Intelligent start-stop methods, systems, and programs for adaptive equipment status in subway stations

By using station operation data and knowledge graphs in subway stations to form dynamic station status diagrams, determining equipment start-up and shutdown procedures, and compensating for deviations in reverse, the adaptiveness problem of equipment start-up and shutdown control is solved, closed-loop control of equipment start-up and shutdown is realized, and the accuracy and adaptability of equipment control are improved.

CN123113404APending Publication Date: 2026-11-17BWTON TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611615225.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-19
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

The existing start-stop control of subway station equipment is difficult to adaptively adjust according to the actual operating effect after the equipment starts and stops, which affects the accuracy and adaptability of equipment control.

Method used

A dynamic station status diagram is generated through collaborative calculation based on station operation data and knowledge graph. The equipment to be controlled is identified and the equipment start-up and shutdown procedure is generated. After execution, the dynamic station status diagram is updated to compensate for deviations in reverse, thereby realizing closed-loop control of equipment start-up and shutdown.

Benefits of technology

It improves the adaptive capability of subway station equipment start-stop control to dynamic operating states, reduces invalid or repeated start-stop operations, and improves the matching degree between equipment control and the actual operating needs of the station.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123113404A_ABST
    Figure CN123113404A_ABST
Patent Text Reader

Abstract

This application provides an intelligent start-stop method, system, and program product for adaptive equipment status in subway stations. The key to this application's embodiments is not simply enabling subway equipment to "automatically start and stop," but rather ensuring the system understands whether the equipment's operation has produced the expected results. Even if the equipment has started, if the passenger flow or traffic conditions in the corresponding area still do not meet expectations, the system will identify this "unresolved status difference" and then search backwards for other equipment that can still function to compensate for it, rather than recalculating all station equipment or blindly adding equipment. In this way, equipment start-stop is transformed from a one-time rule-based control into a closed-loop control that can continuously correct deviations based on actual operational results.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of intelligent rail transit technology, specifically to an intelligent start-stop method, system, and computer program product for adaptive equipment status in subway stations. Background Technology

[0002] With the development of intelligent urban rail transit, subway stations are typically equipped with escalators, turnstiles, ventilation and air conditioning, lighting, and other equipment. The operational status of these devices is monitored and their start / stop control is achieved through integrated monitoring systems or intelligent station management platforms. Currently, the start / stop of equipment is usually executed based on operating hours, preset scenarios, or fixed control conditions.

[0003] However, in actual operation, the demand for equipment start-up and shutdown changes dynamically with passenger flow distribution, regional operating status, and equipment operation. Fixed-time or fixed-condition start-up and shutdown control cannot adapt to changes in station operating status in a timely manner. At the same time, after the equipment has started or stopped, its actual effect on the operating status of the relevant area may deviate from expectations.

[0004] Therefore, existing subway station equipment start-up and shutdown methods struggle to adaptively adjust subsequent control based on the actual operational effects after equipment start-up and shutdown, thus affecting the accuracy and adaptability of equipment control. Improving the adaptability of subway station equipment start-up and shutdown control to dynamic operating states and actual control effects is a technical problem that urgently needs to be solved by existing technologies. Summary of the Invention

[0005] One objective of this application is to improve the adaptive capability of subway station equipment start-stop control to dynamic operating states and actual control effects.

[0006] According to one aspect of the embodiments of this application, a smart start-stop method for adaptive equipment status in subway stations is disclosed, the method comprising: Station status collaborative calculation is performed based on station operation data and station operation knowledge graph to form a dynamic station status map, which represents at least one of the current operation status and future operation status; Based on the operating status of each state node in the dynamic station status diagram, and the interaction relationship between station equipment and state nodes and the equipment start-stop conditions in the station operation knowledge graph, the equipment to be controlled is determined and the corresponding equipment start-stop control equipment start-stop program is generated. Based on the action effect parameters corresponding to the equipment start-stop control in the station operation knowledge graph, the expected state changes of the state nodes that have an interaction with the equipment to be controlled are deduced based on the dynamic station status diagram after the equipment start-stop procedure is executed; The device start-up and shutdown procedure is executed, and the dynamic station status diagram is updated based on the station operation data generated during the execution process to obtain the actual status changes of the status nodes that have an interaction relationship with the device to be controlled. Based on the deviation between the actual state change and the expected state change, starting from the state node that generates the deviation, the device start-stop control for compensating the deviation is determined in reverse along the action relationship, and the device start-stop procedure is updated accordingly.

[0007] According to one aspect of the embodiments of this application, a computer system is disclosed, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method as described in any of the preceding claims.

[0008] According to one aspect of the embodiments of this application, a computer program product is disclosed, including a computer program that, when executed by a processor, implements the steps of the method as described in any of the preceding claims.

[0009] The key to this application's embodiments is not to make subway equipment "automatically turn on and off," but to ensure the system knows whether the expected effect has been achieved after the equipment is turned on or off. Even if the equipment has started, if the passenger flow or traffic conditions in the corresponding area still do not meet expectations, the system will identify this "unresolved state difference" and then look backwards from this location to find other equipment that can still function, providing targeted compensation, rather than recalculating all the equipment in the station or blindly adding equipment. In this way, equipment start-up and shutdown are transformed from one-time rule-based control into a closed-loop control that can continuously correct deviations based on actual operational results.

[0010] In other words, this application embodiment forms the current and future dynamic station status through collaborative calculation of station operation data and station operation knowledge graph, enabling the equipment start-up and shutdown criteria to reflect the actual operating status of the station. The determination of the controlled equipment is constrained by the interaction relationship between equipment and status nodes and the equipment start-up and shutdown conditions. Furthermore, the expected state changes after equipment start-up and shutdown are obtained in advance using action effect parameters, ensuring a clear correspondence between the controlled equipment and its expected effects. After equipment start-up and shutdown are executed, the dynamic station status graph is updated based on actual operating data to obtain the actual state changes. The deviation between the actual state changes and the expected state changes is used to determine compensation control in reverse along the equipment interaction relationships. Thus, equipment start-up and shutdown are transformed from one-time control based on fixed conditions to closed-loop control based on the station's dynamic status and continuous correction of actual control effects. This reduces ineffective or repetitive equipment start-up and shutdown, improves the matching degree between equipment control and the actual operating needs of the station, and enhances the adaptive capability under dynamic operating conditions.

[0011] Other features and advantages of this application will become apparent from the following detailed description, or may be learned in part from practice of this application.

[0012] It should be understood that the above general description and the following detailed description are merely exemplary and do not limit this application. Attached Figure Description

[0013] The above and other objectives, features and advantages of this application will become more apparent from a detailed description of exemplary embodiments thereof with reference to the accompanying drawings.

[0014] Figure 1 This is a flowchart of an intelligent start-stop method for adaptive operation of subway station equipment status, provided according to an exemplary embodiment.

[0015] Figure 2 It is based on Figure 1 The corresponding embodiment shows a flowchart describing the steps of performing station status collaborative calculations based on station operation data and station operation knowledge graphs to form a dynamic station status map.

[0016] Figure 3 It is based on Figure 1 The corresponding embodiment shows a flowchart describing the steps of determining the equipment to be controlled and generating the corresponding equipment start-stop control procedure based on the operating status of each state node in the dynamic station status diagram, the interaction relationship between station equipment and state nodes in the station operation knowledge graph, and the equipment start-stop conditions.

[0017] Figure 4 It is based on Figure 1 The corresponding embodiment shows a flowchart describing the steps of the expected state change of state nodes that have an interaction relationship with the controlled equipment after the execution of the equipment start-stop procedure based on the action effect parameters corresponding to the equipment start-stop control in the station operation knowledge graph and the dynamic station state diagram.

[0018] Figure 5 It is based on Figure 1 The corresponding embodiment shows a flowchart describing the steps of executing the equipment start-up and shutdown procedure, updating the dynamic station status diagram based on the station operation data generated during the execution process, and obtaining the actual state change steps of the state nodes that have an interaction relationship with the equipment to be controlled.

[0019] Figure 6 It is based on Figure 1 The corresponding embodiment shows a flowchart describing a method for determining equipment start-up and shutdown control to compensate for the deviation based on the deviation between the actual state change and the expected state change, starting from the state node that caused the deviation, and updating the equipment start-up and shutdown procedure accordingly.

[0020] Figure 7 It is based on Figure 6The flowchart of the method for determining the target compensation device for reducing the state compensation amount based on the matching relationship between the action effect parameters and the state compensation amount of each candidate compensation device in step S153 of the corresponding embodiment is shown.

[0021] Figure 8 It is based on Figure 6 The flowchart of the method for determining the target compensation device according to the degree of matching between the corresponding state action quantity and the state change quantity in step S203 of the corresponding embodiment is shown in the embodiment.

[0022] Figure 9 This is a schematic diagram illustrating a video surveillance service environment at a rail transit station according to an exemplary embodiment. Detailed Implementation

[0023] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided to make the description of this application more comprehensive and complete, and to fully convey the concept of the exemplary embodiments to those skilled in the art. The drawings are merely illustrative of this application and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted.

[0024] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more exemplary embodiments. Numerous specific details are provided in the following description to give a full understanding of exemplary embodiments of this application. However, those skilled in the art will recognize that the technical solutions of this application can be practiced with one or more of the specific details omitted, or other methods, components, steps, etc., can be employed. In other instances, well-known structures, methods, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of this application.

[0025] Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0026] Figure 1 This is a flowchart of a smart start-stop method for adaptive operation of subway station equipment states, provided according to an exemplary embodiment. The smart start-stop method for adaptive operation of subway station equipment states provided in this application embodiment includes: Step S110: Based on station operation data and station operation knowledge graph, perform station status collaborative calculation to form a dynamic station status map. The dynamic station status map represents at least one of the current operation status and future operation status. Step S120: Based on the operating status of each state node in the dynamic station status diagram, as well as the interaction relationship between station equipment and state nodes and the equipment start-stop conditions in the station operation knowledge graph, determine the equipment to be controlled and generate the corresponding equipment start-stop control equipment start-stop program. Step S130: Based on the action effect parameters corresponding to equipment start-stop control in the station operation knowledge graph, the expected state changes of state nodes that have an interaction with the equipment to be controlled are deduced based on the dynamic station status diagram after the equipment start-stop procedure is executed. Step S140: Execute the equipment start-up and shutdown procedure, and update the dynamic station status diagram based on the station operation data generated during the execution process to obtain the actual status changes of the status nodes that have an interaction with the equipment to be controlled. Step S150: Based on the deviation between the actual state change and the expected state change, starting from the state node that caused the deviation, determine the equipment start-stop control to compensate for the deviation in reverse along the action relationship, and update the equipment start-stop procedure accordingly.

[0027] These steps are explained in detail below.

[0028] In step S110, for example, the station operation data comes from data sources such as video monitoring, ticketing system, train operation, integrated monitoring, and equipment monitoring. The objects described may include station areas, passages, entrances and exits, platforms, turnstiles, escalators, and other station objects involved in station operation.

[0029] The system acquires station operation data for the current station status calculation period. This data is used to determine the operational status of station objects. For example, the described operational status includes, but is not limited to, the number of people in the area, passenger flow direction, passenger flow speed, inbound and outbound traffic volume, estimated train arrival and departure information, the operational status of equipment such as escalators and turnstiles, and environmental conditions. Further, station operation data reflects the changes in the status of space, passenger flow, and equipment within a subway station during operation. This data may include video monitoring data, ticketing data, train operation data, equipment operation data, and environmental monitoring data. For instance, video monitoring data reflects the passenger flow, direction, or speed of the flow control units in the concourse, platform, passageways, and escalator areas; ticketing data reflects inbound and outbound passenger flow; train operation data reflects train arrivals and departures and the resulting changes in passenger flow; and equipment operation data reflects the current operational, stopped, or other operational status of station equipment such as escalators, turnstiles, ventilation and air conditioning, and lighting.

[0030] The station operation knowledge graph pre-stores relatively stable station objects and their relationships during station operation. Corresponding to the aforementioned points, station objects include station halls, platforms, passageways, entrances / exits, equipment service areas, station equipment, and access interfaces connecting different areas. The graph relationships formed by station objects in the station operation knowledge graph include spatial connections, passenger flow transmission relationships, interaction relationships between equipment and status nodes, and equipment service scope relationships, etc., which will not be listed here in detail. It should be understood that the station operation knowledge graph also stores parameters such as basic equipment capabilities, basic throughput capacity of connections, passenger flow propagation paths, path travel time, and the action effects that equipment start / stop control can produce.

[0031] In an exemplary embodiment, station status collaborative computing does not directly use the raw data generated by a certain data source as the basis for starting and stopping the equipment. Instead, it associates the operation information from different data sources with the objects and their operation relationships in the station operation knowledge graph based on the station objects corresponding to the operation data, so as to obtain a reliable dynamic station status that can reflect the actual operation of the corresponding station objects.

