Power plant emergency resource intelligent configuration method and system based on emergency scene

Abstract

The invention discloses a power plant emergency resource intelligent configuration method and system based on an emergency scene, and belongs to the technical field of power plant emergency resource configuration. According to the method, the emergency scene information covering key parameters such as the accident type, the influence range and the severity level is acquired in real time, and the emergency resource optimal configuration algorithm model is called to perform dynamic calculation and analysis in combination with the preset database containing the emergency resource type and quantity. The algorithm model can comprehensively evaluate scene requirements and resource conditions, and generates an optimal emergency resource allocation scheme for the current emergency scene, including resource types, quantity, allocation priority, suggested paths and the like. The emergency response speed, decision making scientificity and resource utilization efficiency of the power plant under emergencies can be remarkably improved, and therefore the comprehensive capacity and the safety guarantee level of the power plant for dealing with various complex emergency scenes are enhanced.

Description

Technical Field

[0001] This invention relates to the field of emergency resource management technology, and more specifically, to a method for intelligent allocation of emergency resources in power plants based on emergency scenarios. Background Technology

[0002] The power system is a critical national infrastructure, and power plants, as the core link in power production, are of paramount importance for their safe and stable operation. However, power plants may face various emergencies during operation, including natural disasters (such as earthquakes, floods, and extreme weather), equipment failures, human sabotage, or safety accidents. These events can all lead to severe power production interruptions, equipment damage, and even casualties. To effectively respond to these emergencies and minimize losses and impacts, power plants need to establish a sound emergency management system, in which the rational allocation of emergency resources is a key factor in the success of emergency response.

[0003] Currently, many power plants rely primarily on pre-established emergency plans and human experience for emergency resource allocation. Traditional emergency plans are often designed for a few typical accident scenarios, with relatively fixed content, lacking adaptability to the complexity and dynamism of actual emergencies. When unexpected or complex accidents involving multiple factors occur, fixed plans are insufficient to provide optimal resource allocation strategies. Human dispatch is easily influenced by the decision-maker's experience, the comprehensiveness of information access, and psychological pressure under emergency conditions, potentially leading to slow response, resource misallocation, or waste, thereby reducing the overall efficiency of emergency response.

[0004] With the development of information technology, some emergency command systems have begun to incorporate information technology, such as building emergency command platforms to share information and resources, and attempting to apply intelligent decision-making support functions. However, many challenges remain in the field of power plant emergency resource allocation:

[0005] (1) Difficulties in data integration and standardization: Power plant emergency resources are diverse, including personnel, equipment, materials, spare parts, etc. The information on these resources may be scattered in different departments or systems, lacking a unified data standard and real-time update mechanism, making it difficult to quickly and accurately grasp the overall status of resources.

[0006] (2) Insufficient adaptability to scenarios: Traditional methods are difficult to quickly analyze and adapt to complex and ever-changing emergency scenarios. For example, extreme weather events have an increasingly significant impact on the power system. They are characterized by a wide range of impacts, strong destructiveness, and are often accompanied by multiple secondary disasters, which puts forward higher requirements for the dynamism and accuracy of resource allocation.

[0007] (3) Low level of intelligent decision-making: Although some algorithms are applied to resource scheduling in certain fields, intelligent configuration algorithms that take into account power plant-specific emergency scenarios, unique power plant resources, and operational constraints are still immature. Existing systems struggle to quickly generate and optimize resource allocation schemes when faced with a large amount of real-time information and multiple constraints.

[0008] (4) Low efficiency in resource coordination and optimization: In complex emergency events, it is often necessary to coordinate the operation of resources of various types and from multiple sources. How to achieve the optimal combination and efficient scheduling of these resources and avoid the phenomenon of "gathering without coordination" is the key to improving emergency response capabilities.

[0009] For example, patent CN113890883B discloses a general human-machine flexible access system and method for dispatching emergency systems, which mainly solves the problems of flexible human-machine access and site mutual backup between primary and backup dispatching systems, but does not involve intelligent optimization and allocation of emergency resources based on specific emergency scenarios. Patent CN109546747B discloses a working method of an intelligent power transmission and distribution network fault early warning management system, which focuses on fault early warning and information reporting. Its emergency response part relies on a preset emergency mechanism and does not elaborate on how to dynamically optimize and allocate resources according to real-time changing scenarios.

[0010] Therefore, there is an urgent need for a method that can intelligently, dynamically, and optimally allocate various existing emergency resources in power plants based on real-time emergency scenario information and advanced algorithm models, in order to improve the timeliness, accuracy, and resource utilization efficiency of emergency response. Summary of the Invention

[0011] The technical problem to be solved by this invention is: in order to overcome the technical problems existing in the background technology, such as slow response, low level of intelligence, inability to effectively adapt to complex and ever-changing emergency scenarios, and low resource utilization rate of power plant emergency resource allocation, a method for intelligent allocation of power plant emergency resources based on emergency scenarios is proposed. Through intelligent means, based on real-time acquired emergency scenario information and the power plant's own emergency resource reserves, a scientific and reasonable emergency resource allocation plan is quickly generated, thereby improving the power plant's ability to respond to emergencies.

[0012] To solve the above technical problems, the present invention adopts the following technical solution:

[0013] First, this invention provides a method for intelligent allocation of emergency resources in power plants based on emergency scenarios. The core of this method lies in its dynamic adaptability and optimization decision-making capabilities. The method includes the following key steps:

[0014] The first step is to obtain emergency scenario information. This information forms the basis for resource allocation decisions and should at least include the type of accident (e.g., fire, equipment damage, natural disaster), the specific location of the accident, the potential scope of impact, and the severity level of the accident. In some specific applications, emergency scenario information can further include more detailed data, such as the specific damage to power equipment, the number and model of generator units that are out of service, the reduction in grid load, or the actual area of ​​power outage. This information can come from internal power plant monitoring systems (e.g., SCADA), various sensors, manual reports, or external information systems (e.g., weather warnings, earthquake monitoring).

