Vehicle intelligent scheduling method and system for coping with extreme weather

Abstract

The invention provides an intelligent vehicle scheduling method and system for dealing with extreme weather, and relates to the technical field of digital traffic management, and the method comprises the steps: obtaining the vehicle position of a to-be-scheduled target vehicle, a user position set corresponding to a plurality of to-be-transported platform users, real-time traffic data and real-time weather data; on the basis of a first path planning model, determining at least one first candidate path, the position of a vehicle passing through the first candidate path and the positions of all users; for each first candidate path, adjusting path segments of the first candidate paths based on a deep reinforcement learning model to obtain corresponding corrected paths; and calculating a path comprehensive score corresponding to each corrected path based on the MCDA evaluation model, and determining a rescue receiving path of the target vehicle from each corrected path according to each path comprehensive score so as to sequentially receive each platform user. Therefore, through intelligent scheduling, it is ensured that resources can be utilized most effectively in extreme weather.

Description

Technical Field

[0001] The present application relates to the field of digital traffic management technology, and in particular to a vehicle intelligent dispatching method and system for coping with extreme weather. Background Art

[0002] As urban traffic networks become increasingly complex and the number of vehicles continues to increase, effective vehicle scheduling becomes increasingly important. Existing vehicle scheduling systems have achieved remarkable results under normal weather conditions, which can reasonably arrange vehicle routes, optimize traffic flow, and reduce congestion. However, when faced with extreme weather conditions, such as heavy rain, heavy snow, freezing rain, typhoons, etc., the effectiveness of these systems is often greatly reduced.

[0003] Extreme weather not only affects traffic flow and vehicle speed, but also leads to poor road conditions and increases the risk of traffic accidents. For example, in heavy rain, roads are prone to waterlogging and visibility is low, which obstructs drivers' vision; heavy snow may cause roads to ice up, increasing the risk of vehicles sliding. In extreme weather conditions, the core of the vehicle dispatch system should no longer be limited to the pick-up and drop-off of specific individual passengers, but should quickly reduce the number and length of stay of passengers in the affected area to protect the lives, health and safety of the people.

[0004] Traditional dispatch systems lack flexibility and adaptability in extreme weather conditions, and they usually still use fixed dispatch logic. However, in extremely bad weather, there are fewer social vehicles and rescue vehicles available for dispatch, while the number of passengers who need to be served is large. How to quickly and effectively transport passengers in disaster areas safely and implement effective rescue at a lower vehicle operating cost is an urgent problem in the industry. Summary of the invention

[0005] The present application provides a vehicle intelligent dispatching method and system for coping with extreme weather conditions, which is used to at least solve the problem that traditional vehicle dispatching solutions cannot meet the transportation needs of passenger users in extreme weather conditions.

[0006] The present application provides a method for intelligent vehicle dispatching in response to extreme weather, comprising: obtaining the vehicle position of a target vehicle to be dispatched, a user position set corresponding to a plurality of platform users to be transported, real-time traffic data and real-time weather data; the distance between each user position in the user position set and the vehicle position is less than a preset distance threshold; the real-time traffic data includes: road closure information, traffic accident information and road congestion information; in the case where it is determined that the real-time weather data includes extreme weather information, at least one first candidate path is determined based on a first path planning model, the first candidate path passes through the vehicle position and all the user positions; the extreme weather information includes any one of the following: heavy rain, heavy snow, freezing rain or typhoon; for each of the A first candidate path, at least one path segment of the first candidate path is adjusted based on a deep reinforcement learning model to obtain a corresponding corrected path; the path segment is determined according to any two user locations in the user location set; the state of the deep reinforcement learning model is defined according to the node location, real-time traffic data and real-time weather data, and the action of the deep reinforcement learning model is defined according to the switching of the path segments; the node location is the vehicle location or the user location; based on the MCDA evaluation model, the path comprehensive scores corresponding to each of the corrected paths are calculated, and according to each of the path comprehensive scores, the rescue pick-up path of the target vehicle is determined from each of the corrected paths to sequentially pick up each of the platform users.

[0007] Optionally, after obtaining the vehicle position of the target vehicle to be dispatched, the user position set of multiple platform users to be transported, real-time traffic data and real-time weather data, the method also includes: when it is determined that the real-time weather data does not contain extreme weather information, based on a second path planning model, determining at least one second candidate path, the second candidate path passing through the vehicle position and any one of the user locations; determining a passenger transport path from each of the second candidate paths to pick up and transport the platform users corresponding to the passenger transport path.

[0008] Optionally, the deep reinforcement learning model adopts a deep Q network, and the main network of the deep Q network includes a cascaded input layer, a hidden layer and an output layer; the input layer is used to normalize the data; the input data includes the real-time traffic data, the real-time weather data and the first candidate path; the hidden layer includes a multi-layer convolutional neural network, which is used to extract feature information from the input data; the output layer is used to determine the expected Q value corresponding to each action in the state space, and to correct the first candidate path according to the path segment corresponding to the action with the maximum expected Q value to determine the corresponding corrected path.

[0009] Optionally, the update formula of the deep Q network is as follows:

[0010] Q(s,a;θ)=Q(s,a;θ)+α[r+γmax a′ Q(s′,a′;θ′)-Q(s,a;θ)]

[0011] r=w time × time +w safety × safety

[0012] In the formula, Q(s,a;θ) represents the current Q value of taking action a in state s, θ represents the current network parameters, α represents the learning rate; γ represents the discount factor, which indicates the importance of future rewards; max a′ Q(s′,a′;θ′) represents the maximum Q value that can be generated by all possible actions a′ in the next state s′, θ′ represents the old network parameters; r represents the immediate reward, which is obtained according to the result of the current action; r time and r safety Represent the timeliness reward item and the safety reward item, and w time and w safety Respectively represent the weight coefficients of the corresponding reward items, and w safety >w time , where w time Encourage the model to choose path segments with shorter distances, and w safety This enables the model to select route segments with a lower probability of abnormal road conditions under real-time weather data.