[0032] Specifically, the execution process of step S110 may include: constructing operation observation knowledge elements for the corresponding observation objects based on the station operation data generated by each data source; the operation observation knowledge elements are used to represent an operation observation result formed by a data source for a specific station object, and enable the operation observation result to be associated with the corresponding observation object in the station operation knowledge graph.

[0033] After associating each operational observation knowledge element with the corresponding observation object, the cooperative state estimation is performed based on the operational observation knowledge elements obtained for the observation object and other observation objects that have operational state transmission relationships with the observation object, so as to obtain the reliable dynamic station state of the corresponding observation object.

[0034] For example, when determining the passenger flow status of a transfer passage, the system not only utilizes passenger flow data obtained from direct monitoring of the passage, but also combines the inflow status of upstream areas, the outflow status of downstream areas with passenger flow transmission relationships with the passage, and the operational status of related passage equipment to jointly determine the operational status of the passage. Therefore, the operational status of a certain node is not determined in isolation by a single sensor data point, but corresponds to the actual passenger flow propagation and equipment operation process within the station.

[0035] A dynamic station status diagram is formed based on the reliable dynamic station status of each observed object and the graph relationships between the observed objects. The state nodes in the dynamic station status diagram correspond to the corresponding station objects and carry the operating status of the station object at the corresponding station status time; state connections are established between different state nodes based on the operating relationships between the corresponding station objects.

[0036] In the scenario of starting and stopping equipment in a subway station, the operating status of a status node includes status attributes that reflect the equipment control requirements or the results of the equipment's operation. For example, spatial status nodes include passenger flow, inflow status, outflow status, passenger flow accumulation status, or capacity status; the status corresponding to passage connection includes passenger flow passage status or effective passage capacity; and equipment status nodes include the current operating status of the equipment.

[0037] For the reliable dynamic station state obtained at the current station state calculation time, the current operating state is formed; for the subsequent operating states that can be determined based on the current operating state and the operating state transmission relationship in the station operating knowledge graph, the future operating state at the future target station state time is formed.

[0038] Therefore, a dynamic station status diagram can represent both "the current actual operating status of the station" and "the expected operating status of the corresponding station as current passenger flow and equipment operation continue to develop." Subsequent equipment start-up and shutdown will be implemented based on existing equipment control needs, and will also be implemented in advance for future equipment control needs.

[0039] Through step S110, the station operation data is first transformed into a dynamic station state that is consistent with the relationship between station space, passenger flow, and equipment operation, thereby providing a unified state basis for subsequent equipment determination, equipment function deduction, and equipment actual effect evaluation.

[0040] See Figure 2 , Figure 2 It is based on Figure 1 The corresponding embodiment shows a flowchart describing the steps of performing station status collaborative calculations based on station operation data and station operation knowledge graphs to form a dynamic station status map.

[0041] The step S110 of this application embodiment, which involves station status collaborative calculation based on station operation data and station operation knowledge graph to form a dynamic station status map, includes: Step S111: Based on the station operation data generated by each data source, construct operation observation knowledge elements for the corresponding observation objects; Step S112: Associate the operation observation knowledge element with the corresponding observation object in the station operation knowledge graph, and perform collaborative state estimation based on the operation state transmission relationship between the operation observation knowledge element and the observation object in the station operation knowledge graph to obtain the credible dynamic station state of each observation object. Step S113: A dynamic station status map is formed based on the credible dynamic station status of each observation object and its map relationship.

[0042] As mentioned above, station operation data is generated from different data sources. The data generation location, data meaning, and collection cycle of each data source are different. Therefore, the embodiments of this application do not directly merge data from different sources into the operation status of the station object. Instead, they first form operation observation knowledge elements for the observed object.

[0043] Specifically, after obtaining a station operation data, the data source that generated the data and the station object that the data actually describes are determined, and observation attributes and corresponding observation values ​​that can characterize the operation of the station object are extracted from the station operation data, thereby constructing an operation observation knowledge element for the observation object.

[0044] The observed objects are those in the station operation knowledge graph that can carry operational status. These can be flow control units such as station halls, platforms, passageways, and entrances / exits, as well as the passage connections between different flow control units, and station equipment such as escalators and turnstiles. The observed attributes correspond to the dynamically changing operational attributes of the respective observed objects.

[0045] For example, if video monitoring equipment obtains the current number of people in a transfer channel, it constructs an operational observation knowledge element with that transfer channel as the observation object and the number of people as the observation attribute; cross-regional passenger flow detection results form a flow observation knowledge element for the corresponding passage connection; escalator operation status messages form an operational observation knowledge element with the escalator as the observation object and the equipment operation status as the observation attribute; subsequent passenger flow input reflected by AFC (Automatic Fare Collection) or train operation data forms an operational observation knowledge element for the corresponding passenger flow generation location or passenger flow propagation object.

[0046] As a result, station operation data that originally came from different sources and had different formats were converted into structured operation information that could at least clearly identify "who observed it, which object was observed, what attributes were observed, and what the observation results were."

[0047] Furthermore, multiple operational observation knowledge elements are associated with the same observed object within the same station status calculation cycle. These operational observation knowledge elements are preferably retained separately rather than being pre-merged during the generation stage. For example, for the same transfer channel, the first camera generates one people observation knowledge element, the second camera generates another, and upstream cross-line counting generates a corresponding inflow observation knowledge element. These operational observation knowledge elements describe the observation results from different sources for the same or related operational objects. Subsequently, step S112 combines their source information and the operational relationships between the objects for collaborative processing. This approach effectively avoids the instantaneous error of one data source directly overwriting other sources during the data generation stage, and retains the basis for subsequently determining which actual operational data supports a reliable dynamic station status.

[0048] In an exemplary embodiment, the operational observation knowledge element records the observation time and source object corresponding to the station operation data to distinguish between the "actual operation time described by the data" and the time when the system receives the data. For example, when different video devices, AFC systems, and equipment control systems send data sequentially due to different acquisition cycles or communication delays, it can be determined whether they belong to the same station state calculation process based on the actual observation time corresponding to each operational observation knowledge element, rather than simply processing the data according to the order in which it arrives at the system.

[0049] Furthermore, quality information characterizing the current availability of operational observation knowledge elements will be determined. This quality information reflects at least one of the following: the temporal validity of the operational data, information completeness, the reliability of the spatial correspondence between the observation results and the target object, the reliability of the data source, or the degree of consistency with other data sources.

[0050] The quality information is preferably not used directly to modify the original observation values ​​in the operational observation knowledge element, but is retained together with the observation values, and the degree of influence of the operational observation knowledge element on the final reliable dynamic station state is controlled when performing cooperative state estimation in step S112.

[0051] Therefore, even if two operational observation knowledge elements give different observation results for the same object, it is not necessary to select one of them in advance in step S111. Instead, the two can be provided as independent operational observation inputs to subsequent cooperative state estimation.

[0052] After obtaining the operational observation knowledge elements, step S112 associates the observed objects recorded in each operational observation knowledge element with the corresponding objects in the station operation knowledge graph. The association operation between the operational observation knowledge elements and the station objects does not merely save the operational observation knowledge elements as subordinate data of the objects, but rather enables the corresponding operational observation knowledge elements to participate in the calculation of the object's operational state and to form computational associations with the states of other objects along the operational state transfer relationship in the station operation knowledge graph.

[0053] Operational state transitivity is used to represent how changes in the operational state of one station object can affect the state of another station object according to the actual operation of the station. In the context of a subway station, operational state transitivity includes passenger flow transit relationships formed between adjacent areas, as well as constraints formed by operational conditions such as equipment status and capacity on changes in the state of related station objects.

[0054] For example, the current operational status of transfer corridor U3 is not only related to the number of passengers directly monitored by video feeds on U3, but also to the passenger flow input to U3 from the upstream area, the passenger flow output from U3 to the downstream area, and the operational status of equipment related to the corresponding passage connections. Therefore, when determining the operational status of U3, it is necessary to simultaneously query the operational observation knowledge elements directly associated with U3 and the operational status of other observation objects associated with U3 along the operational status transmission relationship.

[0055] The specific implementation of collaborative state estimation includes: for each observation object to be estimated, firstly, reading the operational observation knowledge elements associated with the observation object to form the direct observation basis for the object; then, querying the associated objects with operational state transmission relationships with the observation object according to the station operation knowledge graph, and reading the operational states that the associated objects have obtained or are being collaboratively calculated; and then, combining the direct observation results with the operational state transmission relationships between the associated objects to determine the operational state of the observation object at the current station state time.

[0056] The difference between cooperative state estimation and directly using sensor output values ​​is that it determines a set of states that can simultaneously adapt to the actual observation results of the station and the operational relationships between objects, rather than selecting a single sensor result for each object independently.

[0057] For example, for a certain passageway, video monitoring might indicate a low number of people, but upstream there's a sustained large inflow and downstream equipment is operating at low capacity. In this case, there might be a discrepancy between the number of people determined by direct video monitoring and the actual flow of passengers. Therefore, the passageway's operational status should be determined based on both direct observation and the consistency between upstream and downstream conditions, rather than solely on a low number from a single video feed. Conversely, when a direct observation is supported by upstream inflow, downstream outflow, and the operational status of adjacent areas, the acceptance level of the operational status represented by that observation is increased.

[0058] The operational status of the observed object obtained through the above collaborative calculation is the reliable dynamic station status. The term "reliable" does not mean that the status is free of any errors, but rather that the status is not a direct copy of the measurement results from a single data source. Instead, it is formed under the joint constraints of the actual observation of the corresponding station object and the operational relationships with other objects. Therefore, it is more suitable as a unified status input for subsequent equipment start-up and shutdown judgments, equipment function deductions, and equipment function effect evaluations.

[0059] In the adaptive start-stop scenario addressed in this application's embodiments, this processing has further advantages. For example, when the channel status used to determine whether an escalator needs to be started comes from multiple sources of operational information, it reduces the possibility of erroneous start-stop caused by a single video short-term obstruction, an abnormal flow detection, or device data delay. Simultaneously, when the device recalculates the same state object after actual execution, the same collaborative state estimation mechanism can be used to evaluate the actual state changes generated by the device, ensuring that the state before start-stop, the projected state, and the actual state after start-stop have consistent state meanings.

[0060] Furthermore, in some embodiments, before the running observation knowledge elements participate in cooperative state estimation, temporal and spatial correspondence processing is performed on the running observation knowledge elements from different sources based on their actual observation time and corresponding observation objects.

[0061] For example, for data that has been recorded by AFC but the corresponding passenger flow has not yet actually reached the downstream channel, the observation objects that this part of the passenger flow can affect can be determined according to the passenger flow transmission relationship in the station operation knowledge graph, so that it can participate in the state calculation of the corresponding object at the corresponding operation time, instead of immediately writing the current number of people in the downstream area when the AFC data is generated.

[0062] This distinguishes between "passenger flow that has been generated but has not yet spread to the target area" and "passenger flow that already exists in the target area," and further provides a basis for the formation of future operational status.

[0063] Furthermore, when multiple operational observation knowledge elements jointly describe the same state attribute, the relative degree of influence of each operational observation knowledge element on the collaborative state estimation is determined based on the quality information obtained in step S111. Operational observation knowledge elements with higher quality exert stronger constraints on the final state; for operational observation knowledge elements with significantly insufficient quality or that clearly conflict with other high-quality observations and station operation relationships, their influence on the state results is reduced.

[0064] Furthermore, to ensure that the credible dynamic station state conforms to the actual operation of the subway station, the following different types of operational constraints can be used simultaneously during the collaborative state estimation process: First, observation constraints are formed on the corresponding state values ​​based on operational observation knowledge elements, so that the state estimation results are consistent with direct observations with a high degree of credibility; Second, state evolution constraints are formed based on the state changes of the observed object between different station states, so that the changes in regional passenger flow state are consistent with the inflow and outflow within the corresponding time period; Third, graph relationship constraints are formed based on the spatial connections, passenger flow transmission, and equipment operating conditions in the station operation knowledge graph, so that passenger flow only propagates along the actually available operating paths, and the corresponding state changes are restricted by the current connection state and equipment operating state.

[0065] The aforementioned different constraints do not generate multiple independent states separately, but rather participate together in determining the state of the corresponding observed object, thereby obtaining a reliable dynamic station state that can adapt to both operational observations and station operation patterns.

[0066] For observation objects that currently lack direct operational observations, the estimated state of the object is obtained based on its previous reliable dynamic station state and the states of associated objects with operational state transmission relationships. In this way, when a camera or other single sensing device lacks data for a short period of time, the corresponding state node does not have to immediately lose its operational state, and the subsequent device start-up and shutdown control does not have to stop due to a short-term interruption of a single data source.

[0067] It should be noted that the aforementioned quality participation, spatiotemporal correspondence, and graph relationship constraints are all used to enhance the specific implementation of "cooperative state estimation based on operational observation knowledge elements and operational state transmission relationships" in step S112. Under different station data conditions and control requirements, one or more of these implementation methods can be selected.

