A substation multi-source intelligent anti-misoperation scheduling method and system

Abstract

The application provides a substation multi-source intelligent anti-misoperation scheduling method and system, multi-source heterogeneous data acquisition and preprocessing, dynamic identification of intermediate states, construction of a device state time sequence model based on a long short-term memory network and a graph neural network, real-time identification of intermediate states in switching operations, intelligent interlocking logic judgment and updating, adoption of a hybrid rule engine, real-time interlocking logic judgment in combination with pre-defined interlocking logic rules, anti-misoperation decision and early warning, generation of operation suggestions based on a deep reinforcement learning model, real-time early warning for high-risk operations, and prompting of operation personnel to correct steps through an augmented reality interface, so that the application can effectively perform anti-misoperation scheduling and interlocking judgment in an intermediate state caused by insufficient switching closing, and improve the safety and reliability of substation operation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent anti-misoperation of power systems, and in particular to a substation multi-source intelligent anti-misoperation scheduling method and system. BACKGROUND

[0002] With the rapid development of smart substations, the number and complexity of substation equipment are increasing, and the traditional substation anti-misoperation system mainly relies on static rule bases and manual experience, which is difficult to cope with complex dynamic scenarios. Switching operation is an important link in the operation and maintenance of substations, and insufficient switching closing may cause the equipment to be in an intermediate state, causing misoperation risks and hidden dangers to the safe and stable operation of the power grid. The existing anti-misoperation system mainly relies on fixed rules and logic for judgment, and has poor adaptability to complex and variable substation operating states, and has deficiencies in handling the intermediate state caused by insufficient switching closing and interlocking judgment logic, making it difficult to adapt to changing power grid operation modes.

[0003] The patent application with application number 2024117969548 discloses an invention patent of an intelligent scheduling operation ticket anti-misoperation management and control system driven by power scheduling business data, which performs anti-misoperation verification through conditions such as the on or off state of circuit breakers and disconnectors, and whether the grounding wire is removed or not. However, in actual operation in substations, the above-mentioned equipment often enters an unexpected intermediate state (such as a circuit breaker being open but a disconnector not being completely closed) due to insufficient switching closing, and the above-mentioned anti-misoperation verification cannot identify the risk events caused by misoperation.

[0004] The patent application with application number 2024113963576 discloses an invention patent of an anti-misoperation protection system for control elements in a power grid, which includes a plurality of anti-misoperation interlocking devices corresponding to a plurality of control elements in the power grid that need to be interlocked and controlled, and the anti-misoperation interlocking devices are used to collect the current state information of the corresponding control elements, transmit the current state information to a centralized control device, and perform anti-misoperation interlocking protection on the control elements based on the target control instructions from the centralized control device. However, the above-mentioned anti-misoperation interlocking devices mainly consider the cascading of devices as a physical logical relationship, and do not consider dynamic interlocking strategies related to power supply energy-saving strategies and power load rates, such as adjusting the grounding knife switch locking strategy according to real-time load.

[0005] Based on the above analysis, the present application aims to provide a substation multi-source intelligent anti-misoperation scheduling method and system that can integrate multiple data sources and use artificial intelligence technology to accurately prevent misoperation in substation scheduling operations, especially in the intermediate state caused by insufficient switching closing, to effectively prevent misoperation scheduling and interlocking judgment, and improve the safety and reliability of substation operation. SUMMARY

[0006] In view of the deficiencies of the prior art, the technical problem to be solved by the present application is to provide a substation multi-source intelligent anti-misoperation scheduling method and system to solve the problems in the prior art.

[0007] To solve the above technical problems, the first aspect provides a substation multi-source intelligent anti-misoperation scheduling method, comprising the following steps:

[0008] S100, multi-source heterogeneous data acquisition and preprocessing, collecting real-time state data, historical operation records, environmental parameters, video monitoring streams and power grid topology information of the substation, and constructing a dynamic knowledge graph;

[0009] S0200, intermediate state dynamic identification, constructing a device state time series model based on a long short-term memory network and a graph neural network, and identifying an intermediate state in the switching operation in real time, wherein the intermediate state includes a half-closed and ungrounded device working state in the switching operation;

[0010] S300, intelligent interlocking logic judgment and updating, adopting a hybrid rule engine, combining pre-defined interlocking logic rules for real-time interlocking logic judgment, wherein the interlocking logic judgment rules include real-time energy topology information and energy consumption dynamic data adjustment locking conditions, the locking conditions are optimized through real-time topology analysis, and redundant locking is reduced;

[0011] S400, anti-misoperation decision and early warning, generating operation suggestions based on a deep reinforcement learning model, real-time early warning for high-risk operations, and prompting an operator to correct steps through an augmented reality interface.

