Patent Application
20220309912
This application is based on and claims priority to Chinese patent application No. 202110677845.4, filed on Jun. 18, 2021, the entire contents of which are incorporated herein by reference for all purposes.
The disclosure relates to the field of computer technology, specially the field of intelligent transportation technology, and in particular to a method for predicting traffic data, an apparatus for predicting traffic data and an electronic device.
Traffic data prediction is an important part of intelligent transportation technology. Accurate traffic data prediction helps to take appropriate traffic control measures in time to make traffic more orderly.
However, the road traffic system is a time-varying and complex nonlinear system, not only including natural factors, such as seasons and climate, but also including various human factors, such as traffic accidents, emergencies and psychological states of drivers, which affects real-time and accurate prediction of road traffic data.
Embodiments of the first aspect provide a method for predicting traffic data. The method includes:
obtaining traffic data in a plurality of periods, the plurality of periods including a target period and at least one historical period before the target period;
determining a reference period from the at least one historical period based on a time interval between each of the at least one historical period and the target period;
determining a data similarity between the target period and the reference period based on traffic data at a first time point in the target period and traffic data at a time point corresponding to the first time point in the reference period;
Embodiments of a second aspect provide an electronic device. The electronic device includes: at least one processor and a memory communicatively coupled to the at least one processor. The memory is configured to store instructions executable by the at least one processor. When the instructions are executed by the at least one processor, the at least one processor is enabled to execute the method according to the embodiments of the first aspect.
In order to achieve the above objectives, embodiments of a third aspect provide a non-transitory computer readable storage medium stored with computer instructions, in which the computer instructions are configured to cause the computer to perform the method according to the embodiments of the first aspect.
It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the disclosure, nor is it intended to limit the scope of the disclosure. Additional features of the disclosure will be easily understood through the following description.
The drawings are used to better understand the solution and do not constitute a limitation of the disclosure, in which:
The exemplary embodiments of the disclosure are described below in combination with the accompanying drawings, which include various details of the embodiments of the disclosure to aid in understanding, and should be considered merely exemplary. Therefore, those skilled in the art should know that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the disclosure. For the sake of clarity and brevity, descriptions of well-known features and structures have been omitted from the following description.
As shown in
In step 101, traffic data in a plurality of periods is obtained, the plurality of periods including a target period and at least one historical period before the target period.
The traffic data monitored in the plurality of periods is obtained through a video sensing device. By connecting the video sensing device and a traffic signal light device, road traffic flow and light state information of the traffic signal are strongly bound, without aligning a control clock of the signal light with the video sensing device. The traffic data acquired by the video sensing device will be stored in a preset database. When the outside world needs to extract the traffic information of a target road, the required traffic data can be directly extracted from the database. The traffic data includes current real-time traffic data and historical real-time traffic data within various time periods. In a possible implementation, the traffic data includes values of multiple indicators, such as flow, saturation and crossing time.
The period is relative to the sampling point, and according to historical experience, it is only necessary to collect traffic data at 4 sampling points in respective periods before the current time point, and then the traffic data at the next time point can be predicted. For example, the traffic flow at the current time point t is v(t), then the traffic data of the previous sampling points are v(t−1), v(t−2), and v(t−3) respectively. Similarly, the traffic flow at the sampling points of the historical periods can be expressed as: vh(t), vh(t−1), vh(t−2) and vh(t−3). Correspondingly, the traffic flow at the next time point of the current period and at the next time point of the historical period are: v(t+1), vh(t+1) respectively.
It should be noted that 4 sampling points are only used for exemplary description, and do not constitute a limitation to this embodiment. Those skilled in the art may know that other period numbers or sampling point numbers do not affect the implementation of this embodiment.
In step 102, a reference period is determined from the at least one historical period based on a time interval between each of the at least one historical period and the target period.
