Rail transit task processing method and device based on multi-data source processing

Abstract

The application discloses a rail transit task disposal method and device based on multi-data source processing, which comprises the following steps: extracting key semantics in a task, associating and expanding the key semantics according to different semantic relationships to obtain associated semantics; obtaining a multi-source data set associated with the associated semantics, calculating the association measurement value between the multi-source data sets, and screening the multi-semantic multi-source data according to the association measurement value; constructing a semantic data association structure tree based on the multi-source data and according to the association measurement value; dividing the task level according to the associated semantics, and determining the disposal decision scheme of the corresponding level according to the task level; simulating the task through digital twin simulation technology according to the semantic data association structure tree and the disposal decision scheme, and determining the optimal disposal decision scheme according to the simulation result. The application solves the problem that it is difficult to select a suitable disposal decision scheme according to a task scene in a rail transit task.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method and apparatus for multi-task-oriented multi-source data aggregation. Background Technology

[0002] In recent years, with the emergence of concepts such as digital twins and full lifecycle management, the processing of multi-source data and its hidden multi-faceted relationships have become a focus of public attention. To leverage the potential efficiency of multi-source data and improve processing capabilities for various tasks, numerous in-depth studies have been conducted focusing on the intrinsic relationships and evolutionary mechanisms of multi-source data. The rapid development of new technologies, particularly artificial intelligence, big data, cloud computing, and geographic information systems, has provided insights into solving perception and analysis challenges for specific tasks.

[0003] Task-oriented multi-source datasets primarily involve the analysis and processing of data within a limited time and space scope, based on the task content. They can provide decision-making for the current task based on historical processing experience and current data conditions, with the core decision-making process revolving around task-aggregated data. Task-oriented data, from a source perspective, includes spatiotemporal foundational data, IoT sensing data, and business application data. It requires aggregating this vast amount of data, analyzing and judging it according to task requirements, and ultimately proposing a decision-making process. Task-oriented data often comes from diverse sources, is massive in volume, varied in type, and diverse in format, which affects the speed of task processing. Therefore, there is an urgent need to propose a task-oriented multi-source data aggregation and processing mechanism that aggregates relevant data according to task requirements and provides auxiliary processing decisions based on those requirements, thereby improving the emergency handling capability of tasks. Summary of the Invention

[0004] This invention provides a method and apparatus for handling rail transit tasks based on multi-data source processing, which solves the problem of difficulty in selecting appropriate handling decision schemes according to task scenarios in rail transit tasks.

[0005] A method for handling rail transit tasks based on multi-data source processing includes:

[0006] Extract key semantics from the task, and extend these key semantics according to different semantic relationships to obtain associated semantics;

[0007] Obtain multi-source datasets associated with the semantic association, calculate the association metric between the multi-source datasets, and filter out multi-source data with multiple semantic associations based on the association metric.

[0008] Based on the multi-source data, a semantic data association structure tree is constructed according to the association metric value;

[0009] Based on the associated semantics, the task levels are divided, and the corresponding handling decision schemes are determined based on the task levels.

[0010] The task is simulated using digital twin simulation technology based on the semantic data association structure tree and the disposal decision scheme, and the optimal disposal decision scheme is determined based on the simulation results.

[0011] In one embodiment of the present invention, the key semantics in the extraction task are extended and associated according to different semantic relationships to obtain associated semantics. Specifically, this includes: determining the dataset included in the task, wherein the dataset includes text data; extracting the key semantics contained in the text data by means of keyword extraction or machine learning feature extraction; determining different semantic relationships by means of semantic information association dictionary; and extending and associated according to the different semantic relationships.

[0012] In one embodiment of the present invention, calculating the association metric between the multi-source datasets specifically includes: determining a preset minimum support; wherein, support is the frequency of an itemset appearing in the multi-source dataset, each row of records in the multi-source dataset corresponds to a transaction, the elements in each transaction are called items, and an itemset is a set containing one or more items; if it contains k items, it is called a k-itemset; S1: scanning the multi-source dataset to obtain a first candidate data itemset, the first candidate data itemset being a set of 1-itemsets; filtering the first candidate data itemset to obtain frequent 1-itemsets that satisfy the minimum support; S2: scanning the multi-source dataset to determine a set of 2-itemsets containing the frequent 1-itemsets as a second candidate data itemset; filtering the second candidate data itemset to obtain frequent 2-itemsets that satisfy the minimum support; and cyclically scanning the multi-source dataset according to the above scanning steps S1 to S2 until frequent n-itemsets are obtained; wherein, frequent n+1 itemsets that satisfy the minimum support cannot be generated from the frequent n-itemsets.