[0012] Optionally, the step S100, multi-source heterogeneous data acquisition and preprocessing, comprises:

[0013] S101, Internet of Things sensor data acquisition, real-time monitoring of the closing position change of the circuit breaker contact through a high-precision position sensor installed on the circuit breaker, acquisition of pressure data of the circuit breaker operating mechanism by a pressure sensor, and monitoring of vibration data of the circuit breaker in the operation process by a vibration sensor;

[0014] S102, SCADA system data acquisition, acquisition of current, voltage, active power and reactive power data of the circuit breaker, and acquisition of energy storage state signals and opening and closing indication signals of the circuit breaker;

[0015] S103, video monitoring device data acquisition, real-time acquisition of video monitoring data of the circuit breaker and disconnector by a high-definition camera and an infrared thermal imager, display of the closing action speed of the circuit breaker in the video, and display of temperature data at the circuit breaker contact by the infrared thermal imager;

[0016] S104: Data preprocessing, the data cleaning module eliminates the error data points in the position sensor data caused by device vibration beyond the range, the pressure sensor data is processed by sliding window filtering, the filtering window size is 5, and the breaker position, current, closing speed, vibration amplitude, energy and energy consumption data are normalized to the interval [0, 1] by the normalization processing module.

[0017] Optionally, the step S200, the intermediate state dynamic identification comprises:

[0018] S201: LSTM model analysis, input the preprocessed breaker position data sequence into the LSTM-based intermediate state recognition model, analyze the current breaker position data sequence, and output the closing speed and position change curve data;

[0019] S202: GNN topology modeling, combining the topology relationship diagram of the substation equipment, the GNN model analyzes the electrical connection relationship between the circuit breaker and the disconnector, and the association relationship with the bus, the power transmission line and other equipment, and determines the closing state of the circuit breaker through impedance abnormality in the electrical connection;

[0020] S203: Intermediate state judgment: the closing position does not reach 95%, it is judged that the circuit breaker is in the intermediate state, and the misoperation risk assessment is performed.

[0021] Optionally, the real-time energy topology information comprises: when the fan and the fire switch, if the fan output power is insufficient and the fire starting time does not reach the energy saving requirement, the fire starting operation is locked; before the fan grid operation, it is necessary to ensure that the speed is in the range of 90%-110% of the rated speed, and the bus voltage is stable within ±5% of the rated voltage, otherwise the grid operation is locked; the energy consumption dynamic data comprises: in the peak electricity price period, if the energy storage system has sufficient power and the photovoltaic power generation power is greater than or equal to 50% of the real-time load, the power grid access operation is locked.

[0022] Optionally, the deep reinforcement learning model, the input is the current state of the substation, including the device closing position, real-time interlocking topology information, load data, energy supply state.

[0023] Optionally, the deep reinforcement learning model, the output is the Q value of each possible operation:

[0024] Q(s,a)←Q(s,a)+ α×[R(s,a)+ γ×maxQ(s',a')-Q(s,a)]

[0025] Wherein Q(s,a) is the current Q value of state s under action a;

[0026] α is the learning rate, 0 < α < 1;

[0027] R(s, a) is the immediate reward obtained after performing action a in state s;

[0028] γ is a discount factor for balancing future rewards and immediate rewards, 0 < γ < 1;

[0029] maxQ(s', a') is the Q value corresponding to the action pair with the maximum Q value among all executable actions after performing action a into new state s'.

[0030] Optionally, the real-time warning form includes highlighting the closing position of the circuit breaker in red, prompting the operator in the form of animation for the unfinished closing steps, and remotely warning through short message and mobile phone APP push message.

[0031] In a second aspect, the present application further provides a substation multi-source intelligent anti-misoperation scheduling system, comprising:

[0032] The multi-source data acquisition unit is used for multi-source heterogeneous data acquisition and preprocessing, and is used for collecting real-time state data, historical operation records, environmental parameters, video monitoring streams and power grid topology information of the substation, and constructing a dynamic knowledge graph.

[0033] The intermediate state determination module is based on a long short-term memory network and a graph neural network, constructs a device state time sequence model, and identifies the intermediate state in the switching operation in real time, wherein the intermediate state includes the half-closed and ungrounded device working state in the switching operation.

[0034] The interlocking logic real-time updating module is used for intelligent interlocking logic determination and updating, adopts a hybrid rule engine, and combines pre-defined interlocking logic rules for real-time interlocking logic determination, wherein the interlocking logic determination rules include real-time energy topology information and energy consumption dynamic data adjustment locking conditions, the locking conditions are optimized through real-time topology analysis, and the redundant locking is reduced.

