Congestion prediction and path optimization method, device and equipment based on vehicle-road cooperation

Abstract

The invention discloses a congestion prediction and path optimization method, device and equipment based on vehicle-road cooperation, and the method comprises the steps: obtaining vehicle speed time sequence data, vehicle flow time sequence data, traffic event data and meteorological condition data of each road section in a target region, and determining the graph structure data of the target region; inputting the vehicle speed time sequence data, the vehicle flow time sequence data and the graph structure data into a road section congestion prediction model to obtain an average vehicle speed, an average acceleration and a congestion level of each road section in a preset future time period; determining a plurality of initial driving paths, and screening out a plurality of candidate driving paths from the initial driving paths according to the congestion level; and according to the time cost function and the fuel consumption cost function, based on a non-dominated sorting genetic algorithm, carrying out optimization solution on the candidate driving path to obtain an optimal driving path. According to the invention, the accuracy and real-time performance of road congestion prediction and vehicle path optimization are improved, so that the travel experience of the user is improved, and the method can be applied to the technical field of intelligent driving.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of intelligent driving, and in particular to a congestion prediction and path optimization method, device and equipment based on vehicle-road cooperation. BACKGROUND

[0002] Current vehicle navigation and traffic congestion prediction technologies mainly fall into two categories: one is a congestion prediction scheme based on a single data source, such as navigation software from Gaode and Baidu, which mainly relies on massive GPS trajectory data uploaded by user terminals to calculate the average speed of a road section through statistical methods, and then to determine the congestion state and plan a path; the other is a preliminary application scheme of vehicle-road cooperation, such as “Congestion Identification and Prediction Method Based on Vehicle-road Cooperation Fusion Perception and Traffic Simulation” (CN117523834A), which identifies events such as construction and accidents by fusing target detection data from vehicle-side and roadside sensors, and predicts changes in traffic capacity in combination with a traffic simulation model.

[0003] In terms of path optimization, existing technologies mostly use Dijkstra algorithm or A* algorithm to plan a route with “shortest distance” or “shortest time” as a single objective function, and some new energy vehicle navigation systems will additionally consider the distribution of charging piles, but have not formed a multi-dimensional cost collaborative optimization mechanism. In terms of communication, GB / T 45315—2025 “Technical Requirements and Test Methods for Vehicle Information Interaction System Based on LTE-V2X Direct Communication” has been issued and implemented, providing standard support for information interaction between vehicles and roadside facilities, but the existing technology has low utilization rate of V2X data fusion, and has not fully realized the collaborative value of multi-source data.

[0004] In summary, the existing congestion prediction and path optimization schemes have the following shortcomings: 1) Single prediction dimension: Most schemes only rely on traffic flow or GPS data, without fusing key influencing factors such as weather (e.g. rain causing a decrease in vehicle speed) and traffic accidents (e.g. scratches causing an increase in lane occupancy rate), resulting in low accuracy of congestion prediction in extreme weather conditions; 2) Weak model generalization ability: traditional statistical models or single neural networks are difficult to capture the spatio-temporal correlation of traffic flow, and have insufficient adaptability to dynamic scenarios such as “tidal traffic flow” and “sudden accidents”, resulting in high real-time congestion prediction error in a short period of time; 3) One-sided path optimization objective: a route with the shortest time may result in a significant increase in fuel consumption due to frequent acceleration and deceleration, while only considering fuel consumption may sacrifice traffic efficiency, failing to balance the dual core needs of users; 4) Vehicle-road cooperation data value not released: existing schemes only make simple screening of roadside radar data and traffic event information transmitted by V2X, without deep fusion with historical data and environmental data, resulting in low data utilization rate.

[0005] The above-mentioned shortcomings result in that the driver frequently encounters problems such as "navigation display smooth actual congestion" and "route time consumption is short but fuel consumption is high", which not only affects the user's travel experience, but also increases the road traffic pressure and energy waste.

[0006] The above problems need to be solved urgently. SUMMARY

[0007] The present application aims to at least partially solve one of the technical problems existing in the prior art.

[0008] To this end, one object of the embodiments of the present application is to provide a congestion prediction and path optimization method based on vehicle-road cooperation, which improves the accuracy and real-time performance of road congestion prediction and vehicle path optimization, thereby improving the user's travel experience.

[0009] Another object of the embodiments of the present application is to provide a congestion prediction and path optimization device based on vehicle-road cooperation.

[0010] In order to achieve the above technical purpose, the technical solutions adopted by the embodiments of the present application include: On the one hand, the embodiments of the present application provide a congestion prediction and path optimization method based on vehicle-road cooperation, comprising the following steps: Obtaining the vehicle speed time series data, the vehicle flow time series data, the traffic event data and the meteorological condition data of each road section in a target area, and determining the graph structure data of the target area according to the traffic event data, the meteorological condition data and the road topology of the target area; Inputting the vehicle speed time series data, the vehicle flow time series data and the graph structure data into a pre-trained road section congestion prediction model to obtain the average vehicle speed, the average acceleration and the congestion level of each road section in a preset future period; Determining a plurality of initial driving paths according to the starting position and the destination position of a target vehicle, and screening a plurality of candidate driving paths from the initial driving paths according to the congestion level; Constructing a time cost function and a fuel consumption cost function, and optimizing and solving the candidate driving paths based on a non-dominated sorting genetic algorithm according to the time cost function, the fuel consumption cost function, the average vehicle speed and the average acceleration to obtain an optimal driving path.

[0011] Further, in one embodiment of the present application, the obtaining the vehicle speed time series data, the vehicle flow time series data, the traffic event data and the meteorological condition data of each road section in a target area, and determining the graph structure data of the target area according to the traffic event data, the meteorological condition data and the road topology of the target area specifically comprises: Obtaining the vehicle speed time series data of each vehicle in each road section through V2X; The traffic flow time series data of each road section is obtained by a roadside radar device, and the traffic event data of each road section is obtained by a roadside camera device; The meteorological condition data of each road section is obtained by a meteorological platform; According to the road topology, the connection relationship between each road section is determined, the road sections are taken as nodes, the connection relationship between the road sections is taken as edges, the node features are determined according to the traffic event data and the meteorological condition data, and the graph structure data is constructed.

