Flood control and disaster relief dynamic scene modeling method based on meta-model and relation aggregation

Abstract

The invention discloses a flood control and disaster relief dynamic scene modeling method based on meta-models and relation aggregation. The method comprises the following steps: constructing a flood control and disaster relief entity object meta-model, an event object meta-model and a flood control and disaster relief scene meta-model; constructing an entity object-event-scene intra-layer incidence relation model and an inter-layer relation aggregation model; and constructing an entity object-event-scene inter-node state transfer model, a node state evolution model and an inter-layer state aggregation model to realize node state deduction. According to the method, the meta-model is used as an abstract modeling means, the modeling universality and semantic expression consistency of objects of different granularity layers of the scene are enhanced, the flood control and disaster relief panorama is split with different granularities through the metadata-entity-event-scene network model, causal relevance among elements with different granularities can be found conveniently, and the scene flood control and disaster relief panorama is obtained. Therefore, the rapid calculation of the dynamic evolution of the scene state is realized, the development situation of the flood disaster is deduced in time, and a basis is provided for assisting a decision maker to make a rapid and accurate emergency response.

Description

Technical Field

[0001] The present invention relates to a dynamic scenario modeling method for flood control and disaster relief based on meta - model and relationship aggregation, belonging to the technical field of complex scenario modeling for flood control and disaster relief. Background Art

[0002] As an abstraction of models, the meta - model provides a structured method to describe and define the constituent elements, relationships, and their constraints of other models. It has a high degree of abstraction and can clearly express the syntax and semantics of modeling languages or methods. The meta - model supports the generality and reusability of models, making it easier to integrate and interoperate models in different fields and applications. By providing a consistent framework, the meta - model promotes the standardization of models and reduces the complexity in the modeling process. At the same time, the meta - model plays an important role in supporting model verification and consistency checking, ensuring the accuracy and reliability of the constructed models.

[0003] The relationship aggregation model can describe the relationships between objects at multiple levels, capture complex interactions and dependencies, and has good flexibility and scalability. It can be adjusted according to scenarios and requirements to adapt to changing environments. This model can reflect the dynamic changes in the relationships between objects, support real - time data updates and dynamic evolution analysis, and effectively identify and analyze the causal relationships between objects. By aggregating data from different sources, the relationship aggregation model realizes the integration and unification of information, providing a more comprehensive perspective. At the same time, the relationship aggregation model can handle intra - layer and inter - layer relationships, enhance the comprehensiveness of the model, and ensure the semantic consistency between different levels and objects, improving the understandability of the model. Therefore, this method has significant application value in the modeling of complex systems.

[0004] For the dynamic scenario modeling of flood control and disaster relief, the existing methods mainly include ontology, mathematical expressions, and knowledge elements. However, these methods all have some significant deficiencies. First, they fail to effectively consider the influence of factors at different levels in terms of modeling granularity, thus limiting the understanding of complex scenarios. Second, these methods are mostly static models and lack the effective capture of the dynamic changes of disasters, resulting in insufficient real - time response capabilities. At the same time, these methods mainly focus on specific regions or disaster types and also face the challenge of standardizing data from different sources, lacking generality and adaptability, and thus affecting the effectiveness and accuracy of the models. Therefore, the research on using the method of combining meta - model and relationship aggregation to construct the dynamic scenario of flood control and disaster relief has significant application value. Summary of the Invention

[0005] Objective of the Invention: Aiming at the problems existing in the existing methods, the present invention proposes a dynamic scenario modeling method for flood control and disaster relief based on meta-model and relationship aggregation. By constructing a four-layer network model of metadata-entity-event-scenario, the panoramic view of flood control and disaster relief is split into different granularities, facilitating the discovery of causal relationships between elements of different granularities, realizing the rapid calculation of the dynamic evolution of scenario states, timely inferring the development trend of flood disasters, and providing a basis for assisting decision-makers to make rapid and accurate emergency responses.

[0006] Technical Solution: To achieve the above objective of the invention, the present invention adopts the following technical solutions:

[0007] In the first aspect, the present invention provides a dynamic scenario modeling method for flood control and disaster relief based on meta-model and relationship aggregation, including the following steps:

[0008] (1) Construct a meta-model of flood control and disaster relief entity objects, an event object meta-model, and a flood control and disaster relief scenario meta-model; the event objects include disaster event objects and disaster relief event objects, the disaster event objects include the set of disaster-bearing bodies involved in the disaster event, the disaster-causing factors, and the state value representing the emergency level of the disaster event; the disaster relief event objects include the name, the set of disaster-bearing bodies assisted by the disaster relief event, the attribute set, and the rescue effect state value; the flood control and disaster relief scenario includes the disaster event set, the disaster relief event set, and the state value representing the emergency level of the flood control and disaster relief scenario;

[0009] (2) Construct an intra-layer association relationship model and an inter-layer relationship aggregation model of entity object-event-scenario; the intra-layer association relationship includes the relationship between entity objects, and the inter-layer relationship aggregation includes the relationship aggregation between entity and disaster event, the relationship aggregation between entity and disaster relief event, and the inter-layer relationship aggregation between event and scenario; the relationship aggregation between entity and disaster event includes determining the disaster event state value according to the vulnerability of the disaster-bearing body of the disaster event object and the danger of the disaster-causing factor; the relationship aggregation between entity and disaster relief event includes determining the rescue effect state value of the disaster relief event according to the vulnerability of the disaster-bearing body of the disaster relief event object and the rescue effect; the inter-layer relationship aggregation between event and scenario includes determining the scenario state value according to the disaster event state value and the rescue effect state value of the disaster relief event;

[0010] (3) Construct a state transfer model between nodes, a node state evolution model, and an inter-layer state aggregation model of entity object-event-scenario to realize the deduction of node states.

[0011] Further, in step (1), it also includes constructing a metadata meta-model; the metadata includes descriptive metadata, management metadata, and usability metadata obtained by functionally partitioning flood control and disaster relief scenario information; the descriptive metadata includes attribute descriptions, spatial locations, and corresponding permission information of flood control and disaster relief materials and engineering entities; the management metadata includes information providing the entire life cycle of an object; the usability metadata includes query behavior record information.

[0012] Further, the flood control and disaster relief entity object meta-model in step (1) includes a flood-affected area information meta-model, a relocation area information meta-model, a material storage point information meta-model, a rescue agency information meta-model, a rescue and relief material information meta-model, and an emergency measure information meta-model;

[0013] The flood-affected area information meta-model includes a flood-affected area code, name, area, location, number of people, and number of people to be relocated / rescued;

[0014] The relocation area information meta-model includes a relocation area code, name, area, location, grade, capacity, and number of people already resettled;

[0015] The material storage point information meta-model includes a material storage point code, name, location, grade, and category;

[0016] The rescue agency information meta-model includes a rescue agency code, name, location, and category;

[0017] The rescue and relief material information meta-model includes a material storage number, material number, material storage warehouse number where the material is located, quantity of the material, storage unit of the material, and type of the material;

[0018] The emergency measure information meta-model includes an emergency measure number, rescue team number, affiliated agency number of the team, rescue destination, number of people participating in the rescue, type of emergency measure, and completion status of the emergency measure.