[0013] In one embodiment of the present invention, constructing a semantic data association structure tree based on the association metric specifically includes: the association metric includes support and confidence, wherein the confidence is the ratio of the number of transactions containing itemset {Y,X} to the number of transactions containing itemset {Y} or itemset {X}; constructing a semantic data association structure tree based on the support and the confidence; the semantic data association structure tree includes a semantic aggregation node layer and a data operation node layer; the semantic aggregation node layer includes semantic information nodes associated with the task, and the relationship between each semantic information node corresponds to the actual application relationship; the data operation node layer includes one or more data operation nodes associated with the task and the corresponding semantics.

[0014] In one embodiment of the present invention, the simulation using digital twin simulation technology based on the semantic data association structure tree and the disposal decision specifically includes: the task is a rail transit task, and the disposal decision includes the disposal process in the experience base and expert disposal experience; acquiring multi-dimensional operational data of the rail transit task, constructing a digital twin model based on the multi-dimensional operational data; and simulating the task disposal decision scheme for the current task scenario of the rail transit task based on the digital twin model, the multi-dimensional operational data, the disposal process, the expert disposal experience, and the aggregated semantics and data obtained through the semantic data association structure tree.

[0015] In one embodiment of the present invention, determining the optimal disposal decision scheme based on simulation results specifically includes: determining the measurement dimensions of the disposal decision scheme, the measurement dimensions including operation time, economic benefits, and safety performance; determining the weight of each measurement dimension, the sum of the weights of each measurement dimension being equal to 1; calculating the priority of each disposal decision scheme based on the weights of each measurement dimension; and determining the optimal disposal decision scheme based on the priority.

[0016] In one embodiment of the present invention, the semantic relationships include causal relationships, sequential relationships, concurrent relationships, and inclusion relationships; the multi-source datasets include spatiotemporal basic data, IoT sensing data, business application data, and operational evaluation data.

[0017] A device for handling rail transit tasks based on multi-data source processing, comprising:

[0018] The semantic information acquisition module is used to extract key semantics from the task and extend the key semantics according to different semantic relationships to obtain associated semantics.

[0019] The filtering module is used to obtain multi-source datasets associated with the semantic association, calculate the association metric between the multi-source datasets, and filter out multi-source data with multiple semantic associations based on the association metric.

[0020] A semantic data association structure tree construction module is used to construct a semantic data association structure tree based on the multi-source data and the association metric value.

[0021] The task level classification module is used to classify tasks according to the associated semantics and determine the corresponding handling decision scheme according to the task level.

[0022] The simulation module is used to simulate the task based on the semantic data association structure tree and the disposal decision scheme using digital twin simulation technology, and to determine the optimal disposal decision scheme based on the simulation results.

[0023] A device for handling rail transit tasks based on multi-data source processing, comprising:

[0024] At least one processor; and,

[0025] The memory is communicatively connected to the at least one processor via a bus; wherein,

[0026] The memory stores instructions executable by the at least one processor, which are executed to perform:

[0027] Extract key semantics from the task, and extend these key semantics according to different semantic relationships to obtain associated semantics;

[0028] Obtain multi-source datasets associated with the semantic association, calculate the association metric between the multi-source datasets, and filter out multi-source data with multiple semantic associations based on the association metric.

[0029] Based on the multi-source data, a semantic data association structure tree is constructed according to the association metric value;

[0030] Based on the associated semantics, the task levels are divided, and the corresponding handling decision schemes are determined based on the task levels.

[0031] The task is simulated using digital twin simulation technology based on the semantic data association structure tree and the disposal decision scheme, and the optimal disposal decision scheme is determined based on the simulation results.

[0032] A non-volatile storage medium storing computer-executable instructions, which are executed by a processor to perform the following steps:

[0033] Extract key semantics from the task, and extend these key semantics according to different semantic relationships to obtain associated semantics;

[0034] Obtain multi-source datasets associated with the semantic association, calculate the association metric between the multi-source datasets, and filter out multi-source data with multiple semantic associations based on the association metric.

[0035] Based on the multi-source data, a semantic data association structure tree is constructed according to the association metric value;

[0036] Based on the associated semantics, the task levels are divided, and the corresponding handling decision schemes are determined based on the task levels.