[0012] Further, in an embodiment of the present application, the road congestion prediction model is trained by the following steps: The first vehicle speed time series sample and the first traffic flow time series sample of each sample road section in a sample area in a first historical period, the traffic event sample and the meteorological condition sample at a first historical time, and the second vehicle speed time series sample and the second traffic flow time series sample in a second historical period are obtained; The training sample is constructed according to the first vehicle speed time series sample, the first traffic flow time series sample, the traffic event sample and the meteorological condition sample, and the corresponding average vehicle speed label, average acceleration label and congestion level label are determined according to the second vehicle speed time series sample and the second traffic flow time series sample; The training sample is input into a pre-constructed LSTM-GCN hybrid neural network to obtain the predicted average vehicle speed, the predicted average acceleration and the predicted congestion level of each sample road section; The loss value is determined according to the predicted average vehicle speed, the predicted average acceleration, the predicted congestion level, the average vehicle speed label, the average acceleration label and the congestion level label; The parameters of the LSTM-GCN hybrid neural network are updated by a back propagation algorithm according to the loss value to obtain the trained road congestion prediction model; The LSTM-GCN hybrid neural network includes an LSTM branch network, a GCN branch network, a feature fusion layer and a full connection layer, the first historical time is the end time of the first historical period and the start time of the second historical period.

[0013] Further, in an embodiment of the present application, the training sample is input into a pre-constructed LSTM-GCN hybrid neural network to obtain the predicted average vehicle speed, the predicted average acceleration and the predicted congestion level of each sample road section, which specifically includes: The first vehicle speed time series sample and the first traffic flow time series sample are input into the LSTM branch network to obtain the time series feature vector of each sample road section; The sample road sections are taken as nodes, the connection relationship between the sample road sections is taken as edges, and a graph structure sample of the sample area is constructed according to the node features determined according to the traffic event sample and the meteorological condition sample; The graph structure sample is input into the GCN branch network to obtain a spatial correlation feature vector of the sample area; The time sequence feature vector and the spatial correlation feature vector are fused based on an attention mechanism through the feature fusion layer to obtain a spatiotemporal fusion feature; The spatiotemporal fusion feature is mapped into the predicted average speed, the predicted average acceleration and the predicted congestion level of each sample road section through the full connection layer.

[0014] Further, in an embodiment of the present application, the multiple initial driving paths are determined according to the starting position and the destination position of the target vehicle, and the multiple candidate driving paths are screened from the initial driving paths according to the congestion level, which specifically includes: The initial driving paths are obtained by path planning according to the starting position, the destination position and the road topology, and the shortest path distance of the shortest driving path is determined; The congestion road sections are determined according to the congestion level, it is judged whether the initial driving paths contain the congestion road sections, and it is judged whether the ratio of the path distance of the initial driving path to the shortest path distance is greater than a preset first threshold value; When the initial driving path does not contain the congestion road section, and the ratio of the path distance of the initial driving path to the shortest path distance is less than or equal to the first threshold value, the initial driving path is determined as the candidate driving path.

[0015] Further, in an embodiment of the present application, the time cost function is:

[0016] wherein, represents the time cost of the candidate driving path, represents the number of road sections of the candidate driving path, represents the length of the i th road section of the candidate driving path, represents the average speed of the i th road section of the candidate driving path, represents the intersection waiting time of the i th road section of the candidate driving path; The fuel consumption cost function is:

[0017] wherein, represents the fuel consumption cost of the candidate driving path, an average acceleration of an i-th road segment representing a candidate driving path, 、 and a calibrated fuel consumption parameter corresponding to the target vehicle.

[0018] Further, in an embodiment of the present application, the optimization and solution of the candidate driving path based on the time cost function, the fuel consumption cost function, the average vehicle speed and the average acceleration are obtained by a non-dominated sorting genetic algorithm, and the specific steps include: determining the road segment number of each road segment, encoding the candidate driving path according to the road segment number, obtaining a plurality of individuals, and initializing a population according to the individuals; calculating the path time cost and the path fuel consumption cost of each individual according to the time cost function and the fuel consumption cost function, non-dominantly sorting the individuals according to the path time cost and the path fuel consumption cost, and obtaining the Pareto level of each individual; determining the time congestion degree of the individuals in the same Pareto level in the time cost space and the fuel consumption congestion degree in the fuel consumption cost space; determining the time cost weight and the fuel consumption cost weight according to the driver preference coefficient of the target vehicle, and determining the weighted congestion degree of each individual according to the time cost weight, the fuel consumption cost weight, the time congestion degree and the fuel consumption congestion degree; screening the parent individuals by the tournament selection method according to the Pareto level and the weighted congestion degree; performing cross and mutation operations on the parent individuals to generate offspring individuals; combining the parent individuals and the offspring individuals, and screening the next generation population by non-dominant sorting; when the preset convergence condition is reached, determining the optimal driving path according to the Pareto optimal solution of the current population.

[0019] On the other hand, an embodiment of the present application provides a congestion prediction and path optimization device based on vehicle-road cooperation, which comprises: a data acquisition module, configured to acquire vehicle speed time series data, vehicle flow time series data, traffic event data and meteorological condition data of each road segment in a target area, and determine graph structure data of the target area according to the traffic event data, the meteorological condition data and the road topology of the target area; a congestion prediction module, configured to input the vehicle speed time series data, the vehicle flow time series data and the graph structure data into a pre-trained road segment congestion prediction model to obtain average vehicle speed, average acceleration and congestion level of each road segment in a preset future period. a path screening module, configured to determine a plurality of initial driving paths according to a starting position and a destination position of the target vehicle, and screen a plurality of candidate driving paths from the initial driving paths according to the congestion level; a path optimization module, configured to construct a time cost function and an oil consumption cost function, and perform optimization and solution on the candidate driving paths based on a non-dominated sorting genetic algorithm according to the time cost function, the oil consumption cost function, the average vehicle speed and the average acceleration, to obtain an optimal driving path.