Optionally, starting from the historical period closest to the target period, a target number of historical periods are selected as the reference period. In the embodiment, the data prediction error of each historical period may be obtained at first, and the target number is determined according to the number of historical periods having data prediction error within a preset error range in the at least one historical period, and further, the target number of historical periods closest to the target period are selected as the reference periods.
The time interval between the historical period and the target period is different from the time interval between any two sampling points in step 101. The time interval mentioned in this step may be the time interval between the current period, that is, the target period, and the historical period, such as today and the day before yesterday.
In this embodiment, the determination of the reference period may include two implementation schemes.
In the first possible implementation scheme, starting from the target period, the target number of historical periods are selected as the reference period in turn, namely the target number is determined at first, and then the target number of historical periods closest to the target period are searched from a database. The determination of the target number here is determined according to the data prediction error of each historical period and the preset error range.
In the second possible implementation scheme, the target number is determined first according to the data prediction error of each historical period and the preset error range, and then the historical period corresponding to the data prediction error within the preset error range is determined as the reference period.
In step 103, a data similarity between the target period and the reference period is determined based on traffic data at a first time point in the target period and traffic data at a time point corresponding to the first time point in the reference period.
Optionally, a first state vector of the target period is determined based on traffic data at a plurality of first time points in the target period. A second state vector of the reference period is determined based on traffic data at time points corresponding to the plurality of first time points in the reference period. A first correlation coefficient of the flow between two historical periods is determined based on the flow of each of the at least one historical period. A second correlation coefficient of the saturation between two historical periods is determined based on the saturation of each of the at least one historical period. A vector distance between the first state vector and the second state vector is determined based on a difference corresponding to the flow between the first state vector and the second state vector, a difference corresponding to the saturation between the first state vector and the second state vector, the first correlation coefficient and the second correlation coefficient. The data similarity between the target period and the reference period is determined based on the vector distance between the first state vector and the second state vector.
The plurality of first time points may be 4 sampling time points set according to experience, which includes the current time point t, the previous time points t−1, t−2, and t−3.
The state vector is composed of the flow and saturation in the traffic data at the sampling time points of the period. For example, the state vector of the target period is
X=[v(t) . . . v(t−3), o(t) . . . o(t−3)], in which o(t) represents the saturation. Similarly, oh(t) represents the saturation of the historical period.
The first correlation coefficient r1 and the second correlation coefficient r2 are Pearson correlation coefficients, and the calculation formula is:
In the above calculation formula of the correlation coefficient r, x may be expressed as the traffic flow v(t) at the current time point, and y may be expressed as the traffic flow v(t+1) at the next time point; or, x may be expressed as the saturation o(t) at the current time point, y may be expressed as the saturation o(t+1) at the next time point.
It is worth noting that the first correlation coefficient r1 and the second correlation coefficient r2 obtained from the traffic data in each historical period are applicable to all periods including the current period and respective historical periods.
The vector distance adopts the Euclidean distance, and its calculation formula is:
In the embodiment, the traffic data sampled at the 4 sampling points (first time points) in the target period and the traffic data of the corresponding time points in the reference period obtained in step 102 are brought into the calculation formula of the vector distance. Thus, the data similarity between the target period and the reference period is obtained. Generally, the smaller the value of the vector distance is, the higher the data similarity is, and the more meaningful it is for the traffic data prediction of the target period.
In step 104, reference traffic data at a time point in the reference period is obtained, the time point in the reference period being corresponding to a second time point to be predicted in the target period.
The second time point to be predicted in the target period is relative to the first time point, and the time interval between the first time point and the second time point is one cycle. For example, the first time point includes: t−3, t−2, t−1 and t, then the second time point is t+1. In addition, the reference traffic data corresponding to the second time point obtained from the reference period only includes the data of traffic flow.
In step 105, traffic data at the second time point in the target period is predicted based on the data similarity between the target period and the reference period, and the reference traffic data in the reference period.
Optionally, in a possible implementation, a weight of the reference period is determined based on the data similarity between the target period and the reference period, in which the weight has a negative correlation with the data similarity. The traffic data at the second time point in the target period is predicted by weighting the reference traffic data in the reference period based on the weight of the reference period.
