Emergency resource data analysis method based on smart station

Abstract

The invention provides an emergency resource data analysis method based on a smart station. The emergency resource data analysis method comprises the steps that real-time station monitoring time sequence data are collected and preprocessed; determining an analysis time period and a resource state label according to the preprocessed data; generating enhanced scheduling or maintenance reserve strategies for different labels; and detecting resource abnormity based on the scheduling strategy and the abnormity prediction model, and generating a scheduling instruction through the decision tree to execute warehouse-in and warehouse-out control or activate a standby channel. According to the invention, the accuracy and response efficiency of emergency resource allocation can be improved, the resource shortage risk is reduced, and the storage utilization rate is optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of emergency resource management, and more specifically, to an emergency resource data analysis method based on a smart station. Background Art

[0002] In the field of emergency resource management, with the development of society and the increase in various emergencies, the effective scheduling and management of emergency resources has become increasingly important. Traditional emergency resource management relies primarily on manual scheduling and simple information systems, which have many shortcomings in resource monitoring, data analysis, and scheduling decisions. For example, in terms of resource monitoring, traditional monitoring methods often only provide limited real-time data, and the accuracy and completeness of the data cannot be guaranteed. In the data analysis process, the lack of effective data preprocessing and in-depth analysis methods makes it impossible to accurately assess the risk level of resource shortages. In terms of scheduling decisions, traditional scheduling strategies are relatively fixed, difficult to dynamically adjust based on real-time data, and lack effective response mechanisms when faced with abnormal resource situations.

[0003] In the process of implementing the embodiments of the present invention, there are at least the following problems or defects in the existing technology: first, traditional methods are unable to monitor and analyze emergency resources in real time and accurately, resulting in the inability to timely discover the risk of resource shortages; second, there is a lack of effective dynamic scheduling strategies, and the scheduling plan cannot be adjusted in real time according to changes in resource status; finally, when faced with resource anomalies, the existing technology lacks effective anomaly detection and decision-making mechanisms, resulting in low efficiency in the scheduling of emergency resources and an inability to meet actual needs. Summary of the Invention

[0004] The present invention provides a method, system, equipment and medium for analyzing emergency resource data based on a smart station.

[0005] In a first aspect of the present invention, a method for analyzing emergency resource data based on a smart station is provided, which is applied to an emergency command platform or a resource scheduling system, comprising:

[0006] Collect real-time site monitoring time series data sets for emergency resources;

[0007] Performing data preprocessing on the real-time station monitoring time series data set to obtain a preprocessed station monitoring time series data set;

[0008] Determining an analysis period based on the pre-processed station monitoring time series data set;

[0009] determining a resource status label according to a target station monitoring time series data set, wherein the target station monitoring time series data set is pre-processed station monitoring data in the pre-processed station monitoring time series data set and located within the analysis period, and the resource status label represents a shortage risk level of the emergency resource;

[0010] In response to determining that the resource status tag is a first-level shortage tag, generating enhanced scheduling policy information for emergency resources as updated scheduling policy information based on the current scheduling policy information corresponding to the smart station and the resource status tag;

[0011] In response to determining that the resource status tag is a normal reserve tag, generating maintenance reserve strategy information for emergency resources according to the current scheduling strategy information and the resource status tag as updated scheduling strategy information;

[0012] Dynamically scheduling emergency resources according to the current scheduling policy information or the updated scheduling policy information, including:

[0013] Determining resource anomaly information based on the current scheduling policy information or the updated scheduling policy information, the resource status label, and the scheduling anomaly prediction model;

[0014] In response to determining that the resource abnormality information indicates that the current scheduling policy information or the updated scheduling policy information has a control abnormality, determining a scheduling instruction corresponding to the resource abnormality information in the abnormal scheduling decision tree;

[0015] According to the scheduling instructions, the target resource warehouse is controlled for entry and exit, or a backup resource channel is activated.

[0016] Furthermore, the performing of data preprocessing on the real-time station monitoring time series data set to obtain a preprocessed station monitoring time series data set includes:

[0017] For each real-time station monitoring data in the real-time station monitoring time series data set, the following processing steps are performed:

[0018] Performing equipment validity verification on the real-time site monitoring data;

[0019] In response to determining that the real-time station monitoring data passes the validity check, performing data cleansing on the real-time station monitoring data to obtain cleansed station monitoring data;

[0020] Performing data formatting on the cleaned station monitoring data to obtain formatted station monitoring data;

[0021] Determining the collection granularity corresponding to the real-time site monitoring time series data set;

[0022] In response to determining that the acquisition granularity is not a reference acquisition granularity and that the acquisition granularity is smaller than the reference acquisition granularity, the formatted station monitoring data sequence is downsampled at the reference acquisition granularity to obtain the preprocessed station monitoring time series data set.

[0023] Furthermore, determining the analysis period based on the pre-processed station monitoring time series data set includes:

[0024] Initializing the time period length corresponding to the analysis time period to obtain an initial analysis time period;

[0025] Based on the initial period length and the pre-processed station monitoring time series data set, perform the following steps:

[0026] Determine the target station monitoring time series data set;

[0027] Calculate the mean value of resource inventory corresponding to the target station monitoring time series data set;

[0028] Obtain the resource consumption peak and valley values ​​in the target station monitoring time series data set;

[0029] Calculate the period characteristic value of the initial analysis period based on the resource inventory mean, resource consumption peak value and resource consumption valley value;

[0030] In response to the resource inventory mean being less than or equal to the dynamic safety threshold, determining an initial analysis period as the analysis period;

[0031] In response to the resource inventory mean being greater than the dynamic safety threshold, the length of the initial analysis period is increased and the above steps are executed again.

[0032] Furthermore, determining the resource status tag based on the target station monitoring time series data set includes:

[0033] Calculate the deviation rate of real-time inventory relative to the safety stock threshold;

[0034] Calculate the ratio of the sum of the incremental demand for multiple types of materials to the demand response time;

[0035] The deviation rate and the ratio are weighted and summed according to a preset weight coefficient to obtain a resource scarcity index;

[0036] In response to the resource shortage index continuously exceeding the first risk threshold for a first time period threshold, determining the resource status label as a first-level shortage label;

[0037] In response to the resource shortage index being continuously lower than the safety threshold for a second period threshold, the resource status tag is determined to be a normal reserve tag.

[0038] Furthermore, the scheduling anomaly prediction model is constructed by the following steps:

[0039] Calculate the absolute difference between the outbound and inbound quantities, and normalize the difference with the maximum inbound quantity to obtain the abnormal characteristics of resource flow;

[0040] The multi-warehouse coordination coefficient is obtained by summing the ratio of each warehouse's inventory to its distance to the disaster site and dividing it by the maximum inventory.

[0041] The resource flow anomaly characteristics and multi-warehouse coordination coefficient are input into the temporal convolutional network model, and the scheduling anomaly probability is output.

[0042] Furthermore, the abnormal scheduling decision tree includes the following determination steps:

[0043] Determine whether the inventory turnover rate is lower than the safety turnover rate;

[0044] If it is lower than the safe turnover rate, further detection is performed to see whether the abnormal characteristics of resource flow exceed the maximum abnormal threshold;

[0045] If the maximum abnormal threshold is exceeded, a first-level dispatch instruction is generated to execute cross-regional transfers;

[0046] If the maximum abnormal threshold is not exceeded, a secondary scheduling instruction is generated to activate the backup warehouse.

[0047] Furthermore, it also includes path optimization mechanisms:

[0048] Obtain the real-time length, speed and load capacity of each transport section;

[0049] Calculate the weighted sum of transportation time and loading time for each section;

[0050] The optimization algorithm is used to solve the minimum value of the weighted sum of all road sections and determine the optimal transportation route.

[0051] Furthermore, the enhanced scheduling strategy includes:

[0052] Calculate the weighted comprehensive value of resource coverage utility and transportation cost saving rate;

[0053] Under the constraint that the total amount of resource allocation to each warehouse does not exceed its inventory,

[0054] A solution for maximizing the weighted comprehensive value is solved by linear programming.

[0055] Furthermore, it also includes 3D site visualization:

[0056] Integrate GPS positioning data, BIM model data and IoT sensor data to build a digital twin model of the station;

[0057] According to the coordinates and inventory of each resource point, the resource distribution heat map is calculated using the Gaussian kernel function.

[0058] Furthermore, dynamic scheduling execution includes:

[0059] Generate a scheduling instruction set containing resource ID, target location and required arrival time;

[0060] Send the instruction set to the intelligent transport vehicle through the 5G private network;

[0061] Real-time monitoring of the difference between actual arrival time and planned arrival time;

[0062] When the difference exceeds the allowable deviation threshold, the transportation path is replanned.