[0015] The second step involves using the acquired emergency scenario information and a pre-set emergency resource database containing detailed information on power plant emergency resources. A core emergency resource optimization and allocation algorithm model is then invoked for calculation and analysis. This emergency resource database is a crucial support for this method, recording the categories (e.g., emergency repair teams, fire-fighting equipment, medical supplies, backup power supplies, communication equipment), quantities, current status (e.g., availability, location, condition), and other relevant attributes (e.g., professional skills, power parameters) of various emergency resources possessed by the power plant. The emergency resource optimization and allocation algorithm model is the core technology of this invention. It is designed to find the most effective resource combination to meet current emergency needs by performing calculations such as multi-objective optimization based on complex, multi-dimensional emergency scenario information and real-time resource availability data. Here, "optimization" is reflected in the fact that the algorithm model is not merely a simple match, but strives to achieve certain pre-set optimization objectives while meeting basic emergency needs, such as the shortest response time, the most economical resource consumption, or the best handling effect.

[0016] The third step involves generating a specific and executable emergency resource allocation plan based on the calculation and analysis results of the emergency resource optimization and allocation algorithm model. This plan serves as the direct basis for emergency response actions and should at least clearly specify the specific types of emergency resources required (e.g., a fire truck of model XX, an engineering team with YY skills), the accurate required quantities, the allocation priority of each resource (to address situations with limited resources), and suggested dispatch instructions (e.g., resource dispatch routes, estimated arrival times, task assignments, etc.).

[0017] To enable the above method to handle more complex real-world situations and improve its intelligence, further optimization features can be incorporated. For example, the emergency resource optimization allocation algorithm model itself can be designed to include multiple functional modules, such as:

[0018] A scenario requirements analysis module is specifically responsible for deeply analyzing the input emergency scenario information and transforming it into quantitative requirements indicators for various emergency resources.

[0019] A resource matching and evaluation module efficiently searches for matching available resources in the emergency resource database based on these quantitative requirements, and quickly evaluates the suitability (e.g., whether the functions are sufficient, whether the capabilities are adequate) and availability (e.g., whether they are on standby, distance, etc.) of these resources; and

[0020] A multi-objective optimization decision-making module, the core of the algorithm, is required to meet the core emergency objectives (such as controlling the situation, saving lives, and restoring critical functions). It comprehensively weighs multiple possible secondary optimization objectives (e.g., minimizing overall response time, reducing the consumption of valuable resources, or achieving the best long-term outcome) to ultimately generate an optimal or near-optimal emergency resource allocation plan. In its implementation, this multi-objective optimization decision-making module can employ mature heuristic algorithms (such as genetic algorithms and particle swarm optimization) or advanced machine learning models (such as reinforcement learning and deep learning networks) to drive its optimization decision-making process.

[0021] Furthermore, to ensure the continued effectiveness of emergency response, this method can also include a closed-loop mechanism for dynamic feedback and adjustment. Specifically, it can monitor changes in the on-site situation during emergency response (e.g., whether the fire is spreading, whether new fault points are appearing) and the consumption of allocated resources (e.g., whether a repair team has completed its task, whether a certain material is about to run out). Once a significant change in the scenario or an update in the resource status is detected, the emergency resource optimization and allocation algorithm model can be re-triggered to dynamically adjust and optimize the original configuration scheme, thereby ensuring that the emergency strategy always remains consistent with the actual situation.

[0022] This invention combines artificial intelligence algorithms with the specific needs of power plant emergency management through the above technical solution, aiming to achieve rapid response, intelligent decision-making, dynamic optimization and efficient utilization of emergency resources.

[0023] Secondly, this invention provides a smart power plant emergency resource allocation device based on emergency scenarios, comprising:

[0024] The emergency scenario information acquisition unit is used to acquire emergency scenario information including one or more of the following: accident type, accident location, scope of impact, and accident severity level.

[0025] An emergency resource optimization and allocation unit is used to perform calculations and analyses by calling an emergency resource optimization and allocation algorithm model based on the emergency scenario information and a preset emergency resource database that stores emergency resource categories, quantities, and statuses. The emergency resource optimization and allocation algorithm model is configured to perform multi-objective optimization calculations based on the emergency scenario information and the data in the emergency resource database to determine the resource combination that meets emergency needs.

[0026] The scheme generation unit is used to generate an emergency resource configuration scheme for the emergency scenario, including the specific type of emergency resources required, the required quantity, the allocation priority, and the suggested scheduling instructions.

[0027] According to another aspect of the present invention, a terminal device is provided, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. The memory stores the computer program capable of running on the processor, and when the processor loads and executes the computer program, it employs the intelligent configuration method for power plant emergency resources provided by the present invention.

[0028] Meanwhile, the present invention also proposes an electronic system, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to execute the intelligent configuration method for power plant emergency resources provided by the present invention.

[0029] Compared with the prior art, the present invention, employing the above technical solution, has the following technical effects:

[0030] First, this invention improves emergency response speed and decision-making efficiency. Specifically, through automated information acquisition, intelligent algorithm analysis, and rapid solution generation, this invention greatly shortens emergency decision-making time, enabling emergency resources to be allocated and deployed in the shortest possible time, thus gaining valuable time for the successful handling of emergencies.