[0035] The strategy generation and warning module is used for anti-misoperation decision and warning, generates operation suggestions based on a deep reinforcement learning model, warns in real time for high-risk operations, and prompts the operator to correct the steps through an augmented reality interface.

[0036] In a third aspect, the present application further provides an electronic device, comprising:

[0037] One or more processors;

[0038] A memory; and one or more programs stored in the memory, the one or more programs including instructions for executing any of the above-mentioned substation multi-source intelligent anti-misoperation scheduling methods.

[0039] In a fourth aspect, the present application also provides a computer readable storage medium comprising one or more programs for execution by one or more processors of an electronic device, the one or more programs comprising instructions for performing any of the above substation multi-source intelligent anti-misoperation scheduling methods.

[0040] Compared with the prior art, the technical scheme of the present application has the following beneficial effects:

[0041] 1. Breakthrough the limitations of traditional final state detection, real-time identification and early warning of unexpected intermediate state, reduce the probability of misoperation.

[0042] 2. Fusion of rule engine and reinforcement learning, improve the flexibility and adaptability of interlocking strategy, adjust the locking condition through real-time energy topology information and energy consumption dynamic data, realize the improvement of energy saving and power supply efficiency.

[0043] 3. Multi-modal data fusion: integrate multi-source heterogeneous data through knowledge graph, support accurate decision-making in complex scenarios. BRIEF DESCRIPTION OF DRAWINGS

[0044] Figure 1 A substation multi-source intelligent anti-misoperation scheduling method flowchart is provided for the embodiments of the present application.

[0045] Figure 2 A substation multi-source intelligent anti-misoperation scheduling system flowchart is provided for the embodiments of the present application. DETAILED DESCRIPTION

[0046] Obviously, many modifications and variations of the present application can be made by those skilled in the art based on the purpose of the present application, which belong to the protection scope of the present application.

[0047] Those skilled in the art can understand that, unless specifically stated, the singular forms "a", "an" and "the" used herein also include the plural forms. It should be further understood that the use of the phrase "comprising" in the specification of the present application means that the features, integers, steps, operations, elements and / or components described exist, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when an element, component is said to be "connected" to another element or component, it can be directly connected to the other element or component, or there can be intermediate elements or components. The phrase "and / or" used herein includes any one of the associated listed items and all combinations of the associated listed items.

[0048] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application. Embodiment 1

[0049] With reference to Figure 1 For an embodiment of the present application, the embodiment provides a multi-source intelligent anti-misoperation scheduling method for a substation, comprising the following steps.

[0050] S100 multi-source heterogeneous data acquisition and preprocessing: acquiring real-time state data, historical operation records, environmental parameters, video monitoring streams and power grid topology information of the substation, and constructing a dynamic knowledge graph.

[0051] In an optional embodiment, in a certain 500 kV substation, an operator performs a switching operation of “switching the high-voltage side circuit breaker of the main transformer from operation to maintenance”:

[0052] S100 multi-source heterogeneous data acquisition and preprocessing specifically comprises the following steps.

[0053] S101: Internet of Things sensor data acquisition: a high-precision position sensor (precision 0.01 mm) installed on the circuit breaker monitors the closing position change of the circuit breaker contact in real time, and the collected position data is 0.02 mm. A pressure sensor (range 0-1000 Pa) collects pressure data of the circuit breaker operating mechanism, and the current pressure is 500 Pa. A vibration sensor monitors the vibration of the circuit breaker during operation, and the collected vibration frequency is 13.5 Hz.

[0054] S102: SCADA system data acquisition: the SCADA system collects the current data of the circuit breaker as 200 A, the voltage data as 500 kV, the active power as 120 MW, and the reactive power as 80 Mvar. At the same time, the energy storage state signal of the circuit breaker is collected as “energy stored”, and the closing and opening indication signal is collected as “closing”.

[0055] S103: video monitoring device data acquisition: a high-definition camera and an infrared thermal imager collect video monitoring data of the circuit breaker and the disconnector in real time. The closing action of the circuit breaker in the video is slightly slow, and the infrared thermal imaging shows that the temperature of the circuit breaker contact is slightly higher than that of other parts, and the temperature value is 65℃.

[0056] S104: Data preprocessing: The data cleaning module eliminates error data points in the position sensor data caused by device vibration that exceed the range, and performs sliding window filtering on the pressure sensor data with a filter window size of 5, which improves the smoothness of the processed pressure data. The normalization processing module normalizes the circuit breaker position data to the [0, 1] interval, and the normalized position data is 0.02. The current data is normalized to the [0, 1] interval, and the normalized current data is 0.4. The feature extraction module calculates the closing speed of the circuit breaker as 0.02 mm / s, the closing synchronization deviation as 0.001 s, and the vibration amplitude as 0.05 mm.