[0019] Further, the disaster event object meta-model in step (1) is constructed as a triple E = (O, En, F), where O represents the set of disaster-bearing bodies involved in the current disaster event O = {o 1 , o 2 , …, o k}, k represents the number of disaster-bearing bodies involved in the disaster event object, En is the disaster-causing factor of the current disaster event, and F represents the current disaster event status value, which is used to represent the emergency level of the current disaster event;

[0020] The disaster-bearing body meta-model is a five-tuple O = (Nm o , At o , Vul, Hl, I), where Nm o is the name of the disaster-bearing body, Ato is the set of disaster-bearing body attributes, At o = {a o1 , a o2 , …, a ot}, where t represents the current number of disaster-bearing body attributes, Vul is the set of vulnerability scores of disaster-bearing body attributes, Vul = {v 1 , v 2 , …, v t}, v i represents the vulnerability score of a i , Hl is the rescue effect value of the disaster-bearing body, and I represents the comprehensive vulnerability score of the current disaster-bearing body;

[0021] The hazard factor meta-model is a quadruple En = (Nm z , At z , Da, H), where Nm z is the name of the hazard factor, At z is the set of hazard factor attributes At z = {a z1 , a z2 , …, a zl}, l represents the current number of hazard factor attributes, Da is the set of hazard values of hazard factor attributes, Da = {d 1 , d 2 , …, d l}, d i represents the hazard score, and H represents the comprehensive hazard score of the hazard factor of the current disaster event;

[0022] The disaster relief event object meta-model is constructed as a quadruple HE = (Nm j , O, At j , X), where Nm j represents the name of the disaster relief event, O represents the set of disaster-bearing bodies rescued by the current disaster relief event, O = {o 1 , o 2 , …, o m}, m represents the number of disaster-bearing body objects involved in the current disaster relief event, At j is the set of disaster relief event attributes At j = {a j1 , a j2 , …, a jh}, h represents the number of attributes of the current disaster relief event, and X represents the rescue effect status value.

[0023] Furthermore, the flood control and disaster relief scenario meta-model in step (1) is constructed as a triple S = (E, HE, SI), where E represents the set of current disaster events, and E = {e i ∣i = 1, …, |E|}, e iDenote the \(i\)-th disaster event in the current scenario, and \(HE\) represents the set of disaster relief events in the current scenario, with \(HE = \{he i | i = 1, …, |HE| \}\), and \(SI\) represents the scenario state value of the current scenario, which is used to represent the emergency level of the flood control and disaster relief scenario.

[0024] Furthermore, the formula for the disaster event state value \(F\) in step (2) is as follows:

[0025]

[0026] where \(H\) represents the comprehensive risk score of the disaster-causing factors of the disaster event, d i is the risk score corresponding to the \(i\)-th attribute value of the disaster-causing factors of the current event, \(l\) represents the number of attributes of the disaster-causing factors, \(I i represents the vulnerability score of the \(i\)-th disaster-bearing body, \(k\) represents the number of disaster-bearing bodies affected by the disaster event; the vulnerability score of a single disaster-bearing body v i represents the vulnerability score corresponding to the \(i\)-th attribute value of the disaster-bearing body, and \(t\) represents the number of attributes of the current disaster-bearing body;

[0027] The formula for the rescue effect state value \(X\) of the disaster relief event in step (2) is as follows:

[0028]

[0029] where \(h i represents the rescue effect value of the \(i\)-th disaster-bearing body, and \(m\) represents the number of disaster-bearing bodies rescued by the disaster relief event;

[0030] The scenario state value in step (2) is obtained by weighted summation of the disaster event state value and the rescue effect state value of the disaster relief event in the current scenario.

[0031] Furthermore, the node state transfer model, node state evolution model, and inter-layer state aggregation model of entity - event - scenario in step (3) include the following steps:

[0032] (1) Define entity objects, events, and scenarios as nodes at different levels in the network model; define the entity node state partition set \(HO=\{ho 1 , ho 2 , …, ho ns \}, \(ns\) is the number of entity node states, and the entity node state vector is where represents the probability value that the entity node state is \(ho i , define the event node state partition set \(He = \{he 1 , he 2 , …, he ks}, where \(k_s\) is the number of event node states, and the event node state vector is indicating that the event node state is \(he\) i the probability value. Define the scenario node state partition set \(H_s=\{h_s\) 1 , \(h_s\) 2 , …, \(h_s\) ms}\), where \(m_s\) is the number of scenario node states, and the scenario node state vector is indicating that the scenario node state is \(h_s\) i the probability value. The sum of the components of each type of node state vector is 1;

[0033] (2) Construct the state transfer model between nodes, and construct the state transfer model between nodes as a quadruple \(TRO=(O_{nm}, Node, A, Num)\), where \(O_{nm}\) represents the name of the state transfer operation, \(Mode\) represents the set of nodes involved in the node state transition, \(A\) represents the attribute of the node state transfer, and \(Num\) represents the set of numerical values of the node state transfer;

[0034] (3) Construct the node state evolution model. For each node, give the initial state probability distribution and a time-varying state transition matrix to obtain the node state evolution model. The calculation formula is as follows:

[0035] \(v\) ′ \(= v\times P\) t

[0036] \(v = [p\) 1 , \(p\) 2 , …, \(p\) q-1 , \(p\) q \)

[0037]

[0038] where \(v\) is the node state vector at the current moment, representing the probability distribution of being in each state at the current moment, and can be any type of node among entity, event, and scenario nodes; \(P\) t is the state transition matrix at the current \(t\) moment, \(v\) ′ is the evolution state vector at the next moment, \(q\) represents the number of states of the current attribute node, and \(P\) t in represents the probability value that the attribute node changes from state \(i\) to state \(j\) at the current \(t\) moment, and there is

[0039] (4) Construct the inter-layer node state aggregation model. First, construct the node state aggregation function between entity nodes and event nodes. The input of this function is the set of entity nodes involved in the event, and the output is the event aggregation state \(u\). The aggregation calculation formula is as follows:

[0040]

[0041] where e ′ is the aggregated event state vector, W i represents the transition probability matrix of the different states of the i-th entity object affecting event e, and p m,q represents the probability that the entity object in state m affects event state q. The Softmax function normalizes the aggregation result. z i is the i-th component, n is the number of entity nodes. Then, an aggregation calculation function between the event node and the scenario node state is constructed. The input of this function is the set of event nodes involved in the scenario, and the output is the scenario aggregation state. The aggregation calculation formula is as follows:

[0042]

[0043]

[0044] where s ′ is the aggregated scenario state vector, T i represents the transition probability matrix of the different states of the i-th event object affecting scenario s, and p i,j represents the probability that the event object in state i affects scenario state j. The Softmax function normalizes the aggregation result. Y i is the i-th component, and d is the number of event nodes.