[0037] The task is simulated using digital twin simulation technology based on the semantic data association structure tree and the disposal decision scheme, and the optimal disposal decision scheme is determined based on the simulation results.

[0038] This invention provides a method and apparatus for handling rail transit tasks based on multi-data source processing, which includes at least the following beneficial effects: The method of this invention yields a suitable handling decision scheme, enabling priority control of the task site through sensors, controllers, and other components. Personnel quickly arrive at the scene to execute the decisions, and a dual response mechanism of sensor control and manual intervention ensures rapid response and accurate task execution. By employing a multi-data source aggregation strategy, relevant data is gathered according to task requirements, and auxiliary handling decision schemes are provided based on these requirements, thereby improving the emergency handling capability of the task. Attached Figure Description

[0039] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings:

[0040] Figure 1 A schematic diagram illustrating the steps of a method for handling rail transit tasks based on multi-data source processing, provided in an embodiment of the present invention;

[0041] Figure 2 This is a flowchart illustrating the multi-source data aggregation method for rail transit tasks provided in an embodiment of the present invention.

[0042] Figure 3 This is a flowchart of the multi-source data information aggregation process provided in an embodiment of the present invention;

[0043] Figure 4 A schematic diagram illustrating the construction of a semantic data structure tree provided in an embodiment of the present invention;

[0044] Figure 5 A schematic diagram of a rail transit task processing device based on multi-data source processing provided in an embodiment of the present invention;

[0045] Figure 6 This is a schematic diagram of a rail transit task processing device based on multi-data source processing, provided as an embodiment of the present invention. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described clearly and completely below in conjunction with specific embodiments. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0047] It should be noted that those skilled in the art will understand, explicitly and implicitly, that the embodiments described in this invention can be combined with other embodiments without conflict. Unless otherwise defined, the technical or scientific terms used in this invention should be understood in their ordinary sense by those skilled in the art. The terms "a," "an," "an," "the," etc., used in this invention do not indicate quantity limitation and can represent singular or plural. The terms "comprising," "including," "having," and any variations thereof used in this invention are intended to cover non-exclusive inclusion; the terms "first," "second," "third," etc., used in this invention are merely to distinguish similar objects and do not represent a specific ordering of objects.

[0048] This invention proposes a method and apparatus for handling rail transit tasks based on multi-data source processing. Utilizing data analysis methods, it combines task and scenario understanding with data analysis to study a multi-source data analysis and aggregation method for rail transit tasks. Combined with digital twin simulation technology, it provides auxiliary decision-making for practical application scenarios. The details are described below.

[0049] Figure 1 A schematic diagram illustrating the steps of a rail transit task handling method based on multi-data source processing provided in this embodiment of the invention may include the following steps:

[0050] S110: Extract key semantics from the task, and expand the key semantics according to different semantic relationships to obtain related semantics.

[0051] In one embodiment of the present invention, the overall process is as follows: Figure 2 As shown, the key semantics in the task are extracted, and the key semantics are extended according to different semantic relationships to obtain associated semantics. Specifically, this includes: determining the dataset included in the task, which includes text data; extracting the key semantics contained in the text data by extracting keywords or machine learning features; determining different semantic relationships by associating semantic information with a dictionary; and extending the key semantics according to the different semantic relationships.

[0052] Furthermore, key semantics in the task are extracted and further subdivided according to different semantic relationships. Specific relationships include:

[0053] A causal relationship means that the occurrence of semantic A will trigger semantic B.

[0054] Sequential relation means that semantic A will be generated in accordance with common sense, and semantic B will be generated in sequence.

[0055] Concurrency means that semantic A is generated at the same time as semantic B.

[0056] The inclusion relationship indicates that semantic A includes semantic B, semantic C, etc.

[0057] Furthermore, for a given rail transit task scenario, key semantics contained in the task are extracted by extracting keywords or using machine learning feature extraction.

[0058] For example, given the task of "ensuring operational safety under severe weather", the keywords "severe weather", "operations", and "safety" can be extracted by using keyword extraction or machine learning semantic extraction methods.

[0059] Specifically, based on the extracted key semantics, the system expands them according to different association relationships, and finds related semantic information by implementing the constructed rail transit semantic information association dictionary.