[0068] In step S113, after obtaining the credible dynamic station status of each observation object, the credible dynamic station status is matched with the corresponding observation object, and organized into a dynamic station status diagram according to the graph relationship between each observation object in the station operation knowledge graph.

[0069] Among them, the state nodes in the dynamic station status diagram are used to carry the credible dynamic station status of the corresponding observed object at the corresponding station status time; the state connection relationship between the state nodes is established according to the station operation map relationship between the corresponding observed objects.

[0070] For example, transfer passage U3 corresponds to a status node, which carries the current passenger flow status of U3; the passage U2 and the area in front of the escalator U4 adjacent to U3 form corresponding status nodes, and establish corresponding status connections according to their passenger flow transmission relationship in the station operation knowledge graph; escalator E1 forms a corresponding equipment status node, and maintains a graph relationship with the passage position it serves.

[0071] Therefore, the dynamic station status diagram is not a new state data structure that is independent of the station operation knowledge graph. Instead, it organizes the credible dynamic station status of the corresponding objects at specific operating times based on the existing objects and relationships in the station operation knowledge graph, thereby obtaining a state diagram that can continuously change with the actual operation of the station.

[0072] The credible dynamic station state formed at the current station state time is used as the current operating state of the corresponding state node. Furthermore, for the subsequent states determined based on the current operating state and the operating state transmission relationship between the observed objects, the future operating state is formed according to the corresponding future target station state time.

[0073] For example, if upstream AFC and train operation information indicate that passengers will enter the transfer channel along a specific path, and the corresponding passage equipment is currently stopped, the future operating status of the channel at the target time can be determined based on the passenger flow relationship and the current equipment operating status. Subsequent step S120 can then use both the current and future operating statuses to determine whether equipment related to this status node needs to be started or stopped.

[0074] The operational observation knowledge element retains independent observations from different data sources for the corresponding station objects. The operational status transfer relationship places the observations that were originally scattered across different objects under the actual passenger flow and equipment operation process of the station for collaborative judgment. The dynamic station status diagram organizes the finally obtained credible status into a state basis for subsequent equipment start-up and shutdown, expected effect deduction and actual effect evaluation.

[0075] In the adaptive start-stop process of the equipment in this application, the dynamic station state diagram formed in step S110 is not only used to determine whether the equipment should start at once. Step S120 generates equipment start-stop control based on the reliable dynamic station state. After the equipment executes the control, the new station operation data generated in step S140 is reconstructed through operation observation knowledge element construction and collaborative state estimation to form a new reliable dynamic station state. Therefore, the same station state formation mechanism is used simultaneously for state judgment before equipment control and actual effect reconstruction after equipment control. This allows subsequent actual state changes to be compared with the expected state changes deduced in step S130 based on the corresponding state objects and state attributes, providing a state consistency basis for further equipment determination based on state deviation in step S150.

[0076] In step S120, the dynamic station status diagram is used to determine whether the corresponding operating status needs to be changed through station equipment based on the current and / or future operating status of the status nodes.

[0077] In this embodiment, the station operation knowledge graph establishes the interaction relationships between station equipment and state nodes. The interaction relationship is used to indicate which state nodes(s) a certain station equipment can affect when its start / stop state changes, and can further indicate what state attributes of the corresponding state nodes it can affect.

[0078] To further explain, the interaction relationship is a knowledge graph used to clarify the correspondence between station equipment and the state nodes affected by its start-stop control, and further associates it with the corresponding state attributes. For the same station equipment, there can be interaction relationships where it acts on different state nodes or different state attributes; for the same state node, there can also be interaction relationships where multiple station devices can act on it separately. The interaction relationship is used to determine the object and attribute of the equipment's start-stop action, and is distinct from the operational state transmission relationship between state nodes used for state deduction; the specific direction and magnitude of the equipment's start-stop action are further characterized by the corresponding action effect parameters.

[0079] In specific execution, step S120 first uses the state node that needs to be adjusted as the query entry point, and determines the associated station equipment that can affect the corresponding state node and state attributes based on the interaction relationship; step S130 then determines the state node or state connection relationship directly affected by the equipment to be controlled based on the interaction relationship, so as to apply the corresponding graph parameter changes according to the action effect parameters; step S150 then uses the state node that generates the state deviation and its corresponding state attribute as the entry point, and queries the candidate station equipment that can be compensated for it in reverse along the same interaction relationship. Thus, the same interaction relationship participates in both the determination of the initial equipment start / stop object and the determination of the compensation equipment after the execution feedback.

[0080] For example, escalators establish interaction relationships with the status nodes corresponding to their service passages or passenger flow areas before and after the escalator, and the escalator can change the passenger flow capacity of the corresponding passage location after it is started; turnstiles establish interaction relationships with the status nodes corresponding to the entry and exit boundaries; ventilation and air conditioning equipment establish interaction relationships with the environmental status nodes of its service area.

[0081] Therefore, the relationship between devices and status nodes is not merely a spatial attribution relationship of "the device is located in a certain area", but rather a further operational relationship that can be used for device control calculations.

[0082] For the status nodes in the dynamic station status diagram, read their current operating status and / or future operating status, and determine the associated station equipment that can act on the corresponding status node based on the interaction relationship.

[0083] Subsequently, the current and / or future operating states of the status nodes are matched with the equipment start / stop conditions corresponding to the associated station equipment. Equipment start / stop conditions specify under what station operating states the corresponding equipment needs to be started or stopped. For example, for an escalator affecting the throughput of a passageway, its start condition is related to the current or future passenger flow status of that passageway; for currently operating equipment, the corresponding stop condition is met when the operating status in the corresponding area continuously decreases to the point where the equipment no longer needs to be operated.

[0084] It should be noted that equipment start-up and shutdown conditions are state-determination conditions associated with the corresponding station equipment, based on the operating state attributes of the state node it operates on, to determine whether there is a need to start or stop. Equipment start-up and shutdown conditions include corresponding start-up and stop-stop conditions. The judgment object of these conditions is at least one of the current operating state and future operating state of the corresponding state node in the dynamic station status diagram, and each associated station equipment matches its own corresponding start-up and shutdown conditions. Therefore, even if multiple station devices can act on the same state node, it does not mean that these station devices need to be started or stopped simultaneously.

[0085] In specific execution, step S122 reads the start / stop conditions of each associated station device, as well as the status nodes and status attributes involved in these conditions, and obtains the corresponding operating status in the dynamic station status diagram for matching. For example, for an escalator serving a certain passage, when the current or future passenger flow demand of the passage exceeds the existing effective throughput capacity, the start condition of the corresponding escalator can be met; when the operating demand of the relevant area continuously decreases to the point where it is no longer necessary to keep the device running, the corresponding shutdown condition can be met. The above condition matching is first used to determine the target device with start or shutdown requirements; step S123 also needs to compare the current device status of the target device with the required device status, excluding devices that do not need to be repeatedly started and stopped, in order to determine the controllable devices that actually need to be controlled by starting or stopping.

[0086] After identifying the target device that meets the start-up or shutdown conditions, the current device status in the dynamic station status diagram is further read. If the current device status of the target device is already consistent with the device status required by the device start-up or shutdown conditions, there is no need to generate repeated start-up or shutdown; if the two are inconsistent, the target device is identified as the device to be controlled, and the corresponding start-up control or shutdown control is generated.

[0087] Therefore, whether a station device is identified as a device to be controlled is subject to at least three common constraints: whether the corresponding state node has a device function requirement, whether the station device can function on the corresponding state node, and whether the station device currently has a need for a start-stop state switch.

[0088] An equipment start-stop procedure is formed based on the equipment start-stop control corresponding to each piece of equipment to be controlled. The equipment start-stop procedure may include one or more equipment start-stop controls, which are used to switch the corresponding equipment to be controlled from its current equipment state to an equipment state that is adapted to the current or future station operating state.

[0089] Through step S120, the dynamic station status obtained in step S110 is no longer only used for displaying the operating status, but is further converted into specific device start-stop control according to the interaction relationship between the device and the status node. That is to say, step S110 determines "which states are changing or about to change", and step S120 further determines "which devices can respond to these state changes and whether start-stop is required".

[0090] See Figure 3 , Figure 3 It is based on Figure 1 The corresponding embodiment shows a flowchart describing the steps of determining the equipment to be controlled and generating the corresponding equipment start-stop control procedure based on the operating status of each state node in the dynamic station status diagram, the interaction relationship between station equipment and state nodes in the station operation knowledge graph, and the equipment start-stop conditions.

[0091] The embodiment of this application provides a step S120 for determining the equipment to be controlled and generating a corresponding equipment start-stop control procedure based on the operating status of each state node in the dynamic station status diagram, the interaction relationship between station equipment and state nodes in the station operation knowledge graph, and the equipment start-stop conditions. This step includes: Step S121: Based on the current and / or future operating status of each status node in the dynamic station status diagram, determine the station equipment associated with the corresponding status node along the interaction relationship; Step S122: Match the current and / or future operating status of the status node with the equipment start / stop conditions of the associated station equipment to determine the target equipment that meets the start or stop conditions. Step S123: Based on the current equipment status of the target equipment in the dynamic station status diagram, determine the equipment to be controlled that needs to be started or stopped from the target equipment. Step S124: Generate device start / stop program according to the start control or stop control corresponding to each device to be controlled.

[0092] Based on the aforementioned dynamic station status diagram, for status nodes that indicate a need for equipment adjustment in the current and / or future operating states, a query is performed in the station operation knowledge graph along the interaction relationship between the station equipment and the status node to obtain the station equipment that can affect the corresponding status node.

[0093] Therefore, this step first determines the operating status that needs to be changed by the equipment and its corresponding status node, and then uses this status node as the entry point for equipment query to determine the associated station equipment, rather than judging whether each piece of equipment in the entire station needs to be started or stopped.

[0094] For example, when the future operating status of a certain traffic status node indicates that its passenger flow demand will increase after continuing to develop according to the current operating status, the node can be used as the query entry point to obtain related station equipment such as escalators and turnstiles that can change the passenger flow status at that location.

[0095] A single state node can correspond to multiple associated station devices, and different devices can also act on different state attributes of the same state node. Therefore, the associated station devices obtained in step S121 only indicate that the corresponding devices have the ability to change the operating state of the state node. Whether they need to be actually started or stopped within the current control cycle still needs to be further determined based on the start and stop conditions of the corresponding devices.

[0096] In step S122, for each associated station device obtained in step S121, its corresponding device start / stop conditions are read, and the state nodes and state attributes involved in the device start / stop conditions are determined. Then, the current operating state and / or future operating state of the corresponding state node are obtained from the dynamic station status diagram and matched with the device start / stop conditions. When the corresponding operating state meets the start condition of an associated station device, the device is identified as a target device with start requirements; when the corresponding operating state meets the deactivation condition, the device is identified as a target device with deactivation requirements.

[0097] Here, each associated station device is matched according to its own corresponding start-stop conditions. Therefore, even if multiple devices can act on the same state node, different start-stop judgment results may be obtained due to differences in device capabilities, usage methods, or control conditions.

[0098] For example, if escalators E1 and E2 are associated with the same access status node, when the future passenger flow demand of this status node increases, the future operating status will be matched with the respective start conditions of E1 and E2, instead of starting all associated devices simultaneously just because the status node reaches a certain unified threshold.

[0099] Therefore, step S121 first uses the interaction relationship to limit "which devices have the ability to adjust the state", and step S122 then further limits "which devices need to change the operating state in the current operating state" through the start and stop conditions corresponding to each device.

[0100] In step S123, the target equipment obtained in step S122 indicates that a start-up or shutdown requirement has been generated from the perspective of station operation status. However, whether actual equipment start-up / shutdown control needs to be generated still needs to be determined based on the current equipment status of the target equipment. For each target equipment, its current equipment status is read from the dynamic station status diagram, and this current equipment status is compared with the start-up or shutdown requirement determined in step S122. When the target equipment meets the start-up conditions and is currently in an operating state that does not require repeated start-up, the same start-up control is not generated; when it is currently in a stopped state and needs to be started, it is identified as the equipment to be controlled by executing start-up control. Correspondingly, when the target equipment meets the shutdown conditions and is still in an operating state, it is identified as the equipment to be controlled by executing shutdown control; for target equipment that is already in a stopped state, shutdown control is not generated again.

[0101] In summary, through steps S121 to S123, the scope of equipment converges sequentially from associated station equipment that can act on the corresponding state nodes to target equipment that meets the equipment start-up and stop conditions, and then to the controllable equipment that currently needs to undergo equipment state switching. This step-by-step convergence process ensures that the equipment start-up and stop determination is simultaneously constrained by the state application location, the equipment's own start-up and stop applicable conditions, and the equipment's current actual state.

