Emergency resource allocation method and system based on demand prediction

Abstract

The invention provides an emergency resource allocation method and system based on demand prediction, and the method comprises the steps: predicting the demand of a disaster point for various emergency resources through a pre-constructed resource demand prediction model according to the current power failure event information of the disaster point when the power failure of the disaster point occurs; based on the demand quantity of the disaster point for various emergency materials, a pre-constructed power grid domain collaborative emergency resource allocation optimization model is used for allocation, and an emergency resource allocation scheme of the disaster point is obtained; and carrying out emergency resource allocation based on the emergency resource allocation scheme and the emergency material library information of the plurality of emergency rescue points. The resource demand prediction model is adopted to timely and accurately predict the demand quantity of various emergency materials of the disaster point, the power grid domain is utilized to cooperate with the emergency resource allocation optimization model for allocation, the emergency resource allocation scheme of the disaster point is matched, and the problem of power grid emergency resource waste is avoided.

Description

Technical Field

[0001] The present invention belongs to the technical field of emergency rescue for sudden incidents, and in particular relates to an emergency resource allocation method and system based on demand forecasting. Background Art

[0002] The prior art discloses (publication number: CN108734388A) an electric power material allocation system, including: a demand summary module, used to collect and summarize the demander's application requirements for various types of electric power materials, wherein the electric power material application requirements include at least one of the requirements for warehouse reserve materials, first aid kit replenishment requirements and emergency repair materials; a material allocation management module, used to allocate appropriate distribution materials to the corresponding material application requirements through a pre-established material matching model according to the various types of electric power material application requirements collected and summarized by the demand summary module, and generate reasonable distribution plans and loading plans for the distribution materials through a pre-established material distribution model and material loading and unloading model; a monitoring module, used to monitor the supply and distribution of various types of electric power materials in real time. The implementation of the present invention is conducive to improving the efficiency and preparedness of material distribution and supply, and saving distribution costs.

[0003] The prior art also discloses (publication number: CN114626727B) a global material emergency allocation system and its allocation method. It includes: step 1, applying the optimal weighting method to determine the weight of each influencing factor, the influencing factors include transportation distance, warehouse age, comprehensive evaluation of suppliers, and whether it is a historical unqualified supplier; step 2, encoding by encoder (LabelEncoder) to realize the numerical normalization conversion of each influencing factor; step 3, calculating the comprehensive recommended score Wj of the factor by weighted average, the higher the score, the more it meets the allocation requirements, and vice versa.

[0004] The prior art also has the following features: at least one set of material storage devices is set up as the front end for users to obtain materials; at least one material warehouse is set up as a guarantee for the material storage device to have sufficient materials; the horizontal distance between the material storage device and the nearest material warehouse does not exceed the set distance; the allocation management platform determines whether the material storage device has materials of preset quantity and specifications, and timely allocates materials to the storage devices that need to be filled, so as to ensure that the time for the shortage of materials in the storage devices that need to be filled does not exceed the set time. The present invention can allocate the time, specifications and quantity of the materials arriving at the storage devices through the user's material demand information and the time when the user arrives at the material storage device according to the positioning data; it can better arrange and utilize limited facilities and materials, so as to better serve customers and achieve greater social benefits.

[0005] The above-mentioned existing technologies all have the following problems: the power grid emergency resource allocation system in the existing technology is affected by uncertain factors and cannot accurately allocate resources, resulting in a waste of power grid emergency resources. Summary of the invention

[0006] In order to overcome the above-mentioned deficiencies of the prior art, in a first aspect, the present invention proposes an emergency resource allocation method based on demand forecasting, comprising:

[0007] When a power outage occurs at a disaster site, the demand for various emergency resources at the disaster site is predicted using a pre-built resource demand prediction model based on the current power outage event information at the disaster site;

[0008] Based on the demand for various emergency materials at the disaster-stricken site, a pre-built power grid domain collaborative emergency resource allocation optimization model is used to allocate emergency resources to obtain an emergency resource allocation plan for the disaster-stricken site;

[0009] Emergency resource allocation is performed based on the emergency resource allocation plan and the emergency material warehouse information of multiple emergency rescue points.

[0010] Preferably, when a power outage occurs at the disaster site, the demand for various emergency resources at the disaster site is predicted using a pre-built resource demand prediction model based on the current power outage event information at the disaster site, including:

[0011] When a power outage occurs at a disaster site, real-time video is collected through multiple image acquisition devices in the power outage area of ​​the disaster site;

[0012] Identifying changes in various emergency resources at the disaster site in the real-time video;

[0013] According to the current power outage event information of the disaster-stricken point and the changes of various emergency resources of the disaster-stricken point, the demand of the disaster-stricken point for various emergency resources is predicted by using a pre-built resource demand prediction model.

[0014] Preferably, the pre-construction process of the power grid domain collaborative emergency resource allocation optimization model includes:

[0015] Analyze the consumption rate of various emergency resources and the shortest time of dispatching various emergency resources in the historical power outage event information of the disaster-stricken point, and establish an emergency resource deployment time sub-model;

[0016] Analyze the emergency rescue points involved in emergency resource dispatch in the historical power outage event information, and establish a collaborative degree sub-model for emergency resource collaborative dispatch;

[0017] Establishing constraint conditions for the emergency resource allocation plan based on the amount of emergency resources provided by each emergency rescue point to the disaster-stricken point;

[0018] Based on the emergency resource allocation time sub-model, the emergency resource collaborative scheduling coordination degree sub-model and the constraint conditions, a power grid domain collaborative emergency resource allocation optimization model is constructed.

[0019] Preferably, the constraint conditions of the emergency resource allocation plan are established based on the quantity of emergency resources provided by each emergency rescue point to the disaster-stricken point, including:

[0020] The quantity of each type of emergency resource provided by each emergency rescue point to the disaster-stricken point is less than or equal to the storage quantity of the emergency resource of the emergency rescue point, as the first sub-constraint condition.

