Anti-misoperation intelligent control method and system based on integrated control

Abstract

The application discloses an anti-misoperation intelligent control method and system based on integrated control, relates to the technical field of power automation, and comprises the following steps: acquiring first data of a three-party integrated control system, constructing a dynamic instruction pool and an operation time sequence diagram; based on the dynamic instruction pool and the operation time sequence diagram, checking operation consistency through a space-time coupling protocol, combining compliance verification rules and power grid operation data, and constructing a power flow constraint network; according to the power flow constraint network, using a mixed integer programming method and an N-1 risk matrix, generating a risk early warning; according to the risk early warning, generating an anti-misoperation decision, and feeding back to the three-party integrated control system. The application is used to solve the problems that operation instruction management and control of a traditional anti-misoperation system lacks dynamicity and consistency checking is weak.

Description

Technical Field

[0001] The present invention relates to the technical field of power system automation, and more particularly to an intelligent control method and system for preventing misoperation based on centralized control integration. Background Art

[0002] In the centralized control and operation scenarios of power systems, intelligent control with centralized error prevention is a core approach to improving grid dispatching efficiency and operational safety. However, current error prevention control systems have significant shortcomings. As grid scale expands and equipment operation complexity increases, traditional error prevention control relies on manual verification and static rules, making it difficult to adapt to dynamic and changing operational requirements. While some centralized control systems attempt to integrate operational instructions with equipment status data, there are technical gaps in the temporal coordination and spatial impact correlation analysis of multi-source operational instructions. For example, operational instructions from cross-regional centralized control centers often lack unified temporal and spatial verification standards, resulting in command timing conflicts (such as circuit breaker opening and closing sequence violations) or uncoordinated spatial impacts (such as busbar switching without considering the risk of power flow transfer on adjacent lines), leading to a high risk of error.

[0003] The existing public technical solutions have at least the following technical problems:

[0004] Operation instruction control lacks dynamism. The static instruction pool cannot absorb dynamic factors such as equipment status changes and grid topology adjustments in real time. The instruction screening and verification rules are rigid and difficult to cope with complex operation scenarios. The means of operation consistency verification are weak and lack a time-space coupling verification mechanism. It is impossible to accurately identify nanosecond-level timing deviations and equipment dependency conflicts, and the ability to prevent and control operation risks is insufficient.

[0005] Due to the superposition of the above problems, there are still hidden dangers of false operation and refusal to operate in the power grid operation under the centralized control mode, which threatens the safe and stable operation of the power grid. Summary of the Invention

[0006] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present invention provide an intelligent control method and system for preventing misoperation based on integrated centralized control. By acquiring the centralized control data of three parties and constructing a dynamic instruction pool and operation timing diagram, relying on the time-space coupling protocol to verify the consistency of operations, combining the verification rules with the power grid data to build a flow constraint network, and using mixed integer programming and N-1 risk matrix to generate early warnings, decisions and feedback, the problems of lack of dynamics and weak consistency verification in the operation instruction control of traditional anti-misoperation systems are solved.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] The application discloses an anti-misoperation intelligent control method based on integrated control, which comprises the following steps: acquiring first data of a three-party integrated control system, constructing a dynamic instruction pool and an operation time sequence diagram; checking operation consistency based on the dynamic instruction pool and the operation time sequence diagram through a space-time coupling protocol, combining compliance verification rules and power grid operation data to construct a power flow constraint network; generating a risk warning according to the power flow constraint network by using a mixed integer programming method and an N-1 risk matrix; and generating an anti-misoperation decision according to the risk warning and feeding back to the three-party integrated control system.

[0009] In a preferred embodiment, the first data specifically comprises device operation data and instruction metadata; the device operation data is mainly real-time operation information, including device identification, device type and device operation object ID; and the instruction metadata is mainly instruction attribute and time sequence information, including instruction ID, instruction type and time stamp.

[0010] In a preferred embodiment, the construction of the dynamic instruction pool specifically comprises: mapping the first data into instruction data units by using a key-value pair storage structure; performing multi-level sorting on the instruction data units based on the time stamp to generate an instruction sequence; listening to the instruction state change of the instruction sequence and combining atomic operations to realize state machine management of the instruction life cycle; and extracting the device operation object ID and the time stamp in the instruction, and performing space-time double association with the power grid topology information at the corresponding moment to mark the spatial influence range of the operation and generate the dynamic instruction pool.