It should be noted that the negative correlation refers to a correlation in which the weight decreases as the similarity increases, and increases as the similarity decreases, for example, an inverse relationship.
The main function of the weight of the reference period is to measure a contribution of the data used in the reference period, that is, the traffic flow at the second time point, to the traffic flow to be predicted at the second time point in the target period, the calculation formula is:
In the above formula, β is the weight of the reference period, and Di is the vector distance between the target period and the reference period. It can be seen from the formula that the larger the vector distance, the smaller the weight.
According to the weight of the reference period and the reference traffic data at the second time point in the reference period, that is, the traffic flow data at the second time point, the traffic data at the second time point in the target period, that is, the traffic flow at the second time point in the target period, can be predicted, and its prediction formula is:
It can be seen from the above formula that if the weight of the reference period is large, the reference traffic flow data at the corresponding time point in the reference period can be used more as the traffic flow at the second time point in the target period. On the contrary, if the weight of the reference period is small, the amount of data in the reference period used is also small, so that the reference traffic data at the second time point of the reference period is effectively used, to further improve the prediction accuracy.
According to the method for predicting traffic data according to the embodiments of the disclosure, traffic data in a plurality of periods is obtained. The plurality of periods includes the target period and at least one historical period before the target period. The reference period is determined from the at least one historical period based on the time interval between each of the at least one historical period and the target period. The data similarity between the target period and the reference period is determined based on traffic data at the first time point in the target period and traffic data at a time point corresponding to the first time point in the reference period. Reference traffic data at the time point in the reference period is obtained, the time point in the reference period being corresponding to the second time point to be predicted in the target period. Traffic data at the second time point in the target period is predicted based on the data similarity between the target period and the reference period, and the reference traffic data in the reference period. The traffic data in the reference period that meets the requirements is selected from the historical periods by comparing the traffic data in the historical period with the target period data. Therefore, the prediction of the traffic data in the target period is realized, which not only maintains the characteristics of the traffic data in the historical period, but also ensure easy operation and high prediction accuracy.
On the basis of the above embodiment, this embodiment provides another possible implementation of the method for predicting traffic data.
As shown in
In step 201, traffic data in a plurality of periods is obtained, the plurality of periods including a target period and at least one historical period before the target period.
In this embodiment, specific reference may be made to step 101 in the above embodiment, the principle is the same, and details are not repeated here.
In step 202, starting from the historical period closest to the target period, a target number of historical periods are selected from the at least one historical period as the reference period.
Optionally, in a possible implementation, a data prediction error of each of the at least one historical period is obtained, and the target number is determined based on a number of historical periods having data prediction error within a preset error range in the at least one historical period.
In this embodiment, the data prediction error of each of the at least one historical period is actually a difference between the predicted traffic flow and the actual traffic flow at the time point in each of the historical periods corresponding to the next time point in step 101 in the above embodiment. The preset error range is an error value distribution interval obtained according to the data prediction error of each historical period, and this distribution interval can be selected based on error ratios of less than 20% according to experience. For example, in each of the historical periods, the predicted traffic flow value at the next time point is A, and the actual traffic flow value is B, then the data prediction error ratio is (A−B)/A. Then, the data prediction error ratios whose values is less than 20% are counted, and the preset error range is determined according to upper and lower limits of these data error values.
Before determining the target number, since there may be a large amount of historical data before the target period, in order to achieve a better matching degree between the traffic data of each historical period and the traffic data of the target period, according to experience, the time interval between each historical period and the target period is selected to be no more than 20 days at most. Then, starting from the target period, traffic data of 20 periods are sequentially acquired, and the data prediction error of these traffic data is compared with the preset error range, and the historical periods having the data prediction error within the preset error range are selected. The number of these historical periods that meet the requirements is counted to obtain the target number.
According to the target number, starting from the historical period closest to the target period, the target number of historical periods are selected as the reference periods. The selection of the reference period may include two schemes.