[0102] In step S124, based on the devices to be controlled determined in step S123 and the corresponding state transitions that need to occur, a corresponding start control or stop control is generated, and the start control and / or stop control form the device start-stop procedure for this round. When there is only one device to be controlled, the device start-stop procedure is formed by the device start-stop control corresponding to that device; when there are multiple devices to be controlled, the device start-stop controls corresponding to each device to be controlled can be jointly written into the device start-stop procedure. When forming the device start-stop procedure, the correspondence between the device to be controlled, the device start-stop control, and the state node that triggers the device start-stop control is maintained, so that the subsequent step S130 can read its action effect parameters according to the corresponding device start-stop control and deduce the expected state change for the state node that has an interaction with the device to be controlled.

[0103] Therefore, this embodiment of the application, through steps S121 to S124, progressively converts the current and / or future operational needs in the dynamic station status diagram into directly executable equipment start-stop controls. This process is particularly suitable for scenarios where multiple devices in the same area of ​​a subway station can jointly or separately affect passenger flow operation. For the same state node, it is not necessary to uniformly start or stop all related devices simply because their operational status reaches a certain condition. Instead, based on the actual interaction between each device and the state node, the start-stop conditions corresponding to each device, and the current device status, the devices that truly need to undergo state switching can be determined step by step, thereby matching the scope of equipment start-stop with the specific operational status.

[0104] In step S130, after the equipment start-up and shutdown procedure is formed, before evaluating the actual execution effect of the equipment start-up and shutdown procedure, it is first determined what state changes should occur in the relevant state nodes when the corresponding equipment start-up and shutdown control takes effect as expected.

[0105] The station operation knowledge graph stores the action effect parameters corresponding to equipment start-stop control. These action effect parameters are used to establish the correspondence between equipment start-stop control and changes in station operation status, that is, to describe what effect the corresponding station equipment is expected to have on the state nodes that have a relationship with it after switching from its current state.

[0106] To further explain, the action effect parameter is a description of the expected state action associated with the specific start-up or stop control of the equipment. It is used to characterize the anticipated impact of the equipment's start-up / stop state switch as a change in graphical parameters that can be applied to the simulation station state diagram. The action effect parameter can characterize the object of action, the attribute of action, the direction of action, and the magnitude of action. Specifically, the object of action corresponds to the state node or state connection relationship with the equipment; the attribute of action is the expected affected operating state attribute; the direction of action indicates whether the attribute is expected to increase or decrease; and the magnitude of action indicates the expected degree of change. For example, the start-up control of an escalator corresponds to an increase in the effective throughput capacity of the connection it serves, while the stop control corresponds to a decrease in the corresponding effective throughput capacity.

[0107] In specific execution, step S132 first determines the state node or state connection relationship directly affected by the start / stop control based on the interaction relationship between the controlled device and the state node. Then, it reads the action effect parameters corresponding to the start / stop control and determines the graph parameter changes to be applied to the simulation station state diagram based on the action attributes, action direction, and action amount. Step S133 further performs state simulation based on the operating state transmission relationship between state nodes to obtain the expected state changes of the relevant state nodes. Therefore, the action effect parameters are the input of the device start / stop effect simulation and are not equivalent to the expected state changes obtained through state simulation. In addition, in step S153, the expected state action direction and state action amount of the candidate compensation device are determined based on the action effect parameters corresponding to the candidate compensation device, and matched with the required state change direction and state change amount of the state node to be compensated to determine the target compensation device.

[0108] For example, for escalator start control, the action effect parameters can describe the increase in the effective passage capacity of the corresponding passage position after the escalator starts; for equipment shutdown control, the action effect parameters can describe the decrease in the corresponding state attribute.

[0109] Therefore, the action effect parameters are not only used to describe the change of equipment from "stop" to "run" or from "run" to "stop", but also to further characterize the state effect that the change of equipment state is expected to bring to the station operation process.

[0110] Specifically, a simulation station status diagram is generated based on the current dynamic station status diagram for the current equipment start-up and shutdown procedure state simulation. Then, according to the action effect parameters corresponding to the equipment start-up and shutdown control, the equipment start-up and shutdown control is converted into graph parameter changes of the corresponding state nodes and / or state connection relationships in the simulation station status diagram. For example, escalator startup corresponds to a change in the effective throughput capacity of the relevant passage state connection, and turnstile startup corresponds to a change in the passage capacity of the corresponding entry and exit connection.

[0111] After the graph parameters change, state deduction is further performed according to the state transmission relationship between state nodes in the station operation knowledge graph. This allows the state changes of the directly affected locations of the equipment to continue to participate in subsequent state calculations, resulting in the deduced state of the state nodes affected by the equipment's start-stop control. The deduced state is then compared with the operating state of the corresponding state node before the equipment start-stop procedure is executed to obtain the expected state changes of the corresponding state node.

[0112] It should be noted that the expected state change determined in step S130 corresponds to the state node in the dynamic station status diagram of step S110. Therefore, the state node used to determine the equipment in step S120 also serves as the carrier of the equipment's action result in step S130. In this way, the equipment control requirements and the expected action result of the equipment are unified into the same station status object, rather than using two separate sets of control indicators and effect indicators.

[0113] See Figure 4 , Figure 4 It is based on Figure 1 The corresponding embodiment shows a flowchart describing the steps of the expected state change of state nodes that have an interaction relationship with the controlled equipment after the execution of the equipment start-stop procedure based on the action effect parameters corresponding to the equipment start-stop control in the station operation knowledge graph and the dynamic station state diagram.

[0114] The embodiment of this application provides a step S130 for deducing the expected state change of state nodes that have an interaction relationship with the controlled equipment after the execution of the equipment start-up and stop procedure based on the action effect parameters corresponding to equipment start-up and stop control in the station operation knowledge graph and the dynamic station status diagram, including: Step S131: Copy the dynamic station status diagram to form the projected station status diagram; Step S132: Based on the action effect parameters corresponding to the equipment start-stop control, convert the equipment start-stop control into graphical parameter changes of the corresponding state nodes and / or state connection relationships in the simulation station status diagram; Step S133: Based on the changes in graph parameters, perform state deduction on the station state diagram according to the state transmission relationship between state nodes in the station operation knowledge graph to obtain the deduced state of the state nodes affected by the changes in graph parameters. Step S134: Determine the expected state change based on the difference between the simulated state and the operating state of the corresponding state node before executing the equipment start-up and shutdown procedure.

[0115] In step S131, based on the equipment start-up and shutdown procedure already generated in the aforementioned embodiment, the dynamic station status diagram before the execution of the equipment start-up and shutdown procedure is used as the initial state of this round of state deduction, and a deduction station status diagram is formed by copying it.

[0116] Copying does not involve creating a new, independent station object model. Instead, it ensures that the state nodes and state connections in the deduced station status diagram are consistent with the corresponding objects in the current dynamic station status diagram, and uses the running state of each state node before the equipment start-up and shutdown procedure is executed as the initial state for the deduction.

[0117] Therefore, the state actions applied to equipment start-stop control in the simulated station status diagram do not directly modify the dynamic station status diagram used to represent the actual operation of the station. When the equipment has not yet actually been executed, the actual dynamic station status remains as it is formed by the actual operating data of the station, while the state changes expected to occur by equipment control only happen in the simulated station status diagram.

[0118] For example, in the dynamic station status diagram, escalator E1 is currently in a stopped state, its corresponding passage connection has a current effective throughput capacity, and the upstream and downstream of this connection correspond to state nodes U3 and U4, respectively. After copying, the deduced station status diagram maintains the correspondence between E1, U3, U4 and the corresponding state connections, while using the current effective throughput capacity and current passenger flow status as the starting state for the deduced diagram.

[0119] Therefore, step S131 actually establishes two state carriers with different purposes for the same station operating state: the dynamic station status diagram continues to represent the actual operating state, while the deduced station status diagram is used to carry the hypothetical effects of the equipment start-up and shutdown procedures before they are actually executed. This will enable subsequent steps S132 and S133 to determine the expected station operating impact of the equipment start-up and shutdown procedures without changing the actual operating state, and will also preserve an independent actual state benchmark for step S140 to update the dynamic station status diagram using actual operating data.

[0120] In step S132, for each device start / stop control formed in step S124, the corresponding action effect parameters are read from the station operation knowledge graph. These action effect parameters describe the expected operational effects on state nodes or state connections that have a relationship with the corresponding device after it switches between start and stop states. In this embodiment, the action effect parameters are primarily used to complete the transition from "device state change" to "station operation parameter change," rather than simply changing the device node itself from a stopped state to a running state, or vice versa.

[0121] Specifically, based on the interaction between the controlled equipment and the state nodes, the state nodes and / or state connection relationships directly affected by the start-stop control of the corresponding equipment are first determined. Then, based on the action effect parameters corresponding to the start-stop control of the equipment, the graphical parameters that need to be changed are determined, and the corresponding parameter changes are applied to the simulation station state diagram. For example, for escalators that can provide passenger flow transport capacity, start control can be converted into an increase in the effective throughput capacity of the corresponding passage state connection, and stop control can be converted into a decrease in the corresponding effective throughput capacity. For turnstile equipment, starting a certain number of turnstiles can be converted into an increase in the available throughput capacity of the corresponding entry and exit connection. For other equipment that affects the area state, their start-stop control can be converted into changes in the operating parameters of the corresponding state nodes affected by the equipment.

[0122] Therefore, the graph parameter changes mentioned in this step refer to the changes in node status parameters or connection status parameters generated in the station status diagram after the equipment start-stop control is actually applied to the station operation process.

[0123] The action effect parameters further indicate the target, attribute, direction, and magnitude of the changes in the graph parameters, enabling different devices to generate different graph parameter changes based on their respective positions and capabilities, even if they all execute "start control".

[0124] For example, if there are escalators E1 and E2 in the same transfer area, although both can improve passenger flow capacity, their start control is converted into changes in the effective throughput capacity of their respective corresponding state connections because they serve different passage connections or have different equipment capabilities, rather than adding an abstract "equipment capacity value" to the entire transfer area.

[0125] Therefore, the discrete equipment control in the equipment start-up and shutdown procedure is further converted into continuous operating parameter changes that can participate in the station state calculation. This conversion is also the key connection between step S120 and step S133: step S120 determines "which equipment needs to be started or stopped", and step S132 further determines "which operating parameter in the station operation diagram is expected to be changed by the start-up or shutdown of the equipment and by how much".

[0126] The changes in the graph parameters generated in step S132 first reflect the impact of equipment start-stop control on the directly affected position. However, in the actual operation of the subway station, this impact will usually further change other related state nodes through the passenger flow process.

[0127] Therefore, step S133 uses the changes in graph parameters obtained in step S132 as the input for equipment control, and performs state deduction on the station state diagram based on the state transmission relationship between state nodes in the station operation knowledge graph.

[0128] Specifically, firstly, based on the corresponding action relationships of equipment start-stop control, the state nodes or state connections directly affected by changes in the graphical parameters are determined. Then, along the state transmission relationship connected to them, the subsequently affected state nodes are determined, and the relevant states are updated according to the changes in the previous state node or state connection. For example, after escalator E1 is started, step S132 first increases the effective throughput capacity of the corresponding passage connection between U3 and U4. This change in effective throughput capacity will further affect the passenger flow that U3 can output to U4, thereby affecting the passenger flow accumulation state of U3 and the inflow state of U4. Therefore, the state deduction does not stop at the direct effect of "increased throughput capacity," but continues to calculate the corresponding changes in operating state along the passenger flow state transmission relationship between U3 and U4. For another example, when the shutdown of a set of turnstiles causes a decrease in the effective throughput capacity of a certain entrance connection, the passenger flow passage state on that connection changes, which can further affect the queuing or accumulation state of the upstream area of ​​the connection.

[0129] Therefore, in the scenario of starting and stopping subway station equipment, the equipment often directly changes a certain local operating condition, while the system ultimately needs to determine the continuous impact of this change in local conditions on the operating state of the relevant station area. Thus, this step extends the local effect of the equipment to the corresponding state changes during station operation through state transit relationships.

[0130] In one specific embodiment, starting from the position where the graph parameters change, the state nodes with which they have a transmission relationship are updated level by level according to the state transmission direction. For state nodes that are not affected by the graph parameter changes, their original operating state can be maintained or they can continue to be updated according to the original operating process, thereby obtaining the inferred station state after the equipment start-stop procedure is implemented. After completing the state inference, the inferred state corresponding to one or more state nodes affected by the equipment start-stop procedure is obtained.

[0131] It should be noted that the equipment start-up and shutdown procedure can include multiple equipment start-up and shutdown controls. When different equipment start-up and shutdown controls change different state nodes or state connection relationships, the corresponding graph parameter changes can be applied to the simulation station state diagram, and a unified state simulation can be performed based on the transmission relationship between state nodes, thereby obtaining the simulation state under the condition of multiple devices acting together.

