Cloud-Supported Optimized Scheduling System for Driving Resources

Abstract

The present application provides an optimized scheduling system for driving resources combined with cloud support, which relates to the field of computer technology and includes: a traffic road network matching module that matches a target traffic road network in a target area; a first prediction module that analyzes real-time driving information to obtain a first predicted vehicle; a driving visibility generation module that acquires a target position in a target service area and generates a driving visibility view in combination with a first intersection, a first position, and a first mapping relationship of the first predicted vehicle; a second prediction module that activates a vehicle prediction middleware to analyze the driving visibility view to obtain a target predicted vehicle; and a resource scheduling module that analyzes the target predicted vehicle through a resource scheduler to obtain a target resource scheduling strategy and schedules resources for the target service area. The present application solves the technical problem that the existing driving scheduling method lacks real-time dynamic feedback, resulting in low resource scheduling efficiency, and achieves the technical effects of improving the efficiency and accuracy of driving scheduling and optimizing the utilization rate of driving resources.

Description

Technical Field

[0001] This application relates to the field of computer technology, and particularly to an optimized scheduling system for driving resources with cloud support. Background Art

[0002] In urban traffic management, the scheduling of driving resources is an important means to improve road traffic efficiency and reduce traffic congestion. Existing driving scheduling methods mainly rely on historical traffic data and fixed scheduling rules. By analyzing past traffic flows, future traffic conditions are predicted, and the allocation of service area resources is adjusted manually. These methods usually adopt fixed scheduling rules and cannot respond promptly to the dynamic changes in traffic flow. Therefore, when emergencies or real-time traffic conditions change, traditional scheduling methods often cannot quickly adjust resource allocation, resulting in longer vehicle queuing times and traffic bottlenecks. In addition, the prediction accuracy of existing methods is limited by historical data and it is difficult to accurately capture the instantaneous characteristics of traffic flow. The lack of real-time data support makes the scheduling strategy lag behind the actual demand, resulting in low resource scheduling efficiency and thus affecting road traffic. Summary of the Invention

[0003] This application provides an optimized scheduling system for driving resources with cloud support, which solves the technical problem that existing driving scheduling methods rely on static historical data and fixed scheduling rules and lack real-time dynamic feedback, resulting in low resource scheduling efficiency, and achieves the technical effect of dynamically responding to changes in the driving environment, improving the efficiency and accuracy of driving scheduling, and optimizing the utilization rate of driving resources.

[0004] In view of the above problems, this application provides an optimized scheduling system for driving resources with cloud support. The system includes: a traffic road network matching module for matching a target traffic road network of a target area in a cloud database, where the target traffic road network includes multiple intersections with location identifiers, and the target area has a target service area; a first prediction module for analyzing real-time driving information of the target traffic road network to obtain first predicted vehicles at a first intersection among multiple intersections, where the first intersection has an identifier of a first location; a driving view generation module for obtaining a target location of the target service area and generating a driving view in combination with a first mapping relationship between the first intersection, the first location, and the first predicted vehicles; a second prediction module for activating a vehicle prediction platform to analyze the driving view to obtain target predicted vehicles at the target location in a predetermined time interval; and a resource scheduling module for analyzing the target predicted vehicles through a resource scheduler to obtain a target resource scheduling strategy and performing resource scheduling on the target service area according to the target resource scheduling strategy.

[0005] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0006] The traffic road network matching module matches the target traffic road network of the target area in the cloud database, identifies multiple intersections, and clarifies the geographical range of the target service area, providing a basic geographical information basis for subsequent driving data analysis, ensuring that the analyzed traffic environment conforms to the actual situation, and improving the relevance of the data. The first prediction module obtains the first predicted vehicles at the first intersection among multiple intersections by analyzing the real-time driving information of the target traffic road network, introduces real-time traffic dynamic data, and provides an instant feedback on the current traffic flow. The driving view generation module obtains the target location of the target service area and generates a driving view in combination with the first mapping relationship among the first intersection, the first location, and the first predicted vehicles, visually displaying the traffic conditions and flow trends. The second prediction module activates the vehicle prediction center platform to analyze the driving view, obtains the target predicted vehicles at the target location within a predetermined time interval, refines the prediction of the future traffic conditions in the target service area, and provides more targeted data for resource scheduling. The resource scheduling module analyzes the target predicted vehicles through a resource scheduler to obtain a target resource scheduling strategy, and performs resource scheduling on the target service area according to the target resource scheduling strategy, optimizing the utilization of service area resources.

[0007] In summary, through the collaborative work of each module in this application, the traffic road network matching module provides accurate basic data, the first prediction module analyzes real-time traffic flow information, the driving view generation module visualizes traffic conditions, the second prediction module prospectively predicts future traffic flow, and the resource scheduling module optimizes resource allocation according to the prediction results, achieving accurate prediction and optimized scheduling of driving resources, improving the flexibility and accuracy of driving resource scheduling, thus more effectively planning resource allocation, reducing vehicle waiting time, and ultimately improving the utilization rate of driving resources and the overall traffic flow efficiency.