[0057] S0200 Intermediate state dynamic identification: Based on long short-term memory network (LSTM) and graph neural network (GNN), a device state time series model is constructed to identify the intermediate state in the switching operation in real time.

[0058] In an optional embodiment, the intermediate state dynamic model construction specifically includes:

[0059] S201: LSTM model analysis: The preprocessed circuit breaker position data sequence is input into the LSTM-based intermediate state recognition model. The LSTM model is trained on a large amount of historical operation data to learn the time series features of normal closing operations. The model analyzes the current circuit breaker position data sequence and finds that its closing speed is slower than normal operation and the position change curve has abnormal fluctuations.

[0060] S202: GNN topology modeling: Combined with the topology relationship graph of the substation equipment, the GNN model analyzes the electrical connection relationship between the circuit breaker and the disconnector, as well as the association relationship with devices such as busbars and transmission lines. The model finds that there is a certain impedance abnormality between the circuit breaker and the disconnector, which may affect the closing state of the circuit breaker.

[0061] S203: Intermediate state judgment: The model comprehensively judges that the circuit breaker is in an intermediate state (the closing position has not reached 95%) and there is a risk of misoperation (the probability of mis-closing is 85%). The system records this judgment result and prepares for subsequent risk assessment and warning operation.

[0062] S300 Intelligent interlocking logic judgment and update: A hybrid rule engine is used to make real-time interlocking logic judgments combined with predefined interlocking logic rules, which include real-time energy topology information and energy consumption dynamic data adjustment locking conditions. Through real-time topology analysis, the locking conditions are optimized to reduce redundant locking.

[0063] Among them, the intelligent interlocking logic judgment and update specifically includes:

[0064] In an optional embodiment, the anti-misoperation based on scheduling operation in multiple energy supply modes is prevented in a certain smart substation, which adopts a hybrid power supply mode of wind power generation and thermal power generation, while the energy saving demand needs to be met. The substation is equipped with two wind turbine generators (wind turbine A and wind turbine B), one thermal power generator (thermal generator C), and related power distribution equipment.

[0065] High-precision speed sensors and power sensors are installed on the wind turbine A and the wind turbine B to monitor the real-time speed of 1200 rpm and 1150 rpm, and the output power of 1.2 MW and 1.0 MW, respectively. A steam pressure sensor and a temperature sensor are installed on the thermal generator C to monitor the real-time steam pressure of 1500 kPa and the temperature of 500℃, and the output power of 3.5 MW.

[0066] The real-time energy consumption data of each generator set is collected by the energy consumption monitoring equipment, including the unit power generation energy consumption of the wind turbine A and the wind turbine B of 0.12 kg of standard coal / kWh and 0.13 kg of standard coal / kWh, and the unit power generation energy consumption of the thermal generator C of 0.35 kg of standard coal / kWh.

[0067] The substation operation data is collected, and the real-time load data of the substation collected by the SCADA system is 5.0 MW, the bus voltage is 220 kV, and the state signals of each circuit breaker and disconnector are normal.

[0068] The data cleaning module eliminates the error data points in the wind turbine speed sensor data that exceed the range due to equipment vibration, and performs sliding window filtering processing on the thermal generator steam pressure sensor data with a filter window size of 5, thereby improving the smoothness of the processed pressure data. The normalization processing module normalizes the output power data of each generator set to the interval [0, 1], and the normalized power of the wind turbine A, the wind turbine B and the thermal generator C is 0.4, 0.33 and 0.7, respectively. The feature extraction module calculates the power fluctuation feature, the start-up time feature, etc. of each generator set, and the power fluctuation amplitude of the wind turbine A is 0.05 MW, and the fluctuation frequency is 0.2 Hz, and the start-up time of the thermal generator C is 1.2 hours.

[0069] Predefined interlocking logic rules: S301 establishes the following interlocking logic judgment rules according to the operation logic of multiple energy supply equipment and the energy saving demand:

[0070] When the wind turbine and the thermal generator are switched, if the output power of the wind turbine is insufficient and the start-up time of the thermal generator does not meet the energy saving requirement, the start-up operation of the thermal generator is locked.

[0071] Before the wind turbine is connected to the grid, it is necessary to ensure that the speed of the wind turbine is within 90%-110% of the rated speed, and the bus voltage is stable within ±5% of the rated voltage, otherwise the grid connection operation is locked.

[0072] Real-time interlocking logic judgment: According to the current collected data and the predefined interlocking logic rules for judgment:

[0073] The sum of the actual output power of fan A and fan B is 2.2 MW, compared with the real-time load of 5.0 MW, the remaining load of 2.8 MW needs to be borne by fan C. But the current start-up time of fan C does not reach the energy-saving requirement of 1.5 hours, so the start-up operation of fan C is locked out, and a warning prompt is issued to the dispatcher.

