An emergency material management method and system based on big data analysis

Abstract

This invention provides an emergency supplies management method and system based on big data analysis, relating to the field of emergency supplies management technology. The method includes: acquiring multi-source real-time data from the Internet of Things (IoT); establishing a unified virtual inventory model by constructing an IoT sensing layer and a transmission layer; inputting the regional emergency resource situation feature vector and historical disaster data from the virtual inventory model into a trained three-layer BP neural network demand prediction model, outputting the demand rate of various emergency supplies in the future period; classifying emergency supplies according to their characteristics, and using corresponding inventory optimization models to optimize differentiated inventory strategies for different categories of supplies in the classification results; constructing an emergency supplies scheduling model with the goal of minimizing total delivery time; and solving the emergency supplies scheduling model using a heuristic algorithm based on the principle of minimum fill rate priority, generating emergency supplies management decisions.

Description

Technical Field

[0001] This invention relates to the field of emergency supplies management technology, and in particular to an emergency supplies management method and system based on big data analysis. Background Technology

[0002] Against the backdrop of frequent public emergencies, the accurate storage and efficient distribution of emergency supplies are core elements for improving emergency rescue capabilities. The rapid development of big data technology has provided technical support for the intelligent upgrading of emergency supplies management. Building a scientific emergency supplies management system has become an important research direction in the field of emergency management. It can realize accurate prediction of material demand, reasonable control of inventory, and efficient scheduling of distribution, providing solid material support for emergency response to various emergencies.

[0003] Currently, the field of emergency supplies management has gradually introduced technologies such as neural network prediction models and quantitative ordering models. Some solutions can achieve preliminary prediction of material demand and control of basic inventory. At the same time, simple distribution scheduling can be carried out by combining traffic network information. The application of such technologies has promoted the transformation of emergency supplies management from traditional manual decision-making to data-driven decision-making, laying a technical foundation for improving the timeliness and accuracy of emergency supplies support, and has important practical significance for improving the emergency rescue system.

[0004] However, existing technologies have not yet formed an integrated closed loop of "demand forecasting - differentiated inventory optimization - delivery scheduling". There are problems such as inventory strategies not being customized according to material categories, incomplete dimensions of delivery scheduling constraints, and a disconnect between demand forecasting and actual delivery decisions, making it difficult to adapt to the diverse material support needs in emergency scenarios. Summary of the Invention

[0005] In view of the shortcomings of the prior art, the purpose of this invention is to provide an emergency supplies management method based on big data analysis, which can solve the technical problems of existing technologies such as inventory strategies not being customized according to material categories, incomplete dimensions of distribution scheduling constraints, and disconnect between demand forecasting and actual distribution decisions.

[0006] A first aspect of this invention proposes an emergency supplies management method based on big data analysis, comprising:

[0007] S1: Acquire real-time data from multiple IoT sources;

[0008] S2: Based on multi-source real-time data from the Internet of Things (IoT), a unified virtual inventory model is established by constructing an IoT sensing layer and a transmission layer.

[0009] S3: Input the regional emergency resource status feature vector and historical disaster data from the virtual inventory model into the trained three-layer BP neural network demand prediction model, and output the demand rate of various emergency supplies in the future period.

[0010] S4: Classify emergency supplies according to their characteristics, and use the corresponding inventory optimization model to optimize the differentiated inventory strategy for different categories of supplies in the classification results;

[0011] S5: Based on the optimization results of demand rate, virtual inventory model and differentiated inventory strategy, set supply capacity constraints, road traffic constraints, transportation vehicle volume constraints and weight constraints, and construct an emergency material dispatching model with the goal of minimizing total delivery time.

[0012] S6: Based on the principle of prioritizing minimum fill rate, an emergency material dispatching model is solved using a heuristic algorithm to generate emergency material management decisions.

[0013] A second aspect of this invention provides an emergency supplies management system based on big data analysis, comprising: a processor and a memory;

[0014] The memory stores programs or instructions that can run on the processor, which, when executed by the processor, implement the steps of the emergency supplies management method based on big data analysis as described in the first aspect.

[0015] A third aspect of the present invention provides a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the emergency supplies management method based on big data analysis as described in the first aspect.

[0016] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0017] In this embodiment of the invention, a three-layer BP neural network demand forecasting model is used to uniformly output the demand rate of various materials and the specific demand quantity at each demand point, breaking down the dimensional barriers in demand forecasting. Simultaneously, differentiated inventory optimization models are constructed for three categories of emergency materials: important, scarce, and time-sensitive, enabling customized replenishment to compensate for the shortcomings of homogeneous inventory strategies. Furthermore, a multi-dimensional constrained emergency material scheduling model is built and solved based on the minimum fill rate priority principle, forming a complete closed loop from demand forecasting to distribution decisions, thereby comprehensively improving the intelligence and precision of emergency material management. Attached Figure Description

[0018] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts. Obviously, the drawings described below are merely some embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings without any creative effort.

[0019] Figure 1 This is a flowchart illustrating an emergency supplies management method based on big data analysis provided in an embodiment of the present invention.

[0020] Figure 2 This is a schematic diagram of the structure of an emergency supplies management system based on big data analysis provided in an embodiment of the present invention. Detailed Implementation

[0021] To enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. It should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0022] The emergency supplies management method based on big data analysis provided by the present invention will be described in detail below with reference to the accompanying drawings, through specific embodiments and application scenarios.

[0023] Reference manual attached Figure 1 The diagram illustrates a flowchart of an emergency supplies management method based on big data analysis provided by an embodiment of the present invention.

[0024] This invention provides an emergency supplies management method based on big data analysis, which may include the following steps:

[0025] S1: Acquire real-time data from multiple IoT sources.

[0026] Among them, IoT multi-source real-time data refers to various real-time monitoring data related to emergency material management, such as available inventory at supply points, supply status, road status matrix, and collection of transportation vehicles.

[0027] S2: Based on multi-source real-time data from the Internet of Things (IoT), a unified virtual inventory model is established by constructing an IoT sensing layer and a transmission layer.

[0028] The Internet of Things (IoT) sensing layer refers to the equipment layer, including various sensors and data collection terminals, deployed at supply points, demand points, and transportation networks. The IoT transmission layer refers to the communication network layer that enables data transmission from the sensing layer to the data platform. The virtual inventory model is a unified data model that integrates information such as inventory status at various supply points, road accessibility, and transportation availability to visualize the overall regional emergency material reserve capacity.