[0045] Second, the present invention provides a flood control and disaster relief dynamic scenario modeling system based on a meta-model and relationship aggregation, including:

[0046] A meta-model construction module for constructing a flood control and disaster relief entity object meta-model, an event object meta-model, and a flood control and disaster relief scenario meta-model; the event object includes a disaster event object and a disaster relief event object. The disaster event object includes a set of disaster-bearing bodies involved in the disaster event, a disaster-causing factor, and a state value indicating the emergency level of the disaster event; the disaster relief event object includes a name, a set of disaster-bearing bodies assisted by the disaster relief event, an attribute set, and a rescue effect state value; the flood control and disaster relief scenario includes a disaster event set, a disaster relief event set, and a state value indicating the emergency level of the flood control and disaster relief scenario;

[0047] A model aggregation module, configured to construct an intra-layer association relationship model and an inter-layer relationship aggregation model for entity object - event - scenario; the intra-layer association relationship includes the relationship between entity objects, and the inter-layer relationship aggregation includes the aggregation of the relationship between an entity and a disaster event, the aggregation of the relationship between an entity and a disaster relief event, and the inter-layer relationship aggregation between an event and a scenario; the aggregation of the relationship between an entity and a disaster event includes determining a disaster event status value according to the vulnerability of the disaster-bearing body and the danger of the disaster-causing factor of the disaster event object; the aggregation of the relationship between an entity and a disaster relief event includes determining a disaster relief event rescue effect status value according to the vulnerability of the disaster-bearing body and the rescue effect of the disaster relief event object; the inter-layer relationship aggregation between an event and a scenario includes determining a scenario status value according to the disaster event status value and the disaster relief event rescue effect status value;

[0048] And a deduction module, configured to construct an inter-node state transfer model, a node state evolution model, and an inter-layer state aggregation model for entity object - event - scenario, and implement node state deduction.

[0049] In a third aspect, the present invention provides a computer system, including a memory, a processor, and a computer program / instruction stored on the memory and executable on the processor. When the computer program / instruction is executed by the processor, the steps of a flood control and disaster relief dynamic scenario modeling method based on a meta-model and relationship aggregation are implemented.

[0050] In a fourth aspect, the present invention provides a computer program product, including a computer program / instruction. When the computer program / instruction is executed by the processor, the steps of a flood control and disaster relief dynamic scenario modeling method based on a meta-model and relationship aggregation are implemented.

[0051] Beneficial effects: The flood control and disaster relief dynamic scenario modeling method based on a meta-model and relationship aggregation proposed by the present invention improves the generality and compatibility of the flood control and disaster relief dynamic scenario model by using the meta-model method, facilitating the integration of data from different sources; uses the relationship aggregation method to solve the defects of single modeling granularity, unclear causal relationship, and insufficient understanding of complex scenarios in traditional modeling; the present invention realizes the rapid calculation of the dynamic evolution of the flood control and disaster relief scenario state, timely deduces the development trend of flood disasters, and provides a basis for assisting decision-makers to make rapid and accurate emergency responses. Description of the Drawings

[0052] Figure 1 is a schematic diagram of the flood control and disaster relief dynamic scenario modeling process based on a meta-model and relationship aggregation of the present invention.

[0053] Figure 2 is a schematic diagram of the meta-data meta-model structure of an embodiment of the present invention.

[0054] Figure 3 is a schematic diagram of the descriptive meta-data meta-model structure of an embodiment of the present invention

[0055] Figure 4 It is a schematic diagram of the management metadata meta-model structure of the embodiment of the present invention.

[0056] Figure 5 It is a schematic diagram of the usability metadata meta-model structure of the embodiment of the present invention.

[0057] Figure 6 It is a schematic diagram of the entity object - event - scenario inter-layer state aggregation model of the embodiment of the present invention. Detailed implementation manners

[0058] The present invention will be further clarified below in conjunction with the accompanying drawings and specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and not to limit the scope of the present invention. After reading the present invention, those skilled in the art's various equivalent modifications of the present invention all fall within the scope defined by the appended claims of this application.

[0059] In combination with Figure 1 Describing the technical details of the present invention, a flood control and disaster relief dynamic scenario modeling method based on meta-model and relationship aggregation disclosed in the embodiment of the present invention proposes a meta-network model for flood control and disaster relief scenarios, including a metadata meta-model layer, a flood control and disaster relief entity object meta-model layer, an event object meta-model layer, and a flood control and disaster relief scenario meta-model layer. The metadata meta-model, the flood control and disaster relief entity object meta-model, the event object meta-model, and the flood control and disaster relief scenario meta-model are respectively constructed, as well as the intra-layer association relationship model and the inter-layer relationship aggregation model of entity object - event - scenario, and the node - to - node state transfer model, the node state evolution model, and the inter-layer state aggregation model of entity object - event - scenario. The main steps are described in detail below.

[0060] Step 1: Construct a flood control and disaster relief entity object meta-model, an event object meta-model, and a flood control and disaster relief scenario meta-model.

[0061] In the flood control and disaster relief scenario of the present invention, the event objects mainly include disaster event objects and disaster relief event objects. Among them, the disaster event objects include the set of disaster - bearing bodies involved in the disaster event, the disaster - causing factors, and the status value indicating the emergency level of the disaster event; the disaster relief event objects include the name, the set of disaster - bearing bodies assisted by the disaster relief event, the attribute set, and the rescue effect status value; the flood control and disaster relief scenario includes the set of disaster events, the set of disaster relief events, and the status value indicating the emergency level of the flood control and disaster relief scenario.

[0062] Specifically, in this embodiment, first, a flood control and disaster relief metadata meta-model is constructed. According to the scenario information, it is functionally divided and modeled separately for descriptive metadata, management metadata, and usability metadata. The specific modeling process is as follows:

[0063] (1) Functionally divide the flood control and disaster relief scenario information into descriptive metadata, management metadata, and usability metadata. Attribute descriptions, spatial locations, and corresponding permission information of entities such as flood control and disaster relief materials and projects are classified into descriptive metadata. Information providing the entire life cycle of the object is classified into management metadata. Query behavior record information is classified into usability metadata. The specific metadata meta-model structure diagram is as shown in the appendix Figure 2 as follows;

[0064] (2) The descriptive metadata meta-model is constructed as a triple DS(Pr, Pe, Sp), which consists of attribute metadata Pr, permission metadata Pe, and spatial metadata Sp. Among them, the attribute metadata Pr is modeled as a four-dimensional array Pr(Id, Na, Ow, Ot), where Id represents the identification code of the entity or resource, Na represents the name of the entity or information, Ow represents the creator, and Ot represents the creation time. The permission metadata is modeled as a binary array Pe(Pu, Po), which consists of user management permission metadata Pu and resource accessibility permission metadata Po. Among them, the user management permission metadata is further modeled as Pu(Pucr, Puup, Puac, Purm), where Pucr represents the permission information to create, Pucr represents the permission information to modify, Puac represents the permission information to access the resource, and Purm represents the permission information to delete. The resource accessibility permission metadata is further modeled as a triple Po(Por, Pow, Poo), which consists of readable permission information Por, writable permission information Pow, and executable operation permission information Poo. The spatial metadata is modeled as a binary array Sp(Spad, Sprf), where Spad represents the location information of the entity or resource, and Sprf represents the spatial reference system information. The UML diagram of the descriptive metadata meta-model is as shown in the appendix Figure 3 as follows;

[0065] (3) The management metadata meta-model is constructed as a triple MA(St, Tr, Sa), which consists of structural metadata St, traceability metadata Tr, and preservation metadata Sa. Among them, the structural metadata is modeled as a binary array St(Sth, Stb), including predecessor information Sth and successor information Stb. The traceability metadata is represented by a binary array Tr(Ops, Opstm), including the operation record Ops of the life cycle state change of the resource or entity and the operation time Opstm. The preservation metadata is modeled as a binary array SV(Pertm, Dis), which contains the information of the persistent time of the resource or entity information Pertm (the time unit is hours, where the special value -1 means that the resource or entity can be stored in the data repository for a long time), and the presentable information Dis (0 means not displayable, 1 means displayable) is used to represent the logical deletion function of the resource. The UML diagram of the management metadata meta-model is as shown in the appendix Figure 4 as follows;

[0066] (4) The usability metadata meta-model is constructed as a binary tuple US(OP, AG). The behavior record metadata OP contains operation information of various resources or entities. Such operations include the transfer in or out of materials, the discharge of water from the reservoir to the downstream, the transfer of personnel, etc.; the aggregated information metadata AG contains statistical information on these behavior records, and the aggregation functions included are the average value, the maximum value, the minimum value, and other statistical values. The UML diagram of the usability metadata meta-model is as shown in the appendix Figure 5 as follows.

