Emergency material management method for subway stations based on dynamic optimization model of material allocation

Abstract

The present invention relates to the technical field of subway station emergency management, and discloses a subway station emergency material management method based on a dynamic optimization model for material allocation. The method comprises the following steps: constructing a static database of emergency events through historical data of emergency events, establishing a corresponding relationship between the types of emergency events, the types of emergency materials, and the quantities of emergency materials, and generating a basic inventory; combining principal component analysis with a neural network model to screen core prediction factors, collecting their real-time monitoring data, and constructing a dynamic database; establishing a dynamic optimization model for material allocation, calculating the matching coefficients between potential emergency events and historical cases, generating a total potential inventory, mapping the basic inventory to the total potential inventory item by item and selecting the maximum value to generate a dynamic inventory; and finally constructing a digital management warehouse for emergency materials to achieve full life cycle management of procurement, warehousing, use, maintenance, and scrapping. The present invention improves the scientificity and accuracy of emergency material allocation and enhances the subway station's ability to respond to emergencies.

Description

Technical Field

[0001] The present invention relates to the technical field of subway station emergency management, and in particular to a subway station emergency material management method based on a dynamic optimization model of material configuration. Background Art

[0002] As a vital component of urban transportation systems, subways significantly improve travel convenience, expand travel radius, and enhance urban operational efficiency, becoming the main artery of urban transportation in China. However, as enclosed underground transportation hubs, subway stations pose significant challenges in the event of an emergency, leading to paralyzed subway operations and significant passenger delays. The rational allocation of emergency supplies can significantly improve the efficiency of emergency response, ensuring smooth rescue efforts while minimizing the losses and harm caused by emergencies. Therefore, the scientific and efficient management of emergency supplies at subway stations is crucial for ensuring the continuity of subway system operations, protecting people's lives and property, and maintaining social stability.

[0003] However, as the subway industry's requirements for emergency response capabilities and speed increase, and as the subway industry's intelligent and digital transformation becomes increasingly evident, traditional subway station emergency material management methods are increasingly unable to meet actual needs and industry requirements:

[0004] (1) The structural contradiction between static configuration and dynamic risk;

[0005] Traditional subway station emergency material allocation is usually planned in advance according to universal standards, which makes it difficult to adapt to the diversity, randomness, and uncertainty of subway emergency events. This leads to a systematic mismatch between the static mode of emergency material allocation and the dynamically changing risks during operation (such as bad weather, passenger flow fluctuations, equipment failures, etc.).

[0006] (2) The imbalance between supply and demand of emergency supplies;

[0007] The types and quantities of emergency supplies deployed at subway stations are all implemented according to preset plans, without considering the differentiated configuration required by certain stations due to differences in passenger flow density or the special surrounding environment (such as flooding prone to nearby waters). As a result, supplies at low-risk stations are idle for long periods of time until they become ineffective, causing waste and losses, while high-risk stations may face problems such as supply shortages due to sudden demands, resulting in reduced rescue efficiency.

[0008] (3) The stubborn problem of information silos in traditional management methods;

[0009] Due to the low degree of digitization of the current emergency material management methods in subway stations, a large amount of data records rely on manually registered paper ledgers. In the event of an emergency, the low efficiency of information transmission through paper documents can easily lead to delayed or misjudgment. At the same time, data information such as material inventory, equipment systems, and monitoring sensors are fragmented and disconnected from each other, resulting in inefficient collection, sharing, and utilization of information related to material configuration, thereby weakening emergency response capabilities.

[0010] It can be seen that there are currently many unresolved problems in the emergency management of subway stations.

[0011] Therefore, it is urgent to design an emergency material management method for subway stations based on a dynamic optimization model of material allocation, digitally improve the traditional subway station emergency material management method, comprehensively consider the historical data and real-time monitoring data of various emergency events, and drive the dynamic optimization and precise adaptation of emergency resource allocation through data fusion, providing solid guarantees for the safe operation of the subway system. Summary of the Invention

[0012] In order to solve the above technical problems, the present invention provides a subway station emergency material management method based on a dynamic optimization model of material configuration, which significantly improves the response efficiency of subway stations to emergencies, and at the same time reduces redundant inventory and lowers operating costs through an extreme value selection algorithm.