[0013] Optionally, the loss function of the deep Q network is:

[0014] L(θ)=E[(y i -Q(s,a;θ) 2 ]

[0015] Time difference target y i The formula is:

[0016] y i =λ(s,a)×r+γmax a′ Q(s′,a′;θ - )

[0017] Where Q(s,a;θ) represents the Q-value estimate of the current state-action pair, provided by the main network; Q(s′,a′;θ - ) represents the Q value of the next state, estimated by the target network; θ and θ - Represent the parameters of the main network and the target network respectively; E[(y i -Q(s,a;θ) 2] represents the expected value of the square of the prediction error; λ(s,a) represents the environmental adaptability factor, which indicates the difficulty of taking a specific action a in a specific state s, and is used to adjust the weight of the immediate reward r.

[0018] Optionally, the first path planning model includes a Dijkstra algorithm module, an A* algorithm module and a Bellman-Ford algorithm module, wherein the Dijkstra algorithm module, the A* algorithm module and the Bellman-Ford algorithm module are connected in parallel to the deep reinforcement learning model.

[0019] Optionally, the MCDA evaluation model calculates the path comprehensive score corresponding to the modified path in the following manner:

[0020]

[0021] In the formula, S route1 represents the comprehensive score of the modified path route1, w i represents the weight of the i-th evaluation criterion, C i,route1 represents the standardized score of the modified route route1 on the i-th evaluation criteria; the evaluation criteria include: driving time evaluation criteria, path economic cost evaluation criteria and road safety evaluation criteria;

[0022] The calculation formula for the standardized score corresponding to the driving time evaluation standard is:

[0023]

[0024] α=1+k×WL

[0025] Where, T route1 represents the expected travel time of the modified route route1, T min and T max They respectively represent the minimum and maximum of the expected travel times of all the modified routes; α represents the adjustment coefficient, k represents a normal constant, and is used to adjust the degree of influence of the severe weather level on the adjustment coefficient; WL represents the severe weather level, which is determined by querying the real-time weather data and the severe weather level mapping table, and the severe weather level mapping table records multiple severe weather levels and corresponding weather data intervals;

[0026] The calculation formula for the standardized score corresponding to the path economic cost evaluation standard is:

[0027]

[0028] In the formula, E routel represents the expected economic cost of the modified path route1, E maxIt represents the maximum of the expected economic costs of all modified paths; the expected economic cost of the path includes the energy consumption cost and the toll fees of the toll road sections involved in the path;

[0029] The calculation formula for the standardized score corresponding to the road safety assessment standard is:

[0030]

[0031] In the formula, m j is the weight of the jth safety factor, C j,route1,sf represents the score of the modified route route1 for the jth safety factor; the safety factor includes at least one of the following: slippery road condition, historical accident record of the route, route curve information and route slope information.

[0032] The present application also provides a vehicle intelligent dispatching system for coping with extreme weather, comprising: a data acquisition unit, configured to acquire the vehicle position of a target vehicle to be dispatched, a user position set corresponding to a plurality of platform users to be transported, real-time traffic data and real-time weather data; the distance between each user position in the user position set and the vehicle position is less than a preset distance threshold; the real-time traffic data includes: road closure information, traffic accident information and road congestion information; a path planning unit, configured to determine at least one first candidate path based on a first path planning model when determining that the real-time weather data includes extreme weather information, the first candidate path passing through the vehicle position and all the user positions; the extreme weather information includes any one of the following: heavy rain, heavy snow, freezing rain or typhoon; a path correction unit, It is configured to adjust at least one path segment of the first candidate path based on the deep reinforcement learning model for each of the first candidate paths to obtain a corresponding corrected path; the path segment is determined according to any two user locations in the user location set; the state of the deep reinforcement learning model is defined according to the node location, real-time traffic data and real-time weather data, and the action of the deep reinforcement learning model is defined according to the switching of the path segments; the node location is the vehicle location or the user location; the pick-up path determination unit is configured to calculate the path comprehensive score corresponding to each of the corrected paths based on the MCDA evaluation model, and determine the rescue pick-up path of the target vehicle from each of the corrected paths according to each of the path comprehensive scores, so as to sequentially pick up each of the platform users.

[0033] The present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the intelligent vehicle dispatching method for coping with extreme weather as described in any one of the above-mentioned methods is implemented.

[0034] The present application also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the vehicle intelligent scheduling method for dealing with extreme weather as described in any one of the above is implemented.

[0035] The present application also provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned intelligent vehicle scheduling methods for dealing with extreme weather.

[0036] The present application provides a method and system for intelligent vehicle dispatching in response to extreme weather conditions to improve the efficiency and safety of vehicle dispatching in extreme weather conditions, which can produce at least the following technical effects:

[0037] (1) By acquiring real-time traffic and weather data, the system can respond quickly to extreme weather conditions and use preset distance thresholds to ensure that vehicles can efficiently approach users with the most urgent needs, thereby speeding up rescue and improving emergency response speed.

[0038] (2) First, candidate paths are selected through the path planning model, and then the deep reinforcement learning model is used to adjust these paths in real time. This can ensure maximum rescue efficiency while avoiding road closures, traffic accidents, or severe congestion caused by extreme weather, thereby improving the practicality and safety of the rescue route and enabling the system to output the most optimized rescue route.