[0021] The number of each type of emergency resources provided by all emergency rescue points is greater than or equal to the number of emergency resources of the type required by the disaster-stricken point as the second sub-constraint condition;

[0022] The number of each type of emergency resources provided by each emergency rescue point to the disaster-stricken point is set as a non-negative integer as the third sub-constraint condition;

[0023] Based on the first sub-constraint condition, the second sub-constraint condition, and the third sub-constraint condition, a constraint condition of an emergency resource allocation plan is established.

[0024] Preferably, the pre-construction process of the resource demand prediction model includes:

[0025] Analyze the historical power outage event information of the disaster-stricken area to obtain historical power outage information and historical actual demand for various emergency supplies;

[0026] A linear regression analysis algorithm is used to establish a relationship between the historical power outage information and the historical actual demand for various emergency supplies;

[0027] Based on the relationship between the historical power outage information and the historical actual demand for the various types of emergency supplies, the resource demand prediction model is constructed using data analysis and prediction software.

[0028] Preferably, the linear regression analysis algorithm is used to establish the relationship between the historical power outage information and the historical actual demand for various types of emergency supplies, including:

[0029] Based on the historical power outage information, the number of users without power, the length of damaged power grid lines, the number of damaged cables, and the maximum rainfall are obtained;

[0030] The number of users experiencing power outages, the length of damaged power grid lines, the number of damaged cables, and the maximum rainfall are taken as independent variables, and the actual historical demand and types of various emergency supplies are taken as dependent variables. A linear regression analysis algorithm is used to establish the relationship between the historical power outage information and the actual historical demand for various emergency supplies.

[0031] Preferably, the obtaining, based on the historical power outage information, the number of power outage users, the length of damaged power grid lines, the number of damaged cables, and the maximum rainfall includes:

[0032] A deep learning model is used to extract the number of users experiencing power outages, the length of damaged power grid lines, the number of damaged cables, and the maximum rainfall from the ledgers, records, work daily reports of each emergency rescue point, and historical power outage event information of the disaster-stricken point.

[0033] Preferably, the emergency resource allocation scheme for the disaster-stricken point is obtained by using a pre-built power grid domain collaborative emergency resource allocation optimization model to allocate various emergency materials based on the demand for the disaster-stricken point, including:

[0034] Based on the demand for various emergency supplies at the disaster-stricken point, the pre-constructed power grid domain collaborative emergency resource allocation optimization model is solved using the particle swarm algorithm to obtain the emergency resource allocation plan for the disaster-stricken point.

[0035] Preferably, the grid domain collaborative emergency resource allocation optimization model satisfies the following formula:

[0036]

[0037] In the above formula: t i represents the emergency resource deployment time from the ith emergency rescue point to the disaster site, k represents the type of emergency resource, m represents the total number of emergency resource types, V k represents the consumption rate of the kth emergency resource; i represents the emergency rescue point, n represents the total number of emergency rescue points, S ik represents a variable of 0 or 1, P1 represents the required target emergency resources, Z1 represents the coordinated deployment time of the target emergency resources, It represents the quantity of the kth type of emergency resources provided by the i-th emergency rescue point to the disaster-stricken point.

[0038] In a second aspect, the present invention also proposes an emergency resource allocation system based on demand forecasting, comprising:

[0039] The emergency resource demand prediction module is used to predict the demand of the disaster-stricken point for various emergency resources by using a pre-built resource demand prediction model according to the current power outage event information of the disaster-stricken point when a power outage occurs at the disaster-stricken point;

[0040] The emergency resource allocation scheme module is used to allocate the emergency resources based on the demand of the disaster-stricken site for various emergency materials by using the pre-built power grid domain collaborative emergency resource allocation optimization model to obtain the emergency resource allocation scheme for the disaster-stricken site;

[0041] The emergency resource allocation module is used to allocate emergency resources based on the emergency resource allocation plan and the emergency material warehouse information of multiple emergency rescue points.

[0042] Preferably, the emergency resource demand prediction module includes:

[0043] When a power outage occurs at a disaster site, real-time video is collected through multiple image acquisition devices in the power outage area of ​​the disaster site;

[0044] Identifying changes in various emergency resources at the disaster site in the real-time video;

[0045] According to the current power outage event information of the disaster-stricken point and the changes of various emergency resources of the disaster-stricken point, the demand of the disaster-stricken point for various emergency resources is predicted by using a pre-built resource demand prediction model.

[0046] Preferably, the system further comprises: a power grid domain collaborative emergency resource allocation optimization model building module, which is used to:

[0047] Analyze the consumption rate of various emergency resources and the shortest time of dispatching various emergency resources in the historical power outage event information of the disaster-stricken point, and establish an emergency resource deployment time sub-model;

[0048] Analyze the emergency rescue points involved in emergency resource dispatch in the historical power outage event information, and establish a collaborative degree sub-model for emergency resource collaborative dispatch;

[0049] Establishing constraint conditions for the emergency resource allocation plan based on the amount of emergency resources provided by each emergency rescue point to the disaster-stricken point;

[0050] Based on the emergency resource allocation time sub-model, the emergency resource collaborative scheduling coordination degree sub-model and the constraint conditions, a power grid domain collaborative emergency resource allocation optimization model is constructed.

[0051] Preferably, the power grid domain collaborative emergency resource allocation optimization model building module is specifically used to:

[0052] The quantity of each type of emergency resource provided by each emergency rescue point to the disaster-stricken point is less than or equal to the storage quantity of the emergency resource of the emergency rescue point, as the first sub-constraint condition.

[0053] The number of each type of emergency resources provided by all emergency rescue points is greater than or equal to the number of emergency resources of the type required by the disaster-stricken point as the second sub-constraint condition;

[0054] The number of each type of emergency resources provided by each emergency rescue point to the disaster-stricken point is set as a non-negative integer as the third sub-constraint condition;

[0055] Based on the first sub-constraint condition, the second sub-constraint condition, and the third sub-constraint condition, a constraint condition of an emergency resource allocation plan is established.