[0031] Secondly, this invention improves the scientific nature and accuracy of resource allocation. Specifically, based on real-time and specific emergency scenario information and comprehensive resource data, this invention uses an optimized algorithm model to make decisions, avoiding the subjective assumptions and information asymmetry problems that may exist in traditional manual scheduling. It can accurately match and mobilize the most suitable resources according to actual needs, thereby improving the rationality of resource allocation.

[0032] Meanwhile, this invention enhances adaptability to complex and ever-changing scenarios. Specifically, the algorithm model of this invention can handle multi-dimensional and dynamically changing scenario information and supports multi-objective optimization, enabling it to better adapt to various complex and unexpected emergency scenarios, including chain reactions or compound disasters triggered by extreme weather and other factors. The method also possesses dynamic adjustment capabilities, allowing for real-time optimization of configuration schemes based on the development of the situation. Through global optimization using intelligent algorithms, the optimal resource combination and scheduling path can be selected as much as possible while meeting emergency needs, avoiding unnecessary resource mobilization and redundant configuration, thereby effectively saving valuable emergency resources and reducing emergency costs.

[0033] Finally, this invention supports knowledge accumulation and continuous improvement. Specifically, by learning and analyzing historical emergency event data, this invention can continuously optimize the algorithm model and emergency plan database, thereby achieving continuous iterative improvement of emergency management capabilities and making the power plant's emergency response methods increasingly intelligent and efficient.

[0034] In summary, by enhancing the intelligence and automation of emergency resource allocation, this invention helps strengthen the capacity building of power plants in all aspects, including emergency preparedness, prevention and early warning, response and rescue, and recovery and reconstruction. This improves the overall emergency support capability and safe operation level of power plants, thereby comprehensively enhancing their safety resilience and continuous operation guarantee capability. Attached Figure Description

[0035] Figure 1 This is a flowchart of a power plant emergency resource intelligent allocation method according to an embodiment of the present invention.

[0036] Figure 2 This is a detailed flowchart or module diagram of the emergency resource optimization and allocation algorithm model according to an embodiment of the present invention.

[0037] Figure 3 and Figure 4 The illustrations illustratively demonstrate exemplary emergency resource allocation schemes that the method of this invention may generate under two typical and different emergency scenarios. Figure 3 This is an example of a resource allocation plan for an emergency scenario of flooding in a factory. Figure 4 This is an example of a resource allocation plan for a plant-wide power outage emergency scenario. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0039] Example 1: Refer to Figure 1The present invention provides a method for intelligent allocation of power plant emergency resources based on emergency scenarios, which mainly includes the following steps:

[0040] Step S210: Emergency event triggering and scene information acquisition.

[0041] When a power plant experiences an emergency (such as equipment malfunction alarm, fire detector triggering, receiving external disaster warnings, or personnel reports), the emergency response process is activated. Immediately, a comprehensive collection of initial information related to the event begins. As mentioned earlier, this information may include the type of incident, time of occurrence, precise location, preliminary assessment of the impact area, known equipment damage, whether any personnel are trapped or injured, and current weather conditions. This information is then compiled and used for subsequent scenario analysis and requirements assessment.

[0042] Step S220: Emergency scenario analysis and needs assessment.

[0043] Upon receiving the initial scenario information, the emergency resource optimization and allocation algorithm model 131 first activates its scenario analysis and demand assessment submodule (see...). Figure 2 Module 310 (of the module). This submodule uses a pre-defined rule base, case base, or machine learning-based model to analyze and judge the input scene information in order to achieve the following objectives:

[0044] (1) Determine the event level: Based on the nature of the event and its initial impact, assess its severity and determine the corresponding emergency response level.

[0045] (2) Predicting development trends: Combining factors such as the power plant's process flow, equipment layout, and surrounding environment, predict the possible evolution direction of the event and potential chain reactions.

[0046] (3) Quantify resource requirements: Based on the event type, level and predicted trend, decompose the abstract emergency objectives (such as fire extinguishing, leak plugging, emergency repair, evacuation, etc.) into quantitative requirements for specific emergency resources, such as how many tons of fire water, how many specific models of smoke exhaust fans, how many electricians with high voltage operation qualifications, how many sets of chemical protective suits, etc., and preliminarily determine the urgency (priority) of each requirement.

[0047] Step S230: Inquire about and assess available emergency resources.

[0048] Based on the assessed resource needs, the real-time availability status of various relevant emergency resources is retrieved from the emergency resource database, including quantity, location, technical parameters, condition, and estimated availability time. The resource matching and assessment submodule in the emergency resource optimization allocation algorithm model 131 (see...) Figure 2 Module 320 in the module performs an applicability assessment on the queried resources to determine whether they meet the specific needs of the current scenario.

[0049] Step S240: Execute the optimization configuration algorithm to generate a preliminary solution.

[0050] This is the core step of the method. The multi-objective optimization decision-making submodule in the emergency resource optimization allocation algorithm model 131 (see...) Figure 2 Module 330 takes the quantified resource requirements and the assessed available resources as input and performs optimization calculations.

[0051] The design of this algorithm model 131 is the key innovation of this invention. It can be constructed based on various principles, such as:

[0052] Rule-based reasoning: It utilizes expert knowledge and historical experience to build a rule base and matches corresponding configuration rules to different scenario features.

[0053] Case-based reasoning (CBR): Learning from successful resource allocation experiences in similar historical emergency events.

[0054] Mathematical programming models, such as linear programming, integer programming, and dynamic programming, are used to establish and solve mathematical models with objectives such as minimizing response time, maximizing disposal effectiveness, or minimizing resource consumption.