[0008] The above description is only an overview of the technical solutions of this application. In order to be able to understand the technical means of this application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features, and advantages of this application more obvious and understandable, the following specifically gives the specific implementation manners of this application. Brief Description of the Drawings

[0009] Figure 1 It is a schematic structural diagram of a driving resource optimization scheduling system with cloud support provided by an embodiment of this application.

[0010] Figure 2Schematic diagram of the process of obtaining the first predicted vehicle in the vehicle driving resource optimization scheduling system with cloud support provided by the embodiments of the present application.

[0011] Figure 3 Schematic diagram of the process of obtaining the target predicted vehicle in the vehicle driving resource optimization scheduling system with cloud support provided by the embodiments of the present application.

[0012] Explanation of reference numerals: Traffic road network matching module 10, first prediction module 20, vehicle driving view generation module 30, second prediction module 40, resource scheduling module 50. Detailed implementation manners

[0013] By providing a vehicle driving resource optimization scheduling system with cloud support, the embodiments of the present application solve the technical problem that the existing vehicle driving scheduling method relies on static historical data and fixed scheduling rules and lacks real-time dynamic feedback, resulting in low resource scheduling efficiency, and achieve the technical effects of dynamically responding to changes in the vehicle driving environment, improving the efficiency and accuracy of vehicle driving scheduling, and optimizing the utilization rate of vehicle driving resources.

[0014] As Figure 1 shown, the embodiments of the present application provide a vehicle driving resource optimization scheduling system with cloud support, and the system includes:

[0015] A traffic road network matching module 10, which is used to match a target traffic road network of a target area in a cloud database. The target traffic road network includes a plurality of intersections with position identifiers, wherein the target area has a target service area.

[0016] Specifically, the cloud database is used to store and manage a large amount of traffic data, including roads, intersections and real-time traffic flow information, and can be accessed through the Internet. The target area refers to a city or area that needs traffic management and scheduling. The target traffic road network is the traffic road network within the target area, including main roads and intersections, and is used to support traffic flow. The position identifier is a unique identifier assigned to each intersection or road, which is convenient for data matching and processing. The target service area is a comprehensive space that provides traffic services for the target area, including vehicle depots (parking lots), comprehensive maintenance centers, general material warehouses and other supporting facilities, and has various traffic operation and maintenance functions such as information management, scheduling plan management, and crew assignment management.

[0017] The traffic road network matching module 10 first accesses the cloud database through the Internet and extracts information on all main roads and intersections within the target area from the cloud database. This information is usually stored in the form of vector data. Then, using Geographic Information System (GIS) technology, corresponding location identifiers are established based on the specific locations of each intersection. Exemplarily, the location identifier can be the longitude and latitude coordinates of each intersection. By obtaining the target traffic road network, this module provides the basic geographical framework information for subsequent vehicle resource scheduling work.

[0018] The first prediction module 20 is used to analyze the real-time driving information of the target traffic road network to obtain the first predicted vehicles at the first intersection among multiple intersections. The first intersection has an identifier of the first location.

[0019] Specifically, the real-time driving information refers to the traffic data collected in real time, including the number of vehicles, vehicle displacement, vehicle speed, passenger flow, congestion status, etc. This data is usually collected through devices such as sensors and cameras. The first intersection is a specific intersection in the target traffic road network, corresponding to the first location identifier. The first predicted vehicle is the prediction result of the number of vehicles obtained by analyzing the real-time driving information for the first intersection, and is used to evaluate the traffic flow at this intersection.

[0020] The first prediction module 20 obtains the real-time driving information of the target traffic road network through network information interaction. This real-time driving information can be monitored and obtained through sensors, cameras, GPS devices, etc. deployed on the road and is transmitted to the cloud database for storage and management. For example, traffic monitoring cameras can capture the real-time vehicle flow at a certain intersection, and sensors can measure the speed and number of passing vehicles.

[0021] Randomly select any intersection from the target traffic road network and denote it as the first intersection. Taking this first intersection as an example, the calculation process of the corresponding first predicted vehicle information is described. The module uses a data analysis algorithm to filter out the driving information related to the first intersection. This algorithm is based on traffic flow theory and determines which vehicles will drive towards the first intersection according to information such as the driving direction and current location of the vehicles. For example, if the driving direction and the current road of the vehicle are known, and the connection relationship between the road and the first intersection is known, it is possible to predict which vehicles will reach the first intersection. By processing and calculating the filtered driving information, the first predicted vehicle information for the first intersection is obtained. For example, it is predicted that within the next 5 minutes, 5 cars and 2 trucks will reach this first intersection with the first location identifier.

[0022] By analyzing real-time driving information, this module can calculate the predicted vehicle information for any intersection in the target traffic road network according to the current traffic conditions, providing a data basis for subsequent traffic management and resource scheduling.

[0023] A driving view generation module 30, which is used to obtain the target location of the target service area and generate a driving view in combination with the first mapping relationship between the first intersection, the first location, and the first predicted vehicle.