[0063] In a second aspect of the present invention, a smart station-based emergency resource data analysis system is provided, which is applied to an emergency command platform or a resource scheduling system, including:

[0064] Data acquisition module, used to collect real-time site monitoring time series data sets for emergency resources;

[0065] A data preprocessing module, configured to perform data preprocessing on the real-time station monitoring time series data set to obtain a preprocessed station monitoring time series data set;

[0066] An analysis period determination module, configured to determine an analysis period based on the pre-processed station monitoring time series data set;

[0067] a label generation module, configured to determine a resource status label based on a target station monitoring time series data set, wherein the target station monitoring time series data set is data within the analysis period in the preprocessed station monitoring time series data set;

[0068] A first policy generation module is configured to generate enhanced scheduling policy information as updated scheduling policy information based on the current scheduling policy information and the resource status label corresponding to the smart station in response to the resource status label being a first-level shortage label;

[0069] a second policy generating module, configured to generate, in response to the resource status tag being a normal reserve tag, maintenance reserve policy information according to the current scheduling policy information and the resource status tag as updated scheduling policy information;

[0070] The dynamic scheduling execution module is used to dynamically schedule emergency resources based on the current scheduling policy information or the updated scheduling policy information, including:

[0071] Anomaly detection submodule, used to determine resource anomaly information based on current scheduling policy information or updated scheduling policy information, resource status labels and scheduling anomaly prediction model;

[0072] A decision tree processing submodule, configured to determine a corresponding scheduling instruction in an abnormal scheduling decision tree in response to resource abnormality information indicating a control abnormality;

[0073] The resource control submodule is used to control the entry and exit of the target resource warehouse according to scheduling instructions, or to activate backup resource channels.

[0074] Furthermore, the data preprocessing module includes:

[0075] The validity check submodule is used to perform equipment validity check on each data in the real-time site monitoring time series data set;

[0076] A data cleaning submodule is used to remove outliers from the data in response to the data passing the validity check;

[0077] The formatting processing submodule is used to format the cleaned data;

[0078] The granularity analysis submodule is used to determine the collection granularity corresponding to the real-time station monitoring time series data set;

[0079] The downsampling processing submodule is configured to downsample the data at the reference acquisition granularity in response to the acquisition granularity being smaller than the reference acquisition granularity.

[0080] Furthermore, the analysis period determination module includes:

[0081] The period initialization submodule is used to initialize the period length corresponding to the analysis period;

[0082] The data set extraction submodule is used to determine the target station monitoring time series data set;

[0083] The mean calculation submodule is used to calculate the mean value of resource stocks;

[0084] The extreme value acquisition submodule is used to obtain the peak and valley values ​​of resource consumption;

[0085] The characteristic value calculation submodule is used to calculate the characteristic value of a period based on the mean value of resource inventory, peak value of resource consumption and valley value of resource consumption;

[0086] a period determination submodule, configured to determine an initial analysis period as the analysis period in response to the resource inventory mean being less than or equal to the dynamic safety threshold;

[0087] The iterative processing submodule is used to increase the length of the initial analysis period and re-execute the processing flow in response to the resource inventory mean being greater than the dynamic safety threshold.

[0088] In a third aspect of the present invention, an electronic device is provided, comprising: at least one processor, a memory, and an input / output unit; wherein the memory is used to store a computer program, and the processor is used to call the computer program stored in the memory to execute any one of the methods described in the first aspect.

[0089] In a fourth aspect of the present invention, a computer-readable medium is provided, comprising instructions, which, when executed on a computer, causes the computer to execute the method according to any one of the first aspects.

[0090] The above embodiments of the present invention have at least the following beneficial effects:

[0091] 1. By collecting site monitoring time series data in real time and performing multi-level preprocessing, the problems of poor data quality and inconsistent collection granularity in traditional emergency resource monitoring are solved, ensuring the accuracy and reliability of subsequent analysis and providing a high-quality data foundation for intelligent decision-making.

[0092] 2. The analysis period is dynamically determined based on the mean value of resource inventory, consumption peaks, and valleys, and the resource scarcity index is weighted and calculated by combining the deviation rate and demand increment. This achieves accurate classification of emergency resource shortage risks, solves the problem of poor adaptability of the traditional static threshold warning mechanism, and improves the real-time and scientific nature of resource status assessment.

[0093] 3. Through the scheduling anomaly prediction model and abnormal scheduling decision tree, the problem of delayed response in traditional manual scheduling is solved, and automatic identification of resource anomalies and dynamic strategy adjustment are realized. At the same time, combined with path optimization and three-dimensional visualization technology, the efficiency and fault tolerance of emergency resource scheduling are comprehensively improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0094] The above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily apparent by reading the following detailed description with reference to the accompanying drawings, in which several embodiments of the present invention are shown by way of example and not limitation, in which:

[0095] Figure 1 A flowchart of an emergency resource data analysis method based on a smart station provided in one embodiment of the present invention;

[0096] Figure 2 A schematic diagram of the structure of an emergency resource data analysis system based on a smart station provided in one embodiment of the present invention;

[0097] Figure 3 The figure schematically shows the structure of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0098] The principles and spirit of the present invention will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are provided solely to enable those skilled in the art to better understand and implement the present invention, and are not intended to limit the scope of the present invention in any way. Rather, these embodiments are provided to make the present invention more thorough and complete, and to fully convey the scope of the present invention to those skilled in the art.

[0099] Those skilled in the art will appreciate that the embodiments of the present invention may be implemented as a system, apparatus, device, method, or computer program product. Therefore, the present invention may be implemented in the following forms: entirely in hardware, entirely in software (including firmware, resident software, microcode, etc.), or in a combination of hardware and software.

[0100] It should be noted that any number of elements in the drawings is for illustration only and not for limitation, and any naming is only for distinction and does not have any limiting meaning.

[0101] Reference below Figure 1 , Figure 1 The figure is a flow chart of an emergency resource data analysis method based on a smart station provided by an embodiment of the present invention. Figure 1 As shown, a method for analyzing emergency resource data based on a smart station includes:

[0102] S1. Collect real-time site monitoring time series data sets for emergency resources;

[0103] S2. performing data preprocessing on the real-time station monitoring time series data set to obtain a preprocessed station monitoring time series data set;

[0104] S3. Determine an analysis period based on the pre-processed station monitoring time series data set;

[0105] S4. Determine a resource status label based on the target station monitoring time series data set, wherein the target station monitoring time series data set is preprocessed station monitoring data in the preprocessed station monitoring time series data set that is within the analysis period, and the resource status label represents a shortage risk level of the emergency resource;

[0106] S5. In response to determining that the resource status tag is a first-level shortage tag, generating enhanced scheduling policy information for emergency resources as updated scheduling policy information based on the current scheduling policy information corresponding to the smart station and the resource status tag;

[0107] S6. In response to determining that the resource status tag is a normal reserve tag, generating maintenance reserve strategy information for emergency resources according to the current scheduling strategy information and the resource status tag as updated scheduling strategy information;

[0108] S7. Dynamically scheduling emergency resources according to the current scheduling policy information or the updated scheduling policy information, including:

[0109] S8. Determine resource anomaly information based on the current scheduling policy information or the updated scheduling policy information, the resource status label, and the scheduling anomaly prediction model;

[0110] S9. In response to determining that the resource abnormality information indicates that the current scheduling policy information or the updated scheduling policy information has a control abnormality, determining a scheduling instruction corresponding to the resource abnormality information in the abnormal scheduling decision tree;

[0111] S10. According to the scheduling instruction, the target resource warehouse is controlled for entry and exit, or a backup resource channel is activated.

[0112] It should be noted that the present invention proposes an emergency resource data analysis method based on a smart station, which is mainly used in emergency command platforms or resource scheduling systems. Among them, a smart station refers to a station that integrates multiple intelligent technologies, such as the Internet of Things, big data, artificial intelligence, etc., which can realize real-time monitoring and intelligent management of emergency resources. Emergency resources include but are not limited to medical supplies, disaster relief equipment, food, etc., which are crucial when responding to emergencies. The real-time station monitoring time series data set refers to emergency resource-related data collected from smart stations and arranged in chronological order. These data can reflect the real-time status and change trends of resources. Through this method, efficient management and dynamic scheduling of emergency resources can be achieved, and emergency response capabilities can be improved.

[0113] Specifically, collecting real-time site monitoring time series data sets for emergency resources refers to collecting data related to emergency resources, such as inventory, consumption rate, equipment status, etc., through various sensors and monitoring equipment installed in smart sites. These data are stored in the form of time series for subsequent analysis. Data preprocessing includes operations such as cleaning, formatting and downsampling of the collected data to improve data quality and availability. The analysis period is a time period determined based on the preprocessed data, and the data is analyzed in detail during this time period. The resource status label is an indicator calculated based on the data within the analysis period, which is used to characterize the shortage risk level of emergency resources. For example, the first-level shortage label indicates that the resource is extremely scarce, and the normal reserve label indicates that the resource is at a normal reserve level. Scheduling policy information refers to the resource scheduling plan formulated based on the resource status label and the current situation, which is used to guide the allocation and use of emergency resources.

