction on the spatial distribution feature sequence of video surveillance data using the principal component analysis method, and retaining the principal components with a variance contribution rate exceeding 85% as the dimensionality-reduced spatial distribution features; performing dimensionality reduction on the regional-level sensor temporal fluctuation features using the linear discriminant analysis method so that the sensor features of different event categories have the maximum separability in the low-dimensional space; performing dimensionality reduction on the aggregated semantic event label features using the non-negative matrix factorization method, with the constraint that each dimension value of the dimensionality-reduced feature vector is non-negative; uniformly scaling the dimensionality-reduced feature vectors of the three modalities to the same number of dimensions, and the number of dimensions is determined through experimental verification. The specific method is to test the accuracy of cross-modal feature matching under different numbers of dimensions on historical data, and select the number of dimensions with the highest accuracy as the final set value; adjusting the directions of each modality feature vector, and the adjustment method is to normalize each feature vector so that its norm is 1, and rotate the feature vector to maximize the cosine value of the angle between different modality feature vectors under the same event category; after the direction adjustment, the cosine similarity calculation results between different modality feature vectors can be directly used for cross-modal correlation measurement.

[0070] In the above embodiments, the specific network structure parameters (such as the number of layers and the number of neurons) of the pre-trained convolutional neural network, recurrent neural network, and natural language processing model can be adjusted according to the actual data scale, but the consistency of the feature extraction process for different modalities needs to be ensured; in the dimensionality reduction process, the selection basis of technical parameters such as the variance contribution rate, the class separability objective of linear discriminant analysis (the class separability objective of linear discriminant analysis is achieved by maximizing the ratio of between-class scatter to within-class scatter), and the constraint conditions of non-negative matrix factorization needs to be verified by historical data to determine the optimal value; in the process of adjusting the direction of the feature vector, the implementation of normalization processing and rotation operation needs to use a numerical optimization algorithm, such as the gradient descent method to find the optimal rotation matrix, but the specific value of the rotation matrix is not limited to a fixed value. Ensure the comparability of multi-modal data in the unified feature space and the feasibility of correlation analysis, and at the same time avoid the problem of causal confusion caused by inconsistent feature scales and directions.

[0071] S2. Construct an initial event correlation chain based on the cross-modal temporal synchronization and co-occurrence frequency of event features in the multi-dimensional feature space. The specific implementation is as follows:

[0072] The process of calculating the cosine similarity between different modality feature vectors within the same time window as the cross-modal temporal synchronization index based on the spatial distribution feature sequence of video surveillance data, the temporal fluctuation feature of sensor monitoring data, and the temporal alignment relationship of semantic event label features in the multi-dimensional feature space includes: according to the timestamp information of each modality feature vector in the multi-dimensional feature space generated in step S1, divide different modality data into multiple time segments according to a fixed time window length. The setting basis of the time window length is the typical response cycle of urban management events. For example, the response cycle of traffic congestion events is usually 5 minutes, so the time window length is set to 5 minutes; within each time window, align the spatial distribution feature sequence of video surveillance data, the temporal fluctuation feature of sensor monitoring data, and the semantic event label feature at the same moment. The alignment method is to interpolate the feature vector of the missing time point as the feature vector of the previous time point; for different modality feature vectors within each time window, calculate the cosine similarity between each pair, such as the cosine similarity between the spatial distribution feature sequence of video surveillance data and the temporal fluctuation feature of sensor monitoring data. The calculation method of the cosine similarity is to perform a dot product operation on the two feature vectors and then divide by the product of their norms; accumulate and average the cosine similarities of all modality feature vector pairs according to the time window to obtain the cross-modal temporal synchronization index.

[0073] The process of counting the co-occurrence frequency of different modal feature vectors with occurrence probabilities exceeding a preset threshold within the same geographical area includes: grouping the feature vectors in the multi-dimensional feature space according to the geographical location coding based on the same geographical area division rule defined in step S1; for the multi-modal feature vectors within each geographical area, counting the number of times they co-occur within the same time window in the historical data. For example, if the spatial distribution feature sequence of video surveillance data indicates a traffic congestion event, and the temporal fluctuation feature of sensor monitoring data indicates an over-limit traffic flow within the same time window, it is determined as one co-occurrence event; when calculating the co-occurrence frequency, divide the number of co-occurrences by the total number of all events in this geographical area in the historical data. For example, in a certain area, 100 events occurred in the past month, and among them, the spatial distribution feature sequence of video surveillance data and the temporal fluctuation feature of sensor monitoring data co-occurred 30 times, then the co-occurrence frequency is 30%; the preset threshold is set based on determining the lowest co-occurrence significance level of different event types through historical data analysis. For example, the co-occurrence frequency threshold for traffic congestion and over-limit traffic flow is set at 20%, and if it is lower than this threshold, the co-occurrence is considered to have no statistical significance.

