Intelligent scheduling optimization method and system based on uninterrupted fire water supply

Abstract

The application discloses an intelligent scheduling optimization method and system based on uninterrupted fire-fighting water supply, and the method comprises the following steps: obtaining a disaster event identifier, collecting alarm records, on-site feedback information and fire vehicle state data, forming water supply demand data and generating a demand sequence through a preset disaster model; determining a vehicle set to be replenished based on the residual water volume, the instantaneous water flow and the demand sequence, and retrieving available water sources to obtain a water source candidate set; calculating the arrival time, the water replenishment capacity and the water replenishment duration of the vehicle to the water source, threshold filtering the water replenishment candidate set and generating a water replenishment instruction draft; solving a water supply and water replenishment rotation scheduling scheme that meets the continuous water supply constraint based on the demand sequence, the vehicle state data and the water replenishment instruction draft; obtaining on-site feedback after execution, updating the candidate set according to the availability index threshold and rolling the scheduling scheme. The method can reduce the supply-demand gap and the instruction failure risk.

Description

Technical Field

[0001] This invention relates to the field of intelligent scheduling technology, specifically to an intelligent scheduling optimization method and system based on uninterrupted fire water supply. Background Technology

[0002] Existing technologies typically employ experience-based command, static plans, or monitoring and alarm methods oriented towards fixed facilities in fire water supply assurance. These methods involve one-time or phased arrangements for fire truck deployment, water intake point selection, and water supply lines, and post-event monitoring and manual intervention based on indicators such as pressure and flow rate at the fire front. However, during the continuous handling of large-scale fires, fire trucks such as water tankers and pump trucks are generally in a cyclical operation of water intake, water supply, and water intake. Moreover, the water demand at the fire site changes with the stage of the operation, and factors such as road conditions, water availability, queuing, and fire truck malfunctions fluctuate dynamically. This requires dispatch decisions to simultaneously address changes in the remaining water volume of fire trucks, water output rate, arrival time, water replenishment capacity, and the constraints of multiple vehicles supplying and replenishing water within a unified time scale.

[0003] Existing technologies typically fail to organize multi-source information, such as alarm records, on-site feedback information, fire truck status data, road traffic data, and water source status data, into a data set that can be used for dispatch calculations under a unified spatiotemporal reference. They also lack constraints that quantify the difference between demand and supply capacity and trigger water supply and replenishment rotation adjustments using thresholds, and still mainly rely on single dispatch or local adjustments. However, when the water volume of fire trucks decreases simultaneously, water replenishment arrangements overlap, or the status of water sources and roads changes abruptly, dispatching cannot generate executable water replenishment instructions in a timely manner and update the water supply and replenishment sequence synchronously. This can easily lead to problems such as widening supply and demand gaps, water supply interruptions, or failure of dispatch instructions, making it difficult to guarantee uninterrupted water supply at the fire scene.

[0004] In summary, existing technologies lack a scheduling optimization mechanism for fire-fighting circulating water supply scenarios that can generate demand sequences and vehicle status data at a unified time granularity based on multi-source dynamic data, generate water supply and replenishment rotation schemes under continuous water supply constraints, and trigger updates based on on-site feedback.

[0005] Therefore, proposing an intelligent scheduling and optimization method and system based on uninterrupted fire water supply to solve the difficulties of existing technologies is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide an intelligent scheduling and optimization method and system based on uninterrupted fire water supply, so as to solve the shortcomings of the prior art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: an intelligent scheduling and optimization method based on uninterrupted fire water supply, comprising:

[0008] S1, Obtain fire vehicle status data and water supply demand data associated with disaster event identifiers; Generate a demand sequence bound to disaster event identifiers based on water supply demand data;

[0009] S2, based on the status data and demand sequence of fire trucks, determine the set of vehicles to be replenished with water, and retrieve available water sources for the vehicles to be replenished with water to obtain a candidate set of water sources;

[0010] S3, calculate the arrival time and water replenishment capacity of the vehicles to be replenished to each water source candidate, filter the water source candidate set from the water source candidate set according to the threshold comparison, and generate a water replenishment instruction draft corresponding to the water replenishment candidate set.

[0011] S4, based on the demand sequence, fire truck status data and water replenishment instruction draft, solves for a water supply-replenishment rotation scheduling scheme that satisfies the continuous water supply constraint.

[0012] S5: Execute the scheduling plan and obtain on-site feedback; when the availability index corresponding to the on-site feedback does not meet the preset threshold, update the water source candidate set or the set of vehicles to be replenished with water, and return to execute S3 and S4 to update the scheduling plan.

[0013] Furthermore, methods for generating demand sequences bound to disaster event identifiers include:

[0014] The system acquires disaster event identifiers and related alarm records and on-site feedback information. It extracts disaster location, disaster type, disaster scale, response stage markers, and start time to form water supply demand data. Based on the disaster event identifiers, it acquires fire truck status data, which includes at least the fire truck identifier, current location, and remaining water volume. It also adds a collection timestamp to the fire truck status data to form a time-sorted sequence of fire truck status data. It performs consistency processing on the fire truck status data sequence to obtain a standard sequence of fire truck status. Based on the water supply demand data, it outputs a demand sequence through a preset disaster model.

[0015] Furthermore, methods for pre-setting disaster models include:

[0016] Obtain a historical disaster sample set. Each historical disaster sample in the historical disaster sample set shall include at least the historical disaster type, historical disaster scale, historical response phase sequence, historical water supply record and historical water use record.

[0017] By aligning historical water supply records and historical water use records with a uniform time granularity, a historical demand sequence is obtained as a training label.

[0018] Using historical disaster types, scales, and response phases as input features, a supervised learning regression training method is employed to iteratively update the model parameters. The demand error function is used as the objective function, and the demand error is compared with an error threshold. If the demand error exceeds the error threshold, iterative updates continue until a stopping condition is met, at which point the model is solidified to obtain the preset disaster model.

[0019] Furthermore, the methods for determining the assembly of vehicles to be replenished with water include:

[0020] Obtain fire truck status data and demand sequences associated with disaster event identifiers, and align them with a uniform time granularity;

[0021] The sustainable water supply duration of each fire truck is calculated based on the remaining water volume and the instantaneous water flow rate; the demand pressure label is obtained based on the comparison between the demand sequence and the threshold.

[0022] The water replenishment trigger flag is obtained by comparing the continuous water supply duration of a single vehicle with the duration threshold. The water replenishment trigger flag is then merged with the demand pressure flag. The fire truck identifiers that meet the preset conditions are written into the set of vehicles to be replenished.

[0023] Furthermore, methods for retrieving a candidate set of available water sources for vehicles awaiting water replenishment include:

[0024] Obtain a list of water sources associated with the disaster event identifier. The list should include at least the water source identifier, water source location, water supply capacity, and availability marker.

[0025] For each fire truck identifier in the water replenishment vehicle set, an initial water source set is obtained based on its current location within a preset search radius;

[0026] The availability flags of each water source in the initial water source set are compared with the availability threshold. Water sources that do not meet the availability threshold are removed from the initial water source set to obtain the available water source set.

[0027] The set of candidate water sources corresponding to the fire truck identification is determined based on the set of available water sources.

[0028] Furthermore, methods for obtaining the water replenishment candidate set include:

[0029] Obtain a candidate set of water sources that corresponds one-to-one with each fire truck identifier in the set of vehicles to be replenished with water; for each fire truck identifier, calculate the arrival time based on the current location and the water source location of each water source identifier in the candidate set; calculate the amount of water to be replenished based on the upper limit parameter of water volume and the remaining water volume, and determine the water replenishment capacity and calculate the water replenishment duration based on the water supply capacity of the water source identifier; compare the arrival time with the arrival time threshold and the water replenishment capacity with the water replenishment capacity threshold, so as to write the water source identifier that meets the threshold conditions into the water replenishment candidate set corresponding to the fire truck identifier.

[0030] Furthermore, the method for generating a draft water replenishment instruction corresponding to the water replenishment candidate set includes:

[0031] Obtain the water replenishment candidate set corresponding to each fire truck identification, and calculate the comprehensive cost based on the arrival time and water replenishment duration of each water source identification in the water replenishment candidate set;

[0032] The overall cost is compared with the cost threshold to determine the preferred order of each water source identifier in the water replenishment candidate set; a draft water replenishment instruction is generated based on the preferred order, and the draft water replenishment instruction includes at least the preferred water source identifier, the route identifier, the estimated arrival time, the estimated water replenishment amount, and the estimated water replenishment duration.

[0033] Furthermore, methods for obtaining a water supply-replenishment rotation scheduling scheme that satisfies the continuous water supply constraint include:

[0034] Obtain the demand sequence, fire truck status data, and draft water replenishment instructions bound to disaster event identifiers, and align them with a unified time granularity;

[0035] The water supply capacity sequence is calculated based on fire truck status data, and the water replenishment action sequence is calculated based on the draft water replenishment order.

[0036] Within each time granularity, the supply-demand difference is calculated based on the demand sequence and the water supply capacity sequence and compared with the gap threshold to determine the water supply vehicle set;

[0037] Within each time granularity, the set of water replenishment vehicles is determined based on the water replenishment action sequence;

[0038] Based on the water supply vehicle set and the water replenishment vehicle set, a water supply-replenishment rotation scheduling scheme is output. The water supply-replenishment rotation scheduling scheme includes at least the water supply sequence and water replenishment sequence of multiple fire trucks, and is associated with disaster event identifiers.

[0039] Furthermore, methods for updating the candidate set of water sources or the set of vehicles awaiting water replenishment include:

[0040] Obtain the dispatch plan associated with the disaster event identifier and expand it into an action instruction sequence at a uniform time granularity, so that each fire truck identifier can execute the water supply sequence and water replenishment sequence according to the action instruction sequence;

[0041] Obtain on-site feedback associated with disaster event identifiers and align it with a unified time granularity;

[0042] Based on on-site feedback, usability indicators are calculated and compared with preset thresholds. When the comparison does not meet the requirements, the candidate set is updated.

[0043] When the availability index is a water source availability index or a road accessibility index and the corresponding comparison is not met, the corresponding water source identifier is removed from the water source candidate set and the water source candidate set is updated. When the availability index is a vehicle water supply capacity index and the corresponding comparison is not met, the corresponding fire truck identifier is written into the water supply vehicle set to update the water supply vehicle set.