[0029] Specifically, when establishing a virtual inventory model, the available inventory and supply availability status of each supply point are recorded in the virtual inventory platform. At the same time, the road status matrix and transportation tool set are collected simultaneously to reflect network accessibility and transportation capacity. Furthermore, emergency supplies are classified based on historical disaster data and expert rules.

[0030] Furthermore, the virtual inventory model will pre-match corresponding inventory optimization models for different categories of materials, achieving an initial connection between data collection and subsequent inventory strategies.

[0031] In this embodiment of the invention, this step enables real-time network monitoring of emergency material-related nodes within the region. Compared with the traditional model that only focuses on the inventory of a single storage node, it can dynamically perceive the overall emergency material reserve capacity of the region, greatly improve the inventory visibility rate, and provide accurate data support for subsequent inventory optimization and distribution scheduling.

[0032] S3: Input the regional emergency resource status feature vector and historical disaster data from the virtual inventory model into the trained three-layer BP neural network demand prediction model, and output the demand rate of various emergency supplies in the future cycle.

[0033] Among them, the regional emergency resource status feature vector refers to a set of features encompassing multi-dimensional disaster and resource status, including disaster type, intensity, affected population, and seasonal climate. The three-layer BP neural network demand forecasting model is a deep learning forecasting model consisting of an input layer, hidden layers, and an output layer, iteratively optimized through a backpropagation algorithm. Demand rate refers to the expected consumption intensity of various emergency supplies within a future unit of time.

[0034] Specifically, this three-layer BP neural network demand forecasting model completes the nonlinear mapping of features and outputs the demand forecast value through the activation functions of the hidden layer and the output layer, and then combines the time dimension parameter of the future period to obtain the final demand rate.

[0035] Optionally, various emergency supplies include essential supplies, scarce supplies, and time-sensitive supplies.

[0036] Based on the differences in characteristics of various emergency supplies, an inventory optimization model is constructed.

[0037] Among them, the inventory optimization models include the inventory optimization model for important materials, the inventory optimization model for scarce materials, and the inventory optimization model for time-sensitive materials.

[0038] Among them, critical materials refer to those with high priority in emergency response, large demand, and significant impact on rescue effectiveness. Scarce materials refer to those with limited supply sources or long acquisition cycles. Time-sensitive materials refer to those with a clear expiration date, are easily perishable, or whose demand changes rapidly over time. Differentiated inventory strategies refer to using appropriate mathematical models for replenishment and inventory control based on material categories. It should be noted that classification modeling achieves precise matching of inventory strategies, improving the adaptability of overall inventory management and the reliability of material support.

[0039] Optionally, the inventory strategy optimization process of the critical materials inventory optimization model specifically includes:

[0040] Based on the demand rate of key materials output by a three-layer BP neural network demand forecasting model, an objective function is constructed to minimize the total inventory cost per unit time.

[0041]

[0042] in, Q represents the total inventory cost per unit time for essential goods. i C represents the order quantity of the i-th important material. 1i Let C represent the unit storage cost of the i-th type of important material. 3i This represents the cost of each order for the i-th type of essential material. This represents the predicted demand rate for the i-th important commodity.

[0043] Among them, total inventory cost per unit time refers to the sum of material storage cost and ordering cost within a unit period. Forecasted demand rate refers to the projected consumption intensity of key materials per unit period, output by a three-layer BP neural network.

[0044] Specifically, the objective function integrates the two core costs of all types of important materials and uses mathematical modeling to clarify the optimization direction of inventory costs.

[0045] It should be noted that this objective function can anchor the direction of inventory cost optimization and provide a quantitative basis for replenishment decisions of important materials.

[0046] Based on the actual conditions of the warehousing facilities, by clearly defining the upper limit of the total storage capacity and combining the unit storage capacity occupancy of each important material, storage capacity constraints are set:

[0047]

[0048] Where, ω i Let W represent the unit storage capacity occupancy of the i-th type of important material, W represent the total storage capacity limit, and n represent the total number of types of important materials.

[0049] Unit storage capacity refers to the storage space occupied by a unit quantity of important materials. Total storage capacity limit refers to the maximum amount of space that storage facilities can use to store important materials.

[0050] Specifically, this constraint will limit the total storage capacity of all important materials to no more than the upper limit, while ensuring that the order quantity of various materials is not negative, thus conforming to the actual carrying capacity of the warehouse.

[0051] It should be noted that this constraint can prevent order quantities from exceeding the actual capacity of the warehouse, ensuring the feasibility of the inventory plan.

[0052] Based on the economic order quantity without inventory capacity constraints, calculate the theoretical order quantity for each important material:

[0053]

[0054] in, This represents the economic order quantity for the i-th type of important material when there are no inventory constraints.

[0055] The theoretical order quantity refers to the replenishment quantity calculated solely based on cost optimization, without considering warehouse capacity limitations.

[0056] It should be noted that this formula can provide a basic reference value for the optimal cost under no storage capacity constraints, and set a benchmark for subsequent batch adjustments under storage capacity constraints.

[0057] Specifically, the calculation of this batch will be entirely focused on minimizing the total inventory cost per unit time, making it an ideal reference value for inventory cost control.

[0058] For example, if the ordering cost of an important material is high and the storage cost is low, its theoretical order quantity will be relatively large in order to reduce the frequency of ordering.

[0059] It should be noted that calculating the theoretical order quantity first can provide a basic reference value for optimal cost, and set a benchmark for subsequent batch adjustments under inventory capacity constraints.

[0060] Based on the theoretical order quantity and unit storage capacity, the theoretical total storage capacity is compared with the upper limit of total storage capacity to determine whether the storage capacity limit is exceeded. If the storage capacity limit is not exceeded, the theoretical order quantity is used as the final order quantity; otherwise, an inventory cost optimization function under storage capacity constraints is constructed to calculate the final order quantity of each important material under storage capacity constraints.

[0061]

[0062] in, Let λ represent the optimal order quantity for the i-th important material after considering the Lagrange multiplier, and let λ represent the constraint penalty coefficient, which is used to characterize how much the total inventory cost of the system can be reduced when the warehouse capacity increases by 1 unit. This allows the inventory holding amount to be converted into cost, so as to ensure the consistency of dimensions.