[0013] The present invention provides a subway station emergency material management method based on a dynamic optimization model for material allocation, comprising the following steps:

[0014] Step S1, obtaining historical data of emergency events at multiple subway stations in the same area and building a static database of emergency events;

[0015] Step S2: Based on the industry emergency material allocation standards, for subway stations of the same category, a correspondence between emergency event type, emergency material type, and emergency material quantity is established to generate a basic inventory;

[0016] Step S3, determining a set of prediction factors for various emergency events based on the emergency event static database;

[0017] Step S4, collecting real-time monitoring data of the prediction factor set and building a dynamic database of emergency events;

[0018] Step S5: establishing a dynamic optimization model for material allocation, calculating the matching coefficients between potential emergency events and historical cases based on the real-time monitoring data of the emergency event dynamic database, and generating a total potential inventory;

[0019] Step S6, mapping the basic inventory and the total potential inventory item by item and selecting the maximum value to generate a dynamic inventory;

[0020] Step S7: Based on the static database of emergency events, the dynamic inventory and the dynamic optimization model of material allocation, a digital management warehouse for emergency materials is constructed to achieve full life cycle management of emergency materials.

[0021] Furthermore, in step S1, the historical data of emergency events at the subway station specifically includes: the category of the subway station, the type of emergency event, the time of occurrence of the emergency event, the type and quantity of emergency materials used, the triggering factors of the emergency event, the development and evolution process of the emergency event, the termination conditions of the emergency event, environmental related factors and impact loss data.

[0022] Furthermore, the prediction factor set of each type of emergency event is determined based on the emergency event static database, specifically including:

[0023] Step S31: for each type of emergency event, extract event triggering elements, environmental related factors and emergency response data of historical cases from the emergency event static database;

[0024] Step S32: performing a multi-dimensional spatiotemporal correlation analysis on the event triggering factors, environmental related factors, and emergency response data to construct a correlation matrix including equipment status parameters, environmental monitoring time series data, and emergency response;

[0025] Step S33, using principal component analysis to calculate the importance of each candidate parameter, and selecting parameters whose importance exceeds a preset threshold as core prediction factors;

[0026] Step S34: verifying the prediction accuracy of the core prediction factors for the occurrence of emergency events through a neural network model, and forming a final prediction factor set when the accuracy reaches a preset standard.

[0027] Furthermore, the step S5 specifically includes:

[0028] In step S51, based on the real-time monitoring data of the emergency event dynamic database, the matching coefficient between potential emergency events and historical cases is calculated; in step S52, a hierarchical response strategy is executed according to the preset matching coefficient threshold range, including three modes: material configuration adjustment, early warning monitoring, and maintaining the original material configuration; in step S53, according to the hierarchical response strategy, the potential inventory of each emergency event category is aggregated to generate the total potential inventory.

[0029] Furthermore, the step S52 specifically includes:

[0030] When the matching coefficient is in the range of 60%-100%, adjust the material configuration, including: adopt the material demand of the corresponding historical case; when the matching coefficient is in the range of 40%-60%, early warning monitoring is carried out, including: no adjustment for the time being, but marking it as an event that requires dynamic monitoring; when the matching coefficient is lower than 40%, maintain the original material configuration, including: keeping the basic inventory level unchanged.

[0031] Furthermore, the matching coefficient is calculated as follows:

[0032] Step S511: normalize the historical data of prediction factors of various emergency events in the emergency event static database, map the prediction factor values ​​to the [0,1] interval through the range normalization formula, and construct a standardized prediction factor matrix ;

[0033] Step S512: For a certain type of emergency event in the same type of subway station in the emergency event static database, calculate the convergence coefficient C of each prediction factor. l , the specific calculation formula is:

[0034] ;

[0035] Among them, C l represents the convergence coefficient of the lth prediction factor; m represents that there are m historical cases of a certain type of emergency event at the same type of subway station in the static database of emergency events; k represents the kth historical case; represents the lth predictor value of the kth historical case after standardization;

[0036] Step S513: Calculate the weight of each prediction factor based on the convergence coefficient , the specific calculation formula is:

[0037] ;

[0038] in, Indicates the The weight of the predictor; l represents the lth predictor; n represents the total number of predictors;