[0044] The intelligent scheduling and optimization system based on uninterrupted fire water supply implements the aforementioned intelligent scheduling and optimization method based on uninterrupted fire water supply. The system includes:

[0045] Data acquisition module: used to acquire fire vehicle status data and water supply demand data associated with disaster event identifiers; and to generate demand sequences bound to disaster event identifiers based on water supply demand data;

[0046] Candidate generation module: Based on fire truck status data and demand sequence, determine the set of vehicles to be replenished with water, and retrieve available water sources for the vehicles to be replenished with water to obtain a candidate set of water sources;

[0047] Instruction generation module: It is used to calculate the arrival time and water replenishment capacity of the vehicles to be replenished to each water source candidate, filter the water source candidate set from the water source candidate set according to the threshold comparison, and generate a water replenishment instruction draft corresponding to the water replenishment candidate set.

[0048] Scheduling solution module: Based on the demand sequence, fire truck status data and water replenishment instruction draft, it solves for a water supply-replenishment rotation scheduling scheme that satisfies the continuous water supply constraint;

[0049] Feedback Update Module: Used to execute the scheduling plan and obtain on-site feedback; when the availability index corresponding to the on-site feedback does not meet the preset threshold, update the water source candidate set or the set of vehicles to be replenished, and return to execute S3 and S4 to update the scheduling plan.

[0050] The technical effects and advantages of the intelligent scheduling and optimization method and system based on uninterrupted fire water supply provided by this invention are as follows:

[0051] By organizing alarm records, on-site feedback information, fire truck status data, water source lists, and road traffic data associated with disaster event identifiers at a unified time granularity, and performing consistent processing on fire truck status data to form a standard sequence of fire truck status, and then mapping water supply demand data such as disaster location, disaster type, disaster scale, response stage markers, and start time to a demand sequence bound to disaster event identifiers through a preset disaster model, the originally scattered demand-side and supply-side information is transformed into an aligned and computable serialized input. This makes subsequent quantification of gaps based on supply-demand differences and threshold-triggered updates executable steps. This allows dispatch decisions to no longer rely on experience-based criteria or static plans, but instead obtain key quantities such as demand, demand increase, vehicle remaining water volume, and real-time effective water output rate at each time granularity. This reduces the estimation deviations of arrival time, sustainable water supply duration, and water replenishment duration caused by data spatiotemporal inconsistencies, and improves the ability to proactively identify supply-demand gaps and the consistency of response.

[0052] Based on demand sequence and fire truck status data, the sustainable water supply duration and demand pressure marker for a single vehicle are calculated. The water replenishment trigger marker and demand pressure marker are merged to obtain a set of vehicles to be replenished. Then, the vehicles to be replenished are searched within a preset search radius to retrieve the initial water source set and filtered into a usable water source set according to the availability marker. Furthermore, the suitability of the water source and the candidate water source set are calculated by combining the arrival time obtained by accumulating the road segment passage cost, water supply capacity, and occupancy penalty factors. Thus, the selection of water replenishment targets and water sources is transformed from a single-point judgment to a joint screening that is simultaneously constrained by the remaining water consumption rate, demand increase, road dynamics, and water source dynamics. During the time window when the vehicle water volume decreases synchronously or the demand increases rapidly, vehicles that meet the trigger conditions can be included in the set of vehicles to be replenished in advance, and water sources with insufficient availability or excessive passage cost can be eliminated. This reduces the probability of water replenishment failure caused by ineffective travel, queuing, and route delays, and improves the matching degree between water replenishment operations and changes in the fire stage.

[0053] Based on the candidate water source set, the arrival time, water replenishment capacity, and water replenishment duration of the vehicles to be replenished to each candidate water source are calculated. The candidate water source set is then filtered using arrival time thresholds, water replenishment capacity thresholds, and revenue thresholds, generating a draft water replenishment instruction containing water source identifiers, path identifiers, estimated arrival times, estimated water replenishment amounts, and estimated water replenishment durations. The demand sequence, fire truck status data, and the draft water replenishment instruction are then aligned to form a scheduling input table. The water supply capacity sequence, water replenishment action sequence, supply-demand difference sequence, and rotation risk are calculated. Under continuous water supply constraints and rotation risk constraints, the multi-vehicle water supply time sequence is solved and output. The water replenishment sequence is implemented; after execution, feedback on water source status, road status, and fire truck status is introduced to calculate availability indicators, triggering an update of the candidate water source set or the set of vehicles to be replenished, and returning to recalculate the draft water replenishment instruction and rotation scheduling scheme; so that the scheduling output not only provides the water supply or replenishment status mark of the vehicle, but also provides the executable path and time fields, and can immediately correct the input set and recalculate the sequence according to threshold conditions in the event of water source occupation, sudden changes in road traffic cost, and decline in vehicle water supply capacity, reducing the risk of instruction non-executability and water supply chain interruption, and ensuring that the continuous water supply target is sustainably met under dynamic disturbances. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0055] Figure 1 This is a schematic diagram of the intelligent scheduling and optimization method based on uninterrupted fire water supply of the present invention.

[0056] Figure 2 This is a flowchart of the intelligent scheduling and optimization system module based on uninterrupted fire water supply of the present invention. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0058] It should be noted that when a component is said to be "fixed to" another component, it can be directly attached to the other component or there may be an intervening component. When a component is said to be "connected to" another component, it can be directly connected to the other component or there may be an intervening component.

[0059] Example 1

[0060] Please see Figure 1 As shown, this embodiment provides an intelligent scheduling and optimization method based on uninterrupted fire water supply, including:

[0061] S1, Obtain fire truck status data and water supply demand data associated with the disaster event identifier. The fire truck status data includes at least the fire truck identifier, current location, and remaining water volume. Generate a demand sequence bound to the disaster event identifier based on the water supply demand data; as follows:

[0062] Obtain disaster event identifiers and alarm records and on-site feedback information associated with the disaster event identifiers. Extract disaster location, disaster type, disaster scale, response stage markers and start time to form basic fields for water supply demand data. Compare the percentage of missing items in the basic fields with the missing rate threshold. If the percentage of missing items is not less than the missing rate threshold, trigger supplementary data entry to obtain complete water supply demand data.

[0063] Based on the disaster event identifier, fire truck status data is obtained from the vehicle-mounted positioning and water volume acquisition unit. The fire truck status data includes at least the fire truck identifier, current location, and remaining water volume. A collection timestamp is added to the fire truck status data, and the fire truck status data is associated with the disaster event identifier to form a fire truck status data sequence sorted by time.

[0064] The fire truck status data sequence is processed to obtain a standard sequence of fire truck status. The current position is converted into a unified coordinate expression according to a preset coordinate reference. The upper limit parameter of water volume corresponding to the fire truck identification is obtained, and the remaining water volume is converted into a remaining water volume value in a unified unit of measurement. The difference value of remaining water volume between adjacent time moments is calculated. The difference value is compared with the abnormal difference threshold. When the difference value is not less than the abnormal difference threshold, the remaining water volume value at that time moment is replaced by the interpolation result of the adjacent time moment.

[0065] Demand sequences are generated based on water supply demand data and bound to disaster event identifiers. These demand sequences are output from a pre-defined disaster model. The steps for obtaining the pre-defined disaster model include:

[0066] A historical disaster sample set is obtained, and each historical disaster sample in the sample set includes at least the historical disaster type, historical disaster scale, historical response stage sequence, historical water supply record, and historical water use record. The historical water supply record and historical water use record are aligned with a uniform time granularity to obtain a historical demand sequence as training labels. The historical disaster type, scale, and historical response stage sequence are used as input features, and a supervised learning regression training method is used to iteratively update the model parameters, with the demand error function as the objective quantity. The demand error is compared with the error threshold. If the demand error is greater than the error threshold, iterative updates continue until the stopping condition is met, and then the preset disaster model is solidified.

[0067] The supervised learning regression training method can be implemented using at least one of the following regression algorithm types: linear regression model, support vector regression model, random forest regression model, gradient boosting regression model, or multilayer feedforward neural network regression model; the demand error function is used to measure the difference between the model output demand sequence and the training label, and can be in the form of mean squared error, mean absolute error, or weighted error; when using the weighted error form, the difference between the predicted demand value and the label demand value is calculated for each time granularity, and stage weights are set according to the disposal stage markers, while peak weights are set for the demand peak interval to increase the penalty intensity of the deviation between the high-intensity stage and the peak demand; the model parameter iterative update can adopt gradient descent or quasi-Newton optimization methods, and the stopping conditions include at least: the demand error is not greater than the error threshold, the number of iterations reaches the preset upper limit, or the validation set error no longer decreases in several consecutive iterations; after stopping, the preset disaster model is solidified.

[0068] Based on a pre-defined disaster model, water supply demand data is input to output a demand sequence, and the demand sequence is bound and stored with disaster event identifiers.

[0069] Align the fire truck status standard sequence and the demand sequence at a unified time granularity to generate a supply-demand alignment sequence; within each time granularity, calculate the sustainable water supply duration based on the remaining water volume value, compare the sustainable water supply duration with the duration threshold to obtain a water supply tension marker; associate the water supply tension marker with the disaster event identifier so that subsequent steps can determine the set of vehicles to be replenished based on the water supply tension marker.

[0070] To facilitate understanding, this step outputs corresponding instance data to illustrate how this step is executed, as follows:

[0071] A fire occurred in a warehouse building in a certain city. The command center generated a disaster event identifier (E20260215A) for this fire. Upon receiving the alarm, the system retrieved alarm records and on-site feedback information associated with the disaster event identifier E20260215A. The system extracted the following basic fields for water supply demand data: the disaster location was a warehouse park on a certain road in a certain district; the disaster type was a building fire; the disaster scale was an area of ​​approximately 10,000 square meters; the response stage was marked as the initial control stage; and the start time was 02:10. The system then calculated the percentage of missing items in these basic fields. It was found that the missing disaster scale field accounted for 0.2% of the missing items. Comparing 0.2 with the missing item rate threshold of 0.1, supplementary data was triggered when the percentage of missing items was not less than the threshold. Supplementary data was then obtained, showing a disaster scale of approximately 10,000 square meters, thus providing complete water supply demand data.

[0072] Subsequently, based on the disaster event identifier E20260215A, fire truck status data was obtained from the vehicle-mounted positioning and water volume acquisition units. Assuming the deployed vehicles included two water tankers, identified as V01 and V02, the acquired fire truck status data included at least their current location and remaining water volume: At 02:15, V01's current location was at the north gate of the park, with 8 cubic meters of remaining water; V02's current location was at the west gate of the park, with 10 cubic meters of remaining water; at 02:20, V01's current location was at the east side of the warehouse, with 6 cubic meters of remaining water; V02's current location was at the west side of the warehouse, with 9 cubic meters of remaining water. The data collection timestamps 02:15 and 02:20 were added to the above fire truck status data, and the fire truck status data was associated with the disaster event identifier E20260215A, forming a time-sorted sequence of fire truck status data.