[0074] The speed of fan A is 1200 rpm, which is outside the range of 90%-110% (900 rpm-1100 rpm) of its rated speed 1000 rpm, and the bus voltage 220 kV is stable within ±5% (209 kV-231 kV) of the rated voltage 220 kV, so the parallel operation of fan A is locked out, and the operator is prompted to check the reason for the abnormal fan speed.

[0075] In another alternative embodiment, in the energy-saving operation mode of a certain substation, the substation adopts a hybrid power supply mode of photovoltaic power generation and municipal power supply, and is equipped with an energy storage system to improve energy utilization efficiency.

[0076] The real-time load data of the substation, the photovoltaic power generation power, the charge and discharge state of the energy storage system, and the peak and valley electricity price information of the power grid are collected. At a certain time, the real-time load is 3.0 MW, the photovoltaic power generation power is 1.8 MW, the energy storage system is in the charging state, the power is 60%, and the current is in the peak electricity price period, the electricity price is high.

[0077] According to the real-time peak and valley electricity price and the carbon emission limit, the energy-saving demand of the substation is evaluated. Since it is in the peak electricity price period and the carbon emission limit is strict, the system determines that at this time the energy-saving operation strategy should be given priority to, and the photovoltaic power generation and energy storage system should be used as much as possible to reduce the use of municipal power supply, so as to reduce energy cost and carbon emission.

[0078] S302 According to the energy-saving demand, the following interlocking logic judgment rules are established:

[0079] In the peak electricity price period, if the energy storage system has sufficient power and the photovoltaic power generation power is greater than or equal to 50% of the real-time load, the municipal power supply access operation is locked out, and the energy storage system and photovoltaic power generation are used preferentially.

[0080] For the operation of non-critical equipment, if its energy consumption is high and does not affect the safe and stable operation of the power grid, the related operation is locked out on the premise of meeting the energy-saving demand.

[0081] According to the energy-saving demand evaluation result, the interlocking logic judgment rules are optimized:

[0082] The current energy storage system has 60% of its capacity, which is sufficient. The photovoltaic power generation is 1.8 MW, which is greater than or equal to 50% (1.5 MW) of the real-time load of 3.0 MW. Therefore, the operation of locking the power grid access is closed to ensure the priority use of energy storage systems and photovoltaic power generation for power supply, achieving the goal of energy saving.

[0083] For a non-critical device with high energy consumption (such as an auxiliary heating device), the system locks its start operation under the premise of meeting the operation safety to reduce unnecessary energy consumption.

[0084] S303 Based on the real-time state of the device and the interlocking rules in the knowledge graph, the scheduling operation instruction is judged in real time. For example, in the case of insufficient closing of the disconnecting switch, the related circuit breaker closing operation is automatically locked to prevent short circuit accidents caused by misoperation; when the bus is in an intermediate state, the operation of other devices related to the bus is locked. The interlocking judgment logic uses a combination of logic gate circuits and state machines to monitor and judge the real-time state of the device. Once an interlocking condition is found, the corresponding interlocking action is triggered immediately. At the same time, the interlocking judgment logic also has self-checking and self-recovery functions, which can automatically release the interlocking state when the device state returns to normal.

[0085] S400 Anti-misoperation decision and early warning: based on deep reinforcement learning (DRL) to generate operation suggestions, real-time warning for high-risk operations, and through augmented reality (AR) interface to prompt operators to correct steps.

[0086] Optimize the scheduling strategy through deep reinforcement learning and build a neural network model: design a deep Q network (DQN) model, input the current state of the substation, and output the Q value of each possible operation.

[0087] Input layer: input the current state of the substation, including device state (such as circuit breaker position, disconnecting switch position, etc.), real-time topology information, load data, energy supply state, etc. The design of the state space should accurately reflect the operation of the substation and provide sufficient information for the model to make decisions.

[0088] Hidden layer: usually use a multi-layer neural network structure, such as 2-3 layers of hidden layers, each layer containing a certain number of neurons (such as 64, 128, 256 neurons). The activation function can use ReLU (Rectified Linear Unit), which can introduce nonlinearity and enable the model to learn more complex feature representations.

[0089] Output layer: output the Q value of each possible operation. Each operation corresponds to a Q value, representing the expected return of executing the operation in the current state. The number of neurons in the output layer is consistent with the size of the action space.

[0090] In an optional embodiment, assume that in a multi-source intelligent anti-misoperation scheduling system of a certain substation, the current state is that the disconnector is not fully closed (intermediate state), and the scheduling system needs to decide whether to allow the circuit breaker to close operation.