[0074] The process of weighted fusion of the cross-modal temporal synchronization index and the co-occurrence frequency to generate an initial event association chain includes: assigning weight coefficients to the cross-modal temporal synchronization index and the co-occurrence frequency. The weight coefficients are set based on verifying the prediction contribution degree of different indicators to the real event association through historical data. For example, if the accuracy rate of the temporal synchronization index accounts for 60% in the historical data, then its weight coefficient is set at 0.6, and the co-occurrence frequency weight coefficient is set at 0.4; the calculation method of weighted fusion is to multiply the temporal synchronization index by the weight coefficient and add the co-occurrence frequency multiplied by the weight coefficient to generate a comprehensive association strength value; the association direction is determined by analyzing the appearance order of the feature vectors within the time window. For example, if the occurrence time of the traffic accident indicated by the spatial distribution feature sequence of video surveillance data is earlier than the time when the temporal fluctuation feature of sensor monitoring data indicates a decrease in traffic flow, then the association direction is from the video surveillance event to the sensor monitoring event; the generation rule of the initial event association chain is to connect the node pairs with the comprehensive association strength value exceeding the preset strength threshold and the same association direction to form a directed association chain network. The preset strength threshold is determined through experimental verification. For example, in the test data, gradually increase the threshold until the false alarm rate of the association chain is lower than 5%.

[0075] In the above embodiments, the setting of the time window length needs to be adjusted in combination with specific event types. For example, for pipeline leakage events, since the response delay of the time series fluctuation characteristics of sensor monitoring data is relatively long, the time window can be extended to 10 minutes. The co-occurrence frequency statistics need to exclude abnormal co-occurrence situations caused by data acquisition equipment failures. For example, when the time series fluctuation characteristics of all sensor monitoring data are missing within the same time window, this time window is not included in the statistics. The weight coefficients of weighted fusion can be dynamically adjusted. For example, when the historical data accumulation of a certain type of event is insufficient, an equal weight distribution method is adopted. The judgment of the association direction needs to consider the timestamp accuracy of the feature vector. For example, when the timestamps of the spatial distribution feature sequences of video surveillance data are accurate to the second level, and the time series fluctuation characteristics of sensor monitoring data are at the millisecond level, the millisecond-level timestamps are used as the basis for judging the sequence.

[0076] S3. Screen out the interference event chains that are synchronous but lack causal relationship in the initial event association chain to generate a purified association chain. The specific implementation is as follows:

[0077] The process of calculating the causal strength value between event nodes in the initial event association chain by the causal strength evaluation method based on the time series prediction error includes: extracting event node pairs from the initial event association chain generated in step S2, and each event node pair includes a predecessor event node and a successor event node; for each event node pair, obtaining the occurrence time series of the predecessor event node in the historical data, and training a time series prediction model based on this time series. The training objective of the time series prediction model is to predict the occurrence time of the successor event node according to the historical occurrence time of the predecessor event node; comparing the actually observed occurrence time of the successor event node with the predicted time, and calculating the absolute value of the prediction error. The calculation method of the causal strength value is to subtract the absolute value of the prediction error from a preset constant. The setting basis of the preset constant is twice the maximum absolute value of the prediction error in the historical data. For example, if the maximum prediction error in the historical data is 0.5, the preset constant is set to 1.0; the larger the causal strength value, the higher the influence degree of the predecessor event node on the successor event node.

[0078] The process of verifying the causal relationship between event nodes based on counterfactual analysis includes: performing an intervention operation on each pair of event nodes in the initial event association chain. The intervention operation is to intervene and eliminate the occurrence record of the subsequent event node in the historical data; recalculate the occurrence probability of the predecessor event node after the intervention and compare it with the occurrence probability before the intervention; if the decrease amplitude of the occurrence probability of the predecessor event node is lower than the preset threshold, then determine that this association chain is a disturbing event chain. The setting basis of the preset threshold is to statistically calculate the lowest value of the decrease amplitude of the occurrence probability of the true causal relationship events in the historical data. For example, in the true causal relationship between traffic accidents and the decrease in traffic flow, after intervening to eliminate the occurrence of the traffic flow decrease event, the decrease amplitude of the traffic accident occurrence probability is 40%, then the preset threshold is set to 30%; the verification process of counterfactual analysis needs to be repeated multiple times to reduce the influence of random errors. The number of repetitions is set according to the amount of historical data. For example, when the amount of data is greater than 1000, it is repeated 10 times.

