WO2021115320A1

WO2021115320A1 - Traffic evaluation method and system

Info

Publication number: WO2021115320A1
Application number: PCT/CN2020/134868
Authority: WO
Inventors: 王宏伟; 孙伟力; 蒙元; 吴天龙
Original assignee: 北京嘀嘀无限科技发展有限公司
Priority date: 2019-12-11
Filing date: 2020-12-09
Publication date: 2021-06-17
Also published as: CN111815098A

Abstract

A traffic evaluation method used for a target road section, the method comprising: acquiring meteorological data of a target road section within a certain time period and historical traffic flow data within a historical time period prior to the time period (401); on the basis of the historical traffic flow data and the meteorological data, using a trained prediction model to determine predicted traffic flow data of the target road section within the time period (403); and on the basis of the predicted traffic flow data, determining at least one traffic evaluation parameter related to the target road section within the time period (405), wherein the traffic evaluation parameter is used for evaluating the traffic recovery capacity of the target road section under the meteorological data condition.

Description

Traffic evaluation method and system

Priority statement

This application claims the priority of the Chinese patent application with application number 201911267745.3 filed on December 11, 2019, which is incorporated herein by reference in its entirety.

Technical field

This application relates to the field of traffic technology, in particular to a method and system for evaluating traffic performance.

Background technique

The urban traffic management department hopes to know in advance the operation of urban road traffic in the city in which it is located. For example, before, during and after the disaster caused by extreme weather, urban road traffic operation conditions, and then make more suitable vehicles, personnel, and intersection signal lights deployment plan before the arrival of extreme weather. Therefore, it is necessary to propose a traffic performance evaluation program to evaluate the road traffic performance under various weather conditions.

Summary of the invention

Some embodiments of the application provide a system. The system includes at least one storage medium, and the storage medium includes instructions. And at least one processor, the at least one processor is in communication with the at least one storage medium, wherein, when the instruction is executed, the at least one processor is configured to perform the following operations. Obtain the meteorological data of the target road section in a certain time period and the historical traffic flow data in the historical time period before the time period; based on the historical traffic flow data and the meteorological data, use the trained prediction model to determine Predicting traffic flow data of the target road section in the time period; based on the predicted traffic flow data, determining at least one traffic evaluation parameter related to the target road section in the time period, and the traffic evaluation parameter is used for Evaluate the traffic recovery capability of the target road section under the weather data condition.

Some embodiments of the present application provide a method implemented on a computing device. The computing device has at least one storage medium, and the storage medium is used to store instructions. And at least one processor in communication with the at least one storage medium. The method includes the following operations. Obtain the meteorological data of the target road section in a certain time period and the historical traffic flow data in the historical time period before the time period; based on the historical traffic flow data and the meteorological data, use the trained prediction model to determine Predicting traffic flow data of the target road section in the time period; based on the predicted traffic flow data, determining at least one traffic evaluation parameter related to the target road section in the time period, and the traffic evaluation parameter is used for Evaluate the traffic recovery capability of the target road section under the weather data condition.

Some embodiments of the present application provide a computer-readable storage medium that stores computer instructions. After the computer reads the computer instructions in the storage medium, the computer executes the following method. The method includes: obtaining meteorological data of a target road section in a certain period of time and historical traffic flow data of a historical period of time before the period of time; based on the historical traffic flow data and the meteorological data, using training A good prediction model determines the predicted traffic flow data of the target road section in the time period; based on the predicted traffic flow data, determines at least one traffic evaluation parameter related to the target road section in the time period, the The traffic evaluation parameter is used to evaluate the traffic recovery capability of the target road section under the weather data condition.

Description of the drawings

This application will be further described in the form of exemplary embodiments, and these exemplary embodiments will be described in detail with the accompanying drawings. These embodiments are not restrictive. In these embodiments, the same number represents the same structure, in which:

Fig. 1 is a schematic diagram of an exemplary traffic evaluation system according to some embodiments of the present application;

Fig. 2 is a schematic diagram of exemplary hardware components and/or software components of an exemplary computing device according to some embodiments of the present application;

Fig. 3A is a block diagram of an exemplary processing device according to some embodiments of the present application;

Fig. 3B is a block diagram of another exemplary processing device according to some embodiments of the present application;

Fig. 4 is an exemplary flowchart of a method for determining traffic evaluation parameters according to some embodiments of the present application;

Fig. 5 is an exemplary flowchart of a method for determining traffic evaluation parameters according to some embodiments of the present application;

Fig. 6 is an exemplary flowchart of determining a prediction model according to some embodiments of the present application;

FIG. 7 is an exemplary graph for describing weather conditions corresponding to meteorological data according to some embodiments of the present application;

Fig. 8 is an exemplary schematic diagram of a traffic flow speed curve according to some embodiments of the present application;

Fig. 9 is an exemplary flow chart of a method for processing traffic information in extreme weather according to some embodiments of the present application;

FIG. 10 is an exemplary flow chart for obtaining a collection of weather data and a collection of traffic flow data in a predetermined area according to some embodiments of the present application;

FIG. 11 is an exemplary flowchart of determining traffic flow data at a predetermined time according to some embodiments of the present application;

FIG. 12 is an exemplary flowchart of determining the traffic recovery capability according to some embodiments of the present application;

FIG. 13 is a block diagram of a traffic information processing device for extreme weather according to some embodiments of the present application; and

Fig. 14 is a block diagram of an electronic device according to some embodiments of the present application.

Detailed ways

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings that need to be used in the description of the embodiments. Obviously, the drawings in the following description are only some examples or embodiments of the application. For those of ordinary skill in the art, without creative work, the application can be applied to the application according to these drawings. Other similar scenarios. Unless it is obvious from the language environment or otherwise stated, the same reference numerals in the figures represent the same structure or operation.

It should be understood that the “system”, “device”, “unit” and/or “module” used herein is a method for distinguishing different components, elements, parts, parts, or assemblies of different levels. However, if other words can achieve the same purpose, the words can be replaced by other expressions.

As shown in the present application and claims, unless the context clearly suggests exceptional circumstances, the words "a", "an", "an" and/or "the" do not specifically refer to the singular, but may also include the plural. Generally speaking, the terms "include" and "include" only suggest that the clearly identified steps and elements are included, and these steps and elements do not constitute an exclusive list, and the method or device may also include other steps or elements.

In this application, a flowchart is used to illustrate the operations performed by the system according to the embodiment of the application. It should be understood that the preceding or following operations are not necessarily performed exactly in order. Instead, the steps can be processed in reverse order or at the same time. At the same time, you can also add other operations to these processes, or remove a step or several operations from these processes.

The embodiments of the present application can be applied to different transportation systems, and different transportation systems include, but are not limited to, one or a combination of land, ocean, aviation, aerospace, and the like. For example, taxis, private cars, ride-hailing cars, buses, agent driving, trains, high-speed trains, ships, airplanes, hot air balloons, unmanned vehicles, collection/delivery express delivery, etc. which apply management and/or distribution transportation systems . The application scenarios of different embodiments of the present application include, but are not limited to, one or a combination of web pages, browser plug-ins, clients, customized systems, enterprise internal analysis systems, artificial intelligence robots, and the like. It should be understood that the application scenarios of the system and method of the present application are only some examples or embodiments of the present application. For those of ordinary skill in the art, they can also be based on these drawings without creative work. Apply this application to other similar scenarios. For example, other similar guidance users to park systems.

At present, for the evaluation of road traffic performance, methods based on network topology and simulation are mostly used in related technologies. The mathematical complexity is high and there are difficulties in engineering practice. Other solutions are based on data-driven methods to evaluate and predict the performance of the transportation network in extreme weather, but the methods are mainly focused on conventional statistical models, such as logistics regression. There are still few technologies for research on road traffic performance evaluation based on real data and machine learning models.

In the process of extreme weather in a certain area, many studies have modeled and analyzed the resilience of urban road traffic networks based on network and optimization methods. However, this type of research generally includes the researcher's subjective modeling emphasis and assumptions for certain network node attacks. There are some artificial judgments in modeling and restrictions on the type of system impact (such as terrorist attacks, etc.). The interference of extreme weather on the urban road traffic network is relatively random and has many influencing factors. The network and optimization methods are not particularly suitable for this kind of research problems.

In response to this situation, some studies have tried to use big data methods to assess the resilience of urban road traffic networks to extreme weather, so as to cover more types of impacts on the system. For example, it can target different types of extreme weather, while also reducing the influence of some anthropogenic elements in the research conclusions, making it more comprehensive and objective. This type of research uses statistical modeling methods of traffic big data (such as bus travel data and taxi trajectory data, etc.) to estimate the temporal and spatial changes of a city’s traffic recovery capacity under an extreme weather event, and obtain information about the recovery Some quantitative temporal and spatial distribution laws of capacity. However, this type of research still has some limitations. For example, because extreme weather parameters and traffic parameters are not directly related, the modeling of this complex nonlinear relationship is often less effective if the smooth curve fitting of general statistical models is used, which leads to recovery based on conventional statistical models (such as Logistics Regression) Capability research is often only able to carry out case analysis of traffic recovery capabilities in a single city and a single extreme weather event, and the generality of research methods and conclusions is relatively limited.

Therefore, based on the existing network and statistical research perspectives, it is often difficult to analyze different impact types and large sample recovery events, and it is also difficult to obtain a general summary of the recovery capability characteristics of urban road traffic networks under different types of extreme weather. Regular conclusions result in low accuracy.

The traffic performance evaluation method disclosed in the application is based on big data and deep learning models to predict the road traffic flow speed. And use the prediction result combined with the historical traffic flow speed to determine the comprehensive judgment of the road traffic performance, and accurate results can be obtained.

Fig. 1 is a schematic diagram of an exemplary traffic assessment system according to some embodiments of the present application. In some embodiments, the traffic evaluation system 100 may be used to evaluate the traffic performance of a certain road or multiple roads in a certain area under various weather conditions. For example, the traffic evaluation system 100 can be used to evaluate road traffic performance loss and performance recovery under extreme weather such as heavy rain, typhoon, and the like. In some embodiments, the traffic assessment system 100 can predict traffic flow data on the road when extreme weather occurs. For example, the traffic evaluation system 100 may determine the traffic flow data of the road when extreme weather occurs based on historical weather data and historical traffic flow data related to the road. And based on the predicted traffic flow data, the road traffic performance is evaluated.

As shown in FIG. 1, the traffic evaluation system 100 may include a processing device 110, a vehicle 120, a storage device 130, a network 140 and an information source 150. The processing device 110 may process data and/or information obtained from the vehicle 120, the storage device 130, and/or the information source 150. In some embodiments, the processing device 110 may process the acquired information and/or data to perform one or more functions described in this application. For example, the processing device 110 may obtain the driving speed of a plurality of vehicles 120 on a road that needs to be evaluated for traffic performance within a period of time, and weather data related to the road. And train a predictive model based on the acquired data. In the process of evaluating the traffic performance of the road, the processing device 110 uses the prediction model to determine the traffic performance of the road under predicted weather conditions. In some embodiments, the processing device 110 may be an independent server or a server group. The server group may be centralized or distributed (for example, the processing device 110 may be a distributed system). In some embodiments, the processing device 110 may be local or remote. For example, the processing device 110 may access and obtain information and/or data from the vehicle 120, the storage device 130, and/or the information source 150 through the network 140. In some embodiments, the processing device 110 may be directly connected to the vehicle 120, the storage device 130, and/or the information source 150 to obtain information and/or data. In some embodiments, the processing device 110 may be executed on a cloud platform. For example, the cloud platform may include one or any combination of private cloud, public cloud, hybrid cloud, community cloud, decentralized cloud, internal cloud, etc. In some embodiments, the processing device 110 may be integrated in the vehicle 120. In some embodiments, the processing device 110 may be implemented by the computing device 200 shown in FIG. 2.

In some embodiments, the processing device 110 may include one or more sub-processing devices (for example, a single-core processor or a multi-core processor). For example only, the processing device 110 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), an application specific instruction processor (ASIP), a graphics processing unit (GPU), a physical processor (PPU), a digital signal processor ( DSP), Field Programmable Gate Array (FPGA), Editable Logic Circuit (PLD), Controller, Microcontroller Unit, Reduced Instruction Set Computer (RISC), Microprocessor, etc. or any combination thereof.

The vehicle 120 may be a vehicle with the qualification/ability to drive on the road, including a motorcycle, a car, a coupe, a sports car, a pickup truck, a station wagon, a sports utility vehicle (SUV), a truck, a transfer vehicle, and the like. In some embodiments, the vehicle 120 may include an in-vehicle terminal. The vehicle-mounted terminal can obtain the position data and driving data generated by the vehicle 120 during the driving process, for example, positioning position, speed, acceleration, and so on. In some embodiments, relevant personnel of the vehicle 120 (for example, a driver and/or passengers) may be equipped with a user terminal, for example, a smart mobile device such as a smart phone. The user terminal may be equipped with various sensors, such as positioning devices, speed sensors, acceleration sensors, and so on. When the relevant personnel of the vehicle 120 (for example, the driver and/or passenger) carry the user terminal to drive or ride the vehicle 120, these sensors can obtain the location data and movement data of the user terminal as the location data of the vehicle 120 during the driving process. And driving data.

In some embodiments, the positioning function of the vehicle 120 and/or the user terminal can be implemented by multiple positioning technologies, for example, the global positioning system (GPS), the global satellite navigation system (GLONASS), and the Beidou navigation system (BDS) , Galileo positioning system (GALILEO), quasi-zenith satellite system (QZSS), wireless fidelity (Wi-Fi) positioning technology, etc. In some embodiments, the vehicle 120 may transmit the collected data/information via the network 140 to the processing device 110 for subsequent steps. The vehicle 120 may also store the collected data/information in its own memory, or transmit it to the storage device 130 via the network 140 for storage. In some embodiments, multiple vehicles included in the vehicle 120, such as vehicles 120-1, 120-2,..., 120-n, may collectively collect data. For example, multiple vehicles can collectively collect various data generated when driving on the same road.

The storage device 130 may store data and/or instructions. In some embodiments, the storage device 130 may store data/information acquired by the vehicle 120 and/or the processing device 110. For example, the storage device 130 may store driving data (for example, positioning positions, driving speeds, driving accelerations, etc.) generated when a plurality of vehicles are driving. In some embodiments, the storage device 130 may store data and/or instructions used by the processing device 110 to execute or use to complete the exemplary methods described in this application. In some embodiments, the storage device 130 may be part of the processing device 110 or the vehicle 120.