[0073] Next, the fire truck status data sequence is standardized to obtain a standard fire truck status sequence. The current positions of V01 and V02 are converted into a unified coordinate expression according to a preset coordinate reference, such as a set of coordinate points under the same map coordinate system. The upper limit parameters of water volume corresponding to the fire truck identifiers V01 and V02 are obtained, with the upper limit of water volume for V01 set to 10 cubic meters and for V02 set to 12 cubic meters. The remaining water volume is converted into a remaining water volume value in a unified unit of measurement, keeping the unit to cubic meters. The difference in remaining water volume between adjacent times is calculated: the difference for V01 between 02:15 and 02:20 is 2 cubic meters, and the difference for V02 is 1 cubic meter. The difference value is compared with the abnormal difference threshold of 3 cubic meters. If the difference value is not less than the abnormal difference threshold, the remaining water volume value at that time is replaced by the interpolation result of the adjacent time. Since the difference values ​​of V01 and V02 are both less than the abnormal difference threshold, the original remaining water volume values ​​are retained, thus obtaining the standard fire truck status sequence.

[0074] Then, based on the water supply demand data, a demand sequence bound to the disaster event identifier E20260215A is generated. The demand sequence is output by a preset disaster model. The steps to obtain the preset disaster model are as follows: Obtain a historical disaster sample set. Each historical disaster sample in the historical disaster sample set includes at least the historical disaster type, historical disaster scale, historical response stage sequence, historical water supply record, and historical water consumption record. Taking one historical disaster sample as an example, its historical disaster type is a warehouse fire, its historical disaster scale is an area of ​​approximately 8,000 square meters, its historical response stage sequence includes the initial control stage and the full suppression stage, its historical water supply record provides the water supply volume every five minutes, and its historical water consumption record provides the water consumption every five minutes. Align the historical water supply record and historical water consumption record of this historical disaster sample with a uniform time granularity of five minutes to obtain the historical demand sequence. Training labels, for example, in the first six time granularities of the initial control phase, are 20, 22, 25, 26, 24, and 23 cubic meters per five minutes, respectively. Using historical disaster types, historical disaster scales, and historical response phase sequences as input features, a supervised learning regression training method is employed to iteratively update the model parameters, with the demand error function as the objective. The demand error is compared to an error threshold; if the demand error exceeds the threshold, iterative updates continue until a stopping condition is met, at which point the model is solidified to obtain the preset disaster model. Based on the preset disaster model, the current water supply demand data is input to output a demand sequence. For example, with a time granularity of five minutes, the demand sequence starting at 02:10 is 18, 20, 22, 23, and 23 cubic meters per five minutes, and this demand sequence is bound and stored with the disaster event identifier E20260215A.

[0075] Finally, the fire truck status standard sequence and demand sequence are aligned with a unified time granularity to generate a supply-demand aligned sequence. Let's assume the demand at 02:15 is 20 cubic meters per five minutes, and the demand at 02:20 is 22 cubic meters per five minutes. For each time granularity, the sustainable water supply duration is calculated based on the remaining water volume: for example, dividing the remaining water volume by the current outflow rate yields the sustainable water supply duration. If the outflow rate of V01 at 02:15 is 0.5 cubic meters per minute, then the sustainable water supply duration for V01 is 16 minutes; if the outflow rate of V02 is 0.4 cubic meters per minute, then the sustainable water supply duration for V02 is 25 minutes. The sustainable water supply duration is compared with a duration threshold of 20 minutes to obtain a water shortage flag: if the sustainable water supply duration of V01 is less than the duration threshold, a true water shortage flag is generated; if the sustainable water supply duration of V02 is not less than the duration threshold, a false water shortage flag is generated. Associate the water shortage marker with the disaster event identifier E20260215A so that subsequent steps can identify the set of vehicles needing water replenishment based on the water shortage marker.

[0076] S2, based on the status data and demand sequence of fire trucks, determines the set of vehicles to be replenished with water, and retrieves available water sources for the vehicles to be replenished to obtain a candidate set of water sources.

[0077] The steps for determining the group of vehicles to be replenished with water include:

[0078] Obtain fire truck status data and demand sequences associated with disaster event identifiers. Group the fire truck status data by fire truck identifier and align it with the unified time granularity of the collection timestamp and demand sequence to obtain the current location, remaining water volume, and instantaneous water flow of each fire truck at each time granularity, as well as the demand and handling stage markers at the same time granularity.

[0079] For each fire truck identification, the sustainable water supply duration is calculated. The sustainable water supply duration is obtained by dividing the remaining water volume by the real-time effective water output rate, where the real-time effective water output rate is obtained by averaging the real-time water output flow rate within a uniform time granularity. When the real-time water output flow rate is missing, the interpolated result of the water output flow rate within an adjacent time granularity for the same fire truck identification is used as a substitute.

[0080] Demand pressure indicators are calculated based on demand sequences. These indicators include phased demand and demand increase. Phased demand is the demand within a uniform time granularity, while demand increase is the difference between demand in adjacent time granularities. Phased demand is compared with demand thresholds, and demand increase is compared with increase thresholds. If any comparison fails, a demand pressure flag is output.

[0081] The system compares the sustainable water supply duration of a single vehicle with a duration threshold to obtain a water replenishment trigger flag. This flag is then merged with the demand pressure flag. Fire trucks whose water replenishment trigger flag and demand pressure flag are both valid are added to the set of vehicles awaiting water replenishment. When the handling phase flag indicates a high-intensity phase, the duration threshold is tightened according to a preset threshold correction rule to trigger an earlier update of the set of vehicles awaiting water replenishment. A specific example is as follows:

[0082] Obtain the fire truck status data and demand sequence associated with the disaster event identifier E20260215A, and group the fire truck status data according to the fire truck identifier. Taking two time granularities, 02:15 and 02:20, as examples, V01's current position at 02:15 is the coordinate point of the north gate of the park, with a remaining water volume of 8 cubic meters and an instantaneous water flow rate of 0.5 cubic meters per minute; V01's current position at 02:20 is the coordinate point of the east side of the warehouse, with a remaining water volume of 6 cubic meters and an instantaneous water flow rate missing; V02's current position at 02:15 is the coordinate point of the west gate of the park, with a remaining water volume of 10 cubic meters and an instantaneous water flow rate of 0.4 cubic meters per minute; V02's current position at 02:20 is the coordinate point of the west side of the warehouse, with a remaining water volume of 9 cubic meters and an instantaneous water flow rate of 0.4 cubic meters per minute. Align the above-mentioned collection timestamps with the five-minute time granularity of the demand sequence. The demand at the same time granularity is 20 cubic meters every five minutes for 02:15 and 22 cubic meters every five minutes for 02:20. The disposal stage is marked as the initial control stage.

[0083] For each fire truck identification, the sustainable water supply duration for a single vehicle is calculated. For V01 at 02:15, the effective instantaneous water output rate is taken as the average instantaneous water flow rate within those five minutes, which is 0.5 cubic meters per minute. Therefore, the sustainable water supply duration for a single vehicle is 8 divided by 0.5, resulting in 16 minutes. Since the instantaneous water flow rate for V01 at 02:20 is missing, the interpolation result at the adjacent time granularity is used as a substitute. The interpolation is taken as 0.5 cubic meters per minute at 02:15 and 0.6 cubic meters per minute returned at 02:25. If the median value of 0.55 cubic meters per minute is taken as the effective water output rate at 02:20, then the continuous water supply time for a single vehicle is 6 divided by 0.55, which gives approximately 10.9 minutes. For V02, the effective water output rate at 02:15 is 0.4 cubic meters per minute, so the continuous water supply time for a single vehicle is 10 divided by 0.4, which gives 25 minutes. If the effective water output rate at 02:20 is still 0.4 cubic meters per minute, then the continuous water supply time for a single vehicle is 9 divided by 0.4, which gives 22.5 minutes.

[0084] Demand pressure indicators are calculated based on demand sequences. The phased demand is taken from the demand within a uniform time granularity: 20 cubic meters per five minutes at 02:15, and 22 cubic meters per five minutes at 02:20. The demand increase is taken from the difference in demand between adjacent time granularities: 2 cubic meters per five minutes at 02:20 relative to 02:15. The phased demand is compared to the demand threshold of 21 cubic meters per five minutes; 02:15 does not meet the threshold, while 02:20 does. The demand increase is compared to the increase threshold of 2 cubic meters per five minutes; the demand increase at 02:20 meets the threshold. Following the rule of outputting a demand pressure flag when any comparison fails, a demand pressure flag is output as true at 02:15 and as false at 02:20.

[0085] The water replenishment trigger flag is obtained by comparing the sustainable water supply duration of a single vehicle with a time threshold. Assuming the time threshold is 20 minutes, if V01's 16 minutes at 02:15 is no more than 20 minutes, the water replenishment trigger flag is true; if V01's approximately 10.9 minutes at 02:20 is no more than 20 minutes, the water replenishment trigger flag is true; if V02's 25 minutes at 02:15 is more than 20 minutes, the water replenishment trigger flag is false; if V02's 22.5 minutes at 02:20 is more than 20 minutes, the water replenishment trigger flag is false. The water replenishment trigger flag is then merged with the demand pressure flag. Fire trucks whose water replenishment trigger flag and demand pressure flag are both true are added to the set of vehicles awaiting water replenishment. (The last sentence appears to be a separate, unrelated statement.) For example, at 02:15, the demand pressure is marked as true, and the V01 water replenishment trigger is marked as true. Therefore, V01 is written into the set of vehicles to be replenished, while the V02 water replenishment trigger is marked as false and is not written. Furthermore, when the treatment stage marker switches from the initial control stage to the high-intensity stage, the duration threshold is tightened according to the preset threshold correction rule. For example, if the duration threshold is tightened from 20 minutes to 15 minutes, then the condition that V01 is not greater than 15 minutes at 02:15 is not true, the water replenishment trigger is turned as false, and the set of vehicles to be replenished is updated accordingly to exclude V01, in order to reflect the changes in triggering conditions under different treatment stages.

[0086] A method for retrieving a candidate set of available water sources for vehicles awaiting water replenishment:

[0087] Obtain a list of water sources associated with disaster event identifiers. The list of water sources should include at least the water source identifier, water source location, water supply capacity, and availability markers. For each fire truck identifier in the set of vehicles to be replenished with water, perform a spatial search within a preset search radius based on its current location to obtain an initial set of water sources.

[0088] Availability screening is performed on the initial water source set. The availability flag of each water source is compared with the availability threshold. Water sources that do not meet the availability threshold are removed from the initial water source set, resulting in a set of available water sources.

[0089] Water source suitability is calculated for the set of available water sources. Water source suitability is composed of travel time utility, water supply capacity utility, and penalty factor. Travel time utility is calculated from the arrival time from the current location to the water source location. The arrival time is obtained by accumulating the road segment passage costs corresponding to the road traffic data. Water supply capacity utility is calculated from the result of comparing the water source's water supply capacity with the capacity threshold. The penalty factor is synthesized from the water source occupancy marker, fault marker, and distance exceedance marker. The weights in water source suitability are determined according to the treatment stage markers and a preset weight table.

