Emergency reliable path recommendation method based on breakpoint optimization technology, medium and equipment

Abstract

The application discloses an emergency reliable path recommendation method based on breakpoint optimization technology, a medium and equipment, relates to the technical field of city public safety management and transportation, and mainly comprises the following steps: obtaining an optimized road network by optimizing a road network by using a neural network optimized by an activation function; sorting candidate paths by using path reliability to obtain a recommendation result; and periodically updating path reliability and adjusting the recommendation result according to real-time disaster and traffic data. The emergency reliable path recommendation method based on breakpoint optimization technology, the medium and the equipment can provide the most efficient evacuation and rescue paths for social vehicles and rescue vehicles when a sudden event occurs in a city road network.

Description

Technical Field

[0001] This invention relates to the fields of urban public safety management and transportation technology, and more specifically, to an emergency reliable route recommendation method, medium, and equipment based on breakpoint optimization technology. Background Technology

[0002] With the increasing complexity of urban road networks and the continuous growth of vehicle ownership, the frequency of urban road emergencies has risen significantly. These emergencies are often unpredictable, especially in terms of timing and location, posing a significant challenge to emergency response. To reduce casualties and property damage caused by emergencies, it is crucial to quickly and accurately provide top-k emergency evacuation routes for vehicles. This not only helps guide vehicles to avoid disaster areas for effective evacuation but also provides valuable passage space for rescue vehicles, thereby improving the overall efficiency of emergency response.

[0003] Emergency route recommendations play a crucial role in urban road emergency evacuation. Through rational route planning, it can be ensured that vehicles on urban roads can avoid the location of the emergency in the shortest possible time, enabling successful evacuation. More importantly, based on these route recommendations, special rescue vehicles can prioritize using the most efficient routes for rescue operations, reducing traffic congestion and ensuring the efficiency of rescue efforts at disaster sites.

[0004] Traditional path planning problems typically focus on finding the shortest or optimal path under static conditions. However, in emergency path planning, due to the complexity and unpredictability of emergencies, the reliability of the path becomes crucial. Existing technologies generally suffer from problems such as fixed number of segments, inability to adjust the structure, and insufficient optimization mechanisms, making them unable to cope with emergencies and resulting in low reliability. Summary of the Invention

[0005] The purpose of this invention is to provide an emergency reliable route recommendation method, medium, and device based on breakpoint optimization technology, which can provide the most efficient evacuation and rescue routes for social vehicles and rescue vehicles when emergencies occur in urban road networks.

[0006] This invention provides an emergency reliable path recommendation method based on breakpoint optimization technology, comprising the following steps: S1: Based on the road network topology data, historical traffic data, real-time sensor monitoring data, and disaster environmental parameters of the target area, the road network topology data is optimized using a neural network optimized by an activation function to obtain optimized road network topology data; S2: Based on the optimized road network topology data, the candidate paths are sorted by path reliability, and the paths ranked higher are used as the recommended results. S3: Based on real-time disaster and traffic data, periodically update route reliability and adjust recommendation results.

[0007] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described emergency reliable path recommendation method based on breakpoint optimization technology.

[0008] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the emergency reliable path recommendation method based on breakpoint optimization technology described above.

[0009] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described emergency reliable path recommendation method based on breakpoint optimization technology.

[0010] The emergency reliable path recommendation method, medium, and equipment based on breakpoint optimization technology provided by this invention have the following beneficial effects: This invention combines traditional route planning with dynamic analysis of real-time disaster and traffic conditions. By constructing a top-k route recommendation model, it provides reliable escape routes for vehicles in emergencies, reducing casualties and property damage, and improving rescue efficiency. This method considers not only the geometric structure and physical characteristics of roads but also multiple dimensions such as dynamic disaster impacts and traffic operation status. To establish a comprehensive route reliability assessment model, this research combines information from multiple dimensions, including structural reliability, traffic reliability, and temporal reliability, enabling real-time and dynamic route planning in complex road networks. Based on this, a multi-dimensional, multi-objective reliability calculation framework is proposed, and a route recommendation model is developed based on this framework.