In some embodiments, the storage device 130 may include mass memory, removable memory, volatile read-write memory, read-only memory (ROM), etc., or any combination thereof. Exemplary mass storage devices may include magnetic disks, optical disks, solid state disks, and the like. Exemplary removable storage may include flash drives, floppy disks, optical disks, memory cards, compact disks, magnetic tapes, and the like. An exemplary volatile read-only memory may include random access memory (RAM). Exemplary RAM may include dynamic RAM (DRAM), double rate synchronous dynamic RAM (DDR SDRAM), static RAM (SRAM), thyristor RAM (T-RAM), zero capacitance RAM (Z-RAM), and the like. Exemplary ROMs may include mask ROM (MROM), programmable ROM (PROM), erasable programmable ROM (EPROM), electronically erasable programmable ROM (EEPROM), compact disk ROM (CD-ROM), and digital General-purpose disk ROM, etc. In some embodiments, the storage device 130 may be implemented on a cloud platform. For example only, the cloud platform may include private cloud, public cloud, hybrid cloud, community cloud, distributed cloud, internal cloud, multi-layer cloud, etc., or any combination thereof. In some embodiments, specific types of historical data may be uniformly stored on a cloud platform, so that multiple processing devices 110 or vehicles 120 can access or update, and ensure the real-time and cross-platform use of the data.

In some embodiments, the storage device 130 may be connected to the network 140 to communicate with one or more components in the traffic assessment system 100 (for example, the processing device 110, the vehicle 120, the information source 150). One or more components in the traffic evaluation system 100 can access data or instructions stored in the storage device 130 through the network 140. In some embodiments, the storage device 130 may directly connect or communicate with one or more components in the traffic assessment system 100 (for example, the processing device 110, the vehicle 120, and the information source 150).

The network 140 may facilitate the exchange of information and/or data. In some embodiments, one or more components in the traffic assessment system 100 (for example, the processing device 110, the vehicle 120, the storage device 130, and the information source 150) can be sent to/from other components in the traffic assessment system 100 via the network 140 And/or receive information and/or data. For example, the processing device 110 may obtain road-related traffic flow data and/or weather data from the vehicle 120, the storage device 130, and/or the information source 150 through the network 140. For another example, the processing device 110 may also obtain historical service data related to historical transportation services from the storage device 130 via the network 140.

In some embodiments, the network 140 may be any form of wired or wireless network or any combination thereof. For example only, the network 140 may include a cable network, a wired network, an optical fiber network, a telecommunication network, an internal network, the Internet, a local area network (LAN), a wide area network (WAN), a wireless local area network (WLAN), a metropolitan area network (MAN), public switched telephone network (PSTN), Bluetooth ^(TM) network, ZigBee ^TM network, near field communication (NFC) network, global system for mobile communications (GSM) network, Code division Multiple Access (CDMA) networks, Time division Multiple Access (TDMA) network , General Packet Radio Service (GPRS) network, Enhanced Data Rate GSM Evolution (EDGE) Network, Wideband Code Division Multiple Access (WCDMA) Network, High Speed Downlink Packet Access (HSDPA) Network, Long Term Evolution (LTE) Network, User Datagram Protocol (UDP) network, Transmission Control Protocol/Internet Protocol (TCP/IP) network, short message service (SMS) network, wireless application protocol (WAP) network, ultra-wideband (UWB) network, mobile communication (1G, 2G) , 3G, 4G, 5G) networks, Wi-Fi, Li-Fi, Narrowband Internet of Things (NB-IoT), etc. or any combination thereof. In some embodiments, the traffic assessment system 100 may include one or more network access points. For example, the traffic assessment system 100 may include wired or wireless network access points, such as base stations and/or wireless access points, through which one or more components of the traffic assessment system 100 may be connected to the network 140 to exchange data and/or information.

The information source 150 may be used to provide reference information for the traffic evaluation system 100. In some embodiments, the reference information may include weather information, traffic information, geographic information, legal and regulatory information, news events, life information, life guide information, and the like. The information source 150 may be implemented by a single central server, multiple servers connected to each other (for example, connected through a communication link), or multiple personal devices. For example, a personal device can generate content (e.g., text, voice, image, video) and upload it to a cloud server. Correspondingly, the information source 150 may be implemented by multiple personal devices and cloud servers. In some embodiments, the storage device 130, the processing device 110, and/or the vehicle 120 may also serve as information sources. For example, the real-time feedback speed and/or positioning information of the vehicle 120 may be used as reference information for other devices to obtain.

Those of ordinary skill in the art should understand that when the components of the traffic assessment system 100 are executed, the components may be executed by electrical signals and/or electromagnetic signals. For example, when the vehicle 120 processes a task, such as identifying, acquiring, or confirming an object, the vehicle 120 may operate a logic circuit in its processor to process the task. When the vehicle 120 sends driving data to the processing device 110, the processor of the vehicle 120 may generate an electrical signal that encodes the data. Then, the processor of the vehicle 120 may send the electrical signal to the output port. If the vehicle 120 communicates with the processing device 110 via a wired network, the output port may be physically connected to a cable, which further transmits electrical signals to the input port of the processing device 110. If the vehicle 120 communicates with the processing device 110 through a wireless network, the output port of the vehicle 120 may be one or more antennas, and the antennas may convert electrical signals into electromagnetic signals. In an electronic device, when the processor of the electronic device processes instructions, sends instructions, and/or performs actions, the instructions and/or actions are conducted via electrical signals. For example, when the processor retrieves or saves data from a storage medium (for example, the storage device 130), it can send an electrical signal to a read/write device of the storage medium, and the read/write device can read or write in the storage medium. Structured data. The structured data can be transmitted to the processor via the bus of the electronic device in the form of electrical signals. Here, the electrical signal may refer to one electrical signal, a series of electrical signals, and/or multiple discontinuous electrical signals.

Fig. 2 is a schematic diagram of exemplary hardware components and/or software components of an exemplary computing device according to some embodiments of the present application. In some embodiments, the processing device 110 may be implemented by the computing device 200. As shown in FIG. 2, the computing device 200 may include a processor 210, a memory 220, an input/output (I/O) 230, and a communication port 240.

The processor 210 may execute computer instructions (for example, program code) and may perform the functions of the processing device 110 according to the technology described in this application. Computer instructions can be used to perform specific functions described in this application, and computer instructions can include programs, objects, components, data structures, programs, modules, functions, and so on. In some embodiments, the processor 210 may include one or more hardware processors, such as a microcontroller, a microprocessor, a reduced instruction set computer (RISC), and an application specific integrated circuit (application specific integrated circuit). circuit (ASIC)), application-specific instruction-set processor (ASIP), central processing unit (CPU), graphics processing unit (GPU)) , Physical processing unit (physics processing unit (PPU)), digital signal processor (digital signal processor (DSP)), field programmable gate array (FPGA)), advanced RISC machine (advanced RISC machine( ARM)), programmable logic device (PLD), any circuit or processor capable of performing one or more functions, or a combination of several of them.

For illustration only, only one processor is described in the computing device 200. However, it should be noted that the computing device 200 may also include multiple processors. The operations and/or methods described in this application that are executed by one processor may also be executed jointly or separately by multiple processors. For example, if the processor of the computing device 200 described in this application performs operation A and operation B, it should be understood that operation A and operation B can also be shared by two or more different processors in the computing device 200. Or separately (for example, the first processor performs operation A and the second processor performs operation B, or the first processor and the second processor jointly perform operations A and B).

The memory 220 may store data/information acquired from the processing device 110, the vehicle 120, the storage device 130, and/or any other components of the traffic evaluation system 100. In some embodiments, the memory 220 may include mass memory, removable memory, volatile read-write memory, read-only memory (ROM), etc., or any combination thereof. Mass storage can include magnetic disks, optical disks, solid state drives, and mobile storage. Removable storage may include flash drives, floppy disks, optical disks, memory cards, ZIP disks, tapes, and so on. Volatile read-write memory may include random access memory (RAM). RAM can include dynamic random access memory (DRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), static random access memory (SRAM), thyristor random access memory (t-ram), zero capacitance random access memory Take memory (Z-RAM) and so on. ROM can include mask read-only memory (MROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), optical disk read-only memory Storage (CD-ROM), digital versatile disc, etc. In some embodiments, the memory 220 may store one or more programs and/or instructions for executing the exemplary methods described in this application.

The input/output 230 may input and/or output signals, data, information, and the like. In some embodiments, the input/output 230 may implement the interaction between the user and the processing device 110. In some embodiments, the input/output 230 may include an input device and an output device. The input device may include a keyboard, a mouse, a touch screen, a microphone, etc., or any combination thereof. The output device may include a display device, a speaker, a printer, a projector, etc. or any combination thereof. The display device may include a liquid crystal display (LCD), a light emitting diode (LED) display, a flat panel display, a curved screen, a television device, a cathode ray tube (CRT), a touch screen, etc., or any combination thereof.

The communication port 240 may be connected to a network (for example, the network 140) to facilitate data communication. The communication port 240 may establish a connection between the processing device 110 and the vehicle 120, the storage device 130, and/or the vehicle 120. The connection may be a wired connection, a wireless connection, any connection capable of data transmission and/or reception, etc., or any combination thereof. Wired connections can include cables, optical cables, telephone lines, etc., or any combination thereof. The wireless connection may include a Bluetooth ^™ link, a Wi-Fi ^™ link, a WiMAX ^™ link, a wireless local area network link, a ZigBee ^™ link, a mobile network link (for example, 3G, 4G, 5G, etc.), etc., or any combination thereof. In some embodiments, the communication port 240 may be and/or include a standardized communication port, such as RS232, RS485, and the like.

Fig. 3A is a block diagram of an exemplary processing device according to some embodiments of the present application. As shown in FIG. 3A, the processing device 110 may include an acquisition module 310, a traffic data flow determination module 320, a traffic evaluation module 330, and a storage module 340.

The obtaining module 310 may obtain the meteorological data of the target road section in a certain time period and the historical traffic flow data in the historical time period before the time period. The time period may be predetermined. For example, taking a specific time (current time) as a comparison, the time period may be 1 hour, 6 hours, 12 hours, 24 hours or other time periods after or before the current time. The meteorological data may include predicted meteorological data of the target road section in the time period. As an example, the predicted weather data may be determined based on a weather forecast model (for example, an open source WRF model (Weather Research and Forecasting Model)). The historical traffic flow data of the target road section in the historical time period before the time period may include the historical time period adjacent to the time period (may also be referred to as the first historical time period in this specification). The first historical traffic flow data of the target road segment and/or one or more historical time periods separated by one or more time periods from the time period (may also be referred to as the second historical time period in this specification) The second historical traffic flow data related to the target road section.

The traffic data flow determining module 320 may determine the predicted traffic flow data of the target road section in the time period by using a trained prediction model based on the historical traffic flow data and the meteorological data. For example, the traffic data flow determination module 320 may input the historical traffic flow data and the meteorological data into the trained prediction model to generate the predicted traffic flow data. In some embodiments, the traffic data flow determination module 320 may obtain one or more historical time periods adjacent to the time period before the target road section (may also be referred to as the third historical time period in this specification). ) Historical weather data. The traffic data flow determination module 320 may use the prediction model to process the historical weather data, the historical traffic flow data, and the weather data to generate the predicted traffic flow data. In some embodiments, the processing device 110 (for example, the traffic data flow determination module 320) may obtain the characteristic parameters of the target road section and determine the predicted traffic flow data based on the characteristic parameters of the target road section and the meteorological data.

In some embodiments, the prediction model may provide a mapping relationship between weather data and traffic flow data. The prediction model may include a trained deep learning model, for example, a sequence-to-sequence model. In some embodiments, the predictive model may be a sequence-to-sequence model based on attention. The traffic data flow determination module 320 may construct an input sequence adapted to the prediction model based on the historical traffic flow data and the meteorological data. And use the prediction model to process the input sequence to determine the predicted traffic flow data.

The traffic evaluation module 330 may determine at least one traffic evaluation parameter related to the target road section in the time period based on the predicted traffic flow data. The traffic evaluation parameters may be used to evaluate the traffic recovery capability of the target road section under the weather data conditions, and may include traffic performance loss parameters (Loss of Resilience, LOR), response time (Response Time, RST), and recovery time. (Recovery Time, RCT), Response Rate (Response Rate, RSR), Recovery Rate (Recovery Rate, RCR), etc., or combinations thereof. In some embodiments, the traffic evaluation module 330 may use reference traffic flow data, and determine traffic evaluation parameters based on the reference traffic flow data, predicted traffic flow data, and weather data. The reference traffic flow data may be the traffic flow data of the target road section under normal weather data conditions (for example, under fine weather conditions). A reference traffic flow speed curve can be obtained based on the reference traffic flow data. The traffic evaluation module 330 may determine the traffic evaluation parameter based on the reference traffic flow speed curve and the predicted traffic flow speed curve.

In some embodiments, the traffic evaluation module 330 may obtain reference traffic flow data related to the target road section, and determine one or more traffic flow data based on the meteorological data, the predicted traffic flow data, and the reference traffic flow data. Evaluate the moment and with one or more predicted traffic flow speeds and reference traffic flow speeds. The predicted traffic flow data may include a traffic flow velocity prediction curve related to the target road section in the time period. The reference traffic flow data may include a traffic flow speed reference curve related to the target road section in a reference time period having the same time identifier as the time period. The traffic evaluation module 330 may determine one or more predicted traffic flow speeds corresponding to the one or more evaluation moments based on the traffic flow speed prediction curve, and determine the one or more predicted traffic flow speeds corresponding to the one or more evaluation moments based on the traffic flow speed reference curve. Corresponding one or more reference traffic flow speeds. In some embodiments, the traffic evaluation module 330 may determine the at least one traffic evaluation parameter based on the predicted traffic flow speed, the reference traffic flow speed, and the one or more evaluation moments. In order to determine the traffic performance loss parameter LoR, the traffic evaluation module 330 may determine the first integral of the traffic flow speed prediction curve and the traffic flow speed reference curve within one or more evaluation time periods constituted by one or more evaluation moments The second integral. The traffic evaluation module 330 may determine the traffic performance loss parameter based on the first integral and the second integral. For example, the traffic evaluation module 330 may determine the difference between the first integral and the second integral, for example, use the difference obtained by subtracting the first integral from the second integral as the traffic performance loss parameter. In some embodiments, the traffic evaluation module 330 may determine one or more first differences between the one or more evaluation moments, and designate one of the one or more first differences as the Response time or said recovery time. In some embodiments, the traffic evaluation module 330 may determine one or more second differences between the predicted traffic flow speed and the reference traffic flow speed corresponding to the one or more evaluation moments, and determine the one One or more ratios between one or more second differences and the one or more first differences. The one or more ratios may be designated as the response rate or the recovery rate.