In the first scheme, the historical period that meets the requirement is determined as the reference period. For example, the requirement may be: the similarity between the historical period and the target period being greater than or equal to a certain threshold, or the prediction error being less than or equal to a certain error limit.
In the second scheme, starting from the historical period closest to the target period, a target number of historical periods are successively selected as the reference periods.
In the first scheme, the data prediction errors of the reference periods are all within the preset error range, and the reference periods may be discretely or adjacently arranged in the historical time (e.g., the number of days). In the second scheme, the reference periods are only adjacent to each other, and are the target number of historical periods adjacent to the target period.
In step 203, a first state vector of the target period is determined based on the traffic data at a plurality of first time points in the target period.
In the state vector X=[v(t) . . . v(t−3), o(t) . . . o(t−3)] of the target period, for flow v(t), and saturation o(t), when calculating the similarity between the current period data and the historical period data, after weighing various influence factors, the influence factor with the highest correlation coefficient between periods is determined as the indicator of the state vector, that is, flow and saturation. Therefore, the traffic data of the subsequent target period predicted according to the flow and saturation in the historical period is more accurate.
In step 204, a second state vector of the reference period is determined based on traffic data at time points corresponding to the plurality of first time points in the reference period.
The state vector of the reference period is Xh=[vh(t) . . . vh(t−3), oh (t) . . . oh(t−3)]. In step 205, the data similarity between the target period and the reference period is determined based on a vector distance between the first state vector and the second state vector.
In this embodiment, the calculation method of the vector distance is shown in step 103 of the above embodiment, and the principle is the same, which will not be repeated here. Since the target number of reference periods need to be selected, the target number of vector distances need to be calculated.
In step 206, reference traffic data at a time point in the reference period is obtained, the time point in the reference period being corresponding to a second time point to be predicted in the target period.
In this embodiment, specific reference may be made to step 104 in the above embodiment, and the principle is the same, which will not be repeated here.
In step 207, traffic data at the second time point in the target period is predicted based on the data similarity between the target period and the reference period, and the reference traffic data in the reference period.
Optionally, in a possible implementation, the weight of the reference period is determined according to the data similarity between the target period and the reference period. The weight has a negative correlation with the data similarity. According to the weight of the reference period, the reference traffic data in the reference period is weighted to predict the traffic data of the second time point in the target period.
In this embodiment, for the weight of the reference period and the prediction formula and principle of the traffic data at the second time point in the target period, reference may be made to the relevant statement in step 105 in the above embodiment, which will not be repeated here. It is worth noting that, the reference traffic data in the reference period used in this embodiment is real-time traffic flow data at a time point corresponding to the second time point in the target period.
In addition, since this embodiment does not perform any processing on the original data in the reference period, the characteristics of the original data are maintained, which is not only beneficial to the authenticity of the predicted data, but also easy to operate. Since the traffic data may contain other influence factors, other influence factors can also be predicted by using the method for predicting traffic data described in the disclosure, which is not limited herein.
With the method for predicting traffic data according to embodiments of the disclosure, traffic data in a plurality of periods is obtained. The plurality of periods includes the target period and at least one historical period before the target period. The reference periods are determined from the at least one historical period based on a time interval between each of the at least one historical period and the target period. The data similarity between the target period and the reference period is determined based on traffic data at the first time point in the target period and traffic data at the time point corresponding to the first time point in the reference period. Reference traffic data at the time point in the reference period is obtained, the time point in the reference period being corresponding to the second time point to be predicted in the target period. Traffic data at the second time point in the target period is predicted based on the data similarity between the target period and the reference period, and the reference traffic data in the reference period. The traffic data in the reference period that meets the requirements is selected from the historical period by comparing the traffic data in the historical period with the target period data. Therefore, the prediction of the traffic data in the target period is realized, which not only maintains the characteristics of the traffic data in the historical period, but also ensure easy operation and high prediction accuracy.