[0067] Then, various objects in the flood control and disaster relief scenarios are hierarchically constructed according to their membership granularity. Based on the meta-models at each level, objects at different levels are constructed. The specific construction process is as follows:

[0068] (1) Construction of the meta-model for flood control and disaster relief entity objects, specifically including: ① Construction of the information meta-model for flood-stricken areas as struck(id, name, size, pos, pda, trspda), where id represents the flood-stricken area code, name represents the name of the stricken area, size represents the area of the flood-stricken area in square kilometers, pos represents the location of the stricken area, pda represents the number of people in the stricken area, and trspda represents the number of people to be transferred / resued; ② Construction of the information meta-model for relocation areas as relocation(id, name, size, pos, grade, capacity, rpda), where id represents the relocation area code, name represents the name of the relocation area, size represents the area of the relocation area in square kilometers, pos represents the location of the relocation area, grade represents the level of the relocation area, divided into three levels: county, city, and province, capacity represents the number of people that the relocation area can accommodate, and rpda represents the number of people already resettled; ③ Construction of the information meta-model for material reserve points as reserve(id, name, pos, grade, type), where id represents the material reserve point code, name represents the name of the material reserve point, pos represents the location of the material reserve point, grade represents the level of the material reserve point, divided into three levels: county, city, and province, and type represents the category of the material reserve point, divided into self-reserve points and agreement reserve points; ④ Information meta-model for rescue agencies as agency(id, name, type, pos), where id represents the rescue agency code, name represents the name of the rescue agency, pos represents the location of the rescue agency, and type represents the category of the rescue agency, divided into fire departments, military, engineering emergency rescue teams, government departments, and others; ⑤ Construction of the information meta-model for emergency rescue materials as material(mrid, material_id, reserve_id, num, unit, type), where mrid is the material reserve number, material_id is the material number, reserve_id is the number of the storage depot where the material is located, num is the quantity of the material, nuit is the storage unit of the material, and type is the type of the material, divided into necessities of life, emergency materials, emergency equipment, life-saving equipment, and others; ⑥ Construction of the information meta-model for emergency measures as EM(id, team_id, agency_id, destination, pda, type, done), where id is the emergency measure number, team_id is the rescue team number, agency_id is the number of the agency to which the team belongs, destination is the rescue destination, pda is the number of people participating in the rescue, type is the category of the emergency measure, divided into material dispatching, engineering emergency rescue, personnel transfer, personnel rescue, and others, and done represents the completion status of the emergency measure.

[0069] (2) Construction of the event object meta-model, specifically including: ① Dividing event objects into two categories: disaster event objects and disaster relief event objects; ② Constructing the disaster event object meta-model as a triple E = (O, En, F), where O represents the set of disaster-bearing bodies involved in the current disaster event O = {o 1 , o 2 , …, o k}, k represents the number of disaster-bearing bodies involved in the disaster event object, En is the disaster-causing factor of the current disaster event, and F represents the status value of the current disaster event, used to indicate the emergency level of the current disaster event; ③ Constructing the disaster-bearing body meta-model as a five-tuple O = (Nm o , At o , Vul, Hl, I), where Nm is the name of the disaster-bearing body, and the specific information of the disaster-bearing body is the disaster-affected object constructed by the flood disaster area information meta-model. At o is the set of disaster-bearing body attributes, At o = {a o1 , a o2 , …, a ot}, t represents the number of current disaster-bearing body attributes, Vul is the set of vulnerability scores of disaster-bearing body attributes Vul = {v 1 , v 2 , …, v t}, v i represents the vulnerability score of a i , and the vulnerability scores are 9, 6, 3, 1 respectively. 1 represents robust, 3 represents generally vulnerable, 6 represents moderately vulnerable, and 9 represents extremely vulnerable. Hl is the disaster relief effect value of the disaster-bearing body, and the disaster relief effect values are 5, 3, 1 respectively. 1 represents poor disaster relief effect, 3 represents general disaster relief effect, 5 represents excellent disaster relief effect, and I represents the comprehensive vulnerability score of the current disaster-bearing body; ④ Constructing the disaster-causing factor meta-model as a four-tuple En = (Nm z , At z , Da, H), where Nm z is the name of the disaster-causing factor, At z is the set of disaster-causing factor attributes At z = {a z1 , a z2 , …, a zl}, l represents the number of current disaster-causing factor attributes, Da is the set of dangerousness values of disaster-causing factor attributes Da = {d 1 , d 2 , …, d l}, d i represents the dangerousness score, and the dangerousness scores are 9, 6, 3, 1 respectively. 1 represents almost no danger, 3 represents general danger, 6 represents medium danger, and 9 represents serious danger. H represents the comprehensive dangerousness score of the disaster-causing factor of the current disaster event; ⑤ Constructing the disaster relief event object meta-model as a four-tuple HE = (Nmj , O, At j , X), where Nm j represents the name of the disaster relief event, O represents the set of disaster-bearing bodies assisted by the current disaster relief event, O = {o 1 , o 2 , …, o m}; m represents the number of disaster-bearing body objects involved in the current disaster relief event; At j is the set of disaster relief event attributes, At j = {a j1 , a j2 , …, a jh}; h represents the number of attributes of the current disaster relief event. Specific disaster relief event attributes include personnel transfer information, material reserve and scheduling information, etc., which are respectively composed of objects constructed by the information meta-model of the resettlement area, the information meta-model of the material reserve point, the information meta-model of the emergency rescue materials, the information meta-model of the rescue agency, and the information meta-model of the emergency measures. X represents the rescue effect status value.

[0070] (3) Construction of the flood control and disaster relief scenario meta-model, specifically including: constructing the disaster relief scenario meta-model as a triple S = (E, HE, SI), where E represents the set of current disaster events, E = {e i ∣i = 1, …, |E|}, e i represents the i-th disaster event in the current scenario; HE represents the set of disaster relief events in the current scenario, HE = {he i ∣i = 1, …, |HE|}; SI represents the scenario status value of the current scenario, which is used to represent the emergency degree of the flood control and disaster relief scenario.