[0091] The current state s of the system is that the disconnector is not fully closed, and the action a is performed: allow the circuit breaker to close. According to the anti-misoperation rule, closing at this time may cause equipment damage or misoperation, so the immediate reward R(s,a) obtained is -10 (negative reward).

[0092] Optionally, the reward function considers the operating safety margin of the equipment: reward is given for operations that keep the equipment state within the safe range after operation; lower reward or penalty is given for operations that bring the equipment close to the safety limit. For example, for operations where the equipment closing position is greater than 90%, the reward value can be set to +1; for operations where the equipment closing position is less than 70%, the reward value can be set to -2. This helps the agent to maintain sufficient safety margin during scheduling and reduce the risk of equipment failure. The specific form of the reward function can be a weighted sum that considers the above factors comprehensively:

[0093] R = w1 × R correct + w2 × R efficiency + w3 × R energy + w4 × R stability + w5 × R safety

[0094] where R correct represents the operation correctness reward, R efficiency represents the operation efficiency reward, R energy represents the energy utilization efficiency reward, R stability represents the system stability reward, R safety represents the safety margin reward; w1, w2, w3, w4, w5 are the corresponding weight coefficients, which are adjusted according to the specific application scenario and requirements to balance the relative importance of each reward factor.

[0095] After performing action a, the new state s' is entered, which may be a device failure state. The system calculates the Q values of each possible action (such as emergency opening, equipment maintenance, etc.) in the new state s', and assumes that the maximum Q value is -20.

[0096] Assume that the current learning rate α is 0.1 and the discount factor γ is 0.9. According to the Q value update formula, update the Q value of the action of allowing the circuit breaker to close in the current state:

[0097] Q(s,a)←Q(s,a)+0.1×[-10+0.9×(-20)-Q(s,a)]

[0098] Through this process, the system continuously updates the Q values of each state-action pair, gradually learning the optimal scheduling strategy to achieve the goal of substation multi-source intelligent anti-misoperation scheduling, effectively avoiding the risk of misoperation caused by intermediate equipment states.

[0099] In an optional embodiment, the system highlights the unfinished steps through the AR interface, pops up warning information such as "The closing position of the circuit breaker does not meet the requirements, there is a risk of misoperation!", and locks the subsequent operation permission. On the AR interface, the closing position of the circuit breaker is highlighted in red, and the unfinished closing steps are animated to prompt the operator. At the same time, the interface shows that the mis-closing probability is 85%, reminding the operator to pay attention.

[0100] The system synchronously updates the dynamic knowledge graph, marks the circuit breaker as "high-risk intermediate state", and associates the state information of related devices (isolators, buses, etc.). In the knowledge graph, the state attribute of the circuit breaker node is updated to "intermediate state", and the state attributes of the isolator and bus nodes connected to it are also updated accordingly to reflect the current device state and potential risks.

[0101] The system automatically sends a notification to the operation and maintenance personnel through SMS and mobile APP push message: "The high-voltage side circuit breaker of the main transformer is in an intermediate state, please review it in time!" The SMS content includes the specific location of the circuit breaker, the current state, and the recommended measures. The mobile APP push message is accompanied by detailed state data and historical operation records of the circuit breaker, making it easy for operation and maintenance personnel to quickly understand the on-site situation.

[0102] Based on the deep reinforcement learning model, combined with the current device state and the anti-misoperation rules in the knowledge graph, the risks and benefits of different decision schemes are evaluated. The model takes into account the consequences of the circuit breaker being in an intermediate state, such as short-circuit faults, equipment damage, etc., as well as the impact of different decision schemes on the operation of the power grid.

[0103] The model generates an anti-misoperation decision suggestion: "Suggest suspending subsequent power transmission operations, arrange on-site personnel to check the circuit breaker closing mechanism, and continue the operation after confirming that the device state is normal." The decision suggestion is displayed to the dispatch personnel through the human-computer interaction terminal, and provides detailed inspection steps and precautions to guide the operation and maintenance personnel to conduct on-site inspection and treatment.

[0104] In other optional embodiments of the present application, the active power and reactive power data collected by the SCADA system can also be used to calculate the load rate of each circuit breaker. Based on the updated interlocking logic judgment rules, the system automatically adjusts the closing sequence of the circuit breakers to avoid the risk of overload caused by operation. The system suggests closing the circuit breakers with lower load first, and then gradually closing the other circuit breakers to ensure stable operation of the power grid. On the human-computer interaction terminal, the system displays the adjusted closing sequence and provides detailed operation steps and precautions.

