An alarm and event push method for power plant safety management

Abstract

This invention discloses an alarm and event push method for power plant safety management, belonging to the field of power plant monitoring technology. Addressing technical problems such as high redundancy of power plant alarm information and poor adaptability of push targets, this method first constructs a multi-dimensional topology mapping of the power plant and performs continuous monitoring. It then extracts the rate of change of the underlying physical state to solve for the dynamic virtual mass of the initial alarm node, thereby generating an anisotropic gravitational field. Cascaded events satisfying causal conditions are then fused into a unified macroscopic alarm event vector. Simultaneously, based on the concurrency of on-duty personnel tasks, operational response delay, and the diffusion radius of hazardous media on-site, a composite cognitive repulsive field is generated. Finally, in the combined potential energy field of superimposed gravity and repulsion, the steepest descent principle of the gradient is used to optimize the target receiving node and perform directional push. This invention achieves automatic cascading fusion and routing of alarm information, effectively reducing redundant pushes and ensuring the accuracy and safety of alarm handling.

Description

Technical Field

[0001] This invention relates to the field of power plant safety monitoring and industrial information technology, specifically to an alarm and event push method for power plant safety management. Background Technology

[0002] Modern power plants, as complex large-scale industrial systems, encompass multiple processes such as thermal, electrical, and chemical processes. To ensure the safe and stable operation of equipment, power plants deploy a massive number of sensors and distributed control systems to monitor the physical status parameters of each production node in real time. When equipment deviates from normal operating conditions, the monitoring system generates corresponding alarm events and pushes them to maintenance personnel for confirmation and handling. The alarm and event collection, generation, and push mechanism is an indispensable basic component of the power plant's safety management and emergency response system.

[0003] Existing power plant alarm and event push methods mainly rely on preset static threshold judgment logic. In actual operation, the data acquisition system periodically polls the operating parameters of the underlying equipment and compares them with fixed upper and lower safety thresholds. Once the data of a certain measuring point exceeds the limit, the system immediately triggers a single alarm signal. Then, according to static configuration rules, these alarm information are sent directly to the host computer screen in the central control room or the mobile terminal of a specific duty person through broadcast or fixed routing path, so as to prompt personnel to take intervention measures.

[0004] However, power plant process systems are highly interconnected and coupled physically. Initial physical faults in a single device often trigger chain reactions along process pipelines or electrical networks. Existing alarm methods treat over-limit alarms at various measuring points as isolated data events, failing to establish a mapping relationship between them and the multi-dimensional physical topology of the power plant and the dynamic deterioration rate of equipment status. This independent processing method, lacking underlying physical logic support, results in the system being unable to perform causal logic identification and fusion interception of derived alarms based on the conduction time delay characteristics of different physical media when a sudden fault occurs. This leads to a large number of redundant cascading alarm messages flooding the communication network simultaneously in a short period of time. This not only causes serious alarm flooding but also means that the pushed data only provides shallow, single-measuring-point reminders, failing to provide on-site personnel with a macroscopic fault description with physical depth. Summary of the Invention

[0005] This invention provides an alarm and event push method for power plant safety management, comprising the following steps:

[0006] Step S10: To address the technical issues of high redundancy in alarm information and poor compatibility between the pushed targets and the actual status of on-duty personnel during power plant operation, the system constructs a multi-dimensional topology mapping system based on physical space, process logic, and electromagnetic connection relationships. Specifically, the system utilizes the plant-level monitoring information system to analyze the process pipeline and electrical wiring models to generate a multi-dimensional topology diagram, and establishes a data update channel by synchronizing the operating parameters of equipment nodes in real time through the distributed control system data interface.

[0007] Step S20: The method constructs a potential energy field model, transforming the underlying physical state parameters into gravitational parameters in a multidimensional topological space. Specifically, after the system captures the first out-of-limit event, it extracts the rate of change of the underlying physical state of the physical device and solves for the dynamic virtual mass. In the thermal fluid solution domain, the system extracts the absolute value of the time partial derivative of the fluid medium pressure parameter and the absolute value of the thermodynamic enthalpy drop rate, and multiplies them by the pressure change weighting coefficient and the enthalpy drop change weighting coefficient, respectively, and performs a linear superposition operation to obtain the thermodynamic mass component. In the electrical protection solution domain, the system calculates the difference between the steady-state voltage and the transient voltage to obtain the transient voltage drop amplitude, and performs an integral operation on the instantaneous value sequence of the zero-sequence current. The drop amplitude and the integral result are superimposed with the voltage drop weighting coefficient and the zero-sequence current integral weighting coefficient, respectively, to obtain the electrical mass component. The system constructs a state column vector containing the thermodynamic mass component and the electrical mass component, and solves for the dynamic virtual mass through matrix multiplication, realizing the transformation of physical state parameters into the gravitational properties of topological nodes.

[0008] Step S30: During the spatial radiation process of the gravitational field, the system maps the medium time delay constant according to the type of physical transmission medium in the shortest topological path. When the physical transmission medium is a fluid process pipeline, the system extracts the medium flow velocity and specific heat capacity parameters to calculate the thermal inertia index and maps it to the medium time delay constant to reduce the spatial attenuation gradient of the gravitational field. When the physical transmission medium is an electrical wire or communication cable, the medium time delay constant is directly assigned to a preset minimum value to increase the spatial attenuation gradient of the gravitational field. The system calculates the gravitational potential energy experienced by the topologically related nodes in the gravitational field. Its value is equal to the product of the dynamic virtual mass and the negative exponent of the natural constant. The exponent of the negative exponent is the ratio of the square of the connected distance to twice the square of the medium time delay constant. The system compares the gravitational potential energy with the potential energy gradient adsorption threshold. If the gravitational potential energy of the derived alarm node is greater than or equal to the potential energy gradient adsorption threshold and is within the causal tracing time window, it is determined that it meets the gravitational potential energy adsorption condition, confirms it as a cascaded event and merges it into the macroscopic alarm event vector, and cancels the independent push task of the derived alarm node.

[0009] Step S40: To quantify the operational load and generate repulsion conditions by combining environmental hazard parameters, the method constructs a composite cognitive repulsion field. The system identifies operation tickets in the on-site execution state and those awaiting monitoring confirmation and counts them as the number of concurrent operation tasks. Simultaneously, the system retrieves historical logs to calculate the time difference between the first timestamp of the push instruction and the second timestamp of the trigger confirmation control. The arithmetic mean within the historical time sliding window is used to confirm the operation response delay time. The real-time cognitive load value is obtained by weighted summation of the number of concurrent operation tasks and the operation response delay time, combined with the task concurrency load weight coefficient and the fatigue delay load weight coefficient, respectively. The weight coefficient decreases as the job skill level increases. In addition, the system calculates the dynamic diffusion radius of the hazardous medium based on the Gaussian plume meteorological diffusion model. When the physical spatial straight-line distance between the absorption node and the first alarm node is less than or equal to the dynamic diffusion radius, the environmental hazard penalty item is assigned a preset multiple of the maximum allowable boundary value. The system linearly superimposes the real-time cognitive load value and the environmental hazard penalty item to generate a composite cognitive repulsion potential energy, which is then configured around the absorption node.

