Urban fire scene water supply service scheduling method and system in dynamic environment

Abstract

The invention discloses an urban fire scene water supply service scheduling method and system in a dynamic environment. The method comprises the following steps: S1, acquiring fire intensity and fire truck parameters; s2, constructing a dynamic nonlinear service combination model according to the fire intensity and the fire fighting truck parameters; s3, operating a two-stage hybrid optimization algorithm to generate a scheduling instruction according to the dynamic nonlinear service combination model; s4, executing the dispatching instruction, and dynamically collecting fire intensity and fire truck parameters; s5, updating the dynamic nonlinear service combination model according to the fire intensity and the fire fighting truck parameters; and S6, the steps S3 to S5 are repeated until the fire intensity is smaller than or equal to zero. The method can scientifically configure fire-fighting resources, improves the fire extinguishing efficiency, and is of great significance in improving the FaaS fire rescue capability and guaranteeing the life and property safety of people.

Description

Technical Field

[0001] The present invention belongs to the technical field of urban fire fighting, and particularly relates to a method and system for dispatching urban fire field water supply services in a dynamic environment. Background Art

[0002] In the process of fire rescue, a dynamic and efficient Internet of Things service combination optimization strategy is one of the key factors to ensure the success of fire extinguishing operations. Currently, in fire fighting operations, the public security fire forces usually select direct service binding, relay service chains, or dynamic service reorganization methods based on the service performance of fire truck Internet of Things nodes and the network topology relationship between service nodes and the fire field, and at the same time, optimize the QoS parameters of service links as much as possible. However, these scheduling decisions mostly rely on past experience for manual inference and lack the support of scientific service combination optimization algorithms. Traditional experience-based scheduling methods are often difficult to cope with in complex and dynamic fire field environments and fail to achieve the optimal allocation of service resources and the maximization of QoS indicators.

[0003] For example, Chinese Patent No. CN109731264A discloses a fire sprinkler machine, a water supply machine, a sprinkler device, a system, and a control method, which relate to the technical field of fire sprinkler water supply. The fire sprinkler machine includes a sprinkler controller and a sprinkler device communicatively connected to the sprinkler controller. The sprinkler controller can establish communication with a water supply controller. The sprinkler device can be connected to the water supply device of the fire water supply machine, and the sprinkler device can receive the water supplied by the water supply device and be used for spraying water outward. The sprinkler controller is used to collect the spraying state information of the sprinkler device and control the sprinkler device to spray water according to the sprinkler control command and / or according to the water supply state information of the water supply device collected by the water supply controller. The sprinkler controller is also used to send the sprinkler control command and / or the spraying state information to the water supply controller. The fire sprinkler machine, the water supply machine, the sprinkler device, the system, and the control method can achieve the automatic linkage of the water supply part and the sprinkler part and reduce the probability of water supply chaos. However, this patent lacks a detailed distribution strategy for water supply resources, only relies on sensors to judge water supply and spraying, lacks spraying rate adjustment and more efficient water replenishment judgment, is difficult to reduce fire fighting resources, and fails to improve the fire extinguishing efficiency. At the same time, this patent uses a relay service chain for water supply, which requires multiple fire truck water tanks to be connected in series for water supply, requires a large number of devices to be connected, will cause a waste of a large amount of fire fighting resources, and is difficult to achieve effective water supply in the case of insufficient equipment or too far distance between water supply resource points, resulting in the inability to achieve efficient fire extinguishing. Summary of the Invention

[0004] To solve the above technical problems, the present invention provides a method and system for dispatching urban fire field water supply services in a dynamic environment.

[0005] The present invention models the functional and non-functional characteristics of fire trucks using a service paradigm, defines it as Fire Truck-as-a-Service, abbreviated as FaaS in English. This paradigm naturally fits the requirements of the urban fire scene - although existing service-oriented methods have been applied in the transportation field, such as modeling the spatio-temporal and QoS attributes of transportation services like buses and ferries, they have not effectively solved the key problem of dynamic scheduling of fire resources under road network constraints. However, FaaS provides an innovative solution for intelligent fire extinguishing systems by abstracting fire trucks into combinable service units. An optimal service combination strategy of fire-fighting equipment as a service is formulated based on the spatio-temporal dynamic changes of fire risks, and in the case of limited service nodes and resources, quickly responds to and extinguishes the fire.

[0006] The present invention is achieved through the following technical solutions.

[0007] A method for scheduling urban fire scene water supply services in a dynamic environment provided by the present invention includes the following steps:

[0008] S1. Obtain the fire intensity and fire truck parameters, where the fire truck parameters include the water spraying rate and the water storage capacity;

[0009] S2. Construct a dynamic non-linear service combination model based on the fire intensity and fire truck parameters. The dynamic non-linear service combination model includes service nodes, a fire spread sub-model, a fire extinguishing rate sub-model, and a service combination engine;

[0010] S3. Run a two-stage hybrid optimization algorithm to generate a scheduling instruction according to the dynamic non-linear service combination model;

[0011] S4. Execute the scheduling instruction and dynamically collect the fire intensity and fire truck parameters;

[0012] S5. Update the dynamic non-linear service combination model according to the fire intensity and fire truck parameters;

[0013] S6. Repeat steps S3 to S5 until the fire intensity is less than or equal to zero.

[0014] Preferably, the expression of the fire spread sub-model is as follows:

[0015] x = αQ 2 + βQ

[0016] where x is the fire spread rate, Q is the fire intensity, and α and β are fire spread coefficients;

[0017] The expression of the fire extinguishing rate sub-model is as follows:

[0018]

[0019] Among them, q is the fire extinguishing rate, p is the water spraying rate, Q is the fire intensity, and μ, θ, and η are the fire extinguishing coefficients.