[0020] In another aspect, an embodiment of the present application provides an electronic device, comprising: at least one processor; at least one memory configured to store at least one program; when the at least one program is executed by the at least one processor, the at least one processor is caused to implement the above-mentioned method for congestion prediction and path optimization based on vehicle-infrastructure cooperation.

[0021] In another aspect, an embodiment of the present application further provides a computer readable storage medium, wherein a computer program executable by a processor is stored, and the computer program, when executed by the processor, implements the above-mentioned method for congestion prediction and path optimization based on vehicle-infrastructure cooperation.

[0022] In another aspect, an embodiment of the present application further provides a computer program product, comprising a computer program, and the computer program, when executed by a processor, implements the above-mentioned method for congestion prediction and path optimization based on vehicle-infrastructure cooperation.

[0023] The advantages and beneficial effects of the present application will be partially given in the following description, partially will become obvious from the following description, or will be understood by the practice of the present application: The embodiment of the present application obtains the vehicle speed time series data, the vehicle flow time series data, the traffic event data and the meteorological condition data of each road section in the target area, determines the graph structure data of the target area according to the traffic event data, the meteorological condition data and the road topology of the target area, inputs the vehicle speed time series data, the vehicle flow time series data and the graph structure data into the pre-trained road congestion prediction model, obtains the average vehicle speed, the average acceleration and the congestion level of each road section in the preset future period, determines a plurality of initial driving paths according to the starting position and the destination position of the target vehicle, screens a plurality of candidate driving paths from the initial driving paths according to the congestion level, constructs the time cost function and the fuel consumption cost function, and optimizes and solves the candidate driving paths based on the non-dominated sorting genetic algorithm according to the time cost function, the fuel consumption cost function, the average vehicle speed and the average acceleration, to obtain the optimal driving path. The embodiment of the present application predicts the average vehicle speed, the average acceleration and the congestion level of the road section based on the vehicle speed time series data, the vehicle flow time series data, the traffic event data and the meteorological condition data of the road section, optimizes the path based on the non-dominated sorting genetic algorithm according to the time cost and the fuel consumption cost, improves the accuracy and the real-time performance of the road congestion prediction and the vehicle path optimization, and thus improves the travel experience of the user. BRIEF DESCRIPTION OF DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following introduces the drawings needed to be used in the embodiments of the present application. It should be understood that the drawings introduced in the following are only for facilitating the clear description of the technical solutions in some embodiments of the present application, and other drawings can also be obtained according to these drawings without any creative labor for those skilled in the art.

[0025] Figure 1 The step flow chart of the congestion prediction and path optimization method based on vehicle-road cooperation provided by the embodiment of the present application is shown in the figure. Figure 2 The structural block diagram of the congestion prediction and path optimization device based on vehicle-road cooperation provided by the embodiment of the present application is shown in the figure. Figure 3 The structural block diagram of the electronic device provided by the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0026] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific examples described herein are only intended to explain the present application and not to limit the present application. When the following description refers to the accompanying drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following example embodiments do not represent all implementations consistent with embodiments of the present application. They are merely examples of apparatuses and methods consistent with some aspects of embodiments of the present application as detailed in the appended claims.

[0027] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of the present application only and is not intended to be limiting of the present application.

[0028] The congestion prediction and path optimization method based on vehicle-road cooperation provided by the embodiments of the present application can be applied to a terminal, can be applied to a server, and can also be software running in the terminal or the server. In some embodiments, the terminal can be a smart phone, a tablet computer, a notebook computer, a desktop computer, a smart speaker, a smart watch, a vehicle-mounted terminal, and the like, but is not limited thereto. The server end can be configured as a stand-alone physical server, can be configured as a server cluster or a distributed system composed of multiple physical servers, can be configured as a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDNs, and basic cloud computing services such as big data and artificial intelligence platforms, and the server can also be a node server in a blockchain network. The software can be an application that implements the congestion prediction and path optimization method based on vehicle-road cooperation, but is not limited to the above forms.

[0029] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld devices or portable devices, tablet devices, multi-processor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, and the like. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, and the like that perform specific tasks or implement specific abstract data types. The present application can also be practiced in a distributed computing environment, in which tasks are performed by remote processing devices connected by a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media, including storage devices.

[0030] It should be noted that in various specific embodiments of the present application, when relevant processing is required according to user information, user behavior data, user history data, and user location information, etc. related to the identity or characteristics of the user, the user's permission or consent will be obtained first, and the collection, use and processing of these data will comply with relevant laws, regulations and standards. In addition, when the embodiments of the present application need to obtain sensitive personal information of the user, the separate permission or separate consent of the user will be obtained through a pop-up window or by jumping to a confirmation page, and after obtaining the separate permission or separate consent of the user, the necessary user-related data for enabling the embodiments of the present application to operate normally will be obtained.

[0031] With reference to Figure 1 The embodiments of the present application provide a congestion prediction and path optimization method based on vehicle-road cooperation, which specifically comprises the following steps: S101, obtaining vehicle speed time series data, vehicle flow time series data, traffic event data and meteorological condition data of each road section in a target area, and determining graph structure data of the target area according to the traffic event data, meteorological condition data and road topology of the target area; S102, inputting the vehicle speed time series data, vehicle flow time series data and graph structure data into a pre-trained road section congestion prediction model to obtain average vehicle speed, average acceleration and congestion level of each road section in a preset future period; S103, determining a plurality of initial driving paths according to the starting position and destination position of the target vehicle, and screening a plurality of candidate driving paths from the initial driving paths according to the congestion level; S104, constructing a time cost function and a fuel consumption cost function, and optimizing and solving the candidate driving paths based on the non-dominated sorting genetic algorithm according to the time cost function, the fuel consumption cost function, the average vehicle speed and the average acceleration to obtain an optimal driving path.

[0032] The embodiments of the present application predict the average vehicle speed, average acceleration and congestion level of the road section based on the vehicle speed time series data, vehicle flow time series data, traffic event data and meteorological condition data of the road section, and optimize the path based on the non-dominated sorting genetic algorithm according to the time cost and fuel consumption cost, thereby improving the accuracy and real-time performance of road congestion prediction and vehicle path optimization, and improving the travel experience of the user.