In some embodiments, the traffic assessment module 330 may obtain a traffic assessment model based on traffic flow speed and determine traffic assessment parameters based on predicted traffic flow data. The traffic assessment model can reflect the relationship between traffic flow speed and traffic assessment parameters. In some embodiments, the traffic evaluation model may include a trained machine learning model (for example, a trained neural network model).

The post-processing module 340 may obtain at least one reference traffic evaluation parameter of multiple road sections in multiple time periods and the predicted traffic evaluation parameter determined by using the prediction model. In some embodiments, the processing device 110 (for example, the post-processing module 340 may obtain real traffic flow data between the multiple time periods and the multiple road sections. For example, the post-processing module 340 may obtain multiple time periods). The actual traffic flow speeds of multiple target road sections at each time (for example, every thirty seconds) within each time period, and curve fitting is performed to obtain the true traffic flow speed curve of each road section in each time period. The post-processing module 340 also At least one traffic evaluation parameter and at least one reference traffic evaluation parameter related to the target road section and at least one reference traffic evaluation parameter and at least one reference traffic evaluation parameter related to other road sections within the target road section can be acquired for multiple time periods to determine the number of traffic evaluation parameters. Groups of reference traffic evaluation parameters and predicted traffic evaluation parameters. The post-processing module 340 may determine the accuracy of traffic evaluation based on multiple sets of reference traffic evaluation parameters and predicted traffic evaluation parameters. For example, based on multiple sets of reference traffic evaluation parameters and predicted traffic The evaluation parameters calculate the root mean square error (RMSE), mean absolute error (MAE), mean square error (MSE), R-Squared and other regression indicators to determine the accuracy of the traffic evaluation. In some embodiments, the post-processing module 340 may be based on The accuracy updates the traffic evaluation model, for example, updates the parameters of the traffic evaluation model.

In some embodiments, the post-processing module 340 may obtain traffic evaluation parameters of multiple road sections in the target area including the target road section. The target area may be an area where the target road section is located, for example, an administrative district or a city. The post-processing module 340 may use the same processing procedure as that of determining the traffic evaluation parameter of the target road section to determine the corresponding traffic evaluation parameter for each of the multiple road sections in the target area.

In some embodiments, the post-processing module 340 may map the traffic assessment parameters of the multiple road sections to the map data of the target area to obtain the visualized result of the traffic assessment of the target area. As an example, the post-processing module 340 may color the map data of the target area based on the traffic evaluation parameters of multiple road sections.

For other descriptions of the modules in FIG. 3A, reference may be made to other parts of this specification, for example, FIG. 4 to FIG. 5.

It should be understood that the system and its modules shown in FIG. 3A can be implemented in various ways. For example, in some embodiments, the system and its modules may be implemented by hardware, software, or a combination of software and hardware. Among them, the hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those skilled in the art can understand that the above-mentioned methods and systems can be implemented using computer-executable instructions and/or included in processor control codes, for example on a carrier medium such as a disk, CD or DVD-ROM, such as a read-only memory (firmware Such codes are provided on a programmable memory or a data carrier such as an optical or electronic signal carrier. The system and its modules of this application can not only be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc. It may also be implemented by software executed by various types of processors, or may be implemented by a combination of the above hardware circuit and software (for example, firmware).

Fig. 3B is a block diagram of an exemplary processing device according to some embodiments of the present application. As shown in FIG. 3A, the processing device 110 may include a sample acquisition module 350 and a training module 360.

The sample acquisition module 350 can acquire multiple training samples. Each training sample may be determined based on reference weather data, predicted weather data, and reference traffic flow data in a certain period of time with at least one training road section. Each training sample may also be determined based on reference traffic flow data and/or reference weather data in a time period before the time period. The time period represents the time period before the current moment (that is, the historical time period). The reference meteorological data may be the meteorological data of the at least one training road section actually monitored in the time period (for example, within one year or two years). The predicted meteorological data may be meteorological data in the time period obtained through prediction based on the actually monitored meteorological data in the time period before the time period. The reference traffic flow data may be actual traffic flow data of at least one training road segment in the time period. The reference traffic flow data may include actual traffic flow speed. The reference traffic flow data in the time period before the time period may include the traffic flow data actually monitored in the time period before the time period. The reference meteorological data in the time period before the time period may include the meteorological data actually monitored in the time period before the time period.

The training module 360 may train the initial machine learning model by using one or more iterative processes based on the multiple training samples to obtain a trained prediction model. The initial machine model may include a deep learning model. For example, sequence-to-sequence model. In some embodiments, the prediction model may be a sequence-to-sequence model based on attention. The attention-based sequence-to-sequence model may be based on an encoder and decoder architecture. In some embodiments, for each training sample, the training module 360 may input the training input sequence to the prediction model for processing to obtain the corresponding prediction result. The prediction result may be predicted traffic flow data, for example, predicted traffic flow speed. The training module 360 may compare the difference between the prediction result and the training label corresponding to the training sample. The difference may indicate the prediction effect of the prediction model. The greater the difference, the worse the prediction effect of the model. The training module 360 may reversely adjust the parameters of the prediction model based on the difference to reduce the difference. For example, adjusting the hyperparameters of the prediction model, such as the learning rate. After completing one or more iterations, if a preset condition is met, for example, the number of iteration rounds or the number of training reaches the preset number, or the difference is less than the preset threshold, the training can be stopped. The obtained trained machine learning model can be designated as the trained prediction model.

For other descriptions of the modules in FIG. 3B, reference may be made to other parts of this specification, for example, FIG. 6.

It should be understood that the system and its modules shown in FIG. 3B can be implemented in various ways. For example, in some embodiments, the system and its modules can be implemented by hardware, software, or a combination of software and hardware. Among them, the hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those skilled in the art can understand that the above-mentioned methods and systems can be implemented using computer-executable instructions and/or included in processor control codes, for example on a carrier medium such as a disk, CD or DVD-ROM, such as a read-only memory (firmware Such codes are provided on a programmable memory or a data carrier such as an optical or electronic signal carrier. The system and its modules of this application can not only be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc. It may also be implemented by software executed by various types of processors, or may be implemented by a combination of the above hardware circuit and software (for example, firmware).

Fig. 4 is an exemplary flowchart of a method for determining traffic evaluation parameters according to some embodiments of the present application. In some embodiments, the process 400 may be executed by a processing device (for example, the processing device 110 shown in FIG. 1). For example, the process 400 may be stored in a storage device (for example, the storage device 130 or a storage unit of a processing device) in the form of a program or instruction. When the processor 210 or the module shown in FIG. 3 executes the program or instruction, the process may be implemented. 400. In some embodiments, the process 400 may utilize one or more additional operations not described below, and/or not be completed by one or more operations discussed below. In addition, the order of operations shown in FIG. 4 is not restrictive. As shown in FIG. 4, the process 400 may include the following operations.

Step 401: Obtain meteorological data of a target road section in a certain time period and historical traffic flow data in a historical time period before the time period. In some embodiments, step 410 may be performed by the obtaining module 310.

It can be known that a road has two or more nodes. For example, the two or more nodes may include the start point, the end point of the road, and the intersection of the road and other roads. A part of the road between two intersections can be called a link. For example, a part of the road between two adjacent intersections, or a part of the road between two intersections with one or more intersections in between. When a certain road section requires traffic performance prediction and/or evaluation, for example, to determine the traffic recovery capability of the road section in extreme weather such as heavy rain, the road section may be referred to as a target road section. In some embodiments, the target road section may be given an identification (ID) to distinguish it from other road sections. Assuming that the target road section belongs to a road network in a certain area, when all road sections in the road network are numbered, the target road section may also be given a number, for example, a number, a letter, etc. or a combination thereof. The code may be the identification. In some embodiments, the identifier of the target road section may be expressed in coordinates, such as latitude and longitude coordinates. For example, the latitude and longitude coordinates corresponding to the geometric center of the target road section may be used to represent the identifier of the target road section. In some embodiments, the time period may be predetermined. For example, a specific time is used as a control, for example, the current time is used as a control, and the time period may be 1 hour, 6 hours, 12 hours, 24 hours, or other time periods after the current time. For another example, a specific time is used as a control, for example, the current time is used as a control. The time period can be 1 hour, 6 hours, 12 hours, or 24 hours before the current time or any time period before the current time. The time period can be adjusted according to different needs. This manual does not make specific restrictions.

In some embodiments, the meteorological data may include predicted meteorological data of the target road section within the time period. The predicted weather data may be related to a certain type of weather, for example, strong winds (typhoons, hurricanes, etc.), precipitation (for example, rain, drizzle, showers), haze (for example, light fog, heavy fog, dense fog) , Snowfall, sandstorm, thunder and lightning, etc. As an example, the predicted weather data may be determined based on a weather prediction model. The weather prediction model may include an open source WRF model (Weather Research and Forecasting Model). The WRF model can use the meteorological monitoring data of a designated geographic area in combination with a meteorological physical and chemical model to generate forecasted meteorological grid data with a predetermined grid density corresponding to the designated geographic area. The predetermined grid density may be a preset resolution when performing grid division for the designated area, and may be expressed by area. For example, 1km×1km, 3km×3km, 5km×5km, etc. Among them, 5km×5km can indicate that the designated area is divided into multiple 5km×5km grids. The predicted weather grid data may refer to the predicted weather data of the area corresponding to the divided grid. After the WRF model is used to determine the predicted weather grid data for a certain area containing the target road section, the processing device 110 can determine which grid the coordinates of the target road section belong to, and then calculate the predicted weather grid corresponding to the grid The data is determined as the meteorological data of the target road section. In some embodiments, the meteorological data may include the values of meteorological parameters such as temperature, humidity, air pressure, precipitation, wind speed, wind direction, illuminance, radiation intensity, hours of sunshine, or any combination thereof within the time period. In some embodiments, the meteorological data may include changes in meteorological parameters over time within the time period. For example, the meteorological data (such as precipitation) includes a change curve of meteorological parameters over time.

In some embodiments, the historical traffic flow data of the target road segment in the historical time period before the time period may include a historical time period adjacent to the time period (may also be referred to as the first historical time period in this specification). Time period) within the first historical traffic flow data of the target road section and/or one or more historical time periods separated by one or more time periods from the time period (may also be referred to as the second historical time period in this specification) Time period) second historical traffic flow data related to the target road section. Continuing to refer to the foregoing example, taking the current time as a control, the time period may be 24 hours after the current time, and the first historical time period may be 24 hours or more before the current time. The time period can be one day, one week, one month, one year, etc. For example, the second historical time period may be 24 hours after one week from the current time or 24 hours after one day. For example, assuming that the current time is 24:00 on Wednesday evening, and the time period that needs to be predicted is from 0:00 to 24:00 on Thursday, the second historical time period may be the 24 hours and the time period between 0:00 and 24:00 on Thursday last week. / Or Wednesday from 0 o'clock to 24 o'clock. In some embodiments, the historical traffic flow data (including the first historical traffic flow data and the second traffic flow data) may include at least historical traffic flow speeds. The processing device 110 may obtain from a database (for example, the storage device 130) the historical travel speeds of a plurality of vehicles passing the target road segment in the first historical time period and the second historical time period, and combine these The average value of the historical travel speed is used as the historical traffic flow speed.

Step 403: Based on the historical traffic flow data and the meteorological data, the trained prediction model is used to determine the predicted traffic flow data of the target road section in the time period. In some embodiments, step 403 may be performed by the traffic data flow determination module 320. For example, the processing device 110 may input the historical traffic flow data and the weather data into the trained prediction model to generate the predicted traffic flow data.

In some embodiments, the traffic data flow determination module 320 may obtain one or more historical time periods adjacent to the time period before the target road section (may also be referred to as the third historical time period in this specification). ) Historical weather data. Similarly, continue to refer to the example in step 401, and take the current time as a control. The time period may be 24 hours after the current time, and the third historical time period may be 24 hours or 48 hours before the current time. , 72 hours or other time period. The historical weather data may be historical weather grid data including the target road section. In some embodiments, the historical weather grid data of the target road section may be obtained according to the WRF model. When generating historical weather grid data, the resolution of the WRF model may be the same as the resolution when generating the predicted weather grid data. Similarly, the historical meteorological data may include temperature, humidity, air pressure, precipitation, wind speed, wind direction, illuminance, radiation intensity, sunshine hours, etc., or any combination thereof. In some second embodiments, actual weather data of the target road section in the third historical time period can be obtained and designated as historical weather data. The actual meteorological data is collected by a meteorological monitoring agency and can be obtained from an associated database.

In some embodiments, the traffic data flow determination module 320 may use the prediction model to determine the predicted traffic flow data based on the historical weather data, the historical traffic flow data, and the weather data. For example, the traffic data flow determination module 320 may use the prediction model to process the historical weather data, the historical traffic flow data, and the weather data to generate the predicted traffic flow data.

In some embodiments, the processing device 110 (for example, the traffic data flow determination module 320) may obtain the characteristic parameters of the target road section and determine the predicted traffic flow data based on the characteristic parameters of the target road section and the meteorological data. For example, the processing device 110 (for example, the traffic data flow determination module 320) may input the characteristic parameters of the target road section and the meteorological data into the prediction model to generate the predicted traffic flow data. For another example, the processing device may input the characteristic parameters of the target road section, the historical traffic flow data, the weather data, and the historical weather data into the prediction model to generate the predicted traffic flow data. The characteristic parameters of the target road section may include the target road section level, the target road section type, the target road section width, the target road section length, the number of lanes of the target road section, the area where the target road section is located, and the like.