[0010] Step S50: During the alarm push execution phase, the method performs an optimization calculation based on the steepest descent of the potential energy gradient. The system extracts the process system code that triggered the alarm and compares it with the professional qualification parameters for identifier consistency. If the comparison is consistent, an attractive potential energy constant greater than the maximum value of the composite cognitive repulsive potential energy is allocated to the corresponding absorbing node to construct an attractive potential energy field. The comprehensive received potential energy is obtained by subtracting the attractive potential energy constant from the composite cognitive repulsive potential energy. The system uses the spatial coordinates of the first alarm node as the addressing calculation starting point and calculates the potential energy descent slope pointing to each absorbing node. The slope is calculated by dividing the difference between the received potential energy and the initial reference potential energy by the straight-line distance between the two points in physical space. The system iterates through the nodes and finds the node with the smallest slope of potential energy decrease, confirming it as the target receiving node. After encapsulating the macroscopic alarm event vector into a network data packet and sending it, the system receives an acknowledgment message containing a processing confirmation timestamp. Based on the processing confirmation timestamp, the dynamic virtual mass of the first alarm node is forcibly assigned to zero. Simultaneously, the gravitational field potential energy distribution is revoked, and the independent push interception restriction of the derived alarm nodes is lifted, restoring the initial state of the gravitational field in the multidimensional topology database.

[0011] The present invention, by adopting the above technical solution, can bring the following beneficial effects:

[0012] 1. This invention utilizes the power plant pipeline and electrical wiring topology to construct a multi-dimensional mapping and transforms the rate of change of the underlying physical state into a dynamic virtual mass. Combined with the medium time delay constant, it generates an anisotropic gravitational field. Then, within the causal tracing time window, it intercepts independent push commands based on gravitational potential energy comparison to generate macroscopic alarm event vectors. This achieves automatic absorption and causal logic identification of derived cascade alarms, transforms scattered equipment over-limit data into system-level fault descriptions with physical depth, and effectively reduces the bandwidth occupation of redundant alarm information.

[0013] 2. This invention quantifies real-time cognitive load by extracting the number of concurrent tasks in the active state of on-duty personnel and the historical operation response delay time, and calculates the diffusion radius of the hazardous medium based on the Gaussian diffusion model to configure the limit penalty constant for the absorption nodes in the hazardous area. The real-time cognitive load and the environmental hazard penalty term are superimposed to generate a composite cognitive repulsive field, realizing the dynamic quantification of personnel operation burden and physical environmental hazard boundary in a multi-dimensional topological space. This enables the alarm push path to automatically avoid on-duty personnel with high load and in dangerous threat areas to ensure on-site processing safety.

[0014] 3. This invention assigns an attractive potential energy constant by comparing the identifier consistency of the professional qualification parameters bound to the process system code and the absorption node. This constant is then algebraically superimposed with the composite cognitive repulsive field to generate a comprehensive receiving potential energy function. The node with the smallest algebraic value of the potential energy decrease slope along the negative direction of the total potential energy gradient is identified as the target receiving node for network-oriented push flow. This enables the accurate distribution of alarm event vectors to the mobile operation terminal with the least comprehensive resistance and the best response capability, under the conditions of satisfying professional authority constraints, personnel load balancing, and environmental safety obstacle avoidance. Attached Figure Description

[0015] Figure 1 This is a logic diagram of the power plant alarm routing push system of the present invention;

[0016] Figure 2 This is a flowchart illustrating the overall workflow of power plant alarm routing and push notification in this invention.

[0017] Figure 3 This is a flowchart illustrating the construction process of the comprehensive potential energy field based on multi-physics field conditions in this invention.

[0018] Figure 4 This is a flowchart of the optimization and flow control based on the steepest descent of the potential energy gradient in this invention.

[0019] Figure 5 This is a schematic diagram illustrating the principle of multidimensional topology and alarm potential energy field distribution in power plants according to the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Example 1

[0022] See attached document Figure 1-5 This invention provides an underlying system architecture for an alarm and event push method for power plant safety management, comprising:

[0023] The hardware acquisition and execution layer and the software logic operation layer are divided into a hardware acquisition and execution layer and a software logic operation layer. The hardware acquisition and execution layer establishes a data interaction channel between physical devices and operators, while the software logic operation layer processes the potential field transformation and routing addressing operations of alarm data.

[0024] The hardware acquisition and execution layer is configured with a distributed control system data interface and a plant-level monitoring information system data interface. The distributed control system data interface synchronizes the operating status parameters of the underlying physical equipment of the power plant in real time. The plant-level monitoring information system data interface retrieves the power plant process pipeline design model and electrical wiring topology model. The distributed control system data interface uses the OPC UA protocol or Modbus TCP protocol for industrial bus data communication. For the handshake rules and message parsing mechanism of the underlying communication protocol, those skilled in the art can refer to the international standard specifications of the corresponding communication protocol for implementation. The corresponding communication mechanism is a well-known technology in this field and will not be described in detail here.

[0025] The hardware acquisition and execution layer is equipped with personnel positioning terminals and mobile operation terminals. The personnel positioning terminal includes an ultra-wideband positioning base station and a wearable positioning tag. The ultra-wideband positioning base station calculates and outputs the three-dimensional spatial coordinate parameters of the on-duty personnel to the system in real time by measuring the time of flight of the radio signal between the ultra-wideband positioning base station and the wearable positioning tag. The mobile operation terminal includes an explosion-proof industrial tablet computer and a host computer workstation in the control room. The mobile operation terminal receives alarm event vectors and presents an operation confirmation interface. The mobile operation terminal has an embedded electronic operation ticket management module, which records and feeds back to the system the number of concurrent operation tasks that the on-duty personnel are in an active state.

[0026] The software logic operation layer is equipped with a topology mapping module. The topology mapping module communicates with the hardware acquisition and execution layer via industrial Ethernet. The topology mapping module transforms the three-dimensional physical pipeline model and electrical wiring topology model input from the plant-level monitoring information system data interface into a multi-dimensional topology graph. The specific lower-level implementation features of the topology mapping module include a graph database construction unit and an adjacency matrix calculation unit. The graph database construction unit parses the process pipeline design model file, extracts the three-dimensional spatial coordinates of the equipment flange interface and electrical wiring terminal as a vertex set, and extracts the process pipelines connecting different equipment flange interfaces and the wires connecting different electrical wiring terminals as an edge set.

[0027] The adjacency matrix calculation unit constructs a network adjacency matrix that represents the physical connectivity relationships between independent device nodes. For any node in the multidimensional topology graph With nodes Network adjacency matrix matrix elements The value retrieval logic is defined as follows:

[0028]

[0029] In the formula, For connecting nodes With nodes The medium connectivity distance;

[0030] When the connection medium is a process pipeline For nodes With nodes The Euclidean distance between the three-dimensional spatial coordinates of the two points is calculated using the following formula:

[0031] ;

[0032] In the formula, ( , , ) is a node The absolute coordinates in three-dimensional space, , , ) is a node The absolute coordinates in three-dimensional space;

[0033] When the connection medium is an electrical wire For nodes With nodes The equivalent electrical impedance value of the electrical conductors between them;

[0034] The software logic operation layer is equipped with a potential energy field operation module, which includes a multiphysics situation analysis unit and a comprehensive potential field generation unit. The multiphysics situation analysis unit extracts the rate of change of the underlying physical state from the data interface input of the distributed control system, and converts the rate of change of the underlying physical state into thermodynamic mass components and electrical mass components according to the node attributes output by the topology mapping module. The comprehensive potential field generation unit constructs an outward radiating gravitational field based on the thermodynamic mass components and electrical mass components. The comprehensive potential field generation unit combines the three-dimensional spatial coordinate parameters input by the personnel positioning terminal and the number of concurrent operation tasks input by the mobile operation terminal to construct an outward diverging cognitive repulsive field. The multiphysics situation analysis unit and the comprehensive potential field generation unit are deployed in the server array, and the tensor processing unit cluster in the server array executes parallel matrix multiplication instructions to accelerate the solution process of the differential equation of the multidimensional potential energy field.