[0033] Further as an optional embodiment, the vehicle speed time series data, vehicle flow time series data, traffic event data and meteorological condition data of each road section in a target area are obtained, and the graph structure data of the target area is determined according to the traffic event data, meteorological condition data and road topology of the target area, which specifically comprises: S1011, acquire vehicle speed time series data of each vehicle in each road section through V2X; S1012, acquire vehicle flow time series data of each road section through a roadside radar device, and acquire traffic event data of each road section through a roadside camera device; S1013, acquire meteorological condition data of each road section through a meteorological platform; S1014, determine the connection relationship between each road section according to the road topology, take the road section as a node, take the connection relationship between the road sections as an edge, determine the node features according to the traffic event data and the meteorological condition data, and construct to obtain graph structure data.

[0034] Specifically, this step carries out multi-source data collection and standardization processing (data layer), and collects data through three types of channels and carries out unified format conversion (sampling frequency 10Hz): 1) vehicle end data: vehicle speed, position data obtained by vehicle-mounted GPS, and fuel consumption related data such as accelerator pedal opening degree and brake frequency collected by CAN bus; 2) roadside data: vehicle flow data received by roadside millimeter wave radar, and traffic event (accident, construction) identification results of high-definition camera based on LTE-V2X communication; 3) environmental and historical data: real-time precipitation, visibility and other meteorological condition data obtained through a network interface from a meteorological department, and near 3 months of road section tidal flow data stored locally.

[0035] The collected data is subjected to outlier rejection (such as GPS jump data) and normalization processing, and a "time-space-feature" three-dimensional data matrix is generated.

[0036] The road network is abstracted as a graph structure (nodes are road sections, and edges are road section connection relationships), traffic event data and meteorological condition data collected by roadside devices are input, and graph structure data is constructed.

[0037] Further as an optional implementation, the road section congestion prediction model is trained through the following steps: S201, acquire first vehicle speed time series samples and first vehicle flow time series samples of each sample road section in a sample area in a first historical period, traffic event samples and meteorological condition samples at a first historical time, and second vehicle speed time series samples and second vehicle flow time series samples in a second historical period; S202, construct training samples according to the first vehicle speed time series samples, the first vehicle flow time series samples, the traffic event samples and the meteorological condition samples, and determine corresponding average vehicle speed labels, average acceleration labels and congestion level labels according to the second vehicle speed time series samples and the second vehicle flow time series samples; S203, input the training sample into the pre-constructed LSTM-GCN hybrid neural network to obtain the predicted average speed, the predicted average acceleration and the predicted congestion level of each sample link; S204, determine the loss value according to the predicted average speed, the predicted average acceleration, the predicted congestion level, the average speed label, the average acceleration label and the congestion level label; S205, update the parameters of the LSTM-GCN hybrid neural network by the back propagation algorithm according to the loss value to obtain the trained link congestion prediction model; The LSTM-GCN hybrid neural network comprises an LSTM branch network, a GCN branch network, a feature fusion layer and a full connection layer, and the first historical time is the end time of the first historical period and the start time of the second historical period.

[0038] Specifically, the first speed time series sample and the first traffic flow time series sample of each sample link in the sample area in the first historical period, the traffic event sample and the meteorological condition sample at the first historical time, and the second speed time series sample and the second traffic flow time series sample in the second historical period are obtained; the training sample is constructed according to the first speed time series sample, the first traffic flow time series sample, the traffic event sample and the meteorological condition sample, and the corresponding average speed label, average acceleration label and congestion level label are determined according to the second speed time series sample and the second traffic flow time series sample; the training sample is input into the pre-constructed LSTM-GCN hybrid neural network to obtain the predicted average speed, the predicted average acceleration and the predicted congestion level of each sample link; the loss value is determined according to the predicted average speed, the predicted average acceleration, the predicted congestion level, the average speed label, the average acceleration label and the congestion level label, and the parameters of the LSTM-GCN hybrid neural network are updated by the back propagation algorithm according to the loss value, that is, one iteration training is completed; when the iteration number reaches a preset threshold or the loss value is lower than the preset threshold, the training is stopped, and the trained link congestion prediction model is obtained.

[0039] Further as an optional implementation, the training sample is input into the pre-constructed LSTM-GCN hybrid neural network to obtain the predicted average speed, the predicted average acceleration and the predicted congestion level of each sample link, which specifically comprises: S2031, input the first speed time series sample and the first traffic flow time series sample into the LSTM branch network to obtain the time series feature vector of each sample link; S2032, the sample link is taken as a node, the connection relationship between the sample links is taken as an edge, and the node features are determined according to the traffic event sample and the meteorological condition sample to construct the graph structure sample of the sample area; S2033, input the graph structure sample into the GCN branch network to obtain a spatial correlation feature vector of the sample area; S2034, perform feature fusion on the time sequence feature vector and the spatial correlation feature vector based on an attention mechanism through a feature fusion layer to obtain a spatiotemporal fusion feature; S2035, map the spatiotemporal fusion feature to a predicted average vehicle speed, a predicted average acceleration and a predicted congestion level of each sample road section through a fully connected layer.

[0040] Specifically, the embodiment of the application constructs a double-branch network of "time sequence feature extraction + spatial correlation modeling", wherein: the LSTM branch is used to input the historical vehicle flow and speed time sequence data (time step 15 minutes) of the road section, capture the time dependence of the traffic flow through 3-layer LSTM units, learn the time sequence rules such as "vehicle flow surge 5 minutes before the morning peak", and obtain the time sequence feature vector of each road section; the GCN branch is used to input the graph structure data, learn the spatial correlation features of "accident road section and upstream and downstream road sections" through 2-layer graph convolution, and obtain the spatial correlation feature vector reflecting the spatial correlation between each road section in the area; the feature fusion layer is used to assign dynamic weights (for example, the GCN branch weight is 0.6 in the accident scene, and the LSTM branch weight is 0.55 in the normal scene) to the double-branch features through the attention mechanism and perform feature fusion to obtain the spatiotemporal fusion feature; and the fully connected layer is used to map the spatiotemporal fusion feature to the congestion level (1-5 levels), average speed prediction value and average acceleration prediction value of each road section in the next 20 minutes. The model training adopts a "pre-training + fine-tuning" mode: the model is pre-trained using the historical data of the entire city for 3 months, and then the parameters of the sub-models of each area are fused through federated learning to avoid data privacy leakage.