[0011] This invention analyzes the situational characteristics of urban road networks and establishes a road network situational prediction model based on urban road network congestion. It also establishes an emergency reliable route recommendation model based on road network congestion mode, so that when an emergency occurs in the urban road network, it can provide the most efficient evacuation and rescue routes for social vehicles and rescue vehicles. Attached Figure Description

[0012] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings: Figure 1 This is a flowchart of the emergency reliable path recommendation method based on breakpoint optimization technology provided by the present invention; Figure 2 This is a schematic diagram of the activation function provided by the present invention; Figure 3This is a flowchart of the structural reconstruction provided by the present invention; Figure 4 This is a comparison chart of the calculation errors of the breakpoint encryption strategy provided by this invention; Figure 5 This is a comparison chart of the computational efficiency of the structural reconstruction algorithm provided by this invention; Figure 6 This is a structural block diagram of the computer device provided by the present invention. Detailed Implementation

[0013] To provide a clearer understanding of the technical features, objectives, and effects of the present invention, specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0014] Figure 1 A schematic diagram of the emergency reliable path recommendation method based on breakpoint optimization technology in this embodiment is shown. In this embodiment, the emergency reliable path recommendation method based on breakpoint optimization technology includes the following steps: S1: Based on the road network topology data, historical traffic data, real-time sensor monitoring data, and disaster environmental parameters of the target area, the road network topology data is optimized using a neural network with the activation function as the optimization activation function to obtain optimized road network topology data; In one exemplary embodiment, the road network topology data includes lane capacity ratio and road design parameters; the historical traffic data includes historical accident statistics; the real-time sensor monitoring data includes the number of vehicles and vehicle speed; and the disaster environment parameters include weather, earthquake intensity, and water depth. In one exemplary embodiment, the process of generating the optimized activation function includes: The input domain of the activation function in the neural network is divided into multiple sub-intervals, and a linear function segment is constructed in each sub-interval. Construct an initial set of breakpoints in the input domain, obtain feature-sensitive regions based on lane capacity ratio, number of vehicles, vehicle speed, and weather, and encrypt the initial breakpoints based on the feature-sensitive regions; Constraints are applied to adjacent road segments at the breakpoints; the constraints include function value continuity and derivative continuity conditions; the stringency of the constraints is dynamically adjusted according to the structural reliability of the path. Reconstruct the structure of all segments of the linear function segment to obtain the optimized activation function; In one exemplary embodiment, the structural reconstruction of all segments of the linear function segment includes the GADR algorithm and the SRADR algorithm. In one exemplary embodiment, the GADR algorithm is as follows: , If a certain segment gradient variance Then the segment will be... Split into two new linear segments, and reinitialize the slope and bias of the new linear segments; like If so, then merge adjacent segments; in, For gradient variance, Expressing expectations; Indicates segmentation The gradient function; This indicates the preset splitting threshold; This indicates the preset merging threshold; In one exemplary embodiment, the SRADR algorithm is as follows: , , when When, stop gradient update; when When merging, the corresponding segments are merged; in, For segmentation Segment reliability; For segmentation Path time deviation; This refers to the actual travel time. This is the baseline travel time when there is no congestion. For segmentation Related road environment parameters; This is a coefficient that balances error and instability; The preset segment reliability threshold; S2: Based on the optimized road network topology data, the candidate paths are sorted by path reliability, and the paths ranked higher are used as the recommended results. In one exemplary embodiment, the path reliability is as follows:

[0015] , , ,

[0016]

[0017] in, For structural reliability; It is an exponential function; This is a disaster intensity correction factor; and For material coefficients; Earthquake intensity; The depth of the flood; This refers to the actual travel time. For free-flow time; and These are the standard parameters for the BPR function; For real-time traffic; For effective traffic capacity; For reliable passage; Represents probability; This indicates the acceptable time deviation threshold; To improve traffic reliability by utilizing historical congestion data; Historical congestion rate of the road segment; For time reliability; This represents the maximum tolerable time deviation rate. For path reliability; , and These are the weighting coefficients for structural reliability, traffic reliability, and time reliability, respectively. S3: Based on real-time disaster and traffic data, periodically update route reliability and adjust recommendation results.

[0018] In some embodiments, the emergency reliable path recommendation method based on breakpoint optimization technology described above can also be implemented in the following ways.

[0019] In this embodiment, the emergency reliable path recommendation method based on breakpoint optimization technology includes the following steps: Step 1: Input and Preprocessing: Acquire road network topology data, historical traffic data, real-time sensor monitoring data, and disaster environmental parameters.