[0105] The embodiment also provides a substation multi-source intelligent anti-misoperation scheduling system, comprising:

[0106] The multi-source data acquisition unit is used for multi-source heterogeneous data acquisition and preprocessing, and is used for collecting real-time state data, historical operation records, environmental parameters, video monitoring streams and power grid topology information of the substation, and constructing a dynamic knowledge graph;

[0107] The intermediate state determination module is based on a long short-term memory network and a graph neural network, constructs a device state time sequence model, and identifies an intermediate state in the switching operation in real time, wherein the intermediate state includes a half-closed and ungrounded device working state in the switching operation.

[0108] The interlocking logic real-time updating module is used for intelligent interlocking logic judgment and updating, adopts a hybrid rule engine, and combines a pre-defined interlocking logic rule to perform real-time interlocking logic judgment, wherein the interlocking logic judgment rule includes real-time energy topology information and energy consumption dynamic data adjustment locking conditions, the locking conditions are optimized through real-time topology analysis, and the redundant locking is reduced.

[0109] The strategy generation and early warning module is used for anti-misoperation decision and early warning, generates an operation suggestion based on a deep reinforcement learning (DRL) model, performs real-time early warning on a high-risk operation, and prompts an operator to correct steps through an augmented reality (AR) interface.

[0110] The actual application shows that the system can reduce the misoperation rate by more than 60%, improve the operation efficiency by 30%, and successfully avoid 3 major accidents caused by the intermediate state in a 500kV substation pilot. Through dynamic monitoring of the intermediate state, optimization of the interlocking logic and fusion of multi-source data, the safety and reliability of the substation scheduling operation are effectively improved, the power failure accidents and equipment damage caused by misoperation are reduced, and the system has significant economic and social benefits.

[0111] The embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same or similar parts of each embodiment can be referred to each other. For the device disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and the related parts can be referred to the method part.

[0112] Those skilled in the art will further realize that the mechanisms of the various examples described herein are capable of being implemented using electronic hardware, computer software, or any combination of the two, to perform the various steps and actions described herein. To clearly illustrate this interchangeability of hardware and software, various components will be described generally in terms of their functionality, without describing specific mechanisms for implementing that functionality. Those skilled in the art will appreciate that the various mechanisms described herein are capable of being implemented in any number of ways, and that the examples described herein are not intended to be exhaustive, but rather are intended to be illustrative. Any steps, actions, or functions described herein can be implemented directly in hardware, in a software module executed by a processor, or in a combination of the two. As such, the disclosure is not limited by the specifics of the functions described herein.

[0113] Finally, it should be noted that, in this document, relational terms such as first and second, and the like can be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus.

Claims

1. A substation multi-source intelligent anti-misoperation scheduling method, characterized in that, The method comprises the following steps: S100, multi-source heterogeneous data acquisition and preprocessing, collecting real-time state data of the substation, historical operation records, environmental parameters, video monitoring streams and power grid topology information, and constructing a dynamic knowledge graph; S200, dynamic identification of intermediate states, constructing a device state time sequence model based on a long short-term memory network and a graph neural network, and identifying intermediate states in the switching operation in real time, the intermediate states including semi-closed and ungrounded device operating states in the switching operation; S300, intelligent interlocking logic judgment and updating, using a hybrid rule engine to perform real-time interlocking logic judgment in combination with pre-defined interlocking logic rules, wherein the interlocking logic judgment rules include real-time energy topology information and energy consumption dynamic data to adjust the locking condition, the locking condition is optimized through real-time topology analysis to reduce redundant locking; S400, mistake-proofing decision and early warning, generating operation suggestions based on a deep reinforcement learning model, real-time early warning for high-risk operations, and prompting the operator to correct the steps through an augmented reality interface; The real-time energy topology information includes: when a fan and a fire switch are switched, if the fan output power is insufficient and the fire starting time does not meet the energy saving requirement, the fire starting operation is locked; before the fan is connected to the grid, it is necessary to ensure that the fan speed is within 90%-110% of the rated speed and the bus voltage is stable within ±5% of the rated voltage, otherwise the grid connection operation is locked; the energy consumption dynamic data includes: during the peak electricity price period, if the energy storage system has sufficient power and the photovoltaic power generation power is greater than or equal to 50% of the real-time load, the grid connection operation is locked. 2.The substation multi-source intelligent anti-misoperation scheduling method according to claim 1, characterized in that, The step S100, multi-source heterogeneous data acquisition and preprocessing, comprises: S101: Internet of Things sensor data acquisition, real-time monitoring of the closing position change of the circuit breaker contact through a high-precision position sensor installed on the circuit breaker, acquisition of pressure data of the circuit breaker operating mechanism by a pressure sensor, and monitoring of vibration data of the circuit breaker during operation by a vibration sensor; S102: SCADA system data acquisition, acquisition of current, voltage, active power and reactive power data of the circuit breaker, as well as energy storage state signals and opening and closing indication signals of the circuit breaker; S103: video monitoring device data acquisition, real-time acquisition of video monitoring data of the circuit breaker and disconnector by a high-definition camera and an infrared thermal imager, display of the closing action speed of the circuit breaker in the video, and display of temperature data at the circuit breaker contact by the infrared thermal imager; S104: data preprocessing, the data cleaning module removes error data points outside the range caused by device vibration from the position sensor data, the pressure sensor data is processed by a sliding window filter, the filter window size is 5, and the normalization processing module normalizes the circuit breaker position, current, closing speed, vibration amplitude, energy and energy consumption data to the [0, 1] interval. 3.The substation multi-source intelligent anti-misoperation scheduling method of claim 1, wherein, The step S200, dynamic identification of intermediate states, comprises: S201: LSTM model analysis, inputting the pre-processed circuit breaker position data sequence into an intermediate state identification model based on LSTM, analyzing the current circuit breaker position data sequence, and outputting the closing speed and position change curve data; S202: GNN topology modeling, combined with the topology relationship diagram of the substation equipment, the GNN model analyzes the electrical connection relationship between the circuit breaker and the disconnector, and the association relationship with the bus, transmission line and other equipment, and determines the closing state of the circuit breaker through the impedance abnormality in the electrical connection; S203: Intermediate state judgment: the closing position does not reach 95%, it is judged that the circuit breaker is in an intermediate state, and the misoperation risk assessment is carried out.