[0041] The vehicle speed time sequence data, vehicle flow time sequence data and graph structure data obtained in the foregoing steps are input into the pre-trained road congestion prediction model to obtain the average vehicle speed, average acceleration and congestion level of each road section in the preset future period, and then the average vehicle speed, average acceleration and congestion level of each road section are used to perform "time-fuel consumption" double-objective path optimization.

[0042] Specifically, based on the prediction results, a double-objective optimization function is constructed and an improved NSGA-III algorithm is used for solving: 1) A plurality of initial driving paths are determined according to the starting position and destination position of the target vehicle, and a plurality of candidate driving paths are selected from the initial driving paths according to the congestion level, the constraint condition being to avoid road sections with a congestion level greater than or equal to 4 and the maximum detour distance being not more than 20% of the shortest path; 2) Constructing time cost function: calculate the travel time of each road segment based on the predicted vehicle speed, and superimpose the waiting time at the intersection (derived from the signal timing data obtained by V2X), to obtain the total time cost of each candidate driving path (road segment length / predicted vehicle speed + intersection waiting time); 3) Constructing fuel consumption cost function: based on the vehicle dynamics model, establish the "speed-acceleration-fuel consumption" correlation formula, and the fuel consumption cost of each candidate driving path; 4) NSGA-III algorithm optimization solution: introduce "driver preference coefficient" (0-1 adjustable), when the coefficient = 0.7, prefer to guarantee time, and when the coefficient = 0.3, prefer to save fuel, finally output 3 groups of Pareto optimal solutions for user selection.

[0043] Further as an optional implementation, a plurality of initial driving paths are determined according to the starting position and the destination position of the target vehicle, and a plurality of candidate driving paths are selected from the initial driving paths according to the congestion level, which specifically comprises: S1031, path planning is performed according to the starting position, the destination position and the road topology, a plurality of initial driving paths are obtained, and the shortest path distance of the shortest driving path is determined; S1032, a plurality of congestion road segments are determined according to the congestion level, whether each initial driving path contains a congestion road segment is judged, and whether the ratio of the path distance of each initial driving path to the shortest path distance is greater than a preset first threshold value is judged; S1033, when the initial driving path does not contain a congestion road segment, and the ratio of the path distance of the initial driving path to the shortest path distance is less than or equal to the first threshold value, the initial driving path is determined as a candidate driving path.

[0044] Specifically, the city road topology structure (road segment length, lane number, speed limit information) is obtained, and the road segments with congestion level ≥ 4 (which need to be excluded or set as unfeasible paths) are marked; the shortest path from the starting position to the destination position is generated using Dijkstra or A* algorithm, the shortest path distance D0 is calculated, and the maximum detour distance threshold is set as 1.2xD0; a plurality of initial driving paths from the starting position to the destination position are generated, paths containing road segments with congestion level ≥ 4 and paths with path distance ratio greater than 1.2 to the shortest path distance are filtered out, and all candidate driving paths are ensured to meet the constraint conditions.

[0045] Further as an optional implementation, the time cost function is:

[0046] wherein, represents the time cost of the candidate driving path, represents the number of road segments of the candidate driving path, This represents the length of the i-th segment of the candidate driving path. This represents the average speed of the i-th segment of the candidate driving path. This represents the waiting time at the intersection of the i-th segment of the candidate driving route; The fuel consumption cost function is:

[0047] in, This represents the fuel cost of the candidate driving route. This represents the average acceleration of the i-th segment of the candidate driving path. , as well as The calibrated fuel consumption parameters are for the target vehicle.

[0048] Specifically, the traffic light timing schemes (red light duration, green light duration, phase cycle) at each intersection are obtained through the vehicle-road cooperative system to calculate the waiting time at the intersection; and the calibrated fuel consumption parameters of the target vehicle are determined. (Acceleration influence coefficient) (Vehicle speed influence coefficient) (Base fuel consumption constant) can be obtained from the vehicle manual or bench test.

[0049] Constructing a time cost function ,in, The intersection waiting time for the i-th segment of the candidate driving path can be derived from the traffic light timing (e.g., red light duration × vehicle arrival phase probability).

[0050] Constructing the fuel consumption cost function , , as well as The calibrated fuel consumption parameters corresponding to the target vehicle (e.g., small cars) , as well as (The calibration needs to be based on the actual vehicle model).

[0051] As an optional implementation, the candidate driving paths are further optimized using a non-dominated sorting genetic algorithm based on the time cost function, fuel consumption cost function, average vehicle speed, and average acceleration to obtain the optimal driving path, which specifically includes: S1041. Determine the segment number of each road segment, encode the candidate driving path according to the segment number to obtain multiple individuals, and initialize the population according to the individuals; S1042. Calculate the path time cost and path fuel cost of each individual based on the time cost function and the fuel cost function. Sort the individuals non-dominated based on the path time cost and path fuel cost to obtain the Pareto level of each individual. S1043, determine the time congestion degree of the individuals in the same Pareto level in the time cost space and the fuel consumption congestion degree in the fuel consumption cost space; S1044, determine the time cost weight and the fuel consumption cost weight according to the driver preference coefficient of the target vehicle, and determine the weighted congestion degree of each individual according to the time cost weight, the fuel consumption cost weight, the time congestion degree and the fuel consumption congestion degree; S1045, screen the parent individuals by the tournament selection method according to the Pareto level and the weighted congestion degree; S1046, perform the crossover and mutation operations on the parent individuals to generate offspring individuals; S1047, combine the parent individuals and the offspring individuals, and screen the next generation population by the non-dominated sorting. S1048, when the preset convergence condition is reached, determine the optimal driving path according to the Pareto optimal solution of the current population.

[0052] Specifically, the specific process of optimizing and solving the candidate driving path based on the non-dominated sorting genetic algorithm is as follows: 1) Algorithm initialization Population coding: integer coding is used to represent the path (such as road segment number sequence), and N initial paths are randomly generated (which need to meet the detour distance constraint).