[0020] Preferably, the step S2, constructing a dynamic nonlinear service combination model according to fire intensity and fire truck parameters, comprises the following steps:

[0021] S21. Set the fire truck as a service node and construct service QoS attributes based on the fire truck parameters;

[0022] S22. Construct a fire spread sub-model based on fire intensity;

[0023] S23, constructing a fire extinguishing rate sub-model based on fire truck parameters;

[0024] S24. Build a service composition engine.

[0025] Preferably, the step S3 of running the two-stage hybrid optimization algorithm to generate a dispatch instruction according to the dynamic nonlinear service combination model and the fire truck parameters includes the following steps:

[0026] S31. Determine the minimum number of service nodes according to the dynamic nonlinear service combination model;

[0027] S32. Optimize the water replenishment triggering conditions and water spraying rates of the service nodes based on the minimum number of service nodes and the dynamic nonlinear service combination model to obtain the optimal water spraying rate, optimal water replenishment strategy, and shortest fire extinguishing time;

[0028] S33. Generate a scheduling instruction based on the optimal water spraying rate and the optimal water replenishment strategy.

[0029] Preferably, the step S31 of determining the minimum number of service nodes according to the dynamic nonlinear service composition model includes the following steps:

[0030] S311, initializing the number of simulated fire trucks;

[0031] S312, initializing the simulated fire intensity and the simulated fire truck water volume according to the dynamic nonlinear service combination model;

[0032] S313. Calculate the fire spread rate based on the simulated fire intensity;

[0033] S314, calculating the total fire extinguishing rate based on the fire extinguishing rate sub-model, the simulated fire truck water volume, and the simulated number of fire trucks;

[0034] S315, updating the simulated fire intensity according to the fire spread rate and the total fire extinguishing rate;

[0035] S316. Determine whether the fire extinguishing is successful according to the simulated fire intensity. If so, record the current number of simulated fire trucks, obtain the minimum number of service nodes, otherwise increase the number of simulated fire trucks and return to step S312.

[0036] Preferably, the step S32 of optimizing the water replenishment trigger condition and the water spraying rate of the service nodes according to the minimum number of service nodes and the dynamic non - linear service combination model, and obtaining the optimal water spraying rate, the optimal water replenishment strategy and the shortest fire extinguishing time includes the following steps:

[0037] S321. Initialize the population according to the minimum number of service nodes. The population includes several scheduling strategy individuals, and each scheduling strategy individual includes a water replenishment trigger condition and a water spraying rate.

[0038] S322. Set the fitness function.

[0039] S323. Adjust the water spraying rate, the water replenishment trigger condition and the water replenishment priority of the scheduling strategy individuals in the population according to the dynamic non - linear service combination model.

[0040] S324. Perform mutation, crossover and selection operations on the population.

[0041] S325. Set the maximum number of iterations, and repeat step S324 until the number of repetitions reaches the maximum number of iterations or the fitness function converges.

[0042] S326. Obtain the optimal water spraying rate, the optimal water replenishment trigger condition and the shortest fire extinguishing time according to the population.

[0043] S327. Generate the optimal water replenishment strategy according to the optimal water replenishment trigger condition and the water replenishment priority.

[0044] Preferably, the adjustment expression of the water spraying rate of the scheduling strategy individual is as follows:

[0045]

[0046] where p i is the water spraying rate of the i - th service node, p is the water spraying rate of the fire truck, and Q t is the fire intensity.

[0047] A water supply service scheduling system for urban fire scenes in a dynamic environment, which is used to implement the above - mentioned urban fire scene water supply service scheduling method, includes: a central decision - making module, an edge execution module and a fire scene feedback module;

[0048] The central decision - making module is deployed in the cloud, and is used to construct and update the dynamic non - linear service combination model according to the fire intensity, and run the two - stage hybrid optimization algorithm to generate scheduling instructions according to the dynamic non - linear service combination model;

[0049] The edge execution module is deployed in the fire truck cluster and is used to execute scheduling instructions and collect fire truck parameters in real time. The fire truck parameters include the water spraying rate and the water storage capacity;

[0050] The fire scene feedback module is deployed in the fire scene monitoring equipment and is used to sense the fire intensity in real time.

[0051] Preferably, the dynamic non-linear service composition model includes service nodes, a fire spread sub-model, a fire extinguishing rate sub-model, and a service composition engine;

[0052] The service nodes are set as fire trucks, and the fire truck parameters form service QoS attributes;

[0053] The fire spread sub-model is used to summarize the fire spread rate;

[0054] The fire extinguishing rate sub-model is used to summarize the fire extinguishing efficiency;

[0055] The service composition engine is used to evaluate the status of service nodes and trigger the water supply service composition according to the fire spread sub-model and the fire extinguishing rate sub-model.

[0056] Preferably, the edge execution module includes a sensor unit and an edge computing unit;

[0057] The sensor unit is used to collect fire truck parameters;

[0058] The edge computing unit is used to execute scheduling instructions.