[0114] Preferably, in the data preprocessing step of the real-time site monitoring time series data set, the following specific operations can be adopted: first, perform equipment validity verification on each real-time site monitoring data to ensure the reliability of the data source; then, perform data cleaning on the data that has passed the verification to remove outliers and noise; then, format the cleaned data to meet the requirements of subsequent analysis; finally, according to the data collection granularity (i.e., the time interval for data collection), if the collection granularity is smaller than the benchmark collection granularity, for example, the benchmark collection granularity is to collect data once every 10 minutes, then downsample the data to reduce the amount of data and improve processing efficiency. In the step of determining the analysis period, an initial analysis period can be initialized, and then characteristic values ​​such as the mean value of resource inventory, peak and valley values ​​of resource consumption are calculated based on the preprocessed data, and based on these characteristic values, it is determined whether the initial analysis period will be used as the final analysis period. If the mean value of resource stock is less than or equal to the dynamic safety threshold, which is a threshold set based on historical data and experience to determine whether the resource is at a safe level, the initial analysis period is determined as the analysis period; otherwise, the length of the initial analysis period is increased and the characteristic value is recalculated until the conditions are met.

[0115] In some embodiments, performing data preprocessing on the real-time station monitoring time series data set to obtain a preprocessed station monitoring time series data set includes:

[0116] For each real-time station monitoring data in the real-time station monitoring time series dataset, the following processing steps are performed:

[0117] Performing equipment validity verification on the real-time site monitoring data;

[0118] In response to determining that the real-time station monitoring data passes the validity check, performing data cleansing on the real-time station monitoring data to obtain cleansed station monitoring data;

[0119] Performing data formatting on the cleaned station monitoring data to obtain formatted station monitoring data;

[0120] Determining the collection granularity corresponding to the real-time site monitoring time series data set;

[0121] In response to determining that the acquisition granularity is not a reference acquisition granularity and that the acquisition granularity is smaller than the reference acquisition granularity, the formatted station monitoring data sequence is downsampled at the reference acquisition granularity to obtain the preprocessed station monitoring time series data set.

[0122] It should be noted that the data preprocessing step for real-time station monitoring time-series datasets in this invention is intended to ensure data quality and usability, thereby providing accurate data support for subsequent analysis and decision-making. Data preprocessing is a crucial step in the data processing process and includes operations such as data validity verification, cleaning, formatting, and downsampling as needed. These operations can remove noise and errors from the data, unify the data format, and adjust the data collection frequency to make it more suitable for subsequent analysis and processing.

[0123] Specifically, device validity verification involves verifying the collected real-time monitoring data from each site, ensuring that the data originates from properly functioning equipment to ensure data accuracy and reliability. If equipment malfunctions or anomalies occur, the collected data may be biased or erroneous, necessitating device validity verification. Data cleaning involves removing outliers and noise from the data. These outliers may be caused by equipment failure, signal interference, or other external factors. Data cleaning improves data quality and makes it more realistic. Data formatting converts cleaned data into a unified format for subsequent analysis and processing. Different equipment and systems may generate data in different formats. Formatting unifies this data, making it comparable and actionable. Collection granularity refers to the interval at which data is collected. For example, if data is collected every minute, the collection granularity is one minute. If the collection granularity is smaller than the baseline collection granularity, it indicates that data collection is too frequent and the data volume is too large, potentially reducing processing efficiency. Therefore, data downsampling is necessary—reducing the number of data points according to specific rules to reduce the data volume and improve processing efficiency.

[0124] Preferably, when verifying device validity, parameters such as the device's signal strength and data integrity can be used to determine whether the device is functioning properly. For example, if the device's signal strength falls below a certain threshold, or if the data contains a large number of missing or erroneous values, it can be determined that the device may be malfunctioning. During data cleaning, statistical methods can be used to identify and remove outliers. For example, the mean and standard deviation of the data can be calculated, and data points outside the range of the mean plus or minus three times the standard deviation are considered outliers and removed. During data formatting, data can be converted to a unified time format, numerical format, etc. based on analysis requirements. For example, all time data can be converted to a unified 24-hour format, and numerical data can be converted to a unified decimal format. During downsampling, an appropriate downsampling method can be selected based on actual needs and data characteristics. For example, if the data exhibits periodic changes, one data point can be retained within each period, thereby reducing the data volume without losing important information.

[0125] In some embodiments, determining the analysis period based on the pre-processed station monitoring time series data set includes:

[0126] Initializing the time period length corresponding to the analysis time period to obtain an initial analysis time period;

[0127] Based on the initial period length and the pre-processed station monitoring time series data set, perform the following steps:

[0128] Determine the target station monitoring time series data set;

[0129] Calculate the mean value of resource inventory corresponding to the target station monitoring time series data set;

[0130] Obtain the resource consumption peak and valley values ​​in the target station monitoring time series data set;

[0131] Calculate the period characteristic value of the initial analysis period based on the resource inventory mean, resource consumption peak value and resource consumption valley value;

[0132] In response to the resource inventory mean being less than or equal to the dynamic safety threshold, determining an initial analysis period as the analysis period;

[0133] In response to the resource inventory mean being greater than the dynamic safety threshold, the length of the initial analysis period is increased and the above steps are executed again.

[0134] It should be noted that the step of determining the analysis period in the present invention is to find a suitable time period from the pre-processed site monitoring time series data set so that the data within this time period can be analyzed in detail, thereby accurately assessing the status of emergency resources. This process involves initializing the length of the analysis period, calculating the mean value of resource inventory, peak and valley values ​​of consumption, and judging whether to determine the current period as the analysis period based on these characteristic values. If the mean value of resource inventory is lower than or equal to the dynamic safety threshold, it means that resources may be in short supply. In this case, the current period is determined as the analysis period; otherwise, the period length is increased and re-evaluated.

[0135] Specifically, the time period length corresponding to the initial analysis period refers to setting an initial time range for starting the analysis process. This initial time period length can be set based on historical data or experience, for example, it can be set to one day or one week. The target site monitoring time series data set refers to the relevant data subset extracted from the preprocessed data set during the initial analysis period, which will be used for subsequent analysis. The mean resource stock refers to the average stock of emergency resources during the target period, which reflects the overall level of resources. The resource consumption peak and resource consumption valley refer to the highest and lowest points of resource consumption during the target period, respectively. These two values ​​can help understand the fluctuations in resource consumption. The dynamic safety threshold is a threshold set based on historical data and experience, which is used to determine whether the resource stock is at a safe level. If the mean resource stock is lower than or equal to this threshold, it may indicate a risk of resource shortage.

[0136] Preferably, when determining the analysis period, an initial period length can be set first, such as one day. Then, the data within this day is extracted from the preprocessed site monitoring time series data set as the target site monitoring time series data set. Next, the average value of the resource inventory within this day, as well as the highest and lowest points of resource consumption, are calculated. Based on these data, the period characteristic values ​​are calculated, for example, the resource inventory mean can be compared with the dynamic safety threshold. If the resource inventory mean is lower than or equal to the dynamic safety threshold, it means that the resources may be insufficient, and this day is determined as the analysis period. If the resource inventory mean is higher than the dynamic safety threshold, it means that the resource inventory is relatively sufficient, and the initial period length can be increased, for example, to two days, and then the data is re-extracted and the above calculations are performed until a period that meets the conditions is found. This process can be implemented through automated algorithms to improve efficiency and accuracy.

[0137] In some embodiments, determining the resource status tag based on the target station monitoring time series data set includes:

[0138] Calculate the deviation rate of real-time inventory relative to the safety stock threshold;

[0139] Calculate the ratio of the sum of the incremental demand for multiple types of materials to the demand response time;

[0140] The deviation rate and the ratio are weighted and summed according to a preset weight coefficient to obtain a resource scarcity index;

[0141] In response to the resource shortage index continuously exceeding the first risk threshold for a first time period threshold, determining the resource status label as a first-level shortage label;

[0142] In response to the resource shortage index being continuously lower than the safety threshold for a second period threshold, the resource status tag is determined to be a normal reserve tag.

[0143] The deviation rate can be calculated by dividing the absolute value of the difference between the real-time inventory and the safety stock threshold by the safety stock threshold. For example, when the safety stock threshold is 1,000 units and the real-time inventory is 800 units, the deviation rate is 20%. The ratio of the total demand increment to the response time can reflect the growth rate of material demand per unit time. For example, if the total demand increment is 500 units and the response time is 2 hours, the ratio is 250 units / hour. The preset weight coefficient can be dynamically adjusted based on historical data. For example, the deviation rate weight is 0.6 and the demand change rate weight is 0.4. The setting of the first risk threshold and the safety threshold needs to be combined with the resource type and the emergency response level. For example, the trigger condition for the first-level shortage label is that the resource shortage index exceeds the threshold of 80 for 4 consecutive hours, and the trigger condition for the normal reserve label is that the index is continuously below the threshold of 30 for 6 hours.