[0060] For example, semantic information related to "severe weather" includes storms, rain, snow, geographical and geological disasters; semantic information related to "operation" includes train dispatching, safe train operation, and large passenger flow events; and semantic information related to "safety" includes "detection of abnormal passenger behavior" and "detection of abnormal escalator scenarios".

[0061] S120: Obtain multi-source datasets associated with semantic association, calculate the association metric between multi-source datasets, and filter out multi-source data with multiple semantic associations based on the association metric.

[0062] Specifically, based on the extended association semantics, multi-source datasets associated with it are mined. These multi-source datasets include data from different sources such as spatiotemporal basic data, IoT sensing data, business application data, and operational evaluation data. The data from different sources include structured data and unstructured data.

[0063] For example, a storm may include wind speed level records at different time intervals over a period of time, wind information data collected by different devices, and warning data given by a monitoring and early warning system.

[0064] The identified multi-source datasets are used to determine the degree of association between the task and the data, and between different data sets, through association measurement algorithms such as the Apriori algorithm. By comparing the association measurement values ​​with pre-set thresholds, suitable data from different sources and in different formats are selected.

[0065] In one embodiment of the present invention, the association metric between multi-source datasets is calculated, and multi-semantic multi-source data is filtered based on the association metric. Specifically, this includes: determining a preset minimum support; wherein, support is the frequency of an itemset appearing in the multi-source dataset, each row of records in the multi-source dataset corresponds to a transaction, the elements in each transaction are called items, and an itemset is a set containing one or more items. If it contains k items, it is called a k-itemset; S1: scan the multi-source dataset to obtain the first candidate data itemset, which is a set of 1-itemsets; filter the first candidate data itemset to obtain frequent 1-itemsets that satisfy the minimum support; S2: scan the multi-source dataset to determine a set of 2-itemsets containing frequent 1-itemsets as the second candidate data itemset; filter the second candidate data itemset to obtain frequent 2-itemsets that satisfy the minimum support; and repeat the above scanning steps S1 to S2 to scan the multi-source dataset until frequent n-itemsets are obtained; wherein, frequent n+1 itemsets that satisfy the minimum support cannot be generated from frequent n-itemsets.

[0066] Specifically, multi-source datasets related to semantics represent data associations in different directions, which can increase the effectiveness of subsequent analysis to some extent. Therefore, it is necessary to select multi-source data that are closely related to the task.

[0067] like Figure 3 The diagram shows the multi-source data aggregation process. First, by scanning the pre-mined multi-semantic multi-source dataset, the support of each candidate data item in the multi-source dataset in the historical records of its respective data source is calculated. Then, a frequent 1-itemset L1 is generated based on the pre-set minimum support. Then, based on L1 and the data in the multi-source dataset, a frequent 2-itemset L2 that satisfies the pre-set minimum support and minimum confidence is generated. This operation is then repeated to generate a frequent n-itemset. This frequent n-itemset can no longer generate an (n+1)-itemset that satisfies the minimum support. Finally, the multi-source data closely related to the task is determined.

[0068] Support refers to the percentage of transactions in an itemset that contains both A and B out of all transactions in the dataset, denoted as S(AB) = support_count(AB) / B = P(AB); confidence refers to the percentage of itemsets containing A that also contain B, denoted as C(AB) = support_count(AB) / support_count(A) = P(B|A).

[0069] S130: Based on multi-source data, construct a semantic data association structure tree according to the association metric.

[0070] In one embodiment of the present invention, a semantic data association structure tree is constructed based on association metrics, specifically including: association metrics including support and confidence, wherein confidence is the ratio of the number of transactions containing itemset {Y,X} to the number of transactions containing itemset {Y} or itemset {X}; a semantic data association structure tree is constructed based on support and confidence; the semantic data association structure tree includes a semantic aggregation node layer and a data operation node layer; the semantic aggregation node layer includes semantic information nodes associated with the task, and the relationship between each semantic information node corresponds to the actual application relationship; the data operation node layer includes one or more data operation nodes associated with the task and the corresponding semantics.

[0071] Furthermore, a semantic data relationship structure tree is constructed based on the correlation metrics between tasks and data, and between different data sets. For example... Figure 4 The diagram illustrates the construction of a semantic data structure tree. The semantic data association structure tree includes a semantic aggregation node layer and a data operation node layer. The semantic aggregation node layer includes semantic information nodes associated with the task, and the relationships between these semantic information nodes correspond to the relationships in actual applications. The data operation node layer includes one or more data operation nodes associated with the task and its corresponding semantics. All nodes are associated with both the task and the semantics, and can be mapped to operations in actual applications.

