A substation in-station local autonomous control system and method based on WAPI

Abstract

The application discloses a substation in-station local autonomous control system and method based on WAPI, and relates to the technical field of control.The system comprises a WAPI in-station wireless network, a state sensing unit located in the station, an execution unit and an autonomous control terminal, and a device state construction module, a graph reasoning and strategy generation module and an execution feedback and closed-loop correction module on the autonomous control side.The method comprises the following steps: S1, state acquisition and safe uploading;S2, device state construction;S3, risk scene identification;S4, graph reasoning and candidate strategy acquisition;S5, target autonomous strategy determination;S6, safety check;S7, control instruction issuing and execution;S8, execution feedback acquisition;S9, execution effect judgment;and S10, closed-loop correction and graph updating.The autonomous strategy, safety check, feedback and closed-loop correction ensure the in-station operation safety under extreme working conditions.

Description

Technical Field

[0001] This invention relates to the field of control technology, and in particular to a local autonomous control system and method for substations based on WAPI. Background Technology

[0002] With the continuous improvement of intelligent and digital operation and maintenance levels in substations, a large number of terminal devices such as environmental sensing, condition monitoring, video surveillance, and mobile operation equipment have been deployed within the substations to achieve comprehensive condition data collection and auxiliary operation and maintenance. While traditional fiber optic communication offers high bandwidth and stability, it is primarily suitable for wired connections between fixed points and cannot meet the flexible access needs of mobile terminals and distributed nodes in the last mile. Therefore, in complex scenarios such as switchgear, equipment rooms, and protection rooms, there is an urgent need to build a secure and reliable intra-substation wireless communication system.

[0003] However, existing mainstream wireless communication technologies are insufficient to fully meet the stringent production control requirements of substations. Specifically: conventional Wi-Fi networks are mostly used for general access, lacking sufficient security and independent networking capabilities; 4G public networks rely excessively on operator networks, posing security risks for data outbound transmission and business continuity risks; while 5G technology offers superior performance, it faces problems such as excessively high construction and maintenance costs and overly complex systems in scenarios requiring independent networking and local autonomy in specific areas. In contrast, WAPI, as a wireless LAN authentication and confidentiality infrastructure, possesses strong identity authentication and secure transmission capabilities, making it more suitable for localized and independent deployment within the substation, providing a secure and reliable communication foundation for control links.

[0004] While WAPI can address the issue of secure communication networks, existing substation control mechanisms still have significant shortcomings, generally facing over-reliance on the master station. Currently, substation information is mostly collected on-site—centralized transmission—master station analysis—manual or remote command issuance. This model has a long response chain, and especially under extreme abnormal conditions such as upstream link interruption or master station disconnection, the substation system often falls into a state of being monitorable but difficult to handle, or having alarms but no closed-loop paralysis, seriously threatening the continuous security of the core equipment operating environment.

[0005] On the other hand, existing localized backup control methods for communication link failures are often too crude and isolated. For example, relying solely on a single temperature or humidity threshold to trigger the start and stop of a single heater or fan lacks comprehensive judgment of multi-source status information, fails to achieve coordinated linkage between different actuators such as air conditioners, exhaust fans, and dehumidifiers, and lacks a feedback verification mechanism after control execution, making it difficult to cope with complex and ever-changing environmental anomalies.

[0006] Therefore, the safety hazards of station operation under extreme conditions have become a technical problem that urgently needs to be solved. Summary of the Invention

[0007] This invention provides a local autonomous control system and method for substations based on WAPI, which solves the technical problem of safety hazards in substation operation under extreme conditions.

[0008] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0009] A WAPI-based substation local autonomous control system includes a WAPI substation wireless network, a state sensing unit, an execution unit, and an autonomous control terminal located within the substation, as well as an equipment state construction module, a graph reasoning and strategy generation module, and an execution feedback and closed-loop correction module on the autonomous control side. The state sensing unit, the execution unit, and the autonomous control terminal are all connected to and communicate with the WAPI substation wireless network.

[0010] The device status construction module is used to organize and express the status of target devices uploaded in real time via the WAPI intranet wireless network, thereby obtaining a structured device status. ;

[0011] The graph reasoning and strategy generation module is used to determine the current device status. By combining existing historical samples, ontology rules and historical policy records, state matching, risk identification and policy reasoning are performed to obtain the corresponding autonomous policy. The target autonomous policy is subject to security verification before it is issued and executed. After the security verification is passed, it is issued and executed.

[0012] The execution feedback and closed-loop correction module is used to correct subsequent graph inference results and autonomous strategies based on changes in device status after execution.

[0013] A further technical solution is as follows: the state sensing unit is deployed around the target device to collect the state of the target device in real time; the state sensing unit includes a temperature sensor, a humidity sensor, and a partial discharge sensor; the execution unit is deployed around the target device to adjust the operating environment of the target device; the execution unit includes a fan, a heater, and a dehumidifier; the device state includes device identification, status, and auxiliary constraints, the device identification includes device name, device number, device type, and region, the status includes temperature, humidity, and partial discharge level, and the auxiliary constraints include execution unit availability, number of actions, action interval time, local communication status, and operating permission status.

[0014] A further technical solution is that: in the graph reasoning and strategy generation module, based on the current device status... Identify the risk scenarios in which the target device is located, and filter the set of historical state subgraphs corresponding to the risk scenarios in the map data; obtain a set of candidate control strategies based on the similarity between the current device state and each historical state subgraph; the risk scenarios include condensation risk scenarios, local overheating scenarios, high humidity risk scenarios, cooling abnormality scenarios, and communication abnormality backup operation scenarios.

[0015] A further technical solution is that, in the graph reasoning and strategy generation module, the security verification includes execution unit availability verification, action count verification, action interval time verification, local communication status verification, and operation license condition verification.

[0016] A further technical solution is as follows: In the execution feedback and closed-loop correction module, when the execution result shows that the current autonomous strategy is effective, the current state-strategy-result record is written back to the graph data to form a new historical sample record; when the execution result shows that the current autonomous strategy is ineffective or only partially effective, the current feedback state is sent back to the graph reasoning and strategy generation module as a new input state to obtain subsequent autonomous strategies.