The prediction model can provide a mapping relationship from weather data to traffic flow data. The processing device 110 (for example, the traffic data flow determining module 320) may predict traffic flow data based on the mapping relationship and the input data (for example, weather data, historical traffic flow data, and/or historical weather data). In some embodiments, the prediction model may include a trained deep learning model, for example, a sequence-to-sequence model. In some embodiments, the prediction model may be a sequence-to-sequence model based on attention. The attention-based sequence-to-sequence model may be constructed based on an encoder and a decoder. The encoder may include an encoder based on long short-term memory (LSTM), for example, LSTM, Bi-LSTM, CNN-LSTM, Conv-LSTM, TGC-LSTM, etc. The decoder may include a decoder based on a gated recurrent unit (Gated Recurrent Unit, GRU), for example, GRU, Bi-GRU, DCGRU, and so on. Exemplarily, the prediction model may use a deep Bi-LSTM as an encoder and a deep GRU as a decoder. The encoder and decoder respectively contain 3 layers of neural networks, each layer of neural network contains 1379 hidden nodes, and SoftSign is used as the activation function. Regarding the training process of the prediction model, you can refer to other parts of this specification, for example, Figure 6, which will not be repeated here.

In some embodiments, the traffic data flow determination module 320 may construct an input sequence adapted to the prediction model based on the historical traffic flow data and the meteorological data. The format of an exemplary input sequence can be as follows: (Days*ID, Steps*Features). Among them, Days represents the number of days, including the first historical time period and the second historical time period, and the dimension is two-dimensional; ID represents the target road section, and the dimension is one-dimensional; Steps represents the time step, assuming that one hour is set Is a time step, then the time step of a time period of 24 hours is 24, and the dimension is determined according to the step length and the length of the time period; Features represents features, which can include meteorological data features and historical traffic flow speeds, The dimension is determined based on the total number of meteorological data and the total number of historical traffic flow speeds.

In some embodiments, the traffic data flow determination module 320 may use the prediction model to process the input sequence to determine the predicted traffic flow data. For example, the traffic data flow determination module 320 may input the input sequence to the prediction model to obtain the predicted traffic flow data. In some embodiments, the predicted traffic flow data may include a traffic flow velocity prediction curve related to the target road section within the time period. For example, the traffic data flow determination module 320 may use the prediction model to obtain the predicted traffic flow speed of the target road section at each time (for example, every thirty seconds) within the time period, and perform curve fitting to obtain the Traffic flow speed prediction curve. For another example, the traffic data flow determination module 320 may use the prediction model to obtain and directly output the traffic flow speed prediction curve of the target road section in the time period. Since the composition of the input sequence uses meteorological data (including predicted meteorological data and historical meteorological data), the traffic flow velocity prediction curve can reflect to a certain extent that the target road section is under different weather conditions (by the predicted meteorological data). And/or weather conditions determined by historical weather data) traffic flow speed changes, for example, the target road section when there is a heavy rain, the traffic flow speed drops and rises.

Step 405: Based on the predicted traffic flow data, determine at least one traffic evaluation parameter related to the target road section within the time period. In some embodiments, step 405 may be performed by the traffic evaluation module 330.

In some embodiments, the traffic evaluation parameter may be used to evaluate the traffic recovery capability of the target road section under the weather data condition. It can be understood that when the weather data conditions are extreme or severe weather such as typhoons and rainstorms, the traffic performance of the target road section will definitely be affected. Traffic performance can be embodied by traffic flow speed, traffic flow, etc. For example, when the meteorological data conditions are severe weather such as typhoons and heavy rains, the traffic performance is degraded, which is manifested as a decrease in the speed of traffic flow and a small traffic volume. Throughout the entire meteorological process, the traffic performance of the target road section experiences a process of falling to rising and returning to a normal level. The traffic evaluation parameters can be used to express the traffic recovery capability of this process. The traffic evaluation parameters may include traffic performance loss parameters (Loss of Resilience, LOR), response time (Response Time, RST), recovery time (Recovery Time, RCT), response rate (Response Rate, RSR), and recovery rate (Recovery Rate, RCR), etc. or a combination thereof. The traffic performance loss parameter may be used to indicate the traffic performance loss of the target road section. For example, under heavy precipitation weather, from the beginning of the precipitation to the end of the precipitation, the traffic flow speed of the target road section drops from the normal value to the lowest value and then returns to the normal value (for example, the traffic flow speed on a non-precipitation day). Performance loss, for example, a decrease in the flow of passing vehicles. The response time may be used to indicate the length of time that the traffic flow speed of the target road section drops from a normal value to a minimum value. For example, in a heavy rainfall weather, starting from precipitation, the traffic flow speed of the target road section drops from a normal value to a minimum value. The recovery time is used to indicate the length of time for the traffic flow speed of the target road section to recover from the lowest value to the normal value. For example, under heavy rainfall, the time period for the traffic flow velocity of the target road section to recover from the lowest value to the normal value at the end of the heavy rainfall. The response rate may be used to indicate the rate at which the traffic flow speed of the target road section drops from a normal value to a minimum value. For example, starting from precipitation under heavy rainfall weather, the ratio between the decreasing value of the traffic flow speed of the target road section from the normal value to the lowest value and the duration. The recovery rate is used to indicate the rate at which the traffic flow speed of the target road section recovers from the lowest value to the normal value. For example, under heavy rainfall, the traffic flow velocity of the target road section is restored from the lowest value to the normal value after the end of the heavy rainfall. The ratio between the restored value and the duration. The traffic evaluation parameters can not only indicate the size of the traffic performance loss of the target road section under the conditions of the meteorological data, but also the response time and response rate (for example, loss time and loss rate) and recovery time/rate. To evaluate the traffic recovery capability of the target road section under the weather data condition.

In some embodiments, the traffic evaluation module 330 may use reference traffic flow data, and determine traffic evaluation parameters based on the reference traffic flow data, predicted traffic flow data, and weather data. The reference traffic flow data may be the traffic flow data of the target road section under normal weather data conditions (for example, under fine weather conditions). A reference traffic flow speed curve can be obtained based on the reference traffic flow data. The traffic evaluation module 330 may determine the traffic evaluation parameter based on the reference traffic flow speed curve and the predicted traffic flow speed curve. For the description of determining the traffic evaluation parameters, reference may be made to other parts of this specification, for example, FIG. 5, which is not repeated here.

In some embodiments, the traffic assessment module 330 may obtain a traffic assessment model based on traffic flow speed and determine traffic assessment parameters based on predicted traffic flow data. The traffic assessment model can reflect the relationship between traffic flow speed and traffic assessment parameters. In some embodiments, the traffic evaluation model may include a trained machine learning model (for example, a trained neural network model). For example, a processing device that is the same as or different from the processing device 110 may acquire multiple samples, and each sample includes traffic flow data and evaluation parameters. The processing device can use the samples to obtain a traffic evaluation model by training a neural network. In some embodiments, the traffic evaluation model may include a trained regression model. The processing device that is the same as or different from the processing device 110 can obtain multiple samples, and each sample includes traffic flow data and evaluation parameters. The processing device may use the sample to perform fitting through a regression model (for example, linear regression, logistic regression, or polynomial regression) to determine a traffic evaluation model. The training on the traffic evaluation model can be the same or similar to the training of the predictive model. For more information about the training of the traffic evaluation model, please refer to the description in Figure 6.

In some embodiments, the processing device 110 (for example, the post-processing module 340) may also obtain at least one reference traffic assessment parameter of multiple road segments in multiple time periods and the predicted traffic assessment parameter determined by using the prediction model. In some embodiments, the processing device 110 (for example, the post-processing module 340) may obtain real traffic flow data of the multiple time periods and the multiple road segments. For example, the processing device 110 (for example, the post-processing module 340) may obtain the actual traffic flow speeds of multiple target road sections at various times (for example, every thirty seconds) in multiple time periods, and perform curve fitting to obtain each The real curve of the traffic flow speed of each road segment in each time period. The processing device 110 (for example, the post-processing module 340) may determine the at least one reference traffic evaluation parameter by using the same and/or similar method as that of determining the traffic evaluation parameter. For example, reference traffic performance loss parameters, reference response time, reference recovery time, reference response rate, and reference recovery rate. In some embodiments, the processing device 110 (for example, the post-processing module 340) may also acquire at least one traffic evaluation parameter and at least one reference traffic evaluation parameter related to the target road section within the target road section for multiple time periods. At least one traffic evaluation parameter and at least one reference traffic evaluation parameter related to the road section are used to determine multiple sets of reference traffic evaluation parameters and predicted traffic evaluation parameters. The processing device 110 (for example, the post-processing module 340) may determine the accuracy of the traffic assessment based on multiple sets of reference traffic assessment parameters and predicted traffic assessment parameters. For example, calculate the root mean square error (RMSE), mean absolute error (MAE), mean square error (MSE), R-Squared and other regression indicators based on multiple sets of reference traffic evaluation parameters and predicted traffic evaluation parameters to determine the accuracy of traffic evaluation . In some embodiments, the processing device 110 (for example, the post-processing module 340) may update the traffic evaluation model based on the accuracy. For example, update the parameters of the traffic assessment model.

In some embodiments, the processing device 110 (for example, the post-processing module 340) may obtain traffic evaluation parameters of multiple road sections in the target area including the target road section. The target area may be an area where the target road section is located, for example, an administrative district or a city. The processing device 110 (for example, the post-processing module 340) may use the same processing procedure as determining the traffic evaluation parameter of the target road section to determine the corresponding traffic evaluation parameter for each of the multiple road sections in the target area.

In some embodiments, the processing device 110 (for example, the post-processing module 340) may map the traffic evaluation parameters of the multiple road sections to the map data of the target area to obtain the visualization result of the traffic evaluation of the target area . As an example, the processing device 110 (for example, the post-processing module 340) may color the map data of the target area based on the traffic evaluation parameters of multiple road sections. For example, the processing device 110 (e.g., the post-processing module 340) may determine a color contrast. In the descending order of the visible light wavelengths, the order of the traffic evaluation parameters is ascending. In other words, when coloring the map data of the target area, the processing device 110 (for example, the post-processing module 340) may use the color corresponding to the traffic evaluation parameter. In this way, after the processing is completed, a color image corresponding to the target area can be obtained. Different colors correspond to different traffic evaluation parameters. The color map corresponding to the target area may be referred to as the visualization result of the traffic assessment of the target area.

It should be noted that the foregoing description of the process 400 is only for example and description, and does not limit the scope of application of the present application. For those skilled in the art, various modifications and changes can be made to the process 400 under the guidance of this application. However, these amendments and changes are still within the scope of this application.

Fig. 5 is an exemplary flowchart for determining at least one traffic evaluation parameter according to some embodiments of the present application. In some embodiments, the process 500 may be executed by a processing device (for example, the processing device 110 shown in FIG. 1). For example, the process 500 may be stored in a storage device (for example, the storage device 130 or a storage unit of a processing device) in the form of a program or instruction. When the processor 210 or the module shown in FIG. 3 executes the program or instruction, the process may be implemented. 500. In some embodiments, the process 500 may be executed by the traffic evaluation module 330. In some embodiments, the process 500 may utilize one or more additional operations not described below, and/or not be completed by one or more operations discussed below. In addition, the order of operations shown in FIG. 5 is not restrictive. As shown in FIG. 5, the process 500 may include the following operations.

Step 501: Obtain reference traffic flow data related to the target road section.

In some embodiments, the reference traffic flow data may be predetermined. The processing device 110 (or the traffic evaluation module 330) can obtain the historical traffic flow in a historical time period (which may be referred to as the fourth historical time period in this specification) before the time period during which traffic evaluation of the target road section is required. data. The fourth historical time period may be one month, one quarter, six months, one year, etc. The processing device 110 (or the traffic evaluation module 330) can filter these historical traffic flow data to obtain the traffic flow data of the days when the weather condition is good (clear or rainless, etc.) (also referred to as reference traffic flow in this application). data). The processing device 110 (or the traffic evaluation module 330) may determine the reference traffic flow data based on the reference traffic flow data. For example, the processing device 110 may perform averaging processing on the reference traffic flow data in multiple days to determine the reference traffic flow data.

In some embodiments, if the time period during which the traffic evaluation needs to be performed on the target road section is a certain day or a certain period of time in a certain day, the processing device 110 (or the traffic evaluation module 330) may determine whether these dates are in a week. The time location divides these reference traffic flow data into seven batches. For example, the reference traffic flow data of multiple dates that belong to the same Monday belong to the same batch, the reference traffic flow data of multiple dates that belong to the same Tuesday belong to the same batch, and the reference traffic flow data of multiple dates that belong to the same Wednesday belong to the same batch. The data belong to the same batch, and so on. After the batches are divided, the processing device 110 (or the traffic evaluation module 330) can perform averaging processing for the reference traffic flow data of each batch. For example, determine the average value of traffic flow speed per hour. In this way, the average processing result of the reference traffic flow data of each batch can be used as a reference traffic flow data. For example, the reference traffic flow data may include a traffic flow speed curve in a day. In some embodiments, the processing device 110 may determine the time position of the time period in a week, and based on the time position, determine that the average processing result of the reference traffic flow data of the corresponding batch may be used as the target road section. Related benchmark traffic flow data. For example, assuming that the time period belongs to Thursday, the average processing result of a batch of reference traffic flow data corresponding to Thursday may be used as the reference traffic flow data related to the target road section. It should be noted that the division of the reference traffic flow data into seven batches according to the time position of the dates in a week is only an exemplary description, and does not limit the scope of the present invention. For example, the reference traffic flow data can also be divided into multiple batches according to the time position of these dates in a month or a quarter to determine the reference traffic flow data.

Step 503: Based on the meteorological data, the predicted traffic flow data, and the reference traffic flow data, determine one or more evaluation moments and one or more predicted traffic flow speeds and reference traffic flow speeds.

In some embodiments, the meteorological data may include changes in meteorological parameters (for example, precipitation) over time within the time period. The processing device 110 (or the traffic evaluation module 330) may determine, based on the weather data, one or more critical moments of changes in the weather parameters corresponding to the weather data. In some embodiments, the critical moment of the change of the meteorological parameter may include the start moment, the end moment, the moment of the greatest change, the moment of the most severe change (ie the moment of the fastest change), etc. of the change of the meteorological parameter. For example, for rainy weather, the processing device 110 (or the traffic evaluation module 330) may determine the start time of the precipitation, the time of the precipitation peak, and the end of the precipitation. The period of time between the beginning of precipitation and the end of precipitation may be the duration of the change of the meteorological parameter. It can be understood that within the duration corresponding to the precipitation, the traffic flow data (for example, the traffic flow speed) of the target road section will change. For example, the traffic flow speed will begin to decrease after the precipitation begins, and may recover upwards after falling to the lowest point or continue at the lowest point until the precipitation ends. In some embodiments, the predicted traffic flow data and the reference traffic flow data may include changes in traffic flow speed over time during the time period. The processing device 110 (or the traffic evaluation module 330) may determine one or more key moments of the traffic flow speed change based on the predicted traffic flow data and the reference traffic flow data. In some embodiments, the critical moments of the traffic flow speed change may include the start moment, the end moment, the moment when the traffic flow speed is the smallest, the moment when the traffic flow speed is the largest, and the moment when the traffic flow speed changes most drastically (that is, the moment when the traffic flow speed changes). The fastest moment) and so on. The processing device 110 (or the traffic assessment module 330) may use these critical moments of changes in the traffic flow speed as the one or more assessment moments.