[0055] Heuristic optimization algorithms, such as Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), and Simulated Annealing (SA), are used to search for approximate optimal solutions in complex solution spaces. For example, the approach of NSGA-II for handling multi-objective optimization problems can be adopted, but the objective function and constraints need to be redefined for power plant emergency scenarios.

[0056] Machine learning models, especially reinforcement learning (RL), can learn, through interaction with a simulated environment, which resource allocation actions (decision-making) will yield the greatest cumulative reward (such as the fastest power restoration or the least loss) in different states (scenarios).

[0057] When optimizing algorithm model 131, the following factors need to be considered comprehensively:

[0058] (1) Demand satisfaction: Ensure that key demands are met first.

[0059] (2) Timeliness of response: Prioritize mobilizing resources that are close by and have short preparation time.

[0060] (3) Resource effectiveness: Select the resources that best match the task and have the strongest capabilities.

[0061] (4) Cost economy: While meeting the requirements of the desired effect, save resources as much as possible.

[0062] (5) Safety and reliability: Avoid new risks caused by improper resource allocation.

[0063] (6) Efficiency of collaborative work: Consider the coordination between different resources.

[0064] The algorithm outputs one or more preliminary emergency resource allocation schemes, each of which details the list of resources to be called up, their quantity, source, task allocation, and expected time nodes.

[0065] Step S250: Solution evaluation, adjustment and confirmation.

[0066] The initially generated configuration plan can be evaluated by emergency commanders. Commanders can combine their own experience with a more detailed understanding of the situation to evaluate the plan. Auxiliary evaluation tools can also be provided, such as simulation exercises to predict the plan's potential effects and problems.

[0067] If the command personnel deem the plan feasible, it will be approved. If adjustments are deemed necessary, the plan can be modified (e.g., replacing a resource, adjusting priorities, adding backup resources, etc.). This method supports manual adjustments to the plan and allows for re-verification of the feasibility and consistency of the adjusted plan.

[0068] Step S260: Issue scheduling instructions and monitor execution.

[0069] Once the configuration plan is finalized, the specific dispatch instructions included in the plan can be issued to relevant emergency response units and personnel. For example, sending task instructions and location navigation to the head of the repair team, sending material release notices to the material warehouse, and requesting support from the fire department.

[0070] Simultaneously, execution monitoring is initiated to track the availability of resources, the progress of tasks, and changes in the on-site situation in real time by acquiring scenario information and feedback from each execution unit.

[0071] Step S270: Dynamic feedback and scheme adjustment.

[0072] Emergency response is a dynamic process. During execution, if the situation monitored by S260 indicates that:

[0073] Significant changes in the scenario: for example, the fire spreads out of control to a new area, or unexpected secondary disasters occur.

[0074] The planned resources could not be delivered or became invalid: for example, a team was blocked en route or a piece of equipment malfunctioned.

[0075] Resource consumption exceeds expectations: For example, fire extinguishing agent is consumed too quickly. The system will return to step S210 or S220 to obtain the latest scenario information and rerun the emergency resource optimization configuration algorithm model 131 to dynamically adjust and optimize the original configuration scheme, generate new instructions, and form a closed-loop "perception-decision-action-feedback" cycle until the emergency event is resolved.

[0076] Example 2: Detailed Explanation of the Emergency Resource Optimization Algorithm Model

[0077] Reference Figure 2 This paper provides a more detailed explanation of the internal structure and workflow of the emergency resource optimization and allocation algorithm model 131. This model can be viewed as an intelligent decision-making engine composed of multiple collaborative sub-modules.

[0078] 1. Module 310: Scenario Understanding and Requirement Quantification Module.

[0079] This module receives the raw emergency scenario information from step S210. Its core function is to transform unstructured or semi-structured scenario descriptions into structured data that can be processed by the algorithm model, and to quantify the specific needs for emergency resources.

[0080] (1) Scene feature extraction: Natural language processing (NLP) technology is used to parse the text report, image recognition technology is used to analyze the surveillance video, and sensor data is combined to extract key scene features, such as event type (fire, explosion, leakage, flooding, equipment tripping, etc.), accurate location (factory, equipment number, coordinates), scope of impact (preliminary assessment), severity level (according to preset standards), current status (developing, under control, deteriorated), special environmental factors (toxic gas, high temperature, confined space), etc.

[0081] (2) Demand rule matching / model prediction:

[0082] Rule-based: A built-in "scenario-resource requirement" rule base is provided. For example, the scenario "the main transformer's A-phase bushing is on fire, with moderate fire intensity" might match the rule: "Immediately dispatch 1 high-voltage electrical fire brigade, 2 foam fire trucks, and 10 dry powder fire extinguishers; prepare 1 electrical repair team, 1 spare bushing, and X liters of insulating oil." The rule base is pre-built by domain experts and can be dynamically updated.

[0083] Model-based: Employing machine learning models (such as classifiers, regression models, or more complex sequence models), this approach directly predicts the types and quantities of resources needed based on the characteristics of the input scenario. The model is trained on a large number of historical emergency cases (including simulation data). For example, given the input "A magnitude 7 earthquake occurred in a certain area, causing partial collapse of the XX turbine building and rupture of the main steam pipeline," the model can predict the need for heavy lifting equipment, life detectors, medical rescue teams, structural engineers, pipeline repair teams, etc., and estimate their approximate quantities.

[0084] (3) Prioritization of needs: Based on the severity of the incident, its impact on personnel safety and core equipment, and the scarcity of the resources themselves, the quantified resource needs are prioritized. For example, rescuing trapped personnel has the highest priority, followed by controlling the spread of the fire, and then restoring power to critical equipment.

[0085] 2. Module 320: Resource Matching and Availability Assessment Module.