[0017] A further technical solution is that, in the execution feedback and closed-loop correction module, when the feedback state... If the target device is still in an abnormal state and the regenerated autonomous policy is different from the current policy, the new autonomous policy will be executed first; if the regenerated autonomous policy is the same as the current policy but the number of executions has reached the preset limit, it will switch to the minimum operating mode or alarm mode.

[0018] A further technical solution includes a data acquisition module for obtaining the status of the target device via the WAPI intranet wireless network.

[0019] A further technical solution includes a control command issuance and execution module, which converts autonomous policies into specific control commands for the execution unit and issues them via the WAPI intranet wireless network.

[0020] A WAPI-based substation local autonomous control method, based on the aforementioned WAPI-based substation local autonomous control system, includes the following steps:

[0021] Step S1: Collect the status of the target device in real time and securely upload it to the autonomous control terminal within the station via the WAPI intra-station wireless network;

[0022] Step S2: Obtain the real-time device status based on the real-time target device status;

[0023] Step S3: Identify the current risk scenario based on the current device status;

[0024] Step S4: Combine the current device status with historical samples, ontology rules and historical policy records in the graph data, perform state matching and graph reasoning to obtain a set of candidate control policies corresponding to the current risk scenario;

[0025] Step S5: Obtain the target autonomous policy based on the candidate control policy set;

[0026] Step S6: Perform security verification on the target autonomous policy. Once the target autonomous policy passes the security verification, proceed to step S7.

[0027] Step S7: Convert the target autonomous policy that has passed the security verification into specific control instructions for the execution unit, and send them to the corresponding execution unit via the WAPI intranet wireless network;

[0028] Step S8: After the execution unit completes its action, the status sensing unit collects the status of the target device again and uploads it to the autonomous control terminal via the WAPI intra-station wireless network to form the feedback status after execution;

[0029] Step S9: Determine the execution effect of the current target autonomous strategy based on the state changes before and after execution; when the target device state recovers to the preset target range, it is determined that the current autonomous strategy is effective and step S10 is executed; when the target device state does not recover to the preset target range, it is determined that the abnormal state has not been eliminated after the execution of the current autonomous strategy and step S4 is executed.

[0030] Step S10: Write the status, target autonomous strategy, execution result and feedback status corresponding to the current event back to the graph data and historical strategy records to update subsequent graph reasoning and autonomous strategies, realize dynamic autonomous control within the station and closed-loop correction.

[0031] A further technical solution is as follows: In step S5, when the candidate control strategy set contains only one candidate strategy, the candidate strategy is directly determined as the target autonomous strategy; when the candidate control strategy set contains multiple candidate strategies, the candidate control strategies are sorted according to historical success rate, recovery time, number of actions and execution cost, and the candidate strategy with the best sorting result is selected as the target autonomous strategy; when the candidate control strategy set is empty, the local rule generation module is called to generate a supplementary strategy as the target autonomous strategy.

[0032] The beneficial effects of adopting the above technical solution are as follows:

[0033] A WAPI-based substation local autonomous control system includes a WAPI substation wireless network, a state sensing unit, an execution unit, and an autonomous control terminal located within the substation, as well as an equipment state construction module, a graph reasoning and strategy generation module, and an execution feedback and closed-loop correction module on the autonomous control side. The equipment state construction module organizes and expresses the state of target equipment uploaded in real time via the WAPI substation wireless network to obtain structured equipment states. The graph reasoning and strategy generation module performs state matching, risk identification, and strategy reasoning based on the current equipment state combined with existing historical samples, ontology rules, and historical strategy records to obtain corresponding autonomous strategies. The target autonomous strategy undergoes security verification before execution; execution is only initiated after the security verification is passed. The execution feedback and closed-loop correction module corrects subsequent graph reasoning results and autonomous strategies based on changes in equipment state after execution. Through autonomous strategies, security verification, feedback, and closed-loop correction, the system ensures safe operation within the substation even under extreme conditions.

[0034] A WAPI-based substation local autonomous control method includes steps S1 (status acquisition and security upload), S2 (equipment status construction), S3 (risk scenario identification), S4 (graph reasoning and candidate strategy acquisition), S5 (target autonomous strategy determination), S6 (security verification), S7 (control command issuance and execution), S8 (execution feedback acquisition), S9 (execution effect judgment), and S10 (closed-loop correction and graph update). Through autonomous strategies, security verification, feedback, and closed-loop correction, it can ensure the safe operation of the substation even under extreme conditions. Attached Figure Description

[0035] Figure 1 This is a topology diagram of the overall technical architecture of WAPI site-wide secure access + device status upload + graph inference to generate autonomous strategies + execution feedback closed-loop correction;

[0036] Figure 2 It is a data flow graph of graph reasoning and strategy generation;

[0037] Figure 3 This is a flowchart of a closed-loop autonomous control method for substations based on WAPI. Detailed Implementation

[0038] To address potential safety hazards in substation operations under extreme conditions, a closed-loop autonomous control system and method based on WAPI is urgently needed. The aim is to utilize WAPI to construct an absolutely secure and controllable local wireless network, and to endow substation edge nodes with advanced autonomous capabilities on top of the communication foundation. This allows them to autonomously complete a fully closed-loop control process—multi-source sensing, integrated decision-making, coordinated execution, and feedback verification—even in the event of upstream link failure, master station disconnection, or local communication anomalies, thereby ensuring continuous control over the substation's operating environment and equipment status.

[0039] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit this application or its application or use. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0040] Many specific details are set forth in the following description in order to provide a full understanding of this application. However, this application may also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the spirit of this application. Therefore, this application is not limited to the specific embodiments disclosed below.

[0041] Example 1:

[0042] This invention discloses a local autonomous control system for substations based on WAPI, which adopts an overall technical architecture of WAPI secure access within the substation, equipment status uploading, graph inference to generate autonomous strategies, and execution feedback closed-loop correction.