In some embodiments, the predicted traffic flow data may include a traffic flow velocity prediction curve related to the target road section within the time period. The predicted traffic flow data may be determined based on the process 400 (for example, step 403). The reference traffic flow data may include a traffic flow speed reference curve related to the target road section in a reference time period having the same time identifier as the time period. As described in step 501, the time identifier may be used to indicate the time position of the time period in a time period such as a week, a month, a day, or a year, for example, the day of the week. The duration of the reference time period may be the same as the duration of the time period, and the start time and end time are the same. For example, assuming that the time period is from 9:00 to 18:00 on Friday, the reference traffic flow data includes the average processing results from 9:00 to 18:00 of the reference traffic flow data corresponding to a batch of Friday, for example, 9 The traffic flow curve corresponding to 18:00 to 18:00. In some embodiments, the processing device 110 (or the traffic assessment module 330) may determine one or more predicted traffic flow speeds corresponding to the one or more assessment moments based on the traffic flow speed prediction curve, and based on the traffic flow speed The reference curve determines one or more reference traffic flow speeds corresponding to the one or more evaluation moments. For example, the processing device 110 (or the traffic evaluation module 330) can determine the traffic flow speed corresponding to each evaluation time on the speed flow curve.

Referring to FIG. 7, FIG. 7 is an exemplary graph for describing weather conditions corresponding to weather data according to some embodiments of the present application. As shown in Fig. 7, the time period is 24 hours, taking precipitation weather as an example. t ₀ represents the beginning of precipitation; t ₁ represents the moment of precipitation peak; t ₂ represents the end of precipitation; T represents _{the observation time window of the precipitation weather between t 0} and t ₂ , which is the duration of the entire precipitation weather. Referring to FIG. 8, FIG. 8 is an exemplary schematic diagram of a traffic flow speed curve according to some embodiments of the present application. As shown in Figure 8, the darker gray curve L1 represents the traffic flow speed reference curve, and the lighter gray curve L2 represents the traffic flow speed prediction curve. m ₀ represents the traffic flow speed at time t′ ₀ , the predicted traffic flow speed at this moment is the same as the reference traffic flow speed, t′ ₀ is the same as t ₀ ; m ₁ represents the lowest predicted traffic flow speed within the observation time window T; t′ ₁ represents the time corresponding to m ₁ _{; m 2} represents the highest value of the predicted traffic flow speed within the time window (t ₂ , t ₂ _{+ t′), m 2} is equal to or less than m ₀ , and traffic is predicted at this moment The flow speed is the same as the reference traffic flow speed; t′ ₂ represents the time corresponding to m _2. +t' can represent the recovery buffer time, which can be equal to 1, 2, 3, etc. It can be understood that after the rainy weather ends, the traffic flow speed will not return to the normal state immediately, and a buffer time is required. +t' can be preset or adjusted according to different weather conditions. t′ ₀ , t′ ₁ , and t′ ₂ may be the evaluation moments, m ₀ , m ₂ may be reference traffic flow data, and m ₀ , m ₁ , and m ₂ may be predicted traffic flow data.

Step 505: Determine the at least one traffic evaluation parameter based on the predicted traffic flow speed, the reference traffic flow speed, and the one or more evaluation moments.

In some embodiments, in order to determine the traffic performance loss parameter LoR, the processing device 110 (or the traffic evaluation module 330) may determine that the traffic flow velocity prediction curve is within one or more evaluation durations formed by one or more evaluation moments. The first integral and the second integral of the traffic flow speed reference curve. Continuing to refer to FIG. 8, the processing device 110 may determine the first integral of the traffic flow velocity prediction curve and the second integral of the traffic flow velocity reference curve _{within the evaluation duration (t′ 2} -t′ _{0 ).} The first integral may represent the flow of vehicles passing through the target road section under weather conditions corresponding to the meteorological data (for example, precipitation weather). The second integral may indicate the amount of traffic passing through the target road section when the weather condition is good. The processing device 110 (or the traffic evaluation module 330) may determine the traffic performance loss parameter based on the first integral and the second integral. For example, the processing device 110 (or the traffic evaluation module 330) may determine the difference between the first integral and the second integral, for example, use the difference obtained by subtracting the first integral from the second integral as the traffic performance loss parameter . The larger the difference, the more the reduction in the traffic volume passing through the target road section in the same time, and the more the traffic performance (for example, the vehicle passing rate per unit time) of the target road section is reduced. In some embodiments, the processing device 110 (or the traffic evaluation module 330) may determine the traffic performance loss parameter based on the following formula (1):

among them,

It represents the reference curve of traffic flow speed, and R(t) represents the prediction curve of traffic flow speed.

In some embodiments, the processing device 110 (for example, the traffic evaluation module 330) may determine one or more first differences between the one or more evaluation moments, and specify the one or more first differences One of the values is used as the response time or the recovery time. Continuing to refer to FIG. 8, _{at time t′ 0} , the traffic flow speed of the target road section (whether it is the predicted traffic flow speed or the reference traffic speed) is not affected by weather conditions. At time t′ ₁ , the predicted traffic flow speed of the target road section is affected by weather conditions and drops to the lowest value. After time t′ ₁ , the predicted traffic flow speed of the target road section begins to recover until it returns to the maximum value at time _{t′ 2.} Then, _{time t′ 0 to} time t′ ₁ may be the decay duration of the traffic flow velocity of the target road section, and may also be referred to as the response time of the target road section in response to weather changes. t _'1 to time t' ₂ may be a long time to restore traffic flow velocity of the target road section, may also be referred to as the target link length when recovery response weather changes. The processing device 110 (for example, the traffic evaluation module 330) may determine that the first time length formed from _{time t′ 0 to} time t′ ₁ is the response time, and the second time length formed from _{time t′ 1 to} time t′ _{2 is} The recovery time. In some embodiments, the processing device 110 (for example, the traffic assessment module 330) may determine the response time and the recovery time based on the following formulas (2)-(3):

RST=t′ ₁ -t′ ₀ (2)

RCT=t′ ₂ -t′ ₁ (3)

In some embodiments, the processing device 110 (for example, the traffic evaluation module 330) may determine one or more second differences between the predicted traffic flow speed and the reference traffic flow speed corresponding to the one or more evaluation moments. Value, and determine one or more ratios between the one or more second differences and the one or more first differences. The one or more ratios may be designated as the response rate or the recovery rate. With continued reference to FIG. 8, the predicted traffic flow velocity of the target road section 't m ₀ decays to ₀ at time "from the time t m ₁ _1' recovery time to ₁ m ₁ t 'from the time t ₂ m ₂ . The processing device 110 (for example, the traffic evaluation module 330) may determine the response rate and the recovery rate based on the following formulas (4)-(5):

It should be noted that the foregoing description of the process 500 is only for example and description, and does not limit the scope of application of the present application. For those skilled in the art, various modifications and changes can be made to the process 500 under the guidance of this application. However, these amendments and changes are still within the scope of this application.

Fig. 6 is an exemplary flowchart for determining a prediction model according to some embodiments of the present application. In some embodiments, the process 600 may be executed by a processing device (for example, the processing device 110 shown in FIG. 1). For example, the process 600 may be stored in a storage device (for example, the storage device 130 or a storage unit of a processing device) in the form of a program or instruction. When the processor 210 or the module shown in FIG. 3B executes the program or instruction, the process may be implemented. 600. In some embodiments, the process 600 may be executed by the training module 340. In some embodiments, the process 600 may utilize one or more additional operations not described below, and/or not be completed by one or more operations discussed below. In addition, the order of operations shown in FIG. 6 is not restrictive. As shown in FIG. 6, the process 600 may include the following operations.

Step 610: Obtain multiple training samples. In some embodiments, step 610 may be performed by the sample acquisition module 350.

In some embodiments, each training sample may be determined based on reference weather data, predicted weather data, and reference traffic flow data in a certain period of time with at least one training road segment. Each training sample may also be determined based on reference traffic flow data and/or reference weather data in a time period before the time period. The time period represents the time period before the current moment (that is, the historical time period). The reference meteorological data may be the meteorological data of the at least one training road section actually monitored in the time period (for example, within one year or two years). The predicted meteorological data may be meteorological data in the time period obtained through prediction based on the actually monitored meteorological data in the time period before the time period. For example, the forecasted weather data on Sunday in the past week can be determined using the WRF model using the actual weather data monitored from Monday to Saturday in the past week. The reference traffic flow data may be actual traffic flow data of at least one training road segment in the time period. The reference traffic flow data may include actual traffic flow speed. The reference traffic flow data in the time period before the time period may include the traffic flow data actually monitored in the time period before the time period. The reference meteorological data in the time period before the time period may include the meteorological data actually monitored in the time period before the time period.

In some embodiments, the training samples may include training input sequences and training labels. The reference traffic flow speed data in this time period can be used as a training label. The reference weather data and/or predicted weather data in the time period and the reference traffic flow data and/or reference weather data in the time period before the time period can generate an input sequence. In some embodiments, the training label may be added to the input sequence. In some embodiments, the processing device 110 (for example, the sample acquisition module 350) may be based on the reference weather data, the predicted weather data, the reference traffic flow speed data in the time period, and the reference traffic flow in the time period before the time period. The data and/or reference meteorological data generate the training samples according to the number of days. For example, assuming that the time period includes 7 days and the plurality of training road sections include 7 pieces, the processing device 110 (for example, the sample acquisition module 350) may generate 7*7 pieces of training samples. Each training road segment corresponds to 7 training samples. As an example, the training sample for a certain training road segment corresponding to the second day can be based on the reference weather data of the second day, the predicted weather data of the second day, the reference traffic flow data of the second day, and the first day before the second day. Or multiple days of reference traffic flow data and/or reference weather data. The reference weather data of the next day, the forecast weather data of the next day, the reference traffic flow data of one or more days before the next day and/or the features included in the reference weather data, such as traffic flow speed, temperature, Humidity, air pressure, precipitation, wind speed, wind direction, illuminance, radiation intensity, sunshine hours, etc. are arranged in sequence to obtain the training input sequence. The reference traffic flow data of the next day is designated as the training label. Since the model is to obtain the predicted traffic flow data for the second day, in other words, the training input sequence of a certain training road section corresponding to the training sample of the second day is input to the model to be trained to obtain the predicted traffic for the second day Flow data, the training label may be the reference traffic flow data of the next day.

Step 603: Based on the multiple training samples, use one or more iterative processes to train an initial machine learning model to obtain a trained prediction model. In some embodiments, step 603 may be performed by the training module 360.

In some embodiments, the initial machine model may include a deep learning model. For example, sequence-to-sequence model. In some embodiments, the prediction model may be a sequence-to-sequence model based on attention. The attention-based sequence-to-sequence model may be based on an encoder and decoder architecture. The encoder may include an encoder based on long short-term memory (LSTM), including LSTM, Bi-LSTM, CNN-LSTM, Conv-LSTM, TGC-LSTM, etc. The decoder may include a decoder based on Gated Recurrent Unit (GRU), including GRU, Bi-GRU, DCGRU, and so on. Exemplarily, the prediction model may use a deep Bi-LSTM as an encoder and a deep GRU as a decoder. The encoder and decoder respectively contain 3 layers of neural networks, each layer of neural network contains 1379 hidden nodes, and SoftSign is used as the activation function.

In some embodiments, for each training sample, the processing device 110 (for example, the training module 360) may input the training input sequence to the prediction model for processing, and obtain the corresponding prediction result. The prediction result may be predicted traffic flow data, for example, predicted traffic flow speed. In some embodiments, the processing device 110 (for example, the training module 360) may compare the difference between the prediction result and the training label corresponding to the training sample. For example, the processing device 110 (for example, the training module 360) may determine the error between the predicted traffic flow speed and the corresponding historical reference traffic flow speed, for example, based on a loss function. The difference may indicate the prediction effect of the prediction model. The greater the difference, the worse the prediction effect of the model. The processing device 110 (for example, the training module 360) may reversely adjust the parameters of the prediction model based on the difference to reduce the difference. For example, adjusting the hyperparameters of the prediction model, such as the learning rate.

In some embodiments, after completing one or more iterations, if a preset condition is met, for example, the number of iteration rounds or the number of training reaches a preset number, or the difference is less than a preset threshold, the training can be stopped. The obtained trained machine learning model can be designated as the trained prediction model.

As an example, for the training of the prediction model, a dropout technique with a ratio of 0.15 can be used to reduce overfitting during the training process. At the same time, the mean square error (Mean Squared Error) can be used as the loss function, and RMSprop as the optimizer (for example, the optimization loss function has a problem of excessive swing in the update, and further accelerates the convergence speed of the function), where the training batch The size (Batch Size) is 64, the period (Epoch) is 400, and Early Stopping technology is used in the training process to obtain the best training effect.

It should be noted that the foregoing description of the process 600 is only for example and description, and does not limit the scope of application of the present application. For those skilled in the art, various modifications and changes can be made to the process 600 under the guidance of this application. However, these amendments and changes are still within the scope of this application.

One aspect of this specification provides a traffic information processing method for extreme weather. The method is based on extreme weather traffic performance information processing. It can predict the recovery ability of traffic performance after extreme weather based on big data, which is a good way for cities. In the face of extreme weather, emergency evacuation, rescue, and dispatch provide a powerful quantitative decision-making aid, as shown in Figure 9, which specifically includes the following steps:

Step 901: Obtain a set of weather data and a set of traffic flow data in a predetermined area.