[0053] Reference point setting: uniformly generate a reference point set (such as 20 reference points) according to the time-fuel consumption target space, which is used for subsequent non-dominated sorting.

[0054] 2) Non-dominated sorting and congestion calculation The paths in the population are non-dominated sorted according to the time cost T and the fuel consumption cost F, and are divided into Pareto levels (the lower the level, the better).

[0055] The congestion of the individuals in the same level (the size of the blank area around the individual in the target space) is calculated, and the greater the congestion, the better the individual diversity.

[0056] 3) Introducing driver preference coefficient Preference guiding mechanism: when the driver preference coefficient λ = 0.7, increase the weight of the time cost in the congestion calculation, for example, set the time cost weight coefficient to be the same as the driver preference coefficient, that is, 0.7, and the fuel consumption cost weight coefficient is 0.3, and the final weighted congestion = 0.7 x time congestion + 0.3 x fuel consumption congestion); when λ = 0.3, vice versa.

[0057] Selection operation: combine the Pareto level and the weighted congestion, and screen the parent individuals by the tournament selection method.

[0058] 4) Crossover and mutation operations Crossover: Partial Map Crossover (PMX) is used to exchange parts of two paths, ensuring that the paths still meet the topological connectivity after crossover.

[0059] Mutation: Randomly replace a certain segment in the path with an adjacent feasible segment (need to check the detour distance constraint) to avoid the algorithm falling into local optimum.

[0060] 5) Elite preservation strategy: Combine the parent and child generations, select the next generation population through non-dominated sorting and reference point association, and preserve the best individuals and individuals with good diversity.

[0061] 6) Generation and output of Pareto optimal solutions Iteration termination condition (convergence condition): stop calculation when the number of iterations reaches the preset value (such as 100 generations) or the population converges (the optimal solution does not change for 10 consecutive generations).

[0062] Pareto optimal solution screening: extract non-dominated solutions from the final population to form a Pareto optimal solution set. According to the driver preference coefficient λ = 0.3, 0.5, 0.7, select 3 groups of typical solutions from the solution set: (1) Time priority solution (λ = 0.7): T is the smallest, F is slightly high; (2) Balanced solution (λ = 0.5): T and F are balanced; (3) Fuel consumption priority solution (λ = 0.3): F is the smallest, T is slightly high.

[0063] Result output: display the time cost, fuel consumption cost, total distance of the path and recommended route of the three groups of solutions in table form, and mark the predicted vehicle speed and signal light waiting time of each road segment.

[0064] In some optional embodiments, the embodiments of the present application receive the latest roadside data and weather information through V2X every 30 seconds, trigger incremental updating of the model, and when the congestion level of a road segment changes by ≥2 levels, automatically re-plan the path and prompt the driver through the vehicle-mounted HUD and voice.

[0065] The method steps of the embodiments of the present application are described above. It can be recognized that the embodiments of the present application predict the average speed, average acceleration and congestion level of the road segment based on the speed time series data, traffic flow time series data, traffic event data and weather condition data of the road segment, optimize the path based on the non-dominated sorting genetic algorithm according to the time cost and fuel consumption cost, improve the accuracy and real-time performance of road congestion prediction and vehicle path optimization, and thus improve the travel experience of users.

[0066] Compared with the prior art, the embodiments of the present application have the following advantages: 1) More accurate congestion prediction: Breakthrough existing technology data dimension limitation, fusion traffic flow, weather, accident and other multi-class key data, through LSTM-GCN double branch model deep mining traffic flow space-time correlation characteristics, congestion prediction accuracy and stability significantly improved, can effectively avoid the problem of "navigation and actual road condition do not match".

[0067] 2) More optimized travel cost: Abandon single target planning mode, build "time-fuel consumption" double target optimization system, combine with driver preference dynamic balance traffic efficiency and energy consumption, solve the contradiction of traditional route "low efficiency and oil saving" or "high efficiency and fuel consumption", improve travel experience and reduce energy consumption.

[0068] With reference to Figure 2 The embodiment of the application provides a congestion prediction and path optimization device based on vehicle-road cooperation, comprising: A data acquisition module is configured to acquire vehicle speed time series data, traffic flow time series data, traffic event data and meteorological condition data of each road section in a target area, and determine graph structure data of the target area according to the traffic event data, the meteorological condition data and a road topology of the target area. A congestion prediction module is configured to input the vehicle speed time series data, the traffic flow time series data and the graph structure data into a pre-trained road section congestion prediction model to obtain average vehicle speed, average acceleration and congestion level of each road section in a preset future period. A path screening module is configured to determine a plurality of initial driving paths according to a starting position and a destination position of a target vehicle, and screen a plurality of candidate driving paths from the initial driving paths according to the congestion level. A path optimization module is configured to construct a time cost function and a fuel consumption cost function, and optimize and solve the candidate driving paths based on a non-dominated sorting genetic algorithm according to the time cost function, the fuel consumption cost function, the average vehicle speed and the average acceleration to obtain an optimal driving path.

[0069] It can be understood that the contents in the above method embodiments are applicable to the device embodiments, the device embodiments specifically realize the same functions as the above method embodiments, and achieve the same beneficial effects as the above method embodiments.

[0070] With reference to Figure 3 The embodiment of the application provides an electronic device, comprising: At least one processor; At least one memory configured to store at least one program; When the above at least one program is executed by the above at least one processor, the above at least one processor realizes the above congestion prediction and path optimization method based on vehicle-road cooperation.

[0071] It can be understood that the contents in the above method embodiments are all applicable to the present device embodiments, the present device embodiments specifically implement the same functions as the above method embodiments, and achieve the same beneficial effects as the above method embodiments.

[0072] The embodiment of the present application further provides a computer readable storage medium, wherein a computer program executable by a processor is stored, and the computer program implements the above-mentioned congestion prediction and path optimization method based on vehicle-road cooperation when executed by the processor.

[0073] The computer readable storage medium of the embodiment of the present application can execute the congestion prediction and path optimization method based on vehicle-road cooperation provided by the method embodiment of the present application, execute the step of any combination of the method embodiment, and has the corresponding functions and beneficial effects of the method.