[0039] (3) The deep reinforcement learning model can continuously adjust the path selection based on real-time data, allowing the system to maintain a high degree of adaptability and flexibility in extreme weather conditions, especially in extreme weather conditions, where poor road conditions and low visibility can significantly increase the risk of traffic accidents. The system effectively reduces the risk of accidents during the rescue process by avoiding high-risk areas and choosing relatively safe routes. Based on the dynamic adjustment capability of the deep reinforcement learning model, the path adaptability and flexibility are enhanced, enabling vehicles to effectively respond to emergencies and reduce rescue delays.

[0040] (4) Through the MCDA (multi-criteria decision analysis) evaluation model, various factors such as path length, expected time, safety, etc. can be comprehensively considered to score each correction path. By comparing these scores, the system can select the optimal rescue path and pick up platform users in order with the lowest cost, thereby ensuring the user service experience.

[0041] Through the embodiments of the present application, combined with real-time data, intelligent path planning and dynamic adjustment capabilities, the efficiency, safety and adaptability of vehicle dispatch under extreme weather conditions are significantly improved. When vehicles and rescue resources are limited, the rescue effect is maximized through intelligent dispatching, ensuring that resources can be used most effectively in extreme weather conditions, providing an effective solution for vehicle dispatch under extreme weather conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the present application or the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0043] Figure 1 A flowchart showing an example of a vehicle intelligent dispatching method for dealing with extreme weather according to an embodiment of the present application is shown;

[0044] Figure 2 A schematic diagram showing the effect of an example of candidate path planning according to an embodiment of the present application is shown;

[0045] Figure 3 A schematic diagram showing the working principle of an example of a state transfer action in a reinforcement learning model;

[0046] Figure 4 A flowchart showing an example of vehicle intelligent dispatch path planning operation according to an embodiment of the present application is shown;

[0047] Figure 5 A schematic diagram showing the structure of an example of a deep Q network according to an embodiment of the present application is shown;

[0048] Figure 6 A schematic structural block diagram of an example of a first path planning model according to an embodiment of the present application is shown;

[0049] Figure 7 A structural block diagram of an example of a vehicle intelligent dispatching system for dealing with extreme weather according to an embodiment of the present application is shown;

[0050] Figure 8 It is a structural schematic diagram of the electronic device provided by this application. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solutions and advantages of this application clearer, the technical solutions in this application will be clearly and completely described below in conjunction with the drawings in this application. Obviously, the described embodiments are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0052] Figure 1 A flowchart showing an example of a method for intelligent vehicle dispatching in response to extreme weather according to an embodiment of the present application is shown.

[0053] Regarding the execution subject of the method of the embodiment of the present application, it can be any electronic device with processing and computing capabilities, such as a computer terminal, a mobile phone terminal, etc. More specifically, in combination with the business service scenario of centralized vehicle dispatch, its execution subject can be a central server to realize the combination of real-time data, intelligent path planning and dynamic adjustment capabilities, and maximize the rescue effect through intelligent dispatch when vehicles and rescue resources are limited, ensuring that resources can be used most effectively in extreme weather, and providing an effective and new dispatch solution for extreme weather.

[0054] like Figure 1 As shown, in step S110, the vehicle position of the target vehicle to be dispatched, the user position set corresponding to the multiple platform users to be transported, the real-time traffic data and the real-time weather data are obtained.

[0055] Here, the distance between each user location in the user location set and the vehicle location is less than the preset distance threshold, and the real-time traffic data includes: road closure information, traffic accident information and road congestion information. It should be understood that the preset distance threshold can be adjusted or set according to the platform business needs or the personalized requirements of the car owner, such as 2 to 5 kilometers, to ensure that the vehicle can efficiently approach the users with the most urgent needs, thereby speeding up the rescue speed and improving the emergency response speed. In addition, the central server can update the real-time traffic data and real-time weather data by exchanging data with the third-party platform server.

[0056] In step S120, when it is determined that the real-time weather data contains extreme weather information, at least one first candidate path is determined based on a first path planning model, and the first candidate path passes through the vehicle position and all user positions.

[0057] Specifically, the extreme weather information includes any one of the following: heavy rain, heavy snow, freezing rain or typhoon. It should be understood that the extreme weather information can also be other types of information and can be set in the system, such as sandstorm weather, dense fog weather, etc.

[0058] Figure 2A schematic diagram showing the effect of an example of candidate path planning according to an embodiment of the present application is shown.

[0059] like Figure 2 As shown, there are multiple user locations to be transported (e.g., 11, 12…19) within a preset distance threshold near the vehicle location 20, and the first candidate path passing through the vehicle location and all user locations is planned to achieve rapid pick-up of all users. It should be noted that it is different from the online car-hailing route planning for a single passenger (i.e., the vehicle location is the pick-up route planning for any user location), and is similar to the carpooling pick-up route planning that needs to pick up multiple passengers at the same time to a certain extent, so as to achieve rapid pick-up of all passengers in the vicinity and reduce the waiting time of passengers in extreme weather environments.

[0060] In some implementations, a plurality of parallel path planning modules are provided in the first path planning model so as to independently complete the planning and output for the first candidate path.

[0061] In step S130, for each first candidate path, at least one path segment of the first candidate path is adjusted based on the deep reinforcement learning model to obtain a corresponding corrected path.

[0062] In some implementations, the path segment is determined based on any two user locations in the user location set (eg, 11-12, 12-15, 13-14, etc.).

[0063] Figure 3 A schematic diagram showing the working principle of an example of state transition action in a reinforcement learning model.