[0039] Step S514: construct a one-dimensional matrix of prediction factors of potential emergency events based on the real-time monitoring data of prediction factors of similar emergency events in the emergency event dynamic database. , calculate the weighted Euclidean distance D between it and each historical case k , the specific formula is:

[0040] ;

[0041] Among them, D krepresents the weighted Euclidean distance between the predictor of the potential emergency event and the k-th historical case; represents the lth predictor value of the potential emergency event; represents the lth predictor value of the kth historical case;

[0042] Step S515: Calculate the matching coefficient S using the weighted Euclidean distance. The specific calculation formula is:

[0043] ;

[0044] Where S represents the matching coefficient between potential emergency events and historical cases.

[0045] Furthermore, the range standardization formula is:

[0046] ;

[0047] in, represents the lth predictor value of the kth historical case after standardization; f ml represents the lth predictor value of the mth historical case; f kl It represents the lth predictor value of the kth historical case, where min and max are the minimum and maximum values ​​of the predictor column, respectively.

[0048] Furthermore, the generation of the total potential inventory volume adopts an extreme value selection algorithm, specifically: performing a union operation on the material demands of multiple categories of potential emergency events; and retaining the maximum demand for the same type of materials in each event scenario.

[0049] Furthermore, in step S7, building a digital management warehouse for emergency supplies includes: step S71, configuring emergency supplies in actual subway stations according to dynamic inventory quantities; step S72, establishing a full life cycle information ledger including procurement, warehousing, use, maintenance and scrapping; step S73, real-time monitoring of emergency supplies data and triggering at least two warning mechanisms among inventory warning, overdue warning and quality alarm.

[0050] The embodiments of the present invention have the following technical effects:

[0051] 1. By combining the characteristics of emergency events, the multivariate real-time monitoring data is matched and mapped with historical cases in the static database of each emergency event to generate dynamic inventory of materials. On the basis of ensuring sufficient reserves of emergency materials, the material demand under different types of emergency events is quantified, realizing the transformation of material allocation from "rigid design" to "precise data driven", and significantly optimizing the accuracy of emergency material allocation.

[0052] 2. By establishing a dynamic optimization model for material allocation, real-time monitoring data such as the status data of various equipment in the station, sensor data, passenger flow density data, meteorological data, and earthquake data are collected. Based on the matching results with historical data, early warning of emergencies is provided to assist in the rapid adjustment of emergency material allocation. This solves the problem of insufficient adaptation of traditional preset material modes to dynamic risk scenarios, and comprehensively improves the subway station's ability to respond to dynamic risks.

[0053] 3. Through dynamic risk assessment of emergency events, over-allocation of materials is avoided and inventory redundancy is reduced; full life cycle management of materials, through functions such as overdue warning, effectively reduces the loss rate of expired materials; improves emergency response capabilities, reduces losses caused by sudden emergency events, and effectively reduces the operating costs of subway stations.

[0054] 4. By establishing a digital archive of the entire life cycle of emergency materials, a complete material data chain is formed, eliminating the fragmentation problem of traditional paper records; cross-system multi-source data fusion analysis is carried out, and multi-dimensional data correlation relationships are constructed, realizing a comprehensive digital upgrade of emergency material management from information fragmentation and isolated operation to data collaboration. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0056] Figure 1 This is a flow chart of a subway station emergency material management method based on a dynamic optimization model for material allocation provided by an embodiment of the present invention;

[0057] Figure 2 This is a module diagram of a subway station emergency material management system based on a dynamic optimization model for material allocation provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0058] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention are described clearly and completely below. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort are also within the scope of protection of the present invention.

[0059] Figure 1This is a flow chart of a subway station emergency material management method based on a dynamic optimization model for material allocation provided by an embodiment of the present invention. Figure 1 , specifically including:

[0060] Step S1: Obtain historical data of emergency events at multiple subway stations in the same area and build a static database of emergency events.

[0061] In some embodiments, in step S1, the historical data of emergency events at the subway station specifically includes: the category of the subway station, the type of emergency event, the time of occurrence of the emergency event, the type and quantity of emergency materials used, the triggering factors of the emergency event, the development and evolution process of the emergency event, the termination conditions of the emergency event, environmental related factors and impact loss data.