[0144] Specifically, the calculation of the real-time inventory deviation rate directly quantifies the gap between current inventory and the safety baseline, providing a foundation for assessing resource gaps. The ratio of total demand increments to response time combines the rate of material consumption with the timeliness requirements of emergency response, effectively capturing the resource consumption characteristics of emergencies. The deviation rate and demand change rate are weighted and integrated using weight coefficients, preserving the fundamental monitoring value of inventory levels while strengthening the driving effect of demand dynamics on the degree of scarcity. The setting of time periods for consecutively exceeding or falling below the threshold avoids misjudgments caused by single data fluctuations. For example, if the resource scarcity index exceeds the threshold twice within three hours but does not exceed the threshold for four consecutive hours, a label update will not be triggered. The bidirectional judgment logic of the first-level scarcity label and the normal reserve label can simultaneously warn of resource shortage risks and identify when inventory has returned to a safe level, providing a reliable basis for differentiated scheduling strategies. Compared with traditional static threshold comparison methods, this solution, through the integration of multi-dimensional indicators and time series condition judgment, effectively distinguishes between sudden demand and cyclical fluctuations, improving the dynamic and accurate assessment of resource status.

[0145] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0146] Calculate the deviation rate of the real-time inventory level relative to the safety stock threshold. For example, the deviation rate can be calculated by dividing the difference between the real-time inventory level and the safety stock threshold by the safety stock threshold. Assuming the real-time inventory level of a certain emergency supply is 80 units and the safety stock threshold is 100 units, the deviation rate is (80-100) / 100 = -20%.

[0147] Calculate the ratio of the sum of the incremental demand for multiple types of supplies to the demand response time. Specifically, you can first count the incremental demand for various emergency supplies within a certain period of time, add up these increments to get the total, and then divide it by the preset demand response time. For example, a station has an increase of 500 masks, 200 protective suits, and 100 liters of disinfectant in the past 24 hours. The demand response time is 12 hours, then the ratio is (500+200+100) / 12=66.67

[0148] The resource scarcity index is calculated by weighting the deviation rate and the ratio according to the preset weight coefficient. Assuming the weight of the deviation rate is 0.6 and the weight of the ratio is 0.4, the resource scarcity index can be expressed as: -20% * 0.6 + 66.67 * 0.4 = -12% + 26.67 = 14.67.

[0149] Furthermore, in response to the resource scarcity index continuously exceeding the first risk threshold for a first time period threshold, the resource status label is determined to be a level 1 scarcity label. For example, if the first risk threshold is set to 15 and the first time period threshold is set to 6 hours, if the resource scarcity index remains above 15 for 6 consecutive hours, the resource status label is determined to be a level 1 scarcity label.

[0150] Therefore, in response to the resource scarcity index being continuously below the safety threshold for a second period threshold, the resource status label is determined to be a normal reserve label. Specifically, the safety threshold can be set to 5, and the second period threshold can be set to 12 hours. When the resource scarcity index remains below 5 for 12 consecutive hours, the resource status label is determined to be a normal reserve label.

[0151] Through the above technical solution, this application realizes dynamic and accurate judgment of the status of emergency resources. By calculating the ratio of the real-time inventory deviation from the safety stock threshold, it directly reflects the gap between the current inventory level and the safety benchmark, providing a basic quantitative basis for judging the resource gap. At the same time, the ratio of the sum of the incremental demand for multiple types of materials to the response time is calculated, and the rate of change of material demand is combined with the timeliness requirements of emergency response, effectively capturing the resource consumption characteristics of emergencies. By weighting and fusing the deviation rate and the rate of change of demand through a preset weight coefficient, it not only retains the basic monitoring value of the inventory level, but also strengthens the driving effect of demand dynamics on the degree of scarcity, forming a comprehensive resource scarcity index. Setting the time period condition of continuously exceeding the risk threshold to trigger the label judgment mechanism avoids misjudgment caused by single data fluctuations and ensures the stability of the status label. The two-way judgment logic of the first-level shortage label and the normal reserve label can not only timely warn of resource shortage risks, but also identify the stable state of inventory recovery to a safe level, providing a reliable basis for the differentiated generation of subsequent scheduling strategies.

[0152] In some embodiments, the scheduling anomaly prediction model is constructed by the following steps:

[0153] Calculate the absolute difference between the outbound and inbound quantities, and normalize the difference with the maximum inbound quantity to obtain the abnormal characteristics of resource flow;

[0154] The multi-warehouse coordination coefficient is obtained by summing the ratio of each warehouse's inventory to its distance to the disaster site and dividing it by the maximum inventory.

[0155] The resource flow anomaly characteristics and multi-warehouse coordination coefficient are input into the temporal convolutional network model, and the scheduling anomaly probability is output.

[0156] The normalization processing of resource flow anomaly features can use the historical maximum value of inventory input or the maximum value of sliding window statistics, for example, normalization is based on the peak value of inventory input in the past 24 hours. The distance parameter in the multi-warehouse coordination coefficient can be replaced with transportation time or path complexity indicators, and the inventory can use real-time inventory data or predicted inventory data. The dilated convolution layer of the temporal convolutional network model can be configured with an exponentially increasing expansion coefficient. For example, when the number of layers is 4, the expansion coefficient is 1, 2, 4, and 8 respectively. The resource flow anomaly features and the multi-warehouse coordination coefficient are input into the model through splicing or weighted fusion. For example, a fully connected layer is added before the input layer to align the feature dimensions.

[0157] Specifically, the absolute difference in inbound and outbound quantities is calculated and then divided by the maximum inbound quantity to eliminate differences in warehouse size. For example, if the outbound quantity in a certain period is 500 units, the inbound quantity is 300 units, and the maximum inbound quantity is 600 units, the normalized result is (500-300) / 600≈0.33. The ratio of each warehouse's inventory to the distance to the disaster site is summed and divided by the maximum inventory. For example, if three warehouses have inventory levels of 800, 1200, and 1000 units, respectively, and are located 50, 30, and 40 kilometers from the disaster site, respectively, with a maximum inventory of 1200 units, the coordination coefficient is (800 / 50 + 1200 / 30 + 1000 / 40) / 1200≈0.23. The temporal convolutional network model extracts temporal features layer by layer through causal convolutional layers, and the final fully connected layer outputs an anomaly probability value. For example, when the probability exceeds 0.85, an alert is triggered. In this process, dynamic anomaly features quantify the instantaneous fluctuations of resource flow, and the coordination coefficient reflects the spatial allocation capability between warehouses. The combination of the two captures spatiotemporal correlations through a temporal convolutional network, thereby improving the prediction accuracy of resource allocation imbalances.

[0158] The scheduling anomaly prediction model is constructed through the following steps: First, the absolute difference between outbound and inbound quantities is calculated and normalized by the maximum inbound quantity to obtain the resource flow anomaly characteristic. For example, assuming a warehouse's outbound quantity is 100 units, inbound quantity is 80 units, and the maximum inbound quantity is 120 units in a day, the resource flow anomaly characteristic can be calculated as (|100 - 80|) / 120 = 0.167. Second, the ratio of each warehouse's inventory to its distance to the disaster site is summed and divided by the maximum inventory to obtain the multi-warehouse coordination coefficient. For example, if there are three warehouses with inventory quantities of 200, 300, and 400 units, respectively, and distances to the disaster site of 10, 15, and 20 kilometers, respectively, with a maximum inventory of 500 units, the multi-warehouse coordination coefficient can be calculated as (200 / 10 + 300 / 15 + 400 / 20) / 500 = 0.093. Finally, the resource flow anomaly characteristics and multi-warehouse coordination coefficients are used as input and processed by a temporal convolutional network model to output the probability of scheduling anomalies. The temporal convolutional network model can adopt a multi-layer convolutional structure, with each layer containing several convolution kernels to capture temporal characteristics. Model training can use historical scheduling data and optimize network parameters through a backpropagation algorithm.

[0159] In some embodiments, the abnormal scheduling decision tree includes the following determination steps:

[0160] Determine whether the inventory turnover rate is lower than the safety turnover rate;

[0161] If it is lower than the safe turnover rate, further detection is performed to see whether the abnormal characteristics of resource flow exceed the maximum abnormal threshold;

[0162] If the maximum abnormal threshold is exceeded, a first-level dispatch instruction is generated to execute cross-regional transfers;

[0163] If the maximum abnormal threshold is not exceeded, a secondary scheduling instruction is generated to activate the backup warehouse.