[0079] The process of generating a filtering rule based on the causal intensity value and the counterfactual analysis result includes: setting a causal intensity threshold, which is determined by adjusting through a test data set. The test data set consists of 80% of the samples in the historical data, and the remaining 20% is used to verify the misjudgment rate; the adjustment method is to gradually increase the threshold in the test data until the misjudgment rate of the purified association chain is lower than the preset misjudgment rate threshold. For example, when the misjudgment rate threshold is set to 5%, the causal intensity threshold is set to 0.7; mark the association chains in the initial event association chain whose causal intensity value is lower than 0.7 or the decrease amplitude of the occurrence probability in the counterfactual analysis is lower than 30% as disturbing event chains; screen out the marked disturbing event chains and retain the remaining association chains as purified association chains; store the purified association chains in a directed graph structure, where the nodes represent event types and the edges represent causal relationships and association intensity values.

[0080] In the above implementation, the specific type of the time series prediction model can be selected according to the data characteristics. For example, for traffic congestion events with obvious periodicity, an autoregressive integrated moving average model is used. For pipeline leakage events with strong randomness, a random forest regression model is used; the preset constant of the causal intensity value needs to be dynamically updated. For example, recalculate the absolute value of the maximum prediction error monthly based on the new data and adjust the constant; the intervention operation of counterfactual analysis needs to exclude the interference of external factors. For example, when intervening to eliminate the record of the subsequent event node, it is necessary to synchronously exclude the influence of other associated events occurring during the same period; the setting of the filtering rule needs to support manual adjustment. For example, when the urban management strategy changes, it is allowed to manually increase the causal intensity threshold to adapt to the new strategy. It can effectively screen out false association chains and improve the reliability of causal associations.

[0081] S4. Perform causal consistency verification on the purified association chains. The specific implementation is as follows:

[0082] The process of verifying whether the logical trigger order of sub-events in the purification association chain conforms to the preset urban management business rules includes: obtaining the preset urban management business rules from the business rule library of the urban management department. The preset urban management business rules include the logical causal constraint conditions and trigger time interval ranges between event types. For example, after a traffic accident occurs, there must be a traffic guidance event and the trigger time interval does not exceed 10 minutes. Extract the sub-events in the purification association chain in chronological order and match the logical causal constraint conditions one by one. If the subsequent event type of a certain sub-event is not within the range allowed by the constraint conditions, it is determined that the logical trigger order does not conform. For the verification of the trigger time interval, calculate the time difference between the sub-event and its subsequent event. If the time difference exceeds the preset time interval range, for example, the time difference between the traffic accident and the traffic guidance event exceeds 15 minutes, it is determined that it does not conform to the business rules.

[0083] The process of generating counterfactual time series data corresponding to the interference event chain includes: extracting event node pairs from the interference event chain screened in step S3, and performing causal relationship failure simulation on each event node pair. The simulation method is to randomly shuffle the time series relationship between the predecessor event node and the successor event node in the historical data or replace it with the time stamps of irrelevant events. Generate counterfactual time series data through the time series data reconstruction method. The specific operation of the time series data reconstruction method is to retain the original numerical values of the time series fluctuation characteristics of the sensor monitoring data and the spatial distribution characteristic sequence of the video monitoring data, but rearrange their time order to make the causal relationship fail. Extract the cross-modal abnormal window caused by the failure of the causal relationship. The recognition method of the cross-modal abnormal window is to detect the time period when the synchronization of the time series fluctuation characteristics of the sensor monitoring data and the spatial distribution characteristic sequence of the video monitoring data within the time window is lower than the preset synchronization threshold. For example, when the preset synchronization threshold is 0.6, the time period with synchronization lower than 0.6 is determined as the abnormal window.