[0020] Step 2, Activation Function Optimization: By adaptively optimizing the piecewise linear activation function, the accuracy of path reliability calculation and online inference efficiency are improved to support real-time sorting and selection of highly reliable paths in the candidate path set; Limited by computing resources and timeliness requirements, the inference speed of existing models is difficult to adapt to the rapid changes in real-world scenarios, resulting in lags in evacuation and rescue decision-making and response. To address this, this embodiment proposes a model optimization method based on an improved activation function, aiming to reduce online computational overhead and accelerate the inference process, thereby achieving near real-time path recommendation and rescue scheduling support without significantly sacrificing prediction accuracy.

[0021] Activation functions, as the core module introducing nonlinearity in neural networks, directly affect the expressive power and training effect of traffic vehicle models. Traditional activation functions such as ReLU, Tanh, and Sigmoid have fixed structures, making it difficult to adapt to different input feature distributions. In recent years, some works have attempted to improve their flexibility by introducing parameters (such as PReLU) or piecewise linear structures (such as APL), but these still generally suffer from problems such as fixed number of segments, inability to adjust the structure, and insufficient optimization mechanisms. To address this, this embodiment proposes a piecewise linear activation function method that integrates structural adaptation and gradient feedback mechanisms, and combines it with dynamic optimization algorithm design to achieve synchronous optimization and self-adjustment of structural parameters and model training.

[0022] like Figure 2 The diagram shown is a schematic of the activation function; First, the input domain of the activation function in the neural network is divided into several sub-intervals, and a linear function segment is constructed in each sub-interval, with the function expression being Equation (1); (1) in, For the first The slope of the segment For the first Segment offset.

[0023] Secondly, an initial set of breakpoints is constructed in the input domain, and breakpoints are densified in feature-sensitive areas based on lane capacity ratio, number of vehicles, vehicle speed, weather, etc.

[0024] Furthermore, the continuity of function values ​​and derivatives are applied to adjacent segments at the breakpoints, and the strictness of the constraints is dynamically adjusted according to the reliability of the path structure. Structural reliability is quantified using parameters such as road design parameters (width, materials), real-time disaster data (earthquake intensity, water depth), and historical accident statistics.

[0025] Finally, all segments are restructured using the GADR and SRADR algorithms, with the reconstruction period T dynamically adjusted according to the disaster situation.

[0026] The GADR algorithm calculation process is as follows: Gradient fluctuation-based structural reconstruction; The gradient variance calculation function is given by equation (2); if a certain segment of gradient variance If the model error is large or the local structure is complex within this interval, a refinement operation is performed, splitting the segment into two new linear segments and reinitializing the corresponding slope and bias; if T consecutive rounds satisfy Then merge adjacent segments; (2) The SRADR algorithm calculation process is as follows: Segment refactoring based on segment reliability indicators; The segment reliability calculation function is given by equation (3), in the high reliability segment Stop gradient updates; in low-reliability segments The corresponding segments will be automatically merged. (3) in, Path time deviation , To correlate road environmental parameters, This is a coefficient that balances error and instability.

[0027] like Figure 3 The diagram shown is a flowchart of the structural restructuring process; Step 3: Path Reliability Model Calculation: Path reliability is calculated from three dimensions: structural reliability, traffic reliability, and time reliability. The candidate paths are then sorted based on their reliability values, and the top k optimal paths are selected.

[0028] The timeliness and safety of rescue operations are highly dependent on the road network's capacity to provide support, and route reliability, as a core indicator for quantifying this capacity, is crucial in disaster emergency rescue scenarios. It is defined as the probability that rescue vehicles will safely and on time arrive at their destination along a selected route under a specific disaster scenario, and needs to be comprehensively evaluated from three dimensions: structural reliability, traffic reliability, and time reliability.

[0029] Step 1: Path reliability emphasizes the ability of a path to meet both the needs of rescue time window and safe passage under the constraints of disaster environments such as road damage, secondary disaster risks, and traffic chaos.

[0030] First, we will analyze the data from three core dimensions to ensure the comprehensiveness and accuracy of subsequent calculations; 1) Structural Reliability: This dimension focuses on the resilience and structural integrity of the physical infrastructure along the route under disaster conditions.

[0031] Calculation logic: an exponential decay model based on road design parameters (width, slope) and real-time disaster intensity (earthquake intensity, flood depth).