[0062] For example, historical data on subway station emergencies can be collected from internal records of the subway operating company, duty logs and surveillance footage at each station, historical records from the emergency response platform or dispatch center, statistical data released by the government and industry regulatory agencies, and other relevant sources (such as passenger complaints and news reports). For each emergency event, the following detailed information should be collected: the type of subway station (for example, underground station, ground station, elevated station, etc.); the type of emergency event (such as fire, flood, earthquake, equipment failure, etc.); the time of occurrence (timestamp accurate to the minute); the type and quantity of emergency supplies used (including but not limited to fire extinguishers, first aid kits, sandbags, etc.); the triggering factors of the emergency event (such as electrical failure, natural disasters, and human sabotage); the development process (a detailed description of how the event developed from the beginning to the end); the termination conditions (the criteria for determining whether the event is resolved); environmental and related factors (such as weather conditions and surrounding environmental characteristics) that may affect the development of the event; and the impact and losses (including casualties, economic losses, and the duration of operational interruption).

[0063] Standardize the collected data to ensure consistency and accuracy. Missing or incomplete information can be supplemented through supplementary surveys or reasonable estimation methods. Simultaneously, this information is entered into a structured database to form a preliminary static database of emergency events.

[0064] Step S2: Based on the industry emergency material allocation standards, for subway stations of the same category, a correspondence between emergency event types, emergency material types, and emergency material quantities is established to generate a basic inventory.

[0065] Specifically, the industry emergency material allocation standards include but are not limited to the industry standard JT / T 1409-2022 "Basic Requirements for the Construction of Emergency Capacity for Urban Rail Transit Operations", etc. Taking this standard as an example, the basic inventory volume established , i, j ≥ 1; subway stations of the same category include underground stations, ground stations, elevated stations, etc., and the basic inventory R B In, t B,i is the set of emergency event types, v B,j is a collection of emergency supplies types, q B,ij The quantity of the jth type of emergency supplies under the i-th type of emergency event. Basic inventory Some of the fragments are:

[0066] .

[0067] Step S3: determining a set of prediction factors for various emergency events based on the emergency event static database.

[0068] In some embodiments, based on the static database of emergency events, a set of prediction factors for each type of emergency event is determined, specifically including:

[0069] Step S31: for each type of emergency event, extract event triggering elements, environmental related factors and emergency response data of historical cases from the emergency event static database;

[0070] Step S32: performing a multi-dimensional spatiotemporal correlation analysis on the event triggering factors, environmental related factors, and emergency response data to construct a correlation matrix including equipment status parameters, environmental monitoring time series data, and emergency response;

[0071] Step S33, using principal component analysis to calculate the importance of each candidate parameter, and selecting parameters whose importance exceeds a preset threshold as core prediction factors;

[0072] Step S34: verifying the prediction accuracy of the core prediction factors for the occurrence of emergency events through a neural network model, and forming a final prediction factor set when the accuracy reaches a preset standard.

[0073] Specifically, for each type of emergency event, the following data are extracted from the static database: event triggers, such as electrical failures and natural disasters; environmental factors, such as weather conditions and surrounding environmental characteristics; and emergency response data, including the use and quantity of emergency supplies and the implementation of emergency measures. These data serve as the foundation for constructing predictive factors. The extracted data is analyzed through the following steps: Equipment status parameters, which collect status parameters of subway station facilities related to the event, such as power system load and drainage pump efficiency; Environmental monitoring time series data, such as rainfall, wind speed, and temperature changes, are collected and organized from local meteorological and earthquake monitoring agencies; and Emergency Response, which records and analyzes time series data on the various response measures taken after the emergency event. A correlation matrix is ​​then constructed, which aims to reflect the interactions between different parameters in both temporal and spatial dimensions.

[0074] Principal component analysis (PCA) is used to assess the importance of each candidate parameter and select the most representative parameters as core predictors. First, the data for all candidate parameters are standardized to a mean of 0 and a variance of 1 to eliminate the effects of different scales. The original data are then transformed into a new coordinate system, where each new coordinate axis corresponds to a principal component. The proportion of variance explained by each principal component is calculated, and those principal components whose cumulative explained variance reaches a preset threshold (e.g., 80%) are retained. Based on these retained principal components, a reverse mapping is performed back to the original variable space to identify the original variables with higher contributions as core predictors.