[0064] like Figure 3 As shown, the state transition diagram involves multiple basic states f1~f n In the state space composed of the state transitions between different basic states, for example, a1 represents the state transition from f1 to f2, a2 ​​represents the state transition from f2 to f1, a3 represents the state transition from f1 to f3, and so on. Here, the corresponding state transitions can occur based on the state transition strategy, and each state transition strategy can be used to cause different state transitions. For example, based on the state transition strategy for the basic state f1, state transitions a2 or a3 can occur.

[0065] In addition, the state range to which a basic state in the state space can be transferred to another state (also called a transferable state) is generally restricted or conditional. For example, any of f1 to f3 will not be transferred to f4 to f5. nThe state transfer occurs between the nodes, and the states that can be transferred to for state f1 are f2 and f3, etc. In combination with the present technical scenario, the node 12-15 can be adjusted to 12-13 or 12-14, but cannot be adjusted to 12-18.

[0066] In the example of the embodiment of the present application, the state of the deep reinforcement learning model (DQN) is defined according to the node position, real-time traffic data and real-time weather data, and the action of the deep reinforcement learning model is defined according to the switching of the path segment, and the node position is the vehicle position or the user position. Thus, the deep reinforcement learning model can optimize and adjust each path segment in the first candidate path, and realize adaptive state adjustment and transfer according to the state environment variables (node ​​position, real-time traffic data and real-time weather data), so that the adjusted corrected path can effectively meet the pick-up needs in extreme weather.

[0067] In some embodiments, the reward function of the deep reinforcement learning model needs to consider factors such as traffic safety and time efficiency, so as to evaluate the effect of taking a certain action in a specific state through immediate rewards, and guide the deep reinforcement learning model to output the best action in a specific state.

[0068] In step S140, the comprehensive scores of the paths corresponding to the respective correction paths are calculated based on the MCDA evaluation model, and the rescue pick-up paths of the target vehicles are determined from the respective correction paths according to the comprehensive scores of the paths, so as to sequentially pick up the users of each platform.

[0069] In some embodiments, the scoring criteria involved in the MCDA evaluation model are diverse, including, for example, safety, time efficiency, economy, etc., and also allow overlap with the DQN reward function to guide the selection of the best path solution from multiple candidate path solutions. By quantifying the value score of each corrected path, the system can comprehensively evaluate and select the optimal rescue path, pick up platform users in order with the lowest cost, and ensure user service experience.

[0070] Figure 4 A flowchart of an example of a vehicle intelligent scheduling path planning operation according to an embodiment of the present application is shown.

[0071] like Figure 4 As shown, in step S410, the vehicle position of the target vehicle to be dispatched, the user position set corresponding to the multiple platform users to be transported, the real-time traffic data and the real-time weather data are obtained.

[0072] In step S420, it is detected whether the real-time weather data contains extreme weather information.

[0073] In step S431, when it is determined that the real-time weather data contains extreme weather information, at least one first candidate path is determined based on a first path planning model, and the first candidate path passes through the vehicle location and all user locations.

[0074] In step S433, when it is determined that the real-time weather data does not contain extreme weather information, at least one second candidate path is determined based on a second path planning model, and the second candidate path passes through the vehicle position and any user position.

[0075] Here, the second path planning model can adopt the online car-hailing pick-up model to select the best pick-up user location from multiple user locations and plan the corresponding pick-up route. More details can be partially referred to the description in the current relevant technology, which will not be repeated here.

[0076] In step S440, a passenger transport route is determined from each second candidate route to pick up and transport the platform user corresponding to the passenger transport route.

[0077] Through the embodiments of the present application, the central server monitors the current weather conditions in real time, and uses passenger pick-up route planning when the current weather conditions are not extreme, and uses rescue pick-up route planning when the current weather conditions are extreme, providing a new type of vehicle dispatching function to improve the riding service experience of car owners and passengers.

[0078] In some examples of the embodiments of the present application, the deep reinforcement learning model adopts a deep Q network. Figure 5 A schematic structural diagram of an example of a deep Q network according to an embodiment of the present application is shown.

[0079] like Figure 5 As shown, the deep Q network 500 includes a main network 510 and a target network 520. The main network 510 is responsible for generating the Q value of the current action, i.e., Q(s, a; θ), wherein the parameters θ of the main network are continuously updated during the training process. The target network 520 is used to estimate the Q value of the future state, i.e., Q(s′, a′; θ - ), where the parameters of the target network are - It is copied from the main network periodically, but remains unchanged between replications.

[0080] It should be noted that in the deep Q network (DQN), the target network mainly plays a role in the training phase, and is not directly used in the actual prediction or decision phase. During the training process of DQN, the target network provides a stable target value for calculating the loss function and updating the parameters of the main network.

[0081] More specifically, the main network 510 includes a cascaded input layer 511, a hidden layer 513, and an output layer 515. Here, the input layer 511 is used to normalize the data, and the input data includes real-time traffic data, real-time weather data, and the first candidate path. The hidden layer 513 includes a multi-layer convolutional neural network for extracting feature information from the input data. The output layer 515 is used to determine the expected Q value corresponding to each action in the state space, and to correct the first candidate path according to the path segment corresponding to the action with the maximum expected Q value to determine the corresponding corrected path.

[0082] Thus, by using the dual network architecture of DQN, possible feedback loops that may occur during self-updates are avoided, thereby improving the stability of the learning process. When estimating the TD (temporal difference) target, using the same network to evaluate the Q value of the current and future states may lead to large deviations.