[0063] Among them, the theoretical total warehouse capacity refers to the sum of the warehouse capacity occupied by the theoretical order batches of all important materials, and the Lagrange multiplier refers to the auxiliary parameters used to solve constrained optimization problems.

[0064] Specifically, when the theoretical total warehouse capacity exceeds the upper limit, an optimization function is constructed and Lagrange multipliers are introduced to balance the cost and warehouse capacity requirements, and the feasible optimal order quantity is obtained.

[0065] It should be noted that this formula can achieve a balance between cost optimization and storage capacity constraints, taking into account both the economic efficiency of inventorying important materials and the utilization rate of storage space.

[0066] Based on the final order quantity and the demand rate of key materials, calculate the order cycle and inventory utilization rate for each key material:

[0067]

[0068]

[0069] Among them, T i This indicates the ordering cycle for the i-th important material. Let Kr represent the final optimal order quantity for the i-th type of important material, and Kr represent the warehouse capacity utilization rate.

[0070] Among them, the ordering cycle refers to the time interval between two replenishments, and the warehouse capacity utilization rate refers to the ratio of the actual warehouse capacity occupied by important materials to the total warehouse capacity.

[0071] It should be noted that calculating the ordering cycle of each important material can clarify the replenishment schedule for these materials, providing a time-based basis for inventory management decisions. Calculating the warehouse capacity utilization rate of each important material can quantify the utilization status of storage space, providing a core indicator for subsequent evaluation and optimization of inventory strategies.

[0072] Specifically, the order cycle calculation can clarify the replenishment rhythm of important materials, while the warehouse capacity utilization rate can quantify the utilization status of warehouse space.

[0073] Furthermore, these two indicators will serve as the basis for evaluating inventory strategies and provide data support for subsequent parameter recalibration.

[0074] For example, if the cost of each order for a certain important material is high, but the unit storage cost is low, the theoretical order quantity calculated using the above formula will be relatively large to reduce the frequency of orders and lower the total ordering cost. If the unit storage capacity of this material is large and the total storage capacity is tight, the final order quantity after adjusting for storage capacity constraints will be appropriately reduced to ensure that the total storage capacity does not exceed the limit.

[0075] For example, an emergency warehouse has two types of essential supplies with a total capacity of W = 1000 m³. Drinking water (i = 1): projected demand rate D1 = 500 cases / month, unit storage cost C 11 =2 yuan / (box·month), cost per order C 31 =100 yuan, unit storage capacity ω1=0.5m³ / box; Convenience food (i=2): Forecasted demand rate D2=300 boxes / month, unit storage cost C 12 =1.5 yuan / (box·month), cost per order C 32 =80 yuan, unit warehouse capacity occupancy ω2=0.8m³ / box. The theoretical order quantity Q1⁽ without warehouse capacity constraints is calculated. 0 =224 boxes, Q2⁽ 0 =179 boxes, theoretical total warehouse capacity occupancy 254.9m³≤1000m³, final order quantity is the theoretical value, ordering cycle T1=13 days, T2=18 days, warehouse capacity utilization rate Kr=25.5%.

[0076] It should be noted that this step can quantify the replenishment pace and warehouse utilization status of critical materials, providing core indicators for subsequent evaluation of inventory strategies.

[0077] Optionally, the inventory strategy optimization process of the scarce resource inventory optimization model specifically includes:

[0078] Based on the demand rate of scarce resources output by the three-layer BP neural network demand forecasting model, and combined with the single order cost and unit annual storage cost of scarce resources, the economic ordering cycle of scarce resources is calculated:

[0079]

[0080] Where T represents the economic ordering cycle of the scarce resource, S represents the single ordering cost of the scarce resource, C0 represents the unit annual storage cost of the scarce resource, and R represents the average demand rate of the scarce resource.

[0081] Among them, the economic ordering cycle refers to the optimal replenishment interval that balances ordering costs and storage costs, and the average demand rate refers to the average consumption intensity of scarce materials per unit time within a year.

[0082] It should be noted that the calculation of the economic ordering cycle for this scarce resource will balance the ordering frequency and storage duration to avoid increasing costs due to frequent ordering or wasting resources due to long-term storage.

[0083] It should be noted that this economic ordering cycle can balance the replenishment and storage costs of scarce materials, and reduce the overall losses in inventory management of scarce materials.

[0084] Calculate the maximum inventory level of scarce resources based on the economic order cycle and the preset safety stock level:

[0085]

[0086] Among them, Q max T represents the maximum inventory of scarce resources. k Indicates the lead time for ordering, Q s This indicates the safety stock level of scarce resources.

[0087] Among them, the order lead time refers to the time span from initiating an order to the goods entering the warehouse, and the safety stock refers to the minimum reserve reserved to cope with demand fluctuations.

[0088] It should be noted that those skilled in the art can set the size of the preset safety stock according to actual needs, and this invention does not limit this.

[0089] Specifically, the calculation of maximum inventory levels covers projected demand during the economic ordering cycle and order lead time, while also incorporating safety stock to address uncertainties.

[0090] Furthermore, setting a maximum inventory level can prevent the problem of idle resources caused by excessive stockpiling of scarce materials.

[0091] It should be noted that calculating the maximum inventory level can clearly define the upper limit of scarce resources and avoid resource waste caused by excessive stockpiling.

[0092] Based on the maximum inventory level and real-time inventory status of scarce resources, calculate the actual order quantity for the scarce resources:

[0093]

[0094] Among them, Qx i Q represents the actual order quantity of the i-th type of scarce material. ni Q represents the amount of scarce goods in transit. ki Q represents the current inventory of scarce resources. mi This indicates the quantity of scarce goods awaiting shipment.

[0095] Among them, "in transit" refers to the quantity of goods that have been ordered but not yet put into storage, and "pending shipment" refers to the quantity of goods that have been allocated but not yet shipped.

[0096] Specifically, the batch calculation will fully consider various real-time inventory statuses to ensure that the replenishment quantity accurately matches the actual inventory gap.