[0074] The embodiment of the present application further provides a computer program product, comprising a computer program, which, when executed by a processor, implements the above-mentioned congestion prediction and path optimization method based on vehicle-road cooperation.

[0075] It can be understood that the contents in the above method embodiments are all applicable to the present program product embodiments, the present program product embodiments specifically implement the same functions as the above method embodiments, and achieve the same beneficial effects as the above method embodiments.

[0076] The memory is a non-transitory computer readable storage medium, which can be used to store non-transitory software programs and non-transitory computer executable programs. In addition, the memory can include a high-speed random access memory, and can also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state memory device. In some embodiments, the memory can optionally include a memory remotely arranged relative to the processor, and these remote memories can be connected to the processor through 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.

[0077] The embodiments described in the embodiments of the present application are for more clearly illustrating the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application. Those skilled in the art can know that, with the evolution of technology and the appearance of new application scenarios, the technical solutions provided by the embodiments of the present application are also applicable to similar technical problems.

[0078] The terms "first," "second," "third," "fourth," etc. (if present) in the specification and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0079] In some alternative embodiments, the functions / operations mentioned in the block diagrams may not occur in the order shown in the operation diagrams. For example, depending on the functions / operations involved, two consecutively shown blocks may actually be executed substantially simultaneously, or the aforementioned blocks may sometimes be executed in reverse order. Furthermore, the embodiments presented and described in the flowcharts of this invention are provided by way of example to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is altered and sub-operations described as part of a larger operation are executed independently.

[0080] Furthermore, although the invention has been described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the aforementioned functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the invention. Rather, given the properties, functions, and internal relationships of the various functional modules in the apparatus disclosed herein, the actual implementation of the module will be understood within the scope of conventional skill of an engineer. Therefore, those skilled in the art can implement the invention as set forth in the claims using ordinary techniques without excessive experimentation. It is also understood that the specific concepts disclosed are merely illustrative and not intended to limit the scope of the invention, which is determined by the full scope of the appended claims and their equivalents.

[0081] If the above functions are realized in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application or parts of the present application that essentially contribute to the prior art or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0082] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a list of executable instructions for implementing logic functions, which can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus or device, such as a computer-based system, a system including a processor or other system that can fetch instructions from an instruction execution system, apparatus or device and execute the instructions, or in conjunction with these instruction execution systems, apparatus or devices. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transport programs for use by or in connection with an instruction execution system, apparatus or device, or in conjunction with these instruction execution systems, apparatus or devices.

[0083] More specific examples (non-exhaustive list) of the computer-readable medium include the following: an electrical connection having one or more wires (electrical devices), a portable computer diskette (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CD ROM). In addition, the computer readable medium can even be paper or other suitable medium on which the above programs are printed, because the programs can be obtained electronically, for example, by optical scanning of the paper or other medium, followed by editing, interpreting or otherwise processing, if necessary, in other suitable ways, to be electronically obtained and then stored in the computer memory.

[0084] It should be understood that various parts of the present application can be implemented in hardware, software, firmware or a combination thereof. In the above embodiments, various steps or methods can be implemented in software or firmware that is stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any of the following technologies, known in the art, or their combinations can be used: a discrete logic circuit having logic gates for implementing logic functions on data signals, an application specific integrated circuit having appropriate combinational logic gates, a programmable gate array (PGA), a field programmable gate array (FPGA), or the like.

[0085] In the above description of the present specification, the description of the terms "one embodiment / one example", "another embodiment / another example", or "certain embodiments / certain examples" or the like means that the specific features, structures, materials or characteristics described in connection with the embodiments or examples are included in at least one embodiment or example of the present application. In the present specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Also, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0086] Although the embodiments of the present application have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made thereto without departing from the principles and spirit of the present application, the scope of which is defined by the claims and their equivalents.

[0087] The above is a specific description of the preferred embodiments of the present application, but the present application is not limited to the above-described embodiments, and those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present application, and these equivalent modifications or substitutions are included in the scope defined by the claims of the present application.

Claims

1. A congestion prediction and route optimization method based on vehicle-road cooperation, characterized in that, Includes the following steps: Acquire time-series data of vehicle speed, traffic flow, traffic incidents, and meteorological conditions for each road segment within the target area; and determine the graph structure data of the target area based on the traffic incident data, the meteorological conditions data, and the road topology of the target area. The vehicle speed time series data, the traffic flow time series data and the graph structure data are input into a pre-trained road segment congestion prediction model to obtain the average vehicle speed, average acceleration and congestion level of each road segment in a preset future time period. Multiple initial driving routes are determined based on the starting and destination locations of the target vehicle, and multiple candidate driving routes are selected from the initial driving routes based on the congestion level. A time cost function and a fuel consumption cost function are constructed. Based on the time cost function, the fuel consumption cost function, the average vehicle speed, and the average acceleration, the candidate driving paths are optimized using a non-dominated sorting genetic algorithm to obtain the optimal driving path.

2. The congestion prediction and route optimization method based on vehicle-road cooperation according to claim 1, characterized in that, The process of acquiring time-series vehicle speed data, time-series traffic flow data, traffic event data, and meteorological condition data for each road segment within the target area, and determining the graph structure data of the target area based on the traffic event data, the meteorological condition data, and the road topology of the target area, specifically includes: The vehicle speed time sequence data of each vehicle in each of the road segments is obtained through V2X. Traffic flow time-series data for each road segment is acquired through roadside radar equipment, and traffic event data for each road segment is acquired through roadside camera equipment. Meteorological condition data for each of the aforementioned road sections are obtained through a meteorological platform; The connection relationships between the road segments are determined based on the road topology. The road segments are treated as nodes, and the connection relationships between the road segments are treated as edges. The node characteristics are determined based on the traffic event data and the meteorological condition data to construct the graph structure data.