[0083] In some examples of the embodiments of the present application, the update formula of the deep Q network is as follows:

[0084] Q(s,a;θ)=Q(s,a;θ)+α[r+γmax a′ Q(s′,a′;θ′)-Q(s,a;θ)] Formula (1)

[0085] r=w time × time +w safety × safety Formula (2)

[0086] In the formula, Q(s,a;θ) represents the current Q value of taking action a in state s, θ represents the current network parameters, α represents the learning rate; γ represents the discount factor, which indicates the importance of future rewards; max a′ Q(s′,a′;θ′) represents the maximum Q value that can be generated by all possible actions a′ in the next state s′, θ′ represents the old network parameters; r represents the immediate reward, which is obtained according to the result of the current action; r time and r safety Represent the timeliness reward item and the safety reward item, and w time and w safety Respectively represent the weight coefficients of the corresponding reward items, and w safety >w time , where w time Encourage the model to choose path segments with shorter distances, and w safety This enables the model to select route segments with a lower probability of abnormal road conditions under real-time weather data.

[0087] For example, suppose there are three second user locations that can form path segments for the first user location. Through real-time traffic, it is found that there is a traffic accident in one of the path segments, and another path segment is blocked due to heavy rain. Then, the DQN model will evaluate the Q value of each path, which is not only based on the length and estimated time of the path, but also considers the impact of real-time data (such as road closures caused by heavy rain) to decide to choose a bypass but safer and more reliable path, even if it is not the shortest. In this way, effective decision-making on intelligent vehicle scheduling under extreme weather conditions is achieved, and the adaptability and robustness of path planning are improved.

[0088] In some examples of the embodiments of the present application, the loss function of the deep Q network is:

[0089] L(θ)=E[(y i -Q(s,a;θ) 2 Formula (3)

[0090] Time difference target y i The formula is:

[0091] y i =λ(s,a)×r+γmax a′ Q(s′,a′;θ - ) Formula (4)

[0092] Where Q(s,a;θ) represents the Q-value estimate of the current state-action pair, provided by the main network; Q(s′,a′;θ - ) represents the Q value of the next state, estimated by the target network; θ and θ - Represent the parameters of the main network and the target network respectively; E[(y i -Q(s,a;θ) 2 ] represents the expected value of the square of the prediction error; λ(s,a) represents the environmental adaptability factor, which indicates the difficulty of taking a specific action a in a specific state s, and is used to adjust the weight of the immediate reward r.

[0093] It should be noted that DQN usually uses a variety of variants of temporal difference learning as the loss function. In the example of the embodiment of the present application, in order to adapt to the special needs under extreme weather conditions, an environmental adaptability factor λ is introduced in the TD target. This factor dynamically adjusts the weight of the immediate reward according to the complexity and difficulty of the current environment (such as weather conditions, traffic conditions) to complete the optimization adjustment of the loss function. More specifically, as shown in the above formula (4), λ(s, a) represents a function based on the current state and action, which is used to evaluate the risk and difficulty of the current choice. For example, path selection under extreme weather conditions or in high traffic congestion areas may be weighted higher because they have higher complexity.

[0094] In more detail, an example of a function formula for calculating λ(s,a) is as follows:

[0095] λ(s,a)=λ W ×λ T ×λ R Formula (5)

[0096] In the formula, λ W , T and λ R They represent the weather condition influence coefficient, traffic flow influence coefficient and road condition influence coefficient respectively. W The value range of λ is [1,2], where 1 indicates sunny weather and 2 indicates extremely bad weather. T The value range of λ is [1,1.5], where 1 indicates smooth traffic and 1.5 indicates high congestion. R The value range is [1,1.5], where 1 means good and 1.5 means bad (such as ice or water accumulation).

[0097] The above formula (5) can reflect the comprehensive impact of different environmental factors on the immediate reward under a given state s and action a. For example, assuming that in a certain state, the weather is moderate rain (λ W =1.5), traffic is slightly congested (λ T =1.2), the road is slippery but drivable (λ R =1.1). Thus, λ(s,a)=1.5×1.2×1.1=1.98, which means that the immediate reward at this state and action will be increased by 98% to reflect the additional difficulty due to adverse weather and road conditions.

[0098] Through the embodiments of the present application, the above loss function is used to train DQN, and the optimization goal is to minimize the loss function so that the predicted Q value of the network is as close as possible to the TD target weighted by environmental adaptability, aiming to make the network more focused on effective decision-making under extreme weather conditions and improve the system's adaptability and robustness in complex environments.

[0099] Figure 6 A schematic block diagram of the structure of an example of a first path planning model according to an embodiment of the present application is shown.

[0100] like Figure 6 As shown, the first path planning model 600 includes a Dijkstra algorithm module 610 , an A* algorithm module 620 and a Bellman-Ford algorithm module 630 .

[0101] It should be noted that the Dijkstra algorithm is a path search method based on a greedy algorithm, which is used to find the shortest path from one node to all other nodes in a weighted graph. It repeatedly selects the unvisited node closest to the starting point and updates the distances of its adjacent nodes until all nodes are visited. In combination with this technical scenario, the Dijkstra algorithm is suitable for finding the shortest path, especially when there are no negative weight edges in the road network. It can provide an optimal path from the starting point to the end point, and is more efficient when there are many path options.

[0102] The Bellman-Ford algorithm calculates the shortest path from a single source point to all other nodes by performing multiple relaxation operations on all edges in the graph, and the algorithm is still effective even in the presence of negative weight edges. In conjunction with this technical scenario, under extreme weather conditions, some roads may cause sudden increases in tolls or time costs due to emergencies (which can be regarded as negative weights). At this time, the Bellman-Ford algorithm can handle such negative weight edges and find a reasonable path.