[0071] In the flood control and disaster relief scenario of this embodiment, there are a total of 3 disaster events and 3 disaster relief events. The environmental information at time t in this flood control and disaster relief scenario is constructed according to the disaster-causing factor model En = (Nm z , At z , Da, H). At z = {24h_precipitation, water_level, river_flow}, where 24h - precipitaiton represents the 24-hour cumulative precipitation, water_level represents the river and lake water level information, and river_flow represents the river flow information. The units are millimeters, meters, and cubic meters per second respectively;

[0072] Construct the disaster-causing factors of the 3 disaster events respectively:

[0073] En 1 = (Nm_e 1 , At 1 , Da 1 , H1 ),At z1 ={215,486.3,267}

[0074] En 2 =(Nm_e 2 ,At 2 ,Da 2 ,H 2 ),At z2 ={193,233.3,352}

[0075] En 3 =(Nm_e 3 ,At 3 ,Da 3 ,H 3 ),At z3 ={126,59.4,536}

[0076] According to the above hazard factor attribute values and domain knowledge, construct the risk score sets of the hazard factors for each disaster event, where Da 1 ={9,6,3}, Da 2 ={9,3,3}, Da 3 ={6,1,6};

[0077] According to the disaster event object meta-model E=(O, En, F), construct the disaster event object at time t for flood control and disaster relief scenarios, where e 1 is the event of people being trapped in residential areas, e 2 are all events of people being trapped in railway areas, e 3 is the risk event of water conservancy projects, where O 1 ={o 11 ,o 12 ,o 13 ,o 14 ,o 15}, e 1 contains a total of 5 disaster-bearing bodies, O 2 ={o 21 ,o 22 ,o 23}, e 2 contains a total of 3 disaster-bearing bodies, O 3 ={o 31 ,o 32}, e 2 contains a total of 2 disaster-bearing bodies, and the disaster event is modeled as:

[0078] e 1 =(O 1 ,En 1 ,F 1 )

[0079] e2 =(O 2 , En 2 , F 2 )

[0080] e 3 =(O 3 , En 3 , F 3 )

[0081] According to the disaster relief event object meta - model HE = (Nm j , O, At j , X), three disaster relief events are respectively modeled as:

[0082] he 1 =(Nm 1 , O 1 , At 1 , X 1 ), At j1 ={hpda, trpda}

[0083] he 2 =(Nm 2 , O 2 , At 2 , X 2 ), At j2 ={hpda, trpda, road_state}

[0084] he 3 =(Nm 3 , O 3 , At 3 , X 3 ), At j3 ={hpda, hmaterial, trpda}

[0085] Among them, hpda represents the number of people participating in disaster relief, trpda represents the number of people transferred, road_state represents the railway restoration status, 0 means not restored, 1 means restored, hmaterial represents the adequacy index of engineering emergency rescue materials, 0 means insufficient materials, and 1 means sufficient materials;

[0086] According to the disaster - bearing body meta - model O = (Nm o , At o , Vul, Hl, I), different types of disaster - bearing body models are constructed. For the disaster - bearing bodies in O 1 , their At o1={pos, spda, size, communication, traffic_condition}, where pos represents the geographical coordinates of the disaster-affected entity, spda represents the number of trapped people, size represents the area of the disaster-affected entity region, with the unit being square kilometers, communication represents the communication situation, where 0 indicates that the area is completely out of contact, 1 indicates that the area is partially out of contact, 2 indicates that the area has normal communication, and traffic_condition represents the traffic situation in the area, where 0 indicates that the area is completely cut off from the outside world in terms of transportation channels, 1 indicates that the area is partially cut off from the outside world in terms of transportation channels, and 2 indicates that the transportation channels are normal; for the disaster-affected entity in O 2 Among them, its At o2 ={pos, spda, communication, strlength}, where pos represents the geographical coordinates of the disaster-affected entity, spda represents the number of trapped people, communication represents the communication situation, where 0 indicates that the area is completely out of contact, 1 indicates that the area is partially out of contact, 2 indicates that the area has normal communication, and strlength represents the damaged length of the railway, with the unit being meters; for the disaster-affected entity in O 3 Among them, its At o3 ={pos, engineering_scale, pda_affect, engineering_state}, where pos represents the location of the project, engineering_scale represents the scale of the project, 1 represents small-scale, 2 represents medium-scale, 3 represents large-scale, pda_affect represents the number of people that may be affected, and engineering_state represents the state of the project, where 1 indicates safe, 2 indicates low risk, 3 indicates medium risk, and 4 indicates high risk; according to the above various types of disaster-affected entities, a model is constructed to complete the attribute modeling of the 3 types of disaster-affected entities in this embodiment:

[0087] a 11 ={pos 11 , 531, 11.3, 2, 2}

[0088] a 12 ={pos 12 , 1432, 8.3, 1, 2}

[0089] a 13 ={pos 13 , 563, 4.6, 0, 0}

[0090] a 14 ={pos 14 , 1368, 9.6, 0, 1}

[0091] a 15 ={pos 15 , 2651, 15.8, 2, 2}

[0092] a 21 ={pos 21 ,536,0,22}

[0093] a 22 ={pos 22 ,963,2,6}

[0094] a 23 ={pos 23 ,635,2,15}

[0095] a 31 ={pos 31 ,1,2500,4}

[0096] a 32 ={pos 32 ,2,6000,2}

[0097] According to the above values of disaster-bearing body attributes and relevant domain knowledge, construct the vulnerability score sets and rescue effect value sets for each disaster-bearing body, where v 11 ={3,9,1,1}, v 12 ={6,6,6,1}, v 13 ={3,3,9,9}, v 14 ={6,6,9,6}, v 15 ={9,9,1,1}, v 21 ={6,9,9}, v 22 ={9,9,3}, v 23 ={6,1,6}, v 31 ={3,6,9}, v 32 ={6,9,6}; h 11 =5, h 12 =3, h 13 =1, h 14 =3, h 15 =5, h 21 =1, h 22 =5, h 23 =3, h 31 =3, h 32 =5.

[0098] According to the flood control and disaster relief scenario meta-model, the current flood control and disaster relief scenario is modeled as: S=(E, HE, SI), where E={e 1 , e 2 , e 3}, He={he 1 , he 2 , he 3}.

[0099] Step 2: Construct the intra-layer association relationship model and the inter-layer relationship aggregation model of entity object - event - scenario, and calculate the state values of each node. The specific construction process is as follows:

[0100] (1) Construct the intra-layer relationship model of entity objects, and model the relationship between entity objects as a triple R = (Ob, Hs, RT), where Ob is the subject of the relationship between entity objects, Hs is the object of the relationship between entity objects, and Rt is the relationship between entities. RT is modeled as a triple RT = (C, Rs, T), C is the category of the relationship between entities, Rs is the quantity of consumed resources, and T is the duration of the relationship;

[0101] (2) Construct the relationship aggregation model between entities and disaster events. According to the disaster-affected objects when the disaster event occurs, construct the set of disaster-bearing bodies O = {o 1 , o 2 , …, o k} of the disaster event object. According to the vulnerability score calculation model of a single disaster-bearing body:

[0102]

[0103] Calculate the vulnerability scores of each disaster-bearing body, where I represents the comprehensive vulnerability score of the disaster-bearing body, v i represents the vulnerability score corresponding to the i-th attribute value of the disaster-bearing body, t represents the number of attributes of the current disaster-bearing body, and finally the vulnerability scores of each disaster-bearing body are obtained as follows: I 11 = 2.27951, I 12 = 3.83366, I 13 = 5.19615, I 14 = 6.64009, I 15 = 3, I 21 = 7.86222, I 22 = 6.24025, I 23 = 3.30193, I 31 = 5.45136, I 32 = 6.86829;

[0104] According to the calculation model of the comprehensive risk score of disaster-causing factors:

[0105]

[0106] Calculate the comprehensive risk scores of the disaster-causing factors of each disaster event. H represents the comprehensive risk score of the disaster-causing factors of the disaster event, d i is the risk score corresponding to the i-th attribute value of the disaster-causing factor of the current event, l represents the number of attributes of the disaster-causing factor, and finally the comprehensive risk scores of the disaster-causing factors of each disaster event are calculated as follows: H 1 = 5.45136, H2 = 4.32674, H 3 = 3.30193; Given the vulnerability scores of each event's disaster-affected bodies and the comprehensive risk scores of disaster-causing factors, according to the disaster event status value calculation model:

[0107]

[0108] Calculate the status values of each disaster event, where H represents the comprehensive risk of disaster-causing factors, I i represents the vulnerability score of the i-th disaster-affected body, k represents the number of disaster-affected bodies affected by the disaster event, and finally the status values of each event are aggregated as follows: F 1 = 4.77918, F 2 = 5.01013, F 3 = 4.50991.