[0024] Specifically, the driving view generation module 30 obtains the target location information of the target service area from the cloud database, that is, the specific location of the target service area in the geographical space. Then, according to the correspondence between the first intersection, the first location, and the first predicted vehicle, these three types of data are integrated to establish the first mapping relationship, ensuring that the predicted vehicle data can be correctly mapped to the corresponding intersection location, so as to accurately reflect the traffic flow in the view. Then, using a graphics drawing tool or a data visualization tool (such as D3.js, Tableau, or Plotly), according to the integrated information, the relationship between the location of the target service area, the first intersection, and the first predicted vehicle is displayed in an intuitive graphical way to generate a driving view. For example, using a visualization tool based on GIS (Geographic Information System), mark the location of the target service area on the map, display the first intersection with a specific icon, and use arrows or other means to represent the driving direction of the first predicted vehicle and the expected arrival at the first intersection, thus generating a view that can clearly display the driving-related information.

[0025] By generating a driving view, the traffic flow, vehicle distribution, and possible traffic jams at intersections can be intuitively displayed, and at the same time, reference data is provided for subsequent target vehicle prediction and driving resource scheduling. This driving view is dynamically updated according to the changes in real-time traffic flow to reflect the latest traffic conditions at intersections.

[0026] A second prediction module 40, which is used to activate the vehicle prediction middle platform to analyze the driving view and obtain the target predicted vehicles at the target location within a predetermined time interval.

[0027] Specifically, the vehicle prediction middle platform is an integrated analysis platform that uses machine learning and data analysis algorithms to process traffic data to generate future traffic predictions. The predetermined time interval refers to a specific time period for which traffic flow prediction is required. The target predicted vehicle is the result obtained by the second prediction module 40 through the vehicle prediction middle platform's analysis of the driving view, that is, the number of vehicles passing through the target location within the predetermined time interval.

[0028] The second prediction module 40 activates the vehicle prediction center platform by sending specific instructions or request signals, and sends the relevant data of the driving viewable map to the vehicle prediction center platform. This vehicle prediction center platform is usually built using programming languages such as Python, R, or Java, and relies on open-source machine learning libraries (such as TensorFlow or Scikit-learn) for training and prediction of the prediction model.

[0029] The vehicle prediction center platform receives the driving viewable map as input data, identifies and parses various data elements in the driving viewable map, including the target location of the target service area, the first intersection, etc., and uses built-in analysis algorithms and data processing techniques to predict the traffic flow situation within a future predetermined time interval. These algorithms can be traffic flow prediction algorithms based on time series analysis, such as ARIMA (Autoregressive Integrated Moving Average Model) or LSTM (Long Short-Term Memory Network). The vehicle prediction center platform feeds back the analysis result to the second prediction module 40, and this result is the target predicted vehicles at the target location within the predetermined time interval. For example, if the predetermined time interval is the next 30 minutes, the vehicle prediction center platform analyzes the traffic flow data of the past week and combines the real-time information of the current day to predict that 20 cars and 5 buses will arrive at the target location within this time interval.

[0030] The second prediction module 40 analyzes the driving viewable map through the vehicle prediction center platform, providing forward-looking data support for subsequent driving resource scheduling, so as to generate a suitable scheduling strategy and optimize resource allocation to cope with the expected traffic flow.

[0031] A resource scheduling module 50, which is used to analyze the target predicted vehicles through a resource scheduler to obtain a target resource scheduling strategy, and perform resource scheduling on the target service area according to the target resource scheduling strategy.

[0032] Specifically, the resource scheduler is a key component in the resource scheduling module 50, which is used to manage and allocate traffic resources, and formulate a reasonable scheduling strategy based on information such as the predicted number and type of vehicles. The target resource scheduling strategy is a specific strategy formulated according to the prediction data and actual needs, which clarifies how to schedule the resources in the target service area, including the allocation and arrangement of resources such as manpower and material resources, to meet the needs of the target predicted vehicles. For example, according to the arrival time and number of vehicles, decide how many staff to dispatch for guidance and service, and how to allocate materials (such as parking spaces, refueling facilities, etc.).

[0033] The resource scheduler in the resource scheduling module 50 obtains the target predicted vehicle information, which includes the number, type, arrival time, etc. of the target predicted vehicles at the target location within a predetermined time interval provided by the second prediction module 40. Then, the resource scheduler analyzes the information of the target predicted vehicles based on preset algorithms and rules. For example, if there are a large number of trucks, more large parking spaces and loading and unloading equipment resources may be required; if there are a large number of cars and their arrival times are concentrated, more guiding personnel may be needed to direct traffic. The resource scheduler may also consider the intervals between vehicle arrival times to determine the order of resource allocation.

[0034] Based on the analysis results, the resource scheduler formulates and outputs the target resource scheduling strategy. This strategy is a detailed resource allocation plan. For example, for the vehicles about to arrive, arrange 5 guiding personnel to guide the vehicles into the parking lot at different entrances. For trucks, reserve 10 large parking spaces and arrange 2 loading and unloading workers to be on standby at any time; for cars, allocate ordinary parking spaces in the order of expected arrival.

[0035] The resource scheduling module 50 receives and executes the target resource scheduling strategy formulated by the resource scheduler to schedule resources for the target service area. For example, the parking lot management system arranges parking spaces for vehicles according to the allocated parking space plan, the human resource management department dispatches guiding personnel and loading and unloading workers to the corresponding positions as needed, and the material management department ensures that the loading and unloading equipment is in an available state, etc., so as to achieve effective scheduling of resources in the target service area to meet the needs of the target predicted vehicles.