[0011] In a preferred embodiment, the space-time double association with the power grid topology information at the corresponding moment to mark the spatial influence range of the operation specifically comprises: constructing an adjacency matrix according to the power grid topology information and allocating a device identification to each device; establishing a mapping relationship between the device identification and the node index of the adjacency matrix; matching the device operation object ID in the instruction data unit with the device identification of the adjacency matrix and recording the matching result; mapping the device identification in the matching result to the node index of the adjacency matrix to obtain a starting node; and starting from the starting node, identifying adjacent devices influenced by the operation instruction based on the device connection relationship and the electrical topology structure by using a breadth-first search algorithm to obtain a spatial influence range set.

[0012] In a preferred embodiment, the checking of the operation consistency based on the dynamic instruction pool and the operation time sequence diagram through the space-time coupling protocol specifically comprises: screening the dynamic instruction pool through a preset rule to obtain key instructions; extracting the time stamp of the key instructions and synchronously analyzing the device dependency relationship in the operation time sequence diagram; screening out key operation nodes influencing the topology structure based on the time stamp and the device dependency relationship; calculating the deviation in the time dimension and the spatial influence range in the space dimension of the key operation nodes to generate a dynamic coupling factor value; comparing the dynamic coupling factor value with a preset threshold value, and outputting an operation consistency checking result according to the comparison result.

[0013] In a preferred embodiment, the computing of the deviation of the key operation node in the time dimension and the spatial influence range in the spatial dimension generates a dynamic coupling factor value, specifically: based on the timestamp of the key operation node, the absolute value of the deviation of the actual time interval of the adjacent key operation node from the preset standard time interval is calculated; the spatial influence range of the key operation node is quantified by a dynamic distance method based on manifold learning and implicit spatial embedding, and is weighted and merged with the absolute value of the deviation to generate a dynamic coupling factor.

[0014] In a preferred embodiment, the quantification of the spatial influence range of the key operation node by the dynamic distance method based on manifold learning and implicit spatial embedding is specifically: high-dimensional core features of the key operation node are extracted and standardized, including device type, voltage level, real-time electrical parameters and topological connection relationship; a feature correlation matrix is constructed based on the electrical topology structure, and the high-dimensional core features are nonlinearly reduced to a low-dimensional implicit space; in the implicit space, a dynamic distance function is defined based on the Euclidean distance of the nodes; taking the key operation node as the center, the distance to other nodes is calculated by the dynamic distance function, and the initial spatial influence range is determined by combining the preset threshold to divide the neighborhood; the initial spatial influence range is mapped to the actual power grid, and the topological connection relationship is verified to output the quantification result.

[0015] In a preferred embodiment, the construction of the power flow constraint network by combining the compliance verification rules and the power grid operation data is specifically: key parameters are obtained from the power grid operation data; based on the key parameters, a fast decoupling method is used to construct a power flow calculation model; the compliance verification rules are converted into inequality constraint conditions in the power flow calculation, and are embedded into the power grid power flow calculation model to obtain a mixed constraint system; the mixed constraint system is iterated to determine the power flow distribution range that satisfies all the constraint conditions, and a power flow constraint network is generated.

[0016] In a preferred embodiment, the generation of risk warning by using the mixed integer programming method and the N-1 risk matrix is specifically: based on the power flow constraint network, the device operation is set as an integer variable and the power flow parameter is set as a continuous variable to construct a mixed integer programming model; according to the mixed integer programming model, an N-1 fault scenario is generated, the topology of the power flow constraint network is modified, and the post-fault constraint conditions are constructed and introduced into the model to calculate and solve to obtain an optimal operation sequence; based on the optimal operation sequence, the importance and occurrence probability of the fault components are quantified by using the N-1 risk matrix, and a risk warning containing operation feasibility, risk probability and impact degree is comprehensively generated.

[0017] On the second aspect, the present application provides an intelligent control system for preventing misoperation based on integrated centralized control, including: a data modeling and instruction pool construction module, which obtains the first data of the three-party centralized control system and constructs a dynamic instruction pool and operation timing diagram; a time-space constraint verification and network construction module, which verifies the operation consistency through the time-space coupling protocol based on the dynamic instruction pool and operation timing diagram, and combines compliance verification rules and power grid operation data to construct a flow constraint network; a risk assessment and warning generation module, which generates risk warnings based on the flow constraint network using a mixed integer programming method and an N-1 risk matrix; an anti-misoperation decision and closed-loop control module, which generates anti-misoperation decisions based on risk warnings and feeds them back to the three-party centralized control system.