In order to realize the above embodiments, the disclosure also provides an apparatus for predicting traffic data.
As shown in
The obtaining module 31 is configured to obtain traffic data in a plurality of periods, the plurality of periods includes a target period and at least one historical period before the target period.
The screening module 32 is configured to determine a reference period from the at least one historical period based on a time interval between each of the at least one historical period and the target period.
The executing module 33 is configured to determine a data similarity between the target period and the reference period based on traffic data at a first time point in the target period and traffic data at a time point corresponding to the first time point in the reference period.
The matching module 34 is configured to obtain reference traffic data at a time point in the reference period, the time point in the reference period corresponds to a second time point to be predicted in the target period.
The predicting module 35 is configured to predict traffic data at the second time point in the target period based on the data similarity between the target period and the reference period, and the reference traffic data in the reference period.
Further, in a possible implementation, the executing module 33 also includes: a first configuring unit 331, a second configuring unit 332 and a first executing unit 333.
The first configuring unit 331 is configured to determine a first state vector of the target period based on the traffic data at a plurality of first time points in the target period.
The second configuring unit 332 is configured to determine a second state vector of the reference period based on traffic data at time points corresponding to the plurality of first time points in the reference period.
The first executing unit 333 is configured to determine the data similarity between the target period and the reference period based on a vector distance between the first state vector and the second state vector.
Further, in a possible implementation, the predicting module 35 also includes: a second executing unit 351 and a predicting unit 352.
The second executing unit 351 is configured to determine a weight of the reference period based on the data similarity between the target period and the reference period, in which the weight has a negative correlation with the data similarity.
The predicting unit 352 is configured to predict the traffic data at the second time point in the target period by weighting the reference traffic data in the reference period based on the weight of the reference period.
In a possible implementation, the screening module 32 is further configured to:
obtain a data prediction error of each of the at least one historical period; and
determine the target number based on a number of historical periods having data prediction error within a preset error range in the at least one historical period.
In a possible implementation, the first executing unit 333 is further configured to:
determine a first correlation coefficient of the flow between historical periods based on the flow of each of the at least one historical period;
determine a second correlation coefficient of the saturation between historical periods based on the saturation of each of the at least one historical period; and
determine the vector distance based on a difference corresponding to the flow between the first state vector and the second state vector, a difference corresponding to the saturation between the first state vector and the second state vector, the first correlation coefficient and the second correlation coefficient.
With the apparatus for predicting traffic data according to the embodiments of the disclosure, traffic data in a plurality of periods is obtained. The plurality of periods includes a target period and at least one historical period before the target period. A reference period is determined from the at least one historical period based on a time interval between each of the at least one historical period and the target period. A data similarity between the target period and the reference period is determined based on traffic data at a first time point in the target period and traffic data at a time point corresponding to the first time point in the reference period. Reference traffic data at a time point in the reference period is obtained, the time point in the reference period corresponds to a second time point to be predicted in the target period. Traffic data at the second time point in the target period is predicted based on the data similarity between the target period and the reference period, and the reference traffic data in the reference period. The traffic data in the reference period that meets the requirements is selected from the historical period by comparing the traffic data in the historical period with the target period data. Therefore, the prediction of the traffic data in the target period is realized, which not only maintains the characteristics of the traffic data in the historical period, but also ensure easy operation and high prediction accuracy.
It should be noted that the foregoing explanation on the embodiments of the method for predicting traffic data is applicable to the apparatus for predicting traffic data of this embodiment, which are not repeated here.
According to embodiments of the disclosure, the disclosure also provides an electronic device, a readable storage medium and a computer program product.
As illustrated in
Components in the device 400 are connected to the I/O interface 405, including: an inputting unit 406, such as a keyboard, a mouse; an outputting unit 407, such as various types of displays, speakers; a storage unit 408, such as a disk, an optical disk; and a communication unit 409, such as network cards, modems, wireless communication transceivers, and the like. The communication unit 409 allows the device 400 to exchange information/data with other devices through a computer network such as the Internet and/or various telecommunication networks.