[0059] The beneficial effects of the present invention are as follows:

[0060] 1. A dynamic non-linear service composition model is proposed, which focuses on the impact of the urban complex environment on the service chain topology structure, filling the gap that the domestic and foreign related research focuses on forest fires and lacks attention to urban fire service orchestration;

[0061] 2. Combining the advantages of the greedy strategy and the differential evolution algorithm, an efficient two-stage hybrid optimization algorithm is proposed. Among them, the greedy strategy has the ability of rapid decision-making in the local optimization of fire truck service nodes, while the differential evolution algorithm avoids falling into local optimal solutions in the global optimization of multiple service chains. The combination of the two improves the flexibility and accuracy of FaaS water supply scheduling, significantly shortens the fire extinguishing time and optimizes the resource allocation;

[0062] 3. Through the dynamic non-linear service composition model and the two-stage hybrid optimization algorithm, the dynamic optimization of the service composition scheme is successfully realized, effectively improving the service resource utilization efficiency and the QoS comprehensive index;

[0063] 4. By deploying a central decision-making module, an edge execution module, and a fire field feedback module to generate and execute scheduling instructions, fire-fighting resources can be scientifically configured, the fire-fighting efficiency can be improved, which is of great significance for enhancing the FaaS fire rescue ability and ensuring the safety of people's lives and property. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 is a flowchart of the method provided by an embodiment of the present invention;

[0065] Figure 2 is a flowchart of the differential evolution algorithm provided by an embodiment of the present invention;

[0066] Figure 3 is a schematic structural diagram of the system provided by an embodiment of the present invention;

[0067] Figure 4 is a schematic diagram of the fire intensity, the number of vehicles, and the water storage capacity of the vehicle in the scheduling experiment results provided by an embodiment of the present invention;

[0068] Figure 5 is a schematic diagram of the fire extinguishing rate and the total water consumption in the scheduling experiment results provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0069] The technical solutions of the present invention will be further described below, but the scope of protection claimed is not limited thereto.

[0070] As Figure 1 shown, a method for scheduling the water supply service in an urban fire field under a dynamic environment includes the following steps:

[0071] S1. Obtain the fire intensity and the fire truck parameters, where the fire truck parameters include the water spraying rate and the water storage capacity;

[0072] The fire intensity is obtained through the fire field feedback module deployed on the fire field monitoring device, and the fire truck parameters are obtained through the sensor unit deployed on the fire truck.

[0073] S2. Construct a dynamic non-linear service combination model according to the fire intensity and the fire truck parameters, where the dynamic non-linear service combination model includes service nodes, a fire spread sub-model, a fire extinguishing rate sub-model, and a service combination engine;

[0074] The dynamic non - linear service composition model abstracts the urban fire - fighting water supply scheduling problem into a dynamic multi - constraint optimization problem. The core goal is to determine the optimal scheduling strategy to meet the fire - fighting demand by simulating the dynamic changes in the behavior of fires and fire trucks. Among them, the constraint conditions include: fire hydrants, as key service resources, are exclusive, and only support single - node water replenishment services during the same period; the service composition needs to ensure that the overall fire - fighting efficiency is always higher than the fire spread rate. Assume that the fire intensity at the fire scene is represented by Q, and the fire changes dynamically over time: when the fire - fighting efficiency is insufficient, the fire will spread and gradually intensify; when the fire - fighting efficiency is higher than the fire spread rate, the fire will gradually weaken until it is extinguished. The fire truck initially carries a limited amount of water, with a water spraying rate of P. When the water is exhausted, it needs to go to the fire hydrant to replenish water and then return to continue fighting the fire. The number of fire hydrants is limited, and each time only one vehicle can be replenished with water.

[0075] Furthermore, water replenishment takes time. Therefore, the constraint on the water replenishment rate r ∈ [15, 30].

[0076] Thus, the water replenishment time can be obtained as:

[0077]

[0078] where V is the total water storage capacity of the fire truck, W i is the total water consumption of the i - th vehicle, and r is the water replenishment rate.

[0079] Then, the total water replenishment time can be obtained as:

[0080] T = T r + 2×T w

[0081] where T r is the water replenishment time, T w is the round - trip time for water replenishment, which is fixed at T w = 20 in this embodiment.

[0082] The step S2 of constructing a dynamic non - linear service composition model according to the fire intensity and fire truck parameters includes the following steps:

[0083] S21. Set the fire truck as a service node and construct the service QoS attributes according to the fire truck parameters;

[0084] The service node is the FaaS unit. Among them, each fire truck is regarded as an FaaS: and assume that two fire truck services FaaS i and FaaS j are composable. Therefore, when a single FaaS cannot meet the user's fire - fighting demand, multiple FaaS units need to be combined according to the fire intensity Q(t).

[0085] Each service node fire truck has heterogeneous service capabilities, and parameters such as the water cannon flow rate and water storage capacity constitute the service QoS attributes. The service nodes initially carry limited resources and need to dynamically trigger the water replenishment service combination when performing fire extinguishing services. The system evaluates the status of each service node in real time through the service combination engine. When the resources of a certain service node are lower than the threshold, it automatically triggers the reorganization of the "fire extinguishing - water replenishment" service chain. The service combination optimization needs to meet the following constraints: The fire hydrant, as a key service resource, has exclusivity and only supports the water replenishment service of a single node during the same period; the service combination needs to ensure that the overall fire extinguishing efficiency is always higher than the fire spread rate.

[0086] In this scenario, the goal of the fire extinguishing task is to determine the optimal scheduling strategy to complete the fire extinguishing task with the least number of fire trucks within the specified time, and further optimize the scheduling strategy based on the least number of FaaS to minimize the total fire extinguishing time, while counting the total water consumption to evaluate resource consumption.