4. The multi-source intelligent anti-misoperation scheduling method for a substation of claim 1, characterized in that, The deep reinforcement learning model inputs the current state of the substation, including the closing position of the equipment, real-time interlocking topology information, load data, and energy supply state.

5. The multi-source intelligent anti-misoperation scheduling method for a substation of claim 4, characterized in that, The deep reinforcement learning model outputs the Q value of each possible operation: Q(s,a)←Q(s,a)+ α×[R(s,a)+ γ×maxQ(s',a')-Q(s,a)] Where Q(s,a) is the current Q value of state s performing action a; α is the learning rate, 0 < α < 1; R(s,a) is the immediate reward obtained after performing action a in state s; γ is the discount factor, used to balance future rewards and immediate rewards, 0 < γ < 1; maxQ(s',a') is the Q value corresponding to the action with the maximum Q value among all executable actions after performing action a into new state s'.

6. The multi-source intelligent anti-misoperation scheduling method for a substation of claim 1, characterized in that, The real-time warning form includes highlighting the closing position of the circuit breaker in red, and prompting the operator with animation for the incomplete closing steps, and sending messages through SMS and mobile APP for remote warning.

7. A substation multi-source intelligent anti-misoperation scheduling system, characterized in that, It includes: Multi-source data acquisition unit: used for multi-source heterogeneous data acquisition and preprocessing, collecting real-time state data, historical operation records, environmental parameters, video monitoring streams and power grid topology information of the substation, and constructing a dynamic knowledge graph; Intermediate state judgment module: based on long short-term memory network and graph neural network, a device state time series model is constructed to identify the intermediate state in the switching operation in real time, the intermediate state includes the half-closed and ungrounded device working state in the switching operation; Interlocking logic real-time updating module: used for intelligent interlocking logic judgment and updating, a hybrid rule engine is adopted to combine pre-defined interlocking logic rules for real-time interlocking logic judgment, wherein the interlocking logic judgment rules include real-time energy topology information and energy consumption dynamic data to adjust the locking condition, the locking condition is optimized through real-time topology analysis to reduce redundant locking; Strategy generation and warning module, used for misoperation prevention and warning, operation suggestions are generated based on the deep reinforcement learning model, high-risk operations are warned in real time, and the operator is prompted to correct the steps through an augmented reality interface; The real-time energy topology information includes: when the fan and the fire switch, if the fan output power is insufficient and the fire starting time does not meet the energy saving requirement, the fire starting operation is locked; before the fan is connected to the grid, it is necessary to ensure that the fan speed is within 90%-110% of the rated speed, and the bus voltage is stable within ±5% of the rated voltage, otherwise the grid connection operation is locked; the energy consumption dynamic data includes: during the peak electricity price period, if the energy storage system has sufficient power and the photovoltaic power generation power is greater than or equal to 50% of the real-time load, the grid connection operation is locked.

8. An electronic device, comprising: It includes: One or more processors; Memory; and one or more programs stored in the memory, including instructions for performing the substation multi-source intelligent anti-misoperation scheduling method of any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, including one or more programs for execution by one or more processors of an electronic device, the one or more programs including instructions for performing the substation multi-source intelligent anti-misoperation scheduling method of any one of claims 1-6.

Images