[0109] (3) Construct an aggregation model for the relationship between entities and disaster relief events. According to the information of the disaster-affected objects assisted during the rescue event, construct a set of disaster-affected bodies O = {o 1 , o 2 , …, o m} of the disaster relief event object and a set of rescue effect values Hl = {h 1 , h 2 , …, h m} for each disaster-affected body. Combine with the vulnerability scores of the disaster-affected bodies to construct a calculation model for the rescue effect status value of the disaster relief event, and the formula is as follows:

[0110]

[0111] X represents the rescue effect status value of the rescue event, I i represents the vulnerability score of o i , h i represents the rescue effect value of o i , m represents the number of disaster-affected bodies rescued by the disaster relief event. Calculate the rescue effect status values of each disaster relief event according to the above calculation model for the rescue effect status value of the disaster relief event, and finally the rescue effect status values of the three disaster relief events are aggregated as X 1 = 0.95311, X 2 = 0.65481, X 3 = 0.63915.

[0112] (4) Construct an aggregation model for the relationship between events and the scenario layer. The scenario status value is obtained by weighted summation of the disaster event status value and the rescue effect status value of the disaster relief event in the current scenario, and the formula is as follows:

[0113]

[0114] SI represents the scenario state value, F i represents the state value of the i-th disaster event in the flood control and disaster relief scenario, λ i represents F i corresponding weight coefficient, m F represents the number of disaster events in the flood control and disaster relief scenario, X j represents the rescue effect state value of the j-th disaster relief event in the flood control and disaster relief scenario, μ j represents X j corresponding weight coefficient, n X represents the number of disaster relief events in the flood control and disaster relief scenario; determine the relevant weight coefficients λ 1 = 0.2, λ 2 = 0.6, λ 3 = 0.2, μ 1 = 0.25, μ 2 = 0.25, μ 3 = 0.5, and finally obtain the aggregated scenario state value SI = 1.30136.

[0115] Step 3: Construct a state transfer model between nodes, a node state evolution model, and an inter-layer state aggregation model to realize the deduction of node states at the next moment. The specific construction process is as follows:

[0116] (1) Define entity objects, events, and scenarios as nodes at different levels in the network model; define the entity node state partition set HO = {ho 1 , ho 2 , …, ho ns}, ns is the number of entity node states, and the entity node state vector is where represents the probability value that the entity node state is ho i , define the event node state partition set He = {he 1 , he 2 , …, he ks}, ks is the number of event node states, and the event node state vector is represents the probability value that the event node state is he i , define the scenario node state partition set Hs = {hs 1 , hs 2 , …, hs ms}, ms is the number of scenario node states, and the scenario node state vector is represents the probability value that the scenario node state is hs i , and the sum of the components of each type of node state vector is 1;

[0117] (2) Construct a state transfer model between nodes, and construct the state transfer model between nodes as a quadruple TRO = (Onm, Node, A, Num), where Onm represents the name of the state transfer operation, Node represents the set of nodes involved in the node state transition, A represents the attributes of the node state transfer. For example, during the scheduling of relief supplies, Node includes the storage points where supply scheduling operations occur and the set of disaster-affected bodies receiving supplies, and A includes attributes such as food, clothing, and rescue equipment. Num represents the set of numerical values of the state transfer between nodes;

[0118] (3) Construct a node state evolution model. For each node, give an initial state probability distribution and a time-varying state transition matrix to obtain the node state evolution model. The calculation formula is as follows:

[0119] v ′ = v × P t

[0120] v = [p 1 , p 2 , …, p q-1 , p q

[0121]

[0122] Among them, v is the node state vector at the current moment, representing the probability distribution of being in each state at the current moment, and can be any type of node among entity, event, and scenario nodes; P t is the state transition matrix at the current t moment, v ′ is the evolution state vector at the next moment, q represents the number of states of the current attribute node, and t in P represents the probability value that the attribute node at the current t moment changes from state i to state j, and there is

[0123] Taking the disaster-affected body o 1 in the event e 11 ​For example, the attribute nodes of this node are divided into states. Among them, spda divides the number of trapped people from 0 to 500 into state 1, the number of trapped people from 500 to 1500 into state 2, and the number of trapped people over 1500 into state 3; size is divided according to the affected area, where the area from 0 to 5 square kilometers is divided into state 1, 5 to 10 square kilometers into state 2, and over 10 square kilometers into state 3; communication is divided into 3 states, state 1 indicates that the area is completely out of contact, state 2 indicates that the area is partially out of contact, and state 3 indicates that the area has normal communication; traffic_condition is divided into 3 states, state 1 indicates that the area is completely cut off from the outside world in terms of transportation channels, state 2 indicates that the area is partially cut off from the outside world in terms of transportation channels, and state 3 indicates that the transportation channels are normal; at this time, o 11 The node state is that spda is in state 2, size is in state 3, communication is in state 3, and traffic_condition is in state 3; then, combining the knowledge in related fields and the disaster-causing factors of the event at this time En 1 , the state transition matrix at time t is obtained:

[0124]

[0125] According to the node state evolution calculation model: v ′ =P t ×v, the state of the disaster-bearing body o at time t+1 is obtained 11 node state. At the next moment, its spda is in state 2, size is in state 3, communication is in state 3, and traffic_condition is in state 3.

[0126] (4) Construct an inter-layer node state aggregation model. First, construct a node state aggregation function between entity nodes and event nodes. The input of this function is the set of entity nodes involved in the event, and the output is the event aggregation state u. The aggregation calculation formula is as follows:

[0127]

[0128] where e ′ is the aggregated event state vector, W i represents the transition probability matrix of different states of the i-th entity object affecting event e, p m,q represents the probability that the entity object in state m affects event state q. The Softmax function normalizes the aggregation result, z i is The i-th component, and then construct an aggregation calculation function between the event node and the scenario node state. The input of this function is the set of event nodes involved in the scenario, and the output is the aggregated scenario state The aggregation calculation formula is as follows:

[0129]

[0130] where s ′ is the aggregated scenario state vector, T i represents the transition probability matrix of the i-th event object in different states affecting the scenario s, and p i,j represents the probability that the event object in state i affects the scenario state j. The Softmax function normalizes the aggregation result, Y i is the i-th component, and d is the number of event nodes. The specific process is as follows: ① Divide the node states. Divide the current flood control and disaster relief scenario state into a getting better state with a state value of 1 and a deteriorating state with a state value of 2. Divide the disaster event state values between 0 and 4.65 into state 1, and divide the disaster event state values between 4.65 and 9 into state 2; divide the rescue event of the rescue event according to the rescue effect state value in step three into two states. The rescue effect state value of state 1 is between 0 and 0.65, and the rescue state effect value of state 2 is greater than 0.65; each carrier state is divided into 2 states according to its own vulnerability score of the disaster-bearing body. The vulnerability score of state 1 is between 1 and 6, and the vulnerability score of state 2 is between 6 and 9, and give the initial state of the disaster-bearing body node. ② Give the state vectors of each layer of nodes and the inter-layer transition probability matrix. ③ Use the inter-layer aggregation model to calculate the scenario state vector; finally, with 10 disaster-bearing bodies as evidence and 3 events as the middle layer, the probability value of the flood control and disaster relief scenario state obtained by aggregation is P(s = 1) = 0.8165, P(s = 2) = 0.1835, and the finally aggregated current scenario state is the getting better state.