[0087] S22. Construct a fire spread sub - model according to the fire intensity;

[0088] Since the fire intensity Q is a key state variable that changes dynamically, with an initial value Q0, which is used to simulate the fire level of the fire. It is updated with the change of the fire extinguishing rate q total and the fire spread rate x.

[0089] In the real - world scenario, the spread rate of the flame is affected by various factors. In this embodiment, the specific value of the flame spread speed is not explored. Therefore, we assume that there is a non - linear relationship between the fire spread speed and the fire intensity. Therefore, the expression of the fire spread sub - model described in this embodiment is as follows:

[0090] x = αQ 2 +βQ

[0091] where x is the fire spread rate, Q is the fire intensity, and α and β are the fire spread coefficients;

[0092] In this embodiment, α = 0.000028 and β = 0.016.

[0093] Based on the basic law that the fire spread rate accelerates with the increase of the fire intensity, assume an idealized dynamic fire model for the fire spread sub - model.

[0094] Although the specific impacts of environmental conditions and combustion materials need to be considered in practical applications, in the present invention, the fire trend is described by a simplified dynamic formula to facilitate the study of the core issues of the fire extinguishing scheduling strategy. In subsequent practical applications, the fire model can be further corrected by combining real - world fire data to improve its practicality.

[0095] S23. Construct a fire extinguishing rate sub - model according to the fire truck parameters;

[0096] The fire truck is initially fully loaded with water, and the water spraying rate p can be dynamically adjusted, where p ∈ [20, 30].

[0097] It is assumed that the fire extinguishing efficiency is related to the non - linear relationship between the water spraying rate p and the fire intensity Q. The stronger the fire, the lower the fire extinguishing effect per unit of water volume. Therefore, the expression of the fire extinguishing rate sub - model is as follows:

[0098]

[0099] Among them, q is the fire extinguishing rate, p is the water spraying rate, Q is the fire intensity, and μ, θ, η are fire extinguishing coefficients;

[0100] In this embodiment, μ = 15, θ = 0.001, and η = 0.1.

[0101] S24. Build a service composition engine.

[0102] The service composition engine dynamically plans the water spraying rate and water replenishment behavior of the fire truck. When the water volume is insufficient, the fire truck needs to select the water replenishment timing according to the dynamic demand of the fire.

[0103] The water replenishment behavior should minimize the interference to the fire extinguishing efficiency as much as possible. The service composition engine in this embodiment includes: 1. Replenish water probabilistically when the water volume is lower than 20% and 10%; 2. Forcefully replenish water when the water volume is exhausted; 3. Allow early return to fire extinguishing after partial water replenishment.

[0104] Through the dynamic non - linear service composition model, it can be obtained that in the fire extinguishing behavior, when the total fire extinguishing rate is greater than the fire spreading rate, that is, q total >x t the fire intensity decreases: <gro

[0105] <gro <gro <gro

[0106] When the total fire extinguishing rate is less than the fire spreading rate, that is, q total <x t the fire intensity increases: <gro

[0107] <gro <gro <gro

[0108] Among them, δ and ε are fire intensity change coefficients, obtained according to the fire extinguishing rate sub - model and the fire spreading sub - model. In this embodiment, δ = 6 and ε = 3.

[0109] The dynamic change of the fire intensity Q is affected by various factors. To simplify the analysis and highlight the dynamic relationship between the fire extinguishing rate and the fire spreading rate, in this embodiment, it is assumed that the change of Q follows an iterative model based on the physical characteristics of the fire. The quadratic relationship model between the spreading rate x and the fire intensity Q reflects the rapid spreading trend when the fire is strong, and the Sigmoid-type fire extinguishing formula describes the process of gradually suppressing the fire by the fire extinguishing behavior. Although this model is not verified by experimental data in the future, based on its reasonable physical assumptions, it can provide theoretical support for the optimization of fire extinguishing strategies.

[0110] S3. Run the two-stage hybrid optimization algorithm to generate scheduling instructions according to the dynamic non-linear service composition model;

[0111] In this step, run the two-stage hybrid optimization algorithm to optimize the following parameters to reduce the waste of fire-fighting resources, increase the utilization rate of fire-fighting resources and increase the fire extinguishing efficiency:

[0112] 1. The number of nodes N of the least FaaS min :

[0113] N min = min{N: Q t ≤ 0}

[0114] 2. The shortest fire extinguishing time T min :

[0115] T min = min{Q t ≤ 0|N = N min}

[0116] 3. The total water consumption W total :

[0117]

[0118] where Q t is the fire intensity, and p i,t is the water spraying rate of the i-th fire truck running for t time;

[0119] In the present invention, the first stage is the greedy strategy and the second stage is the differential evolution algorithm, so it is called the two-stage hybrid optimization algorithm.

[0120] Among them, the greedy strategy, in English it is Greedy Strategy, is a method of algorithm design. Its core lies in only considering the optimal solution of the current stage in each step of the decision-making process. This method makes local optimal choices in order to achieve the purpose of global optimal solution.

[0121] Differential Evolution Algorithm, abbreviated as DE algorithm in English, is an evolutionary algorithm widely used in solving global optimization problems. It is particularly good at dealing with complex, non-linear and multi-constrained optimization problems. In the context of fire truck fire extinguishing scheduling, the differential evolution algorithm can be used to improve fire extinguishing scheduling and water replenishment strategies, aiming to determine the shortest completion time and optimal strategy for fire extinguishing tasks while minimizing the number of FaaS used. Among many multi-objective optimization algorithms, differential evolution shows its unique advantages: short running time, simplicity of structure and fast convergence speed. In terms of convergence speed and robustness, the performance of the DE algorithm on many commonly used test functions and real-world problems is usually more excellent compared to other optimization algorithms.