[0090] The suitability of a water source is compared with a suitability threshold. Water sources with a suitability score not less than the threshold are added to the candidate water source set corresponding to the fire truck identifier. The candidate water source set is then associated with the fire truck identifier and output. A specific example is as follows:

[0091] Obtain a list of water sources associated with the disaster event identifier E20260215A. The list should include at least the water source identifier, location, supply capacity, and availability marker. Assume the list includes three water sources: W01, W02, and W03. W01 is located at the coordinates of the municipal fire hydrant group on the east side of the park, with a supply capacity of 3 cubic meters per minute and is marked as available. W02 is located at the coordinates of the river intake point on the south side of the park, with a supply capacity of 5 cubic meters per minute and is marked as unavailable. W03 is located at the coordinates of the fire water tank on the north side of the park, with a supply capacity of 2 cubic meters per minute and is marked as available. For the fire truck identifier V01 in the set of vehicles to be replenished with water, obtain its current location as the coordinates of the north gate of the park. Perform a spatial search within a preset search radius of 3 kilometers. If all three water sources fall within the search radius, the initial water source set for V01 is obtained as W01, W02, and W03.

[0092] Availability screening is performed on the initial water source set V01, comparing the availability markers of each water source with the availability threshold. If the availability threshold requires the availability markers to be available, then the availability markers of W02 are unavailable, which does not meet the availability threshold. Therefore, W02 is removed from the initial water source set, resulting in the available water source sets W01 and W03.

[0093] For calculating the water source suitability of the available water source set, to facilitate a one-to-one explanation, it is assumed that the water source suitability is composed of travel time utility, water supply capacity utility, and penalty factor. When the treatment stage is marked as the initial control stage, the preset weight table gives a weight of 0.5 for travel time utility, a weight of 0.4 for water supply capacity utility, and a weight of 0.1 for the penalty factor. For travel time utility, the arrival time from the current location to the water source location is first calculated. The arrival time is obtained by accumulating the road segment toll costs corresponding to the road traffic data. Let's assume the travel time is from the current location (V01)... The path from location V01 to W01 includes road segment A and road segment B. The travel cost for road segment A is 4 minutes, and the travel cost for road segment B is 3 minutes, so the arrival time is 7 minutes. The path from the current location V01 to W03 includes road segment C and road segment D. The travel cost for road segment C is 5 minutes, and the travel cost for road segment D is 5 minutes, so the arrival time is 10 minutes. Converting the arrival time into travel time utility, and assuming that the travel time utility is converted to the reciprocal of the arrival time and normalized to the maximum value, then the travel time utility of W01 is higher than that of W03.

[0094] For the utility of water supply capacity, the water supply capacity of the water source is obtained and compared with the capacity threshold. Assuming the capacity threshold is 2.5 cubic meters per minute, then the water supply capacity of W01 is 3 cubic meters per minute, which meets the capacity threshold, and the utility of water supply capacity is set to 1; the water supply capacity of W03 is 2 cubic meters per minute, which does not meet the capacity threshold, and the utility of water supply capacity is set to 0.6, which is used to reflect the utility reduction when the capacity is insufficient.

[0095] For the penalty factor, the water source occupancy marker, fault marker, and distance exceedance marker are obtained and synthesized. If W01's occupancy marker is unoccupied, fault marker is not faulty, and distance exceedance marker is not exceeded, then the penalty factor is 0. If W03's occupancy marker is occupied, fault marker is not faulty, and distance exceedance marker is not exceeded, then the penalty factor is 0.5, to reflect the increased penalty caused by occupancy. The above-mentioned travel time utility, water supply capacity utility, and penalty factor are weighted and synthesized according to a preset weight table, resulting in W01's water source suitability being higher than W03's water source suitability.

[0096] The suitability of the water source is compared with the suitability threshold. If the suitability threshold is 0.7, then the suitability of W01 is not less than 0.7 and the suitability of W03 is less than 0.7. W01 is written into the water source candidate set corresponding to the fire truck identifier V01. The water source candidate set of V01 is output as W01. The water source candidate set is associated with the fire truck identifier V01 and output for direct use in subsequent steps when generating the water replenishment instruction draft and the water supply and replenishment rotation sequence.

[0097] Methods for obtaining passage costs include:

[0098] Establish a road segment base table, which generates static fields such as road segment identifier, road segment start and end nodes, road segment length, road grade, speed limit, number of lanes, and whether it is closed for construction from electronic maps or road centerline data. This forms the road segment base table, which serves as the benchmark set for calculating traffic costs.

[0099] Acquire dynamic traffic data associated with disaster event identifiers; dynamic traffic data should include at least real-time traffic speed, congestion index, accident and control information, road construction and occupation information, and information on the impact of weather on road surfaces; dynamic traffic data can come from sources such as traffic information services, public security traffic management notices, road cameras and geomagnetic detection, and fire truck positioning trajectory reverse inference, and add timestamps and spatial range identifiers to each dynamic record.

[0100] Dynamic traffic data is mapped to road segment identifiers. Spatial matching is performed on each dynamic record, and its coverage area is geometrically intersected with the road segments in the basic road segment table to obtain a set of associated road segments. When there are multiple dynamic records for the same road segment within the same time granularity, they are weighted and fused according to the credibility weight of the data source, and the dynamic speed value and event tag value of the road segment at that time granularity are output.

[0101] Traffic elements are normalized and thresholded. Using the road segment speed limit as a benchmark, dynamic speed values ​​are converted into speed ratios, which are calculated by dividing the dynamic speed by the speed limit. The speed ratios are compared with congestion thresholds to output congestion flags. Event flag values ​​are compared with closure thresholds to output closure flags. Meteorological impact indicators are compared with meteorological thresholds to output severe weather flags. Subsequent calculations only require reading the flags and ratios.

[0102] Calculate the toll cost; for each road segment, calculate the base travel time at each time granularity, where the base travel time is the road segment length divided by the dynamic speed; then, add a penalty term to obtain the toll cost, which can be calculated in the following form:

[0103] The passage cost equals the base travel time multiplied by the congestion penalty coefficient, plus the event penalty time and the weather penalty time. The congestion penalty coefficient is determined by comparing the speed ratio with the congestion threshold, the event penalty time is determined by the accident and control markers, and the weather penalty time is determined by the severe weather markers. When the closure marker is set, the passage cost is set to the impassable cost, which is used to directly eliminate the road segment during path solving.

[0104] Traffic costs are updated on a rolling basis with a uniform time granularity. When a road segment lacks dynamic traffic data within a continuous update time window and the missing percentage is not less than the missing rate threshold, the basic travel time is calculated by reverting to the speed limit estimate in the road segment's basic table, and a missing penalty item is added to avoid misjudging it as smooth traffic. For example, if a road segment is 1 kilometer long and has a speed limit of 60 kilometers per hour, and the dynamic speed within a certain time granularity is 20 kilometers per hour, then the basic travel time is 1 kilometer divided by 20 kilometers per hour, which equals 0.05 hours, approximately 3 minutes. The speed ratio is 20 divided by 60, which equals 0.33, which is less than the congestion threshold of 0.5. The congestion penalty coefficient is taken as 1.5, so the traffic cost is 3 minutes multiplied by 1.5, which equals 4.5 minutes. If a control event is marked as active during the same time period and the event penalty time is 2 minutes, then the traffic cost is 6.5 minutes. If a closure mark is set, the traffic cost of that road segment is set as an impassable cost and is removed during route selection.

[0105] S3 calculates the arrival time and water replenishment capacity of the vehicles to be replenished to each water source candidate, filters the water source candidate set from the water source candidate set based on threshold comparison, and generates a water replenishment instruction draft corresponding to the water replenishment candidate set.

[0106] The method for calculating the arrival time and replenishment capacity of vehicles needing water replenishment to each candidate water source, and then filtering the candidate water source set from the candidate water source set based on threshold comparison, includes the following steps:

[0107] Obtain the set of vehicles to be replenished with water associated with disaster event identifiers, and obtain the water source candidate set corresponding one-to-one with each fire truck identifier in the set of vehicles to be replenished with water; for each fire truck identifier, obtain its current location, remaining water volume, water volume limit parameters and real-time water flow rate, and obtain the water source location, water supply capacity, available marker and occupied marker of each water source identifier in the water source candidate set, and generate a vehicle water source pairing table.

[0108] For each vehicle water source pair in the vehicle water source pairing table, the arrival time is calculated; using the current location and the water source location as endpoints, a set of candidate paths is retrieved from the road network covered by the road traffic data; for each candidate path, the traffic costs of each road segment included in the path are accumulated segment by segment to obtain the path traffic cost, where the road segment traffic cost is obtained by converting the road segment length and dynamic speed, and adding a road event penalty term; the path cost with the minimum path traffic cost in the candidate path set is determined as the arrival time of the vehicle water source pair, and the corresponding path identifier is recorded.

[0109] The water replenishment capacity is calculated based on the vehicle water source pairing table. For each vehicle water source pair, the amount of water to be replenished is calculated, which is the upper limit of water volume minus the remaining water volume. The available water capacity of the water source identifier is obtained, which is the water replenishment flow rate that can be provided per unit time. When the occupancy mark is established, the available water capacity is converted into effective available water capacity according to the occupancy reduction factor. The effective available water capacity is used as the water replenishment capacity, and the water replenishment duration is calculated. The water replenishment duration is the ratio of the amount of water to be replenished to the effective available water capacity. In order to correlate the water replenishment capacity with the fire status, the real-time water outflow is obtained and converted into the real-time effective water outflow rate. The real-time effective water outflow rate and the effective available water capacity are used to construct a water replenishment benefit index, which is used to reflect the increase in sustainable water supply duration after water replenishment.

[0110] The candidate water supply set is filtered based on threshold comparisons. Arrival time is compared with an arrival time threshold; vehicle water source pairs with arrival times no greater than the arrival time threshold are retained. Water supply capacity is compared with a water supply capacity threshold; vehicle water source pairs with water supply capacity no less than the water supply capacity threshold are retained. Water supply benefit indicators are compared with a benefit threshold; vehicle water source pairs with water supply benefit indicators no less than the benefit threshold are retained. Water source identifiers meeting the threshold conditions are written into the corresponding fire truck identifier's candidate water supply set, and their route identifier, arrival time, and water supply duration are simultaneously retained. A specific example is as follows:

[0111] Retrieve the set of vehicles awaiting water replenishment associated with disaster event identifier E20260215A, and obtain a candidate set of water sources corresponding one-to-one with each fire truck identifier in the set of vehicles awaiting water replenishment. For fire truck identifier V01, its current location is the coordinate point of the north gate of the park, the remaining water volume is 6 cubic meters, the upper limit parameter of the water volume is 10 cubic meters, and the instantaneous water flow rate is 0.5 cubic meters per minute. Retrieve the water source location of water source identifier W01 in the candidate set as the coordinate point of the municipal fire hydrant group on the east side of the park, the water supply capacity is 3 cubic meters per minute, and it is marked as available or occupied. Pair V01 and W01 as vehicle water source pairs and write them into the vehicle water source pairing table.