[0056] Preferably, the system further comprises: a resource demand prediction model building module, which is used to:

[0057] Analyze the historical power outage event information of the disaster-stricken area to obtain historical power outage information and historical actual demand for various emergency supplies;

[0058] A linear regression analysis algorithm is used to establish a relationship between the historical power outage information and the historical actual demand for various emergency supplies;

[0059] Based on the relationship between the historical power outage information and the historical actual demand for the various types of emergency supplies, the resource demand prediction model is constructed using data analysis and prediction software.

[0060] Preferably, the resource demand prediction model building module includes:

[0061] An acquisition unit, configured to acquire the number of users without power, the length of damaged power grid lines, the number of damaged cables, and the maximum rainfall based on the historical power outage information;

[0062] The relationship acquisition unit is used to use the number of power outage users, the length of the damaged power grid lines, the number of damaged cables, and the maximum rainfall as independent variables, and the historical actual demand and type of the various emergency materials as dependent variables, and use a linear regression analysis algorithm to establish the relationship between the historical power outage information and the historical actual demand for the various emergency materials.

[0063] Preferably, the acquisition unit is specifically used to: use a deep learning model to extract the number of power outage users, the length of damaged power grid lines, the number of damaged cables, and the maximum rainfall from the ledgers, records, work daily reports of each emergency rescue point and the historical power outage event information of the disaster-stricken point.

[0064] Preferably, the emergency resource allocation plan module is specifically used to: based on the demand for various emergency materials at the disaster-stricken point, use the particle swarm algorithm to solve the pre-constructed power grid domain collaborative emergency resource allocation optimization model to obtain the emergency resource allocation plan for the disaster-stricken point.

[0065] Preferably, the grid domain collaborative emergency resource allocation optimization model satisfies the following formula:

[0066]

[0067] In the above formula: t irepresents the emergency resource deployment time from the ith emergency rescue point to the disaster site, k represents the type of emergency resource, m represents the total number of emergency resource types, V k represents the consumption rate of the kth emergency resource; i represents the emergency rescue point, n represents the total number of emergency rescue points, S ik represents a variable of 0 or 1, P1 represents the required target emergency resources, Z1 represents the coordinated deployment time of the target emergency resources, It represents the quantity of the kth type of emergency resources provided by the i-th emergency rescue point to the disaster-stricken point.

[0068] In a third aspect, the present invention application further proposes an electronic device, comprising: at least one processor and a memory; the memory and the processor are connected via a bus;

[0069] The memory is used to store one or more programs;

[0070] When the one or more programs are executed by the at least one processor, the method for emergency resource allocation based on demand forecasting is implemented.

[0071] In a fourth aspect, the present invention application further proposes a readable storage medium having an execution program stored thereon, and when the execution program is executed, the emergency resource allocation method based on demand forecasting is implemented.

[0072] Compared with the closest prior art, the present invention has the following beneficial effects:

[0073] The present invention applies for an emergency resource allocation method and system based on demand forecasting, including: when a power outage occurs at a disaster site, based on the current power outage event information of the disaster site, a pre-constructed resource demand prediction model is used to predict the demand of the disaster site for various emergency resources; based on the demand of the disaster site for various emergency materials, a pre-constructed power grid domain collaborative emergency resource allocation optimization model is used to allocate, and an emergency resource allocation plan for the disaster site is obtained; based on the emergency resource allocation plan and the emergency material warehouse information of multiple emergency rescue points, emergency resources are allocated. The current power outage event information of the disaster site is predicted using a resource demand prediction model, so the demand of the disaster site for various emergency materials can be predicted in a timely and accurate manner; because the current power outage event information is used, the predicted demand of the disaster site for various emergency materials will not be affected by uncertain factors; at the same time, the power grid domain collaborative emergency resource allocation optimization model is used for allocation, and an emergency resource allocation plan for the disaster site is matched to avoid the problem of waste of power grid emergency resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 A flowchart of an emergency resource allocation method based on demand forecasting provided by the present invention;

[0075] Figure 2 A flow chart of the particle swarm algorithm provided for the present invention application;

[0076] Figure 3 An architecture diagram of an emergency resource allocation system based on demand forecasting provided by the present invention;

[0077] Figure 4 A schematic diagram of the operation of an electronic device provided in the present invention application. DETAILED DESCRIPTION

[0078] The specific implementation methods of the present invention are further described in detail below with reference to the accompanying drawings.

[0079] Embodiment 1:

[0080] like Figure 1 As shown, the present invention provides an emergency resource allocation method based on demand forecasting, comprising the following steps:

[0081] Step 1: When a power outage occurs at a disaster site, the demand for various emergency resources at the disaster site is predicted using a pre-built resource demand prediction model based on the current power outage event information at the disaster site;

[0082] Step 2: Based on the demand for various emergency materials at the disaster-stricken site, a pre-built power grid domain collaborative emergency resource allocation optimization model is used to allocate and obtain an emergency resource allocation plan for the disaster-stricken site;

[0083] Step 3: Based on the emergency resource allocation plan and the emergency material warehouse information of multiple emergency rescue points, emergency resource allocation is performed.

[0084] The pre-construction process of the resource demand prediction model in step 1 above may include:

[0085] Step a, analyzing the historical power outage event information of the disaster site to obtain historical power outage information and historical actual demand for various emergency supplies;

[0086] Step b, using a linear regression analysis algorithm to establish a relationship between the historical power outage information and the historical actual demand for various types of emergency supplies;

[0087] Step c: Based on the relationship between the historical power outage information and the historical actual demand for the various emergency supplies, a resource demand prediction model is constructed using data analysis and prediction software.

[0088] The resource demand prediction model is based on historical power outage information, and uses ontology development software (Protégé) to build an ontology model, analyze the important concepts of the standard ontology of power emergency materials, and determine the relationship between classes to form a complex network between entities. The ontology concepts and the relationship between each concept are established to form an ontology framework. The ontology concepts include: emergency resource allocation information in historical power outage event information, the geographical location of the disaster site, the number of emergency rescue points, the types of various emergency materials and the actual historical demand quantity, the rescue team, and the damaged data of the power grid in the power outage event; the linear regression analysis algorithm is used to analyze the relationship between historical power outage information and the actual historical demand for various emergency materials, and a resource demand prediction model for emergency rescue in large-scale power outage events is established.