[0097] For example, consider a scarce rescue equipment: a single order cost S = 5000 yuan, an annual unit storage cost C0 = 1200 yuan / unit, an average demand rate R = 10 units / year, and an order lead time T. k =2 months, safety stock Q s =3 units. The calculated economic order cycle T = 0.91 months, and the maximum inventory level Q. max =14 units. If the current inventory Q... k ᵢ=2 units, in-transit quantity Q n =4 units, quantity to be shipped Q m ᵢ=1 unit, actual order quantity Qᵢ=14-4-2+1=9 units.

[0098] It should be noted that calculating order quantities based on real-time inventory status can ensure the accuracy of replenishment quantities and avoid shortages or stockpiling of scarce materials.

[0099] Optionally, the inventory strategy optimization process of the time-sensitive goods inventory optimization model specifically includes:

[0100] Based on the demand rate and lead time of time-sensitive goods output by the three-layer BP neural network demand forecasting model, the lead time demand of time-sensitive goods is calculated:

[0101]

[0102] Where D1 represents the lead time demand for time-sensitive goods, R p This indicates the demand rate for time-sensitive goods.

[0103] Among them, the lead time demand refers to the total expected consumption of time-sensitive materials within the order lead time.

[0104] Specifically, the calculation of this demand will accurately predict the material shortage during the lead time, providing a core basis for setting the replenishment trigger timing.

[0105] It should be noted that the calculation of lead time demand can accurately predict the material shortage during the replenishment cycle, providing a core basis for setting reorder points.

[0106] Determine the reorder point for time-sensitive goods based on their lead time demand:

[0107]

[0108] Among them, Q k This indicates the ordering point for time-sensitive goods.

[0109] The reorder point refers to the inventory threshold that triggers the replenishment process.

[0110] Specifically, by linking the ordering point to the lead time demand, it can be ensured that after the replenishment is initiated, the materials can be put into storage within the lead time, connecting with subsequent demand consumption.

[0111] Furthermore, the establishment of this reorder point will also reserve space for the subsequent accumulation of safety stock, thereby improving supply stability.

[0112] It should be noted that linking the reorder point to the lead time demand can ensure the rationality of the replenishment trigger timing and avoid shortages of time-sensitive materials during the lead time.

[0113] Based on the cost per order, unit storage cost, and demand rate of time-sensitive goods, the economic order quantity for replenishment after the reorder point is calculated using a quantitative ordering model.

[0114]

[0115] Among them, Q * C1 represents the optimal economic order quantity for time-sensitive goods, and C1 represents the unit storage cost of time-sensitive goods.

[0116] Among them, the fixed ordering model refers to an inventory model that uses a fixed replenishment quantity and triggers replenishment based on the reorder point, while the optimal economic order quantity refers to the best replenishment quantity that balances ordering costs and storage costs.

[0117] It should be noted that the calculation of the economic order quantity for replenishment after the reorder point is triggered will balance the ordering frequency and storage duration of time-sensitive materials, reducing expiration losses and storage costs.

[0118] For example, for time-sensitive goods such as pharmaceuticals with short shelf lives, the economic order quantity will be relatively small to reduce the risk of expiration due to long-term storage.

[0119] It should be noted that this economic order quantity can ensure supply while reducing the spoilage and storage costs of time-sensitive goods.

[0120] In one possible implementation, S3 specifically includes sub-steps S301 to S305:

[0121] S301: Based on the virtual inventory model and historical disaster data, identify the key factors affecting the demand for emergency supplies and construct a multi-dimensional feature input vector.

[0122] Among them, the multidimensional feature input vector refers to the integration of information from multiple dimensions such as disaster type, intensity, affected population, season, and climate conditions into a structured data set, which serves as the input to the model.

[0123] Specifically, this step involves extracting all factors strongly correlated with material demand from regional resource status data and historical disaster data in the virtual inventory model, and then integrating them into a standardized input vector.

[0124] It should be noted that this step systematically integrates the complex factors affecting demand, providing a comprehensive and structured information foundation for subsequent accurate forecasting.

[0125] S302: Set the number of neurons in the input layer, hidden layer, and output layer of the three-layer BP neural network demand prediction model.

[0126] Among them, the setting of the number of neurons refers to determining the number of nodes in each layer of the neural network based on the dimension of input features, the total number of material types, and the model complexity requirements, so as to build a suitable network architecture.

[0127] Specifically, the number of input layer nodes corresponds to the feature dimension of the multi-dimensional feature input vector, the number of output layer nodes corresponds to the total number of types of emergency supplies, and the number of hidden layer nodes is reasonably set according to the model's generalization requirements.

[0128] Furthermore, the number of neurons directly affects the model's feature fitting ability and computational efficiency, and both accuracy and feasibility must be considered.

[0129] It should be noted that by setting the network structure appropriately, overfitting can be avoided while ensuring model capacity, thus laying a structural foundation for efficient training and reliable prediction.

[0130] S303: Input the multi-dimensional feature input vector into the pre-defined three-layer BP neural network demand prediction model, and calculate the output values ​​of the hidden layer neurons:

[0131]

[0132] Among them, h j Let f(j) represent the output value of the j-th hidden layer neuron, f() represent the Sigmoid activation function, and w pj x represents the connection weight from the p-th feature in the input layer to the j-th neuron in the hidden layer. p Let b represent the p-th input feature. j This represents the bias term of the j-th neuron in the hidden layer.

[0133] Among them, the output value of hidden layer neurons refers to the intermediate result generated in the hidden layer after the input information is weighted, summed and activated, and is used to extract nonlinear combinations and higher-order patterns of input features.

[0134] Specifically, this step involves summing the feature values ​​of the input vector with their corresponding connection weights, adding a bias term, and then processing the sum with an activation function to obtain the output of the hidden layer.

[0135] It should be noted that this formula introduces a nonlinear transformation through an activation function, enabling the model to learn and express the deep correlation between complex disaster situations and material needs.

[0136] S304: Based on the output values ​​of the hidden layer neurons, calculate the predicted emergency material demand value corresponding to the output layer neurons:

[0137]

[0138] in, Let g represent the predicted demand value for the k-th type of emergency supplies, g() represent the output layer activation function, and v jk c represents the connection weight from the j-th neuron in the hidden layer to the k-th neuron in the output layer. k This represents the bias term of the k-th neuron in the output layer.