[0032] The calculation formulas are shown in (4) and (5) respectively: (4) For the defined disaster intensity correction factor (e.g., when the earthquake intensity is ≥ 7), Flood depth hour ); The calculation formula is based on the inherent vulnerability of roads: (5) The average height of buildings along the route (m) indicates that an increase in high-rise buildings will increase the risk of collapse. Effective road width (m), narrow road section Low value leads to Increase; Building density (buildings / km), dense area The risk increases sharply at times; The formula for calculating the damage exposure level is as follows: Quantify the scope of the disaster's impact; 2) Trafficability Reliability: This dimension focuses on the probability and efficiency of maintaining connectivity and providing effective traffic capacity after a disaster, as shown by Equations (6) and (7), respectively.

[0033] Calculation formula: (6) Free-flow time (baseline travel time when there is no congestion); The actual travel time is dynamically updated using the BRP function: (7) For real-time traffic ( It updates every 5 minutes via IoT sensors; To design traffic capacity ( ); These are the standard parameters for the BPR function; The acceptable time deviation threshold (usually 0.3) is the probability that the actual time does not exceed 30% of the free flow time. 3) Temporal Reliability: This dimension is directly related to the golden time requirement for rescue, referring to the probability that the actual travel time of the rescue vehicle on the route meets or is less than the rescue time limit requirement.

[0034] Key metrics: Time standard deviation Reflects the time-varying characteristics of the road network. Considered high risk; Time guarantee rate: ≥90%; Step 2: Based on the above definition of path reliability and analysis of core dimensions, the structural reliability is first modified. Using the dynamic coupling model of road physical properties and real-time disaster intensity, the probability of road damage follows an exponential decay law, as shown in Equation (8).

[0035] The core formula is: (8) Key parameter analysis: :earthquake intensity( (High-risk correction triggered at times); Flood depth ( (Activate hydrological corrections in a timely manner); Disaster intensity correction factor (earthquake intensity) hour Flood depth hour ); , Material coefficient (concrete road) steel structure bridge ).

[0036] Engineering significance: parameters and High-grade roads demonstrate the disaster resistance and resilience of materials by improving... The value reduces the attenuation rate.

[0037] accomplish: Disaster data is acquired in real time through a GIS system and updated every 5 minutes. and ; like (Medium damage threshold) automatically triggers emergency repair dispatch instructions.

[0038] Step 3: Based on the structural reliability correction, further calculate the traffic reliability. Combining the superimposed impact of real-time traffic flow fluctuations and road physical damage on traffic capacity, an improved BRP function is adopted. Through real-time monitoring and calculation, the current traffic reliability can be obtained, as shown in Equation (9).

[0039] 1) Traffic reliability: (9) Dynamic variables: For real-time traffic (dynamically collected by IoT sensors); For effective traffic capacity ; Acceptable time deviation threshold (default) ).

[0040] accomplish: when The congestion response mechanism is activated in a timely manner, with reliability decreasing in stages: (like ; (like .

[0041] Step 4: In order to improve the accuracy and reliability of the model, historical congestion data (the GIS database stores events from the past 5 years) is combined with the real-time calculation of traffic reliability to make corrections and avoid model distortion caused by occasional congestion, as shown in Equation (10).

[0042] (10) Historical congestion rate of a road segment (e.g., a road segment with a historical congestion rate of 20%) (Reduced by 20%).

[0043] Implementation: Establish a spatiotemporal database of blocking events, and classify and match the current disaster mode according to disaster type (earthquake / flood).

[0044] Step 5: After obtaining the corrected travel reliability, further ensure that the travel time fluctuation strictly meets the golden time window for rescue, and trigger rerouting by monitoring the standard deviation, as shown in Equation (11).

[0045] Time reliability model: (11) Maximum tolerable time deviation rate (default) ; Time guarantee rate: required during major disasters .

[0046] Dynamic adjustment: Hourly updates of travel time standard deviation ; like If the path is identified as high-risk, a rerouting algorithm will be initiated.

[0047] Step 6: After obtaining the three types of index values ​​for the path, use the improved path reliability algorithm model ( A candidate set is generated in the road network, and paths with fragile structures or risks of passage are eliminated during the generation stage to ensure the feasibility of the candidate paths, as shown in Equation (12).