[0132] The processing procedure in step S133 is particularly applicable to scenarios in subway stations where the "equipment's operational location" and the "final operational location to be improved" are not entirely the same. For example, increasing the throughput capacity of equipment at a certain connection location may not ultimately aim to simply increase the capacity value of that connection, but rather to reduce passenger flow accumulation in its upstream area. Through the state transmission on the diagram, this type of indirect effect can also be included in the expected calculation of the equipment control effect.

[0133] In step S134, the projected states of the corresponding state nodes are compared with the operating states of the corresponding state nodes when the dynamic station status diagram is copied in step S131. For each affected state node, the difference in its state attributes before and after the equipment start-up and shutdown procedure can be determined, and the difference can be used as the expected state change of the corresponding state node.

[0134] For example, before the equipment start-up and shutdown procedure is executed, the effective throughput capacity corresponding to a certain passage connection is the first capacity state; after the changes in the graph parameters of the equipment start-up control and the subsequent state deduction, its deduced state corresponds to the second capacity state. Then, the change of the second capacity state relative to the first capacity state forms the expected state change of this state attribute.

[0135] Furthermore, if the change in this capability causes changes in the degree of passenger flow accumulation, outflow status, etc. of the upstream state nodes, the difference between the projected state and the running state before program execution can be determined for the corresponding state attributes.

[0136] An expected state change corresponds to at least a clearly defined state node and state attribute. For operating states with numerical attributes, this is represented by the increase or decrease in the state value; for operating attributes that can be represented by other state forms, this is represented by the change in the corresponding state before and after the execution of the device start-up and shutdown procedure.

[0137] Therefore, the expected state change is not an abstract judgment of whether the equipment control effect is "effective" or "ineffective", but a clear indication of which state node and which operating state is expected to undergo what kind of change under the condition that the equipment start-up and shutdown procedure is working normally in this round.

[0138] This process directly defines the subsequent step S140. After the equipment is actually executed, step S140 obtains the actual state changes from the real station operation data based on the corresponding state nodes and state attributes determined in step S134; thus, the expected state changes and the actual state changes have a consistent comparison object.

[0139] Furthermore, when the actual state change is inconsistent with the expected state change, step S150 can also directly use the state node that produces the difference as the query entry point for reverse determination of the compensation device.

[0140] Therefore, step S130 is not only used to predict the effect of the equipment before the equipment is executed, but also to determine in advance the state object and state attributes to be targeted by the subsequent actual effect evaluation and compensation control, so that S130, S140 and S150 form a continuous state closed loop.

[0141] In summary, this embodiment, through steps S131 to S134, converts the equipment start-up and shutdown procedure obtained in step S120 into an expected state action that can be calculated and compared within a dynamic station status system. In the subway station scenario, this process ensures that "what effect the equipment should theoretically produce after it has been started" is no longer judged solely based on whether the equipment has completed a state switch or its rated capacity, but rather determined based on the equipment's position, actual action parameters, and the transmission relationship between related states. This provides a comparable expected benchmark for subsequently evaluating the actual effect of the equipment based on real station operation data and implementing compensation for insufficient effects.

[0142] In step S140, after obtaining the expected state change of the corresponding state node, the equipment start-up and shutdown procedure is executed, causing the equipment to be controlled to perform the corresponding start-up control or shutdown control. During the execution of the equipment start-up and shutdown procedure, the corresponding station operation data continues to be acquired. The station operation data can include not only the data generated by changes in the operating state of the equipment itself, but also the operation data such as passenger flow, passage, or environment corresponding to the state nodes that have an interaction with the equipment to be controlled.

[0143] This application does not directly use the completion of equipment state switching as a judgment result that the equipment control has achieved the expected effect. This is because, during the actual operation of a subway station, even if the equipment itself has completed startup or shutdown, the actual effect of the corresponding equipment on passenger flow or other station operation status may still be affected by the current passenger flow distribution, the status of adjacent areas, and the operation of other equipment. Therefore, the station operation data generated during the execution of the equipment startup and shutdown procedure is reused for station status collaborative calculation.

[0144] Specifically, based on the station operation data generated during the execution process, corresponding operation observation knowledge elements are constructed, and the new operation observation knowledge elements are associated with the corresponding observation objects in the station operation knowledge graph. Then, based on the operation observation knowledge elements and the operation state transmission relationship between the observation objects, the collaborative state estimation is performed again to obtain the reliable dynamic station state of the corresponding state nodes after the equipment start-up and shutdown procedure is executed, and the dynamic station state graph is updated accordingly.

[0145] Further comparison of the operating states of the corresponding state nodes before and after the execution of the equipment start-up and shutdown procedure yields the actual state changes of the state nodes. The actual state changes correspond to the same or comparable state nodes and state attributes as the expected state changes obtained in step S130. For example, if step S130 determines that the effective throughput capacity of a certain passage state node is expected to change after the escalator starts, then step S140 recalculates the state of that passage state node based on the operating data generated during the actual execution process, and obtains the actual second change. The second change is the actual state change that can be compared with the first change.

[0146] Therefore, this embodiment distinguishes between two different levels: "whether the equipment itself has completed start-up and shutdown" and "whether the start-up and shutdown of the equipment actually produces the expected station operation effect." The current state of the equipment can be used to confirm what kind of state change has occurred in the equipment itself, while the actual state change is used to evaluate what effect the change in the equipment state ultimately has on the relevant station operation state.

[0147] Meanwhile, since the expected state change in step S130 and the actual state change in step S140 are both based on the state node system formed in step S110, it avoids the difficulty of direct comparison due to the different objects or different state meanings used in the prediction results and actual feedback.

[0148] See Figure 5 , Figure 5 It is based on Figure 1 The corresponding embodiment shows a flowchart describing the steps of executing the equipment start-up and shutdown procedure, updating the dynamic station status diagram based on the station operation data generated during the execution process, and obtaining the actual state change steps of the state nodes that have an interaction relationship with the equipment to be controlled.

[0149] The step S140 provided in this application embodiment, which executes the equipment start-up and shutdown procedure and updates the dynamic station status diagram based on the station operation data generated during the execution process to obtain the actual state changes of the state nodes that have an interaction relationship with the equipment to be controlled, includes: Step S141: Construct corresponding operation observation knowledge elements based on the station operation data generated during the execution of the equipment start-up and shutdown procedure; Step S142: Based on the new operational observation knowledge elements and station operational knowledge graph, perform station status collaborative calculations to update the operational status of state nodes that have an interaction relationship with the equipment to be controlled; Step S143: Determine the actual state change based on the difference in the running state of the state nodes before and after the update.

[0150] In step S141, after the equipment start-up and shutdown procedure is executed, station operation data related to the equipment to be controlled and its functional state nodes are acquired, and new operation observation knowledge elements are formed according to the aforementioned construction method of operation observation knowledge elements. The specific construction process will not be elaborated here. The acquired station operation data can at least reflect the actual operation of the state nodes that have a functional relationship with the equipment to be controlled after the equipment start-up and shutdown procedure is executed.

[0151] Furthermore, based on the state nodes and state attributes corresponding to the expected state changes determined in step S130, the station operation data and corresponding observation objects required for this actual effect evaluation are determined, so that the actual effect evaluation object corresponds to the calculation object of the expected state changes.

[0152] In step S142, the new operational observation knowledge element is associated with the corresponding observation object, and the reliable dynamic station state of the state node that has an interaction relationship with the equipment to be controlled is re-determined after the equipment start-up and shutdown procedure is executed, in accordance with the aforementioned station state collaborative calculation method.

[0153] The key point of step S142 is that the actual effect after execution is not to directly adopt the original measurement results after the equipment execution, but to re-enter the newly generated station operation data into the same station state collaborative calculation system as before the equipment execution, thereby obtaining the reliable operation state after execution. Therefore, the operation state before and after the equipment execution has the same state object and state meaning, providing a unified state benchmark for subsequent direct calculation of actual state changes.

[0154] Furthermore, a new state version is generated for the trusted dynamic station state after execution, so as to retain the state results of the same state node before and after the execution of the device start-up and shutdown procedure.

[0155] For the status node updated in step S142, the running state after the execution of its device start / stop procedure is compared with the running state of the corresponding status node before execution to obtain the actual state change of the status node.

[0156] It should be understood that the actual state change corresponds to the same state node and state attribute as the expected state change obtained in step S130. For example, if step S130 determines that a device start-stop control is expected to change the effective throughput capacity of state node P1, then step S143 determines the actual change in effective throughput capacity based on the reliable dynamic station state before and after P1 execution, rather than using another state attribute to replace the actual change.

[0157] In step S150, the actual state change obtained in step S140 is compared with the expected state change obtained in step S130. If the actual state change reaches the corresponding expected state change, it indicates that the current round of equipment start-stop control has produced a state function that is in line with expectations; if there is a deviation between the actual state change and the expected state change, it indicates that the current round of equipment start-stop control still has a state function that has not reached the expected state function at the corresponding state node.

[0158] This embodiment does not re-perform an independent equipment start / stop judgment for all station equipment after the deviation occurs. Instead, it first identifies the state node that caused the deviation and uses this state node as the starting point for subsequent equipment determination. Specifically, based on the actual state change and the expected state change, it determines the state node with the state deviation to be compensated and the corresponding state compensation amount. The state compensation amount is used to represent the state change that still needs to be supplemented relative to the expected state change. Subsequently, using the state node to be compensated as the starting point for reverse lookup, it searches along the interaction relationships between station equipment and state nodes in the station operation knowledge graph to find candidate compensation equipment that can produce a corresponding state effect on the state node to be compensated.

[0159] The reverse determination referred to here is relative to the direction in which the device acts on the state node. In the device action relationship, the device determines the state node it can act on; however, after a state deviation occurs, the process starts from the state node that caused the deviation and searches backward along the same action relationship for the device that can act on that state node. In this way, the state deviation itself is not only used to determine whether the current device control effect has achieved the expected result, but also further limits the query location and query range of subsequent compensation devices.

[0160] Furthermore, based on the matching relationship between the action effect parameters and state compensation amount corresponding to each candidate compensation device, the target compensation device that can reduce the state compensation amount is determined, and the start-stop control of the corresponding compensation device is formed.

[0161] In one implementation, the required direction and amount of state change for the state node to be compensated are determined based on the state compensation amount; the direction and amount of state action that the start-stop control of the corresponding device can generate for the state node to be compensated are determined based on the action effect parameters of the candidate compensation devices; from the candidate compensation devices whose state action direction is consistent with the required state change direction, the target compensation device is determined based on the degree of matching between the state action amount and the required state change amount. Therefore, it is not simply a matter of using a device that can act on the deviation state node in the graph; rather, it is further determined whether the effect that the device can generate is compatible with the state action that has not yet been realized.

[0162] Furthermore, when the state effect generated by a single target compensation device is insufficient to meet the current state compensation amount, compensation device start-stop control is sequentially formed based on multiple candidate compensation devices. Specifically, the required state change amount is used as the initial remaining compensation amount. After determining a target compensation device, the current remaining compensation amount is updated according to the state effect amount corresponding to the target compensation device. If the updated remaining compensation amount is still greater than the preset compensation threshold, the next target compensation device is determined from the candidate compensation devices that have not yet been determined as target compensation devices, until the remaining compensation amount meets the compensation requirements. The determined start-stop control of one or more compensation devices is added to the equipment start-stop procedure to form an updated equipment start-stop procedure. The updated equipment start-stop procedure continues to perform expected state change deduction, equipment execution effect acquisition, and state deviation judgment according to steps S130 to S150, thereby ensuring that subsequent equipment start-stop is based on the actual effect of the previous equipment control.

[0163] See Figure 6 , Figure 6 It is based on Figure 1 The corresponding embodiment shows a flowchart describing a method for determining equipment start-up and shutdown control to compensate for the deviation based on the deviation between the actual state change and the expected state change, starting from the state node that caused the deviation, and updating the equipment start-up and shutdown procedure accordingly.

[0164] The step S150 provided in this application embodiment, which determines the equipment start-up and shutdown control for compensating the deviation based on the deviation between the actual state change and the expected state change, starting from the state node that caused the deviation, and reversing the action relationship, and updates the equipment start-up and shutdown procedure accordingly, includes: Step S151: Based on the actual state change and the expected state change, determine the state nodes with state deviations that need to be compensated and the corresponding state compensation amounts. Step S152: Using the node to be compensated as the starting point for reverse query, determine the candidate compensation devices that can be applied to the node to be compensated along the interaction relationship. Step S153: Based on the matching relationship between the action effect parameters and state compensation amount corresponding to each candidate compensation device, determine the target compensation device for reducing the state compensation amount, and generate the corresponding compensation device start / stop control. Step S154: Add the compensation equipment start-stop control to the equipment start-stop program to form an updated equipment start-stop program.

[0165] The steps described above will be explained in further detail below.

[0166] In step S151, based on the aforementioned embodiments, the expected state change and the actual state change correspond to the corresponding state nodes and their state attributes. Therefore, a comparison is made between the two to determine whether the actual state effect generated by the device start-up and shutdown procedure achieves the expected state effect.