[0018] From the above technical solutions, it can be seen that the present invention constructs a dynamic instruction pool and operation timing diagram through a cache cluster, integrates data in real time, and allows instruction screening and verification to adapt to changes in the power grid, solving the problem that traditional instruction control is static and difficult to cope with complex scenarios; through the time-space coupling protocol, it synchronously analyzes timestamps and device dependencies, accurately identifies the timing deviation and spatial impact of operations, and makes up for the weakness of the operation consistency verification method; by constructing a power flow constraint network, combining mixed integer programming with the N-1 risk matrix, it combines physical laws and operation rules, accurately assesses risks and generates error-proof decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 The figure is a flow chart of the intelligent control method for preventing misoperation based on centralized control integration of the present invention.

[0020] Figure 2 This is a structural diagram of the intelligent control system for preventing misoperation based on centralized control integration of the present invention. DETAILED DESCRIPTION

[0021] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0022] Example 1, Figure 1 The present invention provides an intelligent control method for preventing misoperation based on centralized control integration, which includes the following steps:

[0023] S1: Obtain the first data from the third-party centralized control system and input it into the cache cluster.

[0024] In this embodiment, the first data of the three-party centralized control system is obtained and input into the cache cluster, specifically:

[0025] By deploying multi-protocol data adapters to connect to third-party systems such as the substation integrated automation system, dispatching master station, and distributed power supply monitoring, the Modbus protocol is used to parse equipment operation data and instruction metadata. Abnormal values ​​are filtered out through data cleaning components and standardized packaging is performed in a unified JSON format.

[0026] RedisCluster is used to build a distributed cache cluster, with the device ID and timestamp combination as the key. Standardized data is written to the cache node in real time through an asynchronous communication mechanism. The ordered set (ZSET) structure is used to store instruction data units in order by timestamp, and data quality identifiers and checksums are added to ensure data integrity and timing consistency, providing a high-speed and reliable data base for the subsequent construction of a dynamic instruction pool.

[0027] S2, based on the cache cluster, builds a dynamic instruction pool and operation timing diagram.

[0028] In this embodiment, a dynamic instruction pool is constructed, specifically as follows:

[0029] A hash structure is used to implement key-value mapping. The key is composed of the device ID and timestamp, and the value is encapsulated as a JSON object containing device operation data and instruction metadata. The atomicity of concurrent writes is ensured through the WATCH-MULTI-EXEC transaction mechanism.

[0030] Redis ZSET is used to sort instructions by timestamp. The score field of ZSET stores the timestamp, and the member field is the instruction key. Timing verification is achieved by calculating the deviation between the time interval of adjacent instructions and the preset standard value.

[0031] Instruction lifecycle management uses a state machine model, defining a transition matrix with four states: pending, executing, completed, and failed. Lua scripts are used to encapsulate state update logic to ensure atomicity of event processing.

[0032] Construct an adjacency matrix based on the connection relationship of power grid equipment ,in is the total number of devices, the matrix elements Representation device and There is a direct electrical connection, otherwise it is 0; at the same time, each device is assigned a unique integer identifier (such as starting from 0), forming a mapping table between device identifiers and matrix indexes. Then, the device operation object ID in the instruction data unit (such as "220kV_Breaker_01") is matched to the device identifier of the adjacency matrix through the mapping table. If the match is successful, the corresponding matrix node index is obtained as the starting node ;

[0033] Use the breadth-first search (BFS) algorithm to traverse the adjacency matrix: Initialize the queue and visited collections , each time a node is taken out of the queue Join the queue and collection until the queue is empty, and finally This is the set of directly and indirectly affected devices. In this process, the adjacency matrix accurately describes the topology structure through mathematical form, and the BFS algorithm ensures that the device connection relationship is traversed hierarchically.

[0034] It should be noted that the device identification refers to: an identifier assigned to each device in the power grid, used to identify the device and associated with the adjacency matrix; the device operation object ID refers to the identifier contained in the instruction data unit, pointing to the operated device.

[0035] In this implementation use case, an operation sequence diagram is constructed, specifically:

[0036] An event collection module is deployed on each node of the cache cluster to capture data read and write, cache invalidation and other operational events in real time, and timestamp each event. To address clock inconsistencies in distributed systems, the VectorClock algorithm is used to perform causal sorting on events.