[0103] The A* algorithm is a heuristic search algorithm that finds the most efficient path by comprehensively considering the known path cost (such as the actual distance to the current node) and an estimated remaining cost (such as the estimated distance from the current node to the destination). In combination with this technical scenario, under extreme weather conditions, the A* algorithm can effectively balance the actual state of the path and the predicted future conditions to find an efficient driving route.

[0104] The Dijkstra algorithm module 610, the A* algorithm module 620 and the Bellman-Ford algorithm module 630 are connected in parallel to the deep reinforcement learning model, so that the deep reinforcement learning model can analyze and correct the path segments in the candidate path output by the path planning algorithm module to ensure that the corrected path can at least meet the safety and economy in extreme weather conditions.

[0105] Furthermore, the MCDA evaluation model is used to comprehensively score and sort the various correction paths output by the deep reinforcement learning model to select the final rescue pick-up path.

[0106] More specifically, the MCDA evaluation model calculates the path comprehensive score corresponding to the correction path in the following way:

[0107]

[0108] In the formula, S route1 represents the comprehensive score of the modified path route1, w i represents the weight of the i-th evaluation criterion, C i,route1represents the standardized score of the modified route route1 on the i-th evaluation criterion; the evaluation criteria include: driving time evaluation criterion, path economic cost evaluation criterion and road safety evaluation criterion.

[0109] The calculation formula for the standardized score corresponding to the driving time evaluation standard is:

[0110]

[0111] α=1+k×WL Formula (8)

[0112] Where, T route1 represents the expected travel time of the modified route route1, T min and T max They respectively represent the minimum and maximum expected travel times of all corrected paths; α represents the adjustment coefficient, k represents a normal number, which is used to adjust the influence of the severe weather level on the adjustment coefficient; WL represents the severe weather level, which is determined by querying the real-time weather data and the severe weather level mapping table. The severe weather level mapping table records multiple severe weather levels and corresponding weather data intervals.

[0113] For the above adjustment coefficient α, for example, assuming k = 0.1, this means that for each increase in the level of bad weather, the adjustment coefficient increases by 0.1. If the weather condition is heavy rain (level 3), then α = 1 + 0.1 × 3 = 1.3. In this way, in heavy rain, the original driving time score will increase by 30% to reflect the additional difficulty caused by bad weather.

[0114] In the embodiment of the present application, when calculating the standardized score of the driving time evaluation standard, the adjustment coefficient α, which is in a positive linear relationship with the weather severity level, is comprehensively considered, effectively taking the severity of weather conditions into consideration, so that the driving time scoring is more fair and in line with actual conditions.

[0115] The calculation formula for the standardized score corresponding to the path economic cost evaluation standard is:

[0116]

[0117] In the formula, E routel represents the expected economic cost of the modified path route1, E max It represents the maximum of the expected economic costs of all modified paths; the expected economic cost of the path includes the energy consumption cost and the toll fees of the toll sections involved in the path.

[0118] In this way, the economic score can help the system compare different routes from a cost-effectiveness perspective and select cost-effective paths in extreme weather conditions, helping to control operating costs and improve service efficiency.

[0119] The calculation formula for the standardized score corresponding to the road safety assessment standard is:

[0120]

[0121] In the formula, m j is the weight of the jth safety factor, C j,route1,sf represents the score of the modified route route1 for the jth safety factor, where the safety factor includes at least one of the following: slippery road condition, historical accident record of the route, curve information of the route and slope information of the route.

[0122] In the embodiment of the present application, each path is scored according to its safety factor, and the scores of different safety factors are combined to obtain the overall safety score of each path. Thus, the system can comprehensively evaluate the safety of each path and ensure that the route selected in extreme weather conditions is as safe as possible.

[0123] It should be noted that in some cases, there is overlap between the evaluation criteria of the MCDA evaluation model and the reward items of the DQN reward function, for example, both involve considerations of safety and time efficiency. However, on the one hand, in terms of method, DQN focuses on learning and optimizing dynamic decision-making processes, while MCDA focuses more on comprehensive evaluation and decision-making, which can achieve effective complementarity. On the other hand, DQN is used to dynamically adjust the path segmentation of candidate paths to obtain a corrected path from the candidate path, while MCDA further evaluates and verifies each corrected path based on DQN to ensure that the final rescue pick-up path can fully and adequately meet various evaluation criteria.

[0124] The intelligent vehicle dispatching system for coping with extreme weather provided by the present application is described below. The intelligent vehicle dispatching system for coping with extreme weather described below and the intelligent vehicle dispatching method for coping with extreme weather described above can be referenced to each other.

[0125] Figure 7 A structural block diagram of an example of a vehicle intelligent dispatching system for coping with extreme weather according to an embodiment of the present application is shown.

[0126] like Figure 7 As shown, the vehicle intelligent dispatching system 700 for coping with extreme weather includes a data acquisition unit 710, a path planning unit 720, a path correction unit 730 and a pick-up path determination unit 740.

[0127] The data acquisition unit 710 is configured to acquire the vehicle position of the target vehicle to be dispatched, the user location set corresponding to multiple platform users to be transported, real-time traffic data and real-time weather data; the distance between each user location in the user location set and the vehicle position is less than a preset distance threshold; the real-time traffic data includes: road closure information, traffic accident information and road congestion information.

[0128] The path planning unit 720 is configured to determine at least one first candidate path based on a first path planning model when it is determined that the real-time weather data contains extreme weather information, wherein the first candidate path passes through the vehicle location and all the user locations; the extreme weather information includes any one of the following: heavy rain, heavy snow, freezing rain or typhoon.