[0139] Among them, the demand prediction value corresponding to the output layer neurons refers to the preliminary demand estimate of various materials obtained after the hidden layer output is weighted and activated twice.

[0140] Specifically, this step involves weighting and summing the output values ​​of each neuron in the hidden layer with their corresponding connection weights, adding the output layer bias term, and then processing the sum with an activation function to obtain preliminary demand predictions for various materials.

[0141] For example, if a linear activation function is used, a predicted value that closely matches the actual demand scale can be directly output, which facilitates the formulation of subsequent inventory strategies.

[0142] It should be noted that the function of this formula is to transform the multi-dimensional disaster features and resource status information extracted from the hidden layer into a predicted value of the k-th type of emergency material demand with clear business meaning through weighted fusion, bias correction, and activation function mapping. This achieves a crucial transition from "feature representation" to "quantitative demand output," enabling the model to accurately characterize the nonlinear relationship between complex disaster factors and emergency material demand, and providing direct and usable demand input for subsequent inventory optimization and distribution scheduling.

[0143] S305: Based on the output demand forecasts for various emergency supplies, and by combining the time dimension parameters of the future period, output the demand rate of various emergency supplies in the future period.

[0144] The demand rate refers to the quantity of goods required per unit time (such as daily or weekly) obtained by allocating the predicted total demand over a specific future time period.

[0145] Specifically, this step involves breaking down static demand forecasts into dynamic cyclical demand rates based on future time dimensions, thus adapting them to the cyclical decision-making needs of inventory optimization and delivery scheduling.

[0146] Furthermore, the output demand rate will also be provided to each demand point simultaneously, providing a basis for the micro-level allocation of delivery volume.

[0147] It should be noted that this step transforms static forecasts into dynamic demand sequences that change over time, providing a time-precise basis for inventory replenishment and delivery scheduling decisions.

[0148] In this embodiment of the invention, this step can effectively capture the nonlinear impact of multiple factors such as disaster intensity and population density on material demand. Compared with traditional expert judgment or static average prediction methods, its prediction error rate is greatly reduced, significantly improving the accuracy and generalization ability of demand forecasting, and providing a scientific demand basis for subsequent inventory strategy formulation.

[0149] S4: Classify emergency supplies according to their characteristics, and use the corresponding inventory optimization model to optimize the differentiated inventory strategy for different categories of supplies in the classification results.

[0150] Among these, material characteristics refer to the attributes of emergency supplies in dimensions such as importance, scarcity, and timeliness. Differentiated inventory strategy optimization refers to inventory control methods that adapt different inventory models to different categories of materials.

[0151] For example, for essential goods, the focus is on balancing inventory costs and warehouse capacity utilization, while for time-sensitive goods, the risk of stockouts and spoilage due to expiration are taken into account.

[0152] In this embodiment of the invention, this step can avoid the problems of imbalance between storage capacity utilization and cost, and loss of time-sensitive materials caused by traditional unified inventory models. It can achieve both optimal storage capacity and cost for important materials, achieve periodic and accurate replenishment for scarce materials, and reduce the risk of stockouts and expiration for time-sensitive materials, thus greatly improving the flexibility and scientific nature of inventory management.

[0153] S5: Based on the optimization results of demand rate, virtual inventory model and differentiated inventory strategy, set supply capacity constraints, road traffic constraints, transportation vehicle volume constraints and weight constraints, and construct an emergency material dispatch model with the goal of minimizing total delivery time.

[0154] The minimum total delivery time refers to minimizing the total time of the entire process of emergency supplies from the supply point to the demand point, including travel time, warehousing time, loading and unloading time. The emergency supplies dispatch model is a delivery decision-making model that considers multiple dimensions such as comprehensive supply capacity, demand guarantee rate, road traffic flow, and vehicle loading capacity.

[0155] Specifically, the emergency supplies dispatch model will set various constraints, such as supply capacity, demand guarantee, road traffic flow, and vehicle loading, around the goal of minimum total delivery time, to ensure the model's practicality and compliance.

[0156] In one possible implementation, S5 specifically includes sub-steps S501 to S506:

[0157] S501: Based on the predicted demand rate, virtual inventory model, and optimization results of differentiated inventory strategies, construct an objective function with minimum total delivery time as the core.

[0158] The objective function for minimum total delivery time is to minimize the total time (including travel, outbound, and loading / unloading time) spent on material delivery from all supply points to all demand points as the mathematical optimization objective.

[0159] Specifically, the construction of this objective function integrates the time characteristics of demand rate, the node distribution characteristics of virtual inventory, and the replenishment results of inventory strategy, anchoring the core timeliness requirements in emergency scenarios.

[0160] It should be noted that this step establishes the core direction of scheduling optimization, aiming to improve the overall efficiency of emergency response from a time perspective.

[0161] S502: Based on the objective function, and combining the actual inventory and supply capacity of the supply points in the virtual inventory model, set supply capacity constraints:

[0162]

[0163] Among them, X ijk P represents the quantity of type k goods distributed from the i-th supply point to the j-th demand point. i S represents the supply availability status of the i-th supply point. ik Let m represent the inventory of the k-th type of material at the i-th supply point, m represent the total number of demand points, and z represent the total number of material categories.

[0164] Among them, supply capacity constraints refer to ensuring that the total amount of various materials transferred from any supply point does not exceed the product of its current available inventory and supply status, so as to prevent over-allocation.

[0165] Specifically, this supply capacity constraint formula limits the total amount of various materials that each supply point can distribute to all demand points, ensuring that it does not exceed its available inventory and the supply capacity corresponding to its supply status, thus avoiding over-capacity distribution.

[0166] Furthermore, this supply capacity constraint formula can ensure the feasibility of distribution on the supply side and prevent scheduling plans from failing due to insufficient inventory.

[0167] It should be noted that this supply capacity constraint formula ensures the feasibility of the scheduling plan, guarantees that decisions are based on actual material availability, and avoids invalid or unexecutable scheduling instructions.

[0168] S503: Combining the demand forecasting model output by a three-layer BP neural network, the system predicts the demand quantity at each demand point and sets the forecasted demand and minimum guarantee rate constraints for various materials based on preset intervals for the delivery quantity at each demand point.

[0169]

[0170] Among them, e k This represents the minimum guarantee rate for the k-th type of supplies. This represents the predicted demand for the k-th type of material at the j-th demand point.