[0075] Build or select an appropriate neural network model and train it using existing historical data. Input the selected core predictive factors into the model, run the model, and analyze the output. The effectiveness of these factors is determined based on the model's prediction accuracy. Typically, a preset benchmark (e.g., an accuracy rate exceeding 80%) is set. Once this benchmark is met, these factors are confirmed as the final set of predictive factors.

[0076] For example, let's assume we're dealing with an emergency event involving flooding at a subway station. We extract historical flood disaster cases from the past few years from a static database, including triggering factors (such as heavy rain), environmental factors (such as ground water height), and emergency response data (such as the number of sandbags). We then perform a multi-dimensional spatiotemporal correlation analysis on this data, taking into account factors such as the working status of the subway station's internal drainage system and external real-time rainfall. We use principal component analysis (PCA) to calculate the weights of each factor and select the most important ones as core predictors. Finally, we use a trained neural network model to verify the effectiveness of these predictors, ensuring they can accurately predict possible future flood disasters.

[0077] Based on the historical data of emergency events in the static database, according to the different characteristics of various emergency events, the occurrence patterns, influencing factors and change trends of emergency events are summarized, so as to determine the prediction factor set of various emergency events. , the set of predictors Some excerpts are as follows.

[0078]

[0079] Step S4: collecting real-time monitoring data of the prediction factor set and building a dynamic database of emergency events.

[0080] In this embodiment, the real-time monitoring data of various prediction factors comes from local meteorological departments, local earthquake monitoring departments, local water conservancy departments, water level gauges in stations, cameras in stations, various sensors in stations, and control systems of various equipment in stations.

[0081] Step S5: establishing a dynamic optimization model for material allocation, calculating the matching coefficients between potential emergency events and historical cases based on the real-time monitoring data of the emergency event dynamic database, and generating the total potential inventory.

[0082] In some embodiments, step S5 specifically includes:

[0083] Step S51, calculating the matching coefficient between potential emergency events and historical cases based on the real-time monitoring data of the emergency event dynamic database;

[0084] Furthermore, the matching coefficient is calculated as follows:

[0085] Step S511: normalize the historical data of prediction factors of various emergency events in the emergency event static database, map the prediction factor values ​​to the [0,1] interval through the range normalization formula, and construct a standardized prediction factor matrix ;

[0086] In this embodiment, the range normalization formula is:

[0087] ;

[0088] in, represents the lth predictor value of the kth historical case after standardization; f ml represents the lth predictor value of the mth historical case; f kl It represents the lth predictor value of the kth historical case, where min and max are the minimum and maximum values ​​of the predictor column, respectively.

[0089] Step S512: For a certain type of emergency event in the same type of subway station in the emergency event static database, calculate the convergence coefficient C of each prediction factor. l , the specific calculation formula is:

[0090] ;

[0091] Among them, C l represents the convergence coefficient of the lth prediction factor; m represents that there are m historical cases of a certain type of emergency event at the same type of subway station in the static database of emergency events; k represents the kth historical case; represents the lth predictor value of the kth historical case after standardization;

[0092] Step S513: Calculate the weight of each prediction factor based on the convergence coefficient , the specific calculation formula is:

[0093] ;

[0094] in, Indicates the The weight of the predictor; l represents the lth predictor; n represents the total number of predictors;

[0095] Step S514: construct a one-dimensional matrix of prediction factors of potential emergency events based on the real-time monitoring data of prediction factors of similar emergency events in the emergency event dynamic database. , calculate the weighted Euclidean distance D between it and each historical case k , the specific formula is:

[0096] ;

[0097] Among them, D k represents the weighted Euclidean distance between the predictor of the potential emergency event and the k-th historical case; represents the lth predictor value of the potential emergency event; represents the lth predictor value of the kth historical case;

[0098] Step S515: Calculate the matching coefficient S using the weighted Euclidean distance. The specific calculation formula is:

[0099] ;

[0100] Where S represents the matching coefficient between potential emergency events and historical cases. Step S52: Execute a hierarchical response strategy based on the preset matching coefficient threshold interval, including three modes: material allocation adjustment, early warning monitoring, and maintaining the original material allocation;