[0089] In the above step b, the use of a linear regression analysis algorithm to establish a relationship between the historical power outage information and the historical actual demand for various types of emergency supplies may include:

[0090] Based on the historical power outage information, the number of users without power, the length of damaged power grid lines, the number of damaged cables, and the maximum rainfall are obtained;

[0091] The number of users experiencing power outages, the length of damaged power grid lines, the number of damaged cables, and the maximum rainfall are taken as independent variables, and the actual historical demand and types of various emergency supplies are taken as dependent variables. A linear regression analysis algorithm is used to establish the relationship between the historical power outage information and the actual historical demand for various emergency supplies.

[0092] As mentioned above, when using a linear regression analysis algorithm to analyze the relationship between the historical power outage event information of the disaster-stricken area and the historical demand for various emergency materials at the disaster-stricken area, the data analysis and prediction (Statistical Package for the Social Sciences, SPSS) software can be used to analyze the data in the historical power outage event information, and a multiple regression equation with the power outage area, damaged substation, number of power outage lines, etc. as the independent variables of the multiple regression equation and the emergency resource demand as the dependent variable is obtained. Therefore, the multiple regression equation has the function of predicting resource demand and can dynamically update data in real time.

[0093] In the above, obtaining the number of power outage users, the length of damaged power grid lines, the number of damaged cables, and the maximum rainfall based on the historical power outage information may include:

[0094] A deep learning model is used to extract the number of users experiencing power outages, the length of damaged power grid lines, the number of damaged cables, and the maximum rainfall from the ledgers, records, work daily reports of each emergency rescue point, and historical power outage event information of the disaster-stricken point.

[0095] The above-mentioned deep learning model can be a BERT (Bidirectional Encoder Representation from Transformers) model or a BERT model combined with a BiLSTM-CRF (Bidirectional Long Short Term Memory-Conditional Random Field) deep learning model, which completes the extraction task and forms a knowledge graph from existing ledgers, records, work daily reports and historical event case reports. The extraction tasks include: the number of electricity users, the length of damaged power grid lines, the number of damaged cables, the maximum rainfall, etc. The graph database Neo4j is selected as the storage and query basis of the knowledge graph, and the constructed ontology framework and the extracted knowledge are imported to finally form a knowledge graph. The knowledge graph can be used as a power emergency resource library; wherein the extracted information can include all concepts in the ontology framework.

[0096] In the above step 1, when a power outage occurs at the disaster site, based on the current power outage event information of the disaster site, the demand for various emergency resources at the disaster site is predicted using a pre-built resource demand prediction model, which may include:

[0097] Step 1.1, when a power outage occurs at a disaster site, real-time video is collected by using multiple image acquisition devices in the power outage area of ​​the disaster site;

[0098] Step 1.2, identifying changes in various emergency resources at the disaster site in the real-time video;

[0099] Step 1.3, based on the current power outage event information of the disaster-stricken site and the changes in various emergency resources of the disaster-stricken site, use the pre-built resource demand prediction model to predict the demand of the disaster-stricken site for various emergency resources.

[0100] In the above step 1.1, the image acquisition device can be a battery-powered camera or a solar-powered camera, which can meet the requirements of real-time video acquisition in the event of a power outage;

[0101] The pre-construction process of the power grid domain collaborative emergency resource allocation optimization model in step 2 above may include:

[0102] Step 2.1, analyzing the consumption rate of various emergency resources and the shortest time for dispatching various emergency resources in the historical power outage event information of the disaster-stricken point, and establishing an emergency resource deployment time sub-model;

[0103] Step 2.2, analyzing the emergency rescue points involved in the emergency resource dispatch in the historical power outage event information, and establishing a sub-model of the emergency resource coordinated dispatch;

[0104] Step 2.3, based on the amount of emergency resources provided by each emergency rescue point to the disaster-stricken point, establishing constraint conditions for the emergency resource allocation plan;

[0105] Step 2.4, based on the emergency resource allocation time sub-model, the emergency resource collaborative scheduling coordination degree sub-model and the constraint conditions, construct a power grid domain collaborative emergency resource allocation optimization model.

[0106] In steps 2.1 and 2.2, according to the analysis of the characteristics of coordinated dispatch of emergency resources, the coordinated dispatch of emergency resources for power grid blackout events is a coordinated dispatch problem of emergency resources for single disaster-stricken point, multiple emergency rescue points, multiple types of emergency resources, and continuous consumption. Assume that A is the disaster-stricken point; S is the set of n emergency rescue points, S = {S1,…Si,…Sn}; k is the type of emergency resources required by the disaster-stricken point, k = 1,2,…,m, k represents the kth type of emergency resources, m represents the total number of emergency resource types; xik is the number of the kth type of emergency resources provided by the i-th emergency rescue point to the disaster-stricken point. In order to find an optimal emergency resource dispatching scheme, under the condition of meeting the demand for emergency resources, emergency resources should be reasonably allocated, so that the emergency resource allocation time is the shortest and the emergency resource coordination degree is the largest; the emergency resource coordination degree can be maximized when the number of emergency rescue points with the shortest emergency resource allocation time is the smallest.

[0107] (1) The shortest time for emergency resource deployment can be as follows:

[0108] According to the analysis of coordinated dispatch of power grid emergency resources, the shortest time for dispatching emergency resources should be the primary goal to be considered, that is, how to dispatch m types of emergency resources to the disaster site in the shortest time. Since the dispatch of power grid emergency resources is a multi-type emergency resource dispatch problem, the shortest time for global emergency resource dispatch can also be the maximum number of single-type emergency resource dispatches in the shortest time.