[0167] Specifically, for each state node involved in the expected state change, the expected state change and the actual state change corresponding to that state node are obtained, and the state changes that have not yet been realized relative to the expected state change are determined. When the difference between the two needs to be further adjusted through equipment control, the state node is identified as a state node to be compensated, and the state compensation amount is determined based on the difference. Here, the state compensation amount represents the state change that the state node to be compensated still needs to obtain in order to make the actual state action after equipment start-up and shutdown further approach the corresponding expected state action. Therefore, the state compensation amount is not an error value in the general sense, but rather a further conversion of the deviation in equipment control effect into a state requirement that subsequent equipment control can handle.

[0168] For example, for numerical state attributes, the state compensation amount is determined based on the unrealized difference between the expected state change and the actual state change; for other comparable state attributes, the corresponding state compensation amount is determined based on the state change that still needs to occur relative to the expected state in relation to the actual state. When a single equipment start-up and shutdown procedure affects multiple state nodes, the expected state change and the actual state change of each state node are compared separately, and only the state nodes with state deviations that require further compensation are identified as the state nodes to be compensated. Therefore, subsequent compensation is not a general re-decision of the original equipment start-up and shutdown procedure, but is first located at the specific state nodes where the control effect is insufficient.

[0169] Furthermore, the state compensation amount is associated with and stored in relation to the state node that generates the compensation requirement and its corresponding state attributes. This allows subsequent device queries to not only know "which position needs further adjustment" but also "which operating state at that position needs further change." Further determination of the state change direction and amount corresponding to the state compensation amount will be explained in subsequent embodiments.

[0170] In step S152, the state node to be compensated is used as the starting point for querying in the station operation knowledge graph, and the station equipment that has an interaction relationship with the state node is queried.

[0171] Unlike step S121, which determines associated equipment based on operational requirements, this step further narrows the query criteria to state compensation requirements that still exist after the equipment is actually executed. Specifically, based on the node identifier of the state node to be compensated and the state attribute that causes the state deviation, the system queries the station operation knowledge graph to determine which equipment can be applied to the state node through corresponding equipment start / stop control, and the queried equipment is used as candidate compensation equipment.

[0172] The reverse lookup referred to here is relative to the direction of the relationship between equipment and status nodes in the station operation knowledge graph. The forward relationship is used to indicate "which status nodes a certain equipment can affect when it starts or stops", while when it is known that a certain status node has a state to be compensated, the same relationship is used to look up "which equipment can affect that status node" from the status node side.

[0173] Therefore, this step does not re-traverse all the equipment in the station, but rather uses the location of the deviation determined in step S151 and the existing interaction relationship between the station equipment and the status node to limit the search range of the compensation equipment to the equipment that can actually act on the corresponding status node to be compensated.

[0174] Furthermore, when there are interactions between a device and the same state node that affect different state attributes, the candidate compensation devices are further narrowed down based on the state attribute that caused the deviation. For example, if the effective throughput capacity of a passage state node needs to be compensated, then devices whose start / stop operations can change that effective throughput capacity are prioritized, rather than simply using a device as a candidate compensation device because it has a spatial association with the area.

[0175] The candidate compensation devices obtained in step S152 only indicate that the corresponding devices have the potential to act on the state node to be compensated, and do not mean that all candidate compensation devices are suitable for the current state compensation. In step S153, the device start-stop control and corresponding action effect parameters that can be executed by each candidate compensation device are read, and it is determined whether the state action represented by the action effect parameters matches the state compensation amount obtained in step S151.

[0176] The matching relationship is used to determine whether the expected state effect of a candidate compensation device after implementing corresponding start-stop control can reduce the current state compensation amount. For example, if a candidate compensation device is expected to cause the state node to be compensated to continue changing towards the desired state direction after executing start control, then the start-stop control of this device can be considered as a candidate compensation control; if its expected effect cannot reduce the current state compensation amount, or will further increase the state deviation, then it will not be considered as the current target compensation device. When there are multiple candidate compensation devices that can reduce the state compensation amount, the target compensation device is determined based on the degree of matching between the action effect parameters of each candidate compensation device and the current state compensation amount.

[0177] It should be noted that the action effect parameters used in this step have the same technical meaning as those used in step S130 when performing the expected state deduction; that is, they are both used to characterize the expected state effect that the equipment start-stop control can produce on the corresponding state node. Therefore, the same equipment-state effect knowledge is used both in S130 to forward determine the expected state change from the equipment start-stop control and in S153 to backward determine which equipment start-stop control is suitable for implementing compensation based on the unrealized state compensation requirements.

[0178] After identifying the target compensation device, corresponding compensation device start-stop control is generated based on the current state switching capabilities of that device. The compensation device start-stop control supplements the state actions that were not actually implemented in the previous device start-stop procedure, rather than simply repeating the original device start-stop control. Further matching processes between the state change direction of the state compensation quantity, the state change quantity, and the state action direction and state action quantity of the candidate compensation device will be explained in subsequent embodiments.

[0179] In step S154, the compensation equipment start-stop control is added to the current equipment start-stop procedure to form an updated equipment start-stop procedure. Specifically, the equipment start-stop controls already executed in the original equipment start-stop procedure and their execution results are retained, while the newly added compensation equipment start-stop control is used as subsequent control content based on the current state deviation, thereby maintaining the correspondence between the original equipment control, actual state changes, and the newly added compensation control.

[0180] When a target compensation device is identified, its corresponding compensation device start-stop control is added to the device start-stop procedure. When multiple target compensation devices need to work together to complete compensation, the corresponding multiple compensation device start-stop controls are added to the device start-stop procedure. Further methods for determining multiple compensation devices will be explained in subsequent embodiments. Therefore, the updated device start-stop procedure is not regenerated independently of the previous one, but rather uses the state effects not yet obtained after the actual execution of the previous device start-stop procedure as the basis for new control, establishing a continuous relationship between the two device controls through the state nodes to be compensated and their state compensation amounts. If necessary, the updated device start-stop procedure can continue to enter the aforementioned expected state deduction and execution feedback process to evaluate whether the newly added compensation device start-stop control can further reduce the corresponding state compensation amount.

[0181] Through the execution of steps S151 to S154, when the equipment control effect is insufficient, instead of simply generating an evaluation result of "effect not meeting expectations" or randomly selecting equipment from all station equipment, the process first assigns the unmet expectations to specific state nodes to form state compensation quantities. Then, it uses the existing equipment-state node interaction relationships in the station operation knowledge graph to reversely limit the range of equipment that can implement compensation, and further determines the compensation equipment using the action effect parameters corresponding to the candidate equipment. This allows the actual effect results of the previous equipment start-up and shutdown to be directly transformed into the directional control basis for subsequent equipment start-up and shutdown.

[0182] For further details, please refer to [link / reference]. Figure 7 , Figure 7 It is based on Figure 6 The flowchart of the method for determining the target compensation device for reducing the state compensation amount based on the matching relationship between the action effect parameters and the state compensation amount of each candidate compensation device in step S153 of the corresponding embodiment is shown.

[0183] The execution process of determining the target compensation device for reducing the state compensation amount based on the matching relationship between the action effect parameters corresponding to each candidate compensation device and the state compensation amount in step S153 of this application embodiment includes: Step S201: Determine the required state change direction and state change amount of the state node to be compensated based on the state compensation amount. Step S202: Based on the action effect parameters corresponding to each candidate compensation device, determine the direction and magnitude of the state action generated by the start / stop control of the corresponding device on the state node to be compensated. Step S203: From the candidate compensation devices whose state action direction is consistent with the state change direction, determine the target compensation device according to the degree of matching between the corresponding state action quantity and the state change quantity.

[0184] The aforementioned step S151 has already determined the state node to be compensated and its state compensation amount based on the difference between the actual state change and the expected state change. Building on this, step S201 further converts the state compensation amount into a state change requirement that can be compared with the action effect of the candidate compensation device.

[0185] Specifically, step S201 determines the direction in which the state attribute needs to continue changing in order to reduce the current deviation, based on the state attribute corresponding to the state compensation amount, and determines the required state change amount based on the portion of the actual state change that has not yet been realized relative to the expected state change. Here, the state change direction indicates whether the corresponding state attribute of the state node to be compensated should continue to increase or decrease based on the current state; the state change amount indicates the degree to which the state attribute still needs to be changed to reduce the current state deviation.

[0186] For example, for a passage status node, if the actual increase in effective passage capacity after the equipment start-stop procedure is lower than its expected increase, the state change direction is to continue increasing the effective passage capacity, and the unrealized portion between the actual increase and the expected increase is taken as the corresponding state change amount. Conversely, if the actual effect of a equipment start-stop control exceeds the expectation, causing the corresponding state attribute to need to be reverted in the opposite direction, the state change direction opposite to the original equipment control action direction is determined based on the resulting state compensation amount.

[0187] Therefore, this step does not only use the absolute size of the state compensation amount to select the device, but further expresses it as follows: the state node to be compensated determines the state attribute to be compensated, that is, the required state change direction and the required state change amount. Thus, the subsequent candidate compensation device not only needs to "be able to act on the node", but also needs its specific device start-stop control to make the corresponding state attribute change in the current required direction.

[0188] For each candidate compensation device determined in step S152, step S202 reads the device start-stop control that can be applied to the state node to be compensated and the action effect parameters corresponding to the device start-stop control.

[0189] Since the basic meaning of the action effect parameters has been explained in the previous embodiments, it will not be repeated here. For step S202, the focus is on determining from the action effect parameters the direction and magnitude of the state action that the device start-stop control can produce for the state attribute determined in step S201. The direction of state action indicates whether the corresponding device start-stop control is expected to increase or decrease the state attribute corresponding to the state node to be compensated; the magnitude of state action indicates the degree to which the device start-stop control is expected to change the state attribute.

[0190] For example, when the compensation object of the state node to be compensated is the effective throughput capacity, the action effect parameter corresponding to the standby escalator start control indicates that it can increase the effective throughput capacity of the state node, so its state action direction is increase, and the corresponding expected increase is the state action amount; if another device is deactivated, it will reduce the effective throughput capacity, so the state action direction of the corresponding deactivation control is decrease.

[0191] It should be noted that the same candidate compensation device may correspond to different device start-stop controls, and the resulting state actions of different device start-stop controls may have different directions. Therefore, step S202 compares the action effects of the candidate compensation device under the currently executable start-stop control, rather than judging whether it is suitable for compensation solely based on the device type or rated capacity.

[0192] Furthermore, when a candidate compensation device can act on multiple state attributes, only the action effect corresponding to the current state attribute to be compensated is extracted for subsequent matching. For example, if a device can simultaneously change the available state and effective passage capacity of a passage connection, and the current compensation object is the effective passage capacity, then step S202 uses the device start / stop control to apply the direction and amount of action to the effective passage capacity, instead of replacing it with changes in other state attributes.

[0193] Therefore, the state change requirement obtained in step S201 and the equipment state action obtained in step S202 have the same state nodes and state attributes, enabling a direct comparison of their direction and degree of action. It is particularly noteworthy that the action effect parameters used in step S202 and the action effect parameters used in step S130 when forward deducing the expected state changes of the equipment start-up and shutdown procedure belong to the same equipment-state action description system. That is, in step S130, the expected state change of the equipment is calculated forward using the action effect parameters corresponding to the equipment start-up and shutdown control, while step S202, based on the existing state compensation requirements, reads the similar action effect parameters corresponding to candidate equipment and determines which equipment start-up and shutdown control's state action can compensate for the required state change. Thus, the same equipment-state action knowledge serves both the forward action deduction before equipment execution and the reverse compensation selection after equipment execution.

[0194] In step S203, after obtaining the required state change direction and amount of the state change of the state node to be compensated, as well as the state action direction and amount of each candidate compensation device, state action direction matching is first performed. Candidate compensation devices whose state action direction is inconsistent with the state change direction obtained in step S201 are not considered as target compensation devices for reducing the state compensation amount in this round. For example, if the state node to be compensated currently needs to further increase its effective throughput capacity, the start-stop control of the device that increases the effective throughput capacity can be included in subsequent matching, while the start-stop control of the device that is expected to further reduce the effective throughput capacity will not be included in this round of compensation selection.

[0195] By first performing directional screening, we can avoid selecting a device as a compensation device simply because there is an interaction between it and the state node to be compensated, only to find that the actual execution further increases the state deviation. For candidate compensation devices whose state action direction is consistent with the required state change direction, we then compare the degree of matching between their state action quantity and the current required state change quantity, and determine the target compensation device accordingly.

[0196] The degree of matching reflects how closely a candidate compensation device's expected state effect matches the current state change that still needs compensation. The closer the state effect is to the required state change, the more suitable the candidate compensation device is for compensating the current state deviation. For example, if a traffic node still needs to increase its effective throughput by 80 people / minute, and candidate compensation device E2 is expected to increase it by 70 people / minute after activation, while candidate compensation device E3 is expected to increase it by 30 people / minute, all other things being equal, E2's state effect is considered to have a higher degree of matching with the current state change, thus E2 is prioritized as the target compensation device.