[0171] Among them, the forecast demand and minimum guarantee rate constraint means that the distribution volume of any material at any demand point is required to be no less than the product of the forecast demand and the preset minimum guarantee rate, and no more than the forecast demand itself.

[0172] It should be noted that those skilled in the art can set the size of the preset interval according to actual needs, and this invention does not limit this.

[0173] Specifically, the forecasted demand and minimum guarantee rate constraints for these various types of supplies will define the upper and lower limits of the delivery volume at the demand points. The lower limit ensures the supply of basic relief materials, while the upper limit avoids waste of materials caused by over-delivery.

[0174] It should be noted that the forecasted demand and minimum availability constraints for these various materials have achieved a balance between meeting basic relief needs and avoiding resource waste, prioritizing the minimum supply level of critical materials.

[0175] S504: Based on the traffic capacity of the road network, road flow constraints are set by limiting the total amount of goods delivered on a single route.

[0176]

[0177] Among them, U ij Let represent the maximum capacity of the road from the i-th supply point to the j-th demand point, and n represent the total number of demand points.

[0178] Among them, road flow constraint refers to the restriction that the total amount of all goods delivered from a certain supply point to a certain demand point shall not exceed the maximum traffic capacity of that road segment.

[0179] Specifically, this road traffic constraint will limit the total delivery volume of a single route, avoid road congestion caused by excessive concentration of goods transportation, and ensure the smooth flow of the transportation chain.

[0180] For example, for rural roads with limited traffic capacity, the maximum road capacity will be set at a lower value to prevent congestion and delays in delivery.

[0181] It should be noted that this road flow constraint incorporates road network traffic restrictions into the model, making the scheduling scheme conform to actual traffic conditions and improving the feasibility of the scheme.

[0182] S505: Based on the actual loading capacity of the transport vehicle, set volume and weight constraints for the transport vehicle:

[0183]

[0184]

[0185] Among them, V k W represents the unit volume of the k-th type of material. k J represents the unit weight of the k-th type of material. ih CV represents the number of vehicles of type h owned by the i-th supply point. h CW represents the unit volumetric cargo capacity of the h-th class of transport vehicles. h represents the unit load capacity of the h-th type of transport vehicle, and l represents the total number of transport vehicle categories.

[0186] Among them, the volume and weight constraints of transport vehicles refer to the fact that the total volume and total weight of all goods shipped from a certain supply point shall not exceed the total volumetric cargo capacity and total load capacity of the available transport vehicles, respectively.

[0187] Specifically, these two types of constraints will limit the total delivery volume of the supply points from the dimensions of volume and weight, respectively, to match the actual loading capacity of the transportation vehicles and eliminate transportation risks such as overloading.

[0188] Furthermore, these two types of constraints can ensure the safety and compliance of material transportation, and avoid transportation delays or safety accidents caused by overloading.

[0189] It should be noted that this dual constraint ensures that the scheduling scheme complies with the physical loading limitations of the transportation vehicles, which is a key link in achieving refined logistics management.

[0190] S506: Based on the full-dimensional constraint settings of objective function, supply capacity constraint, forecasted demand and minimum guarantee rate constraint, road traffic constraint, vehicle volume constraint and weight constraint, the overall construction of the scheduling model with minimum total delivery time as the objective is completed.

[0191] Among them, the overall construction of the scheduling model refers to integrating the objective function and all real-world constraints into a complete mathematical programming problem, forming a solvable optimization model.

[0192] Specifically, this step integrates the objective function with various constraints to form a logically closed-loop scheduling model, providing a complete optimization framework for subsequent decision-making.

[0193] It's important to note that, firstly, the introduction of the objective function centered on "minimum total delivery time" in S501 directly quantifies the most critical "timeliness" indicator in emergency response. This shifts the focus of scheduling decisions from solely cost or distance to the core objective of rescue response speed, providing a unified optimization direction for all subsequent constraints. Secondly, the supply capacity constraint formula in S502 effectively avoids issues like "over-stocking" or "false supply capacity" in theoretical scheduling schemes by limiting the total amount delivered by each supply point to the product of its available inventory and supply status. This ensures consistency between the model's solution and the actual inventory status, guaranteeing the feasibility of the plan from the outset. Thirdly, the predicted demand and minimum guarantee rate constraints in S503 directly embed the predicted demand output from the BP neural network into the scheduling model, allowing demand prediction results to truly participate in decision-making calculations. This approach balances fairness and efficiency by both ensuring basic rescue needs at each demand point through lower bound constraints and preventing resource waste through upper bound constraints. Fourth, the road flow constraint proposed in S504 incorporates the traffic network's carrying capacity into the model, limiting the total delivery volume on a single supply-demand path to the constraints of real-world traffic conditions. This avoids scheduling bottlenecks on certain key road sections, thus reducing the risk of delivery delays due to road congestion, landslides, or insufficient capacity. Fifth, S505 sets loading constraints for transport vehicles from both volume and weight dimensions. Compared to traditional models that only consider a single loading indicator, this is closer to real-world logistics scenarios. Its advantage lies in simultaneously preventing unreasonable loading situations such as "volume overload but weight within limits" or "weight overload but volume not full," improving transportation safety and compliance.

[0194] In this embodiment of the invention, compared with the traditional scheduling model that only targets transportation cost or route length, this step can take into account real-world constraints such as outbound efficiency, road congestion, and loading restrictions, ensuring the practicality and accuracy of the scheduling model. It provides an optimized framework that fits emergency scenarios for subsequent delivery decisions and improves the overall response efficiency of the delivery process.

[0195] S6: Based on the principle of prioritizing minimum fill rate, an emergency material dispatching model is solved using a heuristic algorithm to generate emergency material management decisions.

[0196] The minimum fill rate priority principle refers to a scheduling priority rule that prioritizes ensuring the basic fill rate of various materials at each demand point. A heuristic algorithm is an algorithm that finds an approximate optimal solution to a complex scheduling optimization problem within an acceptable timeframe. Emergency material management decisions refer to the final execution plan, which includes the material distribution volume, distribution routes, and transportation vehicle allocation from each supply point to the demand point.

[0197] Specifically, the solution process first sorts materials by minimum fill rate and completes basic demand scheduling, then sorts excess demand by unit time efficiency and optimizes allocation, while verifying time window constraints and adjusting decisions.