In this step, in order to be able to accurately predict the traffic network situation and the recovery ability of traffic performance in a predetermined area within a predetermined time range in the future, first of all, it is necessary to obtain big data about weather conditions and traffic network conditions. Obtain the historical data of the predetermined area within a certain time range in the past. The predetermined area can be delineated according to actual needs. The historical data here includes at least historical data related to weather, etc. and historical data related to the road traffic network. The past certain time here can be determined according to demand, for example, it can be weather data and traffic flow data in the past 12 months, as shown in FIG. 10. In some embodiments, the following steps are specifically included:

Step 1001: Obtain historical weather data and traffic flow historical data of the predetermined area.

In this step, when it is necessary to judge and predict the traffic network conditions of the predetermined area based on meteorological factors, firstly, it is necessary to obtain historical weather data and traffic flow historical data of the predetermined area. For example, it may be obtained that the predetermined area is in Weather grid data within a certain time range and historical traffic flow data within a certain time range.

Specifically, in the process of obtaining meteorological historical data, corresponding meteorological data can be obtained based on the WRF (Weather Research and Forecasting Model) of the Bureau of Meteorology. The meteorological data can be various data representing meteorological conditions such as temperature, air pressure, and precipitation. By adopting this model, it is only necessary to generate a designated geographic range based on the location parameters of the predetermined area and the use requirements for meteorological parameters, and determine the meteorological grid data with a predetermined grid density based on the geographic range. For example, the meteorological grid data can be The historical data of the monitoring station is combined with the meteorological physical and chemical model to generate meteorological grid data. Of course, this model can also be used to predict the future meteorological grid data in the future 6-72 hours.

In the process of obtaining historical traffic flow data, first, determine the actual road information within the predetermined area. The road information can be expressed in the form of road segments, so as to accurately determine the predetermined area or the traffic network coverage area of the predetermined area. The traffic flow information of the road section reflects the traffic network status of the predetermined area. Specifically, the predetermined area is divided according to road sections. Among them, the link between two signal lights can be determined as a road section and based on the actual road section. The situation determines the code or ID of the road section and the road level. The road section here can include roads of various road levels, such as highways, urban highways, national roads, provincial roads, and so on. For example, obtain 1379 roads of different levels for a certain area, including 34 highways (Level 00), 8 urban highways (Level 01), 116 national highways (Level 02), and 521 provincial highways (Level 03) , 700 county roads (Level 04). Further, based on the ID and road level of each road segment, the average traffic flow speed recorded by the date of each road segment ID hourly speed (m/s) is obtained.

Step 1003: Perform feature extraction on the weather history data and traffic flow history data, and obtain a road section-based weather data set and a traffic flow data set.

After obtaining a large amount of weather history data and traffic flow history data through step S201, in this step, these historical basic data are preliminarily sorted and analyzed, and features are extracted from the weather history data and traffic flow history data. First of all, in order to accurately obtain the weather and traffic flow information of a predetermined area, first extract the road segment ID feature in each data item, and sort all road segment IDs in the road traffic network of the predetermined area according to ID or road level to obtain data based on road segment ID set.

Further, temperature, air pressure, etc., characterized as meteorological factors or weather conditions, can be used as meteorological feature parameters, and the traffic flow velocity, characterized as road traffic network conditions, can be extracted as traffic flow feature parameters. For example, in one embodiment, Determine temperature (Temperature), air pressure (Pressure), precipitation (Precipitation), humidity (Humidity), wind direction (Wind_direction), wind speed (Wind_speed), traffic flow speed (Traffic_speed) as characteristic parameters. Feature extraction, and classification and sorting of data by date or even hour according to feature input fields such as temperature, air pressure, precipitation, humidity, wind direction, wind speed, and traffic flow speed.

Finally, the data in the data set based on the road segment ID is matched with the extracted weather and traffic flow data. For example, the historical traffic flow data is sorted into separate traffic flow data based on the road segment ID on a weekly or daily basis. Collection; for historical weather data, the weather data is assigned to the road section of the traffic road network according to the principle of proximity, and the meteorological characteristic parameters determined in the weather historical data obtained through the WRF model, for example, the weather historical data is based on Monday, Tuesday,... , Sunday, etc. are sorted into separate road-based weather data collections, so as to finally obtain the road-segment ID-based weather data collections and traffic flow data collections in the predetermined area.

Step 903: Input the data in the meteorological data set and the traffic flow data set into a machine learning model, so as to determine traffic flow data at a predetermined time.

Through step 901, the meteorological data set and the traffic flow data set of the predetermined area are obtained. In order to facilitate the simulation and prediction of future traffic flow data through the machine learning model, the data in these data sets need to be used as the machine learning model in a certain format. Input data, and use traffic flow data in a predetermined time range in the future as output data. The machine learning model here is a deep sequence model, so that the deep sequence model can be used to simulate and predict the road traffic network after extreme weather. Specifically, in some embodiments, the machine learning model is a deep sequence model based on the attention mechanism, as shown in FIG. 11, including the following steps:

Step 1101: Determine a model data format based on the road section.

In this step, based on the acquired meteorological data collection and traffic flow data collection, the model input data format is constructed according to the preset input data form of the depth sequence model. Specifically, the traffic flow speed data in the traffic flow data set, the weather feature data in the weather data set, etc. are constructed in a certain order into the input data format based on the road section ID, namely Input Sequence, in the form of (Days*ID, Steps* Features); The traffic flow speed that needs to be predicted in the next n hours is used as the output data format based on the road segment ID, namely Output Sequence, in the form of (Days*ID, Steps); finally, the input data format Input Sequence and the output data format Output Sequence are spliced Up, the format is (Days*ID, Steps*(Input Features+Output Features)).

Further, before step 301, the weather data and traffic flow data in the weather data set and the traffic flow data set may also be normalized. Specifically, for example, the input data (Days*ID, Steps*Features) The data becomes (Days, Steps, ID*Features), so that the data is normalized (Normalization), so that the deep sequence model can better process the data.

Step 1103: Determine traffic flow data at a predetermined time through a deep sequence model based on the attention mechanism.

In this step, the attention-based Seq2Seq Model is used to predict the traffic network speed at a predetermined time in the future. The attention mechanism is literally similar to the human attention mechanism. Humans quickly scan the global text to obtain the area that needs to be focused on, which is commonly referred to as the focus of attention, and then invest more attention resources in this area to obtain more detailed information about the target that needs to be focused, and suppress Other useless information. The existence of this mechanism has greatly improved the means for humans to screen out high-value information from a large amount of information. It is a survival mechanism formed by humans in long-term evolution. The attention mechanism in deep learning is essentially similar to the human selective mechanism. The core goal is to select information that is more critical to the current task goal from a large number of information. At present, the attention mechanism has been widely used in various types of deep learning tasks such as natural language processing, image recognition, and speech recognition. It is one of the most worthy of attention and in-depth understanding of the core technology of deep learning technology. Today's mainstream sequence transformation models are based on the encoder and decoder architectures. Behind the encoder architecture is a complex recurrent neural network (RNN) or convolutional neural network (CNN). In order to obtain better performance, an attention mechanism is further added on the basis of the codec architecture.

Specifically, the deep sequence model based on the attention mechanism adopted here uses deep Bidirectional LSTM (BiLSTM) as an encoder, and deep Gated Recurrent Units (GRU) as a decoder to construct an attention vector. From the structure of the model, the encoder and decoder respectively contain 3 layers of neural networks, each layer of neural network contains 1379 hidden nodes, Soft Sign is used as the activation function, and the 01.5 ratio of Dropout technology is used to reduce over-fitting.

Further, in the training of the deep sequence model of the attention mechanism, Mean Squared Error is used as the loss function, and RMSprop is used as the optimizer, where the training Batch Size is 64, the Epoch is 400, and the Early Stopping technology is used in the training process. In order to obtain the best training effect. The structure diagram is shown in Figure 2. By adopting a deep sequence model based on the attention mechanism, the traffic flow data at a predetermined time is determined, so as to accurately evaluate the traffic network conditions at a predetermined time in the future.

Step 905: Determine the traffic recovery capability based on the traffic flow data of the predetermined time.

In step 903, after determining the traffic flow data for a predetermined time through the deep sequence model, it is necessary to predict and evaluate the recovery performance of the traffic network after extreme weather. As shown in Figure 12, in some embodiments, this is achieved through the following steps:

Step 1201: Determine benchmark traffic performance and traffic recovery indicators.

In this step, first of all, it is necessary to determine the benchmark traffic performance. For example, look for days without any extreme weather, such as no precipitation, in the meteorological data collection and traffic flow data collection, and at the same time obtain whether each road ID in the predetermined area on these dates is The vehicle speed data of the day of precipitation can also calculate the average vehicle speed of each week, such as the average vehicle speed of Monday, Tuesday, Wednesday, Thursday, Friday, Saturday, and Sunday. You can also use the average vehicle speed of the day as The average speed per week is used as the reference traffic performance. For example, defining m′ ₀ , m′ ₂ as the reference traffic flow speed at t′ ₀ and t′ ₂ can be implemented in the following ways:

A. Calculate benchmark traffic performance R for all urban roads

for i _th city road:

_{for h th} day in traffic and weather data:

if the daily precipitation is 0:

record daily speed

compute

In this step, traffic recovery performance is represented by traffic recovery indicators. In this embodiment, five traffic recovery indicators for evaluating traffic recovery performance are determined, including Loss of Resilience (LoR) indicators for traffic network performance and response time Response Time (RST) indicator, Recovery Time (RCT) indicator, Response Rate (RSR) indicator, and Recovery Rate (RCR) indicator.

Specifically, Loss of Resilience (LoR) refers to the loss of performance of the transportation network, expressed as the difference between the enclosed area of the speed within a certain time range and the enclosed area of the daily speed when extreme weather occurs, such as heavy rainfall. , The unit is meter (m); Response Time (RST)/Recovery Time (RCT) refers to the time when the traffic network performance decays from the normal value to the lowest value and recovers from the lowest value when the traffic network faces extreme weather The normal value of time, the unit is hour (hour); Response Rate (RSR)/Recovery Rate Recovery Rate (RCR) refers to the rate at which the traffic network performance decays from the normal value to the lowest value when the traffic network faces extreme weather And the rate of returning from the lowest value to the normal value, in units of every square second per meter (m/s2).

Specifically, the method for determining or calculating the above-mentioned traffic restoration index is as follows:

RST=t′ ₁ -t′ ₀ (7)

RCT=t′ ₂ -t′ ₁ (9)

Taking extreme weather as precipitation as an example, the parameter definitions in the above formula are as follows:

The first type of parameter is the time point of the extreme weather time. Take the key time point of the precipitation event as an example, where t ₀ is the start time of precipitation; t ₁ is the time of peak precipitation; t ₂ is the end time of precipitation; T is t ₀ The observation window period of precipitation events between _{t 2} and t 2.

The second type of parameters is about the key points of the traffic speed curve, where m ₀ is the speed value at time _{t 0} _{; t′ 0} is the same as _{t 0} _{; m 1} is the lowest speed value of the traffic network in the observation window period T; t ′ ₁ is the time to capture m ₁ _{; m 2} is determined by the following way, in the period (t′ ₁ , t′ ₁ +3), the speed value of the restored traffic network should be equal to m ₀ , if the traffic in this period The maximum speed of the network is still less than m ₀ , then the maximum speed is selected as the value of _{m 2} _{; t′ 2} is the time to capture and confirm m _2.

The process of calculating the traffic recovery index is realized in the following ways:

B. Calculate traffic recovery indicators

for i _th city road:

_{for h th} day in traffic and weather data:

if the daily precipitation is greater than 0:

search[t ₀ , t ₁ , t ₂ , t ₃ , T]

search[m ₀ , t′ ₀ , m ₁ , t′ ₁ , m ₂ , t′ ₂ ]

search[m′ ₀ ，m′ ₂ ]

return

Save

Step 1203: Obtain the traffic recovery performance of the predetermined area based on the traffic recovery index.

After determining the calculation method of the benchmark traffic performance, the traffic recovery index, and the traffic recovery index through step 1101, the data related to weather and traffic flow can be substituted into the calculation method to obtain the traffic in the predetermined area based on the road section. Recovery performance.

After obtaining the traffic recovery performance of the predetermined area based on the traffic recovery index, a grade label may be generated for each road segment ID based on the calculation result of the traffic recovery index. For example, the recovery time (RCT) index is lower than the first Threshold values can be regarded as road sections that can quickly recover traffic conditions, and recovery Time (RCT) indicators higher than the second threshold can be regarded as road sections that need to slowly recover traffic conditions. In this way, it can be generated for different road sections. The grade labels of different recovery times can of course also be converted into precipitation grade labels, precipitation label grade labels, etc., which are not limited in this embodiment.

In addition, after obtaining the traffic recovery performance of the predetermined area based on the traffic recovery index, the accuracy of the traffic recovery performance index can also be judged according to the date and road section ID, specifically through MSE, RMSE, MAE, R-Squared The other evaluation indicators evaluate the regression algorithm in the depth sequence model.

Finally, after obtaining the traffic recovery performance of the predetermined area based on the traffic recovery index, it can also be combined with the geographic information system software (GIS) developed by ESRI, Intergraph and MapInfo in the prior art to determine the traffic recovery performance index in time. Evaluate with the predicted performance in space.

The present disclosure can match meteorological data and traffic flow data for each road section in a predetermined area, by constructing a deep sequence model to predict the overall traffic flow speed of all road sections in the transportation network of the predetermined area, and at the same time construct a traffic recovery based on traffic performance loss Performance index capture algorithm, in the prediction of traffic flow speed, the traffic recovery performance index capture algorithm is used to obtain the prediction result of traffic recovery performance, and the prediction accuracy of traffic recovery performance index under different extreme weather and different geographical locations Make an evaluation. The present disclosure can effectively use massive meteorological data, travel data, traffic data, etc. to optimize the depth sequence model, and through the depth sequence model to simulate and predict the traffic network conditions after extreme weather and the recovery performance of the traffic network, it is a good way for urban traffic. Before, during and after the occurrence of extreme weather or disasters, the management department should grasp the specific conditions and changing trends of the urban transportation network, and be able to resume more effective emergency preparedness, response and handling plans.

Another aspect of this specification provides a traffic information processing device for extreme weather. As shown in FIG. 13, it includes an acquisition module 1310, a first determination module 1320, and a second determination module 1330. The foregoing modules are coupled with each other, wherein:

The obtaining module 1310 is used to obtain a collection of weather data and traffic flow data in a predetermined area.