[0129] The path correction unit 730 is configured to adjust at least one path segment of the first candidate path for each of the first candidate paths based on the deep reinforcement learning model to obtain a corresponding corrected path; the path segment is determined based on any two user locations in the user location set; the state of the deep reinforcement learning model is defined based on the node position, real-time traffic data and real-time weather data, and the action of the deep reinforcement learning model is defined based on the switching of the path segments; the node position is the vehicle position or the user position.

[0130] The pick-up path determination unit 740 is configured to calculate the path comprehensive scores corresponding to each of the corrected paths based on the MCDA evaluation model, and determine the rescue pick-up path of the target vehicle from each of the corrected paths according to the path comprehensive scores, so as to sequentially pick up each of the platform users.

[0131] In some embodiments, an embodiment of the present application provides a non-volatile computer-readable storage medium, which stores one or more programs including execution instructions, and the execution instructions can be read and executed by electronic devices (including but not limited to computers, servers, or network devices, etc.) to execute the above-mentioned vehicle intelligent scheduling method for responding to extreme weather in the present application.

[0132] In some embodiments, the embodiments of the present application also provide a computer program product, which includes a computer program stored on a non-volatile computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer executes the above-mentioned vehicle intelligent scheduling method for dealing with extreme weather.

[0133] In some embodiments, the embodiments of the present application also provide an electronic device, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute a vehicle intelligent scheduling method for dealing with extreme weather.

[0134] Figure 8 is a schematic diagram of the hardware structure of an electronic device for executing a vehicle intelligent dispatching method for coping with extreme weather provided by another embodiment of the present application, such as Figure 8 As shown, the device includes:

[0135] One or more processors 810 and memory 820, Figure 8 A processor 810 is taken as an example.

[0136] The device for executing the vehicle intelligent dispatching method for coping with extreme weather may further include: an input device 830 and an output device 840 .

[0137] The processor 810, the memory 820, the input device 830 and the output device 840 may be connected via a bus or other means. Figure 8 The example of connecting through bus is taken in the following.

[0138] The memory 820, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer executable programs and modules, such as program instructions / modules corresponding to the vehicle intelligent dispatching method for dealing with extreme weather in the embodiment of the present application. The processor 810 executes various functional applications and data processing of the server by running the non-volatile software programs, instructions and modules stored in the memory 820, that is, the vehicle intelligent dispatching method for dealing with extreme weather in the above method embodiment is implemented.

[0139] The memory 820 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the electronic device, etc. In addition, the memory 820 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 820 may optionally include a memory remotely arranged relative to the processor 810, and these remote memories may be connected to the electronic device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0140] The input device 830 may receive input digital or character information and generate signals related to user settings and function control of the electronic device. The output device 840 may include a display device such as a display screen.

[0141] The one or more modules are stored in the memory 820, and when executed by the one or more processors 810, the vehicle intelligent scheduling method for dealing with extreme weather in any of the above method embodiments is executed.

[0142] The above-mentioned product can execute the vehicle intelligent dispatching method for dealing with extreme weather provided in the embodiment of the present application, and has the corresponding functional modules and beneficial effects of the execution method. For technical details not fully described in this embodiment, please refer to the method provided in the embodiment of the present application.

[0143] The electronic devices of the embodiments of the present application exist in various forms, including but not limited to:

[0144] (1) Mobile communication equipment: This type of equipment is characterized by having mobile communication functions and its main purpose is to provide voice and data communications. This type of terminal includes: smart phones, multimedia phones, functional phones, and low-end phones.

[0145] (2) Ultra-mobile personal computer devices: These devices belong to the category of personal computers, have computing and processing functions, and generally also have mobile Internet access features. These terminals include: PDA, MID and UMPC devices, etc.

[0146] (3) Portable entertainment devices: These devices can display and play multimedia content. They include audio and video players, handheld game consoles, e-books, smart toys, and portable car navigation devices.

[0147] (4) Other onboard electronic devices with data interaction functions, such as on-board devices installed in vehicles.

[0148] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0149] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a general hardware platform, and of course, by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0150] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A vehicle intelligent dispatching method for dealing with extreme weather, comprising: Obtaining the vehicle position of the target vehicle to be dispatched, the user position set corresponding to the multiple platform users to be transported, real-time traffic data and real-time weather data; The distance between each user location in the user location set and the vehicle location is less than a preset distance threshold; the real-time traffic data includes: road closure information, traffic accident information and road congestion information; In the case where it is determined that the real-time weather data includes extreme weather information, determining at least one first candidate path based on a first path planning model, the first candidate path passing through the vehicle location and all the user locations; the extreme weather information includes any one of the following: heavy rain, heavy snow, freezing rain or typhoon; For each of the first candidate paths, adjusting at least one path segment of the first candidate path based on the deep reinforcement learning model to obtain a corresponding corrected path; The path segment is determined according to any two user locations in the user location set; the state of the deep reinforcement learning model is defined according to the node location, real-time traffic data and real-time weather data, and the action of the deep reinforcement learning model is defined according to the switching of the path segment; the node location is the vehicle location or the user location; The comprehensive scores of the paths corresponding to the respective correction paths are calculated based on the MCDA evaluation model, and the rescue pick-up paths of the target vehicles are determined from the respective correction paths according to the comprehensive scores of the paths, so as to sequentially pick up the respective platform users.

2. The method according to claim 1, wherein: After obtaining the vehicle position of the target vehicle to be dispatched, the user position set of multiple platform users to be transported, the real-time traffic data and the real-time weather data, the method further includes: In a case where it is determined that the real-time weather data does not contain extreme weather information, determining at least one second candidate path based on a second path planning model, where the second candidate path passes through the vehicle location and any one of the user locations; A passenger transport route is determined from each of the second candidate routes to pick up and transport the platform user corresponding to the passenger transport route.