[0043] like Figure 1As shown, on the equipment side (left side of the diagram), state sensing units and execution units are deployed around the target equipment switchgear #1. The state sensing units collect equipment status information such as temperature, humidity, and partial discharge. The execution units include at least fans, heaters, and dehumidifiers. On the autonomous control side, the state sensing units securely upload the collected equipment status information via the WAPI intra-site wireless network to form the current status information of the target equipment. Then, based on the current status information, graph inference is performed to generate an autonomous policy corresponding to the current status. The autonomous policy then sends control commands to the execution units via the WAPI intra-site wireless network to drive the fans, heaters, dehumidifiers, and other devices, thereby achieving local closed-loop control of the switchgear's operating status. The status changes after execution are then fed back and uploaded by the state sensing units to correct subsequent graph inference and autonomous policy output. Through this architecture, the system retains local state sensing, policy output, control execution, and feedback write-back capabilities even in scenarios of upper-level link anomalies or master station disconnection, thus achieving interpretable, scalable, and continuously iterative intra-site autonomous control.

[0044] Based on the above architecture, the technical solution of this application can be divided into five parts: a data acquisition module, a device status construction module, a graph reasoning and strategy generation module, a control command issuance and execution module, and an execution feedback and closed-loop correction module. Specifically, the data acquisition module collects the operating status information of the target device; the device status construction module organizes and expresses the real-time status data uploaded via WAPI; the graph reasoning and strategy generation module matches or generates corresponding autonomous strategies based on the current device status information; the control command issuance and execution module converts the autonomous strategies into specific control commands for the execution unit and issues them via the WAPI intra-site wireless network; and the execution feedback and closed-loop correction module corrects subsequent graph reasoning results and strategy output criteria based on changes in the device status after execution.

[0045] 1. Data acquisition module.

[0046] The data acquisition module is preferably implemented by a status sensing unit, which is used to collect real-time operational status information of the target equipment and its surrounding environment, and upload it to the autonomous control side via the WAPI intra-station wireless network. The status sensing unit can be deployed in switchgear, transformers, cable trays, protection rooms, or other target areas that require local autonomous control, and is used to acquire real-time data reflecting the operational status of the target equipment, environmental status, and abnormal characteristics.

[0047] The status sensing unit includes at least one or more of the following: temperature sensor, humidity sensor, partial discharge sensor, smoke sensor, water immersion sensor, door magnetic sensor, oil temperature sensor, hot spot temperature sensor, load rate acquisition unit, and equipment operation status acquisition unit; wherein, for switchgear scenarios, it is preferable to acquire equipment status information such as temperature, humidity, partial discharge quantity, and cabinet door status; for transformer scenarios, it is preferable to acquire equipment status information such as oil temperature, hot spot temperature, load rate, oil level, and cooling device operation status.

[0048] The data acquisition module collects status information including at least one or more of the following: temperature, humidity, partial discharge level, dew point difference, abnormal duration, device availability, number of actions, and execution unit operating status. The status perception unit encapsulates the collected device status information according to a preset data format and securely uploads it via the WAPI intranet wireless network to form the current status information of the target device, providing input for subsequent graph inference and autonomous strategy generation.

[0049] like Figure 1 As shown, taking switch cabinet #1 as an example, the status sensing unit can be composed of a temperature sensor, a humidity sensor and a partial discharge sensor, which are used to collect the temperature, humidity and partial discharge information of switch cabinet #1 respectively. After the collected status information is uploaded through the WAPI station wireless network, the corresponding equipment status information can be formed on the autonomous control side, including equipment identification, temperature value, humidity value, partial discharge value and other extended status parameters, which can then serve as the basic input for subsequent graph inference.

[0050] 2. Device status construction module.

[0051] The device status construction module is used to organize, parse, and express real-time status data uploaded via the WAPI intranet wireless network to form device status information that can be used for subsequent graph inference. Specifically, the device status construction module performs field mapping, status classification, and attribute association on the data uploaded by the status sensing unit according to the target device type, status parameter type, and preset data format, thereby constructing the device status corresponding to the target device at the current moment.

[0052] The device status preferably includes device identification information, status parameter information, and auxiliary constraint information; wherein, the device identification information is used to uniquely identify the target device, and includes at least one or more of the following: device name, device number, device type, and region; the status parameter information is used to characterize the current operating status of the target device, and includes at least one or more of the following: temperature, humidity, partial discharge, dew point difference, abnormal duration, oil temperature, hot spot temperature, load rate, and cooling device status; the auxiliary constraint information is used to characterize the conditions related to the execution of the current strategy, and includes at least one or more of the following: execution unit availability, number of actions, action interval time, and local communication status.

[0053] In a preferred embodiment, the device state construction module organizes the real-time state information of the target device into a structured state expression form to facilitate subsequent graph inference. The structured state expression form can be represented as:

[0054]

[0055] In the formula, This indicates the device identification information of the target device at the current moment. This indicates the current status parameter information of the target device. This indicates the auxiliary constraint information of the target device at the current moment.

[0056] Furthermore, the state parameter information Information used to characterize the operating status, environmental status, and abnormal characteristics of a target device or target area at the current moment can be represented as:

[0057]

[0058] In the formula, Indicates the first corresponding to the target device or target area One state parameter, The number of state parameters is selected based on the target device type, operating scenario, and autonomous control requirements.

[0059] The status parameters may include at least one or more of the following: temperature, humidity, dew point difference, partial discharge amount, abnormal duration, smoke concentration, water immersion status, cabinet door status, oil temperature, hot spot temperature, load rate, oil level, cooling device operating status, equipment availability, number of actions, and communication status.

[0060] Status parameter information corresponding to different target devices or target areas They can be different; taking a switchgear as an example, the status parameter information... This may include temperature, humidity, partial discharge level, dew point difference, and duration of abnormality; taking a transformer as an example, the status parameter information... This may include oil temperature, hot spot temperature, load rate, oil level, and the operating status of the cooling device; taking an equipment room or cable tray as an example, the status parameter information... This may include ambient temperature, ambient humidity, smoke concentration, water immersion status, and the operating status of ventilation devices.

[0061] Furthermore, the auxiliary constraint information is used to characterize the constraints of the current strategy execution, including at least one or more of the following: execution unit availability, number of actions, action interval time, local communication status, control link connectivity status, and device operation permission conditions. Specifically, execution unit availability characterizes whether execution devices such as fans, heaters, dehumidifiers, coolers, or oil pumps are currently operational; the number of actions and action interval time constrain frequent start-stop of execution devices; local communication status and control link connectivity status determine whether control commands can be reliably transmitted via the WAPI intra-station wireless network; and device operation permission conditions characterize manual / automatic mode, interlocking conditions, and other safe execution conditions, which can be specifically expressed as:

[0062]

[0063] In the formula, Indicates the availability of execution units. Indicates the number of times the unit action is executed. This indicates the time interval since the last action. Indicates the status of the local communication or control link. Indicates the running license status.