[0112] Arrival time is calculated for vehicle water source pairs V01 and W01 in the vehicle water source pairing table. Using the current location of V01 and the location of the W01 water source as endpoints, a candidate path set is retrieved from the road network covered by road traffic data. Let the candidate path set include path P1 and path P2. Path P1 includes road segment A and road segment B, and path P2 includes road segment C and road segment D. Road segment A has a length of 1.0 km, a dynamic speed of 15 km / h, and a base travel time of 1.0 divided by 15 hours (approximately 4 minutes), plus a road event penalty of 0 minutes to obtain a road segment travel cost of 4 minutes. Road segment B has a length of 0.8 km, a dynamic speed of 16 km / h, and a base travel time of 1.0 divided by 15 hours (approximately 4 minutes). The basic travel time is approximately 3 minutes. Adding a 0-minute road event penalty, the road segment cost is 3 minutes, resulting in a path cost of 7 minutes for path P1. For path C, the length is 1.2 km, the dynamic speed is 12 km / h, the basic travel time is approximately 6 minutes, and adding a 1-minute road event penalty, the road segment cost is 7 minutes. For path D, the length is 0.5 km, the dynamic speed is 20 km / h, the basic travel time is approximately 2 minutes, and adding a 0-minute road event penalty, the road segment cost is 2 minutes, resulting in a path cost of 9 minutes for path P2. The path with the minimum path cost of 7 minutes from the candidate path set is determined as the arrival time of vehicle water source pairs V01 and W01, and the corresponding path is recorded as P1.

[0113] To calculate the water replenishment capacity based on the vehicle water source pairing table, for vehicle water source pairs V01 and W01, the required water replenishment volume is calculated as the upper limit water volume parameter of 10 cubic meters minus the remaining water volume of 6 cubic meters, resulting in 4 cubic meters. The water supply capacity of water source identifier W01 is obtained as 3 cubic meters per minute, and its occupancy marker is set to unoccupied. Therefore, the occupancy reduction factor is taken as 1, and the effective water supply capacity is 3 cubic meters per minute. This effective water supply capacity is used as the replenishment capacity. The replenishment duration is calculated by dividing the required water replenishment volume of 4 cubic meters by the effective water supply capacity. The water supply capacity is 3 cubic meters per minute, which yields approximately 1.33 minutes. To correlate the water supply capacity with the fire situation, the instantaneous outflow rate of V01 is obtained as 0.5 cubic meters per minute and converted into an instantaneous effective outflow rate of 0.5 cubic meters per minute. The instantaneous effective outflow rate and the effective water supply capacity are used to construct a water supply benefit index. Assuming that the water supply benefit index is the ratio of the effective water supply capacity to the instantaneous effective outflow rate, the water supply benefit index is 3 divided by 0.5, which yields 6. This index is used to reflect the increasing trend of sustainable water supply duration after water replenishment.

[0114] Based on threshold comparisons, a water replenishment candidate set is selected. The arrival time of 7 minutes is compared with the arrival time threshold of 12 minutes. If the arrival time is not greater than the arrival time threshold, vehicle water sources V01 and W01 are retained. The water replenishment capacity of 3 cubic meters per minute is compared with the water replenishment capacity threshold of 2.5 cubic meters per minute. If the water replenishment capacity is not less than the water replenishment capacity threshold, vehicle water sources V01 and W01 are retained. The water replenishment benefit index 6 is compared with the benefit threshold 4. If the water replenishment benefit index is not less than the benefit threshold, vehicle water sources V01 and W01 are retained. After the threshold conditions are met, the water source identifier W01 is written into the water replenishment candidate set corresponding to the fire truck identifier V01, and the path identifier P1, arrival time of 7 minutes, and water replenishment duration of approximately 1.33 minutes are simultaneously retained for direct use when generating a draft water replenishment instruction later.

[0115] Method for generating draft water replenishment instructions corresponding to the water replenishment candidate set:

[0116] Obtain the water replenishment candidate set corresponding to each fire truck identification, and calculate the comprehensive cost of each water source identification in the water replenishment candidate set. The comprehensive cost is obtained by summing the arrival time and water replenishment duration according to the preset weight. The preset weight is determined according to the treatment stage mark according to the preset weight table. The comprehensive cost is compared with the cost threshold. Water source identification with a comprehensive cost not greater than the cost threshold is determined as the preferred water replenishment candidate, and the preferred order is generated according to the comprehensive cost from small to large.

[0117] It should be noted that the preset weight table maps the disposal stage markers to a set of adjustable weights, which are used to allocate the influence of indicators such as arrival time and water replenishment duration in the comprehensive cost: the weight items are defined to include at least arrival time weight, water replenishment duration weight, and penalty item weight, so that the weights sum to one; the stage set is divided according to the disposal stage markers. For example, the initial control stage focuses on rapid arrival to avoid short-term supply interruption, so the arrival time weight is increased and the water replenishment duration weight is correspondingly decreased; the comprehensive suppression stage focuses on continuous supply capacity, so the water replenishment duration weight is increased and the arrival time weight is correspondingly decreased; when the demand pressure marker obtained from the demand sequence calculation is valid, the arrival time weight or penalty item weight is further increased on the basic weight of the corresponding stage according to the preset correction rules to reduce the delay risk caused by path uncertainty or water source occupation; finally, the above stage set and weight reorganization are solidified into the preset weight table, so that only the disposal stage marker needs to be obtained to select the corresponding weight reorganization to participate in the comprehensive cost calculation.

[0118] A draft water replenishment instruction is generated corresponding to the water replenishment candidate set. For each fire truck identifier, the preferred water source identifier and alternative water source identifier are written in the preferred order. The path identifier, arrival time, water replenishment capacity, amount of water to be replenished, and water replenishment duration corresponding to the preferred water source identifier are written into the draft water replenishment instruction, and the granularity of the instruction effective time and the treatment stage mark are written. The amount of water to be replenished is the difference between the upper limit parameter of the water intake and the remaining water volume, and the water replenishment duration is the ratio of the amount of water to be replenished to the effective water supply capacity.

[0119] The draft water replenishment instructions are associated with disaster event identifiers and fire truck identifiers, and a set of instruction fields is output for direct use in subsequent steps. The instruction field set includes at least the water source identifier, route identifier, estimated arrival time, estimated water replenishment volume, and estimated water replenishment duration, thereby providing input boundaries for subsequent water supply and replenishment rotation arrangements. A specific example is as follows:

[0120] Obtain the water replenishment candidate set corresponding to fire truck identifier V01, and calculate the comprehensive cost for water source identifier W01 within the water replenishment candidate set. Preset weights are assigned according to a preset weight table based on the treatment stage markers. Assuming the weight for arrival time in the initial control stage is 0.6 and the weight for water replenishment duration is 0.4, the comprehensive cost for water source identifier W01 is the arrival time (7 minutes) multiplied by 0.6 plus the water replenishment duration (1.33 minutes) multiplied by 0.4, resulting in approximately 4.2 plus 0.532, which equals approximately 4.732 minutes. Comparing the comprehensive cost of approximately 4.732 minutes with the cost threshold of 6 minutes, the comprehensive cost is not greater than the cost threshold. Therefore, water source identifier W01 is determined as the preferred water replenishment candidate for V01. Since the water replenishment candidate set only includes W01, W01 is selected as the preferred candidate based on the comprehensive cost from smallest to largest.

[0121] A draft water replenishment instruction is generated corresponding to the candidate water replenishment set. For fire truck identifier V01, the preferred water source identifier is written as W01 according to the preferred order. Since no other water source meets the cost threshold, the candidate water source identifier is left blank or written with an empty set marker. The path identifier P1, arrival time of 7 minutes, water replenishment capacity of 3 cubic meters per minute, water volume to be replenished, and water replenishment duration corresponding to the preferred water source identifier W01 are written into the draft water replenishment instruction. The difference between the upper limit parameter of water volume to be replenished of 10 cubic meters and the remaining water volume of 6 cubic meters is 4 cubic meters. The water replenishment duration is the ratio of the water volume to be replenished of 4 cubic meters to the effective water supply capacity of 3 cubic meters per minute, which is approximately 1.33 minutes. At the same time, the granularity of the instruction effective time is written into the draft water replenishment instruction as five minutes, and the disposal stage is marked as the initial control stage, so that the draft instruction corresponds to the time granularity and stage status.

[0122] The draft water replenishment instruction is associated with the disaster event identifier E20260215A and the fire truck identifier V01, and a set of instruction fields is output for direct use in subsequent steps. The instruction field set includes at least the water source identifier W01, the route identifier P1, the estimated arrival time of 7 minutes, the estimated water replenishment volume of 4 cubic meters, and the estimated water replenishment duration of approximately 1.33 minutes. This set of instruction fields is recorded together with the disaster event identifier E20260215A so that subsequent steps can directly obtain this set of fields to determine the water replenishment destination and estimated time arrangement of V01 when scheduling water supply and replenishment rotation.

[0123] S4, based on the demand sequence, fire truck status data and water replenishment instruction draft, solves for a water supply-replenishment rotation scheduling scheme that satisfies the continuous water supply constraint. The scheduling scheme includes at least the water supply sequence and water replenishment sequence of multiple fire trucks.

[0124] Based on demand sequences, fire truck status data, and draft water replenishment instructions, methods for obtaining water supply-replenishment rotation scheduling schemes that satisfy continuous water supply constraints include:

[0125] Obtain the demand sequence, fire truck status data, and water replenishment instruction draft bound to the disaster event identifier, and align them according to a unified time granularity to form a dispatch input table. The dispatch input table includes at least the demand quantity and disposal stage marker in each time granularity, as well as the current location, remaining water quantity, water quantity limit parameter, and real-time water flow rate of each fire truck identifier, and includes the water source identifier, path identifier, estimated arrival time, estimated water replenishment quantity, and estimated water replenishment duration associated with each fire truck identifier in the water replenishment instruction draft.