[0197] In one implementation, the degree of matching is determined based on the difference between the state action quantity and the state change quantity; the smaller the difference, the closer the expected effect of the equipment start-stop control is to the current compensation requirement; alternatively, the degree of matching can also be determined based on the proportional relationship between the two. This application does not limit this, as long as it reflects the degree of fit between the state action quantity of the candidate compensation equipment and the current state change quantity to be compensated.

[0198] Furthermore, candidate compensation devices whose state action is less than the current state change but can still significantly reduce the state compensation amount can also be identified as target compensation devices. In this case, the unfinished portion after the target compensation device has executed can continue to generate remaining compensation requirements, and the further processing method will be described in subsequent embodiments.

[0199] Therefore, this step does not require a single target compensation device to completely eliminate all state compensation quantities at once, thus reserving processing space for subsequent determination of other compensation devices based on the remaining compensation quantities.

[0200] Compared to selecting a device solely based on the existence of an interaction relationship between the device and the state node, this application embodiment further determines whether the device's direction of action is correct and whether its degree of action is compatible with the unrealized state change. Therefore, the candidate compensation device is further narrowed down from "able to influence the state" to "a device capable of reducing the state deviation in the direction and degree currently required."

[0201] This approach is particularly suitable for scenarios in subway stations where the same state node is affected by multiple devices. For example, the same passage location may be simultaneously affected by multiple escalators, turnstiles, or other access devices. Simply relying on the interaction between devices and state nodes is insufficient to determine which device is suitable for the current compensation requirement. However, by correlating the direction and magnitude of state changes with the effects of device actions, the determination of the compensation device can be further aligned with the actual, unrealized operational effects.

[0202] See Figure 8 , Figure 8 It is based on Figure 6 The flowchart of the method for determining the target compensation device according to the degree of matching between the corresponding state action quantity and the state change quantity in step S203 of the corresponding embodiment is shown in the embodiment.

[0203] The execution process of determining the target compensation device according to the matching degree between the corresponding state action quantity and the state change quantity in step S203 of this application embodiment includes: Step S2031: Determine the state change amount as the initial residual compensation amount; Step S2032: From the candidate compensation devices whose state action direction is consistent with the state change direction, determine the target compensation device according to the matching degree between the state action amount of each candidate compensation device and the current remaining compensation amount, and update the current remaining compensation amount according to the state action amount of the determined target compensation device. Step S2033: When the updated remaining compensation amount is greater than the preset compensation threshold, continue to determine the target compensation device from the candidate compensation devices that have not been determined as the target compensation device and update the remaining compensation amount until the remaining compensation amount is not greater than the preset compensation threshold.

[0204] The aforementioned step S201 has obtained the required state change direction and state change amount for the state node to be compensated, and step S202 has obtained the state action direction and state action amount for the corresponding device start / stop control of each candidate compensation device. Based on this, when the current state change requirement may need to be completed by one or more devices, the state change amount is used as the initial remaining compensation amount at the start of this round of device selection.

[0205] In this embodiment, the remaining compensation amount represents the state change requirements that are not yet covered by the expected state action of the determined target compensation device during the target compensation device selection process. Therefore, since the target compensation device has not yet been determined initially, all state change amounts required by the state nodes to be compensated are uncovered compensation requirements, and thus the state change amounts are determined as the initial remaining compensation amount.

[0206] It should be noted that the remaining compensation amount in this embodiment is a calculation amount in the compensation equipment selection stage. It is updated based on the expected state action amount represented by the action effect parameters of the candidate compensation equipment, and is used to determine how many more compensation equipment need to be selected; it is different from the actual state deviation obtained again based on the new station operation data after the actual execution of equipment start-stop control. The latter is still re-determined by the aforementioned steps S140 and S150 based on the actual operation results.

[0207] This distinction allows for a progressive determination of the number and combination of compensation devices to be added to the equipment start-up and shutdown procedure, based on the state functions that multiple candidate compensation devices can provide, before the actual execution of the equipment start-up and shutdown procedure.

[0208] Step S2032 uses the current remaining compensation amount as the state action requirement that the equipment selection in this round needs to meet. Among the candidate compensation equipment whose state action direction is consistent with the required state change direction, the matching degree between the state action amount of each candidate compensation equipment and the current remaining compensation amount is compared. The state action amount and the current remaining compensation amount involved in the comparison correspond to the same state node to be compensated and the same state attribute, so they can directly indicate the degree to which the start-stop control of the corresponding equipment is expected to cover the current compensation requirement.

[0209] The matching degree characterizes how close the state action of a candidate compensation device is to the current remaining compensation amount. For example, the matching degree can be determined based on the difference or proportional relationship between the two. A higher matching degree is given to candidate compensation devices that can cover a large portion of the remaining compensation amount without excessively deviating from the current compensation requirement.

[0210] Based on the matching results between each candidate compensation device and the current remaining compensation amount, the current target compensation device is determined. After determining the target compensation device, the current remaining compensation amount is updated according to the state action quantity corresponding to the target compensation device. The update is used to deduct the state action that the target compensation device is expected to achieve, so that the updated remaining compensation amount represents the state changes that still need to be achieved by other compensation devices after considering the target compensation device.

[0211] For example, if a node in a state to be compensated currently needs to increase its effective throughput by 170 units, then the initial remaining compensation amount is 170. If the target compensation device determined for the first time is expected to increase the effective throughput by 90 units after startup, then the current remaining compensation amount will be updated to the 80 units that still need to be compensated during the device selection phase.

[0212] Once the target compensation device is determined, it is excluded from the set of subsequent candidate compensation devices to avoid repeatedly selecting the same device based on the same start-stop control in the same round of device start-stop procedure.

[0213] It is particularly important to note that the determination of the second and subsequent target compensation devices is not based on the initial state change, but rather on the current remaining compensation amount updated after the previous target compensation device is determined, and then re-matched.

[0214] For example, if the state action values ​​of three candidate compensation devices are 90, 70, and 40, and the initial remaining compensation amount is 170, the device with a state action value of 90 can be selected in the first round; at this point, the remaining compensation amount becomes 80. In the second round, the matching object for device selection changes from 170 to 80, and the device with a state action value of 70 is a better match for the new compensation requirement. Therefore, each time a target compensation device is determined, the state compensation requirement faced by the next round of device selection changes.

[0215] After each update of the remaining compensation amount, step S2033 compares the updated remaining compensation amount with a preset compensation threshold. The preset compensation threshold indicates that the current compensation requirement has been reduced to an acceptable range where no further compensation equipment needs to be added. When the updated remaining compensation amount is not greater than the preset compensation threshold, the determination of the target compensation equipment for this round ends. When the updated remaining compensation amount is still greater than the preset compensation threshold, it indicates that the combined effect of one or more target compensation equipment is still insufficient to meet the current compensation requirement.

[0216] At this point, the updated remaining compensation amount is used as the state change requirement corresponding to the next round of equipment selection. From the candidate compensation devices that have not yet been determined as target compensation devices, the matching degree between the state action of each device and the current remaining compensation amount is re-compared, and the next target compensation device is determined.

[0217] Subsequently, the remaining compensation amount is updated again based on the state action of the newly determined target compensation device, and the above judgment process is repeated until the remaining compensation amount is not greater than the preset compensation threshold. For example, the required state change of the state node to be compensated is 170, the preset compensation threshold is 20, and the three candidate compensation devices that meet the state action direction requirements can generate state action amounts of 90, 70, and 35, respectively. First, 170 is determined as the initial remaining compensation amount, and the first target compensation device with a state action amount of 90 is determined, and the updated remaining compensation amount is 80. Since 80 is greater than the compensation threshold of 20, matching continues from the remaining candidate compensation devices. At this time, the candidate device with a state action amount of 70 has a high degree of matching with the current remaining compensation amount of 80, and it is determined as the second target compensation device, and the updated remaining compensation amount is 10. Since 10 is no longer greater than the preset compensation threshold of 20, the selection of devices is stopped, and the remaining candidate device with a state action amount of 35 is no longer added to the compensation control in this round.

[0218] In this way, even if multiple station devices can act on the same state node to be compensated, it is not necessary to directly activate all candidate devices because a single device is insufficient. Instead, it is determined one by one whether it is still necessary to add more devices based on the remaining state requirements after each device is expected to act.

[0219] Furthermore, in one implementation, for candidate compensation devices whose state action is significantly greater than the current remaining compensation, a preset compensation threshold is used to determine whether the expected over-action is within an acceptable range. For candidate compensation devices that are expected to cause significant over-compensation, their matching degree is reduced or they are not considered as the current target compensation device. For example, if there are only 20 units of compensation demand remaining, and a candidate device will increase the corresponding state attribute by 100 units upon startup, then even if its action direction is correct, it may not be suitable for the current remaining state demand. By re-matching the remaining compensation amount with the state action, excessive device control can be avoided based solely on "the device can increase the state attribute".

[0220] Furthermore, when all candidate compensation devices have been selected and judged, but the remaining compensation amount is still greater than the preset compensation threshold, the remaining compensation amount that is not currently satisfied is retained as an indicator that the current set of candidate devices is insufficient to fully achieve state compensation. After the device start-up and shutdown procedure is actually executed, the aforementioned state deviation and compensation device determination process is re-entered based on the new actual state changes.

[0221] In steps S2031 to S2033, this embodiment transforms the process of determining multiple compensation devices from a one-time device combination selection to a progressive selection process based on remaining state requirements.

[0222] Of particular importance is that the state effect of the preceding target compensation device directly changes the selection criterion for the subsequent target compensation device. Therefore, different compensation devices are not selected independently based on their initial state compensation values, but rather form a correlation within the same compensation process.

[0223] This approach is particularly suitable for scenarios in subway stations where the same operating state can be changed by multiple discrete devices. For example, the same passage area can be provided with throughput capacity by multiple escalators or multiple sets of turnstiles, and the throughput capacity that each device can increase is discrete. When the required compensation capacity cannot be fully provided by a single device, this embodiment can add devices one by one according to the remaining capacity demand, and stop activating other devices when the allowable compensation range has been reached, thereby matching the number of compensation devices with the degree of effect of the required state.

[0224] Once all target compensation devices are identified, the corresponding compensation device start-stop control can be generated according to the aforementioned step S153, and the corresponding compensation device start-stop control can be added to the device start-stop program according to step S154. This will not be repeated here.

[0225] The following example, using the transfer passage and its associated escalator equipment, illustrates the interaction between the aforementioned steps.

[0226] During a certain operating period, step S110 performs station status collaborative calculations based on the current station operating data to obtain the current passenger flow status of the corresponding status node of the transfer channel, and forms the corresponding future operating status based on the upstream status with which it has a passenger flow transfer relationship. At the current moment, the channel can still meet the passenger flow requirements, but the future operating status indicates that if the current equipment operation continues, the passenger flow demand of the channel will exceed its effective throughput capacity.

[0227] Step S120 determines the escalator E1 that can change the passage capacity of the channel based on the future operating state of the channel status node and the equipment interaction relationship in the station operation knowledge graph. When the future operating state meets the start-up conditions of escalator E1 and escalator E1 is currently in a stopped state, escalator E1 is identified as the device to be controlled, and a device start-up and stop procedure including escalator E1 start-up control is generated.

[0228] Step S130: Based on the action effect parameters corresponding to the start-up control of escalator E1, apply the expected change in passage capacity brought about by the start-up of escalator E1 to the corresponding state node or state connection relationship, and deduce the expected state change of the relevant channel state node after the start-up of escalator E1 based on the state transmission relationship between state nodes.

[0229] After the escalator E1 is actually started, step S140 continues to acquire station operation data and re-perform station state collaborative calculation to obtain the reliable operation state of the same channel state node after the escalator E1 is started, and obtain the actual state change of the escalator E1 based on the state difference before and after execution.

[0230] If the actual state change is less than the expected state change, for example, although escalator E1 starts normally, the actual increase in the effective throughput capacity of the corresponding channel is lower than the expected increase due to actual passenger flow organization, the reception status of adjacent areas, or other operational factors. In step S150, the state change of the channel state node that still needs to be compensated is obtained based on the difference between the two.

[0231] At this point, instead of restarting the device selection process from all devices in the entire station, the system starts with the current channel status node and searches backwards along the interaction relationships between devices and status nodes to find other devices that can affect the channel status. For example, if the query reveals that the backup escalator E2 can also improve the relevant passage capacity, then the system further matches the direction and amount of action represented by the action effect parameters of escalator E2 with the current status compensation amount, and generates the start control for escalator E2 when the compensation requirements are met. If the effect generated by escalator E2 alone is still insufficient, the next candidate compensation device is determined based on the updated remaining compensation amount.