[0086] This module receives a quantified and prioritized list of resource requirements from module 310 and queries and evaluates available resources from the emergency resource database.

[0087] (1) Resource retrieval: Based on the resource type, specifications, skill requirements, etc., perform precise or fuzzy matching in the database to find all potentially available resource instances.

[0088] (2) Status and capability verification: For the retrieved resource instances, verify their current status (whether they are idle, intact, or under maintenance), actual capabilities (such as the water capacity of fire trucks, the power of generators, and the validity period of personnel qualification certificates), geographical location, and estimated arrival time (considering road conditions and transportation tools).

[0089] (3) Constraint check: Consider the power plant’s specific safety constraints, operating procedure constraints, environmental constraints, etc. For example, certain types of equipment are prohibited in certain areas, or certain operations require specific permits.

[0090] (4) Generate a candidate resource set: For each resource requirement, generate a set containing all eligible and available candidate resource instances, along with their relevant attributes (such as cost, efficiency, reliability and other evaluation indicators).

[0091] 3. Module 330: Multi-objective optimization decision module. This is the core of the algorithm model, responsible for selecting the optimal resource combination from the candidate resource set to form the final configuration scheme. Due to the complexity of emergency scenarios, it is often necessary to consider multiple, even conflicting, objectives simultaneously; therefore, a multi-objective optimization method is appropriate.

[0092] (1) Definition of Objective Function: Based on the power plant's emergency strategy and specific scenarios, define one or more optimization objective functions. Common objectives may include:

[0093] min(Tresponse): Minimize the overall emergency response time (the time from the occurrence of the event to the availability and commencement of critical resources). The specific calculation formula is as follows:

[0094]

[0095] in, Indicates the first The resource reached the first The time required to locate the task site and begin operations. A set of resource indexes for all assigned tasks.

[0096] max(Eeffect): Maximizes the effectiveness of emergency response (e.g., minimizing downtime, minimizing economic losses, maximizing the number of people rescued). The specific calculation formula is as follows:

[0097]

[0098] in, For the first The resource for the first The performance matching coefficient for each task is determined based on the matching degree between the historical disposal score of the resource and the current scenario.

[0099] min(Cresource): Minimize the cost of emergency resources consumed (including material consumption, equipment depreciation, and labor costs). The specific calculation formula is as follows:

[0100]

[0101] in, For the first The unit price or consumption cost of a resource.

[0102] min(Rrisk): Minimizes secondary risk or new risk introduced due to improper resource allocation. The specific formula for calculating the minimum secondary risk R is as follows:

[0103]

[0104] in, For the first The resource is executing the first The potential risk coefficient for each task.

[0105] max(Ffairness): In some cases, it may be necessary to consider the fairness of resource allocation (e.g., when multiple disaster sites need resources simultaneously), and maximize the fairness of resource allocation. The specific calculation formula is as follows:

[0106]

[0107] in, For the first The supply of each resource For the first Total demand for each task.

[0108] (2) Setting constraints:

[0109] The resource quantity constraint means that the number of resources for each scheduling method cannot exceed the number available in the database:

[0110]

[0111] Represents any resource Only one task can be assigned at a time.

[0112] Demand satisfaction constraints indicate that critical, high-priority resource requirements must be met:

[0113]

[0114] in, For a set of critical tasks, For the first The capability value of each resource. For the first The minimum requirement for each task;

[0115] Time window constraints mean that certain tasks must be completed within a specific time window:

[0116]

[0117] in, For the first The required processing time for each task This is the latest deadline required for the task.

[0118] Logical order constraints indicate that there are sequential dependencies between certain tasks:

[0119]

[0120] in, A set of pre- and post-tasks with dependencies. For prerequisite tasks The completion time, For post-task The start time;

[0121] Budget constraints mean that the total cost of emergency resource mobilization cannot exceed the preset budget.

[0122]

[0123] in, This is the upper limit of the resource cost budget for a single emergency response action.

[0124] (3) Algorithm selection and implementation:

[0125] If the objective and constraints are relatively simple, an exact algorithm such as an extension of linear programming (e.g., goal programming) can be used.

[0126] For more complex, nonlinear, and large-scale problems, heuristic or metaheuristic algorithms are typically used, such as:

[0127] NSGA-II (Non-dominated Sorting Genetic Algorithm II): A classic and efficient multi-objective genetic algorithm that can find a set of Pareto optimal solutions for decision-makers to choose from based on their preferences.

[0128] MOPSO (Multi-Objective Particle Swarm Optimization): Another multi-objective optimization algorithm based on swarm intelligence.

[0129] A multi-objective version of Simulated Annealing or Tabu Search.

[0130] Reinforcement learning (RL) can model the resource allocation problem as a Markov decision process (MDP), where the states are the current scenario and resource status, the actions are choosing a resource allocation, and the reward function is related to the objective function. Through training, the agent can learn a strategy to make optimal allocation decisions under different states.

[0131] (4) Scheme generation and sorting:

[0132] Multi-objective optimization algorithms typically produce a Pareto front solution set, which is a set of mutually independent solutions. The output of module 330 allows decision-makers to select one or a few recommended solutions based on preferences, or automatically based on preset preferences (e.g., assigning higher weight to response time). These solutions are then formatted into an executable contingency resource allocation scheme.

[0133] 4. Module 340: Dynamic Adjustment and Learning Module for Solutions (optional).

[0134] This module corresponds to step S270 in the method flow and gives the method the ability to continuously learn and improve.

[0135] (1) Real-time monitoring interface: Receives feedback information from the execution monitoring process, including changes in the on-site situation, actual arrival time of resources, task completion status, resource consumption, etc.