[0109] If the shortest time for dispatching a single type of emergency resource can be obtained, then the shortest time for dispatching all emergency resources can also be obtained. The calculation method for the shortest time for dispatching a single type of emergency resource is as follows;

[0110]

[0111] In the above formula, t i represents the emergency resource dispatching time from the i-th emergency rescue point to the disaster site A; v k represents the consumption rate of the kth type of emergency resources; S ikA variable representing 0 or 1, which takes 1 when the i-th emergency rescue point dispatches the k-th type of emergency resources to the disaster-stricken point A, otherwise it takes 0; P1 represents the required target emergency resources, and Z1 represents the coordinated deployment time of the target emergency resources.

[0112] The shortest time T of the entire emergency resource coordinated scheduling plan is as follows:

[0113] T=max{T k |k=1,2,......,m}

[0114] In the above formula, T k It represents the shortest time for dispatching the k-th emergency resource.

[0115] (2) The maximum coordination degree of emergency resources is as follows:

[0116] The degree of emergency resource coordination can be expressed by the number of emergency rescue points. The fewer the number of emergency rescue points participating in the emergency rescue, the greater the degree of emergency resource coordination. When an emergency rescue point dispatches emergency resources to the disaster site, it can be considered that the emergency rescue point has participated in the emergency rescue work. The calculation of the minimum number of emergency rescue points participating in the emergency rescue is as follows:

[0117]

[0118] In the above formula, S i It represents a 0-1 variable, which takes 1 when the emergency rescue point dispatches emergency resources to the disaster site A, otherwise it takes 0. S represents the number of emergency rescue points participating in the emergency rescue.

[0119] In step 2.3, in order to meet the demand for emergency resources at the disaster site in the power outage event, the constraint conditions of the emergency resource allocation plan are established based on the quantity of emergency resources provided by each emergency rescue point to the disaster site, including:

[0120] The quantity of each type of emergency resource provided by each emergency rescue point to the disaster-stricken point is less than or equal to the storage quantity of the emergency resource of the emergency rescue point, as the first sub-constraint condition.

[0121] The number of each type of emergency resources provided by all emergency rescue points is greater than or equal to the number of emergency resources of the type required by the disaster-stricken point as the second sub-constraint condition;

[0122] The number of each type of emergency resources provided by each emergency rescue point to the disaster-stricken point is set as a non-negative integer as the third sub-constraint condition;

[0123] Based on the first sub-constraint condition, the second sub-constraint condition, and the third sub-constraint condition, a constraint condition of an emergency resource allocation plan is established.

[0124] In the above second sub-constraint, the variability of accident development is taken into account, and the emergency resource demand is uncertain. The size of the emergency resource demand can be expressed in the form of an interval.

[0125] In the above, based on the demand of the disaster-stricken site for various emergency materials, the emergency resource allocation plan for the disaster-stricken site is obtained by using the pre-built power grid domain collaborative emergency resource allocation optimization model, including:

[0126] Based on the demand for various emergency supplies at the disaster-stricken point, the pre-constructed power grid domain collaborative emergency resource allocation optimization model is solved using the particle swarm algorithm to obtain the emergency resource allocation plan for the disaster-stricken point.

[0127] like Figure 2 As shown, the particle swarm optimization (PS0) introduced in the present invention takes the shortest collaborative scheduling time and the maximum emergency resource coordination as the goal of the particle swarm algorithm. Under the circumstances of multiple emergency rescue points, disaster-stricken points, and continuous consumption of multiple emergency materials, the pre-constructed power grid domain collaborative emergency resource allocation optimization model is solved to obtain the optimal solution for emergency resource allocation.

[0128] The particle swarm optimization algorithm is a simulation method of global random search based on swarm intelligence, which promotes the overall evolution of individuals by simulating the migration and social behavior of bird flocks and the social sharing of information. The basic idea of ​​the particle swarm optimization algorithm is to seek the best solution for the group through cooperation and information sharing among individuals in the group. In particle swarm optimization, any particle in the search space can be regarded as a feasible solution to the optimization problem. First, a particle swarm consisting of N particles is initialized, and the optimal solution search is performed in the D-dimensional search space. Among them, the first r The position vector of a particle is expressed as Xr = (x r1 , x r2 , …x rD ), x rD Indicates r The D-th dimensional position vector element of the particle; r The velocity vector of a particle is expressed as Vr = (V r1 ,V r2 ,…,V rD ), V rD Indicates r The D-th dimension velocity vector element of the particle. r The best adaptive position vector experienced by a particle is denoted as P r =(p r1 ,p r2 ,…p rD ), p rD Indicatesr The best adaptive position vector element of the Dth dimension experienced by each particle, the optimal adaptive position vector of the entire particle swarm is P g , also called the global extremum (g best ), when the termination condition is reached, the optimal solution for the group is obtained.

[0129] The above-mentioned grid domain collaborative emergency resource allocation optimization model established under the conditions of multiple emergency rescue points, disaster-stricken points, and continuous consumption of multiple types of emergency materials satisfies the following formula:

[0130]

[0131] In the above formula: t i represents the emergency resource deployment time from the ith emergency rescue point to the disaster site, k represents the type of emergency resource, m represents the total number of emergency resource types, V k represents the consumption rate of the kth emergency resource; i represents the emergency rescue point, n represents the total number of emergency rescue points, S ik represents a variable of 0 or 1, P1 represents the required target emergency resources, Z1 represents the coordinated deployment time of the target emergency resources, It represents the quantity of the kth type of emergency resources provided by the i-th emergency rescue point to the disaster-stricken point.

[0132] In the above step 3: according to the emergency resource allocation plan and the emergency material warehouse information of multiple emergency rescue points obtained in real time, the emergency resources of the disaster-stricken points and multiple emergency rescue points are coordinated and allocated. During the emergency resource allocation process, the quantity and type of emergency resources in the emergency material warehouse information are dynamically updated in real time. It can timely and accurately predict the quantity of emergency materials required in the power outage event, and dynamically update the data. By searching the inventory and geographical location of the material warehouse, according to the different needs of users, the optimal material allocation plan can be matched, thereby improving the efficiency and scientificity of emergency rescue.