[0084] The process of calculating the decay rate of the temporal dependence strength between the time series fluctuation characteristics of the sensor monitoring data and the spatial distribution characteristic sequence of the video monitoring data within the cross-modal abnormal window includes: In the real data, the calculation process of the temporal dependence strength includes calculating the covariance of the time series fluctuation characteristics of the sensor monitoring data and the spatial distribution characteristic sequence of the video monitoring data and normalizing it to between 0 and 1. In the counterfactual data, perform the same calculation on the same time window to obtain the counterfactual temporal dependence strength. The calculation method of the decay rate is to divide the counterfactual temporal dependence strength by the real temporal dependence strength. If the decay rate exceeds 1, it means that the dependence relationship is enhanced, and if it is lower than 1, it means that the dependence relationship is weakened. The preset threshold tolerance range is determined according to the decay rate distribution interval of the normal event association chain in the historical data. For example, if the decay rate of normal events is distributed between 0.85 and 1.15, the threshold tolerance range is set to 0.8 to 1.2.

[0085] If the logical trigger order conforms to the preset urban management business rules and the decay rate of the temporal sequence dependence intensity is within the preset threshold tolerance range, the process of determining that the causal consistency verification passes includes: double-checking each sub-event chain in the purification association chain. First, verify whether the logical trigger order conforms to the business rules, and second, verify whether its corresponding decay rate is within the tolerance range; if the sub-event chain meets both conditions, it is marked as a causal association chain that passes the verification; for the sub-event chain that fails the verification, it is returned to step S3 to re-perform interference chain screening and parameter adjustment.

[0086] In the above implementation, the preset urban management business rule library needs to be updated regularly. For example, when new night construction control rules are added to urban traffic management, corresponding event type constraints and time interval ranges need to be supplemented in the business rule library; the timestamp replacement operation in the temporal data reconstruction method needs to ensure that the geographical location encoding of the replacement event is the same as that of the original event to avoid introducing spatial dimension interference; the preset synchronization threshold of the cross-modal abnormal window is set through historical data analysis. For example, the lowest synchronization of the normal event association chain in the historical data is selected as the threshold benchmark. If the lowest synchronization of the normal event is 0.6, the threshold is set to 0.6; the statistical range of the decay rate distribution needs to exclude extreme outliers. For example, only the data within 3σ of the normal distribution is used to calculate the upper and lower limits of the interval. If the decay rate mean is 1.0 and the standard deviation is 0.1, the interval is set to 0.7 to 1.3.

[0087] Step S4 improves the reliability of causal associations through a double-check mechanism (logical trigger order and counterfactual data verification). Traditional methods only rely on statistical correlations or single-modal verification, which easily ignore the business logic constraints between events and the risk of dynamic failure of causal relationships; through preset urban management business rule verification, it forces the causal chain to conform to actual management experience (such as there must be a traffic guidance event after a traffic accident) to avoid logical contradictions; at the same time, counterfactual temporal data is introduced to simulate the causal relationship failure scenario, and the cross-modal dependence intensity decay rate is quantified to identify false associations caused by data noise or accidental co-occurrence; compared with existing technologies, integrating the prior knowledge of business rules and data-driven verification solves the problems of disconnection between causal associations and business logic and weak anti-interference ability, thus significantly improving the interpretability of urban management decisions and the accuracy of emergency responses.

[0088] S5. Adjust the spatio-temporal weight parameters and coverage density association factors of the purification association chain, and generate dynamic correction weights in combination with the feedback priority of text work order data. The specific implementation is as follows:

[0089] The process of calculating the time decay factor and spatial proximity based on the occurrence time and geographical location encoding of event nodes in the purification association chain includes: extracting the timestamp and geographical location encoding of event nodes from the purification association chain verified in step S4. The timestamp is accurate to the second level, and the geographical location encoding adopts the longitude-latitude grid encoding rule, and the grid division is consistent with the urban management geographic information system; the calculation method of the time decay factor is to dynamically select the decay coefficient according to the event type label. For example, the coefficient 0.85 is used for traffic accidents, and the coefficient 0.7 is used for pipeline leaks, and the rationality of the coefficient is verified by regression analysis of historical data; the time decay factor is calculated according to the exponential function of the interval days. For example, when the interval of a traffic accident is 3 days, the factor is 0.85^3≈0.614; the calculation method of spatial proximity is to use the Haversine formula to calculate the geographical spherical distance between event nodes, which is limited to the urban-level geographical range, and the distance is normalized to the range of 0 to 1. The smaller the value, the higher the proximity.

[0090] The process of generating the coverage density correlation factor by counting the distribution density of event types in the same geographical area in historical data includes: dividing the urban geographical area into square kilometer grids, counting the occurrence times of specific event types in each grid within a preset time period, and setting the minimum occurrence times threshold to 1 to avoid too small a denominator; the calculation method of the event type distribution density is (number of occurrences + 1) / (grid area + 1), and the Laplace smoothing is used to process the distortion of sparse data; the coverage density correlation factor automatically grades the historical density data through the K-means clustering algorithm. For example, the density is clustered into three categories: low, medium, and high, and mapped to 0.3, 0.6, and 0.9, and the threshold generation does not require manual intervention.