[0072] Specifically, based on the selected multi-semantic, multi-source data, a semantic data association structure tree is constructed using support and confidence.

[0073] For example, the task node is "Ensure operational safety under severe weather". The first layer includes weather conditions such as "strong wind", "heavy rain", "heavy snow", "mudslide", and "landslide". In the second layer, "strong wind" includes "wind level", "wind speed", "direction", "duration", and the speed of vehicles traveling in different time periods. Based on the selected data and the calculated support and confidence of each set, a complete semantic data association structure tree is finally constructed.

[0074] S140: Based on the associated semantics, classify the task levels and determine the corresponding handling decision-making schemes based on the task levels.

[0075] Furthermore, based on the key and related semantics extracted from the task, the task is classified into levels, and corresponding handling decisions are given according to the expert experience and event handling database of the corresponding level.

[0076] Specifically, based on the key semantics and related semantics mined out, different levels of event handling are divided, and handling procedures and expert handling experience in the experience base of different levels are given.

[0077] For example, the severity of "strong winds" is divided into three levels: Level 1, Level 2, and Level 3, with Level 3 being the most severe. These levels correspond to different equipment and data in rail transit mission scenarios, as well as different handling methods. For instance, under Level 1 strong winds, sensors and other equipment detect low bridge vibration frequencies, high train speeds, and high passenger density inside the trains. Under Level 3 strong winds, sensors and other equipment detect high bridge vibration frequencies, low train speeds, and low passenger density inside the trains.

[0078] S150: Simulate the task using digital twin simulation technology based on the semantic data association structure tree and disposal decision scheme, and determine the optimal disposal decision scheme based on the simulation results.

[0079] In one embodiment of the present invention, simulation is performed using digital twin simulation technology based on a semantic data association structure tree and disposal decisions. Specifically, the task is a rail transit task, and the disposal decisions include disposal processes and expert disposal experience from an experience base. Multi-dimensional operational data of the rail transit task is acquired, and a digital twin model is constructed based on the multi-dimensional operational data. Based on the digital twin model, and according to the multi-dimensional operational data, disposal processes, expert disposal experience, and aggregated semantics and data derived from the semantic data association structure tree, the task disposal decision scheme for the current task scenario of the rail transit task is simulated.

[0080] Specifically, by combining multi-dimensional operational data of rail transit, a digital twin model is constructed. By utilizing multi-source data experimental simulation models and data collected on-site, simulation technology is used to simulate the operational status and data interaction of various equipment and infrastructure in the rail transit operation field. The specific details and technologies of model construction can be implemented using existing digital twin technologies.

[0081] In one embodiment of the present invention, determining the optimal disposal decision scheme based on simulation results specifically includes: determining the measurement dimensions of the disposal decision scheme, including operation time, economic benefits, and safety performance; determining the weight of each measurement dimension, wherein the sum of the weights of each measurement dimension is equal to 1; calculating the priority of each disposal decision scheme based on the weight of each measurement dimension; and determining the optimal disposal decision scheme based on the priority.

[0082] Furthermore, by combining the data aggregated in the constructed semantic data association structure tree and providing corresponding handling decision-making schemes based on expert experience and an event handling library, digital twin simulation technology is used to simulate the current task scenario online, incorporating historical digital twin data corresponding to multi-faceted operational data. This simulates the semantic data association structure tree and handling decision-making schemes, and based on the simulation results, prioritizes the schemes from three perspectives: operation time, economic benefits, and safety performance. The formula is as follows:

[0083] Score = αs (operation time) + βs (economic efficiency) + γs (safety performance)

[0084] The most suitable decision-making plan can be quickly determined based on this priority selection formula.

[0085] The operation time, economic benefits, and safety performance were all normalized. α, β, and γ are parameter values ​​obtained through experimental verification, where α+β+γ=1.

[0086] In one embodiment of the present invention, by using data from various sensors, controllers, and other auxiliary digital twins deployed at the mission site, corresponding measures can be provided before human intervention, ensuring the rapid execution and accuracy of the mission.

[0087] Based on the final appropriate handling decision, priority is given to controlling the mission site through components such as sensors and controllers. Personnel arrive at the scene quickly to execute the decision. The dual response mechanism of sensor control and manual intervention ensures rapid response and accuracy of mission execution.