[0036] By formulating and executing the target resource scheduling strategy, this module can not only effectively respond to the predicted traffic flow, but also optimize the resource utilization in the target service area, improving the overall traffic management efficiency and resource utilization rate.

[0037] Furthermore, as Figure 2 shown, the first prediction module 20 in the embodiment of the present application is further used to execute the following steps:

[0038] Determine the first connecting road of the first intersection according to the target traffic road network, where the first connecting road includes multiple roads; obtain the first road among the multiple roads, and combine the target traffic road network to obtain the traveling road directions of the first road with the first number of travels; combine the real-time driving information to obtain the first real-time driving information of the first road; calculate the first predicted vehicle according to the first real-time driving information and the first number of travels.

[0039] Specifically, the first connecting road is a set of roads connected to the first intersection, usually including multiple roads entering and leaving the intersection. For example, at a crossroads, the four roads in the east, west, south, and north directions are the first connecting roads of this crossroads (the first intersection).

[0040] First, by querying the topological structure data of the target traffic road network, determine the first connecting road of the first intersection. For example, in a database storing urban traffic road network information, search for road records directly connected to the first intersection, and these records constitute the first connecting road.

[0041] Select a specific road from the multiple roads of the first connecting road, denoted as the first road. The selection process can be carried out according to the priority of the roads (such as giving priority to main roads) or randomly. Then, read the road attributes in the traffic road network data to determine the driving direction of the first road and the first number of directions. Among them, the first number of directions represents the number of driving directions of vehicles on the first road.

[0042] Filter out the real-time driving information corresponding to the first road from the real-time driving information, denoted as the first real-time driving information, including information such as vehicle speed, vehicle displacement, number of vehicles, and number of passengers. According to the first real-time driving information and the first number of directions, calculate the first predicted vehicle.

[0043] Through the above series of steps, the first prediction module 20 can accurately predict the vehicle flow at a specific intersection, provide reliable data support for the subsequent generation of the driving viewable graph and the scheduling of driving resources, and help optimize traffic flow and resource allocation.

[0044] Furthermore, the first real-time driving information described in the embodiment of the present application includes multiple vehicles with the identification of unit vehicle displacement, and the first prediction module 20 is further used to perform the following steps:

[0045] Perform clustering analysis on the multiple vehicles with the unit vehicle displacement as the clustering constraint to obtain a clustering result. The clustering result includes a first clustering cluster, and the first clustering cluster corresponds to the first unit vehicle displacement; count the first clustering cluster to obtain the first number of clustering vehicles, and combine the first number of directions to calculate the first predicted number of vehicles with the first unit vehicle displacement; sum up the first predicted numbers to obtain the first predicted vehicle.

[0046] Specifically, the unit vehicle displacement is a unit value used to measure the exhaust volume of a vehicle's engine and can be used to distinguish different types or levels of vehicles. The first clustering cluster is a clustering set in the results obtained from clustering analysis. The vehicles in this set have similar unit vehicle displacement characteristics and correspond to a specific first unit vehicle displacement. The first clustering vehicle count refers to the number of vehicles included in the first clustering cluster, which is obtained by counting the first clustering cluster. This number reflects the proportion of vehicles with the first unit vehicle displacement among multiple vehicles. The first prediction count is a predicted number of vehicles with the first unit vehicle displacement under specific conditions calculated by combining the first clustering vehicle count and the first travel count, and it is an intermediate result for calculating the first predicted vehicles.

[0047] Read multiple vehicles passing through the first road and their corresponding unit vehicle displacements from the first real-time vehicle information. Using clustering analysis with the unit vehicle displacement size as the clustering constraint, classify the multiple vehicles and group the vehicles with similar unit vehicle displacements into one category to form a clustering cluster. Commonly used clustering algorithms include K-means clustering, hierarchical clustering, or DBSCAN, etc.

[0048] Taking the K-means algorithm as an example, consider the unit vehicle displacement as a feature dimension. First, determine how many clusters need to be formed. Exemplarily, K = 3, that is, all vehicles are divided into three categories: small displacement, medium displacement, and large displacement. Then, iteratively group the vehicles. For each vehicle, based on the value of its unit vehicle displacement, assign it to different clustering clusters so that the vehicles in the same cluster have the most similar displacements, and finally obtain the clustering result. It includes multiple clustering clusters, and each clustering cluster corresponds to a unit vehicle displacement. For example, if vehicles with unit vehicle displacements of 1.5L, 1.6L, and 1.8L are relatively similar under a certain distance metric, they are grouped into the first clustering cluster, and the first unit vehicle displacement corresponding to this clustering cluster can be set to 1.5 - 1.8L (the specific value is determined according to the actual situation and analysis accuracy).