[0064] 3. Graph Reasoning and Strategy Generation Module.

[0065] like Figure 2 As shown, the graph reasoning and strategy generation module is used to receive the device status of the target device at the current moment. It combines existing historical samples, ontology rules, and historical policy records in the graph data module to perform state matching, risk identification, and policy reasoning, and outputs an autonomous policy corresponding to the current state.

[0066] In a preferred embodiment, the graph reasoning and strategy generation module first determines the current device state. The risk scenario of the target device is identified, and a set of historical state subgraphs corresponding to the risk scenario is selected from the map data. Subsequently, a set of candidate control strategies is determined based on the similarity between the current device state and each historical state subgraph. The risk scenario includes at least one or more of the following: condensation risk scenario, local overheating scenario, high humidity risk scenario, cooling anomaly scenario, and communication anomaly backup operation scenario.

[0067] Furthermore, the step of basing the current device state on the similarity between the current device state and each historical state subgraph, i.e., basing the current device state on the similarity between the current device state and each historical state subgraph, is... With historical state subgraph The matching distance between them can be expressed as:

[0068]

[0069] In the formula, This represents the k-th state parameter at the current time. This represents the k-th state parameter corresponding to the historical state subgraph. This represents the normalized boundary parameter corresponding to the k-th state parameter. It satisfies... The historical state subgraph is determined to match the current device state, where, This indicates the preset matching threshold.

[0070] A set of candidate control strategies can be obtained from the historical state subgraph that matches the current device state. :

[0071]

[0072] In the formula, This represents the control policy associated with the i-th historical state subgraph. When... When, directly output the control strategy as the target autonomous strategy; when When candidate control strategies are ranked based on historical success rate, recovery time, number of actions, and execution cost, the control strategy with the best ranking result is selected as the target autonomous strategy; when When necessary, the local rule generation module is invoked to generate supplementary strategies.

[0073] In a preferred embodiment, the local rule generation module generates a corresponding control strategy based on the current risk scenario, current state parameters, and safety boundary set. Taking a switch cabinet scenario as an example, when the humidity is higher than a preset humidity threshold and the dew point difference is lower than a preset safety threshold, an anti-condensation linkage strategy of turning on the heater and dehumidifier can be generated; when the temperature is higher than a preset temperature threshold, a cooling strategy of turning on the fan can be generated; when the local discharge abnormally increases and the duration exceeds a preset duration, a protection strategy of alarm + backup operation can be generated.

[0074] The target autonomous policy must pass a security verification before being issued and executed. This security verification includes at least one or more of the following: execution unit availability verification, action count verification, action interval verification, local communication status verification, and runtime permission condition verification. Only when the target autonomous policy meets all preset security conditions can it be converted into a control command and issued to the execution unit via the WAPI intranet wireless network.

[0075] In a preferred embodiment, the security verification function can be expressed as:

[0076]

[0077] In the formula, Indicating a goal-oriented autonomous strategy The security verification results Indicates the availability of execution units. Indicates the current number of actions. Indicates the maximum number of actions allowed. This indicates the time interval since the last action. Indicates the minimum action interval. Indicates the status of the local communication or control link. Indicates the running license status. When... When, the target autonomous strategy is determined to be executable; when If the target autonomous strategy cannot be executed directly, return to rematch or generate another strategy.

[0078] 4. Control command issuance and execution module.

[0079] The control command issuance and execution module is used to convert the target autonomous policy that has passed security verification into specific control commands for the execution unit, and then issue them to the corresponding execution unit via the WAPI intranet wireless network to achieve local control of the target device's operating status.

[0080] In a preferred embodiment, the target autonomy strategy After instruction mapping, a set of control instructions is formed. :

[0081]

[0082] In the formula, This indicates a control instruction acting on the j-th execution unit, wherein the control instruction includes at least one or more of the following: start instruction, stop instruction, gear adjustment instruction, mode switching instruction, and alarm linkage instruction.

[0083] The execution unit includes at least one or more of the following: a fan, a heater, a dehumidifier, an air conditioner, a cooler assembly, an oil pump, and an audible and visual alarm device. The execution unit may differ for different target devices; for example, in a switchgear cabinet, the execution unit includes at least a fan, a heater, and a dehumidifier; in a transformer cabinet, the execution unit includes at least a cooler assembly, an oil pump, and an audible and visual alarm device.

[0084] Furthermore, the control command issuance process is completed via the WAPI intra-site wireless network. The autonomous control side generates a control message based on the execution object, execution order, and execution parameters corresponding to the target autonomous strategy, and sends it to the corresponding execution unit via the WAPI intra-site wireless network. After receiving the control message, the execution unit executes the corresponding operation according to the preset action logic. By using the WAPI intra-site wireless network for command transmission, secure access, reliable issuance, and local closed-loop execution of control commands in intra-site scenarios can be guaranteed.

[0085] In a preferred embodiment, the target autonomous strategy and the control commands can be mapped one-to-one or one-to-many. For example, in a switchgear scenario, when the target autonomous strategy is an anti-condensation linkage strategy, the set of control commands... This may include two control commands: turning on the heater and turning on the dehumidifier; when the target autonomous strategy is a local overheating cooling strategy, the set of control commands... This may include one or both of turning on the fan and turning on the air conditioner; when the target autonomous strategy is a minimum operating strategy, the set of control commands... It may include one or both of the following: issuing audible and visual alarms and restricting certain actions.

[0086] To avoid conflicts or unsafe operating conditions caused by concurrent actions of multiple execution units, the control command issuance and execution module can also preset execution order constraints according to the target autonomous strategy. These execution order constraints include at least one or more of the following: sequential execution relationships, mutually exclusive execution relationships, and synchronous execution relationships. For example, in a transformer scenario, the cooler unit can be started first, and then the oil pump can be started depending on the operating status; in a switchgear scenario, the heater can be turned on first, followed by the dehumidifier, or both can be started synchronously according to preset conditions.