[0035] The software logic operation layer deploys a routing control and push module, which is connected to the potential energy field operation module. The routing control and push module includes a gradient optimization unit and a message sending unit. The gradient optimization unit calculates the total potential energy gradient vector of the macro alarm event vector in the comprehensive potential energy field and plans the route path with the fastest descent of the total potential energy gradient. The message sending unit encapsulates the macro alarm event vector into a network data packet according to the route path with the fastest descent of the total potential energy gradient and sends a push request to the target mobile operating terminal.

[0036] The alarm and event push system for power plant safety management provided in this embodiment includes the following steps when performing the information distribution task from the physical site to personnel terminals:

[0037] Step S10: Construct multidimensional topology mapping and continuous status monitoring. The system receives the three-dimensional physical pipeline model and electrical wiring topology model output by the plant-level monitoring information system. The system parses the three-dimensional physical pipeline model and electrical wiring topology model, extracts equipment flange interfaces and electrical terminals as equipment nodes, extracts process pipelines and wires as connection edges, and generates a multidimensional topology map. After the multidimensional topology map is generated, the system continuously synchronizes the real-time operating parameters of the physical equipment corresponding to each equipment node in the multidimensional topology map through the distributed control system data interface.

[0038] Step S20: Capture the first over-limit event and solve the dynamic virtual mass. When the real-time operating parameters of any device node in the multi-dimensional topology graph exceed the preset safety threshold, the system marks the device node that triggers the over-limit alarm as the first alarm node. The system extracts the rate of change of the underlying physical state of the physical device corresponding to the first alarm node. If the first alarm node is a thermal fluid device, the system extracts the pressure partial derivative and thermodynamic enthalpy drop rate of the fluid medium and calculates the thermodynamic mass coefficient. If the first alarm node is an electrical device;

[0039] The system extracts the transient voltage drop amplitude and zero-sequence current, and calculates the electrical quality coefficient. The system inputs the thermodynamic quality coefficient or electrical quality coefficient into a preset mapping matrix to solve for the dynamic virtual quality of the first alarm node in the multi-dimensional topology graph. For the sampling and filtering of the underlying physical state data, those skilled in the art can use the Kalman filter algorithm or the moving average filter algorithm to eliminate high-frequency noise interference. The corresponding filtering algorithms are well-known technologies in the field and will not be described in detail here.

[0040] Step S30: Generate an anisotropic gravitational field and perform cascaded alarm fusion. The system generates an outward-radiating gravitational field in the multidimensional topology map based on the dynamic virtual mass, with the first alarm node as the center. When the connecting medium is a thermo-fluid pipeline, the system assigns a medium time delay constant greater than one to form a smooth gravitational space decay characteristic.

[0041] When the connection medium is an electrical wire, the system assigns a time delay constant of less than one to the medium to form a steep gravitational space attenuation characteristic. When the surrounding associated device nodes of the first alarm node exceed the state limit and generate derivative alarms, the system calculates the gravitational potential energy of the derivative alarm node in the gravitational field. If the calculated gravitational potential energy is greater than the preset adsorption threshold and the triggering time of the derivative alarm is within the preset causal tracing time window, the system determines that the corresponding derivative alarm is a cascade event. The system merges the alarm information data fields carried by the cascade event into the data packet of the first alarm node, generates a unified macroscopic alarm event vector, and cancels the independent push command of the derivative alarm node.

[0042] Step S40: Quantify personnel cognitive load and generate a composite repulsive field. The system registers the mobile operation terminal of the on-duty personnel as an absorption node in the multi-dimensional topology map. The system extracts the number of concurrent operation tasks currently active for each on-duty personnel through the electronic operation ticket management module, extracts the operation response delay time from the historical system log, calculates the real-time cognitive load value of each on-duty personnel, and simultaneously analyzes the environmental hazard parameters of the physical area where the first alarm node is located. The system adds an environmental hazard penalty item to the absorption node whose physical location is within the diffusion radius of the hazardous medium, and generates a composite cognitive repulsive field around each absorption node.

[0043] Step S50: Calculate the comprehensive potential energy gradient and execute adaptive push flow. The system superimposes the gravitational field and the composite cognitive repulsive field to construct a global comprehensive potential energy field in the multidimensional topology map. The system plans the routing addressing path of the macroscopic alarm event vector along the negative direction of the total potential energy gradient, finds the target mobile operation terminal corresponding to the minimum value of the total received potential energy in the multidimensional topology map, and after the target mobile operation terminal is established, the system encapsulates the macroscopic alarm event vector into a network data packet and sends it to the target mobile operation terminal. The target mobile operation terminal receives the network data packet and returns an acknowledgment message containing a processing confirmation timestamp to the system backend. After receiving the acknowledgment message, the system clears the dynamic virtual mass corresponding to the first alarm node, resets the comprehensive potential energy field, and ends the current alarm flow cycle.

[0044] Furthermore, this embodiment discloses an alarm and event push method for power plant safety management, including the following steps:

[0045] Step S110: Construct a multi-dimensional topology node set. The system retrieves the power plant's 3D design model file through the plant-level monitoring information system data interface. The system parses the 3D design model file and extracts the absolute 3D spatial coordinates of the center points of each physical device within the power plant. The system reads the equipment attribute tags from the 3D design model file and extracts the process system code to which each physical device belongs. The system combines the absolute coordinates in three-dimensional space with the process system code to generate a topological node coordinate vector containing four dimensions. The system records the topology node coordinate vectors corresponding to all extracted physical devices into the graph database to form a set of vertices of a multidimensional topology graph. For the parsing and spatial coordinate extraction algorithm of the three-dimensional design model file, those skilled in the art can use geometric model parsing tools based on industrial basic standards to implement it. The aforementioned basic geometric parsing operation is a well-known technology in this field and will not be described in detail here.

[0046] Step S120: Construct a multidimensional topology edge set and calculate connectivity distances. The system extracts process pipeline layout data and electrical system wiring data from the 3D design model file. The system maps process pipelines and electrical cables connecting adjacent physical devices to network edges in the multidimensional topology graph, forming an edge set. Based on the mapped network edge attributes, the system calculates the medium connectivity distance between each topology node. When two topology nodes are connected by a fluid process pipeline, the system retrieves the 3D spatial layout data of the fluid process pipeline, performs line integral calculations along the 3D spatial layout, and sets the actual physical total length of the pipeline as the medium connectivity distance of the corresponding network edge. When two topology nodes are connected by an electrical cable, the system retrieves the single-line diagram ledger data of the electrical system and extracts the equivalent resistance parameters of the corresponding electrical cable. With equivalent reactance parameters And using the calculation formula:

[0047]

[0048] Solve for the magnitude of the total impedance The system will convert the magnitude of the total impedance. The mapping is to the medium connectivity distance of the corresponding network edge. The aforementioned design establishes an equivalent relationship between the network topology distance and the physical laws of fluid and electromagnetic transmission at the bottom layer of the power plant, providing physical boundary support for the subsequent calculation of gravitational field attenuation.