[0037] Events are grouped by node based on the cache cluster topology. Node communication links are mapped using an adjacency matrix, and event timing associations are supplemented based on inter-node data transmission relationships. When constructing a timing graph, operation events are abstracted into nodes (including attributes such as event type and timestamp). Directed edges are constructed based on dependencies determined by causal algorithms. For concurrent operations, conflict detection and consistency protocols are used to define the final execution path.

[0038] S3, based on the dynamic instruction pool and operation sequence diagram, verifies the consistency of operations through the time-space coupling protocol, and builds a power flow constraint network by combining compliance verification rules and power grid operation data.

[0039] In this embodiment, based on the dynamic instruction pool and the operation sequence diagram, the operation consistency is checked through the time-space coupling protocol, specifically:

[0040] Construct an instruction filter through preset rules to filter the instructions in the dynamic instruction pool.

[0041] It should be noted that the preset rules build a multi-layer screening mechanism based on power system safety regulations and real-time operating status:

[0042] Prioritize operations based on their type. Operations that directly change the topology, such as circuit breaker opening and closing and busbar switching, are forcibly included as key instructions (weight 1.0). High-risk operations, such as main transformer switching, are weighted ≥ 0.8. Low-risk operations, such as data queries, are weighted ≤ 0.2 and filtered by default.

[0043] Then, based on the equipment electrical centrality design rules, the equipment betweenness centrality is calculated using the adjacency matrix. The operational weight of key equipment such as hub busbars and tie lines = (number of associated lines / number of lines in the entire network × 0.8) + voltage level coefficient (220kV corresponds to 1.2). For example, when a 220kV busbar connects 8 lines (out of 100 lines in the entire network), the weight is 1.264.

[0044] It is also necessary to link with the real-time status of the power grid. When the load rate exceeds 80%, the weight of flow regulation operations will be automatically increased by 0.3. Within 30 minutes after a system failure, only fault isolation-related operations will be retained.

[0045] After extracting the timestamps of key instructions, we parse the adjacency matrix of the operation timing graph Get device dependencies, where Representation device The operation depends on the device Based on timestamp and dependency, and according to the filter formula, filter out the items that meet the formula and key operation nodes.

[0046] The screening formula is:

[0047]

[0048] In the formula Indicates the Timestamps of key operation nodes, The deviation is in nanoseconds. is the standard time interval.

[0049] When calculating the dynamic coupling factor, the deviation in the time dimension Processing is done through normalization formula;

[0050] The normalization formula is:

[0051]

[0052] Where, is the actual time interval.

[0053] The influence range of key nodes in the spatial dimension is quantified by a dynamic distance method based on manifold learning and implicit spatial embedding. The specific steps are as follows:

[0054] Obtain the original data of key operation nodes, extract equipment types (such as transformers, circuit breakers, busbars), voltage levels, real-time electrical parameters and topological connection relationships. Perform Z-score normalization on the extracted features to obtain the standardized feature vector ;

[0055] Based on the electrical topology, the physical connection relationship between devices is converted into an adjacency matrix , and combined with the standardized feature vector , calculate the cosine similarity between devices and generate the feature correlation matrix ( Representative equipment and equipment the strength of the association between

[0056] The Laplace eigenmap algorithm is used to map the feature correlation matrix Perform dimensionality reduction to obtain a low-dimensional matrix ;

[0057] In the implicit space, the dynamic distance between nodes i and j is defined as:

[0058] In the formula is the operation time interval, is the time window; For equipment the importance of For equipment Voltage level; is the weight coefficient (determined by training with historical fault data);

[0059] Based on historical operation data, the impact range of all key operations is counted, and the 95% percentile of the dynamic distance is calculated as the preset threshold , for the key operation nodes in the implicit space ;

[0060] like , then the device Belongs to the initial impact range ,right Apply the DBSCAN algorithm to remove isolated points and form a connected influence area;

[0061] described Key operation nodes With equipment The "dynamic distance" between

[0062] Traversal The equipment in the grid is removed, and nodes without electrical connection in the actual grid are eliminated; the power flow distribution after the operation is calculated by the Newton-Raphson method, and the indirect impact equipment caused by power transfer is supplemented;

[0063] The quantitative result is obtained through weighted calculation, that is, the influence .