[0136] (2) Deviation analysis: Compare the actual implementation with the original plan and analyze the reasons for the deviation.

[0137] (3) Trigger replanning: When the deviation exceeds the preset threshold or new major changes occur, a signal is sent to module 310 or module 330 to request a re-evaluation of the scenario or optimization decision and to generate an adjusted configuration scheme.

[0138] (4) Case Study and Model Update: The complete process of each emergency event (initial scenario, configuration plan, execution process, handling effect, lessons learned) is stored as a case in the knowledge base. Using this case data, the prediction model in module 310, the optimization algorithm parameters or rule base in module 330 are iteratively updated and optimized through machine learning methods (such as supervised learning, unsupervised learning or offline training of reinforcement learning) to continuously improve decision-making capabilities. For example, it can be analyzed which types of resource combinations are more effective in specific scenarios, or which scheduling strategies lead to delays.

[0139] Through the collaborative work of the above modules, the emergency resource optimization allocation algorithm model 131 can achieve efficient, intelligent, dynamic and optimized allocation of power plant emergency resources.

[0140] Example 3: Application examples in typical emergency scenarios

[0141] To illustrate the working process and effects of the method of the present invention more specifically, two typical power plant emergency scenarios are listed below and described in conjunction with the above embodiments. These scenarios focus more on disasters such as flooding of the plant, dam emergencies (indirectly related, such as water conditions caused by extreme weather), power outages throughout the plant, earthquakes, floods, and extreme weather.

[0142] Table 1: Examples of Typical Emergency Scenarios and Their Key Input Parameters

[0143]

[0144] refer to Figure 3 The resource allocation process for SCEN-A (severe flooding of the factory due to heavy rain) is shown below:

[0145] 1. Scenario Analysis and Requirements Assessment (Module 310):

[0146] The "factory flooding" incident was identified as a "natural disaster (flood)" and classified as "severe." The main threats are extensive equipment damage leading to prolonged production stoppages, personnel safety, and the potential for the disaster to escalate.

[0147] Quantitative Requirements: Water rescue team (at least one, equipped with inflatable boats and life jackets), high-power submersible pumps (at least five units, with a total flow rate of no less than 2000 cubic meters per hour), sandbags (at least 2000), waterproof barriers (several), emergency lighting equipment, communication equipment (waterproof walkie-talkies, satellite phones), temporary emergency power supply (for drainage pumps and lighting), medical rescue team (on standby), equipment repair team (electrical and mechanical professionals, on standby). Priority: Rescuing trapped personnel > Controlling water ingress and organizing drainage > Protecting core equipment > Restoring plant functionality.

[0148] 2. Available Resource Query and Evaluation (Module 320):

[0149] Database feedback: The factory's emergency rescue team (8 people, including 2 certified divers, equipped with 1 rubber boat) is on duty and 10 minutes away from the site; the warehouse has 3 large submersible pumps (500 cubic meters / hour / unit) and 4 medium-sized submersible pumps (200 cubic meters / hour / unit); 5,000 sandbags are in stock; there are sufficient emergency lighting fixtures; and 2 diesel generators (200kW and 500kW respectively) are available.

[0150] 3. Optimize configuration and generate solutions (Module 330):

[0151] Algorithm model 131 takes "rescuing trapped personnel as quickly as possible and minimizing flood damage" as its core objective, and optimizes it by taking into account resource transportation and deployment time.

[0152] The generated preliminary scheme may look like this (see Table 2 for example format):

[0153] Table 2: Example of Emergency Resource Allocation Plan for SCEN-A (Factory Flooding)

[0154]

[0155] 1. Plan Evaluation and Application: Command personnel can review this plan and may make minor adjustments based on the actual situation on site (such as water flow direction, specific location of trapped personnel, etc.). Once confirmed, the dispatch instructions in the plan can be used for emergency response.

[0156] 2. Dynamic Adjustment: If the rainstorm continues and the upstream water flow exceeds expectations, causing the drainage speed to be unable to keep up with the water accumulation speed, or if new areas are flooded, after receiving feedback, module 340 will trigger replanning. It may be necessary to request external support (such as calling municipal drainage vehicles or coordinating more large water pumps) or adjust the sandbag containment strategy to prioritize the protection of higher-level equipment.

[0157] refer to Figure 4 As shown, the resource allocation process for SCEN-B (extreme weather causing a plant-wide power outage) is similar:

[0158] 1. Scenario Analysis and Needs Assessment: The core is to restore power to critical systems (such as DCS monitoring, emergency lighting, fire pumps, and communications) as soon as possible, maintain the basic safety of the power plant, and create conditions for black start or waiting for external power to be restored.

[0159] 2. Available Resources Inquiry: Inquire about the status and fuel reserves of emergency diesel generators, mobile emergency power vehicles, black start backup power supplies and equipment, communication equipment (especially those independent of mains power), lighting equipment, and inspection and operation personnel.

[0160] 3. Optimize configuration:

[0161] The primary task is to immediately start the emergency diesel generator and, according to the preset load sequence list, prioritize restoring power to first-level critical loads (such as DCS, SIS, emergency communications, emergency lighting, fire protection system power supply, battery charging power supply, etc.).

[0162] Personnel deployment: Organize operation personnel to check the status of generators and plant power systems, and confirm the grid disconnection status; organize inspection personnel to conduct special inspections of key equipment throughout the plant (especially outdoor equipment such as GIS and transformers) to check for any direct damage caused by extreme weather.

[0163] Black start preparation: If the conditions for black start are met and the external power supply is difficult to restore in a short time, then according to the black start plan, allocate black start power sources (such as gas turbines, dedicated diesel generators) and related operators, equipment, and materials to prepare for the execution of the black start procedure.