[0049] Step S130: Establish a mapping between multi-dimensional topology nodes and measurement point data channels. The system establishes a data communication connection with the measurement point sensors in the industrial field through the distributed control system data interface. The system retrieves a preset measurement point matching dictionary and reads the data acquisition channel identifier from the measurement point matching dictionary. The system binds the data acquisition channel identifier with the coordinate vectors of each topology node in the vertex set of the graph database, establishing a parameter update mapping channel from the measurement point sensor to the graph database node. Step S240: Perform continuous acquisition of state parameters and extraction of temporal differential features. The measurement point sensor acquires the real-time operating parameter sequence of the physical device at a preset sampling frequency. The system receives the real-time operating parameter sequence, performs timestamp alignment and outlier removal operations, extracts the transient absolute value of the real-time operating parameter sequence in the current sampling period, and uses a backward difference algorithm to calculate the rate of change of the corresponding real-time operating parameter over time. The calculation formula is:

[0050]

[0051] In the formula, At the current sampling time, For the previous sampling time, These are the real-time operating parameter values ​​collected at the current sampling time. The system extracts the transient absolute values ​​and calculates the rate of change over time from the real-time operating parameter values ​​collected at the previous sampling time. Write the data into the dynamic attribute fields of the corresponding topological nodes to complete the attribute field update of the multidimensional topological graph, providing a computational input source for solving the potential energy field mass problem.

[0052] Furthermore, this embodiment discloses the implementation details of converting the degradation rate of the underlying physical device into the intensity of the gravitational source in the comprehensive potential energy field. The specific steps include:

[0053] Step S210: Identify the process attribute boundary of the first alarm node. The system reads the attribute fields of the first alarm node in the multidimensional topology database and extracts the process system code to which the first alarm node belongs. The system is based on the process system code. The field value determines the process category of the physical equipment, and divides the data processing flow of the first alarm node into the thermal fluid solution domain or the electrical protection solution domain. For the solution domain that has not been assigned a data processing flow, the system forces the corresponding mass component to be initialized to zero. The aforementioned processing logic avoids non-identical physical parameters from participating in the calculation and reduces the calculation dimension of the feature matrix.

[0054] Step S220: Calculate the thermodynamic mass components within the thermal fluid solution domain, when the process system is coded. When the system indicates that the first alarm node belongs to the thermal system, it extracts the current sampling time from the dynamic attribute field of the first alarm node. Fluid medium pressure parameters With temperature parameters The system calls the preset reference formulas for the thermodynamic properties of industrial water and steam, and applies the pressure parameters. With temperature parameters Mapped to the real-time thermodynamic enthalpy value under the corresponding state For the specific process of looking up tables or performing polynomial fitting calculations for thermodynamic enthalpy, those skilled in the art can refer to the published documents of the IF97 industrial standard for implementation. The corresponding conversion calculation is a well-known technique in this field and will not be elaborated here.

[0055] After acquiring the fluid medium pressure parameters and real-time thermodynamic enthalpy sequence, the system calls the differential module to calculate the time partial derivatives of the fluid medium pressure parameters. With thermodynamic enthalpy drop rate The system uses linear superposition to obtain the thermodynamic mass components. The specific differential equation is as follows:

[0056]

[0057] In the formula, The pressure change weighting coefficient is used. As the weighting coefficient for enthalpy drop change, the system reads the power plant process design parameter table, assigns the pressure change weighting coefficient to the reciprocal of the pipeline's rated design pressure, and assigns the enthalpy drop change weighting coefficient to the reciprocal of the medium's specific heat capacity parameter value. This weighting assignment method essentially performs dimensionless normalization on physical parameters with huge differences in dimensions and values, and quantitatively converts the rate of deterioration of the thermodynamic state into the mass benchmark of the gravitational field.

[0058] Step S230: Calculate the electrical quality components within the electrical protection solution domain, when the process system is coded. When the initial alarm node is assigned to the electrical system, the system extracts the corresponding electrical circuit at the initial moment of the over-limit trigger. up to the current sampling time The system extracts the initial moment of the out-of-limit trigger from the timing parameters during the process. steady-state voltage With the current sampling time transient voltage The transient voltage drop amplitude is obtained by performing a subtraction operation. The system synchronously extracts the zero-sequence current instantaneous value sequence for the aforementioned time interval. ;

[0059] The system calculates the electrical quality component by integrating the transient voltage drop amplitude with the zero-sequence current. The calculation formula is:

[0060]

[0061] In the formula, This is the voltage sag weighting factor. As the zero-sequence current integral weighting coefficient, the system retrieves the setting value file of the relay protection device bound to the corresponding electrical circuit, assigns the voltage drop weighting coefficient to the reciprocal of the absolute value of the low voltage protection trip setting, and assigns the zero-sequence current integral weighting coefficient to the reciprocal of the absolute value of the ground fault protection current action setting. After performing dimensionless normalization processing through the reciprocal of the setting value, the aforementioned weighting assignment method makes the degree of exceeding the limit of the electrical parameter positively correlated with the electrical quality component obtained by the solution.

[0062] Step S240: Generate dynamic virtual mass matrix parameters; the system constructs a state column vector containing thermodynamic mass components and electrical mass components. The system initializes the dimension in memory space as follows: Normalized mapping row vectors The system performs matrix multiplication:

[0063]

[0064] The system will solve for the dynamic virtual mass. The initial alarm node is written to the potential field gravity source feature region in the graph database, and the potential energy field calculation module extracts the dynamic virtual mass in the next calculation cycle. The numerical load drives the spatial radiation calculation of the anisotropic gravitational field, and completes the quantitative conversion of the rate of change of the physical state of the equipment into the virtual quality of the network topology.

[0065] Furthermore, this embodiment discloses the underlying mathematical model and implementation steps for constraining the spatial radiation distribution of the potential energy field using the properties of a physical conduction medium, including:

[0066] Step S310: Traverse the multidimensional topology branches and extract connected paths. The system reads the multidimensional topology graph database, starting from the first alarm node, and traverses the surrounding associated device nodes outward along the network edges of the multidimensional topology graph. The traversed device nodes are marked as topology-related nodes. For any topology-related node, the system extracts the shortest topology path connecting the first alarm node and the topology-related node, and extracts the connected distance value corresponding to the shortest topology path. For the search and calculation of the shortest topological path, those skilled in the art can use Dijkstra's algorithm or A* pathfinding algorithm to implement it. The corresponding path search operation is a well-known technology in the field and will not be described in detail here.