[0088] For example, in severe weather conditions, based on the final decision-making process, the first step is to guide passenger flow through evacuation by controlling escalators, roller shutters, and public announcements to avoid panic. Simultaneously, staff are deployed to the scene to investigate problems and manage passenger flow, ensuring the task is completed quickly and accurately through this combined approach.

[0089] The above describes a method for handling rail transit tasks based on multi-data source processing, as provided in this invention. Based on the same inventive concept, this invention also provides a corresponding device for handling rail transit tasks based on multi-data source processing, such as… Figure 5 As shown.

[0090] The semantic information acquisition module 510 is used to extract key semantics from the task and extend the key semantics according to different semantic relationships to obtain associated semantics; the filtering module 520 is used to acquire multi-source datasets associated with associated semantics, calculate the association metric between multi-source datasets, and filter out multi-source data with multiple semantics based on the association metric; the semantic data association structure tree construction module 530 is used to construct a semantic data association structure tree based on multi-source data and the association metric; the task level classification module 540 is used to classify task levels according to associated semantics and determine the corresponding disposal decision scheme based on the task level; the simulation module 550 is used to simulate the task using digital twin simulation technology based on the semantic data association structure tree and the disposal decision scheme, and determine the optimal disposal decision scheme based on the simulation results.

[0091] This invention also provides a corresponding device for handling rail transit tasks based on multi-data source processing, such as... Figure 6 As shown, it includes:

[0092] The system includes at least one processor 610, a communication interface 620, a memory 630, and a communication bus 640; wherein the processor 640, the communication interface 620, and the memory 630 communicate with each other via the communication bus 640; the processor 640 can call logical instructions stored in the memory 630 to cause at least one processor 610 to execute:

[0093] Key semantics are extracted from the task, and these key semantics are extended according to different semantic relationships to obtain associated semantics. Multi-source datasets associated with the associated semantics are acquired, and the association metric between the multi-source datasets is calculated. Multi-source data with multiple semantics are selected based on the association metric. Based on the multi-source data, a semantic data association structure tree is constructed according to the association metric. Based on the associated semantics, task levels are divided, and corresponding disposal decision schemes are determined according to the task levels. The task is simulated using digital twin simulation technology based on the semantic data association structure tree and disposal decision schemes. The optimal disposal decision scheme is determined based on the simulation results.

[0094] Based on the same idea, some embodiments of the present invention also provide media corresponding to the above methods.

[0095] Some embodiments of the present invention provide a storage medium storing computer-executable instructions, which are executed by a processor to perform the following steps:

[0096] Key semantics are extracted from the task, and these key semantics are extended according to different semantic relationships to obtain associated semantics. Multi-source datasets associated with the associated semantics are acquired, and the association metric between the multi-source datasets is calculated. Multi-source data with multiple semantics are selected based on the association metric. Based on the multi-source data, a semantic data association structure tree is constructed according to the association metric. Based on the associated semantics, task levels are divided, and corresponding disposal decision schemes are determined according to the task levels. The task is simulated using digital twin simulation technology based on the semantic data association structure tree and disposal decision schemes. The optimal disposal decision scheme is determined based on the simulation results.

[0097] The various embodiments in this invention are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the device and medium embodiments are relatively simple in description because they are fundamentally similar to the method embodiments; relevant parts can be referred to the descriptions in the method embodiments.

[0098] The devices, media, and methods provided in the embodiments of the present invention are one-to-one correspondences. Therefore, the devices and media also have similar beneficial technical effects as their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be repeated here.

[0099] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process method or product that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process method or product. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process method or product that includes that element.

[0100] The above are merely embodiments of the present invention and are not intended to limit the invention. Although the present invention has been described in detail above with general descriptions and specific embodiments, modifications or improvements can be made to it, which will be obvious to those skilled in the art. Therefore, all such modifications or improvements made without departing from the spirit of the present invention fall within the scope of protection claimed by the present invention.