[0164] Communication support: Utilize independent communication methods such as satellite phones to maintain contact with the power grid dispatch center and higher-level units, report on the situation within the plant, and obtain information on the progress of external power grid restoration.

[0165] Material support: Ensure that the emergency generator has sufficient fuel and prepare necessary repair materials to deal with any possible equipment damage.

[0166] 4. Issuance and Dynamic Adjustment of the Plan:

[0167] (1) The scheduling instructions in the plan are issued to each professional team.

[0168] (2) Dynamic adjustment:

[0169] If an emergency diesel generator fails to start or runs out of fuel, a backup generator or mobile power vehicle must be immediately activated to support the critical load, and fuel must be replenished urgently.

[0170] If the inspection finds that the equipment is severely damaged due to extreme weather (such as the collapse of switch station equipment or damage to the main transformer), the recovery strategy must be adjusted immediately. It may not be possible to perform a black start immediately. Resources should be concentrated to deal with the damaged equipment, and a report should be made to the power grid dispatch center to request external support.

[0171] Based on the estimated time for the external power grid to be restored, a dynamic decision is made: whether to continue maintaining the current power supply status and prepare for a black start, or to prepare for power receiving operations.

[0172] If extreme weather continues and poses further threats to equipment in the plant (such as strong winds or persistent icing), it is necessary to strengthen inspections and protective measures.

[0173] These embodiments demonstrate that the method of the present invention can intelligently and efficiently allocate various emergency resources of a power plant according to specific, real-time emergency scenarios, especially complex scenarios caused by natural disasters or extreme weather, such as flooding of the plant or power outages, thereby significantly improving the overall effectiveness of emergency response.

[0174] The following table compares the main differences between the present invention and existing technologies in solving the problem of emergency resource allocation in power plants, and highlights the advantages of the present invention:

[0175] Table 3: Comparison of the present invention and existing technologies in terms of emergency resource allocation in power plants

[0176]

[0177] Example 4: This example provides an intelligent allocation device for power plant emergency resources based on emergency scenarios, including:

[0178] The emergency scenario information acquisition unit is used to acquire emergency scenario information including one or more of the following: accident type, accident location, scope of impact, and accident severity level.

[0179] An emergency resource optimization and allocation unit is used to perform calculations and analyses by calling an emergency resource optimization and allocation algorithm model based on the emergency scenario information and a preset emergency resource database that stores emergency resource categories, quantities, and statuses. The emergency resource optimization and allocation algorithm model is configured to perform multi-objective optimization calculations based on the emergency scenario information and the data in the emergency resource database to determine the resource combination that meets emergency needs.

[0180] The scheme generation unit is used to generate an emergency resource configuration scheme for the emergency scenario, including the specific type of emergency resources required, the required quantity, the allocation priority, and the suggested scheduling instructions.

[0181] Example 5: This example proposes a terminal device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. The memory stores the computer program capable of running on the processor. When the processor loads and executes the computer program, it adopts the intelligent configuration method for power plant emergency resources provided by this invention.

[0182] It should be noted that the terminal device can be a computer device such as a desktop computer, a laptop computer, or a cloud server, and the terminal device includes, but is not limited to, a processor and a memory. For example, the terminal device may also include input / output devices, network access devices, and buses.

[0183] Furthermore, the processor can be a central processing unit (CPU). Of course, depending on the actual use, other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), off-the-shelf programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. can also be used. The general-purpose processor can be a microprocessor or any conventional processor, etc., and this application does not limit it in this regard.

[0184] Furthermore, the memory can be an internal storage unit of the terminal device, such as the hard disk or RAM of the terminal device, or an external storage device of the terminal device, such as a plug-in hard disk, smart memory card (SMC), secure digital card (SD), or flash memory card (FC) equipped on the terminal device. In addition, the memory can also be a combination of the internal storage unit and the external storage device of the terminal device. The memory is used to store computer programs and other programs and data required by the terminal device. The memory can also be used to temporarily store data that has been output or will be output. This application does not limit this.

[0185] Furthermore, through this terminal device, any one of the intelligent configuration methods for power plant emergency resources in the above embodiments can be stored in the memory of the terminal device, and loaded and executed on the processor of the terminal device for convenient use.

[0186] Example 6: This example proposes an electronic system, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, characterized in that the instructions are executed by the at least one processor to enable the at least one processor to execute the intelligent configuration method for power plant emergency resources proposed in this invention.

[0187] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0188] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. field Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the present invention. Various changes and modifications can be made to the present invention without departing from the spirit and scope thereof, and all such changes and modifications fall within the scope of the present invention as claimed.

Claims

1. A method for intelligent allocation of emergency resources in power plants based on emergency scenarios, characterized in that, Includes the following steps: Acquire emergency scenario information, which includes at least one or more of the following: accident type, accident location, scope of impact, and accident severity level; Based on the emergency scenario information and according to a preset emergency resource database that stores emergency resource categories, quantities, and statuses, an emergency resource optimization and configuration algorithm model is invoked for calculation and analysis. The emergency resource optimization and configuration algorithm model is configured to perform multi-objective optimization calculations based on the emergency scenario information and the data in the emergency resource database to determine the resource combination that meets emergency needs. Generate an emergency resource configuration plan for the emergency scenario, including the specific types of emergency resources required, the required quantity, the allocation priority, and the suggested scheduling instructions.

2. The method according to claim 1, characterized in that, The emergency resource optimization allocation algorithm model includes: The scenario requirements analysis module is used to analyze the emergency scenario information and quantify the specific requirements for emergency resources; The resource matching and evaluation module is used to search for available resources in the emergency resource database based on quantified needs, and to evaluate their applicability and availability; The multi-objective optimization decision-making module is used to generate the emergency resource allocation scheme by comprehensively considering multiple objectives, including the shortest response time, the least resource consumption, or the best handling effect, while meeting the core emergency objectives.