[0126] Based on the status data of fire trucks, the water supply capacity sequence is calculated. For each fire truck identification, the available water volume is calculated at each time granularity. The available water volume is the immediate effective water discharge rate multiplied by the time granularity duration. The immediate effective water discharge rate is obtained by averaging the immediate water discharge flow rate within that time granularity. The predicted remaining water volume at the end of the granularity is calculated. The predicted remaining water volume is the remaining water volume minus the available water volume. The predicted remaining water volume is compared with the lower limit threshold. The time granularity where the predicted remaining water volume is not greater than the lower limit threshold is marked as the water supply restricted time granularity, so that the fire truck identification will no longer be assigned to water supply in subsequent allocations.

[0127] Based on the draft water replenishment instruction, the sequence of water replenishment actions is calculated. For each fire truck identification, the granularity of the water replenishment start time is determined according to the expected arrival time, the number of granularities of the water replenishment duration is determined according to the expected water replenishment duration, and the amount of water backfilled after the water replenishment is determined according to the expected water replenishment amount. The granularity of the water replenishment start time is compared with the water replenishment start threshold, and the expected water replenishment duration is compared with the water replenishment duration threshold. Water replenishment actions that do not meet the threshold conditions are not included in the subsequent solution, thereby limiting the executable boundary of the draft water replenishment instruction.

[0128] Construct a continuous water supply constraint and calculate the supply-demand difference sequence. Within each time granularity, calculate the required water supply according to the demand in the scheduling input table; accumulate the available water supply for that granularity according to the set of fire trucks assigned to water supply to obtain the supplied water supply; calculate the supply-demand difference as the required water supply minus the supplied water supply, and compare the supply-demand difference with the gap threshold. The time granularity where the supply-demand difference is not greater than the gap threshold is determined to satisfy the continuous water supply constraint.

[0129] Construct rotation constraints and calculate rotation risk. Within each time granularity, count the number of fire trucks designated for water replenishment and the number of fire trucks designated for water supply restriction time granularities. Simultaneously, calculate the demand increase based on the demand sequence, where the demand increase is the difference in demand between adjacent time granularities. Combine the number of water replenishment vehicles, the number of restricted vehicles, and the demand increase according to a preset weight table corresponding to the disposal stage markers to synthesize the rotation risk. Compare the rotation risk with a risk threshold. Time granularities where the rotation risk is not greater than the risk threshold are allowed to schedule water replenishment vehicles to enter the water replenishment sequence.

[0130] Solve the water supply-replenishment rotation scheduling scheme, with the objective of minimizing the cumulative value of the supply-demand difference and the cumulative value of the rotation risk over all time periods. Within each time granularity, execute the allocation rules: First, select fire truck identifiers in the non-water-restricted time granularity from the water supply capacity sequence as the water supply candidate set. Then, add water supply vehicles to the water supply candidate set in descending order of available water capacity for that granularity, until the supply-demand difference meets the gap threshold or the water supply candidate set is exhausted. Next, select the water replenishment action sequence where the water replenishment start time granularity falls within that granularity. Fire truck identification marks are used as the candidate set for water replenishment, and the number of vehicles that can enter the water replenishment set is limited according to the comparison results of rotation risk and risk threshold. When the supply-demand difference does not meet the gap threshold, and the ratio of the number of water replenishment vehicles to the number of vehicles in service is not less than the water replenishment ratio threshold in the current time granularity, the water replenishment start time granularity of at least one water replenishment vehicle is postponed to the next time granularity, and the supply-demand difference and rotation risk are recalculated until the supply-demand difference is not greater than the gap threshold or the preset iteration termination condition is reached.

[0131] The system outputs a water supply-replenishment rotation scheduling plan, marking the status of each fire truck at various time granularities as either supplying or replenishing water. This forms a scheduling plan that includes the water supply and replenishment sequences of at least multiple fire trucks. The scheduling plan is then stored in association with disaster event identifiers for direct retrieval in subsequent steps. A specific example is shown below:

[0132] Obtain the demand sequence, fire truck status data, and draft water replenishment instructions bound to the disaster event identifier E20260215A, and align them according to a unified time granularity to form a dispatch input table. Taking 02:15 as an example, the dispatch input table includes a demand of 20 cubic meters every five minutes and the disposal stage marked as the initial control stage; it also includes the current location, remaining water volume, water volume limit parameter, and real-time water flow rate of fire truck identifiers V01 and V02 at this granularity; and it also includes the water source identifier W01, path identifier P1, estimated arrival time of 7 minutes, estimated water replenishment volume of 4 cubic meters, and estimated water replenishment duration of approximately 1.33 minutes associated with fire truck identifier V01 in the draft water replenishment instruction.

[0133] Based on the fire truck status data, the water supply capacity sequence is calculated. For fire truck identification V01, the available water volume is calculated within the time granularity of 02:15. The available water volume is calculated by multiplying the immediate effective water outflow rate by the time granularity duration. The immediate effective water outflow rate is taken as the average immediate water flow rate of 0.5 cubic meters per minute for this granularity, and the time granularity duration is 5 minutes. Therefore, the available water volume for V01 at this granularity is 2.5 cubic meters. The predicted remaining water volume at the end of this granularity is calculated as 6 cubic meters minus 2.5 cubic meters, resulting in 3.5 cubic meters. For fire truck identification V02, the available water volume is calculated using the same method: 0.4 cubic meters per minute multiplied by 5 minutes, resulting in 2.0 cubic meters. The predicted remaining water volume is 10 cubic meters minus 2.0 cubic meters, resulting in 8.0 cubic meters. Comparing the predicted remaining water volume with the lower limit threshold of 3 cubic meters, the predicted remaining water volume of 3.5 cubic meters for V01 is not greater than 3 cubic meters, which is not valid. The predicted remaining water volume of 8.0 cubic meters for V02 is not greater than 3 cubic meters, which is also not valid. Therefore, the time granularity of 02:15 is not marked as the water supply restricted time granularity for either V01 or V02, and both are allowed to be arranged as water supply in the subsequent allocation.

[0134] Based on the draft water replenishment instruction, the water replenishment action sequence is calculated. For fire truck V01, the water replenishment start time granularity is determined according to the estimated arrival time of 7 minutes. Converted to a 5-minute time granularity, the water replenishment start time granularity is 02:25. The water replenishment duration granularity is determined according to the estimated water replenishment duration of approximately 1.33 minutes. Rounded up, the water replenishment duration granularity is 1. The water backfill volume after water replenishment is determined according to the estimated water replenishment volume of 4 cubic meters. The water replenishment start time granularity of 02:25 minutes is compared with the water replenishment start threshold of 02:35 minutes. The water replenishment start time granularity is earlier than the water replenishment start threshold. The estimated water replenishment duration of 1.33 minutes is compared with the water replenishment duration threshold of 5 minutes. The estimated water replenishment duration is not greater than the water replenishment duration threshold. Therefore, the water replenishment action of V01 participates in the subsequent solution to limit the executable boundary of the draft water replenishment instruction.

[0135] A continuous water supply constraint is constructed, and the supply-demand difference sequence is calculated. Taking the time granularity of 02:15 as an example, the required water supply is calculated as 20 cubic meters every five minutes according to the demand in the dispatch input table. Let the set of fire truck identifiers assigned to water supply within this granularity be V01 and V02. Then the supplied water volume is 4.5 cubic meters, which is the sum of the available water volume of this granularity (2.5 cubic meters and 2.0 cubic meters). The supply-demand difference is calculated as the required water supply of 20 cubic meters minus the supplied water volume of 4.5 cubic meters, resulting in 15.5 cubic meters. This supply-demand difference is compared with the gap threshold of 16 cubic meters. Since the supply-demand difference is not greater than the gap threshold, the 02:15 time granularity is determined to satisfy the continuous water supply constraint. Taking the time granularity of 02:20 as an example, the supply-demand difference is calculated similarly based on the demand of 22 cubic meters every five minutes and compared with the gap threshold to form a supply-demand difference sequence.

[0136] A rotation constraint is constructed and the rotation risk is calculated. Taking 02:15 as an example, the number of fire trucks assigned to water replenishment is 0, and the number of fire trucks corresponding to the water supply restriction time granularity is also 0. Based on the demand sequence, the demand increase is calculated. The demand difference between 02:20 and 02:15 is 2 cubic meters per five minutes, which is taken as the demand increase for that period. The number of water replenishment vehicles, the number of restricted vehicles, and the demand increase are combined into a rotation risk by marking the disposal stage as the initial control stage according to a preset weight table. The weight of the number of water replenishment vehicles is set to 0.5, the weight of the number of restricted vehicles is set to 0.3, and the weight of the demand increase is set to 0.5. If the value is 0.2, then the rotation risk is 0 multiplied by 0.5, plus 0 multiplied by 0.3, plus 2 multiplied by 0.2, which equals 0.4. Comparing the rotation risk of 0.4 with the risk threshold of 1.0, the rotation risk is not greater than the risk threshold, thus the 02:15 time granularity allows water replenishment vehicles to enter the water replenishment sequence. Furthermore, at the 02:25 time granularity, if V01 is arranged for water replenishment, the number of water replenishment vehicles is 1. The number of restricted vehicles is determined according to the aforementioned lower limit threshold for water volume, and the increase in demand is obtained by the difference between adjacent granularities. Then, the rotation risk is calculated according to the same preset weight table and compared with the risk threshold to determine whether V01 is allowed to enter the water replenishment sequence at this time granularity.

[0137] S5: Execute the scheduling plan and obtain on-site feedback; when the availability index corresponding to the on-site feedback does not meet the preset threshold, update the water source candidate set or the set of vehicles to be replenished with water, and return to execute S3 and S4 to update the scheduling plan.

[0138] The methods for executing the scheduling plan and obtaining on-site feedback, and updating the water source candidate set or the set of vehicles to be replenished when the availability index corresponding to the on-site feedback does not meet the preset threshold, include:

[0139] Obtain the dispatch plan associated with the disaster event identifier and expand it into an action instruction sequence at a uniform time granularity. The action instruction sequence for each fire truck identifier includes at least the status markers at that time granularity, such as water supply or water replenishment, water supply location or water source identifier, route identifier, estimated arrival time and estimated water replenishment duration. The action instruction sequence is then associated with the fire truck identifier and output to execute the water supply sequence and water replenishment sequence.

[0140] Acquire on-site feedback associated with disaster event identifiers and align it with a unified time granularity; on-site feedback should include at least water source status feedback, road status feedback, and fire truck status feedback, where water source status feedback should include at least the availability marker, occupancy marker, and waiting time corresponding to the water source identifier; road status feedback should include at least the actual arrival time or road segment passage cost update value corresponding to the route identifier; and fire truck status feedback should include at least the current location, remaining water volume, and real-time water flow rate corresponding to the fire truck identifier; attach a collection timestamp and object identifier to each on-site feedback, so that it can be written back to the corresponding time granularity.