[0049] Select one of the multiple clustering clusters, denoted as the first clustering cluster, which corresponds to the first unit vehicle displacement. Taking the first clustering cluster as an example, the subsequent calculation of the prediction count is illustrated. Count the total number of vehicles in this first clustering cluster to obtain the first clustering vehicle count. Combine the first travel count to calculate the first prediction count of the vehicles with the first unit vehicle displacement. The calculation process can be through simple average calculation or weighted average calculation. Taking the average as an example, the formula for calculating the first prediction count is: First prediction count = N / M, where the first clustering vehicle count is N and the first travel count is M. Exemplarily, if the first clustering vehicle count is 50 and the first travel count is 2, then the first prediction count is 50 / 2 = 25 vehicles.

[0050] For multiple different clusters, after calculating the first prediction numbers corresponding to each cluster respectively, adding these first prediction numbers together gives the total number of first predicted vehicles. For example, if the first prediction number corresponding to the above-mentioned first unit vehicle displacement is 25 vehicles, the first prediction number corresponding to the second unit vehicle displacement is 15 vehicles, and the first prediction number corresponding to the third unit vehicle displacement is 20 vehicles, then the number of first predicted vehicles is 25 + 15 + 20 = 60 vehicles.

[0051] Classifying and counting driving based on unit vehicle displacement can more accurately predict the number of different vehicle types, so as to obtain the first predicted vehicles at the first intersection accurately and provide reliable data support for subsequent modules.

[0052] Furthermore, as Figure 3 shown, the second prediction module 40 in the embodiment of the present application is further configured to perform the following steps:

[0053] The vehicle prediction middle platform determines the first-level neighborhood of the target location according to the driving visible view; extracts any first-level intersection in the first-level neighborhood and determines the second-level neighborhood of the any first-level intersection; samples the second-level intersections in the second-level neighborhood to obtain a set of target second-level intersections; fuses the predicted vehicles of each intersection in the set of target second-level intersections to obtain the any predicted vehicle of the any first-level intersection; and records the fusion result of the any predicted vehicle as the target predicted vehicle.

[0054] Specifically, the first-level neighborhood refers to a specific area range around the target location, which is directly related to the target location and is usually relatively close. The second-level neighborhood refers to the area range determined with the first-level intersection as the center. The set of target second-level intersections is a set of intersections obtained by sampling the second-level intersections in the second-level neighborhood, and the intersections in this set are specific intersections selected for subsequent analysis after screening. The any predicted vehicle refers to the number of vehicles expected to pass through the corresponding first-level intersection within a certain time period, which is a comprehensive prediction result of the vehicle situation at any first-level intersection.

[0055] First, the vehicle prediction middle platform determines the first-level neighborhood of the target location by analyzing elements such as geographical location information and road connection relationships in the driving visible view. For example, in the driving visible view, with the target location as the center, a region is circled as the first-level neighborhood according to a certain distance radius or road connection rule, and this region may include information such as roads directly connected to the target location and surrounding buildings.

[0056] Next, after determining the first-level neighborhood, find the intersections within the first-level neighborhood area, that is, the first-level intersections. Randomly extract one intersection from them, that is, any first-level intersection. Taking this intersection as the center, determine the second-level neighborhood of this arbitrary first-level intersection according to a method similar to determining the first-level neighborhood (such as considering factors like distance and road connections). For example, if the first-level neighborhood is a block range, then the crossroads within the block are arbitrary first-level intersections. Then, taking this crossroads as the center, circle a slightly smaller surrounding area as the second-level neighborhood again.

[0057] Then, sample the intersections (i.e., second-level intersections) in the second-level neighborhood to obtain the target second-level intersection set. There may be multiple second-level intersections in the second-level neighborhood. Select a part of the intersections to form the target second-level intersection set through a certain sampling method (such as random sampling, sampling according to a certain weight ratio, sampling according to a predetermined rule, etc.). For example, in a second-level neighborhood containing 10 second-level intersections, select 5 intersections according to the method of equal-probability random sampling to form the target second-level intersection set.

[0058] After that, fuse the predicted vehicles of each intersection in the target second-level intersection set to obtain the arbitrary predicted vehicle of any first-level intersection. The fusion process can use methods such as weighted average method and summation method. Exemplarily, there are 3 intersections in the target second-level intersection set, which respectively predict 5 vehicles, 3 vehicles, and 4 vehicles. Using the simple summation method for fusion, then the number of arbitrary predicted vehicles is 5 + 3 + 4 = 12 vehicles. Organize and fuse the arbitrary predicted vehicle results obtained for each arbitrary first-level intersection to obtain the target predicted vehicle.

[0059] Furthermore, the second prediction module 40 in the embodiment of the present application is further configured to execute the following steps:

[0060] Descend the multiple second-level intersections in the second-level neighborhood based on the predicted number of vehicles for driving to obtain a descending list of intersections; take 50% of the intersections in the descending list of intersections to form the target second-level intersection set.

[0061] Specifically, in order to improve the accuracy of the target predicted vehicle and the accuracy of driving resource scheduling, use the predicted number of vehicles for driving as the sampling standard to sample the second-level intersections in the second-level neighborhood. The specific process includes: obtaining the predicted number of vehicles for driving of each second-level intersection through the driving visible graph, and arranging these second-level intersections according to these predicted numbers using a sorting algorithm (such as bubble sort, quick sort, etc.) to obtain a descending list of intersections.