[0164] It should be noted that the abnormal scheduling decision tree in this invention is a logical structure used to quickly determine the response strategy when resource anomaly information is detected. It generates corresponding scheduling instructions based on a series of preset judgment conditions, such as whether the inventory turnover rate is lower than the safe turnover rate and whether the abnormal characteristics of resource flow exceed the maximum abnormality threshold. These instructions can be first-level scheduling instructions for executing cross-regional allocations or second-level scheduling instructions for activating backup warehouses. This decision tree structure can quickly make reasonable scheduling decisions based on different abnormal situations, improving the efficiency and reliability of emergency resource scheduling.

[0165] Specifically, inventory turnover refers to the number of times emergency resources are turned over within a given period. It reflects the efficiency and speed of resource utilization. Typically, inventory turnover is calculated by dividing the outbound inventory by the average inventory level over a given period. The safety turnover rate is a threshold set based on historical data and experience, used to determine whether the current inventory turnover rate is within normal limits. If the inventory turnover rate is lower than the safety turnover rate, it indicates a potential resource backlog or supply shortage, requiring further investigation. The resource flow anomaly characteristic is calculated by calculating the absolute difference between outbound and inbound inventory and normalizing it with the maximum inbound inventory. It reflects the degree of anomaly in the resource flow process. The maximum anomaly threshold is a set upper limit used to determine whether the resource flow anomaly characteristic is within an acceptable range. If the resource flow anomaly characteristic exceeds this threshold, it indicates a serious resource flow problem and requires urgent action. Level 1 and level 2 dispatch instructions are generated at different levels based on the severity of the anomaly. Level 1 dispatch instructions typically involve more urgent measures, such as cross-regional transfers, while level 2 dispatch instructions are more routine, such as activating backup warehouses.

[0166] The safety turnover rate of inventory turnover can be set to a dynamically adjusted value, for example, determined by multiplying the historical average turnover rate by a correction factor of 0.8-1.2. The calculation of abnormal resource flow characteristics may include the percentage of the difference between the outbound and inbound quantities relative to the historical peak, and the threshold comparison is triggered when the percentage exceeds the preset range of 60%-80%. The execution of cross-regional transfer instructions needs to be associated with the transportation route optimization mechanism, such as giving priority to the route with the shortest weighted transportation time. The activation conditions of the backup warehouse can be combined with its straight-line distance from the target warehouse, for example, it is activated when it is within 50 kilometers and the inventory volume meets 1.2 times the demand.

[0167] Specifically, a comparison of inventory turnover rate and safety turnover rate serves as a primary screening criterion, identifying inventory liquidity anomalies through quantitative indicators. When this condition is met, the system automatically extracts resource flow anomaly characteristics for secondary verification. The calculation of these anomaly characteristics incorporates real-time inventory inflow and outflow data. A sliding window mechanism is used to set the maximum anomaly threshold, dynamically adjusting the threshold baseline based on, for example, the last seven days of data. When both judgment conditions are met, a cross-regional transfer instruction is triggered, and its execution leverages a path optimization algorithm to generate an optimal transportation plan. When only the primary condition is met, the activation instruction for the backup warehouse prioritizes the use of a pre-set adjacent storage point and automatically opens storage channels via IoT devices. This tiered judgment mechanism, through dual-condition filtering, categorizes resource anomalies into different response levels. Level one instructions correspond to cross-regional resource reallocation, while level two instructions enable localized rapid replenishment, thereby ensuring scheduling accuracy while avoiding resource waste.

[0168] As a preferred embodiment, the solution of the present application is specifically implemented as follows: the abnormal scheduling decision tree includes multiple judgment steps. First, determine whether the inventory turnover rate is lower than the safe turnover rate. For example, the safe turnover rate can be set to 3 times / month. When it is detected that the inventory turnover rate of a certain emergency material is lower than 3 times / month, the next judgment is triggered. Secondly, further detect whether the abnormal characteristics of resource flow exceed the maximum abnormal threshold. The abnormal characteristics of resource flow can be measured by calculating the difference between the outbound volume and the inbound volume, and the maximum abnormal threshold can be set to 30%. If the maximum abnormal threshold is exceeded, a first-level scheduling instruction is generated to perform cross-regional transfers. Specifically, 50% of the inventory can be transferred from neighboring areas. If the maximum abnormal threshold is not exceeded, a second-level scheduling instruction is generated to enable the spare warehouse. For example, the nearest spare warehouse can be enabled and 20% of its inventory can be called.

[0169] In some embodiments, a path optimization mechanism is also included:

[0170] Obtain the real-time length, speed and load capacity of each transport section;

[0171] Calculate the weighted sum of transportation time and loading time for each section;

[0172] The optimization algorithm is used to solve the minimum value of the weighted sum of all road sections and determine the optimal transportation route.

[0173] It should be noted that the path optimization mechanism in this invention is designed to dynamically determine the optimal transport route during emergency resource scheduling based on real-time road and transportation conditions, thereby improving the efficiency and reliability of resource scheduling. This mechanism obtains information such as the real-time length, vehicle speed, and load capacity of each transport route, calculates the weighted sum of the transport time and loading time for each route, and uses an optimization algorithm to find the minimum value of this weighted sum across all routes to determine the optimal transport route. This process ensures that emergency resources reach their destination in the shortest possible time, significantly improving emergency response capabilities, particularly in emergency situations.

[0174] Specifically, the real-time length of each transport route refers to the actual length of each road section during the transport process. This data can be obtained in real time through the traffic monitoring system. Vehicle speed refers to the actual speed of the vehicle on each road section, which can be affected by factors such as road conditions and traffic regulations. Loading capacity refers to the maximum load capacity of the transport vehicle, which determines the amount of resources the vehicle can transport. Transportation time refers to the time required for the vehicle to travel on each road section, which is related to the length of the road section and the vehicle speed. Loading time refers to the time required for the vehicle to load and unload resources, which depends on the efficiency of the loading equipment and the type of resources. The weighted sum is the value obtained by adding the transportation time and loading time after assigning different weights according to their importance. This value is used to evaluate the advantages and disadvantages of different transportation routes. Optimization algorithms are mathematical methods used to find the optimal solution among multiple possible solutions. For example, genetic algorithms and simulated annealing algorithms can be used.

[0175] Preferably, when implementing a route optimization mechanism, the real-time length, vehicle speed, and load capacity of each transport route segment are first acquired through a traffic monitoring system. The transport time for each route segment is then calculated based on this information. For example, the transport time can be calculated by dividing the route segment length by the vehicle speed. Loading time is also calculated based on the efficiency of the loading equipment and the type of resource. Next, transport time and loading time are weighted according to their importance. For example, a weight of 0.7 for transport time and 0.3 for loading time are assigned. A weighted sum of these sums is then calculated for each route segment. Finally, an optimization algorithm, such as a genetic algorithm, is used to evaluate all possible transport routes, and the path with the lowest weighted sum is identified as the optimal route. During this process, the weights and optimization algorithm parameters can be adjusted based on the actual situation to ensure the optimal route is found. For example, in an emergency, the weight of transport time can be increased to ensure that resources reach their destination as quickly as possible.

[0176] In some embodiments, the enhanced scheduling strategy includes:

[0177] Calculate the weighted comprehensive value of resource coverage utility and transportation cost saving rate;

[0178] Under the constraint that the total amount of resource allocation to each warehouse does not exceed its inventory,

[0179] A solution for maximizing the weighted comprehensive value is solved by linear programming.

[0180] It should be noted that the enhanced scheduling strategy in this invention is designed to ensure that resources can reach the required locations quickly and efficiently by optimizing resource allocation and transportation routes when a primary shortage of emergency resources occurs. This strategy calculates a weighted composite value of resource coverage utility and transportation cost savings, and then uses linear programming to find a solution that maximizes this weighted composite value, subject to the constraint that the total amount of resource allocation to each warehouse does not exceed its inventory. This approach can maximize the utilization of existing resources and improve emergency response efficiency in situations of extreme resource shortages.

[0181] Specifically, resource coverage utility refers to the number of demand points or the size of demand that can be covered by reasonable allocation of resources, which reflects the effectiveness of resource allocation. The transportation cost saving rate refers to the proportion of costs saved by optimizing transportation routes and methods compared with conventional transportation plans, which reflects the improvement in transportation efficiency. The weighted comprehensive value is the value obtained by weighting the resource coverage utility and the transportation cost saving rate according to the preset weight coefficient, which is used to comprehensively evaluate the pros and cons of the scheduling plan. Linear programming is a mathematical optimization method used to solve the maximum or minimum value of the objective function under a set of linear constraints. In the present invention, linear programming is used to solve the weighted comprehensive value maximization solution under the constraint that the total amount of resource allocation to each warehouse does not exceed its inventory.