[0232] It can be seen that the steps in this application embodiment do not independently complete a data calculation, a device control, and a feedback evaluation. Instead, the output of the previous step continuously limits the processing object of the next step: the state node formed in step S110 and its current and / or future operating state limit the device function requirements to be processed in step S120; the device to be controlled and the device start-stop control determined in step S120 limit the action effect parameters to be called in step S130 and the state node to be deduced; the expected state change obtained in step S130 further limits the state object and state attribute to be evaluated by actual station operation data in step S140; after the actual state change obtained in step S140 deviates from the corresponding expected state change, step S150 converts the state node where the deviation occurs into the reverse query entry of the compensation device.

[0233] Therefore, the embodiments of this application form a closed-loop processing process from "station operation status" to "equipment start-stop control", then from "equipment start-stop control" back to "station operation status change", and further determine the subsequent equipment start-stop control from the state change that does not meet the expectation.

[0234] In this way, equipment start-up and shutdown are no longer based solely on fixed operating times or single monitoring conditions, nor are they solely based on whether the equipment itself has started or stopped as the effect of equipment control. Instead, the basis for equipment control, the expected function of the equipment, the actual function of the equipment, and subsequent compensation control are all linked to specific state nodes in the dynamic station status diagram. This allows the equipment start-up and shutdown procedure to be continuously adjusted according to the current and / or future operating status of the station and the actual effect of the preceding equipment control, thereby improving the matching degree and adaptability between the start-up and shutdown control of subway station equipment and the actual operating status.

[0235] Further, see Figure 9 , Figure 9 This is a schematic diagram illustrating a video surveillance service environment at a rail transit station according to an exemplary embodiment. Figure 9 The interface shown includes entry points for functions such as video surveillance, equipment ledger, and maintenance statistics. It also provides selection options for lines, stations, and monitoring screen items. The displayed live video feed shows the scene when a train is stopped at the platform. This interface demonstrates that during daily station operations, video data from relevant areas can be obtained based on monitoring points and integrated with station equipment management and operation monitoring.

[0236] In a specific application example, when applying the method of this application to a subway station with the aforementioned video surveillance business environment, step S110 first obtains the video data and observation time corresponding to each monitoring point in the station, and obtains operational observation results such as the number of people in the area, passenger flow direction, or passenger flow status from video analysis; simultaneously, it obtains station operational data such as passenger flow entering and exiting the station, expected train arrivals and departures, and equipment start / stop status from the ticketing system, train operation system, and station equipment monitoring interface. Based on the station objects and status attributes corresponding to each data source, operational observation knowledge elements are constructed, and these operational observation knowledge elements are associated with the corresponding observation objects in the station operational knowledge graph.

[0237] For example, platform video surveillance points are associated with the corresponding state nodes of the platform, and passageway U3, escalator service area U4, and escalators E1 and E2, which have passenger flow transfer relationships with this platform, are mapped to the corresponding objects in the station operation knowledge graph. When train arrival and departure information and video observations indicate that subsequent passenger flow may transfer from the platform to passageway U3, collaborative state estimation is performed based on direct observations of the monitoring points and passenger flow transfer relationships in the station operation knowledge graph. This forms the current operating state of state nodes such as passageway U3 and U4, and the future operating state at the corresponding future target station state time is deduced, resulting in a dynamic station state diagram.

[0238] In step S120, when the dynamic station status diagram indicates that the future passenger flow demand for passage U3 will exceed the current effective throughput capacity, the corresponding status node is used as the query entry point. Following the interaction relationships between devices and status nodes in the station operation knowledge graph, associated devices such as escalators E1 and E2 that can adjust the throughput capacity of the passage are identified. Then, the current or future operating status of the corresponding status node is matched with the start / stop conditions of each associated device. If escalator E1 meets the start conditions and is currently in a stopped state, escalator E1 is identified as the device to be controlled, and a device start / stop program containing the start control of escalator E1 is generated. For devices already in the required operating state, the same start / stop control is not generated repeatedly.

[0239] In step S130, based on the action effect parameters corresponding to the escalator E1 start control, the expected effects of E1 start on traffic connection, effective throughput capacity attributes, and their direction and magnitude of change are determined. A dynamic station status diagram is copied to form a deduced station status diagram. The action effect parameters are converted into graphical parameter changes of the corresponding state connection relationships. State deduction is performed according to the state transmission relationships between state nodes such as channel U3 and escalator service area U4 to obtain the expected state changes of relevant state nodes after escalator E1 start in terms of passenger outflow, passenger accumulation, and inflow status.

[0240] In step S140, the equipment start-up and shutdown procedure is executed, and the actual start-up and shutdown status of escalator E1, as well as new station operation data such as video monitoring and passenger flow monitoring of related areas, are continuously received. Using the same operation observation knowledge element construction and station status collaborative calculation method as in step S110, the state nodes that interact with escalator E1 are updated to determine their actual state changes before and after the equipment start-up and shutdown procedure is executed. Therefore, whether the equipment has successfully started up and whether it has actually achieved the passenger flow improvement effect required for the relevant passageway are evaluated separately.

[0241] In step S150, the actual state change and expected state change corresponding to the same state node and the same state attribute are compared. If escalator E1 has been started, but the actual effective throughput capacity increase of the relevant state node of passage U3 is still lower than the expected increase, then the state node to be compensated and the state compensation amount are determined based on the state changes that have not yet been realized between the two. Starting from the state node to be compensated, candidate compensation devices are queried in reverse along the interaction relationship between the equipment and the state node, for example, the standby escalator E2 is found; then, based on the action effect parameters of escalator E2, the state action direction and state action amount of its start control are determined, matched with the requirements corresponding to the state compensation amount, the target compensation device that can reduce the deviation is determined, and the start-stop control of the compensation device is generated, and the equipment start-stop program is updated accordingly.

[0242] In the aforementioned operational environment of station video surveillance and equipment operation management, video surveillance footage and other operational data are used to form a dynamic station status corresponding to specific observed objects. The interaction relationships and equipment start-stop conditions are used to convert relevant state requirements into controllable equipment and its start-stop control. Action effect parameters are used to deduce the expected effects of equipment start-stop. The new station operational data after execution then locates any unintended state changes to the corresponding state nodes and determines the compensation equipment in reverse order along the same interaction relationship. Therefore, this application can update the equipment start-stop procedure based on the difference between the actual and expected effects of the equipment, rather than simply performing a one-time control based solely on the equipment's own start-stop status.

[0243] In one exemplary embodiment, this application also provides a computer system including a memory, a processor, and a computer program stored in the memory, the processor executing the computer program to implement the steps of the method as described above.

[0244] In one exemplary embodiment, this application also provides a computer program product including a computer program that, when executed by a processor, implements the steps of the method as described above.

[0245] In one exemplary embodiment, this application also provides a computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the steps of the method as described above.

[0246] Through the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions according to the embodiments of this application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, terminal device, or network device, etc.) to execute the methods according to the embodiments of this application.

Claims

1. A smart start-stop method for subway station equipment with adaptive status, characterized in that, The method includes: Station status collaborative calculation is performed based on station operation data and station operation knowledge graph to form a dynamic station status map, which represents at least one of the current operation status and future operation status; Based on the operating status of each state node in the dynamic station status diagram, and the interaction relationship between station equipment and state nodes and the equipment start-stop conditions in the station operation knowledge graph, the equipment to be controlled is determined and the corresponding equipment start-stop control equipment start-stop program is generated. Based on the action effect parameters corresponding to the equipment start-stop control in the station operation knowledge graph, the expected state changes of the state nodes that have an interaction with the equipment to be controlled are deduced based on the dynamic station status diagram after the equipment start-stop procedure is executed; The device start-up and shutdown procedure is executed, and the dynamic station status diagram is updated based on the station operation data generated during the execution process to obtain the actual status changes of the status nodes that have an interaction relationship with the device to be controlled. Based on the deviation between the actual state change and the expected state change, starting from the state node that generates the deviation, the device start-stop control for compensating the deviation is determined in reverse along the action relationship, and the device start-stop procedure is updated accordingly.

2. The method according to claim 1, characterized in that, The process of performing collaborative station status calculations based on station operation data and station operation knowledge graphs to form a dynamic station status map includes: Based on the station operation data generated from various data sources, construct operation observation knowledge elements for the corresponding observation objects; The operation observation knowledge element is associated with the corresponding observation object in the station operation knowledge graph, and the collaborative state estimation is performed based on the operation observation knowledge element and the operation state transmission relationship between the observation objects in the station operation knowledge graph to obtain the credible dynamic station state of each observation object. The dynamic station status map is formed based on the credible dynamic station status of each observation object and its spectral relationship.

3. The method according to claim 1, characterized in that, The process of determining the equipment to be controlled and generating corresponding equipment start-up and shutdown procedures based on the operating status of each state node in the dynamic station status diagram, and the interaction relationships and equipment start-up and shutdown conditions between station equipment and state nodes in the station operation knowledge graph, includes: Based on the current and / or future operating status of each status node in the dynamic station status diagram, determine the station equipment associated with the corresponding status node along the interaction relationship; The current and / or future operating states of the status nodes are matched with the equipment start-up and stop conditions of the associated station equipment to determine the target equipment that meets the start-up or stop-up conditions. Based on the current device status of the target device in the dynamic station status diagram, determine the device to be controlled that needs to be started or stopped. The device start / stop program is generated based on the start control or stop control corresponding to each device to be controlled.

4. The method according to claim 1, characterized in that, The step of deducing the expected state changes of state nodes that interact with the controlled equipment after executing the equipment start-stop procedure based on the dynamic station status diagram, according to the action effect parameters corresponding to the equipment start-stop control in the station operation knowledge graph, includes: The dynamic station status diagram is copied to form the inferred station status diagram; Based on the action effect parameters corresponding to the equipment start-stop control, the equipment start-stop control is converted into graph parameter changes of the corresponding state nodes and / or state connection relationships in the simulation station status diagram; Based on the changes in the graph parameters, the state of the station state diagram is deduced according to the state transmission relationship between state nodes in the station operation knowledge graph, so as to obtain the deduced state of the state nodes affected by the changes in the graph parameters. The expected state change is determined based on the difference between the simulated state and the operating state of the corresponding state node before the execution of the device start-up and shutdown procedure.

5. The method according to claim 1, characterized in that, The execution of the equipment start / stop procedure, and the updating of the dynamic station status diagram based on the station operation data generated during the execution process, to obtain the actual state changes of the state nodes that have an interaction relationship with the equipment to be controlled, includes: Based on the station operation data generated during the execution of the equipment start-up and shutdown procedure, construct corresponding operation observation knowledge elements; Based on the new operational observation knowledge elements and the station operational knowledge graph, station status collaborative calculation is performed to update the operational status of state nodes that have an interaction relationship with the equipment to be controlled; The actual state change is determined based on the difference in the running state of the state nodes before and after the update.

6. The method according to claim 1, characterized in that, The step of determining equipment start-up and shutdown control to compensate for the deviation based on the deviation between the actual state change and the expected state change, starting from the state node that generated the deviation, and working backwards along the interaction relationship, and updating the equipment start-up and shutdown procedure accordingly, includes: Based on the actual state changes and the expected state changes, determine the state nodes with state deviations that need to be compensated and the corresponding state compensation amounts. Using the node to be compensated as the starting point for reverse query, candidate compensation devices that can be applied to the node to be compensated are determined along the interaction relationship. Based on the matching relationship between the action effect parameters of each candidate compensation device and the state compensation amount, the target compensation device for reducing the state compensation amount is determined, and the corresponding compensation device start-stop control is generated. The compensation equipment start-stop control is added to the equipment start-stop program to form an updated equipment start-stop program.

7. The method according to claim 6, characterized in that, The step of determining the target compensation device for reducing the state compensation amount based on the matching relationship between the action effect parameters corresponding to each candidate compensation device and the state compensation amount includes: The required state change direction and state change amount of the state node to be compensated are determined based on the state compensation amount. Based on the action effect parameters corresponding to each candidate compensation device, determine the direction and magnitude of the state action generated by the start-stop control of the corresponding device on the state node to be compensated. The target compensation device is determined from among the candidate compensation devices whose state action direction is consistent with the state change direction, according to the degree of matching between the corresponding state action quantity and the state change quantity.

8. The method according to claim 7, characterized in that, Determining the target compensation device based on the degree of matching between the corresponding state action and the state change includes: The state change amount is determined as the initial residual compensation amount; From the candidate compensation devices whose state action direction is consistent with the state change direction, the target compensation device is determined according to the matching degree between the state action quantity of each candidate compensation device and the current remaining compensation quantity, and the current remaining compensation quantity is updated according to the state action quantity of the determined target compensation device. When the updated remaining compensation amount is greater than the preset compensation threshold, the target compensation device is determined from the candidate compensation devices that have not been determined as the target compensation device, and the remaining compensation amount is updated, until the remaining compensation amount is not greater than the preset compensation threshold.

9. A computer system comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the method according to any one of claims 1-8.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1-8.