[0133] The method of the present invention analyzes the actual emergency teams, emergency materials, and geographical locations and inventory of emergency resource libraries during the power outage rescue process based on the current emergency rescue situation, and proposes a method for predicting and allocating emergency resources, thereby assisting emergency command decision-making and realizing the scientific nature of emergency resource allocation. First, it solves the problem of low efficiency caused by the current reliance on manual labor in the process of emergency rescue. Secondly, the method has the function of dynamically updating data, and combines the on-site conditions reflected by the monitoring equipment to update the material demand in real time, solving the problem of resource waste caused by inaccurate material quantities due to uncertain factors, and realizing rapid and accurate response of various teams and materials in emergencies, scientific and reasonable matching of various resource elements, and maximizing the prevention and reduction of emergencies and the losses and impacts caused by them. It provides an efficient and reliable emergency material allocation plan for enterprises.

[0134] For example, the method of the present invention can be applied to the following scenarios: natural disasters frequently occur in the jurisdiction of the East China Power Supply Bureau, and extreme weather such as typhoons and rainstorms frequently occur, which has a huge impact on the safe production and operation of the power grid, and it is necessary to urgently deploy emergency materials to support power supply restoration, which puts great pressure on the emergency material guarantee work. It can also be used in scenarios where the power material supply plan has slow manual material demand aggregation, low efficiency in finding goods, difficult matching, and unscientific vehicle scheduling.

[0135] Embodiment 2:

[0136] like Figure 3 As shown, the present invention provides an emergency resource allocation system based on demand forecasting, comprising:

[0137] The emergency resource demand prediction module is used to predict the demand of the disaster-stricken point for various emergency resources by using a pre-built resource demand prediction model according to the current power outage event information of the disaster-stricken point when a power outage occurs at the disaster-stricken point;

[0138] The emergency resource allocation scheme module is used to allocate the emergency resources based on the demand of the disaster-stricken site for various emergency materials by using the pre-built power grid domain collaborative emergency resource allocation optimization model to obtain the emergency resource allocation scheme for the disaster-stricken site;

[0139] The emergency resource allocation module is used to allocate emergency resources based on the emergency resource allocation plan and the emergency material warehouse information of multiple emergency rescue points.

[0140] Furthermore, the emergency resource demand prediction module includes:

[0141] When a power outage occurs at a disaster site, real-time video is collected through multiple image acquisition devices in the power outage area of ​​the disaster site;

[0142] Identifying changes in various emergency resources at the disaster site in the real-time video;

[0143] According to the current power outage event information of the disaster-stricken point and the changes of various emergency resources of the disaster-stricken point, the demand of the disaster-stricken point for various emergency resources is predicted by using a pre-built resource demand prediction model.

[0144] Furthermore, the system further comprises: a power grid domain collaborative emergency resource allocation optimization model building module, which is used to:

[0145] Analyze the consumption rate of various emergency resources and the shortest time of dispatching various emergency resources in the historical power outage event information of the disaster-stricken point, and establish an emergency resource deployment time sub-model;

[0146] Analyze the emergency rescue points involved in emergency resource dispatch in the historical power outage event information, and establish a collaborative degree sub-model for emergency resource collaborative dispatch;

[0147] Establishing constraint conditions for the emergency resource allocation plan based on the amount of emergency resources provided by each emergency rescue point to the disaster-stricken point;

[0148] Based on the emergency resource allocation time sub-model, the emergency resource collaborative scheduling coordination degree sub-model and the constraint conditions, a power grid domain collaborative emergency resource allocation optimization model is constructed.

[0149] Furthermore, the power grid domain collaborative emergency resource allocation optimization model construction module is specifically used to:

[0150] The quantity of each type of emergency resource provided by each emergency rescue point to the disaster-stricken point is less than or equal to the storage quantity of the emergency resource of the emergency rescue point, as the first sub-constraint condition.

[0151] The number of each type of emergency resources provided by all emergency rescue points is greater than or equal to the number of emergency resources of the type required by the disaster-stricken point as the second sub-constraint condition;

[0152] The number of each type of emergency resources provided by each emergency rescue point to the disaster-stricken point is set as a non-negative integer as the third sub-constraint condition;

[0153] Based on the first sub-constraint condition, the second sub-constraint condition, and the third sub-constraint condition, a constraint condition of an emergency resource allocation plan is established.

[0154] Furthermore, the system also includes: a resource demand prediction model building module, which is used to:

[0155] Analyze the historical power outage event information of the disaster-stricken area to obtain historical power outage information and historical actual demand for various emergency supplies;

[0156] A linear regression analysis algorithm is used to establish a relationship between the historical power outage information and the historical actual demand for various emergency supplies;

[0157] Based on the relationship between the historical power outage information and the historical actual demand for the various types of emergency supplies, the resource demand prediction model is constructed using data analysis and prediction software.

[0158] Furthermore, the resource demand prediction model construction module includes:

[0159] An acquisition unit, configured to acquire the number of users without power, the length of damaged power grid lines, the number of damaged cables, and the maximum rainfall based on the historical power outage information;

[0160] The relationship acquisition unit is used to use the number of power outage users, the length of the damaged power grid lines, the number of damaged cables, and the maximum rainfall as independent variables, and the historical actual demand and type of the various emergency materials as dependent variables, and use a linear regression analysis algorithm to establish the relationship between the historical power outage information and the historical actual demand for the various emergency materials.

[0161] Furthermore, the acquisition unit is specifically used to: use a deep learning model to extract the number of power outage users, the length of damaged power grid lines, the number of damaged cables, and the maximum rainfall from the ledgers, records, work daily reports of each emergency rescue point and the historical power outage event information of the disaster-stricken point.

[0162] Furthermore, the emergency resource allocation plan module is specifically used to: based on the demand for various emergency materials at the disaster-stricken point, use the particle swarm algorithm to solve the pre-constructed power grid domain collaborative emergency resource allocation optimization model to obtain the emergency resource allocation plan for the disaster-stricken point.