[0101] Optionally, step S52 specifically includes:

[0102] When the matching coefficient is between 60% and 100%, the material allocation is adjusted, including adopting the material requirements of the corresponding historical case. When the matching coefficient is between 40% and 60%, early warning monitoring is implemented, including temporarily not adjusting but marking the event as requiring dynamic monitoring. When the matching coefficient is below 40%, the original material allocation is maintained, including keeping the base inventory unchanged. Step S53: Based on the hierarchical response strategy, the potential inventory of each emergency event category is aggregated to generate the total potential inventory.

[0103] In some embodiments, the total potential inventory is generated using an extreme value selection algorithm, specifically: performing a union operation on the material demands of multiple categories of potential emergency events; and retaining the maximum demand for the same type of materials in each event scenario.

[0104] For example, taking the underground station flood emergency as an example, the matching coefficient The calculation process is:

[0105] For underground station flood emergency events, there are = 5 prediction factors, namely {maximum rainfall per hour, 24-hour rainfall, water depth at the entrance and exit, water depth inside the station, and drainage pump load rate}. =10 historical cases, construct a prediction factor matrix based on historical data ,but

[0106] ;

[0107] The predicted factor values ​​are mapped to the [0,1] interval through the range standardization formula to construct a standardized predicted factor matrix ;

[0108] ;

[0109] The data volatility of each predictor after standardization is judged and the The convergence coefficient C of the standardized predictors l ,

[0110] ;

[0111] Calculate the The weights of the standardized predictors ,

[0112] ;

[0113] get , , , , .

[0114] Based on the real-time monitoring data of the prediction factors of similar emergency events (underground station flood control emergency events) in the emergency event dynamic database, a one-dimensional matrix of prediction factors of potential emergency events is constructed.

[0115] , calculate the weighted Euclidean distance D between it and each historical case k ;

[0116] ;

[0117] We get D1=9.29, D2=3.03, D3=4.17, D4=0.57, D5=13.35, D6=17.01, D7=4.76, D8=2.08, D9=27.53, D 10 =33.30.

[0118] By calculating the matching coefficients using weighted Euclidean distance, we get S1=9.72%, S2=24.84%, S3=19.32%, S4=63.72%, S5=6.97%, S6=5.55%, S7=17.36%, S8=32.48%, S9=3.50%, S 10 =2.92%.

[0119] For underground station flood emergency events, the matching coefficient The largest value is in the fourth case, 60%< =63.72%<100%, then the material demand situation of the fourth case is used as the potential inventory of the underground station flood prevention emergency event.

[0120] Calculate the potential inventory of each category of emergency events in turn , the total potential inventory generated , some fragments are as follows:

[0121] .

[0122] Step S6: Map the basic inventory and the total potential inventory item by item and select the maximum value to generate a dynamic inventory.

[0123] Specifically, if the total potential inventory has been generated , the basic inventory Total potential inventory Map each item one by one, select the element-level maximum value, and use the result as the dynamic inventory , dynamic inventory Some snippets are shown below:

[0124] .

[0125] Step S7: Based on the static database of emergency events, the dynamic database of emergency events and the dynamic optimization model of material allocation, a digital management warehouse for emergency materials is constructed to realize the full life cycle management of emergency materials.

[0126] In some embodiments, in step S7, building a digital management warehouse for emergency supplies includes: step S71, allocating emergency supplies at actual subway stations according to dynamic inventory; that is, according to dynamic inventory , determine the types and quantities of emergency supplies, and promptly make corresponding arrangements in actual subway stations.

[0127] Step S72: Establish a full life cycle information ledger covering procurement, warehousing, use, maintenance, and scrapping;

[0128] Specifically, dynamic tracking is carried out on emergency materials from procurement, warehousing, use, inventory, maintenance to scrapping, and the entire life cycle information ledger of emergency materials is recorded in the background database. The ledger mainly includes but is not limited to information such as the type, specifications, quantity, functional use, quality, spatial distribution, service life, and previous usage of emergency materials.

[0129] Step S73: monitor emergency supplies data in real time and trigger at least two warning mechanisms among inventory warning, overdue warning and quality warning.