[0062] Next, select 50% of the intersections in the descending list of intersections at the intersection to form the target secondary intersection set. For example, if there are 10 intersections in the list, then select the first 5 intersections. For cases where 50% of the quantity is not an integer, the method of rounding up is used for sampling. Through this sampling method, secondary intersections with relatively more predicted driving vehicles can be selected to more accurately analyze and fuse the predicted vehicles at any primary intersection, thereby obtaining the target predicted vehicles.

[0063] Further, the resource scheduling module 50 in the embodiment of the present application is further configured to perform the following steps:

[0064] Introduce the first service resource prediction and evaluation function to predict and evaluate the resource requirements of the target predicted vehicles, and obtain the first target predicted resource requirements; the resource scheduler performs resource scheduling on the target service area based on the target predicted energy requirements and target predicted parking requirements in the first target predicted resource requirements.

[0065] Specifically, the first service resource prediction and evaluation function is a mathematical or algorithmic model used to evaluate and predict the service resource requirements of vehicles. The first target predicted resource requirements are the results calculated by the first service resource prediction and evaluation function, which include the resource requirements of the target predicted vehicles in the target service area, including the first target predicted energy requirements and the first target predicted parking requirements. Among them, the first target predicted energy requirements are the predicted values of the energy required by the target predicted vehicles in the target service area (such as the fuel filling amount, charging amount, etc.). The first target predicted parking requirements are the demand predictions for the parking of the target predicted vehicles in the target service area (such as the number of parking spaces required, the duration distribution of parking, etc.).

[0066] First, introduce the first service resource prediction and evaluation function to predict and evaluate the resource requirements of the target predicted vehicles, and obtain the first target predicted resource requirements. In this process, the first service resource prediction and evaluation function processes the information related to the target predicted vehicles input. This information may include the type of vehicle (sedan, truck, etc.), quantity, expected arrival time, expected stay duration, etc. For example, if there are a large number of electric vehicles among the target predicted vehicles and they are expected to stay for a long time, the function will calculate the total charging amount requirement based on factors such as the battery capacity and charging efficiency of the vehicles, and at the same time calculate the number of parking space requirements based on the vehicle quantity and parking habits (such as whether a large parking space is required), thereby obtaining the first target predicted resource requirements.

[0067] Then, the resource scheduler performs resource scheduling for the target service area based on the target predicted energy demand and the target predicted parking demand in the first target predicted resource demand. For the target predicted energy demand, if the target service area is a comprehensive service area, the resource scheduler may check whether the refueling facilities (such as the number of fuel guns, refueling speed, etc.) and the charging facilities (such as the number of charging piles, charging power, etc.) can meet the demand. If the charging demand is large and the number of charging piles is insufficient, measures such as arranging to add temporary charging piles or adjusting the charging power may be taken. For the target predicted parking demand, the resource scheduler will check the existing number of parking spaces, layout, etc. of the parking lot. If the number of parking spaces is insufficient, it may consider opening up temporary parking areas or guiding vehicles to park reasonably to improve the utilization rate of parking spaces, etc., so as to achieve effective scheduling of resources in the target service area.

[0068] Further, the first real-time driving information described in the embodiments of the present application includes multiple vehicles with the identification of the number of passengers on board, and the resource scheduling module 50 is further used to perform the following steps:

[0069] Analyze the multiple vehicles with the identification of the number of passengers on board to obtain the second predicted number of passengers in the first unit vehicle displacement; obtain the first total number of passengers in the first predicted vehicle according to the second predicted number; sum up the first total number of passengers to obtain the target predicted number of passengers in the target predicted vehicle; introduce a second service resource prediction evaluation function to perform resource demand prediction evaluation on the target predicted number of passengers to obtain a second target predicted resource demand; the resource scheduler performs resource scheduling for the target service area based on the target predicted energy demand and the target predicted rest demand in the second target predicted resource demand.

[0070] Specifically, the second service resource prediction evaluation function is a function used to perform resource demand prediction evaluation on the target predicted number of passengers. The second target predicted resource demand is the result obtained through the second service resource prediction evaluation function, which includes the resource demand situation based on the target predicted number of passengers, including the target predicted energy demand and the target predicted rest demand, where the target predicted energy demand is the energy used by the passengers, such as the energy supply of food and beverages in the service area, and the target predicted rest demand is the demand for rest rooms, seats, etc.

[0071] First, read the number of passengers identification corresponding to each vehicle from the first real-time driving information, that is, the total number of people that each vehicle can carry. According to the aforementioned clustering analysis results, count the number of people carried by the vehicles with a unit displacement corresponding to each clustering cluster to obtain the corresponding second prediction number. Taking the vehicle with the first unit vehicle displacement as an example, read the number of passengers identification of all vehicles in the first clustering cluster, and calculate their sum to obtain the second prediction number of the first unit vehicle displacement. According to the clustering analysis results, sum up the second prediction numbers corresponding to the vehicles with all unit vehicle displacements to obtain the first total number of passengers, which is the total number of people that the first predicted vehicle can carry. Then, sum up the first total number of passengers to obtain the target predicted number of passengers of the target predicted vehicle.