[0067] Step S320: Parse the physical transmission medium category and map the medium time delay constant. The system parses the network edge attributes contained in the shortest topology path, identifies the corresponding physical transmission medium category in the physical field, and establishes a physical transmission medium category and medium time delay constant mapping in the database. The numerical mapping relationship is such that when the physical transport medium corresponding to the shortest topological path is a fluid process pipeline, the system extracts the medium velocity from the 3D design model file. With specific heat capacity parameter and use the formula Calculate thermal inertia index The system uses a linear mapping formula. Calculate the medium time delay constant corresponding to the fluid medium. In the formula As a preset scaling factor, the aforementioned parameter calculation method reduces the spatial attenuation gradient of the gravitational field in the direction of the fluid process pipeline. Due to the large thermal inertia and transmission time delay of the fluid medium, its fault evolution physically manifests as a large-scale, slowly varying diffusion in space. Therefore, a smoother gravitational field is needed to attract related derivative alarms with a greater topological distance. When the physical transmission medium corresponding to the shortest topological path is an electrical wire or communication cable, due to the extremely low time delay characteristics of electromagnetic wave signals, the fault usually manifests as a transient and highly localized protection action. The system will use the medium time delay constant... The value is directly assigned to the preset minimum constant value. The aforementioned constant assignment method increases the spatial attenuation gradient of the gravitational field in the branch direction of the electrical network, ensuring that only instantaneous over-limits occurring within the immediate vicinity of the electrical node will be adsorbed and fused.

[0068] Step S330: Calculate the spatial potential energy distribution of the anisotropic gravitational field. The system reads the dynamic virtual mass stored in the multidimensional topology database of the first alarm node. The system uses a decay function to calculate the gravitational potential energy of topologically connected nodes in a gravitational field. The specific equation for gravitational potential energy spatial radiation is as follows:

[0069]

[0070] In the formula, This represents the gravitational potential energy aggregated at topologically related nodes. The dynamic virtual quality of the first alarm node, This represents the shortest topological path connection distance between the node that first issued an alarm and the nodes associated with it in the topology. The medium time delay constant is obtained by mapping according to the physical transmission medium category. For the natural constant An exponential function with base 0;

[0071] The aforementioned calculations result in an asymmetric decay distribution of the gravitational field as it radiates outward along different network branches in the multidimensional topological space, providing a quantitative comparison benchmark for subsequent determination of whether the derived alarm nodes belong to cascade events.

[0072] Furthermore, this embodiment discloses the implementation steps for identifying and generating unified push events using the gravitational field potential energy distribution, including:

[0073] Step S340: Set the potential energy gradient adsorption threshold and the causal tracing time window. The system presets the potential energy gradient adsorption threshold. The potential energy gradient adsorption threshold is used to define the effective adsorption range of the gravitational field, and the system synchronously presets a causal tracing time window. The system extracts the initial moment when the first alarm node triggers the over-limit alarm. The time interval of the causal tracing time window is determined as follows: When a peripheral device node within the gravitational field's radiation range experiences a state exceeding its limit and generates a derivative alarm, the system extracts the trigger time of the derivative alarm. The system determines the trigger time. Whether it falls within the time interval of the causal tracing time window, if the trigger time If the time frame does not fall within the specified time range, the system determines the corresponding derivative alarm as an independent event; if the trigger time is... Once the time interval is entered, the system initiates the spatial dimension gravitational potential energy comparison logic;

[0074] Step S350: Perform spatial dimension potential energy comparison and cascading event determination. The system extracts the coordinate vectors of the derivative alarm nodes that triggered the derivative alarms in the multidimensional topology graph, and extracts the connectivity distance between the derivative alarm nodes and the initial alarm nodes in the multidimensional topology graph. The system calls the aforementioned gravitational potential energy spatial radiation equation to calculate the gravitational potential energy experienced by the derivative alarm nodes in the gravitational field generated by the initial alarm nodes. The system will use gravitational potential energy Adsorption threshold with potential energy gradient Numerical comparisons were performed, and if gravitational potential energy... Adsorption threshold less than the potential energy gradient The system determines that the derived alarm node does not meet the gravitational potential energy adsorption condition and marks the derived alarm as an independent alarm event. If the gravitational potential energy... Adsorption threshold greater than or equal to the potential energy gradient The system determines that the derived alarm node meets the gravitational potential energy adsorption condition. The system officially confirms the derived alarm that simultaneously meets the time interval determination condition and the potential energy value comparison condition as a cascade event triggered by the first alarm node.

[0075] Step S360: Intercept independent push commands and generate a macroscopic alarm event vector. After confirming that the derived alarm is a cascading event, the system cancels the push task of the derived alarm node sending independent push data packets to the external network in the routing control push module to avoid redundant alarm information occupying communication bandwidth. The system extracts the underlying physical state parameters of the initial alarm node and the state parameters of all derived alarm nodes confirmed as cascading events, and the system constructs a unified macroscopic alarm event vector. The mathematical encapsulation structure of the macroscopic alarm event vector is defined as follows:

[0076]

[0077] In the formula, The multidimensional topological coordinate vector of the first alarm node. The dynamic virtual quality of the first alarm node, The timestamp for the generation of macroscopic alarm event vectors. Let be the set of coordinate vectors of all cascading events, and ,in The total number of cascading events. For the first The multidimensional topological coordinate vector of each derived alarm node;

[0078] The aforementioned encapsulation structure merges scattered local alarm event data into a single data packet, which serves as the basic computational unit for force optimization in the subsequent integrated potential energy field.

[0079] Furthermore, this embodiment discloses the process of extracting personnel operation status data and converting it into the fundamental parameters of the cognitive repulsive field in the comprehensive potential energy field, specifically including the following steps:

[0080] Step S410: Identify and count the concurrent execution of tasks in the active state. The system establishes a communication connection with the electronic operation ticket management module through the underlying data interface. The system extracts the operation ticket task ledger bound to each on-duty personnel at the current moment. The system reads the execution status field in the operation ticket task ledger. The system filters operation tickets with the execution status field in the "in progress" state and "awaiting monitoring confirmation" state. The filtered operation tickets are classified as active state tasks. The system counts the total number of active state tasks and assigns the total number to the corresponding on-duty personnel's concurrent operation task count. The aforementioned processing logic eliminates operation ticket data that has been terminated or not started, ensuring that the extracted task data reflects the actual workload of personnel at the current moment.

[0081] Step S420: Extract historical system logs and calculate operation response delay time. The system reads the historical operation logs stored on the system's backend server from the mobile terminal. The system has a preset historical time window of fixed length and traces back along the historical timeline starting from the current system time. The system extracts all historical alarm event commands received by the corresponding on-duty personnel within the historical time window. For each historical alarm event command, the system extracts the first timestamp of the push command sent by the backend server to the mobile terminal and the second timestamp of the on-duty personnel triggering the confirmation control on the mobile terminal's touch interface. The system calculates the time difference between the second and first timestamps to generate a single operation response delay record. The system summarizes all single operation response delay records within the historical time window and calculates the arithmetic mean. If there are no historical alarm event commands within the historical time window, the system assigns the preset standard response time constant as the operation response delay time. If historical alarm event commands exist within the historical time sliding window, the system will use the calculated arithmetic mean as the corresponding on-duty personnel's operation response delay time. For the data extraction and timestamp parsing rules of the system background log files, those skilled in the art can refer to the query syntax of standard relational databases to implement them. The corresponding log parsing operation is a well-known technology in this field and will not be described in detail here.