[0064] The formula for the weighted calculation is:

[0065]

[0066] In the formula, is the attenuation coefficient of dynamic distance.

[0067] Calculate the spatial impact factor (N is the total number of system devices). Dynamic coupling factor Using the weighted synthesis formula.

[0068] The weighted synthesis formula is:

[0069]

[0070] In the formula is the time weight coefficient (which can be dynamically adjusted according to the real-time operation state of the power grid).

[0071] The dynamic coupling factor is compared with the preset threshold If the operation timing and spatial impact are both compliant, output a green identifier; if there is a slight risk, output a yellow warning and mark the specific deviation item; if the operation is determined to be high risk, output a red warning and trigger the locking mechanism.

[0072] In this embodiment, combined with the compliance verification rules and power grid operation data, a power flow constraint network is constructed, specifically:

[0073] Get the power grid operation data and extract the injected power ( node number), branch impedance , voltage amplitude and phase angle key parameters.

[0074] When constructing the power grid power flow calculation model using the fast decoupling method, the node types (such as balanced nodes, PQ nodes, and PV nodes) also need to be divided and the topology data is arranged to form a classified parameter set containing various node parameters.

[0075] Based on the classified parameter set, the power flow equation in polar form is constructed

[0076] The active power equation is:

[0077]

[0078] The reactive power equation is:

[0079]

[0080] in and is a node The injected active and reactive power, and is a node and The voltage amplitude, and is the node admittance matrix element, is a node and The voltage phase angle difference is obtained by generating a node admittance matrix The nonlinear system of equations.

[0081] Using the core assumption of the fast decoupling method (in a high-voltage power grid, when the resistance is much smaller than the reactance and the voltage phase angle difference is not large, the active power is mainly related to the voltage phase angle difference, and the reactive power is mainly related to the voltage amplitude), the nonlinear equations are decoupled and split into active and reactive subsystems.

[0082] The corresponding equation of the active subsystem is:

[0083]

[0084] Where, is the active power correction value, is the susceptance matrix corresponding to the active power, is the voltage phase angle correction.

[0085] The corresponding equation of the reactive subsystem is:

[0086]

[0087] Where, is the reactive power correction value, is the susceptance matrix corresponding to reactive power, is the voltage amplitude correction value.

[0088] Based on the above active subsystem and reactive subsystem modeling, the Jacobian matrix for iterative calculation is generated (the Jacobian matrix in the fast decoupling method is simplified, and the active iteration corresponds to Matrix, reactive iteration corresponding to matrix).

[0089] The iterative algorithm is designed based on the characteristics of the Jacobian matrix. The voltage amplitude and phase angle are initialized. The voltage phase angle correction is iteratively calculated by the active subsystem and the phase angle is updated. Then the voltage amplitude correction is iteratively calculated by the reactive subsystem and the amplitude is updated. This process is repeated until the deviation of the results of two adjacent iterations satisfies the power deviation less than , and obtain the power flow distribution solution.

[0090] The solved power flow distribution is constrained and checked to see whether it meets the physical constraints of the equipment (such as line capacity, node voltage upper and lower limits, etc.) and operating rule constraints (such as the N-1 criterion, etc.). Rationality verification is also carried out (compared with historical power flow data, theoretical stability range, etc.). If all pass, the power flow calculation model is output.

[0091] Compliance verification rules are divided into two categories:

[0092] Physical constraints on the equipment include the line transmission power not exceeding the thermal stability limit and node voltage within the rated range ;

[0093] Where, is the apparent power actually transmitted by the line, is the thermal stability limit of the line (maximum allowable transmission power), For the The actual voltage of each grid node, For the The minimum allowable voltage of a grid node, For the The maximum allowable voltage of a grid node

[0094] The operation rule constraints include the power flow transfer after the fault does not exceed the limit under the N-1 criterion , k is the safety factor;

[0095] Where, It is the transfer power borne by other lines after a line equipment fails. is the maximum allowable transmission power of the line;

[0096] These rules are converted into inequality constraints by introducing slack variables and penalty functions, and then embedded into the power flow calculation model to form a hybrid constraint system.