[0072] After that, introduce a second service resource prediction and evaluation function to predict and evaluate the resource requirements for the target predicted number of passengers, and obtain the second target predicted resource requirements. For example, the energy requirement for catering supply may be proportional to the number of people. More passengers require more lounges and seats. According to the number of passengers and the corresponding demand ratio, the corresponding energy requirement and rest requirement can be calculated. Exemplarily, every 10 people require 1000 kcal of catering energy supply and 5 rest seats. Then, according to the target predicted number of passengers, the target predicted energy requirement and the target predicted rest requirement can be calculated, so as to obtain the second target predicted resource requirements.

[0073] The resource scheduler performs resource scheduling for the target service area based on the target predicted energy requirement and the target predicted rest requirement in the second target predicted resource requirements. For the target predicted energy requirement, if it is the catering energy supply, the resource scheduler will check the food reserves in the restaurants in the service area, the supply capacity of the cooking equipment, etc. If the food reserves are insufficient, replenishment may be arranged; if the supply capacity of the cooking equipment is limited, the menu may be adjusted or temporary cooking equipment may be added, etc. For the target predicted rest requirement, the resource scheduler will check the number of lounges and the number of seats in the service area. If the number of seats is insufficient, temporary seats may be considered to be added or passengers may be guided to use the rest facilities reasonably, etc., to meet the resource scheduling requirements of the target service area.

[0074] Furthermore, the resource scheduling module 50 described in the embodiment of the present application performs environmental resource scheduling for the target service area based on the target predicted vehicle and the target predicted number of passengers.

[0075] Specifically, the resource scheduling module 50 will also allocate and arrange the environmental related resources of the target service area according to the situation of the target predicted vehicle and the target predicted number of passengers. The environmental resources include but are not limited to the operation of the air purification equipment in the service area, temperature regulation (such as the use of air conditioners), environmental sanitation maintenance (such as the number of trash cans, the arrangement of cleaning personnel, etc.), and greening resources (such as the maintenance and layout adjustment of green plants, etc.).

[0076] First, consider the number of target predicted vehicles. If the number of target predicted vehicles is large, the vehicle exhaust emissions and the like may have a greater impact on the air quality of the service area. At this time, the resource scheduling module 50 will increase the operating power of the air purification equipment or increase the number of operating air purification equipment. For example, by analyzing and predicting that a large number of trucks (a part of the target predicted vehicles) will enter the service area, since the truck exhaust emissions are large, the resource scheduling module 50 can arrange in advance to increase the power of the air purification equipment from the normal 50% to 80% to ensure good air quality in the service area.

[0077] Secondly, analyze the target predicted number of passengers. If the target predicted number of passengers is large, the demand for temperature adjustment resources will increase. The resource scheduling module 50 needs to reasonably arrange the cooling or heating power of the air conditioner according to the number of people. For example, when it is predicted that a large number of passengers will come to the service area, if it is summer, the cooling power of the air conditioner may need to be increased from 30 kilowatts to 50 kilowatts to ensure a suitable temperature in the service area.

[0078] At the same time, the target predicted number of passengers will also affect the resource scheduling in environmental sanitation maintenance. More people mean more garbage generation. The resource scheduling module 50 may increase the number of trash cans and the shift arrangement of cleaning staff. For example, normally 1 trash can is placed per 50 square meters. When the predicted number of passengers increases, it may be adjusted to 1 trash can per 30 square meters; originally 2 cleaning staff are arranged to clean per hour, and it may be increased to 3 cleaning staff per hour.

[0079] In addition, for the greening resources in the service area, they will also be adjusted according to the target predicted vehicles and the target predicted number of passengers. If the number of vehicles and people is large, the maintenance of green plants needs to be strengthened to better absorb carbon dioxide, purify the air and beautify the environment. For example, increase the frequency of watering and fertilizing the green plants, or adjust the layout of the green plants, and place more green plants in areas with frequent personnel and vehicle flows to improve the overall environmental quality of the service area.

[0080] In summary, the driving resource optimization scheduling system combined with cloud support provided by the embodiments of the present application has the following technical effects:

[0081] The traffic road network matching module 10 matches the target traffic road network of the target area in the cloud database, identifies multiple intersections, clarifies the geographical scope of the target service area, provides a basic geographical information basis for subsequent driving data analysis, ensures that the analyzed traffic environment conforms to the actual situation, and improves the relevance of the data. The first prediction module 20 obtains the first predicted vehicles at the first intersection among the multiple intersections by analyzing the real-time driving information of the target traffic road network, introduces real-time traffic dynamic data, and provides an immediate feedback on the current traffic flow. The driving view generation module 30 obtains the target position of the target service area, and generates a driving view in combination with the first mapping relationship among the first intersection, the first position, and the first predicted vehicles, visually displaying the traffic conditions and flow trends. The second prediction module 40 activates the vehicle prediction middle platform to analyze the driving view, obtains the target predicted vehicles at the target position in a predetermined time interval, refines the prediction of the future traffic conditions in the target service area, and provides more targeted data for resource scheduling. The resource scheduling module 50 analyzes the target predicted vehicles through the resource scheduler, predicts the target predicted energy demand, target predicted parking demand, target predicted energy demand, and target predicted rest demand in combination with the service resource prediction and evaluation function, formulates a target resource scheduling strategy, and performs resource scheduling on the target service area according to the target resource scheduling strategy, optimizing the utilization of service area resources.