[0141] The on-site feedback is converted into availability indicators and threshold comparisons are performed. Availability indicators include at least water source availability indicators, road accessibility indicators, and vehicle water supply capacity indicators. The water source availability indicator is calculated from available markers and occupied markers. When an available marker does not meet the availability threshold or an occupied marker meets the occupancy threshold, the water source availability indicator is determined to be unsatisfactory. The road accessibility indicator is the difference between the actual arrival time and the estimated arrival time, or the updated arrival time. The difference is compared with the delay threshold or the arrival time is compared with the arrival time threshold. If the difference is not met, the road accessibility indicator is determined to be unsatisfactory. The vehicle water supply capacity indicator is the ratio of the remaining water volume to the immediate effective water output rate to obtain the sustainable water supply duration. The sustainable water supply duration is compared with the duration threshold. If the threshold is not met, the vehicle water supply capacity indicator is determined to be unsatisfactory. When any availability indicator is determined to be unsatisfactory, the candidate set is updated.

[0142] The candidate set of water sources or the set of vehicles awaiting water replenishment are updated according to the type of indicators that are not met. When the water source availability indicator or the road accessibility indicator is not met, the corresponding water source identifier is removed from the candidate set of water sources, and available water sources are searched again within a preset search radius based on the current location of the corresponding fire truck identifier to replenish the candidate set of water sources. When the vehicle water supply capacity indicator is not met, the corresponding fire truck identifier is added to the set of vehicles awaiting water replenishment, and the water supply restriction time granularity corresponding to the fire truck identifier is updated to be true in subsequent allocations. A specific example is as follows:

[0143] Obtain the dispatch plan associated with disaster event identifier E20260215A and expand it into an action instruction sequence at a unified time granularity; for fire truck identifier V01, the action instruction sequence at the time granularity of 02:25 should at least include a status marker of water replenishment, a water source identifier of W01, a path identifier of P1, an estimated arrival time of 7 minutes, and an estimated water replenishment duration of approximately 1.33 minutes; for fire truck identifier V02, the action instruction sequence at the same time granularity should be marked as water supply and the water supply location should be given; associate the action instruction sequence with fire truck identifiers V01 and V02 and output it so that it is executed according to the water supply sequence and the water replenishment sequence.

[0144] Acquire on-site feedback associated with disaster event identifier E20260215A and align it with a unified time granularity. Within the time granularity of 02:25, acquire water source status feedback, with water source identifier W01 marked as available and occupied, with a waiting time of 8 minutes; acquire road status feedback, with the actual arrival time corresponding to path identifier P1 being 15 minutes, and simultaneously report the updated road segment traffic cost value to increase the arrival time; acquire fire truck status feedback, with fire truck identifier V01 corresponding to the current location near the east intersection of the park, remaining water volume of 4.5 cubic meters, and immediate water flow rate of 0.5 cubic meters per minute; attach the collection timestamp of 02:25 and object identifiers W01, P1, and V01 to the above on-site feedback so that it can be written back to the corresponding time granularity.

[0145] The on-site feedback was converted into availability indicators and threshold comparisons were performed. For the water source availability indicator, it was calculated from available and occupied markers. The availability threshold required available markers to be available, and the occupancy threshold required occupied markers to be unoccupied. Since W01's available markers met the availability threshold but its occupied markers met the occupancy threshold, the water source availability indicator was deemed unsatisfactory. For the road accessibility indicator, the difference between the actual arrival time and the estimated arrival time was taken. The difference was 15 minutes minus 7 minutes, resulting in 8 minutes. This difference was compared to the delay threshold of 5 minutes. Since the difference was not less than the delay threshold, the road accessibility indicator was deemed satisfactory. The requirement is not met. For the vehicle water supply capacity indicator, the ratio of remaining water volume to the instantaneous effective water output rate is used to obtain the sustainable water supply duration. The instantaneous effective water output rate is taken as the average instantaneous water flow rate within this time granularity, which is 0.5 cubic meters per minute. Therefore, the sustainable water supply duration is 4.5 cubic meters divided by 0.5 cubic meters per minute, which gives 9 minutes. The sustainable water supply duration is compared with the duration threshold of 20 minutes. The sustainable water supply duration is not greater than the duration threshold. Therefore, the vehicle water supply capacity indicator is determined to be unmet. Since there are unmet items among the water source availability indicator, road accessibility indicator, and vehicle water supply capacity indicator, the candidate set is updated.

[0146] The candidate set of water sources or the set of vehicles awaiting water replenishment are updated according to the type of indicators that are not met. If the water source availability indicator or the road accessibility indicator is not met, the corresponding water source identifier W01 is removed from the water source candidate set of fire truck identifier V01. Based on the current location of fire truck identifier V01, available water sources are searched again within a preset search radius of 3 kilometers to supplement the water source candidate set. If water source identifier W03 is found and its availability marker meets the availability threshold and its occupancy marker does not meet the occupancy threshold, then W03 is added to the water source candidate set of V01. At the same time, if the vehicle water supply capacity indicator is not met, fire truck identifier V01 is added to the set of vehicles awaiting water replenishment. In subsequent allocation, the water supply restriction time granularity corresponding to fire truck identifier V01 is updated to meet the requirement, so that V01 will not be arranged as a water supplier at this time granularity in subsequent solutions.

[0147] The methods for returning to execute S3 and S4 to update the scheduling scheme include:

[0148] Based on the updated set of vehicles awaiting water replenishment and the set of candidate water sources, execute S3. For each vehicle awaiting water replenishment, calculate the arrival time, water replenishment capacity, and water replenishment duration for each candidate water source. Compare the arrival time with the arrival time threshold, the water replenishment capacity with the water replenishment capacity threshold, and the water replenishment benefit index with the benefit threshold to filter and obtain a water replenishment candidate set. Generate or update a draft water replenishment instruction based on the water replenishment candidate set, so that the draft water replenishment instruction includes water source identifier, route identifier, estimated arrival time, estimated water replenishment volume, and estimated water replenishment duration.

[0149] Based on the updated draft water replenishment instruction, execution S4 is returned. The scheduling input table is re-formed at a unified time granularity. The water supply capacity sequence, water replenishment action sequence, supply-demand difference sequence, and rotation risk quantity are recalculated. New water supply and replenishment time sequences are output, constrained by the satisfaction of the supply-demand difference and gap threshold comparison results, and the satisfaction of the rotation risk quantity and risk threshold comparison results. This forms the updated scheduling scheme and is associated with the disaster event identifier. A specific example is as follows:

[0150] Based on the updated set of vehicles awaiting water replenishment and the candidate set of water sources, execution S3 is returned. For fire truck identifier V01 in the set of vehicles awaiting water replenishment, its current location is obtained as near the intersection on the east side of the park, its remaining water volume is 4.5 cubic meters, its maximum water volume parameter is 10 cubic meters, and its instantaneous water flow rate is 0.5 cubic meters per minute. Its candidate set of water sources is obtained as water source identifier W03, and the water source location of W03 is obtained as the coordinate point of the fire water tank on the north side of the park, its water supply capacity is 2 cubic meters per minute, its availability is marked as available, and its occupation is marked as unoccupied. This forms the vehicle water source pair V01 and W03 in the vehicle water source pairing table. Calculations are then performed separately. Arrival time, water replenishment capacity, and water replenishment duration: Taking the current location and the water source location as endpoints, a candidate path set is retrieved in the road network covered by road traffic data. Assuming the path cost for path identifier P3 is 10 minutes, the arrival time is 10 minutes, and path identifier P3 is recorded. The amount of water to be replenished is calculated as the upper limit parameter of 10 cubic meters minus the remaining water volume of 4.5 cubic meters, resulting in 5.5 cubic meters. The occupancy marker indicates that the effective water supply capacity when unoccupied is equal to the supply capacity of 2 cubic meters per minute; this effective supply capacity is used as the water replenishment capacity. The water replenishment duration is the amount of water to be replenished (5.5 cubic meters) divided by 2 cubic meters per minute, resulting in 2.75 minutes. A water replenishment benefit index is constructed, assuming it is the ratio of effective supply capacity to the immediate effective outflow rate. The immediate effective outflow rate is taken as the average immediate outflow rate within that time granularity, 0.5 cubic meters per minute. Therefore, the water replenishment benefit index is 2 divided by 0.5, resulting in 4. The arrival time of 10 minutes is compared with the arrival time threshold of 12 minutes, and the arrival time is not greater than the arrival time threshold. The water replenishment capacity of 2 cubic meters per minute is compared with the water replenishment capacity threshold of 2 cubic meters per minute, and the water replenishment capacity is not less than the water replenishment capacity threshold. The water replenishment benefit index 4 is compared with the benefit threshold 4, and the water replenishment benefit index is not less than the benefit threshold. Based on this, the water replenishment candidate set for V01 is obtained, which includes the water source identifier W03. The water replenishment instruction draft is generated or updated based on the water replenishment candidate set, so that the water replenishment instruction draft includes the water source identifier W03, the route identifier P3, the estimated arrival time of 10 minutes, the estimated water replenishment volume of 5.5 cubic meters, and the estimated water replenishment duration of 2.75 minutes, and is associated with the disaster event identifier E20260215A and the fire truck identifier V01.

[0151] Based on the updated water replenishment instruction draft, execution S4 is returned, and the scheduling input table is re-formed according to a unified time granularity. This ensures that the input table includes the demand quantity and handling stage markers corresponding to the demand sequence within the time granularity of 02:25 and subsequent time granularities. This includes the current location, remaining water volume, water volume limit parameters, and instantaneous water flow rate of fire truck identifiers V01 and V02, as well as the water source identifier W03, path identifier P3, estimated arrival time of 10 minutes, estimated water replenishment volume of 5.5 cubic meters, and estimated water replenishment duration of 2.75 minutes corresponding to V01 in the updated water replenishment instruction draft. When recalculating the water supply capacity sequence, the constraints that V01 meets at the water supply restriction time granularity are incorporated, preventing V01 from being allocated water supply at that time granularity. When recalculating the water replenishment action sequence, the water replenishment start time granularity is determined based on the estimated arrival time of 10 minutes, using a five-minute time granularity. The water replenishment start time granularity is converted to 02:35, and the water replenishment duration granularity is determined to be 1 based on the expected water replenishment duration of 2.75 minutes. Within each time granularity, the supply and demand difference sequence and the rotation risk quantity are recalculated, and a new water supply sequence and water replenishment sequence are output with the constraints that the supply and demand difference and the gap threshold comparison result are satisfied, and the rotation risk quantity and the risk threshold comparison result are satisfied. For example, the updated scheduling scheme arranges fire truck identifier V02 to maintain water supply from 02:25 to 02:35, and fire truck identifier V01 to go to water source identifier W03 according to path identifier P3 from 02:25 to 02:35 and enter the water replenishment sequence at 02:35, so that the supply and demand difference at the corresponding time granularity is not greater than the gap threshold and the rotation risk quantity is not greater than the risk threshold. The updated scheduling scheme is then associated with disaster event identifier E20260215A.