[0131] Based on the same inventive concept, an embodiment of the present invention discloses a flood control and disaster relief dynamic scenario modeling system based on a meta-model and relationship aggregation, including:

[0132] A meta-model construction module for constructing a meta-model of flood control and disaster relief entity objects, event objects, and flood control and disaster relief scenario meta-models; the event objects include disaster event objects and disaster relief event objects, the disaster event objects include a set of disaster-affected bodies involved in the disaster event, disaster-causing factors, and a status value indicating the emergency level of the disaster event; the disaster relief event objects include a name, a set of disaster-affected bodies assisted by the disaster relief event, a set of attributes, and a rescue effect status value; the flood control and disaster relief scenario includes a set of disaster events, a set of disaster relief events, and a status value indicating the emergency level of the flood control and disaster relief scenario;

[0133] A model aggregation module for constructing an intra-layer association relationship model and an inter-layer relationship aggregation model of entity object - event - scenario; the intra-layer association relationship includes the relationship between entity objects, and the inter-layer relationship aggregation includes the aggregation of the relationship between entities and disaster events, the aggregation of the relationship between entities and disaster relief events, and the inter-layer relationship aggregation between events and scenarios; the aggregation of the relationship between entities and disaster events includes determining the status value of the disaster event according to the vulnerability of the disaster-affected bodies of the disaster event object and the danger of the disaster-causing factors; the aggregation of the relationship between entities and disaster relief events includes determining the rescue effect status value of the disaster relief event according to the vulnerability of the disaster-affected bodies of the disaster relief event object and the rescue effect; the inter-layer relationship aggregation between events and scenarios includes determining the scenario status value according to the status value of the disaster event and the rescue effect status value of the disaster relief event;

[0134] And a deduction module for constructing a node state transfer model, a node state evolution model, and an inter-layer state aggregation model of entity object - event - scenario to realize node state deduction.

[0135] An embodiment of the present invention also discloses a computer system, including a memory, a processor, and a computer program / instructions stored on the memory and executable on the processor. When the computer program / instructions are executed by the processor, the steps of a flood control and disaster relief dynamic scenario modeling method based on meta-model and relationship aggregation are implemented.

[0136] An embodiment of the present invention also discloses a computer program product, including computer program / instructions. When the computer program / instructions are executed by the processor, the steps of a flood control and disaster relief dynamic scenario modeling method based on meta-model and relationship aggregation are implemented.

[0137] The above is the preferred embodiment of the present invention. It should be noted that for those of ordinary skill in the art in this technical field, without departing from the technical principle of the present invention, several improvements and deformations can be made, and these improvements and deformations should also be regarded as the protection scope of the present invention.

Claims

1. A method for dynamic scenario modeling of flood control and disaster relief based on metamodel and relational aggregation, characterized in that: The steps include: (1) Constructing a flood control and disaster relief entity object metamodel, an event object metamodel, and a flood control and disaster relief scenario metamodel; the event object includes a disaster event object and a disaster relief event object; the disaster event object includes a set of disaster-affected objects involved in the disaster event, disaster-causing factors, and a state value indicating the urgency of the disaster event; the disaster relief event object includes a name, a set of disaster-affected objects rescued by the disaster relief event, an attribute set, and a rescue effect state value; the flood control and disaster relief scenario includes a set of disaster events, a set of disaster relief events, and a state value indicating the urgency of the flood control and disaster relief scenario; (2) constructing an intra-layer association relationship model and an inter-layer relationship aggregation model of entity object-event-scenario; the intra-layer association relationship includes the relationship between entity objects, and the inter-layer relationship aggregation includes the aggregation of the relationship between entity and disaster event, the aggregation of the relationship between entity and disaster relief event, and the aggregation of the relationship between event and scenario; The aggregation of the relationship between the entity and the disaster event includes determining the disaster event state value according to the vulnerability of the disaster-bearing body of the disaster event object and the danger of the disaster-causing factor; the aggregation of the relationship between the entity and the disaster relief event includes determining the rescue effect state value of the disaster relief event according to the vulnerability of the disaster-bearing body of the disaster relief event object and the rescue effect; the aggregation of the relationship between the event and the scenario layer includes determining the scenario state value according to the disaster event state value and the rescue effect state value of the disaster relief event; (3) Construct the node state transfer model, node state evolution model and inter-layer state aggregation model of entity object-event-scenario to realize node state deduction.

2. The method for dynamic scenario modeling of flood prevention and disaster relief based on metamodel and relationship aggregation according to claim 1 is characterized in that: Step (1) also includes constructing a metadata meta-model; the metadata includes descriptive metadata, management metadata, and usability metadata obtained by functionally dividing the flood control and disaster relief scene information; the descriptive metadata includes attribute descriptions, spatial locations, and corresponding authority information of flood control and disaster relief materials and engineering entities; The management metadata includes information on the entire life cycle of the object; the usability metadata includes query behavior record information.

3. The method for dynamic scenario modeling of flood prevention and disaster relief based on metamodel and relationship aggregation according to claim 1 is characterized in that: The flood prevention and disaster relief entity object metamodel in step (1) includes a flood disaster area information metamodel, a transfer and resettlement area information metamodel, a material storage point information metamodel, a rescue agency information metamodel, a rescue material information metamodel and an emergency measures information metamodel; The flood-affected area information meta-model includes the flood-affected area code, name, area, location, number of people, and number of people to be transferred / rescued; The transfer and resettlement area information element model includes the transfer and resettlement area code, name, area, location, grade, number of people that can be accommodated and number of people who have been resettled; The material reserve point information meta-model includes the material reserve point code, name, location, level and category; The rescue agency information meta-model includes rescue agency code, name, location and category; The emergency rescue material information meta-model includes material reserve number, material number, material storage warehouse number, material quantity, material storage unit and material type; The emergency measures information element module includes the emergency measures number, the rescue team number, the team's affiliated organization number, the rescue destination, the number of people involved in the rescue, the emergency measures category and the completion status of the emergency measures.

4. The method for dynamic scenario modeling of flood prevention and disaster relief based on metamodel and relationship aggregation according to claim 1 is characterized in that: In step (1), the disaster event object metamodel is constructed as a triple E = (O, En, F), where O represents the set of hazard-bearing objects involved in the current disaster event O = {o1, o2, ..., o k }, k represents the number of disaster-bearing bodies involved in the disaster event object, En is the disaster factor of the current disaster event, and F represents the current disaster event state value, which is used to indicate the urgency of the current disaster event; The disaster-bearing volume element model is a five-tuple O = (Nm o ,At o , Vul, Hl, I), where Nm o is the name of the disaster-bearing body, At o is the attribute set of the disaster-bearing body, At o ={a o1 , a o2 , ..., a ot }, t represents the number of attributes of the current hazard-prone body, Vul is the set of vulnerability scores of the hazard-prone body attributes Vul = {v1, v2, ..., v t },v i Indicates a i Hl is the rescue effect value of the disaster-prone body, and I represents the comprehensive vulnerability score of the current disaster-prone body; The hazard factor metamodel is a four-tuple En = (Nm z ,At z , Da, H), where Nm z is the name of the hazard factor, At z is the attribute set of disaster factors At z ={a z1 , a z2 , ..., a zl }, l represents the number of current disaster factor attributes, Da is the set of hazard value of disaster factor attributes Da = {d1, d2, ..., d l }, d i represents the risk score, H represents the comprehensive risk score of the hazard factors of the current disaster event; The disaster relief event object metamodel is constructed as a four-tuple HE=(Nm j ,O,At j , X), where Nm j represents the name of the disaster relief event, O represents the set of disaster-affected objects rescued by the current disaster relief event O = {o1, o2, ..., o m }, m represents the number of disaster-affected objects involved in the current disaster relief event, At j is the attribute set At of disaster relief events j ={a j1 , a j2 , ..., a jh }, h represents the number of attributes of the current disaster relief event, and X represents the rescue effect status value.