[0097] The hybrid constraint system is iteratively optimized and a network is generated. The hybrid constraint system is iteratively solved using algorithms such as the interior point method and sequential quadratic programming. The flow state variables are updated in each iteration, and the constraint satisfaction is checked at the same time. The variables are continuously adjusted to minimize the objective function and ensure that the constraints are satisfied until convergence. The flow distribution range that satisfies all constraints is determined. This range is presented in the form of a multidimensional feasible domain (such as voltage amplitude range, phase angle range, power transmission limit, etc.), which is the flow constraint network.

[0098] S4, based on the power flow constraint network, uses the mixed integer programming method and the N-1 risk matrix to generate risk warnings.

[0099] In this embodiment, a risk warning is generated based on the power flow constraint network using a mixed integer programming method and an N-1 risk matrix, specifically:

[0100] Based on the power flow constraint network, the various devices and power flow parameters in the network are clearly defined. Device operations are set as variables. For example, device switching, start and stop operations can be set as integer variables to represent the status of the device operation. At the same time, power flow parameters such as line power and node voltage are set as continuous variables. Using mixed integer programming methods, combined with the basic characteristics of the power flow constraint network, an optimization model is constructed.

[0101] Using the power flow constrained network information contained in the optimization model, N-1 fault scenarios are generated, simulating the failure of a single component in the network. For each generated N-1 fault scenario, the topology of the power flow constrained network is adjusted. For example, if a line fails, the line is removed from the network topology.

[0102] Add corresponding constraints to the original optimization model to adapt the model solution to fault scenarios, such as operating restrictions on equipment after a fault and adjusting the range of power flow parameters.

[0103] By solving the model after supplementing the constraints, an operation plan that is suitable for the model is obtained.

[0104] Based on the operational plan, the importance and probability of the faulty components are quantified using the N-1 risk matrix. The importance and probability of the faulty components quantified by the N-1 risk matrix correspond one-to-one with the N-1 fault scenarios.

[0105] At the same time, a feasibility analysis is conducted on the action plan to determine whether it can be successfully executed in the actual scenario. Combined with the feasibility analysis results of the action plan, an early warning is generated, including risk probability, impact level, and operational feasibility. The generation of this early warning is directly based on the action plan.

[0106] S5 generates error-proof decisions based on risk warnings and feeds back to the three-party centralized control system.

[0107] Risk warnings are graded and categorized, and response strategies are automatically matched based on pre-set rules. For example, for high-risk warnings (such as main transformer overload), the system immediately triggers emergency plans, prioritizing load shedding or power transfer. For medium-risk warnings (such as single line overload), alternative action plans are generated and assessed for feasibility. Low-risk warnings simply prompt operations and maintenance personnel to pay attention to changes in equipment status.

[0108] Optimize decision-making, considering system safety margin, operation cost, and recovery time, and select the optimal strategy from multiple alternatives. For example, when the bus voltage exceeds the limit, the system automatically compares the effects of "adjusting the capacitor", "adjusting the transformer tap", "transferring the load", etc. and selects the scheme with the smallest disturbance to the power grid and the lowest implementation cost.

[0109] Perform operation sequence deduction, generate anti-misoperation steps according to device dependency and operation timing requirements. For example, before performing line outage operation, the system will automatically check the operation sequence of "first disconnect the circuit breaker, then open the disconnecting switch" and verify whether the operation time interval meets the safety regulations. At the same time, risk secondary assessment is carried out for each operation step to avoid new risks in the operation process.

[0110] Form a hierarchical feedback mechanism: for emergency risks, the system automatically performs predefined anti-misoperation (such as automatic removal of faulty equipment); for risks requiring human intervention, generate decision-making suggestions including operation steps, expected effects, and risk prompts and push them to the operator terminal; for low-risk events, only mark the abnormal points on the monitoring interface and provide trend prediction. The feedback information also includes decision-making basis and backup schemes to ensure that the operation and maintenance personnel have a comprehensive understanding of the risk situation.

[0111] Laplacian Eigenmap is a commonly used nonlinear dimensionality reduction algorithm, belonging to the manifold learning category. The core is to preserve the local geometric structure of high-dimensional data by constructing its adjacency graph and Laplacian matrix, and map it to a low-dimensional space, which is suitable for processing data with complex nonlinear correlations.

[0112] Redis ZSET is a data structure that automatically sorts elements by associated scores and ensures element uniqueness, supporting efficient range queries, ranking retrieval, and dynamic updates, suitable for scenarios such as rankings, timelines, and priority queues.