[0082] Step S430: Quantitatively assess the real-time cognitive load value. The system generates absorption node mapping elements for the mobile operation terminals of each on-duty personnel in the multi-dimensional topology graph database. The system synchronously inputs the number of concurrent operation tasks obtained in the previous steps. Operation response delay time The system calls a linear weighted model to calculate the real-time cognitive load values ​​for each absorption node. The specific calculation formula is as follows:

[0083]

[0084] In the formula, To understand load values ​​in real time, The number of concurrent operations to be extracted. To calculate the operation response delay time, This is the task concurrency load weighting coefficient. To determine the fatigue delay load weighting coefficient, the system retrieves the job qualification table linked to the employee ID of the on-duty personnel, extracts the job skill level parameters of the corresponding on-duty personnel, and establishes a negative correlation mapping rule between the job skill level parameters and the weighting coefficient. That is, the higher the level corresponding to the job skill level parameter, the higher the weighting coefficient for the task concurrency load. Weighting coefficient for fatigue delay load The smaller the assigned coefficient value, the more the aforementioned quantitative assessment method quantitatively transforms the personnel's task processing burden into a numerical input parameter for generating a cognitive repulsive field in the comprehensive potential energy field.

[0085] Furthermore, the technical implementation process of constructing a composite cognitive repulsive field by combining the spatial location of on-duty personnel with the physical state of the environment specifically includes the following steps:

[0086] Step S440: Extract the diffusion radius of the hazardous medium and the spatial distance to personnel. The system reads the environmental safety attribute field of the first alarm node in the multidimensional topology database. The system determines whether the environmental safety attribute field contains a hazardous medium label. If the environmental safety attribute field does not contain a hazardous medium label, the system will determine the dynamic diffusion radius of the hazardous medium. The value is set to zero; when the environmental safety attribute field contains a hazardous medium label, the system extracts the environmental wind speed parameters, medium leakage rate parameters, and atmospheric stability level parameters of the corresponding physical area. The system then inputs these parameters into the Gaussian plume meteorological diffusion model to calculate the dynamic diffusion radius of the hazardous medium. For the internal calculation formula of the Gaussian plume meteorological diffusion model, those skilled in the art can refer to the well-known meteorological diffusion calculation specifications for implementation. The internal mechanism of the diffusion model is a well-known technology in this field and will not be described in detail here.

[0087] The system obtains the real-time physical three-dimensional absolute coordinates of each absorption node in the multi-dimensional topology map through personnel positioning terminals. The system synchronously extracts the three-dimensional absolute coordinates of the device center point in physical space of the first alarm node. The system uses the Euclidean distance formula to calculate the physical spatial straight-line distance between each absorption node and the first alarm node. The specific distance calculation formula is as follows:

[0088]

[0089] In the formula, The straight-line distance in physical space. To absorb the real-time physical three-dimensional absolute coordinates of the node, The three-dimensional absolute coordinates of the device center point of the first alarm node;

[0090] Step S450: Calculate the environmental hazard penalty term based on spatial location. The system calculates the straight-line physical distance between each absorption node. Dynamic diffusion radius with hazardous media The system performs numerical comparisons and configures environmental hazard penalty items for each absorption node based on the comparison results. The configuration logic for environmental hazard penalty items is set as a piecewise function:

[0091]

[0092] In the formula, For environmental hazard penalties, To absorb the physical straight-line distance between the node and the first alarm node, The dynamic diffusion radius of the hazardous medium. As a preset limit penalty constant, the system retrieves the system's global configuration parameter table, extracts the maximum allowable boundary value of the real-time cognitive load value, and sets the limit penalty constant accordingly. The value is assigned as a preset multiple of the maximum allowable boundary value. The aforementioned assignment method enables the absorption node, which is within the diffusion radius of the hazardous medium, to obtain a large repulsive potential energy in a straight line distance from the physical space.

[0093] Step S460: Generate an asymmetric composite cognitive repulsive field with coupled multidimensional parameters; the system extracts the real-time cognitive load value of the on-duty personnel calculated by the previous calculation module. The system will recognize the load values ​​in real time. Environmental hazard penalties By performing linear superposition and fusion, the system generates a composite cognitive repulsive field around each absorbing node based on the superposition and fusion results. The specific equation for the composite potential energy is as follows:

[0094]

[0095] In the formula, To absorb the combined cognitive repulsive potential energy generated by the nodes, The preset repulsive field gain coefficient, To understand load values ​​in real time, This is a penalty item for environmental hazards;

[0096] The system will calculate the composite cognitive repulsive potential energy. The dynamic feature fields of the corresponding absorption nodes are written into the multidimensional topology database. Since the environmental hazard penalty term only acts on absorption nodes whose physical spatial straight-line distance is less than or equal to the dynamic diffusion radius, the generated composite cognitive repulsive field presents an asymmetric potential energy distribution state in the multidimensional topological coordinate system. The aforementioned calculation steps quantitatively transform the physical environmental hazard factors of the industrial site into repulsive boundary conditions that block the push of alarm information to personnel in the dangerous area.

[0097] Furthermore, this embodiment discloses the process of calculating the overall situational energy distribution and performing alarm information routing and addressing, specifically including the following steps:

[0098] Step S510: Extract professional matching features and construct an attractive potential energy field. The system reads the attribute fields of the macroscopic alarm event vector, extracts the process system code of the physical equipment that triggered the alarm, and synchronously traverses each absorption node in the multi-dimensional topology diagram. The system retrieves the professional qualification parameters bound to the mobile operation terminal corresponding to each absorption node. The system compares the identifier consistency between the process system code and the professional qualification parameters. When the comparison result is consistent, the system determines that the corresponding absorption node has the professional authority to handle the current alarm event and assigns a preset attractive potential energy constant to the corresponding absorption node. The system retrieves the global configuration parameter table and sets the attraction potential energy constant. The numerical setting is greater than the maximum value of the composite cognitive repulsive potential energy recorded in the multidimensional topology database. This numerical setting ensures that the priority of professional matching is always higher than the repulsive effect caused by spatial distance and cognitive load in the potential energy calculation, avoiding the optimization deadlock phenomenon where alarms cannot be issued due to excessive repulsive force from all personnel in extremely high-load or high-risk environments. When the comparison results are inconsistent, the system will set the attractive potential energy constant of the corresponding absorbing node. By forcibly assigning a value of zero, the aforementioned comparison operation constructs an attractive potential field that guides alarm information to the corresponding professional terminal.

[0099] Step S520: The system generates a comprehensive receiving potential energy function through algebraic superposition. It extracts the composite cognitive repulsive field potential energy parameters corresponding to each absorbing node in the multidimensional topology graph. The system then algebraically superimposes the composite cognitive repulsive field with the attractive potential energy field to generate a comprehensive potential energy function. Finally, the system calculates the comprehensive receiving potential energy of each absorbing node using this comprehensive potential energy function. The specific calculation formula is as follows:

[0100]

[0101] In the formula, For the first The combined received potential energy of each absorbing node For the first The composite cognitive repulsive potential energy of each absorbing node For the first The system assigns an attractive potential energy constant to each absorption node and writes the calculated comprehensive received potential energy into the potential energy status register of each absorption node.