[0087] In a preferred embodiment, after the control command is issued, the execution unit returns execution confirmation information to the autonomous control side. This execution confirmation information includes at least one or more of the following: command reception status, execution start status, execution completion status, and execution exception status. The autonomous control side uses the execution confirmation information to determine whether the control command has been correctly issued and executed, thus providing a basis for subsequent execution feedback and closed-loop correction.

[0088] 5. Execution feedback and closed-loop correction module.

[0089] The execution feedback and closed-loop correction module is used to re-collect the operating status information of the target device after the execution unit completes the corresponding control action, and judge the execution effect of the target autonomous strategy based on the status changes before and after execution, thereby realizing the closed-loop correction of the subsequent graph inference results and autonomous strategy output.

[0090] In a preferred embodiment, after the execution unit performs its action, the state sensing unit again collects the state of the target device and uploads the feedback state information to the autonomous control side via the WAPI intra-station wireless network, forming the feedback state of the target device at the next moment:

[0091]

[0092] In the formula, This indicates the device identification information for the next moment. This indicates the state parameter information for the next moment. This indicates the auxiliary constraint information for the next time step.

[0093] Furthermore, the state parameter information This includes at least one or more of the following: post-execution temperature, humidity, partial discharge level, dew point difference, duration of abnormality, oil temperature, hot spot temperature, load rate, and operating status of the actuator. By comparing the changes in state parameters before and after execution, the mitigation effect of the target autonomous strategy on the abnormal state can be evaluated.

[0094] In a preferred embodiment, the execution effect judgment can be achieved through a recovery criterion function. accomplish:

[0095]

[0096] In the formula, when When, the current autonomous strategy is deemed to be effective; when If the current autonomous strategy fails to eliminate the abnormal state, it is necessary to re-perform graph reasoning and strategy generation.

[0097] The preset target range is set according to different equipment types and different operating scenarios. Taking switchgear as an example, the preset target range may include one or more of the following: temperature drops below the preset upper temperature limit, humidity drops below the preset humidity threshold, partial discharge is lower than the preset alarm value, and dew point difference is higher than the preset safety threshold. Taking transformer as an example, the preset target range may include one or more of the following: oil temperature drops to a safe range, hot spot temperature is lower than the preset upper limit, load rate returns to normal, and cooling device operates normally.

[0098] When the execution result indicates that the current autonomous strategy is effective, the execution feedback and closed-loop correction module writes the current state-strategy-result record back into the graph data to form a new historical sample record. When the execution result indicates that the current autonomous strategy is ineffective or only partially effective, the execution feedback and closed-loop correction module sends the current feedback state as a new input state back into the graph reasoning and strategy generation module to rematch or generate subsequent autonomous strategies.

[0099] In a preferred embodiment, the current processing event can form a new sample:

[0100]

[0101] In the formula, Indicates the state before execution. This indicates the current moment's target autonomous strategy. This indicates the feedback status after execution. This indicates a label representing the strategy execution result. The strategy execution result label includes at least one of the following: normal recovery, partial mitigation, and execution failure.

[0102] Furthermore, based on the new sample It can update the map data or historical strategy sub-map library:

[0103]

[0104] In the formula, This represents a newly added historical state subgraph formed by the current event being handled. Through this update method, the autonomous control system can continuously accumulate new state-policy-outcome relationships during long-term operation, thereby improving the accuracy of subsequent similar state matching and autonomous policy output.

[0105] To ensure the continuity of closed-loop correction, in a preferred embodiment, when the feedback state... If the target device is still in an abnormal state and the regenerated autonomous policy is different from the current policy, the system will prioritize the execution of the new autonomous policy. If the regenerated autonomous policy is the same as the current policy but the number of executions has reached the preset limit, the system will switch to a backup operation mode or an alarm mode to avoid new risks caused by frequent actions of the execution unit.

[0106] Through the aforementioned execution feedback and closed-loop correction modules, this application can not only achieve one-time state matching and strategy output, but also continuously correct subsequent control processes based on state changes after execution, enabling the system to have local adaptive capabilities and continuous optimization capabilities in chain-break scenarios, thereby forming a complete in-station closed-loop autonomous control link.

[0107] Example 2:

[0108] Example 2 differs from Example 1 in that the system includes a state perception unit, an execution unit, a WAPI intra-station wireless network, and a device state construction module, a graph reasoning and strategy generation module, and an execution feedback and closed-loop correction module on the autonomous control side. The state perception unit includes a temperature sensor, a humidity sensor, and a partial discharge sensor, and the execution unit includes a fan, a heater, and a dehumidifier.

[0109] This invention discloses a local autonomous control system for substations based on WAPI. In this embodiment, switch cabinet #1 in the substation is used as the target device to construct a closed-loop autonomous control system for the substation based on WAPI.

[0110] like Figure 1As shown, the system mainly includes a WAPI in-station wireless network, a state awareness unit located in the station, an execution unit, an autonomous control terminal, and an autonomous control side device state construction module, a graph reasoning and strategy generation module, and an execution feedback and closed-loop correction module. The state awareness unit is connected to and communicates with the WAPI in-station wireless network, the execution unit is connected to and communicates with the WAPI in-station wireless network, and the autonomous control terminal is connected to and communicates with the WAPI in-station wireless network. The similarities between Embodiment 2 and Embodiment 1 will not be repeated.

[0111] The status sensing unit is deployed around switchgear #1 to collect real-time operating status information of switchgear #1. In this embodiment, the status sensing unit includes at least a temperature sensor, a humidity sensor, and a partial discharge sensor. The execution unit is deployed around switchgear #1 to adjust the operating environment of switchgear #1. In this embodiment, the execution unit includes at least a fan, a heater, and a dehumidifier.

[0112] The operational status information collected by the status awareness unit is uploaded to the autonomous control side via the WAPI intra-site wireless network. The autonomous control side organizes, parses, and expresses the uploaded data to form the current device status information of switchgear #1. This device status information includes at least device identification information, status parameter information, and auxiliary constraint information; wherein, the device identification information is used to identify the target device as switchgear #1, the status parameter information is used to characterize the real-time operational status of switchgear #1, and the auxiliary constraint information is used to characterize the relevant conditions for the execution of the current strategy.