[0091] The process of extracting the feedback priority in the semantic event label features of text work order data includes: extracting the urgency classification score (high 1.0, medium 0.5, low 0.2) and historical processing time efficiency score (if the time efficiency ≤ 30 minutes, the score is 1.0, and if it is > 30 minutes, it decreases proportionally) from the semantic event label features generated in step S1; in the cold start stage, the feedback priority is calculated with equal weights (urgency 50%, processing time efficiency 50%), and when the accumulated historical data exceeds 100 pieces, it switches to dynamic weights (selecting the optimal ratio through A / B testing, such as 70% and 30%); the final value of the feedback priority is the weighted score. For example, for a traffic accident, the urgency score of 1.0×0.7 + the processing time efficiency score of 1.0×0.3 = 1.0.

[0092] The process of conditionally fusing the spatiotemporal weight parameter, coverage density correlation factor and feedback priority includes: the preset emergency threshold is dynamically set by the top 15% quantile of historical data (for example, 0.9), and a smooth transition interval of ±5% of the threshold is introduced (0.85-0.95); when the feedback priority is ≥0.95, the dynamic correction weight = spatiotemporal weight parameter × coverage density correlation factor; when the feedback priority is ≤0.85, the dynamic correction weight = coverage density correlation factor / feedback priority; within the transition interval (0.85<priority<0.95), linear interpolation is used to mix the two calculation results, for example, when the priority is 0.90, the weight = 0.5×product+0.5×ratio; the final weight is normalized to the range of 0-1 to ensure compatibility with the visualization display rule of step S6.

[0093] In the above implementation, the time decay coefficient is dynamically updated every month through sliding window regression analysis of newly added data; spatial proximity calculation is limited to city-level application to avoid spherical error; coverage density clustering is automatically performed every quarter to ensure that the threshold adapts to the latest data; feedback priority weights are optimized through A / B testing every two weeks after cold start; and smooth transition rules eliminate the risk of conditional jumps.

[0094] Step S5 optimizes the accuracy of urban management decisions through a multi-dimensional dynamic parameter fusion mechanism. Traditional methods use static weight allocation, which cannot adapt to the spatiotemporal heterogeneity and dynamic changes in the urgency of urban events. This step introduces an adaptive attenuation coefficient for event types (such as 0.85 for traffic accidents and 0.7 for pipe network leaks), and quantifies the time-effect attenuation law by regressing historical data to solve the response deviation caused by unified parameters; combines Haversine spherical distance calculation and coverage density clustering and grading to improve geographic correlation accuracy; and designs a threshold smooth transition interval in conditional rule fusion to avoid the weight jump problem caused by traditional hard thresholds. Through dynamic parameter adaptation, spatial precise calculation, and rule smoothing, the real-time and accuracy of the visualization guidance scheme in resource scheduling and emergency response are significantly improved.

[0095] S6. Integrate the dynamic correction weights with the purified association chain verified by causal consistency, build a cross-modal causal chain network and generate a visual guidance plan. The specific implementation is as follows:

[0096] The process of constructing a cross-modal causal chain network with a directed weighted graph structure by mapping the weights of nodes and edges between the dynamically corrected weights and the purified association chains verified by causal consistency includes: obtaining the dynamically corrected weights from step S5, which are generated by fusing through the conditional rules in step S5, and the weight values are normalized values from 0 to 1; obtaining the purified association chains verified by causal consistency from step S4, and the purified association chains include the causal relationships and association directions between event nodes; defining the nodes of the directed weighted graph as event types (such as traffic accidents, pipeline network leaks), and the node attributes include the event occurrence time and geographical location coding; defining the edges as causal relationships, the starting point of the edge is the predecessor event node, the end point is the successor event node, and the edge attributes include the dynamically corrected weight value and the association direction; the construction rule is to only retain the edges with a dynamically corrected weight exceeding 0.3 to exclude low-weight noise interference, and the threshold 0.3 is determined according to the false alarm rate statistics of low-weight edges in historical data.