[0182] Preferably, when implementing an enhanced scheduling strategy, it is first necessary to determine the calculation methods for resource coverage utility and transportation cost savings. For example, resource coverage utility can be measured by calculating the ratio of the amount of resources allocated to each demand point to the amount of demand at that point, and transportation cost savings can be calculated by comparing the cost difference between the optimized transportation route and the conventional route. Then, weight coefficients are set according to the actual situation, and the resource coverage utility and transportation cost savings are weighted and summed to obtain a weighted comprehensive value. Next, a linear programming model is constructed, using the total amount of resources allocated to each warehouse as the decision variable, the weighted comprehensive value as the objective function, and the inventory level of each warehouse as a constraint. A linear programming solver, such as the simplex method or the interior point method, is used to find a solution that maximizes the objective function. In practical applications, the weight coefficients and constraints can be adjusted according to different emergency scenarios and resource types to ensure the rationality and effectiveness of the scheduling plan. For example, in cases of extreme resource shortages, the weight of resource coverage utility can be increased to prioritize meeting the basic needs of more demand points.

[0183] In some embodiments, a 3D site visualization is also included:

[0184] Integrate GPS positioning data, BIM model data and IoT sensor data to build a digital twin model of the station;

[0185] According to the coordinates and inventory of each resource point, the resource distribution heat map is calculated using the Gaussian kernel function.

[0186] It should be noted that the three-dimensional station visualization in this invention is intended to provide an intuitive visualization platform for emergency resource distribution and management. By integrating GPS positioning data, BIM model data, and IoT sensor data, a digital twin model of the station is constructed, enabling real-time monitoring and management of emergency resources. Furthermore, a Gaussian kernel function is used to calculate a resource distribution heat map based on the coordinates and inventory levels of each resource point, visually displaying the distribution of resources and helping managers make quick decisions.

[0187] Specifically, GPS positioning data refers to the geographic location information of resource points obtained through the global positioning system, which is used to determine the precise location of resource points. BIM model data refers to building information model data, which contains detailed information such as the building structure and facility layout of the station, and is used to construct a three-dimensional model of the station. IoT sensor data refers to real-time data collected by IoT sensors, such as the inventory and status of resources, which is used to reflect the real-time situation of resources. The digital twin model of the station is a virtual station model. By integrating the above data, it can reflect the actual situation of the station in real time, including the location, quantity and status of resources. The resource distribution heat map is a visualization tool that uses the depth of color to represent the distribution density of resources, helping managers to intuitively understand the distribution of resources. The Gaussian kernel function is a mathematical function used to calculate the color intensity on the heat map based on the coordinates and inventory of resource points to generate a heat map.

[0188] Preferably, when implementing 3D site visualization, first obtain the precise location information of each resource point using a GPS device. Simultaneously, extract the site's building structure and facility layout data from the BIM system. Real-time resource inventory and status data is then collected from IoT sensors. This data is integrated to construct a digital twin model of the site. When building the model, professional 3D modeling software, such as Revit or SketchUp, can be used to import and visualize the BIM data. To generate a resource distribution heat map, a Gaussian kernel function is used to calculate the color intensity of each location based on the resource point's coordinates and inventory. Specifically, the parameters of the Gaussian kernel function, such as the kernel width, can be set to control the heat map's smoothness. The calculated color intensity is then mapped onto the heat map to generate an intuitive resource distribution map. In practical applications, the Gaussian kernel function parameters can be adjusted to meet specific needs to generate a heat map that better reflects actual requirements. For example, in areas with dense resource distribution, the kernel width can be reduced to more accurately reflect resource distribution.

[0189] In some embodiments, dynamically scheduling execution includes:

[0190] Generate a scheduling instruction set containing resource ID, target location and required arrival time;

[0191] Send the instruction set to the intelligent transport vehicle through the 5G private network;

[0192] Real-time monitoring of the difference between actual arrival time and planned arrival time;

[0193] When the difference exceeds the allowable deviation threshold, the transportation path is replanned.

[0194] It should be noted that the dynamic scheduling execution in the present invention is to ensure that emergency resources can reach the designated location efficiently and accurately according to the predetermined scheduling strategy. This process includes generating a scheduling instruction set containing the resource ID, target location and required arrival time, and sending the instruction set to the intelligent transport vehicle through the 5G private network. At the same time, the difference between the actual arrival time and the planned arrival time is monitored in real time. When the difference exceeds the allowable deviation threshold, the transportation route is replanned. This method can effectively respond to various emergencies during transportation and ensure the timely supply of emergency resources.

[0195] Specifically, a scheduling instruction set refers to a set of instructions containing information such as resource ID, target location, and required arrival time, which is used to guide the actions of intelligent transport vehicles. The resource ID is a number that uniquely identifies each batch of emergency resources and is used to distinguish different resources. The target location refers to the specific place where the resource needs to arrive, usually a geographic coordinate or address. The required arrival time refers to the latest time by which the resource must arrive at the target location. The 5G private network is a high-speed, low-latency communication network used to quickly and accurately transmit the scheduling instruction set to intelligent transport vehicles. Intelligent transport vehicles refer to transportation vehicles with autonomous navigation and communication capabilities, such as self-driving vehicles or drones. The actual arrival time refers to the time when the resource actually arrives at the target location, while the planned arrival time refers to the arrival time scheduled according to the scheduling instruction set. The allowable deviation threshold is a set time range used to determine whether the deviation between the actual arrival time and the planned arrival time is within an acceptable range.

[0196] Preferably, when implementing dynamic scheduling, a detailed set of scheduling instructions is first generated based on the resource type and location of demand. These instructions include the resource's unique identifier (resource ID), the precise coordinates of the target location, and the required arrival time. These instructions are then transmitted to the intelligent transport vehicle via the 5G private network, ensuring rapid transmission and accurate receipt of the instructions. Upon receiving the instructions, the intelligent transport vehicle proceeds according to the predetermined route and time. During transportation, onboard sensors and communication equipment monitor the transport status, including the actual arrival time, in real time. If the difference between the actual arrival time and the planned arrival time exceeds a permissible deviation threshold, for example, more than 10 minutes, the system automatically triggers a rerouting mechanism. This rerouting process can take into account factors such as current road conditions and the vehicle's remaining battery or fuel, using an optimization algorithm to select a new optimal route. For example, if traffic is congested, the system can automatically adjust the route to select a more unobstructed route to ensure the resource arrives on time.

[0197] The above embodiments of the present invention have the following beneficial effects:

[0198] 1. By collecting site monitoring time series data in real time and performing multi-level preprocessing, the problems of poor data quality and inconsistent collection granularity in traditional emergency resource monitoring are solved, ensuring the accuracy and reliability of subsequent analysis and providing a high-quality data foundation for intelligent decision-making.

[0199] 2. The analysis period is dynamically determined based on the mean value of resource inventory, consumption peaks, and valleys, and the resource scarcity index is weighted and calculated by combining the deviation rate and demand increment. This achieves accurate classification of emergency resource shortage risks, solves the problem of poor adaptability of the traditional static threshold warning mechanism, and improves the real-time and scientific nature of resource status assessment.

[0200] 3. Through the scheduling anomaly prediction model and abnormal scheduling decision tree, the problem of delayed response in traditional manual scheduling is solved, and automatic identification of resource anomalies and dynamic strategy adjustment are realized. At the same time, combined with path optimization and three-dimensional visualization technology, the efficiency and fault tolerance of emergency resource scheduling are comprehensively improved.

[0201] like Figure 2 As shown, some embodiments provide an emergency resource data analysis system based on a smart station, the system comprising:

[0202] Data collection module 201, used to collect real-time site monitoring time series data sets for emergency resources;

[0203] A data preprocessing module 202 is configured to perform data preprocessing on the real-time station monitoring time series data set to obtain a preprocessed station monitoring time series data set;

[0204] An analysis period determination module 203 is configured to determine an analysis period based on the pre-processed station monitoring time series data set;

[0205] a label generation module 204 for determining a resource status label based on a target station monitoring time series data set, wherein the target station monitoring time series data set is data in the pre-processed station monitoring time series data set within the analysis period;

[0206] The first policy generating module 205 is configured to generate enhanced scheduling policy information as updated scheduling policy information based on the current scheduling policy information corresponding to the smart station and the resource status label in response to the resource status label being a first-level shortage label;

[0207] The second policy generating module 206 is configured to generate, in response to the resource status tag being a normal reserve tag, maintenance reserve policy information according to the current scheduling policy information and the resource status tag as updated scheduling policy information;

[0208] The dynamic scheduling execution module 207 is used to dynamically schedule emergency resources according to the current scheduling policy information or the updated scheduling policy information, including:

[0209] Anomaly detection submodule 208, configured to determine resource anomaly information based on current scheduling policy information or updated scheduling policy information, resource status tags, and a scheduling anomaly prediction model;

[0210] The decision tree processing submodule 209 is configured to determine a corresponding scheduling instruction in the abnormal scheduling decision tree in response to the resource abnormality information indicating a control abnormality;

[0211] The resource control submodule 210 is used to control the entry and exit of the target resource warehouse according to the scheduling instructions, or to activate the backup resource channel.