[0130] Real-time monitoring of emergency material data is carried out, while multi-dimensional statistics and intelligent analysis are carried out, and early warnings and alarms are automatically issued, including inventory warnings, overdue warnings, maintenance expiration warnings, environmental quality warnings, etc.

[0131] The present invention achieves accurate mapping of emergency material demand and historical cases by constructing a static database and a dynamic database of emergency events, combining predictive factor analysis, matching coefficient calculation and hierarchical response strategy, and transforming the configuration of emergency materials from "static setting" to "dynamic adjustment". It also generates dynamic inventory covering multiple event scenarios by combining an extreme value selection algorithm. Based on the intelligent matching of real-time monitoring data and historical cases, it reduces material configuration errors, breaks through the traditional static configuration mode, and improves the scientificity and accuracy of emergency material configuration; through a hierarchical early warning mechanism and full-process data drive, it improves the response speed of emergencies, that is, the risk coverage rate; using a redundancy elimination algorithm and full-life cycle digital management, it effectively avoids the problem of material redundancy or shortage; and also realizes the full life cycle management of emergency materials through a digital management warehouse, enhancing the subway station's response capability to emergencies and resource utilization efficiency, with significant safety benefits and management value.

[0132] The embodiment of the present invention also provides a subway station emergency material management system based on a dynamic optimization model for material allocation, see Figure 2 , including the following modules:

[0133] Static database construction module, used to obtain historical data of emergency events at multiple subway stations in the same area and build a static database of emergency events;

[0134] The basic inventory generation module is used to establish a correspondence between emergency event types, emergency material types, and emergency material quantities for subway stations of the same category based on industry emergency material allocation standards, and generate basic inventory quantities;

[0135] A prediction factor set construction module, configured to determine a prediction factor set for each type of emergency event based on the emergency event static database;

[0136] A dynamic database construction module is used to collect real-time monitoring data of the prediction factor set and build a dynamic database of emergency events;

[0137] A material allocation dynamic optimization model establishment module is used to establish a material allocation dynamic optimization model, calculate the matching coefficient between potential emergency events and historical cases based on the real-time monitoring data of the emergency event dynamic database, and generate the total potential inventory;

[0138] A dynamic inventory generation module is used to map the basic inventory and the total potential inventory item by item and select the maximum value to generate a dynamic inventory;

[0139] The emergency material digital management warehouse construction module is used to build an emergency material digital management warehouse based on the emergency event static database, the emergency event dynamic database and the material configuration dynamic optimization model to realize the full life cycle management of emergency materials.

[0140] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. A subway station emergency material management method based on a dynamic optimization model of material allocation, characterized in that: The steps include: Step S1, obtaining historical data of emergency events at multiple subway stations in the same area and building a static database of emergency events; Step S2: Based on the industry emergency material allocation standards, for subway stations of the same category, a correspondence between emergency event type, emergency material type, and emergency material quantity is established to generate a basic inventory; Step S3, based on the static database of emergency events, determines a set of prediction factors for various emergency events; specifically, it includes: Step S31: for each type of emergency event, extract event triggering elements, environmental related factors and emergency response data of historical cases from the emergency event static database; Step S32: performing a multi-dimensional spatiotemporal correlation analysis on the event triggering factors, environmental related factors, and emergency response data to construct a correlation matrix including equipment status parameters, environmental monitoring time series data, and emergency response; Step S33, using principal component analysis to calculate the importance of each candidate parameter, and selecting parameters whose importance exceeds a preset threshold as core prediction factors; Step S34, verifying the prediction accuracy of the core prediction factors for the occurrence of emergency events through a neural network model, and forming a final prediction factor set when the accuracy reaches a preset standard; Step S4, collecting real-time monitoring data of the prediction factor set and building a dynamic database of emergency events; Step S5: establishing a dynamic optimization model for material allocation, calculating the matching coefficients between potential emergency events and historical cases based on the real-time monitoring data of the emergency event dynamic database, and generating a total potential inventory; Step S6, mapping the basic inventory and the total potential inventory item by item and selecting the maximum value to generate a dynamic inventory; Step S7: constructing a digital management warehouse for emergency supplies based on the static database of emergency events, the dynamic inventory and the dynamic optimization model for supply configuration.