[0113] In this embodiment, the status parameter information includes at least temperature, humidity, dew point difference, partial discharge amount, and abnormal duration; the auxiliary constraint information includes at least execution unit availability, number of actions, action interval time, local communication status, and operation permission status.

[0114] like Figure 2 As shown, when the status sensing unit collects the current status of switch cabinet #1 as follows: temperature 18℃, humidity 92%, dew point difference 1.2℃, partial discharge 0pC, the autonomous control side forms the corresponding equipment status information based on the above status information and further performs graph inference.

[0115] In this embodiment, the graph reasoning and strategy generation module combines historical samples, ontology rules, and historical strategy records to identify risk scenarios for the current equipment status. Since the current status simultaneously satisfies the conditions of humidity being higher than a preset humidity threshold, dew point difference being lower than a preset safety threshold, and partial discharge being within the normal range, the current scenario of switch cabinet #1 is identified as a high humidity + low dew point difference condensation risk scenario, and corresponding risk judgment results are generated, namely high humidity risk, low dew point difference, and condensation risk.

[0116] After identifying the aforementioned risk scenarios, the graph reasoning and strategy generation module further combines historical similar state records in the graph data to output an autonomous strategy corresponding to the current risk scenario. In this embodiment, the output autonomous strategy is: turn on heater #1 and turn on dehumidifier #1. If necessary, the fan can be further controlled in conjunction with environmental changes, but in this embodiment, the coordinated control of the heater and dehumidifier is prioritized.

[0117] Before issuing the autonomous policy, the autonomous control side also performs a security verification on the autonomous policy. The security verification includes at least the following: whether heater #1 and dehumidifier #1 are currently available; whether the number of actions in the recent period has exceeded the preset limit; whether the time interval since the last action meets the minimum interval requirement; whether the WAPI local communication link is normal; and whether the current operating mode allows automatic control. Only when all of the above conditions are met is it allowed to convert the autonomous policy into a control command and issue it via the WAPI intranet wireless network.

[0118] After passing the security verification, the autonomous control side converts the autonomous strategy into specific control commands for the execution units and sends them to the corresponding execution units via the WAPI intra-station wireless network. Upon receiving the control commands, the execution units start heater #1 and dehumidifier #1 to adjust the internal operating environment of switch cabinet #1 in a coordinated manner, thereby suppressing the further development of condensation risk.

[0119] After the execution unit completes its actions, the status awareness unit again collects the status of switch cabinet #1 and uploads the feedback status to the autonomous control side via the WAPI intra-station wireless network. If the feedback result indicates that the humidity of switch cabinet #1 has fallen below the preset humidity threshold and the dew point difference has recovered to above the preset safety threshold, the current autonomous strategy is deemed to be effective, and the status information, strategy information, and execution result corresponding to this handling process are written back into the graph data as one of the historical samples for subsequent graph inference.

[0120] For example, in this embodiment, after a period of time, if the humidity of switch cabinet #1 decreases from 92% to 78%, the dew point difference increases from 1.2℃ to 3.0℃, and the partial discharge remains normal, the autonomous control side determines that the current condensation risk has been eliminated, the current autonomous strategy has been successfully executed, and the current closed-loop control process ends.

[0121] If the feedback indicates that switchgear #1 has not yet returned to the preset target range, for example, if the humidity is still higher than the threshold or the dew point difference is still lower than the safety threshold, the autonomous control side will send the feedback status as a new input to the graph inference module to re-execute risk scenario identification and candidate strategy acquisition in order to re-match or generate subsequent autonomous strategies. If the regenerated autonomous strategy is different from the current strategy, the new autonomous strategy will be executed first. If the regenerated autonomous strategy is the same as the current strategy but the number of actions has reached the preset limit, the system will switch to the minimum operating mode or alarm mode to avoid frequent start-stop of the execution unit.

[0122] In this embodiment, the WAPI intra-station wireless network is used to realize the uploading of status information, the distribution of autonomous policies, and the feedback of execution. This enables switch cabinet #1 to still complete local status perception, graph reasoning, policy output, control execution, and feedback correction by relying on the intra-station autonomous control side even in scenarios such as failure of the upper-level communication link, loss of connection of the master station, or unavailability of the external network, thereby forming a complete closed-loop autonomous control process.

[0123] It should be noted that the temperature threshold, humidity threshold, dew point difference safety threshold, upper limit of number of actions, and minimum action interval used in this embodiment are only illustrative examples. Those skilled in the art can adjust them according to different equipment types, operating scenarios, and actual engineering needs without affecting the substantive content of this application.

[0124] The purpose of Example 2 is to address the problem that existing substation wireless systems typically only have status monitoring and alarm functions when the upstream communication link fails, the master station is disconnected, or the external network is unavailable. They are unable to autonomously complete policy generation, linkage execution, and feedback correction based on the substation wireless network. The aim is to achieve secure uploading of equipment status information, graph reasoning and policy output based on historical policy models, local distribution of control commands, and closed-loop write-back of execution feedback.

[0125] The beneficial technical effects of Example 2 are explained below.

[0126] The status awareness unit collects the operating status information of the target equipment and securely uploads it via the WAPI intra-station wireless network. The autonomous control side constructs the equipment status from the real-time data, performs graph reasoning by combining historical samples, ontology rules, and historical policy records, outputs the corresponding autonomous policy, and sends control commands to the execution unit via the WAPI intra-station wireless network. After execution, feedback status information is collected again to judge the effect of policy execution, and the status, policy, and results are written back to the graph data, realizing closed-loop correction and continuous updating. This application can realize local status awareness, policy generation, linkage execution, and feedback correction of equipment within the substation in scenarios such as failure of the upper-level communication link, loss of connection of the master station, or unavailability of the external network, thereby improving the continuity, reliability, and adaptability of autonomous control within the substation.

[0127] Example 3:

[0128] like Figure 3 As shown, this invention discloses a WAPI-based substation local autonomous control method, and a WAPI-based substation local autonomous control system based on Embodiment 1, which includes the following steps:

[0129] Step S1: Status Acquisition and Secure Upload.

[0130] The operating status information of the target device is collected by the status sensing unit deployed around the target device and uploaded to the autonomous control side via the WAPI station wireless network; the operating status information includes at least one or more of the following: temperature, humidity, partial discharge, dew point difference, abnormal duration, device availability and execution unit operating status.