[0212] It is understandable that the modules and references recorded in the emergency resource data analysis system based on the smart station Figure 1 The steps in the described method for analyzing emergency resource data based on smart stations correspond to each other. Therefore, the operations, features, and beneficial effects described above for the method for analyzing emergency resource data based on smart stations are also applicable to the emergency resource data analysis system based on smart stations and the modules contained therein, and will not be repeated here.

[0213] In some embodiments, the data preprocessing module includes:

[0214] The validity check submodule is used to perform equipment validity check on each data in the real-time site monitoring time series data set;

[0215] A data cleaning submodule is used to remove outliers from the data in response to the data passing the validity check;

[0216] The formatting processing submodule is used to format the cleaned data;

[0217] The granularity analysis submodule is used to determine the collection granularity corresponding to the real-time station monitoring time series data set;

[0218] The downsampling processing submodule is configured to downsample the data at the reference acquisition granularity in response to the acquisition granularity being smaller than the reference acquisition granularity.

[0219] It's important to note that the data preprocessing module in this invention ensures that the collected real-time station monitoring time series data meets the requirements of subsequent analysis and processing. This module includes a validity verification submodule, a data cleaning submodule, a formatting submodule, a granularity analysis submodule, and a downsampling submodule. These submodules work together to perform a series of preprocessing operations on the data, thereby improving its quality and usability. These steps effectively remove noise and errors from the data, unify the data format, and adjust the data collection frequency, providing reliable data support for subsequent analysis and scheduling.

[0220] Specifically, the validity verification submodule performs device validity verification on each collected data point to ensure that the data originates from a properly functioning device. If a device experiences a malfunction or anomaly, the collected data may contain bias or errors. The data cleaning submodule removes outliers and noise from the data. These outliers may be caused by device failure, signal interference, or other external factors. Data cleaning improves data quality and makes it more realistic. The formatting submodule converts the cleaned data into a unified format for subsequent analysis and processing. Different devices and systems may generate data in different formats. Formatting unifies this data, making it comparable and operational. The granularity analysis submodule determines the data collection granularity, that is, the interval between data collections. If the collection granularity is smaller than the baseline collection granularity, it indicates that data collection is too frequent and the data volume is too large, which may lead to reduced processing efficiency. The downsampling submodule downsamples the data, reducing the number of data points according to specific rules to reduce the data volume and improve processing efficiency.

[0221] Preferably, when implementing the data preprocessing module, each data point is first verified using the validity verification submodule. For example, the device's signal strength, data integrity, and other parameters can be tested to determine whether the device is functioning properly. If the device's signal strength falls below a certain threshold, or if the data contains a large number of missing or erroneous values, it can be determined that the device is likely malfunctioning. In the data cleaning submodule, statistical methods can be used to identify and remove outliers. For example, the mean and standard deviation of the data can be calculated, and data points outside the range of the mean plus or minus three times the standard deviation are considered outliers and removed. In the formatting submodule, data can be converted to a unified time format, numerical format, etc. based on analysis requirements. For example, all time data can be converted to a unified 24-hour format, and numerical data can be converted to a unified decimal format. In the granularity analysis submodule, an appropriate downsampling method can be selected based on actual requirements and data characteristics. For example, if the data exhibits periodicity, one data point can be retained within each period to reduce the data volume without losing important information. These steps ensure data quality and availability, providing reliable data support for subsequent analysis and scheduling.

[0222] In some embodiments, the analysis period determination module includes:

[0223] The period initialization submodule is used to initialize the period length corresponding to the analysis period;

[0224] The data set extraction submodule is used to determine the target station monitoring time series data set;

[0225] The mean calculation submodule is used to calculate the mean value of resource stocks;

[0226] The extreme value acquisition submodule is used to obtain the peak and valley values ​​of resource consumption;

[0227] The characteristic value calculation submodule is used to calculate the characteristic value of a period based on the mean value of resource inventory, peak value of resource consumption and valley value of resource consumption;

[0228] a period determination submodule, configured to determine an initial analysis period as the analysis period in response to the resource inventory mean being less than or equal to the dynamic safety threshold;

[0229] The iterative processing submodule is used to increase the length of the initial analysis period and re-execute the processing flow in response to the resource inventory mean being greater than the dynamic safety threshold.

[0230] It should be noted that the analysis period determination module in the present invention is designed to find a suitable time period from the preprocessed site monitoring time series data set, so that the data within this time period can be analyzed in detail, thereby accurately assessing the status of emergency resources. This module includes a time period initialization submodule, a data set extraction submodule, a mean calculation submodule, an extreme value acquisition submodule, an eigenvalue calculation submodule, a time period determination submodule, and an iterative processing submodule. These submodules work together to determine the final analysis period through a series of calculations and judgments. This process ensures that the selection of the analysis period is based on the actual characteristics and requirements of the data, thereby improving the accuracy and reliability of the analysis.

[0231] Specifically, the period initialization submodule is used to initialize the length of the analysis period. This initial length can be set based on historical data or experience, for example, to one day or one week. The dataset extraction submodule is used to extract the target station monitoring time series dataset from the preprocessed station monitoring time series dataset. This data will be used for subsequent analysis. The mean calculation submodule is used to calculate the mean resource inventory corresponding to the target station monitoring time series dataset. This mean reflects the overall resource level. The extreme value acquisition submodule is used to obtain the peak and valley resource consumption values ​​in the target station monitoring time series dataset. These two values ​​can help understand the fluctuations in resource consumption. The characteristic value calculation submodule is used to calculate period characteristic values ​​based on the mean resource inventory, peak resource consumption, and valley resource consumption values. These characteristic values ​​are used to assess resource status. The period determination submodule is used to determine whether the initial analysis period should be designated as the final analysis period based on a comparison of the mean resource inventory value with the dynamic safety threshold. The iterative processing submodule is used to increase the length of the initial analysis period if the mean resource inventory value exceeds the dynamic safety threshold and re-execute the above steps until a qualified analysis period is found.

[0232] Preferably, when implementing the analysis period determination module, the period initialization submodule first sets an initial period length, for example, one day. The dataset extraction submodule then extracts data from the preprocessed site monitoring time-series dataset for that day as the target site monitoring time-series dataset. The mean calculation submodule then calculates the average resource inventory for that day, and the extreme value acquisition submodule obtains the highest and lowest points of resource consumption. The characteristic value calculation submodule calculates the characteristic value of the period based on this data. For example, the mean resource inventory value can be compared with a dynamic safety threshold. If the mean resource inventory value is lower than or equal to the dynamic safety threshold, the period determination submodule determines that day as the analysis period. If the mean resource inventory value is higher than the dynamic safety threshold, the iterative processing submodule increases the initial period length, for example, to two days, then re-extracts data and performs the above calculations until a period that meets the requirements is found. In practical applications, parameters such as the initial period length and dynamic safety threshold can be adjusted according to different emergency scenarios and resource types to ensure that the selected analysis period accurately reflects the actual resource status.

[0233] Reference below Figure 3 , which shows a schematic structural diagram of an electronic device 300 suitable for implementing some embodiments of the present invention. The electronic devices in some embodiments of the present invention may include, but are not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 3 The terminal device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0234] like Figure 3 As shown, the electronic device 300 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. Various programs and data required for the operation of the electronic device 300 are also stored in the RAM 303. The processing device 301, the ROM 302, and the RAM 303 are connected to each other via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0235] Typically, the following devices may be connected to the I / O interface 305: an input device 306 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 307 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 308 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 309. The communication device 309 may allow the electronic device 300 to communicate with other devices wirelessly or by wire to exchange data. Figure 3 The electronic device 300 is shown with various devices, but it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed instead. Figure 3 Each block shown in the figure may represent one device, or may represent multiple devices as needed.

[0236] Furthermore, the storage medium of the embodiment of the present application stores program instructions that can implement all the above methods, wherein the program instructions can be stored in the above storage medium in the form of a software product, including a number of instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, or a terminal device such as a computer, a server, a mobile phone, or a tablet.

[0237] The above descriptions are merely some preferred embodiments of the present invention and an illustration of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present invention is not limited to the technical solutions formed by a specific combination of the above-mentioned technical features, but should also encompass other technical solutions formed by any combination of the above-mentioned technical features or their equivalents without departing from the above-mentioned inventive concept. For example, a technical solution formed by mutually replacing the above-mentioned features with (but not limited to) technical features having similar functions disclosed in the embodiments of the present invention.