Claims

1. A method for handling rail transit tasks based on multi-data source processing, characterized in that, include: Extract key semantics from the task, and extend these key semantics according to different semantic relationships to obtain associated semantics; Obtain a multi-source dataset associated with the semantic association, wherein the multi-source dataset includes spatiotemporal basic data, IoT sensing data, business application data, and operational evaluation data; Calculate the correlation metric between the multi-source datasets, and filter out multi-semantic multi-source data based on the correlation metric. Based on the multi-source data, a semantic data association structure tree is constructed according to the association metrics, specifically including: the association metrics include support and confidence; the confidence is the ratio of the number of transactions containing itemset {Y,X} to the number of transactions containing itemset {Y}, or the ratio of the number of transactions containing itemset {Y,X} to the number of transactions containing itemset {X}; support is the frequency of an itemset appearing in the multi-source dataset, where each row of records in the multi-source dataset corresponds to a transaction, and the elements in each transaction are called items; an itemset is a set containing one or more items; a semantic data association structure tree is constructed according to the support and the confidence; the semantic data association structure tree includes a semantic aggregation node layer and a data operation node layer; the semantic aggregation node layer includes semantic information nodes associated with the task, and the relationship between each semantic information node corresponds to the actual application relationship; the data operation node layer includes one or more data operation nodes associated with the task and its corresponding semantics; Based on the associated semantics, the task levels are divided, and the corresponding handling decision schemes are determined based on the task levels. The task is simulated using digital twin simulation technology based on the semantic data association structure tree and the disposal decision scheme, and the optimal disposal decision scheme is determined based on the simulation results.

2. The method according to claim 1, characterized in that, The key semantics extracted in the task are further extended and associated according to different semantic relationships to obtain associated semantics, specifically including: Identify the dataset included in the task, wherein the dataset includes text data; Key semantic elements contained in the text data are extracted by keyword extraction or machine learning feature extraction. Different semantic relationships are determined by associating semantic information with a dictionary, and the key semantics are extended by association based on the different semantic relationships.

3. The method according to claim 1, characterized in that, The calculation of the correlation metric between the multi-source datasets, and the filtering of multi-semantic multi-source data based on the correlation metric, specifically includes: Determine the preset minimum support; An itemset containing k items is called a k-itemset. S1: Scan the multi-source dataset to obtain the first candidate data itemset, which is a set of 1 itemsets; The first candidate data itemset is filtered to obtain frequent 1-itemsets that satisfy the minimum support. S2: Scan the multi-source dataset to determine the set of 2-itemsets containing the frequent 1-itemsets, as the second candidate data itemset; The second candidate data itemset is filtered to obtain frequent 2-itemsets that satisfy the minimum support. The multi-source dataset is scanned cyclically according to the above scanning steps S1~S2 until frequent n-item sets are obtained; Specifically, the frequent n-itemsets cannot generate a more frequent n+1-itemsets that satisfy the minimum support.

4. The method according to claim 1, characterized in that, The simulation, performed using digital twin simulation technology based on the semantic data association structure tree and the decision-making scheme, specifically includes: The task is a rail transit task, and the handling decision-making scheme includes the handling process in the experience base and the handling experience of experts. Acquire multi-dimensional operational data of rail transit tasks, and construct a digital twin model based on the multi-dimensional operational data; Based on the digital twin model, and according to the multi-dimensional operational data, the handling process, the expert handling experience, and the aggregated semantics and data obtained through the semantic data association structure tree, the task handling decision-making scheme for the current task scenario of the rail transit task is simulated.

5. The method according to claim 1, characterized in that, The process of determining the optimal handling decision based on simulation results specifically includes: Determine the measurement dimensions for the disposal decision-making plan, including operation time, economic benefits, and safety performance; Determine the weights of each measurement dimension, wherein the sum of the weights of each measurement dimension is equal to 1; The priority of each decision-making option is calculated based on the weight of each of the aforementioned measurement dimensions; The optimal handling decision is determined based on the priorities mentioned above.

6. The method according to claim 1, characterized in that, The semantic relationships include causal relationships, sequential relationships, concurrent relationships, and inclusion relationships; The multi-source dataset includes spatiotemporal basic data, IoT sensing data, business application data, and operational evaluation data.