[0102] Step S530: Perform steepest gradient descent calculation and target node optimization. The system extracts the spatial coordinates of the first alarm node contained in the macro alarm event vector, sets the spatial coordinates of the first alarm node as the addressing calculation starting point, and reads the multi-dimensional topology graph database to obtain the physical spatial straight-line distance between each absorbing node and the addressing calculation starting point. The system is based on the steepest gradient descent principle, calculating the slope of potential energy decrease from the addressing computation starting point to each absorbing node. The calculation formula is:

[0103]

[0104] In the formula, The starting point for addressing calculation points to the first The slope of potential energy decrease along the direction of each absorbing node. For the first The combined received potential energy of each absorbing node The initial reference potential energy for addressing the starting point of the calculation (the system default value is zero). For the first The physical spatial straight-line distance between each absorbing node and the addressing calculation start point;

[0105] The system establishes an optimization comparison array, traverses the potential energy decrease slope values ​​corresponding to all absorbing nodes, and finds the node with the smallest potential energy decrease slope value. The node with the smallest potential energy decrease slope value is identified as the target receiving node. The optimal decision equation is:

[0106]

[0107] The aforementioned mathematical calculation process transforms the network routing task into an equivalent energy gradient descent process in a potential energy field, enabling the system to solve for the push path with the minimum overall resistance under multidimensional constraints.

[0108] Furthermore, this embodiment discloses the implementation process of performing underlying data encapsulation, network transmission, and potential field state reset after establishing the target receiving node, specifically including the following steps:

[0109] Step S540: Execute the serialization and encapsulation of the macro alarm event vector. The system reads the network configuration attributes of the target receiving node determined by the aforementioned gradient optimization calculation. The system extracts the network protocol address and communication port number bound to the target receiving node. The system extracts the multi-dimensional topology coordinate vector of the first alarm node, the dynamic virtual quality, and the set of coordinate vectors of all cascaded events contained in the macro alarm event vector. The system calls the preset data serialization algorithm to convert the aforementioned extracted data parameters into a binary byte stream. The system adds a routing header containing the network protocol address of the target receiving node to the beginning of the binary byte stream to generate an application layer network data packet.

[0110] Step S550: Establish a communication link and execute network-directed push stream. The system establishes a data communication link with the mobile operating terminal bound to the target receiving node through the underlying network communication interface. The system sends application layer network data packets to the mobile operating terminal through the data communication link. For the underlying transmission control and redundancy verification of application layer network data packets, those skilled in the art can refer to the international standard specifications of transmission control protocols for implementation. The corresponding underlying transmission communication mechanism is a well-known technology in this field and will not be described in detail here.

[0111] Step S560: Execute terminal deserialization parsing and operation confirmation command acquisition. The mobile operating terminal receives application layer network data packets sent by the system. The mobile operating terminal deserializes and decodes the application layer network data packets, extracting the multi-dimensional topology coordinate vector of the initial alarm node, the dynamic virtual quality, and the coordinate vector set of all cascading events. The mobile operating terminal renders an alarm prompt window in the graphical user interface and marks the physical fault location on the built-in electronic map of the mobile operating terminal based on the multi-dimensional topology coordinate vector of the initial alarm node. The mobile operating terminal acquires the touch confirmation signal triggered by the on-duty personnel in the graphical user interface and generates a processing confirmation timestamp based on the touch confirmation signal. The system sends a receipt message with the terminal device identifier and uploads the receipt message to the system backend.

[0112] Step S570: Trigger potential field state reset and data cleanup. The system background receives the receipt message uploaded by the mobile operation terminal. The system verifies whether the terminal device identifier in the receipt message matches the target receiving node. The system extracts the processing confirmation timestamp from the receipt message. The system confirms the processing timestamp. The potential field state reset logic is triggered. The system locates the first alarm node that triggered the current push in the multi-dimensional topology graph database. The system forcibly assigns the dynamic virtual mass of the first alarm node to zero. The corresponding state update expression is:

[0113]

[0114] In the formula, For system time The corresponding dynamic virtual quality, The initial dynamic virtual mass is obtained by solving the initial time of the limit-breaking trigger. The processing confirmation timestamp returned by the mobile operating terminal;

[0115] After the system detects that the dynamic virtual mass has returned to zero, it synchronously cancels the gravitational field potential energy distribution radiating from the first alarm node. The system cleans up the merged state parameters of the derivative alarm nodes that are confirmed as cascaded events in the multidimensional topology graph database and removes the independent push interception restriction of the derivative alarm nodes. The aforementioned data cleanup operation completes the processing cycle of the current alarm event and restores the initial state of the gravitational field in the multidimensional topology graph database.

Claims

1. A method for alarm and event push for power plant safety management, characterized in that, Includes the following steps: Step S10, Constructing a multi-dimensional topology mapping and continuous status monitoring: Generate a multi-dimensional topology map and continuously synchronize the real-time operating parameters of the physical devices corresponding to each device node in the multi-dimensional topology map; Step S20: Capture the first over-limit event and solve the dynamic virtual quality: When the real-time operating parameters corresponding to the device node exceed the preset safety threshold, the device node that triggers the over-limit alarm is marked as the first alarm node, and the underlying physical state change rate of the physical device corresponding to the first alarm node is extracted to solve the dynamic virtual quality of the first alarm node in the multi-dimensional topology graph. Step S30: Generate an anisotropic gravitational field and perform cascaded alarm fusion: Generate an outward radiating gravitational field in the multidimensional topology map based on dynamic virtual mass, and when it is determined that the surrounding associated device nodes generate derivative alarms and meet the cascaded event conditions, merge the alarm information data fields to generate a unified macroscopic alarm event vector. Step S40, quantify personnel cognitive load and generate a composite repulsive field: extract the number of concurrent operation tasks currently active for each on-duty personnel and the operation response delay time, calculate the real-time cognitive load value of each on-duty personnel, and combine it with the environmental hazard parameters of the physical area where the first alarm node is located to generate a composite cognitive repulsive field around the absorption node corresponding to the mobile operation terminal of each on-duty personnel. Step S50: Calculate the comprehensive potential energy gradient and perform adaptive push flow: Superimpose the gravitational field and the composite cognitive repulsive field to construct a global comprehensive potential energy field, find the target mobile operating terminal corresponding to the minimum value of the total received potential energy along the negative direction of the total potential energy gradient, and encapsulate the macroscopic alarm event vector into a network data packet and send it to the target mobile operating terminal.

2. The alarm and event push method for power plant safety management according to claim 1, characterized in that, Step S20 includes: Step S210: Identify the process attribute boundary of the first alarm node: extract the process system code to which the first alarm node belongs, and divide the data processing flow into the thermal fluid solution domain or the electrical protection solution domain. Step S220: Calculate the thermodynamic mass components within the thermal fluid solution domain: extract the fluid medium pressure and temperature parameters and map them to real-time thermodynamic enthalpy values; obtain the thermodynamic mass components using differential equations. : In the formula, The pressure change weighting coefficient is used. The enthalpy drop change weighting coefficient is used. The time partial derivative of the fluid medium pressure parameter, The rate of thermodynamic enthalpy drop; Step S230: Calculate the electrical quality components within the electrical protection solution domain: extract the steady-state voltage at the initial moment of over-limit triggering and perform a subtraction operation with the transient voltage at the current sampling moment to obtain the transient voltage drop amplitude. And extract the instantaneous value sequence of zero-sequence current. Calculate electrical quality components : In the formula, This is the voltage sag weighting factor. The zero-sequence current integral weighting coefficient is... To trigger the initial time of exceeding the limit, This is the current sampling time; Step S240: Generate dynamic virtual mass matrix parameters: Construct a state column vector containing thermodynamic mass components and electrical mass components, and perform matrix multiplication to obtain the dynamic virtual mass.