[0163] Furthermore, the grid domain collaborative emergency resource allocation optimization model satisfies the following formula:

[0164]

[0165] In the above formula: t i represents the emergency resource deployment time from the i-th emergency rescue point to the disaster site, k represents the type of emergency resources, V k represents the consumption rate of the kth emergency resource; i represents the emergency rescue point, n represents the total number of emergency rescue points, S ik represents a variable of 0 or 1, P1 represents the required target emergency resources, Z1 represents the coordinated deployment time of the target emergency resources, It represents the quantity of the kth type of emergency resources provided by the i-th emergency rescue point to the disaster-stricken point.

[0166] The system of the present invention is based on the principles of multivariate regression linear computer algorithm and the like, and can build an intelligent allocation platform. The intelligent allocation platform may include: data collection layer, data transmission layer, data processing layer, intelligent middle platform layer, user interface layer, etc.; the system can summarize and manage the geographical location, inventory, types and quantity of emergency resources, personnel team and other information of the emergency resource library of each emergency resource point, and has a storage and query basis, i.e., an electric power emergency material allocation information library. The functions of each module can be combined to establish a user end, and the system can realize the dynamic update and scientific optimization of the emergency material allocation plan.

[0167] The above-mentioned intelligent recognition platform can set up a data processing center, and the real-time video stream collected by the image acquisition device transmits the accident information to the data processing center through a secure network connection. The data processing center ensures high-speed and secure data transmission and supports large amounts of video information streams. After that, the video stream is first pre-processed (such as format conversion, compression, etc.), and then sent to the power grid domain collaborative emergency resource allocation optimization model for analysis. After that, the particle swarm algorithm, genetic algorithm, etc. can be used to analyze the processed material demand data, and the optimal allocation plan of each resource library can be updated in real time.

[0168] The above-mentioned intelligent identification platform can be provided with a user interface layer: providing an interactive interface for system users, including real-time video monitoring, historical data query, system status display and other functions. Managers can monitor the system operation status and adjust system settings through the user interface layer. An information management module can also be established to summarize and manage the geographical location, inventory, material types and quantities, personnel teams and other information of each emergency resource library, with a storage and query basis, that is, an electric power emergency material allocation information library; the present invention makes good arrangements for the use of limited facilities and emergency materials, so as to better serve customers and achieve greater social benefits.

[0169] Data collection layer: This layer includes multiple high-definition cameras installed in the blackout area to collect real-time video streams. The cameras cover key emergency rescue areas to ensure that the situation of personnel and materials can be captured.

[0170] Data transmission layer: Real-time video streams transmit accident information to the data processing layer through a secure network connection. This layer ensures high-speed and secure data transmission and supports large-volume video streams.

[0171] Data processing layer: The received video stream is first preprocessed (such as format conversion, compression, etc.) and then sent to the deep learning model for analysis. The main task of this layer is to use the constructed regression equation to predict the amount of emergency supplies required.

[0172] Intelligent middle platform layer: The core algorithm layer uses particle swarm algorithms, genetic algorithms, etc. to analyze the processed material demand data and update the optimal allocation plan for each resource library in real time. This layer also includes a learning management module, which is responsible for model training, optimization and updating.

[0173] User interface layer: provides an interactive interface for system users, including real-time video monitoring, historical data query, system status display and other functions. Managers can monitor the system operation status and adjust system settings through this interface.

[0174] The system of the present invention can achieve the following effects:

[0175] (1) The system of the present invention solves the problem of low efficiency caused by relying on manual labor during emergency rescue operations.

[0176] (2) The system has the function of dynamically updating data. It combines the on-site conditions reflected by the monitoring equipment to update the demand for emergency materials in real time, solving the problem of resource waste caused by inaccurate emergency material quantities due to uncertain factors. It enables rapid and accurate response of various emergency materials in emergencies, scientific and reasonable matching of various emergency resource elements, and maximally preventing and reducing emergencies and the losses and impacts they cause.

[0177] (3) It provides enterprises with an efficient and reliable emergency material allocation plan. It can timely and accurately predict the quantity of materials needed in a power outage event and dynamically update the data. By searching the inventory and geographical location of the material warehouse, it can match the optimal material allocation plan according to the different needs of users, thereby improving the efficiency and scientificity of emergency rescue.

[0178] Embodiment 3:

[0179] like Figure 4 As shown, the present invention also provides an electronic device, which may be a computer device, a single-chip device, an intelligent mobile device, etc. The electronic device in this embodiment may include a processor, a memory, a transceiver component, etc. The memory, the processor, and the transceiver component are connected via a bus; the memory may be used to store an execution program, and an exemplary execution program may include instructions; the processor is used to execute the instructions stored in the memory. The memory may also be used to store data, which may be called and / or modified when the instructions are executed.

[0180] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in a storage medium to implement the corresponding method flow or corresponding functions, so as to implement the steps of an emergency resource allocation method based on demand forecasting in the above-mentioned embodiment.

[0181] Embodiment 4:

[0182] Based on the same inventive concept, the present invention also provides a readable storage medium, specifically an electronic device readable storage medium (Memory), which is a memory device in an electronic device for storing programs and data. It can be understood that the storage medium here can include both built-in storage media in electronic devices and, of course, extended storage media supported by electronic devices. The storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by a processor are also stored in the storage space, and these instructions can be one or more execution programs (including program codes). It should be noted that the storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk storage. The processor loads and executes one or more instructions stored in the storage medium, which can implement the steps of an emergency resource allocation method based on demand forecasting in the above embodiment.

[0183] Those skilled in the art will appreciate that the embodiments of the present invention application can be provided as methods, systems, or computer program products. Therefore, the present invention application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present invention application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0184] The present invention application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present invention application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0185] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1A function specified in one or more boxes.

[0186] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0187] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention application rather than to limit its protection scope. Although the present invention application is described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that after reading the present invention application, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the application, but these changes, modifications or equivalent substitutions are all within the protection scope of the claims to be approved.