The computing unit 401 may be various general-purpose and/or dedicated processing components with processing and computing capabilities. Some examples of computing unit 401 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units that run machine learning model algorithms, and a digital signal processor (DSP), and any appropriate processor, controller and microcontroller. The computing unit 401 executes the various methods and processes described above, such as the method for predicting traffic data. For example, in some embodiments, the method may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as the storage unit 408. In some embodiments, part or all of the computer program may be loaded and/or installed on the device 400 via the ROM 402 and/or the communication unit 409. When the computer program is loaded on the RAM 403 and executed by the computing unit 401, one or more steps of the method described above may be executed. Alternatively, in other embodiments, the computing unit 401 may be configured to perform the method for predicting traffic data in any other suitable manner (for example, by means of firmware).
Various implementations of the systems and techniques described above may be implemented by a digital electronic circuit system, an integrated circuit system, Field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), System on Chip (SOCs), Load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or a combination thereof. These various embodiments may be implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general programmable processor for receiving data and instructions from the storage system, at least one input device and at least one output device, and transmitting the data and instructions to the storage system, the at least one input device and the at least one output device.
The program code configured to implement the method of the disclosure may be written in any combination of one or more programming languages. These program codes may be provided to the processors or controllers of general-purpose computers, dedicated computers, or other programmable data processing devices, so that the program codes, when executed by the processors or controllers, enable the functions/operations specified in the flowchart and/or block diagram to be implemented. The program code may be executed entirely on the machine, partly executed on the machine, partly executed on the machine and partly executed on the remote machine as an independent software package, or entirely executed on the remote machine or server.
In the context of the disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), electrically programmable read-only-memory (EPROM), flash memory, fiber optics, compact disc read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.
In order to provide interaction with a user, the systems and techniques described herein may be implemented on a computer having a display device (e.g., a Cathode Ray Tube (CRT) or a Liquid Crystal Display (LCD) monitor for displaying information to a user); and a keyboard and pointing device (such as a mouse or trackball) through which the user can provide input to the computer. Other kinds of devices may also be used to provide interaction with the user. For example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or haptic feedback), and the input from the user may be received in any form (including acoustic input, voice input, or tactile input).
The systems and technologies described herein can be implemented in a computing system that includes background components (for example, a data server), or a computing system that includes middleware components (for example, an application server), or a computing system that includes front-end components (for example, a user computer with a graphical user interface or a web browser, through which the user can interact with the implementation of the systems and technologies described herein), or include such background components, intermediate computing components, or any combination of front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local area network (LAN), wide area network (WAN), the Internet and Block-chain network.
The computer system may include a client and a server. The client and server are generally remote from each other and interacting through a communication network. The client-server relation is generated by computer programs running on the respective computers and having a client-server relation with each other. The server can be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system, to solve the traditional physical host with a Virtual Private Server (VPS) service, which has the defects of difficult management and weak business expansibility. The server can also be a server for a distributed system, or a server that incorporates a block-chain.
It is noted that artificial intelligence is a discipline that studies certain thinking processes and intelligent behaviors (such as learning, reasoning, thinking and planning) that allow computers to simulate life, which has both hardware-level technologies and software-level technologies. Artificial intelligence hardware technology generally includes technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, and big data processing. Artificial intelligence software technology generally includes computer vision technology, speech recognition technology, natural language processing technology, and its learning/deep learning, big data processing technology, knowledge map technology and other aspects.
It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the disclosure could be performed in parallel, sequentially, or in a different order, as long as the desired result of the technical solution disclosed in the disclosure is achieved, which is not limited herein.
The above specific embodiments do not constitute a limitation on the protection scope of the disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent replacement and improvement made within the spirit and principle of the disclosure shall be included in the protection scope of the disclosure.
| Number | Date | Country | Kind |
|---|---|---|---|
| 202110677845.4 | Jun 2021 | CN | national |