[0152] Example 2

[0153] See Figure 2 As shown, this embodiment provides an intelligent scheduling and optimization system based on uninterrupted fire water supply, including:

[0154] Data acquisition module: used to acquire fire vehicle status data and water supply demand data associated with disaster event identifiers; and to generate demand sequences bound to disaster event identifiers based on water supply demand data;

[0155] Candidate generation module: Based on fire truck status data and demand sequence, determine the set of vehicles to be replenished with water, and retrieve available water sources for the vehicles to be replenished with water to obtain a set of water source candidates;

[0156] Instruction generation module: It is used to calculate the arrival time and water replenishment capacity of the vehicles to be replenished to each water source candidate, filter the water source candidate set from the water source candidate set according to the threshold comparison, and generate a water replenishment instruction draft corresponding to the water replenishment candidate set.

[0157] The scheduling solution module: Based on the demand sequence, fire truck status data and water replenishment instruction draft, it solves for a water supply-replenishment rotation scheduling scheme that satisfies the continuous water supply constraint;

[0158] Feedback Update Module: Used to execute the scheduling plan and obtain on-site feedback; when the availability index corresponding to the on-site feedback does not meet the preset threshold, update the water source candidate set or the set of vehicles to be replenished, and return to execute S3 and S4 to update the scheduling plan.

[0159] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A smart scheduling and optimization method based on uninterrupted fire-fighting water supply, characterized in that, include: S1, obtain fire truck status data and water supply demand data associated with disaster event identifiers; Demand sequences are generated based on water supply demand data and bound to disaster event identifiers; S2, based on the status data and demand sequence of fire trucks, determine the set of vehicles to be replenished with water, and retrieve available water sources for the vehicles to be replenished with water to obtain a candidate set of water sources; S3, calculate the arrival time and water replenishment capacity of the vehicles to be replenished to each water source candidate, filter the water source candidate set from the water source candidate set according to the threshold comparison, and generate a water replenishment instruction draft corresponding to the water replenishment candidate set. S4, based on the demand sequence, fire truck status data and water replenishment instruction draft, solves for a water supply-replenishment rotation scheduling scheme that satisfies the continuous water supply constraint. Methods for obtaining a water supply-replenishment rotation scheduling scheme that satisfies the continuous water supply constraint include: Obtain the demand sequence, fire truck status data, and draft water replenishment instructions bound to disaster event identifiers, and align them with a unified time granularity; The water supply capacity sequence is calculated based on fire truck status data, and the water replenishment action sequence is calculated based on the draft water replenishment order. Within each time granularity, the supply-demand difference is calculated based on the demand sequence and the water supply capacity sequence and compared with the gap threshold to determine the water supply vehicle set; Within each time granularity, the set of water replenishment vehicles is determined based on the water replenishment action sequence; Based on the water supply vehicle set and water replenishment vehicle set, the water supply-replenishment rotation scheduling scheme is output. The water supply-replenishment rotation scheduling scheme includes at least the water supply sequence and water replenishment sequence of multiple fire trucks, and is associated with the disaster event identifier. When calculating the water replenishment action sequence based on the draft water replenishment directive, the water replenishment start time granularity is determined according to the expected arrival time, and the number of water replenishment duration time granularities is determined according to the expected water replenishment duration. Within each time granularity, the number of fire trucks designated for water replenishment is counted, and the number of fire trucks designated for water supply restriction time granularity is also counted. At the same time, the demand increase is calculated based on the demand sequence. The number of water replenishment vehicles, the number of restricted vehicles, and the demand increase are combined into a rotation risk quantity according to a preset weight table corresponding to the disposal stage markers. The rotation risk quantity is compared with the risk threshold. When the supply-demand difference does not meet the gap threshold, and the ratio of the number of water replenishment vehicles to the number of vehicles in service within the current time granularity is not less than the water replenishment ratio threshold, the water replenishment start time granularity of at least one water replenishment vehicle is postponed to the next time granularity, and the supply-demand difference and rotation risk quantity are recalculated. S5: Execute the scheduling plan and obtain on-site feedback; when the availability index corresponding to the on-site feedback does not meet the preset threshold, update the water source candidate set or the set of vehicles to be replenished, and return to execute S3 and S4 to update the scheduling plan; the methods for updating the water source candidate set or the set of vehicles to be replenished include: Obtain the dispatch plan associated with the disaster event identifier and expand it into an action instruction sequence at a uniform time granularity, so that each fire truck identifier can execute the water supply sequence and water replenishment sequence according to the action instruction sequence; Obtain on-site feedback associated with disaster event identifiers and align it with a unified time granularity; Based on on-site feedback, usability indicators are calculated and compared with preset thresholds. When the comparison does not meet the requirements, the candidate set is updated. When the availability index is a water source availability index or a road accessibility index and the corresponding comparison is not met, the corresponding water source identifier is removed from the water source candidate set and the water source candidate set is updated. When the availability index is a vehicle water supply capacity index and the corresponding comparison is not met, the corresponding fire truck identifier is written into the water supply vehicle set to update the water supply vehicle set.

2. The intelligent scheduling and optimization method based on uninterrupted fire-fighting water supply according to claim 1, characterized in that, Methods for generating demand sequences that are bound to disaster event identifiers include: The system acquires disaster event identifiers and related alarm records and on-site feedback information. It extracts disaster location, disaster type, disaster scale, response stage markers, and start time to form water supply demand data. Based on the disaster event identifiers, it acquires fire truck status data, which includes at least the fire truck identifier, current location, and remaining water volume. It also adds a collection timestamp to the fire truck status data to form a time-sorted sequence of fire truck status data. It performs consistency processing on the fire truck status data sequence to obtain a standard sequence of fire truck status. Based on the water supply demand data, it outputs a demand sequence through a preset disaster model.

3. The intelligent scheduling and optimization method based on uninterrupted fire-fighting water supply according to claim 2, characterized in that, Methods for pre-setting disaster models include: Obtain a historical disaster sample set. Each historical disaster sample in the historical disaster sample set shall include at least the historical disaster type, historical disaster scale, historical response phase sequence, historical water supply record and historical water use record. By aligning historical water supply records and historical water use records with a uniform time granularity, a historical demand sequence is obtained as a training label. Using historical disaster types, scales, and response phases as input features, a supervised learning regression training method is employed to iteratively update the model parameters. The demand error function is used as the objective function, and the demand error is compared with an error threshold. If the demand error exceeds the error threshold, iterative updates continue until a stopping condition is met, at which point the model is solidified to obtain the preset disaster model.

4. The intelligent scheduling and optimization method based on uninterrupted fire-fighting water supply according to claim 1, characterized in that, Methods for determining the group of vehicles to be replenished with water include: Obtain fire truck status data and demand sequences associated with disaster event identifiers, and align them with a uniform time granularity; The sustainable water supply duration of each fire truck is calculated based on the remaining water volume and the instantaneous water flow rate; the demand pressure label is obtained based on the comparison between the demand sequence and the threshold. The water replenishment trigger flag is obtained by comparing the continuous water supply duration of a single vehicle with the duration threshold. The water replenishment trigger flag is then merged with the demand pressure flag. The fire truck identifiers that meet the preset conditions are written into the set of vehicles to be replenished.

5. The intelligent scheduling and optimization method based on uninterrupted fire-fighting water supply according to claim 4, characterized in that, Methods for retrieving available water sources and obtaining a candidate set of water sources for vehicles awaiting water replenishment include: Obtain a list of water sources associated with the disaster event identifier. The list should include at least the water source identifier, water source location, water supply capacity, and availability marker. For each fire truck identifier in the water replenishment vehicle set, an initial water source set is obtained based on its current location within a preset search radius; The availability flags of each water source in the initial water source set are compared with the availability threshold. Water sources that do not meet the availability threshold are removed from the initial water source set to obtain the available water source set. The set of candidate water sources corresponding to the fire truck identification is determined based on the set of available water sources.

6. The intelligent scheduling and optimization method based on uninterrupted fire-fighting water supply according to claim 1, characterized in that, Methods for obtaining a water replenishment candidate set include: Obtain a candidate set of water sources that corresponds one-to-one with each fire truck identifier in the set of vehicles to be replenished with water; for each fire truck identifier, calculate the arrival time based on the current location and the water source location of each water source identifier in the candidate set; calculate the amount of water to be replenished based on the upper limit parameter of water volume and the remaining water volume, and determine the water replenishment capacity and calculate the water replenishment duration based on the water supply capacity of the water source identifier; compare the arrival time with the arrival time threshold and the water replenishment capacity with the water replenishment capacity threshold, so as to write the water source identifier that meets the threshold conditions into the water replenishment candidate set corresponding to the fire truck identifier.

7. The intelligent scheduling and optimization method based on uninterrupted fire-fighting water supply according to claim 6, characterized in that, Methods for generating draft water replenishment instructions corresponding to the water replenishment candidate set include: Obtain the water replenishment candidate set corresponding to each fire truck identification, and calculate the comprehensive cost based on the arrival time and water replenishment duration of each water source identification in the water replenishment candidate set; The overall cost is compared with the cost threshold to determine the preferred order of each water source identifier in the water replenishment candidate set; a draft water replenishment instruction is generated based on the preferred order, and the draft water replenishment instruction includes at least the preferred water source identifier, the route identifier, the estimated arrival time, the estimated water replenishment amount, and the estimated water replenishment duration.

8. An intelligent scheduling and optimization system based on uninterrupted fire water supply, implementing the intelligent scheduling and optimization method based on uninterrupted fire water supply as described in any one of claims 1-7, characterized in that the system... include: Data acquisition module: used to acquire fire truck status data and water supply demand data associated with disaster event identifiers; Demand sequences are generated based on water supply demand data and bound to disaster event identifiers; Candidate generation module: Based on fire truck status data and demand sequence, determine the set of vehicles to be replenished with water, and retrieve available water sources for the vehicles to be replenished with water to obtain a candidate set of water sources; Instruction generation module: It is used to calculate the arrival time and water replenishment capacity of the vehicles to be replenished to each water source candidate, filter the water source candidate set from the water source candidate set according to the threshold comparison, and generate a water replenishment instruction draft corresponding to the water replenishment candidate set. The scheduling solution module: Based on the demand sequence, fire truck status data and water replenishment instruction draft, it solves for a water supply-replenishment rotation scheduling scheme that satisfies the continuous water supply constraint; Feedback Update Module: Used to execute the scheduling plan and obtain on-site feedback; when the availability index corresponding to the on-site feedback does not meet the preset threshold, update the water source candidate set or the set of vehicles to be replenished, and return to execute S3 and S4 to update the scheduling plan.

Images