Claims

1. An emergency resource allocation method based on demand forecasting, characterized in that: include: When a power outage occurs at a disaster site, the demand for various emergency resources at the disaster site is predicted using a pre-built resource demand prediction model based on the current power outage event information at the disaster site; Based on the demand for various emergency materials at the disaster-stricken site, a pre-built power grid domain collaborative emergency resource allocation optimization model is used to allocate emergency resources to obtain an emergency resource allocation plan for the disaster-stricken site; Emergency resource allocation is performed based on the emergency resource allocation plan and the emergency material warehouse information of multiple emergency rescue points.

2. The method according to claim 1, characterized in that When a power outage occurs at the disaster site, the demand for various emergency resources at the disaster site is predicted using a pre-built resource demand prediction model based on the current power outage event information at the disaster site, including: When a power outage occurs at a disaster site, real-time video is collected through multiple image acquisition devices in the power outage area of ​​the disaster site; Identifying changes in various emergency resources at the disaster site in the real-time video; According to the current power outage event information of the disaster-stricken point and the changes of various emergency resources of the disaster-stricken point, the demand of the disaster-stricken point for various emergency resources is predicted by using a pre-built resource demand prediction model.

3. The method according to claim 1, characterized in that The pre-construction process of the power grid domain collaborative emergency resource allocation optimization model includes: Analyze the consumption rate of various emergency resources and the shortest time of dispatching various emergency resources in the historical power outage event information of the disaster-stricken point, and establish an emergency resource deployment time sub-model; Analyze the emergency rescue points involved in emergency resource dispatch in the historical power outage event information, and establish a collaborative degree sub-model for emergency resource collaborative dispatch; Establishing constraint conditions for the emergency resource allocation plan based on the amount of emergency resources provided by each emergency rescue point to the disaster-stricken point; Based on the emergency resource allocation time sub-model, the emergency resource collaborative scheduling coordination degree sub-model and the constraint conditions, a power grid domain collaborative emergency resource allocation optimization model is constructed.

4. The method according to claim 3, characterized in that The constraint conditions for establishing the emergency resource allocation plan based on the amount of emergency resources provided by each emergency rescue point to the disaster-stricken point include: The quantity of each type of emergency resource provided by each emergency rescue point to the disaster-stricken point is less than or equal to the storage quantity of the emergency resource of the emergency rescue point, as the first sub-constraint condition. The number of each type of emergency resources provided by all emergency rescue points is greater than or equal to the number of emergency resources of the type required by the disaster-stricken point as the second sub-constraint condition; The number of each type of emergency resources provided by each emergency rescue point to the disaster-stricken point is set as a non-negative integer as the third sub-constraint condition; Based on the first sub-constraint condition, the second sub-constraint condition, and the third sub-constraint condition, a constraint condition of an emergency resource allocation plan is established.

5. The method according to claim 2, characterized in that: The pre-construction process of the resource demand prediction model includes: Analyze the historical power outage event information of the disaster-stricken area to obtain historical power outage information and historical actual demand for various emergency supplies; A linear regression analysis algorithm is used to establish a relationship between the historical power outage information and the historical actual demand for various emergency supplies; Based on the relationship between the historical power outage information and the historical actual demand for the various types of emergency supplies, the resource demand prediction model is constructed using data analysis and prediction software.

6. The method according to claim 5, characterized in that The linear regression analysis algorithm is used to establish the relationship between the historical power outage information and the historical actual demand for various emergency supplies, including: Based on the historical power outage information, the number of users without power, the length of damaged power grid lines, the number of damaged cables, and the maximum rainfall are obtained; The number of users experiencing power outages, the length of damaged power grid lines, the number of damaged cables, and the maximum rainfall are taken as independent variables, and the actual historical demand and types of various emergency supplies are taken as dependent variables. A linear regression analysis algorithm is used to establish the relationship between the historical power outage information and the actual historical demand for various emergency supplies.

7. The method according to claim 6, characterized in that The obtaining, based on the historical power outage information, the number of power outage users, the length of damaged power grid lines, the number of damaged cables, and the maximum rainfall includes: A deep learning model is used to extract the number of users experiencing power outages, the length of damaged power grid lines, the number of damaged cables, and the maximum rainfall from the ledgers, records, work daily reports of each emergency rescue point, and historical power outage event information of the disaster-stricken point.

8. The method according to claim 1, characterized in that: Based on the demand of the disaster-stricken site for various emergency materials, the emergency resource allocation plan for the disaster-stricken site is obtained by using a pre-built power grid domain collaborative emergency resource allocation optimization model, including: Based on the demand for various emergency supplies at the disaster-stricken point, the pre-constructed power grid domain collaborative emergency resource allocation optimization model is solved using the particle swarm algorithm to obtain the emergency resource allocation plan for the disaster-stricken point.

9. The method according to any one of claims 1 to 8, characterized in that: The grid domain collaborative emergency resource allocation optimization model satisfies the following formula: In the above formula: t i represents the emergency resource deployment time from the ith emergency rescue point to the disaster site, k represents the type of emergency resource, m represents the total number of emergency resource types, V k represents the consumption rate of the kth emergency resource; i represents the emergency rescue point, n represents the total number of emergency rescue points, S ik represents a variable of 0 or 1, P1 represents the required target emergency resources, Z1 represents the coordinated deployment time of the target emergency resources, and x ik It represents the quantity of the kth type of emergency resources provided by the i-th emergency rescue point to the disaster-stricken point.

10. An emergency resource allocation system based on demand forecasting, characterized in that: include: The emergency resource demand prediction module is used to predict the demand of the disaster-stricken point for various emergency resources by using a pre-built resource demand prediction model according to the current power outage event information of the disaster-stricken point when a power outage occurs at the disaster-stricken point; The emergency resource allocation scheme module is used to allocate the emergency resources based on the demand of the disaster-stricken site for various emergency materials by using the pre-built power grid domain collaborative emergency resource allocation optimization model to obtain the emergency resource allocation scheme for the disaster-stricken site; The emergency resource allocation module is used to allocate emergency resources based on the emergency resource allocation plan and the emergency material warehouse information of multiple emergency rescue points.

Images