[0097] The process of extracting the critical path based on the cross-modal causal chain network includes: presetting the display threshold, which is determined by analyzing the distribution of the total path weights corresponding to effective management decisions in historical data. For example, in the traffic accident handling scenario, if the total path weights of 80% of the effective decision paths exceed 2.5, then the display threshold is set to 2.5; the minimum number of event chain nodes is required to be set according to the urban management business rules. For example, traffic accident handling requires at least 3 nodes including accident reporting, traffic control, and on-site cleaning; traversing all possible paths in the directed weighted graph, calculating the total path weight (such as if the path contains 3 edges with weights of 0.8, 0.7, and 0.6 respectively, then the total is 2.1), and screening out the paths with a total exceeding 2.5 and the number of nodes ≥ 3 as the critical path.

[0098] The process of overlaying and rendering the critical path on the urban geographic information system map includes: accessing the map layer of the urban geographic information system, and the map layer contains the spatial data of roads, buildings, and municipal facilities; mapping the event nodes in the critical path to the corresponding coordinate points on the map according to the geographical location coding, and the display rule of dynamic marking is: nodes with a weight value ≥ 0.7 are displayed as red icons, 0.5 ≤ weight value < 0.7 are displayed as orange icons, and weight value < 0.5 are displayed as yellow icons. The color gradient is verified through human-computer interaction experiments to ensure the recognition rate; the flow arrows of the causal edges are drawn according to the association direction, and the thickness of the arrows is proportional to the dynamically corrected weight value. For example, an edge with a weight of 0.9 uses a 3-pixel thick arrow, and an edge with a weight of 0.5 uses a 1-pixel thin arrow; the weight heat map is generated by superimposing the weight values of event nodes in the same geographical area, and the color gradient transitions from blue (low weight) to red (high weight). The color scale division of the heat map is based on the equal-frequency binning method.

[0099] The process of outputting the visualization guidance plan and synchronously generating the cross-departmental task priority list includes: outputting the visualization guidance plan to the display terminal of the urban management command center, supporting users to view event details through touch interaction (such as clicking on an icon to display time, location, and related events); the generation rule of the cross-departmental task priority list is: extract all event nodes in the critical path, and sort them from high to low according to the dynamically corrected weight values. For example, if the weight of node A is 0.9, the weight of node B is 0.8, and the weight of node C is 0.8, then the priority order is A > B > C; when the weight values are the same, determine the execution order according to topological sorting. The rule of topological sorting is to give priority to executing nodes with an in-degree of 0 (i.e., nodes without predecessor event constraints). For example, if the weights of nodes B and C are both 0.8 and there is no topological dependence, then sort them according to the chronological order of event occurrence.

[0100] In the above embodiments, the construction of the directed weighted graph needs to exclude cyclic dependency paths (such as event A → event B → event A). The detection method for cyclic paths is to mark the visited nodes in depth-first search and automatically eliminate them when a cycle is detected; the update period of the preset display threshold is synchronized with the adjustment of urban management strategies. For example, recalculate the threshold according to the new strategy every quarter; the color gradient and size parameters of the visualization elements need to comply with the human-computer interaction design specifications (such as using a blue-yellow gradient instead of a red-green contrast for a color-blind-friendly color scheme); the output format of the task priority list supports docking with the interfaces of third-party scheduling systems. The interface protocol uses JSON format to transmit fields such as event types, weight values, and execution orders. The naming of JSON fields is consistent with the urban management data dictionary. Convert complex causal association relationships into actionable decision-making plans while ensuring the efficiency and consistency of cross-departmental collaborative execution.

[0101] Embodiment 2: Figure 2 The structural schematic diagram of the intelligent urban management information visualization integrated management system of the present invention is given. The intelligent urban management information visualization integrated management system includes the following modules:

[0102] Data acquisition module: used to obtain the multi-modal time-series data stream of the urban management system and extract event features respectively, and map the event features to a multi-dimensional feature space;

[0103] Association construction module: used to construct an initial event association chain based on the cross-modal time-series synchronization and co-occurrence frequency of event features in the multi-dimensional feature space;

[0104] Interference screening module: used to screen out interference event chains with insufficient causal relationships but synchronization in the initial event association chain, and generate a purified association chain;

[0105] Causal verification module: used to verify the causal consistency of the purified association chain;

[0106] Weight adjustment module: used to adjust the spatio-temporal weight parameters and coverage density correlation factors of the purification association chain, and generate dynamic correction weights by combining the feedback priorities of text work order data;

[0107] Solution generation module: used to integrate the dynamically corrected weights and the purification association chains verified by causal consistency, construct a cross-modal causal chain network, and generate a visual guidance solution.

[0108] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to obtain a formula closest to the actual situation. The preset parameters and threshold selection in the formulas are set by those skilled in the art according to the actual situation.