5. The method for dynamic scenario modeling of flood prevention and disaster relief based on metamodel and relationship aggregation according to claim 1 is characterized in that: The flood prevention and disaster relief scenario metamodel in step (1) is constructed as a triple S = (E, HE, SI), where E represents the current disaster event set, and E = {e i |i=1,…,|E|},e i represents the i-th disaster event in the current scenario, HE represents the set of disaster relief events in the current scenario, and HE = {he i |i=1,…,|HE|}, SI represents the situation state value of the current situation, which is used to indicate the urgency of the flood control and disaster relief situation.

6. The method for dynamic scenario modeling of flood prevention and disaster relief based on metamodel and relationship aggregation according to claim 1 is characterized in that: The formula for the disaster event status value F in step (2) is as follows: Where H represents the comprehensive risk score of disaster factors. d i is the risk score corresponding to the i-th attribute value of the hazard factor of the current event, l represents the number of hazard factor attributes, I i represents the vulnerability score of the ith hazard-prone body, k represents the number of hazard-prone bodies affected by the disaster event; the vulnerability score of a single hazard-prone body is v i represents the vulnerability score corresponding to the i-th attribute value of the hazard-prone body, and t represents the number of attributes of the current hazard-prone body; The formula for the disaster relief event rescue effect status value X in step (2) is as follows: where h i represents the rescue effect value of the i-th disaster-stricken object, and m represents the number of disaster-stricken objects rescued by the disaster relief event; The scenario state value in step (2) is obtained by weighted summing the disaster event state value and the disaster relief event rescue effect state value in the current scenario.

7. The method for dynamic scenario modeling of flood prevention and disaster relief based on metamodel and relationship aggregation according to claim 1 is characterized in that: The entity-event-scenario node state transfer model, node state evolution model and inter-layer state aggregation model in step (3) include the following steps: (1) Define entity objects, events, and scenarios as nodes at different levels in the network model; define the entity node state partition set HO = {ho1, ho2, ..., ho ns }, ns is the number of entity node states, and the entity node state vector is in Indicates that the entity node state is ho i The probability value of event node state partition set is defined ks is the number of event node states, and the event node state vector is Indicates that the event node state is he i The probability value of the scenario node state partition set Hs = {hs1, hs2, ..., hs ms }, ms is the number of scenario node states, and the scenario node state vector is Indicates that the scenario node status is hs i The probability value of each type of node state vector is 1; (2) Constructing a state transfer model between nodes. The state transfer model between nodes is constructed as a four-tuple TRO = (Onm, Node, A, Num), where Onm represents the name of the state transfer operation, Node represents the set of nodes involved in the node state transfer, A represents the attribute of the node state transfer, and Num represents the set of number values ​​of the state transfer between nodes; (3) Construct a node state evolution model. For each node, give the initial state probability distribution and a time-varying state transfer matrix to obtain the node state evolution model. The calculation formula is as follows: v′=v×P t v=[p1,p2,…,p q-1 ,p q ] Where v is the node state vector at the current moment, which indicates the probability distribution of each state at the current moment. It can be any type of node among entity, event and scenario nodes; P t is the state transfer matrix at the current time t, v′ is the evolution state vector at the next time, q represents the state number of the current attribute node, P t In represents the probability value of the attribute node changing from state i to state j at the current time t, and (4) Construct an inter-layer node state aggregation model. First, construct a node state aggregation function between entity nodes and event nodes. The input of this function is the set of entity nodes involved in the event, and the output is the event aggregation state u. The aggregation calculation formula is as follows: Where e′ is the event state vector after aggregation, W i represents the transition probability matrix of the event e affected by different states of the i-th entity object, p m，q It indicates the probability of affecting the event state q when the entity object is in state m. The Softmax function normalizes the aggregation result. z i for The i-th component, n is the number of entity nodes, and then the aggregation calculation function between the event node and the scenario node state is constructed. The input of this function is the set of event nodes involved in the scenario, and the output is the scenario aggregation state The aggregation calculation formula is as follows: where s ′ is the aggregated scenario state vector, T i represents the transition probability matrix of different states of the i-th event object affecting scenario s, p i,j It represents the probability of affecting the situation state j when the event object is in state i. The Softmax function normalizes the aggregation result. Y i for The i-th component, d is the number of event nodes.

8. A dynamic scenario modeling system for flood prevention and disaster relief based on metamodel and relational aggregation, characterized in that: include: A metamodel construction module is used to construct a flood control and disaster relief entity object metamodel, an event object metamodel and a flood control and disaster relief scenario metamodel; the event object includes a disaster event object and a disaster relief event object, the disaster event object includes a set of disaster-affected bodies involved in the disaster event, disaster-causing factors and a state value indicating the urgency of the disaster event; the disaster relief event object includes a name, a set of disaster-affected bodies rescued by the disaster relief event, an attribute set and a rescue effect state value; the flood control and disaster relief scenario includes a set of disaster events, a set of disaster relief events and a state value indicating the urgency of the flood control and disaster relief scenario; A model aggregation module is used to construct an intra-layer association relationship model and an inter-layer relationship aggregation model of entity object-event-scenario; the intra-layer association relationship includes the relationship between entity objects, and the inter-layer relationship aggregation includes the aggregation of the relationship between entity and disaster event, the aggregation of the relationship between entity and disaster relief event, and the aggregation of the relationship between event and scenario; The aggregation of the relationship between the entity and the disaster event includes determining the disaster event state value according to the vulnerability of the disaster-bearing body of the disaster event object and the danger of the disaster-causing factor; the aggregation of the relationship between the entity and the disaster relief event includes determining the rescue effect state value of the disaster relief event according to the vulnerability of the disaster-bearing body of the disaster relief event object and the rescue effect; the aggregation of the relationship between the event and the scenario layer includes determining the scenario state value according to the disaster event state value and the rescue effect state value of the disaster relief event; And the deduction module is used to construct the node state transfer model, node state evolution model and inter-layer state aggregation model of entity object-event-scenario to realize node state deduction.

9. A computer system comprising a memory, a processor, and a computer program / instruction stored in the memory and executable on the processor, characterized in that: When the computer program / instructions are executed by the processor, the steps of a method for dynamic scenario modeling of flood control and disaster relief based on metamodel and relationship aggregation are implemented according to any one of claims 1 to 7.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by the processor, the steps of a method for dynamic scenario modeling of flood control and disaster relief based on metamodel and relationship aggregation are implemented according to any one of claims 1 to 7.

Images