3. The congestion prediction and route optimization method based on vehicle-road cooperation according to claim 1, characterized in that, The road congestion prediction model is trained through the following steps: The sample area obtains the first vehicle speed time series sample and the first traffic flow time series sample, the traffic event sample and the meteorological condition sample at the first historical moment, and the second vehicle speed time series sample and the second traffic flow time series sample for each sample road segment in the first historical time period. Training samples are constructed based on the first vehicle speed time series sample, the first traffic flow time series sample, the traffic event sample, and the meteorological condition sample, and corresponding average vehicle speed labels, average acceleration labels, and congestion level labels are determined based on the second vehicle speed time series sample and the second traffic flow time series sample. The training samples are input into a pre-constructed LSTM-GCN hybrid neural network to obtain the predicted average vehicle speed, predicted average acceleration, and predicted congestion level for each sample road segment. The loss value is determined based on the predicted average vehicle speed, the predicted average acceleration, the predicted congestion level, the average vehicle speed label, the average acceleration label, and the congestion level label; The parameters of the LSTM-GCN hybrid neural network are updated using the backpropagation algorithm based on the loss value to obtain the trained road segment congestion prediction model. The LSTM-GCN hybrid neural network includes an LSTM branch network, a GCN branch network, a feature fusion layer, and a fully connected layer. The first historical moment is the end time of the first historical period and the start time of the second historical period.

4. The congestion prediction and route optimization method based on vehicle-road cooperation according to claim 3, characterized in that, The step of inputting the training samples into a pre-constructed LSTM-GCN hybrid neural network to obtain the predicted average vehicle speed, predicted average acceleration, and predicted congestion level for each sample road segment specifically includes: The first vehicle speed time series sample and the first traffic flow time series sample are input into the LSTM branch network to obtain the time series feature vector of each sample road segment; The sample road segments are used as nodes, the connections between the sample road segments are used as edges, and the node features are determined based on the traffic event samples and the meteorological condition samples to construct a graph structure sample of the sample area. The graph structure samples are input into the GCN branch network to obtain the spatial correlation feature vector of the sample region; The feature fusion layer performs feature fusion on the temporal feature vector and the spatial correlation feature vector based on an attention mechanism to obtain spatiotemporal fused features. The spatiotemporal fusion features are mapped to the predicted average vehicle speed, the predicted average acceleration, and the predicted congestion level for each of the sample road segments through the fully connected layer.

5. The congestion prediction and route optimization method based on vehicle-road cooperation according to claim 1, characterized in that, The process of determining multiple initial driving routes based on the starting and destination locations of the target vehicle, and then selecting multiple candidate driving routes from these initial routes based on the congestion level, specifically includes: Based on the starting position, the destination position, and the road topology, path planning is performed to obtain multiple initial driving paths, and the shortest path distance of the shortest driving path is determined. Based on the congestion level, several congested road segments are determined, and it is determined whether each initial driving path includes the congested road segment, and whether the ratio of the path distance of each initial driving path to the shortest path distance is greater than a preset first threshold. If the initial driving path does not include the congested road segment, and the ratio of the path distance of the initial driving path to the shortest path distance is less than or equal to the first threshold, the initial driving path is determined as the candidate driving path.

6. The congestion prediction and route optimization method based on vehicle-road cooperation according to claim 1, characterized in that, The time cost function is: in, This represents the time cost of the candidate driving route. This indicates the number of road segments in the candidate driving route. This represents the length of the i-th segment of the candidate driving path. This represents the average speed of the i-th segment of the candidate driving path. This represents the waiting time at the intersection of the i-th segment of the candidate driving route; The fuel consumption cost function is: in, This represents the fuel cost of the candidate driving route. This represents the average acceleration of the i-th segment of the candidate driving path. , as well as The calibrated fuel consumption parameters are those corresponding to the target vehicle.

7. A congestion prediction and route optimization method based on vehicle-road cooperation according to any one of claims 1 to 6, characterized in that, The step of optimizing the candidate driving paths based on the time cost function, the fuel consumption cost function, the average vehicle speed, and the average acceleration using a non-dominated sorting genetic algorithm to obtain the optimal driving path specifically includes: Determine the segment number of each road segment, encode the candidate travel path according to the segment number to obtain multiple individuals, and initialize the population according to the individuals; The path time cost and path fuel cost of each individual are calculated based on the time cost function and the fuel cost function. The individuals are then non-dominated and sorted according to the path time cost and the path fuel cost to obtain the Pareto level of each individual. Determine the time congestion degree of individuals of the same Pareto level in the time cost space and the fuel consumption congestion degree in the fuel consumption cost space; The time cost weight and fuel consumption cost weight are determined based on the driver preference coefficient of the target vehicle, and the weighted congestion degree of each individual is determined based on the time cost weight, the fuel consumption cost weight, the time congestion degree, and the fuel consumption congestion degree. Parent individuals are selected using a tournament selection method based on the Pareto rank and the weighted crowding degree. Crossover and mutation operations are performed on the parent individuals to generate offspring individuals; The parent individuals are merged with the offspring individuals, and the next generation population is obtained by screening through non-dominated sorting. When the preset convergence condition is met, the optimal travel path is determined based on the Pareto optimal solution of the current population.

8. A congestion prediction and route optimization device based on vehicle-road cooperation, characterized in that, include: The data acquisition module is used to acquire time-series data of vehicle speed, traffic flow, traffic events, and meteorological conditions of each road segment within the target area, and to determine the graph structure data of the target area based on the traffic event data, the meteorological conditions, and the road topology of the target area. The congestion prediction module is used to input the vehicle speed time series data, the traffic flow time series data and the graph structure data into the pre-trained road segment congestion prediction model to obtain the average vehicle speed, average acceleration and congestion level of each road segment in a preset future time period. The route filtering module is used to determine multiple initial driving routes based on the starting position and destination position of the target vehicle, and to filter multiple candidate driving routes from the initial driving routes based on the congestion level. The path optimization module is used to construct a time cost function and a fuel consumption cost function. Based on the time cost function, the fuel consumption cost function, the average vehicle speed, and the average acceleration, it optimizes the candidate driving paths using a non-dominated sorting genetic algorithm to obtain the optimal driving path.

9. An electronic device, characterized in that, include: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements a vehicle-road cooperative congestion prediction and route optimization method as described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements a congestion prediction and path optimization method based on vehicle-road cooperation as described in any one of claims 1 to 7.

Images