2. The subway station emergency material management method based on the material allocation dynamic optimization model according to claim 1 is characterized in that: In step S1, the historical data of emergency events at the subway station specifically includes: the category of the subway station, the type of emergency event, the time of occurrence of the emergency event, the type and quantity of emergency materials used, the triggering factors of the emergency event, the development and evolution process of the emergency event, the termination conditions of the emergency event, environmental related factors and impact loss data.

3. The subway station emergency material management method based on the material allocation dynamic optimization model according to claim 1 is characterized in that: The step S5 specifically includes: In step S51, based on the real-time monitoring data of the emergency event dynamic database, the matching coefficient between potential emergency events and historical cases is calculated; in step S52, a hierarchical response strategy is executed according to the preset matching coefficient threshold range, including three modes: material configuration adjustment, early warning monitoring, and maintaining the original configuration; in step S53, according to the hierarchical response strategy, the potential inventory of each emergency event category is aggregated to generate the total potential inventory.

4. The subway station emergency material management method based on the material allocation dynamic optimization model according to claim 3 is characterized in that: The step S52 specifically includes: When the matching coefficient is in the range of 60%-100%, adjust the material configuration, including: adopt the material demand of the corresponding historical case; when the matching coefficient is in the range of 40%-60%, early warning monitoring is carried out, including: no adjustment for the time being, but marking it as an event that requires dynamic monitoring; when the matching coefficient is lower than 40%, maintain the original material configuration, including: keeping the basic inventory level unchanged.

5. The subway station emergency material management method based on the material allocation dynamic optimization model according to claim 3 is characterized in that: The matching coefficient is calculated as follows: Step S511: normalize the historical data of prediction factors of various emergency events in the emergency event static database, map the prediction factor values ​​to the [0,1] interval through the range normalization formula, and construct a standardized prediction factor matrix ; Step S512: For a certain type of emergency event at the same type of subway station in the emergency event static database, calculate the convergence coefficient of each prediction factor. C l , the specific calculation formula is: ； in, C l Indicates the l The convergence coefficient of the prediction factors; m means that there are m historical cases of a certain type of emergency event at the same type of subway station in the static database of emergency events; k means the kth historical case; represents the lth predictor value of the kth historical case after standardization; Step S513: Calculate the weight of each prediction factor based on the convergence coefficient , the specific calculation formula is: ； in, Indicates the The weight of the predictor; l represents the lth predictor; n represents the total number of predictors; Step S514: construct a one-dimensional matrix of prediction factors of potential emergency events based on the real-time monitoring data of prediction factors of similar emergency events in the emergency event dynamic database. , calculate the weighted Euclidean distance D between it and each historical case k , the specific formula is: ； Among them, D k represents the weighted Euclidean distance between the predictor of the potential emergency event and the k-th historical case; represents the lth predictor value of the potential emergency event; represents the lth predictor value of the kth historical case; Step S515: Calculate the matching coefficient S using the weighted Euclidean distance. The specific calculation formula is: ； Where S represents the matching coefficient between potential emergency events and historical cases.

6. The subway station emergency material management method based on the material allocation dynamic optimization model according to claim 5 is characterized in that: The range standardization formula is: ； in, represents the lth predictor value of the kth historical case after standardization; f ml represents the lth predictor value of the mth historical case; f kl It represents the lth predictor value of the kth historical case, where min and max are the minimum and maximum values ​​of the predictor column, respectively.

7. The subway station emergency material management method based on the material allocation dynamic optimization model according to claim 3 is characterized in that: The total potential inventory is generated using an extreme value selection algorithm, specifically: performing a union operation on the material demands of multiple categories of potential emergency events; and retaining the maximum demand for the same type of materials in each event scenario.

8. The subway station emergency material management method based on the material allocation dynamic optimization model according to claim 1 is characterized in that: In step S7, building a digital management warehouse for emergency supplies includes: step S71, configuring emergency supplies in actual subway stations based on dynamic inventory; step S72, establishing a full life cycle information ledger including procurement, warehousing, use, maintenance and scrapping; step S73, real-time monitoring of emergency supplies data and triggering at least two warning mechanisms among inventory warning, overdue warning and quality warning.

Images