[0122] The greedy strategy shows excellent performance in solving local optimization problems. For example, in scenarios such as determining the priority of water replenishment and dynamically adjusting the water spraying rate, the greedy strategy can quickly make the current optimal decision, with high real-time performance and decision-making efficiency. The differential evolution algorithm shows significant advantages in solving global optimization problems. When optimizing the overall scheduling strategy, such as determining the optimal water spraying rate and water replenishment trigger conditions, this algorithm can explore the global solution space with the goal of minimizing the fire extinguishing time T and avoid falling into local optimal solutions.

[0123] Combining the greedy strategy with the differential evolution algorithm can not only use the greedy strategy to quickly solve sub-problems and improve the computational efficiency of the algorithm, but also optimize the global strategy through the differential evolution algorithm to enhance the practicality of global optimization. The greedy strategy makes the algorithm closer to actual needs through simple and efficient local decisions, while the differential evolution algorithm effectively expands the global search ability of the solution and avoids the algorithm falling into local optima. The combination of the two can significantly accelerate the convergence process of the algorithm and improve the overall quality of the solution.

[0124] The step S3 of running the two-stage hybrid optimization algorithm to generate scheduling instructions according to the dynamic non-linear service composition model and fire truck parameters includes the following steps:

[0125] S31. Determine the minimum number of service nodes according to the dynamic non-linear service composition model;

[0126] In this embodiment, the minimum number of service nodes is determined by the greedy strategy.

[0127] The step S31 of determining the minimum number of service nodes according to the dynamic non-linear service composition model includes the following steps:

[0128] S311. Initialize the number of simulated fire trucks;

[0129] In this embodiment, the number of initialized fire trucks is 1;

[0130] S312. Initialize the simulated fire intensity and the water volume of the simulated fire trucks according to the dynamic non - linear service composition model;

[0131] Initialize the simulated fire intensity as the fire intensity simulation in the dynamic non - linear service composition model, denoted as Q0; initialize the water volume of the simulated fire trucks as the water volume of the fire trucks in the dynamic non - linear service composition model, denoted as W0.

[0132] S313. Calculate the fire spread rate according to the simulated fire intensity;

[0133] Use the fire spread sub - model in the dynamic non - linear service composition model to calculate, and obtain the fire spread rate x t .

[0134] S314. Calculate the total fire - extinguishing rate according to the fire - extinguishing rate sub - model, the water volume of the simulated fire trucks and the number of simulated fire trucks;

[0135] According to the fire - extinguishing rate sub - model, the water volume of the simulated fire trucks and the number of simulated fire trucks, the fire - extinguishing rate q of each fire truck can be obtained i , so as to obtain the total fire - extinguishing rate q total = ∑q i .

[0136] S315. Update the simulated fire intensity according to the fire spread rate and the total fire - extinguishing rate;

[0137] In the dynamic non - linear service composition model, the calculation formula for fire update has been determined. Then, in this embodiment, the expression for updating the simulated fire intensity is as follows:

[0138]

[0139] S316. Judge whether the fire is extinguished successfully according to the simulated fire intensity. If so, record the current number of simulated fire trucks and obtain the minimum number of service nodes. Otherwise, increase the number of simulated fire trucks and return to step S312.

[0140] Among them, the specific judgment value for judging whether the fire is extinguished successfully is that if Q T ≤0 and T≤1000, then the fire is extinguished successfully, and record the current number of simulated fire trucks as the minimum number of service nodes N min ; otherwise, if the time limit is exceeded or the fire spreads before the fire is extinguished, it is a fire - extinguishing failure. Increase the number of fire trucks N and return to step S312.

[0141] Repeat the calculation until the number of simulated fire trucks can extinguish the fire successfully, and the minimum number of service nodes is obtained.

[0142] S32. Optimize the water replenishment trigger conditions and water spraying rates of service nodes according to the minimum number of service nodes and the dynamic non-linear service combination model to obtain the optimal water spraying rate, the optimal water replenishment strategy, and the shortest fire extinguishing time;

[0143] In this embodiment, the global scheduling strategy of all service nodes is optimized by the differential evolution algorithm, and at the same time, the greedy strategy is combined to dynamically adjust the water spraying rate and the water replenishment trigger conditions to adapt to the real-time changes of the fire situation, so as to achieve the dual optimization of scheduling efficiency and fire extinguishing effect.

[0144] As Figure 2 shown, the S32. Optimize the water replenishment trigger conditions and water spraying rates of service nodes according to the minimum number of service nodes and the dynamic non-linear service combination model to obtain the optimal water spraying rate, the optimal water replenishment strategy, and the shortest fire extinguishing time includes the following steps:

[0145] S321. Initialize the population according to the minimum number of service nodes. The population includes several scheduling strategy individuals, and each scheduling strategy individual includes a water replenishment trigger condition and a water spraying rate;

[0146] Generate a set according to the minimum number of service nodes:

[0147]

[0148] Among them, X i is an individual, representing a scheduling strategy, p i is the water spraying rate of each service node, and the threshold i is the water replenishment trigger condition of each service node, that is, the water volume percentage.

[0149] Then randomly generate the population:

[0150]

[0151] Among them, N p is the population size.

[0152] S322. Set the fitness function;

[0153] In this embodiment, the fitness function f(x) is set to represent the fire extinguishing completion time T: f(x) = T, subjecto Q T ≤0; if Q T >0, the fire extinguishing fails, and a large penalty value is given.