[0048] (12) in, , , The weighting coefficient is determined based on the specific disaster scenario and rescue needs, and generally meets the following requirements: ).

[0049] Subsequently, the overall reliability of each candidate path is calculated and ranked, and the top K paths are selected as the recommended results. The value of K can be flexibly set according to the urgency of the rescue mission and environmental uncertainties. When the disaster evolves rapidly or traffic conditions are complex, the system will output more redundant paths to enhance robustness; while when resources are limited or time is tight, only a few high-reliability paths will be recommended to simplify decision-making. Compared with the traditional method of providing only a single path, this mechanism not only ensures the timeliness and safety of the optimal path, but also provides alternative solutions, thereby improving the robustness and fault tolerance of emergency rescue operations.

[0050] Step 4: Dynamic Updates and Output: Based on real-time disaster and traffic data, the reliability indicators are updated periodically, and the recommendation results are adjusted.

[0051] Two types of simulation experiments will be conducted below. The first experiment verifies the practical effect and quantization improvement of the proposed sensitive breakpoint encryption strategy. The second part verifies the feasibility and efficiency improvement of the structure reconstruction algorithm.

[0052] (1) Sensitive breakpoint encryption strategy: In urban emergency reliable route recommendation and intelligent traffic evacuation scenarios, this embodiment constructs a feature-sensitive area breakpoint encryption strategy based on multiple factors such as lane capacity ratio, number of vehicles, speed, and weather. By dynamically and adaptively dividing the input domain, it accurately captures key feature changes in complex traffic environments, achieving a dual improvement in path reliability calculation accuracy and emergency response efficiency.

[0053] The experiment utilizes real-time weather data from the Gaode Map API (such as rainwater accumulation depth and typhoon wind force levels) and floating car trajectory speed-density distribution (dynamically reflecting traffic flow status). It integrates the road's inherent vulnerability index (statically characterizing road disaster resistance) and building density (related to personnel evacuation pressure) to construct a four-dimensional feature-sensitive area identification model. For typical scenarios such as earthquakes, rainwater accumulation, typhoon impacts, and complex disasters, the path reliability calculation errors of uniformly spaced and densely spaced breakpoints are tested respectively.

[0054] from Figure 4The results show that in earthquake scenarios, increasing the number of breakpoints reduced the error from 0.32 to 0.18, a reduction of 43.75%; in rainstorm and waterlogging scenarios, the error was reduced from 0.28 to 0.15, a reduction of 46.43%; in typhoon impact scenarios, the error was reduced from 0.35 to 0.22, a reduction of 37.14%; and in complex disaster scenarios (multi-factor superposition interference), increasing the number of breakpoints compressed the error from 0.41 to 0.25, a reduction of 39.02%.

[0055] Experimental results show that by fusing multi-source data and dynamically encrypting breakpoints, this strategy effectively improves the accuracy of path reliability calculation in complex disaster scenarios, providing more reliable data support for subsequent emergency path planning and intelligent traffic evacuation decisions, and solving the technical pain point that traditional uniform breakpoints are difficult to adapt to multi-factor mutations. (2) Structural Reconstruction Algorithm: In urban emergency route planning, the GADR and SRADR algorithms proposed in this invention achieve dynamic optimization of the segmented structure of the activation function through a dual-driven mechanism of gradient variance and segment reliability. For example, when the lane volume ratio exceeds... At that time, the GADR algorithm uses gradient variance The condition splits the linear segment into two sub-segments, accurately capturing sudden changes in traffic density and speed; while when And segment reliability At the same time, the SRADR algorithm automatically merges low-reliability segments and applies continuity constraints on function values ​​and derivatives at breakpoints to ensure the mathematical rigor of the model calculation. Table 1 shows the experimental data based on 174G taxi trajectory data from Beijing from 2020 to 2023, extracting four types of sensitive scenarios.

[0056] Table 1: Sensitive Scenarios and Sample Size

[0057] Experimental results show that the computational efficiency of GADR and SRADR algorithms is improved compared with traditional methods (e.g., Figure 5 As shown in the figure, compared to fixed segments, it takes less time and demonstrates higher efficiency.