3. The method according to claim 1, wherein: The deep reinforcement learning model adopts a deep Q network, and the main network of the deep Q network includes a cascaded input layer, a hidden layer and an output layer; The input layer is used to normalize the data; the input data includes the real-time traffic data, the real-time weather data and the first candidate path; The hidden layer includes a multi-layer convolutional neural network, which is used to extract feature information from the input data; The output layer is used to determine the expected Q value corresponding to each action in the state space, and to correct the first candidate path according to the path segment corresponding to the action with the maximum expected Q value to determine the corresponding corrected path.

4. The method according to claim 3, wherein: The update formula of the deep Q network is as follows: Q(s,a;θ)=Q(s,a;θ)+α[r+γmax a′ Q(s′,a′;θ′)-Q(s,a;θ)] r=w time ×r time +w safety ×r safety In the formula, Q(s,a;θ) represents the current Q value of taking action a in state s, θ represents the current network parameters, α represents the learning rate; γ represents the discount factor, which indicates the importance of future rewards; max a′ Q(s′,a′;θ′) represents the maximum Q value that can be generated by all possible actions a′ in the next state s′, θ′ represents the old network parameters; r represents the immediate reward, which is obtained according to the result of the current action; r time and r safety Represent the timeliness reward item and the safety reward item, and w time and w safety Respectively represent the weight coefficients of the corresponding reward items, and w safety >w time , where w time Encourage the model to choose path segments with shorter distances, and w safety This enables the model to select route segments with a lower probability of abnormal road conditions under real-time weather data.

5. The method according to claim 4, wherein: The loss function of the deep Q network is: L(θ)=E[(y i -Q(s,a;θ)) 2 ] Time difference target y i The formula is: y i =λ(s,a)×r+γmax a′ Q(s′,a′;θ - ) Where Q(s,a;θ) represents the Q-value estimate of the current state-action pair, provided by the main network; Q(s′,a′;θ - ) represents the Q value of the next state, estimated by the target network; θ and θ - Represent the parameters of the main network and the target network respectively; E[(y i -Q(s,a;θ) 2 ] represents the expected value of the square of the prediction error; λ(s,a) represents the environmental adaptability factor, which indicates the difficulty of taking a specific action a in a specific state s, and is used to adjust the weight of the immediate reward r.

6. The method according to any one of claims 1 to 5, wherein: The first path planning model includes a Dijkstra algorithm module, an A* algorithm module and a Bellman-Ford algorithm module, wherein the Dijkstra algorithm module, the A* algorithm module and the Bellman-Ford algorithm module are connected in parallel to the deep reinforcement learning model.

7. The method according to claim 6, wherein: The MCDA evaluation model calculates the path comprehensive score corresponding to the modified path in the following way: In the formula, S route1 represents the comprehensive score of the modified path route1, w i represents the weight of the i-th evaluation criterion, C i,route1 represents the standardized score of the modified route route1 on the i-th evaluation criteria; the evaluation criteria include: driving time evaluation criteria, path economic cost evaluation criteria and road safety evaluation criteria; The calculation formula for the standardized score corresponding to the driving time evaluation standard is: α=1+k×WL Where, T route1 represents the expected travel time of the modified route route1, T min and T max They respectively represent the minimum and maximum of the expected travel times of all the modified routes; α represents the adjustment coefficient, k represents a normal constant, and is used to adjust the degree of influence of the severe weather level on the adjustment coefficient; WL represents the severe weather level, which is determined by querying the real-time weather data and the severe weather level mapping table, and the severe weather level mapping table records multiple severe weather levels and corresponding weather data intervals; The calculation formula for the standardized score corresponding to the path economic cost evaluation standard is: In the formula, E routel represents the expected economic cost of the modified path route1, E max It represents the maximum of the expected economic costs of all modified paths; the expected economic cost of the path includes the energy consumption cost and the toll fees of the toll road sections involved in the path; The calculation formula for the standardized score corresponding to the road safety assessment standard is: In the formula, m j is the weight of the jth safety factor, C j,route1,sf represents the score of the modified route route1 for the jth safety factor; the safety factor includes at least one of the following: slippery road condition, historical accident record of the route, route curve information and route slope information.

8. A vehicle intelligent dispatching system for dealing with extreme weather, comprising: A data acquisition unit is configured to acquire a vehicle position of a target vehicle to be dispatched, a user position set corresponding to a plurality of platform users to be transported, real-time traffic data, and real-time weather data; the distance between each user position in the user position set and the vehicle position is less than a preset distance threshold; the real-time traffic data includes: road closure information, traffic accident information, and road congestion information; A path planning unit is configured to determine at least one first candidate path based on a first path planning model when it is determined that the real-time weather data includes extreme weather information, wherein the first candidate path passes through the vehicle location and all the user locations; the extreme weather information includes any one of the following: heavy rain, heavy snow, freezing rain or typhoon; a path correction unit, configured to adjust, for each of the first candidate paths, at least one path segment of the first candidate path based on a deep reinforcement learning model to obtain a corresponding corrected path; The path segment is determined according to any two user locations in the user location set; the state of the deep reinforcement learning model is defined according to the node location, real-time traffic data and real-time weather data, and the action of the deep reinforcement learning model is defined according to the switching of the path segment; the node location is the vehicle location or the user location; The pick-up path determination unit is configured to calculate the path comprehensive scores corresponding to each of the corrected paths based on the MCDA evaluation model, and determine the rescue pick-up path of the target vehicle from each of the corrected paths according to the path comprehensive scores, so as to sequentially pick up each of the platform users.

Images