7. A device for handling rail transit tasks based on multi-data source processing, characterized in that, include: The semantic information acquisition module is used to extract key semantics from the task and extend the key semantics according to different semantic relationships to obtain associated semantics. The filtering module is used to obtain multi-source datasets associated with the semantic association, wherein the multi-source datasets include spatiotemporal basic data, IoT sensing data, business application data, and operational evaluation data; Calculate the correlation metric between the multi-source datasets, and filter out multi-semantic multi-source data based on the correlation metric. A semantic data association structure tree construction module is used to construct a semantic data association structure tree based on the multi-source data and the association metric, specifically including: the association metric includes support and confidence, wherein the confidence is the ratio of the number of transactions containing itemset {Y,X} to the number of transactions containing itemset {Y}, or the ratio of the number of transactions containing itemset {Y,X} to the number of transactions containing itemset {X}; support is the frequency of an itemset appearing in the multi-source dataset, where each row of records in the multi-source dataset corresponds to a transaction, and the elements in each transaction are called items, and an itemset is a set containing one or more items; a semantic data association structure tree is constructed based on the support and the confidence; the semantic data association structure tree includes a semantic aggregation node layer and a data operation node layer; the semantic aggregation node layer includes semantic information nodes associated with the task, and the relationship between each semantic information node corresponds to the actual application relationship; the data operation node layer includes one or more data operation nodes associated with the task and its corresponding semantics; The task level classification module is used to classify tasks according to the associated semantics and determine the corresponding handling decision scheme according to the task level. The simulation module is used to simulate the task based on the semantic data association structure tree and the disposal decision scheme using digital twin simulation technology, and to determine the optimal disposal decision scheme based on the simulation results.

8. A processing device for rail transit tasks based on multi-data source processing, characterized in that, include: At least one processor; as well as, The memory is communicatively connected to the at least one processor via a bus; wherein, The memory stores instructions executable by the at least one processor, which are executed to perform: Extract key semantics from the task, and extend these key semantics according to different semantic relationships to obtain associated semantics; Obtain a multi-source dataset associated with the semantic association, the multi-source dataset including spatiotemporal basic data, IoT sensing data, business application data and operation evaluation data; calculate the association metric between the multi-source datasets, and filter out the multi-source data with multiple semantics based on the association metric; Based on the multi-source data, a semantic data association structure tree is constructed according to the association metrics, specifically including: the association metrics include support and confidence; the confidence is the ratio of the number of transactions containing itemset {Y,X} to the number of transactions containing itemset {Y}, or the ratio of the number of transactions containing itemset {Y,X} to the number of transactions containing itemset {X}; support is the frequency of an itemset appearing in the multi-source dataset, where each row of records in the multi-source dataset corresponds to a transaction, and the elements in each transaction are called items; an itemset is a set containing one or more items; a semantic data association structure tree is constructed according to the support and the confidence; the semantic data association structure tree includes a semantic aggregation node layer and a data operation node layer; the semantic aggregation node layer includes semantic information nodes associated with the task, and the relationship between each semantic information node corresponds to the actual application relationship; the data operation node layer includes one or more data operation nodes associated with the task and its corresponding semantics; Based on the associated semantics, the task levels are divided, and the corresponding handling decision schemes are determined based on the task levels. The task is simulated using digital twin simulation technology based on the semantic data association structure tree and the disposal decision scheme, and the optimal disposal decision scheme is determined based on the simulation results.

9. A non-volatile storage medium storing computer-executable instructions, characterized in that, The computer-executable instructions are executed by the processor to perform the following steps: Extract key semantics from the task, and extend these key semantics according to different semantic relationships to obtain associated semantics; Obtain a multi-source dataset associated with the semantic association, wherein the multi-source dataset includes spatiotemporal basic data, IoT sensing data, business application data, and operational evaluation data; Calculate the correlation metric between the multi-source datasets, and filter out multi-semantic multi-source data based on the correlation metric. Based on the multi-source data, a semantic data association structure tree is constructed according to the association metrics, specifically including: the association metrics include support and confidence; the confidence is the ratio of the number of transactions containing itemset {Y,X} to the number of transactions containing itemset {Y}, or the ratio of the number of transactions containing itemset {Y,X} to the number of transactions containing itemset {X}; support is the frequency of an itemset appearing in the multi-source dataset, where each row of records in the multi-source dataset corresponds to a transaction, and the elements in each transaction are called items; an itemset is a set containing one or more items; a semantic data association structure tree is constructed according to the support and the confidence; the semantic data association structure tree includes a semantic aggregation node layer and a data operation node layer; the semantic aggregation node layer includes semantic information nodes associated with the task, and the relationship between each semantic information node corresponds to the actual application relationship; the data operation node layer includes one or more data operation nodes associated with the task and its corresponding semantics; Based on the associated semantics, the task levels are divided, and the corresponding handling decision schemes are determined based on the task levels. The task is simulated using digital twin simulation technology based on the semantic data association structure tree and the disposal decision scheme, and the optimal disposal decision scheme is determined based on the simulation results.

Images