[0058] The SRADR algorithm introduces a segment reliability index R, which is used when the segment reliability... (like Furthermore, when the road vulnerability index is high, low-reliability segments are automatically merged, and the model structure is optimized through function continuity constraints. This mechanism avoids the consumption of training resources by invalid segments in traditional fixed segments. For example, when training a model involving 174G of taxi trajectory data, it can reduce redundant parameter iterations by approximately 41%. Combined with the gradient feedback mechanism of the GADR algorithm, the two simultaneously optimize the segmented structure of the activation function, reducing model training time by 40.9%, significantly shortening the model iteration cycle in emergency scenarios, and improving the system's adaptability to dynamic disaster environments.

[0059] The piecewise linear activation function based on gradient feedback and segment reliability optimization significantly improves the accuracy of path reliability calculation and online inference efficiency, enabling the system to generate highly reliable candidate paths in real time under complex disaster environments, and provide optimal and alternative paths through path reliability recommendation, thereby enhancing the robustness and response speed of emergency rescue path selection.

[0060] Model implementation results: 1. Precise Route Recommendation: In emergency rescue scenarios, this model accurately recommends the most reliable routes for rescue vehicles based on real-time data and comprehensive evaluation results. It integrates multi-source data, including road physical attributes, real-time disaster intensity, and traffic flow, and uses advanced algorithms to dynamically calculate the reliability of each route. The model can quickly identify road damage and congestion, selecting safe and unobstructed routes to ensure rescue vehicles avoid dangerous and congested sections, thereby significantly improving the timeliness and safety of rescue operations and saving valuable time for rescue efforts.

[0061] 2. Enhanced Road Network Resilience: By quantifying risk tolerance and integrating redundant routes with emergency repair coordination mechanisms, this model effectively enhances the road network's resilience under disaster conditions. It comprehensively assesses the structural reliability, traffic reliability, time reliability, and risk tolerance of routes, accurately identifying weak points in the road network. Based on the assessment results, redundant routes are planned in advance to ensure that rescue missions can continue via alternative routes when critical road sections are interrupted. Simultaneously, the model coordinates with emergency repair resources in real time to quickly repair damaged road sections and promptly restore their traffic capacity, significantly reducing the impact of localized interruptions on rescue missions and effectively ensuring the smooth flow of rescue lifelines.

[0062] 3. Decision Support Optimization: This model provides comprehensive and accurate decision support for emergency response decision-makers. It collects and analyzes massive amounts of data in real time, including road damage levels, traffic congestion, disaster types and intensities, generating an intuitive visualization interface that clearly displays the reliability information of each route and the overall road network status. Based on this data, decision-makers can quickly formulate scientific and reasonable emergency evacuation plans and resource allocation schemes, optimize the allocation of rescue forces, and improve rescue efficiency. Furthermore, the model's dynamic update mechanism ensures that decision-makers always make judgments based on the latest information and adjust rescue strategies in a timely manner, thereby significantly improving overall emergency response efficiency and management level, and minimizing disaster losses. Table 2 shows the comparison results of various indicators for reliable routes.

[0063] Table 2: Four-Dimensional Quantitative Comparison of Path Reliability

[0064] Key advantages: MIPR dynamically couples disaster data (such as aftershock probability updates) To avoid static models being affected by... Overestimation leads to distortion of traffic capacity; Traffic reliability improved by 28% due to damage to the BRP function's associated structure and reduced traffic capacity. Table 3 shows the comparison results of various indicators of resource allocation and life-saving effectiveness.

[0065] Table 3: Resource Allocation and Life-Saving Efficiency

[0066] This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the emergency reliable path recommendation method based on breakpoint optimization technology described above. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc.; the storage medium may also include combinations of the above types of memory.

[0067] This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps of the emergency reliable path recommendation method based on breakpoint optimization technology described above.

[0068] like Figure 6As shown, the computer device 120 may include: at least one processor 121, such as a central processing unit (CPU), at least one communication interface 123, memory 124, and at least one communication bus 122. The communication bus 122 is used to enable communication between these components. The communication interface 123 may include a display screen and a keyboard; optionally, the communication interface 123 may also include a standard wired interface or a wireless interface. The memory 124 may be high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Optionally, the memory 124 may also be at least one storage device located remotely from the aforementioned processor 121. The memory 124 stores application programs, and the processor 121 calls the program code stored in the memory 124 to execute any of the aforementioned method steps. The communication bus 122 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The communication bus 122 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 6The term 124 is represented by a single line, but this does not imply a single bus or a single type of bus. The memory 124 may include volatile memory, such as random-access memory (RAM); it may also include non-volatile memory, such as flash memory, hard disk drive (HDD), or solid-state drive (SSD); or it may include combinations of the above types of memory. The processor 121 may be a central processing unit (CPU), a network processor (NP), or a combination of a CPU and an NP. The processor 121 may further include hardware chips. These hardware chips may be application-specific integrated circuits (ASICs), programmable logic devices (PLDs), or combinations thereof. The PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof. Optionally, the memory 124 is also used to store program instructions. The processor 121 can invoke the program instructions to implement the emergency reliable path recommendation method based on breakpoint optimization technology as described in this embodiment.