[0154] S323. Adjust the water spraying rate, water replenishment trigger condition, and water replenishment priority of the scheduling strategy individuals in the population according to the dynamic non-linear service combination model;

[0155] In this embodiment, the expression for adjusting the water spraying rate of the scheduling strategy individual is as follows:

[0156]

[0157] Among them, p i is the water spraying rate of the i-th service node in the scheduling policy individual, p is the water spraying rate of the fire truck parameter, and Q t is the fire intensity.

[0158] The water replenishment trigger condition is adjusted to:

[0159] W i ≤ threshold i × W0

[0160] Among them, W i is the water storage volume of the current service node, and W0 is the initial water storage volume of the service node.

[0161] The water replenishment priority is adjusted to:

[0162] c i = arg minW i

[0163] Among them, c i represents the water replenishment priority.

[0164] S324. Perform mutation, crossover, and selection operations on the population;

[0165] In this embodiment, the mutation operation is:

[0166] Generate a mutation vector according to the differential evolution formula:

[0167]

[0168] Among them, are different individuals randomly selected from the population.

[0169] The crossover operation is:

[0170] Cross the mutation vector V i with the current individual X i to generate a trial vector:

[0171]

[0172] Compare the fitness values of the current individual X i and the experimental individual and the experimental individual U i and select the better one to enter the next generation:

[0173]

[0174] S325. Set the maximum number of iterations, and repeat step S324 until the number of repetitions reaches the maximum number of iterations or the fitness function converges;

[0175] In this embodiment, the maximum number of iterations is set to 100.

[0176] S326. Obtain the optimal water spraying rate, the optimal water replenishment trigger condition, and the shortest fire extinguishing time according to the population;

[0177] At this time, the population has reached the optimal state. Selecting the optimal individual from the population can obtain the optimal water spraying rate the optimal water replenishment trigger condition and the shortest fire extinguishing time T.

[0178] S327. Generate the optimal water replenishment strategy according to the optimal water replenishment trigger condition and the water replenishment priority.

[0179] Combining the optimal water replenishment trigger condition and the water replenishment priority to obtain the optimal water replenishment strategy P, that is, the optimal water replenishment strategy includes the water replenishment trigger condition and the water replenishment priority of each service node.

[0180] S33. Generate a scheduling instruction according to the optimal water spraying rate and the optimal water replenishment strategy.

[0181] The scheduling instruction only needs to control the water spraying rate and the water replenishment strategy of the service individual.

[0182] S4. Execute the scheduling instruction, and dynamically collect the fire intensity and the fire truck parameters;

[0183] Dynamically collect the fire intensity and the fire truck parameters for real-time updating of the dynamic non-linear service composition model.

[0184] S5. Update the dynamic non-linear service composition model according to the fire intensity and the fire truck parameters;

[0185] After updating the non-linear composition model, the latest situation of the fire scene can be grasped in real time. When the service resources are sufficient, it is not necessary to maintain the original scheduling instruction and there is no need to update, so as to maintain the fire extinguishing efficiency. In the case of lack of service resources, the two-stage hybrid optimization algorithm can be run in real time to further save service resources.

[0186] S6. Repeat steps S3 to S5 until the fire intensity is less than or equal to zero.

[0187] When the fire intensity is less than or equal to zero, the fire extinguishing is successful and water supply scheduling is no longer required.

[0188] In this method, aiming to minimize the number of FaaS and its scheduling strategy is a reasonable choice that meets the requirements of the actual scenario. Although increasing the number of FaaS nodes without limit can achieve rapid fire extinguishing, this assumption does not conform to the resource constraints in reality. In actual operation, resources such as fire trucks, rescue personnel, and water sources are limited. Therefore, by optimizing the scheduling strategy to minimize the number of FaaS to complete the fire extinguishing task, not only can limited resources be reasonably utilized, but also the impact of resource waste on the handling of other emergencies can be avoided. In addition, this optimization strategy can significantly reduce the operation costs, including fuel consumption, equipment maintenance costs, and human resource investment, thus improving the economy. At the same time, the scheduling strategy of minimizing the number of FaaS has good scalability and can be used as a basic solution. When resource conditions permit, the number of FaaS can be appropriately increased to further improve the fire extinguishing speed and adapt to the dynamic changes of the fire situation. Through this step-by-step optimization method, not only can the scientific nature of the basic strategy be ensured, but also its applicability in various scenarios can be enhanced. If the goal is not to minimize the number of FaaS, it may lead to redundant resource scheduling, reduce efficiency, and even affect the fire protection demand allocation in other areas. Aiming to minimize the number of FaaS can ensure the efficiency and fairness of the scheduling plan and provide a theoretical basis for subsequent optimization. On this basis, appropriately expanding the FaaS service nodes and resources can achieve a balance between the fire extinguishing time and resource investment. Therefore, studying the minimization of the number of FaaS and its scheduling strategy not only meets the fire extinguishing requirements, but also can effectively reduce resource occupancy, improve the scheduling efficiency, provide a more economical, reasonable, and operable solution for practice, and at the same time leave room for subsequent optimization to further enhance the emergency response ability.