3. The alarm and event push method for power plant safety management according to claim 1, characterized in that, Step S30, which generates an anisotropic gravitational field, includes: Step S310, Traverse the multidimensional topology branches and extract connected paths: Starting from the first alarm node, traverse the multidimensional topology graph and extract the shortest topology path connecting the first alarm node and topology-related nodes, as well as the corresponding connected distance values. ; Step S320, parse the physical transmission medium category and map the medium delay constant: parse the physical transmission medium category corresponding to the network edge attributes contained in the shortest topology path to map the medium delay constant. When the physical transmission medium is a fluid process pipeline, the fluid velocity and specific heat capacity parameters are extracted to calculate the thermal inertia index and mapped to the fluid medium's corresponding time delay constant. When the physical transmission medium is an electrical conductor, the medium's time delay constant is directly assigned to a preset minimum value. Step S330: Calculate the spatial potential energy distribution of the anisotropic gravitational field: Calculate the gravitational potential energy experienced by the topologically connected nodes in the gravitational field using the gravitational potential energy spatial radiation equation. : In the formula, For dynamic virtual quality.

4. The alarm and event push method for power plant safety management according to claim 1, characterized in that, The process of performing cascading alarm fusion includes: Step S340: Set the potential energy gradient adsorption threshold and the causal tracing time window: Determine whether the triggering time of the derived alarm falls within the time interval of the causal tracing time window; Step S350: Perform potential energy comparison and cascade event determination in the spatial dimension: calculate the gravitational potential energy of the derived alarm node in the gravitational field. If the gravitational potential energy is greater than or equal to the potential energy gradient adsorption threshold and the triggering time falls within the time interval of the causal tracing time window, then the derived alarm is confirmed as a cascade event. Step S360: Intercept independent push commands and generate macro alarm event vectors: Cancel the push task of sending independent push data packets by derivative alarm nodes, and extract the status parameters of the initial alarm node and the derivative alarm nodes confirmed as cascaded events to construct macro alarm event vectors. ， In the formula This is the coordinate vector of the first alarm node. For the set of coordinate vectors of cascading events, To generate a timestamp.

5. The alarm and event push method for power plant safety management according to claim 1, characterized in that, Step S40, which calculates the real-time cognitive load value, includes: Step S410, identify and count the number of concurrent tasks in the active state: count the total number of operation tickets in the on-site execution state and the state waiting for monitoring confirmation, and assign the value as the number of concurrent operation tasks. ; Step S420: Extract historical system logs and calculate operation response latency: Calculate the time difference between the first timestamp of the push command arriving and the second timestamp of the confirmation control being triggered, and use the arithmetic mean within the historical time slider as the operation response latency. ; Step S430, Quantitatively assess real-time cognitive load: Call the linear weighted model Calculate the real-time cognitive load value of the absorption node. Furthermore, the higher the job skill level parameter, the higher the weighting coefficient. and The smaller the coefficient value, the better.

6. The alarm and event push method for power plant safety management according to claim 5, characterized in that, The process of generating a composite cognitive repulsive field includes: Step S440, extract the diffusion radius of hazardous media and spatial distance from personnel: When the initial alarm node contains a hazardous media tag, calculate the dynamic diffusion radius of the hazardous media. And solve for the physical spatial straight-line distance between the absorbing node and the first alarm node. ; Step S450, calculate the environmental hazard penalty term based on spatial location: when the physical space straight-line distance At that time, environmental hazard penalties will be applied. The value is assigned to the limit penalty constant. ; Step S460, Generate an asymmetric composite cognitive repulsive field with coupled multidimensional parameters: This involves applying real-time cognitive load values... Environmental hazard penalties Linear superposition generates composite cognitive repulsive potential energy .

7. The alarm and event push method for power plant safety management according to claim 1, characterized in that, Step S50, constructing the global integrated potential field, includes: Step S510, extract professional matching features and construct an attractive potential energy field: compare the identifier consistency between the process system code of the physical equipment that triggered the alarm and the professional qualification parameters bound to the absorption node. If the comparison is consistent, assign an attractive potential energy constant. The value of the attractive potential energy constant is set to be greater than the maximum value of the composite cognitive repulsive potential energy; Step S520, algebraic superposition to generate the comprehensive receiving potential energy function: The composite cognitive repulsive field and the attractive potential energy field are algebraically superimposed, with the following formula: In the formula, For the first The combined received potential energy of each absorbing node.

8. The alarm and event push method for power plant safety management according to claim 7, characterized in that, The process of performing adaptive streaming also includes: Step S530: Perform steepest gradient descent calculation and target node optimization: Set the spatial coordinates of the first alarm node as the starting point of the addressing calculation, and calculate the potential energy descent slope pointing towards the absorbing node. ， In the formula The initial reference potential energy is used to determine the starting point for addressing calculations; the node with the smallest algebraic value of the potential energy decrease slope is identified as the target receiving node. .

9. The alarm and event push method for power plant safety management according to claim 1, characterized in that, Step S10, the process of constructing the multidimensional topology mapping, includes: Step S110, construct a multi-dimensional topology node set: extract the three-dimensional spatial absolute coordinates of the center point of the physical device and the process system code, and generate a topology node coordinate vector containing four dimensions as a vertex set; Step S120, Constructing a multidimensional topological edge set and calculating the connectivity distance: When nodes are connected by electrical cables, extract the equivalent resistance parameters and equivalent reactance parameters to calculate the magnitude of the total impedance. ， The magnitude of the total impedance is then mapped to the dielectric connectivity distance of the corresponding network edge.

10. The alarm and event push method for power plant safety management according to claim 8, characterized in that, After step S530, which optimizes the target receiving node, the following steps are also included: Step S540: Perform serialization and encapsulation of macro alarm event vectors: convert the multi-dimensional topological coordinate vector of the first alarm node, the dynamic virtual mass and the set of cascaded event coordinate vectors into a binary byte stream and encapsulate it into an application layer network data packet; Step S550: Establish a communication link and perform network-directed push: Establish a data communication link with the mobile operating terminal bound to the target receiving node through the underlying network communication interface, and send application layer network data packets to the mobile operating terminal through the data communication link; Step S560: Perform terminal deserialization parsing and operation confirmation instruction acquisition: The mobile terminal deserializes and parses application layer network data packets, and uploads a data packet containing a processing confirmation timestamp after acquiring the touch confirmation signal. The receipt message; Step S570, trigger potential field state reset and data cleanup: based on the processing confirmation timestamp. The dynamic virtual mass of the first alarm node is forcibly assigned to zero, the gravitational field potential energy distribution is canceled, and the independent push interception restriction of the derived alarm nodes is removed.

Images