Claims

1. A method for analyzing emergency resource data based on a smart station, applied to an emergency command platform or a resource dispatching system, characterized in that: include: Collect real-time site monitoring time series data sets for emergency resources; Performing data preprocessing on the real-time station monitoring time series data set to obtain a preprocessed station monitoring time series data set; Determining an analysis period based on the pre-processed station monitoring time series data set; determining a resource status label according to a target station monitoring time series data set, wherein the target station monitoring time series data set is pre-processed station monitoring data in the pre-processed station monitoring time series data set and located within the analysis period, and the resource status label represents a shortage risk level of the emergency resource; In response to determining that the resource status tag is a first-level shortage tag, generating enhanced scheduling policy information for emergency resources as updated scheduling policy information based on the current scheduling policy information corresponding to the smart station and the resource status tag; In response to determining that the resource status tag is a normal reserve tag, generating maintenance reserve strategy information for emergency resources according to the current scheduling strategy information and the resource status tag as updated scheduling strategy information; Dynamically schedule emergency resources according to the current scheduling policy information or the updated scheduling policy information.

2. The method according to claim 1, characterized in that Dynamically scheduling emergency resources according to the current scheduling policy information or the updated scheduling policy information includes: Determining resource anomaly information based on the current scheduling policy information or the updated scheduling policy information, the resource status label, and the scheduling anomaly prediction model; In response to determining that the resource abnormality information indicates that the current scheduling policy information or the updated scheduling policy information has a control abnormality, determining a scheduling instruction corresponding to the resource abnormality information in the abnormal scheduling decision tree; According to the scheduling instructions, the target resource warehouse is controlled for entry and exit, or a backup resource channel is activated.

3. The method according to claim 2, characterized in that The performing data preprocessing on the real-time station monitoring time series data set to obtain a preprocessed station monitoring time series data set includes: For each real-time station monitoring data in the real-time station monitoring time series data set, the following processing steps are performed: Performing equipment validity verification on the real-time site monitoring data; In response to determining that the real-time station monitoring data passes the validity check, performing data cleansing on the real-time station monitoring data to obtain cleansed station monitoring data; Performing data formatting on the cleaned station monitoring data to obtain formatted station monitoring data; Determining the collection granularity corresponding to the real-time site monitoring time series data set; In response to determining that the acquisition granularity is not a reference acquisition granularity and that the acquisition granularity is smaller than the reference acquisition granularity, the formatted station monitoring data sequence is downsampled at the reference acquisition granularity to obtain the preprocessed station monitoring time series data set.

4. The method according to claim 3, characterized in that Determining an analysis period based on the pre-processed station monitoring time series data set includes: Initializing the time period length corresponding to the analysis time period to obtain an initial analysis time period; Based on the initial period length and the pre-processed station monitoring time series data set, perform the following steps: Determine the target station monitoring time series data set; Calculate the mean value of resource inventory corresponding to the target station monitoring time series data set; Obtain the resource consumption peak and valley values ​​in the target station monitoring time series data set; Calculate the period characteristic value of the initial analysis period based on the resource inventory mean, resource consumption peak value and resource consumption valley value; In response to the resource inventory mean being less than or equal to the dynamic safety threshold, determining an initial analysis period as the analysis period; In response to the resource inventory mean being greater than the dynamic safety threshold, the length of the initial analysis period is increased and the above steps are executed again.

5. The method according to claim 4, characterized in that Determining the resource status tag based on the target station monitoring time series data set includes: Calculate the deviation rate of real-time inventory relative to the safety stock threshold; Calculate the ratio of the sum of the incremental demand for multiple types of materials to the demand response time; The deviation rate and the ratio are weighted and summed according to a preset weight coefficient to obtain a resource scarcity index; In response to the resource shortage index continuously exceeding the first risk threshold for a first time period threshold, determining the resource status label as a first-level shortage label; In response to the resource shortage index being continuously lower than the safety threshold for a second period threshold, the resource status tag is determined to be a normal reserve tag.

6. The method according to claim 5, characterized in that The scheduling anomaly prediction model is constructed by the following steps: Calculate the absolute difference between the outbound and inbound quantities, and normalize the difference with the maximum inbound quantity to obtain the abnormal characteristics of resource flow; The multi-warehouse coordination coefficient is obtained by summing the ratio of each warehouse's inventory to its distance to the disaster site and dividing it by the maximum inventory. The resource flow anomaly characteristics and multi-warehouse coordination coefficient are input into the temporal convolutional network model, and the scheduling anomaly probability is output.

7. The method according to claim 2, characterized in that The abnormal scheduling decision tree includes the following decision steps: Determine whether the inventory turnover rate is lower than the safety turnover rate; If it is lower than the safe turnover rate, further detection is performed to see whether the abnormal characteristics of resource flow exceed the maximum abnormal threshold; If the maximum abnormal threshold is exceeded, a first-level dispatch instruction is generated to execute cross-regional transfers; If the maximum abnormal threshold is not exceeded, a secondary scheduling instruction is generated to activate the backup warehouse.

8. The method according to claim 7, characterized in that Also includes path optimization mechanism: Obtain the real-time length, speed and load capacity of each transport section; Calculate the weighted sum of transportation time and loading time for each section; The optimization algorithm is used to solve the minimum value of the weighted sum of all road sections and determine the optimal transportation route.

9. The method according to claim 8, characterized in that Enhanced scheduling strategies include: Calculate the weighted comprehensive value of resource coverage utility and transportation cost saving rate; Under the constraint that the total amount of resource allocation to each warehouse does not exceed its inventory, A solution for maximizing the weighted comprehensive value is solved by linear programming.

10. The method according to claim 2, characterized in that Also includes 3D site visualization: Integrate GPS positioning data, BIM model data and IoT sensor data to build a digital twin model of the station; According to the coordinates and inventory of each resource point, the resource distribution heat map is calculated using the Gaussian kernel function.

11. The method according to claim 10, characterized in that The method further comprises: Generate a scheduling instruction set containing resource ID, target location and required arrival time; Send the instruction set to the intelligent transport vehicle through the 5G private network; Real-time monitoring of the difference between actual arrival time and planned arrival time; When the difference exceeds the allowable deviation threshold, the transportation path is replanned.

12. An emergency resource data analysis system based on a smart station, applied to an emergency command platform or resource dispatch system, comprising: Data acquisition module, used to collect real-time site monitoring time series data sets for emergency resources; A data preprocessing module, configured to perform data preprocessing on the real-time station monitoring time series data set to obtain a preprocessed station monitoring time series data set; An analysis period determination module, configured to determine an analysis period based on the pre-processed station monitoring time series data set; a label generation module, configured to determine a resource status label based on a target station monitoring time series data set, wherein the target station monitoring time series data set is data within the analysis period in the preprocessed station monitoring time series data set; A first policy generation module is configured to generate enhanced scheduling policy information as updated scheduling policy information based on the current scheduling policy information and the resource status label corresponding to the smart station in response to the resource status label being a first-level shortage label; a second policy generating module, configured to generate, in response to the resource status tag being a normal reserve tag, maintenance reserve policy information according to the current scheduling policy information and the resource status tag as updated scheduling policy information; The dynamic scheduling execution module is used to dynamically schedule emergency resources according to current scheduling policy information or updated scheduling policy information.

13. The system according to claim 12, wherein the dynamic scheduling execution module comprises: Anomaly detection submodule, used to determine resource anomaly information based on current scheduling policy information or updated scheduling policy information, resource status labels and scheduling anomaly prediction model; A decision tree processing submodule, configured to determine a corresponding scheduling instruction in an abnormal scheduling decision tree in response to resource abnormality information indicating a control abnormality; The resource control submodule is used to control the entry and exit of the target resource warehouse according to scheduling instructions, or to activate backup resource channels.

14. The system according to claim 13, wherein the data pre-processing module comprises: The validity check submodule is used to perform equipment validity check on each data in the real-time site monitoring time series data set; A data cleaning submodule is used to remove outliers from the data in response to the data passing the validity check; The formatting processing submodule is used to format the cleaned data; The granularity analysis submodule is used to determine the collection granularity corresponding to the real-time station monitoring time series data set; The downsampling processing submodule is configured to downsample the data at the reference acquisition granularity in response to the acquisition granularity being smaller than the reference acquisition granularity.

15. The system according to claim 14, wherein the analysis period determination module comprises: The period initialization submodule is used to initialize the period length corresponding to the analysis period; The data set extraction submodule is used to determine the target station monitoring time series data set; The mean calculation submodule is used to calculate the mean value of resource stocks; The extreme value acquisition submodule is used to obtain the peak and valley values ​​of resource consumption; The characteristic value calculation submodule is used to calculate the characteristic value of a period based on the mean value of resource inventory, peak value of resource consumption and valley value of resource consumption; a period determination submodule, configured to determine an initial analysis period as the analysis period in response to the resource inventory mean being less than or equal to the dynamic safety threshold; The iterative processing submodule is used to increase the length of the initial analysis period and re-execute the processing flow in response to the resource inventory mean being greater than the dynamic safety threshold.

16. An electronic device comprising: one or more processors; a storage device having one or more programs stored thereon, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 11.

17. A computer-readable medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the method according to any one of claims 1 to 11 is implemented.

Images