Through the acquisition module 1310, in order to be able to accurately predict the traffic network conditions and the recovery ability of traffic performance in the predetermined area within the predetermined time range in the future, first, it is necessary to obtain big data about the weather conditions and the traffic network conditions. Obtain the historical data of the predetermined area within a certain time range in the past. The predetermined area can be delineated according to actual needs. The historical data here includes at least historical data related to weather, etc. and historical data related to the road traffic network. The past certain time here can be determined according to demand, for example, it can be weather data and traffic flow data for the past 12 months. In some embodiments, it specifically includes the following parts:

The first acquiring unit is used to acquire historical weather data and traffic flow historical data of the predetermined area.

Through the first obtaining unit, when it is necessary to judge and predict the traffic network conditions of the predetermined area based on meteorological factors, firstly, it is necessary to obtain historical weather data and traffic flow historical data of the predetermined area. For example, the predetermined area may be obtained. Weather grid data within a certain time range and historical traffic flow data within a certain time range.

The second acquisition unit is used for feature extraction of the weather history data and traffic flow history data, and acquires a road section-based weather data collection and a traffic flow data collection.

After acquiring a large amount of historical weather data and traffic flow historical data through the first acquisition unit, through the second acquisition unit, these historical basic data are preliminarily sorted and analyzed, and the weather historical data and traffic flow historical data are characterized extract. First of all, in order to accurately obtain the weather and traffic flow information of a predetermined area, first extract the road segment ID feature in each data item, and sort all road segment IDs in the road traffic network of the predetermined area according to ID or road level to obtain data based on road segment ID set.

The first determining module 1320 is configured to input the data in the meteorological data set and the traffic flow data set into a machine learning model, so as to determine traffic flow data at a predetermined time.

Through the acquisition module 1310, the meteorological data set and traffic flow data set of the predetermined area are acquired. In order to facilitate the simulation and prediction of future traffic flow data through the machine learning model, the data in these data sets need to be used as a machine learning model in a certain format For example, the traffic flow data of a predetermined time range in the future is used as the output data. The machine learning model here is a deep sequence model, so that the deep sequence model can be used to simulate and predict road traffic network conditions after extreme weather. Specifically, in some embodiments, the machine learning model is a deep sequence model based on an attention mechanism, including the following parts:

The first determining unit is used to determine the model data format based on the road section.

Through the first determining unit, based on the acquired meteorological data set and traffic flow data set, the model input data format is constructed according to the preset input data form of the depth sequence model. Specifically, the traffic flow speed data in the traffic flow data set, the weather feature data in the weather data set, etc. are constructed in a certain order into the input data format based on the road section ID, namely Input Sequence, in the form of (Days*ID, Steps* Features); The traffic flow speed that needs to be predicted in the next n hours is used as the output data format based on the road segment ID, namely Output Sequence, in the form of (Days*ID, Steps); finally, the input data format Input Sequence and the output data format Output Sequence are spliced Up, the format is (Days*ID, Steps*(Input Features+Output Features)).

Further, the first determining unit can also be used to normalize the weather data and traffic flow data in the weather data set and the traffic flow data set. Specifically, for example, the input data (Days*ID, Steps*Features ) The data becomes (Days, Steps, ID*Features), so that the data is normalized (Normalization), so that the deep sequence model can better process the data.

The second determining unit is used to determine traffic flow data at a predetermined time through a deep sequence model based on the attention mechanism.

Through the second determination unit, the attention-based Seq2Seq Model is used to predict the traffic network speed at a predetermined time in the future. The attention mechanism is literally similar to the human attention mechanism. Humans quickly scan the global text to obtain the area that needs to be focused, which is commonly referred to as the focus of attention, and then invest more attention resources in this area to obtain more detailed information about the target that needs to be focused, and suppress Other useless information. The existence of this mechanism has greatly improved the means for humans to screen out high-value information from a large amount of information. It is a survival mechanism formed by humans in long-term evolution. The attention mechanism in deep learning is essentially similar to the human selective mechanism. The core goal is to select information that is more critical to the current task goal from a large number of information. At present, the attention mechanism has been widely used in various types of deep learning tasks such as natural language processing, image recognition, and speech recognition. It is one of the most worthy of attention and in-depth understanding of the core technology of deep learning technology. Today's mainstream sequence transformation models are based on the encoder and decoder architectures. Behind the encoder architecture is a complex recurrent neural network (RNN) or convolutional neural network (CNN). In order to obtain better performance, an attention mechanism is further added on the basis of the codec architecture.

The second determining module 1330 is configured to determine the traffic recovery capability based on the traffic flow data of the predetermined time.

After the second determination module 1330 determines the traffic flow data for a predetermined time through the deep sequence model, it is necessary to predict and evaluate the recovery performance of the traffic network after the occurrence of extreme weather. As shown in the figure, in some embodiments, the following parts are included:

The third determining unit is used to determine the benchmark traffic performance and the traffic recovery index.

Through the third determination unit, first, it is necessary to determine the benchmark traffic performance, for example, look for days without any extreme weather, such as no precipitation, in the weather data collection and traffic flow data collection, and at the same time obtain the ID of each road in the predetermined area on these dates. Vehicle speed data on days without precipitation can also calculate the average vehicle speed every week, such as Monday, Tuesday, Wednesday, Thursday, Friday, Saturday, Sunday, and the average vehicle speed of the day. As the average speed per week, as the reference traffic performance, for example, define m′ ₀ , m′ ₂ as the reference traffic flow speed at time t′ ₀ and t′ ₂ , which can be implemented in the following ways:

A. Calculate benchmark traffic performance R for all urban roads

for i _th city road:

_{for h th} day in traffic and weather data:

if the daily precipitation is 0:

record daily speed

compute

Through the third determining unit, the traffic recovery performance is expressed by traffic recovery indicators. In this embodiment, five traffic recovery indicators used to evaluate the traffic recovery performance are determined, including the Loss of Resilience (LoR) indicator and response time of the traffic network. Response Time (RST) indicator, Recovery Time (RCT) indicator, Response Rate (RSR) indicator, and Recovery Rate (RCR) indicator.

Specifically, Loss of Resilience (LoR) refers to the loss of performance of the transportation network, expressed as the difference between the enclosed area of the speed within a certain time range and the enclosed area of the daily speed when extreme weather occurs, such as heavy rainfall. , The unit is meter (m); Response Time (RST)/Recovery Time (RCT) refers to the time when the traffic network performance decays from the normal value to the lowest value and recovers from the lowest value when the traffic network faces extreme weather The normal value of time, the unit is hour (hour); Response Rate (RSR)/Recovery Rate (RCR) refers to the rate at which the traffic network performance decays from the normal value to the lowest value when the traffic network faces extreme weather And the rate of returning from the lowest value to the normal value, in units of every square second per meter (m/s2).

RST=t′ ₁ -t′ ₀ (7)

RCT=t′ ₂ -t′ ₁ (9)

B. Calculate traffic recovery indicators

for i _th city road:

_{for h th} day in traffic and weather data:

if the daily precipitation is greater than 0:

search[t ₀ , t ₁ , t ₂ , t ₃ , T]

search[m ₀ , t′ ₀ , m ₁ , t′ ₁ , m ₂ , t′ ₂ ]

search[m′ ₀ ，m′ ₂ ]

return

Save

The third acquiring unit is configured to acquire the traffic recovery performance of the predetermined area based on the traffic recovery index.

After the reference traffic performance and the traffic recovery index are determined by the third determining unit, the traffic recovery performance of the predetermined area based on the road section can be obtained

In addition, the second determining module 1330 also includes a generating unit, which is used to generate a grade label for each road segment ID based on the calculation result of the traffic recovery index. For example, if the Recovery Time (RCT) index is lower than the first threshold, it can be considered It is a road section that can quickly recover the traffic condition, and the Recovery Time (RCT) indicator is higher than the second threshold, which can be considered as a road section that needs to slowly recover the traffic condition. In this way, different road sections are generated with different recovery time grade labels Of course, it is also possible to form a precipitation grade label, a precipitation label grade label, etc. through conversion, which is not limited in this embodiment.

In addition, it may also include a judging unit, which is used to judge the accuracy of the traffic recovery performance index according to the date and road section ID. Specifically, the regression algorithm in the depth sequence model is evaluated by evaluation indexes such as MSE, RMSE, MAE, R-Squared, etc. Evaluation.

Finally, it also includes an evaluation unit, which is used to evaluate the prediction performance of traffic recovery performance indicators in time and space in combination with geographic information system software (GIS) developed by ESRI, Intergraph, and MapInfo in the prior art.

The present disclosure can match meteorological data and traffic flow data for each road section in a predetermined area, by constructing a deep sequence model to predict the overall traffic flow speed of all road sections in the transportation network of the predetermined area, and at the same time construct a traffic recovery based on traffic performance loss Performance index capture algorithm, in the prediction of traffic flow speed, the traffic recovery performance index capture algorithm is used to obtain the prediction result of traffic recovery performance, and the prediction accuracy of traffic recovery performance index under different extreme weather and different geographical locations Make an evaluation. The present disclosure can effectively use massive meteorological data, travel data, traffic data, etc. to optimize the depth sequence model, and through the depth sequence model to simulate and predict the traffic network conditions after extreme weather and the recovery performance of the traffic network, it is a good way for urban traffic Before, during and after the occurrence of extreme weather or disasters, the management department should grasp the specific conditions and changing trends of the urban transportation network, and be able to resume more effective emergency preparedness, response and handling plans.

Another aspect of this specification provides a storage medium, which is a computer-readable medium and stores a computer program. When the computer program is executed by a processor, the method provided by any embodiment of the present disclosure is implemented, including the following steps S11 to S13 :

S11. Obtain a collection of weather data and a collection of traffic flow data in a predetermined area;

S12: Input the data in the meteorological data set and the traffic flow data set into a machine learning model, so as to determine traffic flow data at a predetermined time;

S13: Determine the traffic recovery capability based on the traffic flow data of the predetermined time.

When the computer program is executed by the processor to obtain the meteorological data set and the traffic flow data set of the predetermined area, the processor specifically executes the following steps: obtain the meteorological historical data and the traffic flow historical data of the predetermined area; Feature extraction is performed on historical traffic flow data, and meteorological data collections and traffic flow data collections based on road sections are obtained.

When the computer program is executed by the processor to determine the traffic flow data for a predetermined time through the deep sequence model, the machine learning model is a deep sequence model based on the attention mechanism. Specifically, the processor executes the following steps: Determine the model based on the road section Data format: Determine the traffic flow data at a predetermined time through a deep sequence model based on the attention mechanism.

When the computer program is executed by the processor to determine the traffic recovery capability based on the traffic flow data for the predetermined time, the processor specifically executes the following steps: determining the benchmark traffic performance and the traffic recovery index; and obtaining the predetermined area based on the traffic recovery index Traffic recovery performance.

Another aspect of this specification provides an electronic device. The schematic structural diagram of the electronic device may be as shown in FIG. 14, at least including a memory 1410 and a processor 1420. The memory 1410 stores a computer program, and the processor 1420 executes the memory 1410. The computer program implements the method provided in any embodiment of the present disclosure. Exemplarily, the steps of the computer program of the electronic device are as follows S21 to S23:

S21: Obtain a collection of meteorological data and a collection of traffic flow data in a predetermined area;

S22: Input the data in the meteorological data set and the traffic flow data set into a machine learning model, so as to determine traffic flow data at a predetermined time;

S23: Determine a traffic recovery capability based on the traffic flow data of the predetermined time.

When the processor executes the weather data collection and traffic flow data collection of the predetermined area stored on the memory, it specifically executes the following computer program: Obtain the weather history data and traffic flow history data of the predetermined area; Feature extraction is performed on historical traffic flow data, and meteorological data collections and traffic flow data collections based on road sections are obtained.

When the processor determines the traffic flow data for a predetermined time through the deep sequence model stored on the execution memory, the machine learning model is a deep sequence model based on the attention mechanism, and specifically executes the following computer program: determine the model based on the road section Data format: Determine the traffic flow data at a predetermined time through a deep sequence model based on the attention mechanism.

When determining the traffic recovery capability based on the traffic flow data stored on the memory for the predetermined time, the processor specifically executes the following computer program: determining a reference traffic performance and a traffic recovery index; and obtaining the predetermined area based on the traffic recovery index Traffic recovery performance.

The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is only an example, and does not constitute a limitation to the application. Although it is not explicitly stated here, those skilled in the art may make various modifications, improvements and amendments to this application. Such modifications, improvements, and corrections are suggested in this application, so such modifications, improvements, and corrections still belong to the spirit and scope of the exemplary embodiments of this application.

At the same time, this application uses specific words to describe the embodiments of the application. For example, "one embodiment", "an embodiment", and/or "some embodiments" mean a certain feature, structure, or characteristic related to at least one embodiment of the present application. Therefore, it should be emphasized and noted that “one embodiment” or “one embodiment” or “an alternative embodiment” mentioned twice or more in different positions in this specification does not necessarily refer to the same embodiment. . In addition, some features, structures, or characteristics in one or more embodiments of the present application can be appropriately combined.

In addition, those skilled in the art can understand that various aspects of this application can be explained and described through a number of patentable categories or situations, including any new and useful process, machine, product, or combination of substances, or a combination of them. Any new and useful improvements. Correspondingly, various aspects of the present application can be completely executed by hardware, can be completely executed by software (including firmware, resident software, microcode, etc.), or can be executed by a combination of hardware and software. The above hardware or software can all be referred to as "data block", "module", "engine", "unit", "component" or "system". In addition, various aspects of the present application may be embodied as a computer product located in one or more computer-readable media, and the product includes computer-readable program codes.

The computer storage medium may contain a propagated data signal containing a computer program code, for example on a baseband or as part of a carrier wave. The propagated signal may have multiple manifestations, including electromagnetic forms, optical forms, etc., or a suitable combination. The computer storage medium may be any computer readable medium other than the computer readable storage medium, and the medium may be connected to an instruction execution system, device, or device to realize communication, propagation, or transmission of the program for use. The program code located on the computer storage medium can be transmitted through any suitable medium, including radio, cable, fiber optic cable, RF, or similar medium, or any combination of the above medium.