[0189] As Figure 3 shown, a water supply service scheduling system for an urban fire scene in a dynamic environment is used to implement the above-mentioned water supply service scheduling method for an urban fire scene, including: a central decision-making module, an edge execution module, and a fire scene feedback module;

[0190] The central decision-making module is deployed in the cloud and is used to construct and update a dynamic non-linear service combination model according to the fire intensity, and run a two-stage hybrid optimization algorithm to generate scheduling instructions according to the dynamic non-linear service combination model;

[0191] The dynamic non-linear service combination model includes service nodes, a fire spread sub-model, a fire extinguishing rate sub-model, and a service combination engine;

[0192] The service nodes are set as fire trucks, and the fire truck parameters form service QoS attributes;

[0193] Set the service node as a fire truck, and form the service QoS attributes with the fire truck parameters, which is convenient for optimization through the two-stage hybrid optimization algorithm, so as to generate scheduling instructions. Each service node has heterogeneous service capabilities, and its parameters such as water cannon flow rate and water storage capacity form the service QoS attributes. The service node initially carries limited resources and needs to dynamically trigger the water supply service combination when performing the fire extinguishing service.

[0194] The fire spread sub-model is used to summarize the fire spread rate;

[0195] The fire extinguishing rate sub-model is used to summarize the fire extinguishing efficiency;

[0196] The service composition engine is used to evaluate the status of the service node and trigger the water supply service combination according to the fire spread sub-model and the fire extinguishing rate sub-model.

[0197] The system evaluates the status of each node in real time through the service composition engine. When the resources of a certain node are lower than the threshold, the "fire extinguishing - water supply" service chain is automatically triggered for reorganization. The service composition engine is optimized through the two-stage optimization algorithm to improve the fire extinguishing efficiency.

[0198] The edge execution module is deployed in the fire truck cluster and is used to execute scheduling instructions and collect fire truck parameters in real time. The fire truck parameters include the water spraying rate and the water storage capacity;

[0199] The edge execution module includes a sensor unit and an edge computing unit;

[0200] The sensor unit is used to collect fire truck parameters;

[0201] The edge computing unit is used to execute scheduling instructions.

[0202] The fire scene feedback module is deployed on the fire scene monitoring equipment and is used to sense the fire intensity in real time.

[0203] The present invention verifies the effectiveness of a method and system for urban fire scene water supply service scheduling in a dynamic environment through simulation experiments and real-scene tests, and focuses on evaluating the QoS comprehensive index and system robustness of the service composition scheme.

[0204] For the convenience of experimental simulation, the data in reality is appropriately simplified, a typical scene with an initial fire of 2000 and a water storage capacity of 6000 is selected, and the time for the fire scene to reach the fire hydrant is set as t = 20.

[0205] The experimental results are as Figure 4 and Figure 5The results show that the FaaS model has successfully achieved fire control and final extinguishment by dynamically scheduling nodes and resources. Specifically, the fire intensity Q gradually decreases over time and is completely extinguished at approximately 700 seconds, indicating that the service composition mechanism of FaaS can efficiently complete the fire extinguishment task in a short time. The fire extinguishment rate q is always higher than the fire spread rate x throughout the process. Although there are certain fluctuations in the fire extinguishment rate, these fluctuations are closely related to the node dynamic scheduling strategy and the water replenishment cycle, effectively ensuring that the fire continues to weaken. The dynamic change in the number of FaaS further shows that in the case of a fire intensity of 2000, relying on the on-demand service supply characteristics of FaaS, only 1 to 3 nodes are required to complete the fire extinguishment task, which fully verifies the ability of the FaaS framework in elastic scheduling of node resources. At the same time, the total water consumption increases linearly with time and reaches approximately 12000 liters when the fire is extinguished, fully reflecting the guarantee role of the non-functional constraints of FaaS on the efficiency of the water replenishment mechanism and the rationality of the scheduling strategy under the premise of a water storage capacity of 6000. The periodic fluctuations in the node water levels demonstrate the alternating execution of the water replenishment and fire extinguishment tasks under the FaaS service unit coordination mechanism. The water levels of all nodes always remain within the safe range, and no resource exhaustion occurs.

[0206] The experimental results show that the present invention exhibits good adaptability and optimization performance. The optimized scheduling strategy significantly improves the water resource utilization rate, shortens the fire extinguishment time, reduces the work pressure and risks of firefighters, and provides a scientific basis and practical guidance for the dispatching, resource allocation, and management of fire trucks in urban fires.

[0207] The present invention proposes a differential evolution algorithm based on a two-stage multi-objective hybrid strategy for the dynamic non-linear FaaS service composition optimization problem, and conducts a comprehensive verification and analysis in the complex dynamic environment of an urban fire scene. By establishing an Internet of Things service composition optimization model, combining the greedy strategy and the differential evolution algorithm, the dynamic optimization of the service composition plan is successfully achieved, effectively improving the service resource utilization efficiency and the comprehensive QoS index. Generating and executing scheduling instructions can scientifically configure fire-fighting resources, improve the fire extinguishment efficiency, and is of great significance for enhancing the FaaS fire rescue ability and ensuring the safety of people's lives and property.

Claims

1. A method for scheduling urban fire water supply services in a dynamic environment, characterized in that, It includes the following steps: S1. Obtain the fire intensity and fire truck parameters, where the fire truck parameters include the water spraying rate and the water storage capacity; S2. Construct a dynamic non-linear service combination model based on the fire intensity and fire truck parameters. The dynamic non-linear service combination model includes service nodes, a fire spread sub-model, a fire extinguishing rate sub-model, and a service combination engine; S3. Run a two-stage hybrid optimization algorithm to generate a scheduling instruction according to the dynamic non-linear service combination model; S4. Execute the scheduling instruction and dynamically collect the fire intensity and fire truck parameters; S5. Update the dynamic non-linear service combination model according to the fire intensity and fire truck parameters; S6. Repeat steps S3 to S5 until the fire intensity is less than or equal to zero.