[0109] It should be noted that the present invention can be deployed on the device itself to achieve embedded applications, or can also run on a PC or other terminals with a user interface, so as to meet various hardware environments and usage requirements.

[0110] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that the computer can access, or a data storage device such as a server or data center that includes one or more collections of available media. The available media can be magnetic media (such as floppy disks, hard disks, magnetic tapes), optical media (such as DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0111] Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, devices, and modules described above can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0112] In several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling, direct coupling, or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of the devices or modules can be in electrical, mechanical, or other forms.

[0113] The modules described as separate components may or may not be physically separated. The components shown as modules may or may not be physical modules. They can be located in one place or distributed to multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0114] In addition, in each embodiment of the present application, the functional modules can be integrated in a processing module, or each module can exist physically alone, or two or more modules can be integrated in one module.

[0115] If the above functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs.

[0116] As described above, it is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present application, and all of them should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0117] Finally, the above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. An intelligent urban management information visualization integrated management method, characterized in that It includes the following steps: S1. Obtain the multi-modal time-series data stream of the urban management system, extract event features respectively, and map the event features to a multi-dimensional feature space; Among them, the multi-modal time-series data stream includes video surveillance data, sensor monitoring data, and text work order data; S2. Construct an initial event association chain based on the cross-modal time-series synchronization and co-occurrence frequency of event features in the multi-dimensional feature space; S3. Screen out the interference event chains with insufficient causal relationship but synchronization in the initial event association chain to generate a purified association chain; S4. Conduct causal consistency verification on the purified association chain; S5. Adjust the spatio-temporal weight parameters and coverage density association factors of the purified association chain, and generate dynamic correction weights by combining the feedback priorities of text work order data; S6. Integrate the dynamic correction weights and the purified association chain passed through causal consistency verification, construct a cross-modal causal chain network, and generate a visual guidance plan.

2. The intelligent urban management information visualization integrated management method according to claim 1, wherein Obtain the multi-modal time-series data stream of the urban management system, extract event features respectively, and map the event features to a multi-dimensional feature space, including: Obtain video surveillance data and extract the spatial distribution features in video frames through a pre-trained convolutional neural network; Obtain sensor monitoring data and extract the time-series fluctuation features of sensor values through a recurrent neural network; Obtain text work order data and extract the semantic event label features in work order texts through a natural language processing model; Perform dimensionality reduction processing on the spatial distribution features, time-series fluctuation features, and semantic event label features respectively, map them to a multi-dimensional feature space with the same dimension, and adjust the directions of each feature vector so that the calculation results of the cosine similarity between different modal feature vectors are comparable.

3. The intelligent urban management information visualization integrated management method according to claim 1, characterized in that Construct an initial event association chain based on the cross-modal time-series synchronization and co-occurrence frequency of event features in the multi-dimensional feature space, including: Based on the time-series alignment relationship of the spatial distribution feature sequence of video surveillance data, the time-series fluctuation features of sensor monitoring data, and the semantic event label features in the multi-dimensional feature space, calculate the cosine similarity of different modal feature vectors within the same time window as the cross-modal time-series synchronization index; Statistical co-occurrence frequency of different modal feature vectors with occurrence probability exceeding a preset threshold within the same geographical area. The co-occurrence frequency is calculated according to the ratio of the number of times of simultaneous occurrence of multi-modal events to the total number of events in historical data; Perform weighted fusion on the cross-modal time-series synchronization index and the co-occurrence frequency to generate an initial event association chain. The initial event association chain includes the association strength value and association direction between event nodes, and the association direction is determined by the chronological order of feature vectors within the time window.

4. The intelligent urban management information visualization integrated management method according to claim 1, wherein Screen out the interference event chains with insufficient causal relationship but synchronization in the initial event association chain to generate a purified association chain, including: Calculate the causal strength value between event nodes in the initial event association chain through a causal strength evaluation method based on time-series prediction error. The causal strength value is a quantitative index of the influence degree of the predecessor event node on the successor event node; Verify the causal relationship between event nodes based on counterfactual analysis. If the decrease amplitude of the occurrence probability of the predecessor event node after the successor event node is intervened and eliminated is lower than the preset threshold, the corresponding association chain is determined as an interference event chain; Generate filtering rules based on the causal strength value and the counterfactual analysis result, filter out the interfering event chains with causal strength values lower than the preset causal strength threshold or those that do not meet the conditions in the counterfactual analysis, and retain the remaining associated chains as the purified associated chains.