[0131] Step S2: Device status construction.

[0132] The autonomous control side organizes, parses, and expresses the real-time status data uploaded via the WAPI intranet wireless network to form the device status information of the target device at the current moment. The device status information includes at least device identification information, status parameter information, and auxiliary constraint information. The auxiliary constraint information includes at least one or more of the following: execution unit availability, number of actions, action interval time, local communication status, control link connectivity status, and operating permission conditions.

[0133] Step S3: Risk scenario identification.

[0134] Based on the current equipment status information, identify the risk scenario in which the target equipment is located; the risk scenario includes at least one or more of the following: condensation risk scenario, local overheating scenario, high humidity risk scenario, cooling abnormality scenario, and communication abnormality backup operation scenario.

[0135] Step S4: Graph reasoning and candidate strategy acquisition.

[0136] The current device status information is input into the graph inference module. Combined with historical samples, ontology rules and historical strategy records in the graph data, status matching and graph inference are performed to obtain a set of candidate control strategies corresponding to the current risk scenario.

[0137] Step S5: Determine the autonomous strategy for the target.

[0138] When the candidate control strategy set contains only one candidate strategy, that candidate strategy is directly determined as the target autonomous strategy. When the candidate control strategy set contains multiple candidate strategies, the candidate control strategies are sorted according to historical success rate, recovery time, number of actions, and execution cost, and the candidate strategy with the best sorting result is selected as the target autonomous strategy. When the candidate control strategy set is empty, the local rule generation module is called to generate a supplementary strategy as the target autonomous strategy.

[0139] Step S6: Security verification.

[0140] The target autonomous policy is subjected to security verification, which includes at least one or more of the following: execution unit availability verification, action count verification, action interval time verification, local communication status verification, and operation license condition verification; the control command issuance step is only initiated when the target autonomous policy passes the security verification.

[0141] Step S7: Issuance and execution of control commands.

[0142] The target autonomous strategy that has passed the security verification is converted into specific control instructions for the execution unit and sent to the corresponding execution unit via the WAPI intranet wireless network. The execution unit then performs at least one action from the fan, heater, dehumidifier, air conditioner, cooler group, oil pump, and audible and visual alarm device.

[0143] Step S8: Perform feedback collection.

[0144] After the execution unit completes its actions, the status awareness unit collects the target device's operating status information again and uploads it to the autonomous control side via the WAPI intranet wireless network, forming post-execution feedback status information.

[0145] Step S9: Execution effect judgment.

[0146] The effectiveness of the current autonomous strategy is determined based on the state changes before and after execution. When the state of the target device recovers to the preset target range, the current autonomous strategy is deemed to be effective. When the state of the target device does not recover to the preset target range, it is determined that the abnormal state has not been eliminated after the execution of the current autonomous strategy, and the process re-enters the graph reasoning and candidate strategy acquisition steps.

[0147] Step S10: Closed-loop correction and map update.

[0148] The status information, target autonomous strategy, execution results, and feedback status of the current event are written back to the graph data and historical strategy records to update the basis for subsequent graph reasoning and autonomous strategy output, thereby realizing the continuous correction and iterative optimization of the station's closed-loop autonomous control capabilities.

[0149] In a preferred embodiment, taking a switch cabinet as an example, when the collected humidity is higher than the preset humidity threshold and the dew point difference is lower than the preset safety threshold, it can be identified as a condensation risk scenario, and a linkage autonomous strategy of turning on the heater and turning on the dehumidifier can be output; when the collected temperature is higher than the preset temperature threshold, it can be identified as a local overheating scenario, and a cooling autonomous strategy of turning on the fan can be output; when the local discharge amount abnormally increases and the duration exceeds the preset duration, a protection strategy of alarm + backup operation can be output.

[0150] In a preferred embodiment, taking a transformer as an example, when the collected oil temperature or hot spot temperature is higher than the preset upper limit, it can be identified as an overheating risk scenario, and an enhanced cooling strategy of starting the cooler group or starting the oil pump can be output; when the collected cooling device is in an abnormal operating state and the temperature rise continues, a protection strategy of alarm + backup operation can be output.

[0151] Compared with existing technologies, this application organizes status acquisition, graph inference, strategy output, command issuance, execution feedback and graph update into a unified closed-loop process, enabling the substation to still complete local autonomous control by relying on the WAPI intra-station wireless network in scenarios such as failure of the upper-level communication link, loss of connection of the master station or unavailability of the external network; at the same time, by continuously writing the execution results back to the graph data and historical strategy records, the system has the ability to continuously learn and continuously optimize.

[0152] Compared with the prior art, this application has at least the following beneficial effects.

[0153] 1. By constructing a secure wireless communication link within the substation through the WAPI intra-station wireless network, the status sensing unit and the execution unit can maintain local communication capabilities even in scenarios such as failure of the upper-level communication link, loss of connection to the master station, or unavailability of the external network. This ensures the reliable uploading of target device status information and the secure issuance of control commands, thereby improving the continuity and reliability of autonomous control within the substation.

[0154] 2. By structuring the target device's operating status information, a device status information system is formed, including device identification information, status parameter information, and auxiliary constraint information. This enables the status data of different target devices and different operating scenarios to be uniformly integrated into the subsequent graph inference process, thereby improving the compatibility and scalability of autonomous control for multiple types of devices and multiple scenarios.

[0155] 3. By combining historical samples, ontology rules, and historical policy records through graph reasoning and policy generation mechanisms, corresponding autonomous policies are matched or generated based on the current device status. This avoids the problem of over-reliance on single threshold triggers or human experience judgment in existing technologies, and makes the autonomous policy output process more interpretable and adaptable.

[0156] 4. By setting up a security verification step, the availability, number of actions, action interval, local communication status, and operating permission conditions of the execution unit are verified before the autonomous strategy is issued and executed. This can effectively avoid problems such as frequent start-up and shutdown of the execution unit, control command conflicts, or malfunctions when the operating conditions are not met, thereby improving the security and stability of the autonomous control process within the station.

[0157] 5. By implementing feedback and closed-loop correction mechanisms, the target device's operating status information is re-collected after the execution unit completes its actions, and the execution effect of the current autonomous strategy is judged based on the status changes before and after execution. When the current autonomous strategy fails to achieve the preset goal, the graph reasoning and strategy generation process can be re-triggered, thereby enabling the system to have local adaptive correction capabilities in the event of a chain failure.