[0082] Overall, in the embodiments of the present application, through the collaborative work of each module, the traffic road network matching module 10 provides accurate basic data, the first prediction module 20 analyzes real-time traffic flow information, the driving view generation module 30 visualizes traffic conditions, the second prediction module 40 prospectively predicts future traffic flow, and the resource scheduling module 50 optimizes resource allocation according to the prediction results, achieving accurate prediction and optimized scheduling of driving resources, improving the flexibility and accuracy of driving resource scheduling, thus more effectively planning resource allocation, reducing vehicle waiting time, and ultimately improving the utilization rate of driving resources and the overall traffic flow efficiency.

[0083] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A driving resource optimization scheduling system with cloud support, characterized in that, Including: A traffic road network matching module, which is used to match the target traffic road network of the target area in the cloud database. The target traffic road network includes multiple intersections with location identifiers. Among them, the target area has a target service area; A first prediction module, which is used to analyze the real-time driving information of the target traffic road network to obtain the first predicted vehicle of the first intersection among the multiple intersections. The first intersection has an identifier of the first location; A driving view generation module, which is used to obtain the target location of the target service area and generate a driving view by combining the first intersection, the first location, and the first mapping relationship of the first predicted vehicle; A second prediction module, which is used to activate the vehicle prediction middle platform to analyze the driving view and obtain the target predicted vehicle at the target location in a predetermined time interval; A resource scheduling module, which is used to analyze the target predicted vehicle through a resource scheduler to obtain a target resource scheduling strategy, and perform resource scheduling on the target service area according to the target resource scheduling strategy; The execution steps of the second prediction module further include: The vehicle prediction middle platform determines the first-level neighborhood of the target location according to the driving view; Extract any first-level intersection in the first-level neighborhood and determine the second-level neighborhood of the any first-level intersection; Sample the second-level intersections in the second-level neighborhood to obtain a target set of second-level intersections; Fuse the predicted vehicles of each intersection in the target set of second-level intersections to obtain the arbitrary predicted vehicle of the any first-level intersection; Record the fusion result of the arbitrary predicted vehicle as the target predicted vehicle.

2. The vehicle travel resource optimization scheduling system combined with cloud support according to claim 1, wherein The execution steps of the first prediction module further include: Determine the first connecting road of the first intersection according to the target traffic road network. The first connecting road includes multiple roads; Obtain the first road among the multiple roads, and combine the target traffic road network to obtain the driving road direction of the first road with the first number of progressions; Combine the real-time driving information to obtain the first real-time driving information of the first road; Calculate the first predicted vehicle according to the first real-time driving information and the first number of progressions.

3. The vehicle travel resource optimized scheduling system combined with cloud support according to claim 2, characterized in that The first real-time driving information includes multiple vehicles with identifiers of unit vehicle displacement. The execution steps of the first prediction module further include: Perform clustering analysis on the multiple vehicles with the unit vehicle displacement as the clustering constraint to obtain a clustering result. The clustering result includes a first clustering cluster, and the first clustering cluster corresponds to the first unit vehicle displacement; Count the first clustering cluster to obtain the first number of clustering vehicles, and combine the first number of progressions to calculate the first predicted number of the vehicles with the first unit vehicle displacement; Sum the first predicted numbers to obtain the first predicted vehicle.

4. The cloud-supported driving resource optimization and scheduling system according to claim 1, wherein The execution steps of the second prediction module further include: Sort the multiple second-level intersections in the second-level neighborhood in descending order based on the predicted number of vehicles to obtain a list of sorted intersections; Select 50% of the intersections in the descending list of the intersections to form the target secondary intersection set.

5. The vehicle driving resource optimized scheduling system combined with cloud support according to claim 3, characterized in that The execution steps of the resource scheduling module further include: Introduce a first service resource prediction and evaluation function to predict and evaluate the resource requirements of the target predicted vehicles, and obtain the first target predicted resource requirements; The resource scheduler performs resource scheduling on the target service area based on the target predicted energy requirements and target predicted parking requirements in the first target predicted resource requirements.

6. The vehicle driving resource optimized scheduling system combined with cloud support according to claim 5, characterized in that, The first real-time driving information includes multiple vehicles with the identification of the number of passengers on board. The execution steps of the resource scheduling module further include: Analyze the multiple vehicles with the identification of the number of passengers on board to obtain the second predicted number of passengers in the first unit vehicle displacement; Obtain the first total number of passengers in the first predicted vehicle according to the second predicted number; Sum up the first total number of passengers to obtain the target predicted number of passengers in the target predicted vehicle; Introduce a second service resource prediction and evaluation function to predict and evaluate the resource requirements of the target predicted number of passengers, and obtain the second target predicted resource requirements; The resource scheduler performs resource scheduling on the target service area based on the target predicted energy requirements and target predicted rest requirements in the second target predicted resource requirements.

7. The vehicle driving resource optimized scheduling system combined with cloud support according to claim 6, wherein The resource scheduling module performs environmental resource scheduling on the target service area based on the target predicted vehicle and the target predicted number of passengers.

Images