[0198] Furthermore, the generated management decisions will also provide execution data support for subsequent adaptive system updates.

[0199] In one possible implementation, S6 specifically includes sub-steps S601 to S604:

[0200] S601: Based on the principle of prioritizing minimum filling rate, various emergency supplies are sorted according to preset sorting rules.

[0201] Among them, the preset sorting rule refers to determining the priority of material scheduling and processing according to the preset minimum fill rate of various materials from high to low.

[0202] Specifically, this step will sort the various supplies from highest to lowest according to their minimum availability rate, and clarify the priority order for dispatching core relief supplies.

[0203] It should be noted that this step clarifies the logical order of scheduling, ensuring that the most critical and urgent material needs are met first when resources are limited.

[0204] S602: Based on the ranking results of various emergency supplies, prioritize the allocation of supplies at each demand point to meet the minimum fill rate.

[0205] Among them, the material demand to meet the minimum fill rate refers to the minimum guaranteed quantity that each demand point must achieve for each type of material, which is the product of the predicted demand and the minimum guarantee rate.

[0206] Specifically, this step will prioritize the allocation of resources to ensure that the supply of core relief supplies at each point of need reaches the minimum fill rate standard, so as to avoid delays in critical relief efforts due to shortages of supplies.

[0207] Furthermore, the scheduling results of this step will serve as the basis for subsequent optimization of excess demand, ensuring the bottom line of rescue and relief support.

[0208] It should be noted that this step first ensures that the basic material needs of all points of need are met, providing the most basic resource guarantee for the rescue operation.

[0209] S603: Based on the remaining excess material demand in the scheduling results, sort them according to preset indicator rules, and complete the distribution optimization through heuristic algorithms.

[0210] Among them, the remaining excess material demand refers to the demand that exceeds the minimum fill rate and is not met at each demand point after all minimum guarantee requirements have been met.

[0211] Specifically, this step first sorts the excess demand by the rescue efficiency per unit time, and then uses a heuristic algorithm to complete the optimal allocation of the remaining materials, maximizing the utilization efficiency of the overall rescue resources.

[0212] For example, for areas with a large affected population and severe disaster, excess demand will be given higher priority and will receive material replenishment first.

[0213] It should be noted that this step, while ensuring basic needs are met, optimizes the allocation of remaining resources to maximize overall rescue effectiveness.

[0214] S604: Based on the upper limit of the preset material delivery time window for each demand point, verify the actual arrival time, and adjust the scheduling decision and delivery strategy according to the verification result until an emergency material management decision that meets all constraints is generated.

[0215] Among them, the time window upper limit verification refers to determining whether the actual arrival time of materials calculated based on the current scheduling plan exceeds the latest time limit that each demand point can accept.

[0216] It should be noted that those skilled in the art can set the upper limit of the preset material delivery time window according to actual needs, and this invention does not limit it.

[0217] Specifically, this step verifies the arrival time of materials at each demand point, and adjusts the route or loading strategy for schemes that exceed the upper limit of the time window until all constraints are met.

[0218] For example, in a flood relief effort, the minimum fill rate for essential supplies such as food and medicine is set at 0.9, while the minimum fill rate for scarce supplies such as tents is set at 0.8. Priority is given to meeting the basic needs of each demand point. For demand points in townships with large affected populations and severe disasters, excess demand exceeding the minimum fill rate will be given higher priority, and remaining supplies will be allocated first. If the actual arrival time of supplies to a remote demand point exceeds the upper limit of the preset time window, the transportation route will be adjusted, choosing a detour through unobstructed sections of road to ensure that time constraints are met.

[0219] For example, after an earthquake, there are 2 supply points (A with 100 tents in stock, B with 90 tents in stock) and 3 demand points (A predicts a demand of 50 tents, B predicts a demand of 80 tents, and C predicts a demand of 60 tents). The minimum tent availability rate is e. k =0.8. First, ensure the basic demand of A (40 units), B (64 units), and C (48 units). The remaining 38 units are allocated according to the disaster-affected population weight: 26 units for B and 12 units for C. A uses 3 trucks (unit volumetric load capacity CV). h =30 units) were delivered to A, 40 units to B, and 64 units to C. B used two trucks to deliver 60 units to C. The arrival time of all materials did not exceed the preset time window of 4 hours.

[0220] It should be noted that this step, through iterative adjustments, ensures that the final scheduling scheme meets both quantity and resource constraints as well as critical time requirements, thereby generating a fully feasible final decision.

[0221] In this embodiment of the invention, this step can solve the problems of high computational complexity and difficulty in implementation of traditional scheduling models. At the same time, by prioritizing the minimum fill rate, it ensures the supply of basic materials to key demand points, which greatly improves the material arrival time satisfaction rate of each demand point, forming a complete decision-making closed loop from demand forecasting to delivery execution, thus ensuring the effectiveness and fairness of emergency rescue.

[0222] Reference manual attached Figure 2 The diagram shows a structural schematic of an emergency supplies management system based on big data analysis provided by an embodiment of the present invention.

[0223] This invention provides an emergency supplies management system 20 based on big data analysis, comprising: a processor 201 and a memory 202;

[0224] The memory 202 stores programs or instructions that can run on the processor 201. When the program or instructions are executed by the processor 201, they implement the steps of the above-mentioned emergency material management method based on big data analysis and achieve the same technical effect. To avoid repetition, the present invention will not elaborate further.

[0225] It should be understood that the processor 201 in this embodiment of the invention may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0226] It should also be understood that the memory 202 in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus RAM (DR RAM).

[0227] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0228] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0229] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0230] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0231] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0232] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0233] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0234] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0235] This invention provides a readable storage medium comprising: storing a program or instructions on the readable storage medium, wherein when the program or instructions are executed by a processor, the program or instructions implement the steps of the above-described emergency material management method based on big data analysis and achieve the same technical effect. To avoid repetition, this invention will not elaborate further.

[0236] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the embodiments of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the protection scope of the present invention.