[0158] 6. By continuously writing the status information, autonomous strategy, execution result, and feedback status corresponding to the current event back into the graph data and historical strategy records, the system can continuously accumulate new status-strategy-result relationships during long-term operation, thereby improving the accuracy of subsequent similar status matching and autonomous strategy output, and realizing the continuous optimization and iterative update of the station's closed-loop autonomous control capabilities.

Claims

1. A local autonomous control system for substations based on WAPI, characterized in that: It includes a WAPI on-site wireless network, a state awareness unit, an execution unit, and an autonomous control terminal located within the site, as well as a device state construction module, a graph reasoning and policy generation module, and an execution feedback and closed-loop correction module on the autonomous control side. The state awareness unit, the execution unit, and the autonomous control terminal are all connected to and communicate with the WAPI on-site wireless network. The device status construction module is used to organize and express the status of target devices uploaded in real time via the WAPI intranet wireless network, thereby obtaining a structured device status. ; The graph reasoning and strategy generation module is used to determine the current device status. By combining existing historical samples, ontology rules and historical policy records, state matching, risk identification and policy reasoning are performed to obtain the corresponding autonomous policy. The target autonomous policy is subject to security verification before it is issued and executed. After the security verification is passed, it is issued and executed. The execution feedback and closed-loop correction module is used to correct subsequent graph inference results and autonomous strategies based on changes in device status after execution.

2. The substation local autonomous control system based on WAPI according to claim 1, characterized in that: The state sensing unit is deployed around the target device to collect the target device's state in real time; the state sensing unit includes a temperature sensor, a humidity sensor, and a partial discharge sensor; the execution unit is deployed around the target device to adjust the target device's operating environment; the execution unit includes a fan, a heater, and a dehumidifier; the device state includes device identification, status, and auxiliary constraints. The device identification includes device name, device number, device type, and region; the status includes temperature, humidity, and partial discharge level; the auxiliary constraints include execution unit availability, number of actions, action interval time, local communication status, and operating permission status.

3. The substation local autonomous control system based on WAPI according to claim 1, characterized in that: In the graph reasoning and strategy generation module, based on the current device status... Identify the risk scenarios in which the target device is located, and filter the set of historical state subgraphs corresponding to the risk scenarios in the map data; obtain a set of candidate control strategies based on the similarity between the current device state and each historical state subgraph; the risk scenarios include condensation risk scenarios, local overheating scenarios, high humidity risk scenarios, cooling abnormality scenarios, and communication abnormality backup operation scenarios.

4. A substation local autonomous control system based on WAPI according to claim 1, characterized in that: In the graph reasoning and strategy generation module, the security verification includes execution unit availability verification, action count verification, action interval time verification, local communication status verification, and operation license condition verification.

5. A substation local autonomous control system based on WAPI according to claim 1, characterized in that: In the execution feedback and closed-loop correction module, when the execution result shows that the current autonomous strategy is effective, the current state-strategy-result record is written back to the graph data to form a new historical sample record; when the execution result shows that the current autonomous strategy is ineffective or only partially effective, the current feedback state is sent back to the graph reasoning and strategy generation module as a new input state to obtain subsequent autonomous strategies.

6. A substation local autonomous control system based on WAPI according to claim 1, characterized in that: In the execution feedback and closed-loop correction module, when the feedback state... If the target device is still in an abnormal state and the regenerated autonomous policy is different from the current policy, the new autonomous policy will be executed first; if the regenerated autonomous policy is the same as the current policy but the number of executions has reached the preset limit, it will switch to the minimum operating mode or alarm mode.

7. A substation local autonomous control system based on WAPI according to claim 1, characterized in that: It also includes a data acquisition module, which is used to collect the status of the target device via the WAPI in-station wireless network.

8. A substation local autonomous control system based on WAPI according to claim 1, characterized in that: It also includes a control command issuance and execution module, which is used to convert autonomous policies into specific control commands for execution units and issue them via the WAPI intranet wireless network.

9. A local autonomous control method for substations based on WAPI, characterized in that: The substation local autonomous control system based on WAPI as described in claim 1 includes the following steps: Step S1: Collect the status of the target device in real time and securely upload it to the autonomous control terminal within the station via the WAPI intra-station wireless network; Step S2: Obtain the real-time device status based on the real-time target device status; Step S3: Identify the current risk scenario based on the current device status; Step S4: Combine the current device status with historical samples, ontology rules and historical policy records in the graph data, perform state matching and graph reasoning to obtain a set of candidate control policies corresponding to the current risk scenario; Step S5: Obtain the target autonomous policy based on the candidate control policy set; Step S6: Perform security verification on the target autonomous policy. If the target autonomous policy passes the security verification, proceed to step S7. Step S7: Convert the target autonomous policy that has passed the security verification into specific control instructions for the execution unit, and send them to the corresponding execution unit via the WAPI intranet wireless network; Step S8: After the execution unit completes its action, the status sensing unit collects the status of the target device again and uploads it to the autonomous control terminal via the WAPI intra-station wireless network to form the feedback status after execution; Step S9: Determine the execution effect of the current target autonomous strategy based on the state changes before and after execution; when the target device state recovers to the preset target range, it is determined that the current autonomous strategy is effective and step S10 is executed; when the target device state does not recover to the preset target range, it is determined that the abnormal state has not been eliminated after the execution of the current autonomous strategy and step S4 is executed. Step S10: Write the status, target autonomous strategy, execution result and feedback status corresponding to the current event back to the graph data and historical strategy records to update subsequent graph reasoning and autonomous strategies, realize dynamic autonomous control within the station and closed-loop correction.

10. A substation local autonomous control method based on WAPI according to claim 9, characterized in that: In step S5, when the candidate control strategy set contains only one candidate strategy, the candidate strategy is directly determined as the target autonomous strategy; when the candidate control strategy set contains multiple candidate strategies, the candidate control strategies are sorted according to historical success rate, recovery time, number of actions and execution cost, and the candidate strategy with the best sorting result is selected as the target autonomous strategy; when the candidate control strategy set is empty, the local rule generation module is called to generate a supplementary strategy as the target autonomous strategy.

Images