2. The urban fire scene water supply service dispatching method according to claim 1, characterized in that, The expression of the fire spread sub-model is as follows: x = αQ 2 + βQ Where x is the fire spread rate, Q is the fire intensity, and α and β are the fire spread coefficients; The expression of the fire extinguishing rate sub-model is as follows: Where q is the fire extinguishing rate, p is the water spraying rate, Q is the fire intensity, and μ, θ, η are the fire extinguishing coefficients.

3. The urban fire scene water supply service scheduling method according to claim 1, characterized in that, The S2. Constructing a dynamic non-linear service combination model according to the fire intensity and fire truck parameters includes the following steps: S21. Set the fire truck as a service node and constitute the service QoS attribute according to the fire truck parameters; S22. Construct a fire spread sub-model according to the fire intensity; S23. Construct a fire extinguishing rate sub-model according to the fire truck parameters; S24. Construct a service combination engine.

4. The urban fire scene water supply service dispatching method according to claim 1, wherein, The S3. Running a two-stage hybrid optimization algorithm to generate a scheduling instruction according to the dynamic non-linear service combination model and fire truck parameters includes the following steps: S31. Determine the minimum number of service nodes according to the dynamic non-linear service combination model; S32. Optimize the water replenishment trigger condition and water spraying rate of the service nodes according to the minimum number of service nodes and the dynamic non-linear service combination model to obtain the optimal water spraying rate, the optimal water replenishment strategy, and the shortest fire extinguishing time; S33. Generate a scheduling instruction according to the optimal water spraying rate and the optimal water replenishment strategy.

5. The urban fire scene water supply service dispatching method according to claim 1, characterized in that The S31. Determining the minimum number of service nodes according to the dynamic non-linear service combination model includes the following steps: S311. Initialize the number of simulated fire trucks; S312. Initialize the simulated fire intensity and the water volume of the simulated fire trucks according to the dynamic non-linear service combination model; S313. Calculate the fire spread rate according to the simulated fire intensity; S314. Calculate the total fire extinguishing rate according to the fire extinguishing rate sub-model, the water volume of the simulated fire trucks, and the number of simulated fire trucks; S315. Update the simulated fire intensity according to the fire spread rate and the total fire extinguishing rate; S316. Judge whether the fire extinguishing is successful according to the simulated fire intensity. If so, record the current number of simulated fire trucks and obtain the minimum number of service nodes. Otherwise, increase the number of simulated fire trucks and return to step S312.

6. The urban fire scene water supply service dispatching method according to claim 1, characterized in that The S32. Optimizing the water replenishment trigger condition and water spraying rate of the service nodes according to the minimum number of service nodes and the dynamic non-linear service combination model to obtain the optimal water spraying rate, the optimal water replenishment strategy, and the shortest fire extinguishing time includes the following steps: S321. Initialize the population according to the minimum number of service nodes. The population includes several scheduling strategy individuals, and the scheduling strategy individuals include the water replenishment trigger condition and the water spraying rate; S322. Set the fitness function; S323. Adjust the water spraying rate, water replenishment trigger condition, and water replenishment priority of the scheduling strategy individuals in the population according to the dynamic nonlinear service composition model; S324. Perform mutation, crossover, and selection operations on the population; S325. Set the maximum number of iterations, and repeat step S324 until the number of repetitions reaches the maximum number of iterations or the fitness function converges; S326. Obtain the optimal water spraying rate, optimal water replenishment trigger condition, and shortest fire extinguishing time according to the population; S327. Generate the optimal water replenishment strategy according to the optimal water replenishment trigger condition and water replenishment priority.

7. The urban fire ground water supply service dispatching method according to claim 6, characterized in that, The adjustment expression of the water spraying rate of the scheduling strategy individual is as follows: where p i is the water spraying rate of the i-th service node, p is the water spraying rate of the fire truck, and Q t is the fire intensity.

8. An urban fire scene water supply service scheduling system in a dynamic environment, characterized in that, The method for scheduling urban fire field water supply services for implementing any one of claims 1-7 includes: a central decision-making module, an edge execution module, and a fire field feedback module; The central decision-making module is deployed in the cloud, and is used to construct and update the dynamic nonlinear service composition model according to the fire intensity, and run the two-stage hybrid optimization algorithm to generate scheduling instructions according to the dynamic nonlinear service composition model; The edge execution module is deployed in the fire truck cluster, and is used to execute the scheduling instructions and collect the fire truck parameters in real time. The fire truck parameters include the water spraying rate and the water storage capacity; The fire field feedback module is deployed in the fire field monitoring device, and is used to sense the fire intensity in real time.

9. The urban fire scene water supply service dispatching system according to claim 8, wherein The dynamic nonlinear service composition model includes service nodes, a fire spread sub-model, a fire extinguishing rate sub-model, and a service composition engine; The service nodes are set as fire trucks, and the fire truck parameters are used to form the service QoS attributes; The fire spread sub-model is used to summarize the fire spread rate; The fire extinguishing rate sub-model is used to summarize the fire extinguishing efficiency; The service composition engine is used to evaluate the state of the service nodes, and trigger the water replenishment service composition according to the fire spread sub-model and the fire extinguishing rate sub-model.

10. The urban fire ground water supply service dispatching system according to claim 8, characterized in that, The edge execution module includes a sensor unit and an edge computing unit; The sensor unit is used to collect the fire truck parameters; The edge computing unit is used to execute the scheduling instructions.

Images