3. The method according to claim 2, characterized in that, The multi-objective optimization decision-making module is used to select the optimal resource combination from the candidate resource set to form the final configuration scheme, specifically including: (1) Based on the power plant's emergency strategy and specific scenarios, define one or more optimization objective functions, including: Minimize overall emergency response time, maximize emergency response effectiveness, minimize the cost of emergency resources consumed, minimize secondary risks or new risks introduced due to improper resource allocation, and ensure the fairness of resource allocation; (2) Setting constraints, including: Resource quantity constraints mean that the number of resources for each scheduling method cannot exceed the number available in the database. Demand satisfaction constraints mean that critical, high-priority resource demands must be met. Time window constraints indicate that certain tasks must be completed within a specific time window; Logical order constraints indicate that there are sequential dependencies between certain tasks; Budget constraints mean that the total cost of emergency resource mobilization cannot exceed the preset budget; (3) An optimization strategy based on heuristic algorithms or machine learning is adopted, specifically: Optimization is performed using heuristic or metaheuristic algorithms, including: NSGA-II algorithm, MOPSO multi-objective optimization algorithm; or Optimization is achieved using reinforcement learning: the resource allocation problem is modeled as a Markov decision process (MDP), where the state is the current scenario and resource status, the action is to choose a resource allocation, and the reward function is related to the objective function. Through training, the agent learns a strategy to make the optimal allocation decision under different states.

4. The method according to claim 3, characterized in that, The specific optimization objective function is as follows: Let the set of task requirements in an emergency scenario be . Available emergency resources are: Define decision variables Indicates the first Is the resource allocated to the first...? One task: ， in Indicates allocation, This indicates no allocation; Objective 1 is to minimize the overall emergency response time. The specific calculation formula is as follows: ， in, Indicates the first The resource reached the first The time required to locate the task site and begin operations. A set of resource indexes for all assigned tasks; Objective 2 is to maximize the effectiveness of emergency response. The specific calculation formula is as follows: ， in, For the first The resource for the first The performance matching coefficient of each task is determined based on the matching degree between the historical disposal score of the resource and the current scenario; Objective 3 is to minimize emergency resource costs. The specific calculation formula is as follows: ， in, For the first The unit price or consumption cost of a resource; Objective four is to minimize the secondary risk R, and the specific calculation formula is as follows: ， in, For the first The resource is executing the first The potential risk factor for each task; Objective 5 is to maximize the fairness of resource allocation. The specific calculation formula is as follows: ， in, For the first The supply of each resource For the first Total demand for each task.

5. The method according to claim 3, characterized in that, The constraints are specifically set as follows: The resource quantity constraint means that the number of resources for each scheduling method cannot exceed the number available in the database: ， Represents any resource Only one task can be assigned at a time. Demand satisfaction constraints indicate that critical, high-priority resource requirements must be met: ， in, For a set of critical tasks, For the first The capability value of each resource. For the first The minimum requirement for each task; Time window constraints mean that certain tasks must be completed within a specific time window: ， in, For the first The required processing time for each task This is the latest deadline required for the task; Logical order constraints indicate that there are sequential dependencies between certain tasks: ， in, A set of pre- and post-tasks with dependencies. For prerequisite tasks The completion time, For post-task The start time; Budget constraints mean that the total cost of emergency resource mobilization cannot exceed the preset budget. ， in, This is the upper limit of the resource cost budget for a single emergency response action.

6. The method according to claim 1, characterized in that, The emergency scenario information also includes information on power equipment damage, out-of-service units, power grid load reduction or outage range. The emergency resource categories in the emergency resource database include at least emergency personnel, emergency equipment, emergency supplies, spare parts and external support forces.

7. The method according to claim 1, characterized in that, The method further includes: The system monitors real-time changes in the emergency response process and resource consumption information, and dynamically adjusts and updates the emergency resource allocation plan based on the monitored information.

8. A smart allocation device for emergency resources in power plants based on emergency scenarios, characterized in that, include: The emergency scenario information acquisition unit is used to acquire emergency scenario information including one or more of the following: accident type, accident location, scope of impact, and accident severity level. An emergency resource optimization and allocation unit is used to perform calculations and analyses by calling an emergency resource optimization and allocation algorithm model based on the emergency scenario information and a preset emergency resource database that stores emergency resource categories, quantities, and statuses. The emergency resource optimization and allocation algorithm model is configured to perform multi-objective optimization calculations based on the emergency scenario information and the data in the emergency resource database to determine the resource combination that meets emergency needs. The scheme generation unit is used to generate an emergency resource configuration scheme for the emergency scenario, including the specific type of emergency resources required, the required quantity, the allocation priority, and the suggested scheduling instructions. The emergency resource optimization and allocation unit includes: The scenario requirements analysis module is used to analyze the emergency scenario information and quantify the specific requirements for emergency resources; The resource matching and evaluation module is used to search for available resources in the emergency resource database based on quantified needs, and to evaluate their applicability and availability; The multi-objective optimization decision-making module is used to generate an emergency resource allocation plan by comprehensively considering multiple objectives, including the shortest response time, the least resource consumption, or the best handling effect, while meeting the core emergency objectives.

9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that, The memory stores a computer program that can run on a processor, and when the processor loads and executes the computer program, it employs the method described in any one of claims 1 to 7.

10. An electronic system comprising: At least one processor; And a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, characterized in that the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1-7.

Images