Claims

1. An emergency material management method based on big data analysis, characterized in that, include: S1: Acquire real-time data from multiple IoT sources; S2: Based on the aforementioned multi-source real-time data from the Internet of Things, a unified virtual inventory model is established by constructing an Internet of Things sensing layer and transmission layer; S3: Input the regional emergency resource status feature vector and historical disaster data from the virtual inventory model into the trained three-layer BP neural network demand prediction model, and output the demand rate of various emergency supplies in the future period. S4: Classify emergency supplies according to their characteristics, and use the corresponding inventory optimization model to optimize the differentiated inventory strategy for different categories of supplies in the classification results; S5: Based on the demand rate, the virtual inventory model, and the optimization results of the differentiated inventory strategy, set supply capacity constraints, road traffic constraints, transportation vehicle volume constraints, and weight constraints, and construct an emergency material dispatching model with the goal of minimizing total delivery time. S6: Based on the principle of prioritizing minimum fill rate, the emergency material scheduling model is solved using a heuristic algorithm to generate emergency material management decisions.

2. The emergency material management method based on big data analysis according to claim 1, characterized in that, The various emergency supplies include essential supplies, scarce supplies, and time-sensitive supplies; Based on the differences in characteristics of various emergency supplies, an inventory optimization model is constructed; The inventory optimization models include an important materials inventory optimization model, a scarce materials inventory optimization model, and a time-sensitive materials inventory optimization model.

3. The emergency material management method based on big data analysis according to claim 2, characterized in that, The inventory strategy optimization process of the aforementioned important material inventory optimization model specifically includes: Based on the demand rate of important materials output by the three-layer BP neural network demand forecasting model, an objective function is constructed with the goal of minimizing the total inventory cost per unit time. Based on the actual conditions of the warehousing facilities, storage capacity constraints are set by clearly defining the total upper limit of the storage space and combining the unit storage capacity occupancy of each important material. Calculate the theoretical order quantity for each important material based on the economic order quantity without inventory constraints. Based on the theoretical order quantity and unit storage capacity, the theoretical total storage capacity is compared with the upper limit of the total storage capacity to determine whether the storage capacity limit is exceeded. If the storage capacity limit is not exceeded, the theoretical order quantity is taken as the final order quantity. Otherwise, an inventory cost optimization function under storage capacity constraints is constructed to calculate the final order quantity of each important material under storage capacity constraints. Based on the final order quantity and the demand rate of the key materials, the ordering cycle and warehouse capacity utilization rate of each key material are calculated respectively.

4. The emergency material management method based on big data analysis according to claim 2, characterized in that, The inventory strategy optimization process of the scarce resource inventory optimization model specifically includes: Based on the demand rate of scarce resources output by the three-layer BP neural network demand forecasting model, and combined with the single order cost and unit annual storage cost of scarce resources, the economic ordering cycle of the scarce resources is calculated. Based on the economic ordering cycle and the preset safety stock level, calculate the maximum inventory level of the scarce material; The actual order quantity of the scarce material is calculated based on the maximum inventory level and the real-time inventory status of the scarce material.

5. The emergency material management method based on big data analysis according to claim 2, characterized in that, The inventory strategy optimization process of the time-sensitive goods inventory optimization model specifically includes: Based on the demand rate of time-sensitive materials output by the three-layer BP neural network demand forecasting model and the order lead time of the time-sensitive materials, the lead time demand of the time-sensitive materials is calculated. Based on the lead time demand of the time-sensitive goods, determine the ordering point for the time-sensitive goods; Based on the cost per order, unit storage cost, and demand rate of the time-sensitive goods, the economic order quantity for replenishment after the order point is triggered is calculated using a quantitative ordering model.

6. The big data analysis based emergency material management method according to claim 1, wherein, S3 specifically includes: S301: Based on the virtual inventory model and historical disaster data, identify the key factors affecting the demand for emergency supplies and construct a multi-dimensional feature input vector; S302: Set the number of neurons in the input layer, hidden layer, and output layer of the three-layer BP neural network demand prediction model; S303: Input the multidimensional feature input vector into the set three-layer BP neural network demand prediction model, and calculate the output value of the hidden layer neurons; S304: Based on the output values ​​of the hidden layer neurons, calculate the predicted emergency material demand value corresponding to the output layer neurons; S305: Based on the output demand forecasts for various emergency supplies, and by combining the time dimension parameters of the future period, output the demand rate of various emergency supplies in the future period. 7.The emergency material management method based on big data analysis of claim 1, wherein, S5 specifically includes: S501: Based on the predicted demand rate, the virtual inventory model, and the optimization results of the differentiated inventory strategy, construct an objective function with minimum total delivery time as the core. S502: Based on the objective function, and combined with the actual inventory and supply capacity of the supply points in the virtual inventory model, set supply capacity constraints; S503: Combines the demand point prediction quantity output by the three-layer BP neural network demand forecasting model, and sets the predicted demand and minimum guarantee rate constraints for various materials through the preset range of demand point delivery quantity. S504: Based on the traffic capacity of the transportation network, road flow constraints are set by limiting the total amount of goods delivered on a single route; S505: Based on the actual loading capacity of the transport vehicle, set volume and weight constraints for the transport vehicle; S506: Based on the full-dimensional constraint settings of the objective function, the supply capacity constraint, the predicted demand and the minimum guarantee rate constraint, the road traffic flow constraint, the vehicle volume constraint and the weight constraint, the overall construction of the scheduling model with the minimum total delivery time as the objective is completed. 8.The emergency material management method based on big data analysis of claim 1, wherein, S6 specifically includes: S601: Based on the principle of prioritizing minimum fill rate, various emergency supplies are sorted according to preset sorting rules; S602: Based on the ranking results of various emergency supplies, prioritize the allocation of supplies at each demand point to meet the minimum fill rate. S603: Based on the remaining excess material demand in the scheduling results, sort them according to preset indicator rules, and complete the distribution allocation optimization through the heuristic algorithm. S604: Based on the upper limit of the preset material delivery time window for each demand point, verify the actual arrival time, and adjust the scheduling decision and delivery strategy according to the verification result until the emergency material management decision that satisfies all constraints is generated.

9. An emergency material management system based on big data analysis, characterized by, include: Processor and memory; The memory stores programs or instructions that can run on the processor, which, when executed by the processor, implement the steps of the emergency supplies management method based on big data analysis as described in any one of claims 1 to 8.

10. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the emergency supplies management method based on big data analysis as described in any one of claims 1 to 8.

Images