[0069] This embodiment provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the emergency reliable path recommendation method based on breakpoint optimization technology described above.

[0070] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. An emergency reliable path recommendation method based on breakpoint optimization technology, characterized in that, Includes the following steps: S1: Based on the road network topology data, historical traffic data, real-time sensor monitoring data, and disaster environmental parameters of the target area, the road network topology data is optimized using a neural network optimized by an activation function to obtain optimized road network topology data; S2: Based on the optimized road network topology data, the candidate paths are sorted by path reliability, and the paths ranked higher are used as the recommended results. S3: Based on real-time disaster and traffic data, periodically update route reliability and adjust recommendation results; The activation function optimization process includes: The input domain of the activation function in the neural network is divided into multiple sub-intervals, and a linear function segment is constructed in each sub-interval. Construct an initial set of breakpoints in the input domain, obtain feature-sensitive regions based on lane capacity ratio, number of vehicles, vehicle speed, and weather, and encrypt the initial breakpoints based on the feature-sensitive regions; Constraints are applied to adjacent road segments at the breakpoints; the constraints include function value continuity and derivative continuity conditions; the stringency of the constraints is dynamically adjusted according to the structural reliability of the path. The structure of all segments of the linear function segment is reconstructed to obtain an optimized activation function; The reliability of the path is as follows: ， ， ， in, For structural reliability; It is an exponential function; This is a disaster intensity correction factor; and Material coefficient; Earthquake intensity; The depth of the flood; This refers to the actual travel time. For free-flow time; and These are the standard parameters for the BPR function; For real-time traffic; For effective traffic capacity; For reliable passage; Represents probability; This indicates the acceptable time deviation threshold; To improve traffic reliability by utilizing historical congestion data; Historical congestion rate of the road segment; For time reliability; This represents the maximum tolerable time deviation rate. For path reliability; , and These are the weighting coefficients for structural reliability, traffic reliability, and time reliability, respectively.

2. The emergency reliable path recommendation method based on breakpoint optimization technology according to claim 1, characterized in that, The road network topology data includes lane capacity ratio and road design parameters; the historical traffic data includes historical accident statistics; the real-time sensor monitoring data includes the number of vehicles and vehicle speed; and the disaster environment parameters include weather, earthquake intensity, and water depth.

3. The emergency reliable path recommendation method based on breakpoint optimization technology according to claim 1, characterized in that, The structural reconstruction of all segments of the linear function segment includes the GADR algorithm and the SRADR algorithm.

4. The emergency reliable path recommendation method based on breakpoint optimization technology according to claim 3, characterized in that, The GADR algorithm is as follows: ， If a certain segment gradient variance Then the segment will be... Split into two new linear segments, and reinitialize the slope and bias of the new linear segments; like If so, then merge adjacent segments; in, For gradient variance, Expressing expectations; Indicates segmentation The gradient function; This indicates the preset splitting threshold; This indicates the preset merging threshold.

5. The emergency reliable path recommendation method based on breakpoint optimization technology according to claim 3, characterized in that, The SRADR algorithm is as follows: ， ， when When, stop gradient update; when When merging, the corresponding segments are merged; in, For segmentation Segment reliability; For segmentation Path time deviation; This refers to the actual travel time. This is the baseline travel time when there is no congestion. For segmentation Related road environment parameters; This is a coefficient that balances error and instability; This is the preset segment reliability threshold.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the emergency reliable path recommendation method based on breakpoint optimization technology as described in any one of claims 1-5.

7. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the emergency reliable path recommendation method steps based on breakpoint optimization technology as described in any one of claims 1-5.

8. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the emergency reliable path recommendation method based on breakpoint optimization technology as described in any one of claims 1-5.

Images