[0113] Lua script is a lightweight, simple, and efficient embedded scripting language with simple syntax and extensibility, commonly embedded in applications (such as Redis, game engines, etc.) for flexible logic customization. In Redis, Lua scripts can combine multiple commands into an atomic operation, executed through commands such as EVAL, ensuring operation atomicity and reducing network interaction, commonly used to implement complex business logic (such as distributed locks, atomic counters, etc.).

[0114] Modbus protocol defines the message structure that controllers can recognize and use, regardless of the network they communicate over, so it does not depend on specific network hardware and can run on multiple media such as serial ports and Ethernet.

[0115] VectorClock (Vector Clock) is an algorithm used to determine event causality and detect concurrent conflicts in distributed systems, by maintaining a vector (array) for each node, recording the logical clock values of all nodes in the system.

[0116] The WATCH-MULTI-EXEC transaction mechanism is an atomic operation implementation based on optimistic locking, which monitors the changes of the key through WATCH, opens the transaction and caches the commands through MULTI, and atomically executes all commands in the queue through EXEC (if the monitored key is not modified by other clients), ensuring data consistency and operation atomicity.

[0117] Embodiment 2, Figure 2 The application provides an anti-misoperation intelligent control system based on integrated control, which comprises the following steps:

[0118] The data access and caching module is used for obtaining the first data of the three-party integrated control system, and constructing a dynamic instruction pool and an operation time sequence diagram.

[0119] The space-time coupling verification module is used for verifying the operation consistency based on the dynamic instruction pool and the operation time sequence diagram, and constructing a power flow constraint network in combination with compliance verification rules and power grid operation data.

[0120] The risk early warning generation module is used for generating a risk early warning by using a mixed integer programming method and an N-1 risk matrix according to the power flow constraint network.

[0121] The anti-misoperation decision feedback module is used for generating an anti-misoperation decision according to the risk early warning and feeding back the anti-misoperation decision to the three-party integrated control system.

[0122] The above formulas are all dimensionless numerical calculations, the formulas are obtained by collecting a large amount of data to simulate a formula of the nearest real situation, and the preset parameters in the formula are set by a person skilled in the art according to the actual situation.

[0123] The above embodiments can be realized wholly or partially by software, hardware, firmware or any combination thereof. When realized by software, the above embodiments can be realized in the form of a computer program product wholly or partially.

[0124] Those skilled in the art can realize that the modules and algorithm steps of the examples described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized by hardware or software depends on the specific application and design constraints of the technical solutions. The person skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0125] In addition, each function module in each embodiment of the present application can be integrated in one processing module, or each module can be physically present alone, or two or more modules can be integrated in one module.

[0126] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0127] Finally: the above is only a preferred embodiment of the present application and is not used to limit the present application, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. An intelligent control method for preventing misoperation based on centralized control integration, characterized in that: The following steps are involved: Obtain the first data of the three-party centralized control system and build a dynamic instruction pool and operation sequence diagram; The dynamic instruction pool is constructed by mapping the first data into instruction data units using a key-value pair storage structure, performing multi-level sorting on the instruction data units based on timestamps, generating an instruction sequence, monitoring instruction state changes in the instruction sequence, implementing state machine management of the instruction life cycle in combination with atomic operations, extracting the device operation object ID and timestamp in the instruction, performing spatiotemporal dual correlation with the power grid topology information at the corresponding moment, marking the spatial impact range of the operation, and generating a dynamic instruction pool; Based on the dynamic instruction pool and the operation sequence diagram, the operation consistency is verified through the spatiotemporal coupling protocol. Specifically, the dynamic instruction pool is screened according to preset rules to obtain key instructions, the timestamps of the key instructions are extracted, and the device dependencies in the operation sequence diagram are synchronously parsed. Based on the timestamps and device dependencies, the key operation nodes that affect the topological structure are screened out, the deviation of the key operation nodes in the time dimension and the spatial influence range in the spatial dimension are calculated, and the dynamic coupling factor value is generated. The dynamic coupling factor value is compared with the preset threshold, and the operation consistency verification result is output based on the comparison result. And combine compliance verification rules and grid operation data to build a power flow constraint network; Based on the power flow constraint network, a mixed integer programming method and N-1 risk matrix are used to generate risk warnings; Based on risk warnings, error-proof decisions are generated and fed back to the three-party centralized control system.