The computer program codes required for the operation of each part of this application can be written in any one or more programming languages, including object-oriented programming languages such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB.NET, Python Etc., conventional programming languages such as C language, Visual Basic, Fortran 2003, Perl, COBOL 2002, PHP, ABAP, dynamic programming languages such as Python, Ruby and Groovy, or other programming languages. The program code can be run entirely on the user's computer, or run as an independent software package on the user's computer, or partly run on the user's computer and partly run on a remote computer, or run entirely on the remote computer or server. In the latter case, the remote computer can be connected to the user's computer through any network form, such as a local area network (LAN) or a wide area network (WAN), or connected to an external computer (for example, via the Internet), or in a cloud computing environment, or as a service Use software as a service (SaaS).

In addition, unless explicitly stated in the claims, the order of processing elements and sequences, the use of numbers and letters, or the use of other names in this application are not used to limit the order of the procedures and methods of this application. Although the foregoing disclosure uses various examples to discuss some embodiments of the invention that are currently considered useful, it should be understood that such details are only for illustrative purposes, and the appended claims are not limited to the disclosed embodiments. On the contrary, the rights are The requirements are intended to cover all modifications and equivalent combinations that conform to the essence and scope of the embodiments of the present application. For example, although the system components described above can be implemented by hardware devices, they can also be implemented only by software solutions, such as installing the described system on an existing server or mobile device.

For the same reason, it should be noted that, in order to simplify the expressions disclosed in this application and help the understanding of one or more embodiments of the invention, in the foregoing description of the embodiments of this application, multiple features are sometimes combined into one embodiment. In the drawings or its description. However, this method of disclosure does not mean that the subject of the application requires more features than those mentioned in the claims. In fact, the features of the embodiment are less than all the features of the single embodiment disclosed above.

In some embodiments, numbers describing the number of ingredients and attributes are used. It should be understood that such numbers used in the description of the embodiments use the modifier "about", "approximately" or "substantially" in some examples. Retouch. Unless otherwise stated, "approximately", "approximately" or "substantially" indicates that the number is allowed to vary by ±20%. Correspondingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, and the approximate values can be changed according to the required characteristics of individual embodiments. In some embodiments, the numerical parameter should consider the prescribed effective digits and adopt the method of general digit retention. Although the numerical ranges and parameters used to confirm the breadth of the range in some embodiments of the present application are approximate values, in specific embodiments, the setting of such numerical values is as accurate as possible within the feasible range.

For each patent, patent application, patent application publication and other materials cited in this application, such as articles, books, specifications, publications, documents, etc., the entire contents of which are hereby incorporated into this application by reference. The application history documents that are inconsistent or conflicting with the content of this application are excluded, and documents that restrict the broadest scope of the claims of this application (currently or later attached to this application) are also excluded. It should be noted that if there is any inconsistency or conflict between the description, definition, and/or term usage in the attached materials of this application and the content described in this application, the description, definition and/or term usage of this application shall prevail .

Finally, it should be understood that the embodiments described in this application are only used to illustrate the principles of the embodiments of this application. Other variations may also fall within the scope of this application. Therefore, as an example and not a limitation, the alternative configuration of the embodiment of the present application can be regarded as consistent with the teaching of the present application. Correspondingly, the embodiments of the present application are not limited to the embodiments explicitly introduced and described in the present application.

Claims

A system including:

At least one storage medium, the storage medium including instructions;

At least one processor, the at least one processor is in communication with the at least one storage medium, wherein, when the instruction is executed, the at least one processor is configured to:

Acquiring meteorological data of the target road section in a certain time period and historical traffic flow data in a historical time period before the time period;

Based on the historical traffic flow data and the meteorological data, using a trained prediction model to determine the predicted traffic flow data of the target road section in the time period;

Based on the predicted traffic flow data, determine at least one traffic evaluation parameter related to the target road section within the time period, where the traffic evaluation parameter is used to evaluate the traffic recovery capability of the target road section under the weather data condition .
The system according to claim 1, wherein the historical traffic flow data includes at least one of the following historical traffic flow data:

The first historical traffic flow data of the target road section in the first historical time period adjacent to the time period, or

Second historical traffic flow data related to the target road segment in a second historical time period separated by one or more time periods from the time period.
The system according to claim 2, wherein, based on the historical traffic flow data and the meteorological data, a trained prediction model is used to determine the predicted traffic flow data of the target road section within the time period, the At least one processor is configured as:

Acquiring historical weather data of the target road section in a third historical time period adjacent to the time period before the time period; and

Based on the historical weather data, the historical traffic flow data, and the weather data, the predicted traffic flow data is determined using the prediction model.
The system according to claim 1, wherein the prediction model includes a sequence-to-sequence model based on attention; in order to determine the time based on the historical traffic flow data and the meteorological data, the trained prediction model is used to determine For the predicted traffic flow data of the target road section in the segment, the at least one processor is configured to:

Based on the historical traffic flow data and the meteorological data, construct an input sequence adapted to the prediction model; and

Use the prediction model to process the input sequence to determine the predicted traffic flow data.
The system according to claim 4, wherein the prediction model comprises an encoder based on long short-term memory (LSTM) and a decoder based on Gated Recurrent Unit (GRU).
The system according to claim 1, wherein, to determine at least one traffic evaluation parameter related to the target road section in the time period based on the predicted traffic flow data, the at least one processor is configured to:

Acquiring reference traffic flow data related to the target road section;

Obtain one or more evaluation moments;

Determine one or more predicted traffic flow speeds and reference traffic flow speeds corresponding to one or more assessment moments; and

The at least one traffic evaluation parameter is determined based on the predicted traffic flow data, the reference traffic flow data, the one or more evaluation moments, and the predicted traffic flow speed and the reference traffic flow speed.
The system according to claim 6, wherein the traffic evaluation parameter includes a traffic performance loss parameter; the predicted traffic flow data includes a traffic flow velocity prediction curve related to the target road section within the time period, the The reference traffic flow data includes a traffic flow speed reference curve related to the target road section in a reference time period having the same time identifier as the time period; to determine the traffic performance loss parameter, the at least one processor is configured for:

Determine the first integral of the traffic flow velocity prediction curve and the second integral of the traffic flow velocity reference curve within one or more assessment durations constituted by one or more assessment moments; the first integral indicates that The flow of vehicles passing through the target road section within the one or more evaluation time lengths under the condition of the meteorological data, and the second integral represents passing the target road section within the one or more evaluation time lengths under good weather conditions Traffic volume; and

Based on the first integral and the second integral, the traffic performance loss parameter is determined.
The system according to claim 6, wherein the traffic evaluation parameters include response time and recovery time, and to determine the response time and the recovery time, the at least one processor is configured to:

Determine one or more first differences between the one or more evaluation moments;

Specify one of the one or more first difference values as the response time or the recovery time.
The system according to claim 8, wherein the traffic evaluation parameter includes a response rate and a recovery rate, and to determine the response rate and the recovery rate, the at least one processor is configured to:

Determining one or more second differences between the predicted traffic flow speed and the reference traffic flow speed corresponding to the one or more evaluation moments;

Determining one or more ratios between the one or more second differences and the one or more first differences;

One of the one or more ratios is designated as the response rate or the recovery rate.
The system according to claim 1, wherein, to determine at least one traffic evaluation parameter related to the target road section in the time period based on the predicted traffic flow data, the at least one processor is configured to:

Obtaining a traffic assessment model that reflects the relationship between traffic flow speed and traffic assessment parameters; and

The traffic assessment parameter is determined based on the traffic assessment model and the predicted traffic flow data.
The system of claim 10, wherein the at least one processor is further configured to:

Acquiring at least one reference traffic assessment parameter of multiple road sections in multiple time periods and a predicted traffic assessment parameter determined by using the traffic assessment model;

as well as

Determine the accuracy of the traffic assessment model based on at least one reference traffic assessment parameter and predicted traffic assessment parameter of the multiple road sections in multiple time periods; and

The traffic evaluation model is updated based on the accuracy.
The system of claim 1, wherein the at least one processor is further configured to:

Acquiring traffic evaluation parameters of multiple road sections in the target area including the target road section; and

The traffic evaluation parameters of the multiple road sections are mapped to the map data of the target area, and the visualized result of the traffic evaluation of the target area is obtained.
The system according to claim 1, wherein the predictive model is determined by the following process, the process comprising:

Obtain multiple training samples, each training sample includes a training input sequence and a training label; the training input sequence is determined based on historical reference weather data, historical forecast weather data, and historical reference traffic flow data related to multiple training road sections. The training label is determined based on the historical reference traffic flow data;

Based on the multiple training samples, one or more iterations are used to train the prediction model to obtain a trained prediction model, where one iteration includes:

For each training sample,

Determine the prediction result corresponding to the training input sequence;

Determine the difference between the prediction result and the training label corresponding to the training sample;

Based on the difference, the parameters of the prediction model are adjusted to reduce the difference.
A method implemented on a computing device, wherein the computing device has at least one storage medium for storing instructions, and at least one processor that communicates with the at least one storage medium, and Methods include:

Acquiring meteorological data of the target road section in a certain time period and historical traffic flow data in a historical time period before the time period;

Based on the historical traffic flow data and the meteorological data, using a trained prediction model to determine the predicted traffic flow data of the target road section in the time period;

Based on the predicted traffic flow data, determine at least one traffic evaluation parameter related to the target road section within the time period, where the traffic evaluation parameter is used to evaluate the traffic recovery capability of the target road section under the weather data condition .
The method according to claim 14, wherein the historical traffic flow data includes at least one of the following historical traffic flow data:

The first historical traffic flow data of the target road section in the first historical time period adjacent to the time period, or

Second historical traffic flow data related to the target road segment in a second historical time period separated by one or more time periods from the time period.
15. The method according to claim 15, wherein the determining the predicted traffic flow data of the target road section within the time period based on the historical traffic flow data and the meteorological data using a trained prediction model comprises:

Acquiring historical weather data of the target road section in a third historical time period adjacent to the time period before the time period; and

Based on the historical weather data, the historical traffic flow data, and the weather data, the predicted traffic flow data is determined using the prediction model.
The method according to claim 14, wherein the prediction model comprises a sequence-to-sequence model based on attention; and the prediction model based on the historical traffic flow data and the meteorological data is used to determine the The predicted traffic flow data of the target road section in the time period includes:

Based on the historical traffic flow data and the meteorological data, constructing an input sequence adapted to the prediction model; and

Use the prediction model to process the input sequence to determine the predicted traffic flow data.
The method according to claim 17, wherein the prediction model comprises an encoder based on Long Short-term Memory (LSTM) and a decoder based on Gated Recurrent Unit (GRU).
The method according to claim 14, wherein the determining at least one traffic evaluation parameter related to the target road section within the time period based on the predicted traffic flow data comprises:

Acquiring reference traffic flow data related to the target road section;

Obtain one or more evaluation moments;

Determine one or more predicted traffic flow speeds and reference traffic flow speeds corresponding to one or more assessment moments; and

The at least one traffic evaluation parameter is determined based on the predicted traffic flow data, the reference traffic flow data, the one or more evaluation moments, and the predicted traffic flow speed and the reference traffic flow speed.
The method according to claim 19, wherein the traffic evaluation parameter includes a traffic performance loss parameter; the predicted traffic flow data includes a traffic flow velocity prediction curve related to the target road section within the time period, the The reference traffic flow data includes a traffic flow speed reference curve related to the target road section in a reference time period having the same time identifier as the time period; the determining the traffic performance loss parameter includes:

Determine the first integral of the traffic flow velocity prediction curve and the second integral of the traffic flow velocity reference curve within one or more assessment durations constituted by one or more assessment moments; the first integral indicates that The flow of vehicles passing through the target road section within the one or more evaluation time lengths under the condition of the meteorological data, and the second integral represents passing the target road section within the one or more evaluation time lengths under good weather conditions Traffic volume; and

Based on the first integral and the second integral, the traffic performance loss parameter is determined.
The method according to claim 19, wherein the traffic evaluation parameter includes a response time and a recovery time, and the determining the response time and the recovery time comprises:

Determine one or more first differences between the one or more evaluation moments;

Specify one of the one or more first difference values as the response time or the recovery time.
The method according to claim 21, wherein the traffic evaluation parameter includes a response rate and a recovery rate, and the determining the response rate and the recovery rate comprises:

Determining one or more second differences between the predicted traffic flow speed and the reference traffic flow speed corresponding to the one or more evaluation moments;

Determining one or more ratios between the one or more second differences and the one or more first differences;

One of the one or more ratios is designated as the response rate or the recovery rate.
The method according to claim 14, wherein the determining at least one traffic evaluation parameter related to the target road section in the time period based on the predicted traffic flow data comprises:

Obtaining a traffic assessment model that reflects the relationship between traffic flow speed and traffic assessment parameters; and

The traffic assessment parameter is determined based on the traffic assessment model and the predicted traffic flow data.
The method of claim 23, wherein the method further comprises:

Acquiring at least one reference traffic assessment parameter of multiple road sections in multiple time periods and the predicted traffic assessment parameter determined by using the model; and

Determine the accuracy of the traffic assessment model based on at least one reference traffic assessment parameter and predicted traffic assessment parameter of the multiple road sections in multiple time periods; and

The traffic evaluation model is updated based on the accuracy.
The method according to claim 14, wherein the method further comprises:

Acquiring traffic evaluation parameters of multiple road sections in the target area including the target road section; and

The traffic evaluation parameters of the multiple road sections are mapped to the map data of the target area, and the visualized result of the traffic evaluation of the target area is obtained.
The method according to claim 14, wherein the prediction model is determined by the following process, the process comprising:

Obtain multiple training samples, each training sample includes a training input sequence and a training label; the training input sequence is determined based on historical reference weather data, historical forecast weather data, and historical reference traffic flow data related to multiple training road sections. The training label is determined based on the historical reference traffic flow data;

Based on the multiple training samples, one or more iterations are used to train the prediction model to obtain a trained prediction model, where one iteration includes:

For each training sample,

Determine the prediction result corresponding to the training input sequence;

Determine the difference between the prediction result and the training label corresponding to the training sample;

Based on the difference, the parameters of the prediction model are adjusted to reduce the difference.
A computer-readable storage medium that stores computer instructions. After the computer reads the computer instructions in the storage medium, the computer executes the following method;

Acquiring meteorological data of the target road section in a certain time period and historical traffic flow data in a historical time period before the time period;

Based on the historical traffic flow data and the meteorological data, using a trained prediction model to determine the predicted traffic flow data of the target road section in the time period;

Based on the predicted traffic flow data, determine at least one traffic evaluation parameter related to the target road section within the time period, where the traffic evaluation parameter is used to evaluate the traffic recovery capability of the target road section under the weather data condition .