5. The intelligent urban management information visualization comprehensive management method according to claim 1, characterized in that Perform causal consistency verification on the purified associated chains, including: Verify whether the logical trigger order of sub-events in the purified associated chains conforms to the preset urban management business rules, where the preset urban management business rules include the logical causal constraint conditions between event types and the trigger time interval range; Generate counterfactual time-series data corresponding to the interfering event chains, simulate the causal relationship failure scenario through the time-series data reconstruction method, and extract the cross-modal abnormal window caused by the causal relationship failure; Calculate the temporal dependence intensity attenuation rate between the temporal fluctuation characteristics of sensor monitoring data and the sequence of spatial distribution characteristics of video surveillance data within the cross-modal abnormal window; If the logical trigger order conforms to the preset urban management business rules and the temporal dependence intensity attenuation rate is within the preset threshold tolerance range, it is determined that the causal consistency verification passes.

6. The intelligent urban management information visualization integrated management method according to claim 1, wherein, Adjust the spatio-temporal weight parameters and coverage density correlation factors of the purified associated chains, and generate dynamic correction weights in combination with the feedback priority of text work order data, including: Calculate the time decay factor and spatial proximity based on the occurrence time and geographical location coding of event nodes in the purified associated chains to generate spatio-temporal weight parameters; Statistically analyze the event type distribution density within the same geographical area in historical data to generate a coverage density correlation factor, where the event type distribution density is the ratio of the number of occurrences of an event type within a preset time period to the area of the region; Extract the feedback priority from the semantic event label features of text work order data, and the feedback priority is calculated based on the weighted average of the classification score of the urgency marked in the work order text and the historical processing time limit; Perform conditional rule fusion on the spatio-temporal weight parameters, coverage density correlation factors, and feedback priority to generate dynamic correction weights.

7. The intelligent urban management information visualization integrated management method according to claim 6, wherein, The time decay factor is calculated by exponential decay according to the interval between the event occurrence time and the current time, and the spatial proximity is calculated based on the geographical location coding distance between event nodes.

8. The intelligent urban management information visualization integrated management method according to claim 6, characterized in that The conditional rule for conditional rule fusion is: when the feedback priority reaches the preset emergency threshold, the dynamic correction weight is determined by the product of the spatio-temporal weight parameter and the coverage density correlation factor; when the feedback priority does not reach the emergency threshold, the dynamic correction weight is determined by the ratio of the coverage density correlation factor to the feedback priority.

9. The intelligent urban management information visualization integrated management method according to claim 1, wherein, Integrate the dynamic correction weights and the purified associated chains that have passed the causal consistency verification, and construct a cross-modal causal chain network and generate a visual guidance plan, including: Perform weight mapping of nodes and edges on the dynamic correction weights and the verified purified associated chains to construct a cross-modal causal chain network with a directed weighted graph structure, where the nodes represent event types, the edges represent causal relationships, and the weight value is the dynamic correction weight; Extract the critical path based on the cross-modal causal chain network, and the screening condition for the critical path is that the total path weight exceeds the preset display threshold and the path length meets the minimum number of event chain nodes requirement; Overlay and render the critical path on the urban geographic information system map to generate a multi-dimensional interactive visual guidance plan; Output a visualization guidance plan and synchronously generate a cross-departmental task priority list, where the task priorities are arranged in descending order according to the weight values of the event nodes in the critical path. For event nodes with the same weight value, the execution order is determined by topological sorting.

10. An intelligent urban management information visualization integrated management system for implementing the intelligent urban management information visualization integrated management method according to any one of claims 1-9, characterized in that It includes the following modules: Data acquisition module: used to obtain the multi-modal time-series data stream of the urban management system, extract event features respectively, and map the event features to a multi-dimensional feature space; Association construction module: used to construct an initial event association chain based on the cross-modal time-series synchronization and co-occurrence frequency of event features in the multi-dimensional feature space; Interference screening module: used to screen out the interference event chains with sufficient synchronization but insufficient causal relationship in the initial event association chain to generate a purified association chain; Causal verification module: used to verify the causal consistency of the purified association chain; Weight adjustment module: used to adjust the spatio-temporal weight parameters and coverage density association factors of the purified association chain, and generate a dynamically corrected weight in combination with the feedback priority of the text work order data; Solution generation module: used to integrate the dynamically corrected weight and the purified association chain passing the causal consistency verification, construct a cross-modal causal chain network and generate a visualization guidance plan.

Images