2. The intelligent control method for preventing misoperation based on centralized control integration according to claim 1 is characterized in that: The first data includes device operation data and instruction metadata; The device operation data includes device identification, device type and device operation object ID; The instruction metadata includes an instruction ID, an instruction type, and a timestamp.

3. The intelligent control method for preventing misoperation based on centralized control integration according to claim 2 is characterized in that: The spatial impact range of the marking operation is specifically performed by performing a temporal and spatial dual association with the grid topology information at the corresponding moment: Construct an adjacency matrix based on the grid topology information and assign a device ID to each device; Establish a mapping relationship between device identification and adjacency matrix node index; Match the device operation object ID in the instruction data unit with the device identifier in the adjacency matrix and record the matching result; Map the device ID in the matching result to the node index of the adjacency matrix to obtain the starting node; The breadth-first search algorithm is adopted, starting from the starting node, based on the device connection relationship and electrical topology structure, to identify the adjacent devices affected by the operation instruction and obtain the spatial impact range set.

4. The intelligent control method for preventing misoperation based on centralized control integration according to claim 3 is characterized in that: The calculation of the deviation of the key operation node in the time dimension and the spatial influence range in the space dimension to generate the dynamic coupling factor value is specifically: Based on the timestamps of the key operation nodes, calculate the absolute value of the deviation between the actual time interval of adjacent key operation nodes and the preset standard time interval; The spatial influence range of key operation nodes is quantified by a dynamic distance method based on manifold learning and implicit spatial embedding, and is weighted and merged with the absolute value of the deviation to generate a dynamic coupling factor.

5. The intelligent control method for preventing misoperation based on centralized control integration according to claim 4 is characterized in that: The dynamic distance method based on manifold learning and implicit spatial embedding is used to quantify the spatial influence range of key operation nodes. Specifically: Extract and standardize high-dimensional core features of key operation nodes, including device type, voltage level, real-time electrical parameters, and topological connection relationships; Construct a feature correlation matrix based on the electrical topology; Based on the feature correlation matrix, the high-dimensional core features are nonlinearly reduced to a low-dimensional implicit space; In the implicit space, a dynamic distance function is defined based on the Euclidean distance of nodes; Taking the key operation node as the center, the distance to other nodes is calculated through the dynamic distance function, and the neighborhood is divided based on the preset threshold to determine the initial spatial influence range; The initial spatial influence range is mapped to the actual power grid, combined with the topological connection relationship verification, and the quantitative results are output.

6. The intelligent control method for preventing misoperation based on centralized control integration according to claim 5 is characterized in that: The flow constraint network is constructed by combining compliance verification rules and grid operation data, specifically: Obtain key parameters from power grid operation data; According to key parameters, a fast decoupling method is used to build a power grid flow calculation model; The compliance verification rules are converted into inequality constraints in power flow calculation and embedded into the power grid power flow calculation model to obtain a hybrid constraint system. The mixed constraint system is iterated to determine the power flow distribution range that satisfies all constraint conditions and generate a power flow constraint network.

7. The intelligent control method for preventing misoperation based on centralized control integration according to claim 6 is characterized in that: The mixed integer programming method and the N-1 risk matrix are used to generate risk warnings, specifically: Based on the power flow constraint network, the equipment operation is set as integer variables and the power flow parameters are set as continuous variables to construct a mixed integer programming model; Based on the mixed integer programming model, N-1 fault scenarios are generated, the power flow constraint network topology is modified, and post-fault constraints are constructed and introduced into the model to calculate and solve the optimal operation sequence. Based on the optimal operation sequence, the N-1 risk matrix is ​​used to quantify the importance and occurrence probability of faulty components, and a comprehensive risk warning is generated that includes operation feasibility, risk probability and impact degree.

8. A system using the intelligent control method for preventing misoperation based on centralized control integration according to any one of claims 1 to 7, comprising: The data modeling and instruction pool construction module obtains the first data of the three-party centralized control system and builds a dynamic instruction pool and operation sequence diagram; The spatiotemporal constraint verification and network construction module, based on a dynamic instruction pool and operation sequence diagram, verifies operational consistency through a spatiotemporal coupling protocol and builds a power flow constraint network by combining compliance verification rules and grid operation data. The risk assessment and warning generation module generates risk warnings based on the power flow constraint network using mixed integer programming and the N-1 risk matrix; The error-prevention decision-making and closed-loop control module generates error-prevention decisions based on risk warnings and feeds them back to the three-party centralized control system.

Images