CN117391257A - Road congestion condition prediction method and device - Google Patents
Road congestion condition prediction method and device Download PDFInfo
- Publication number
- CN117391257A CN117391257A CN202311506953.0A CN202311506953A CN117391257A CN 117391257 A CN117391257 A CN 117391257A CN 202311506953 A CN202311506953 A CN 202311506953A CN 117391257 A CN117391257 A CN 117391257A
- Authority
- CN
- China
- Prior art keywords
- congestion
- data
- target area
- related data
- prediction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 74
- 230000008569 process Effects 0.000 claims abstract description 33
- 238000004088 simulation Methods 0.000 claims abstract description 28
- 238000012545 processing Methods 0.000 claims description 51
- 238000000605 extraction Methods 0.000 claims description 38
- 238000010801 machine learning Methods 0.000 claims description 29
- 238000012549 training Methods 0.000 claims description 16
- 238000005259 measurement Methods 0.000 claims description 15
- 238000003066 decision tree Methods 0.000 claims description 10
- 238000007781 pre-processing Methods 0.000 claims description 9
- 230000002159 abnormal effect Effects 0.000 claims description 6
- 125000004122 cyclic group Chemical group 0.000 claims description 5
- 230000000694 effects Effects 0.000 claims description 5
- 238000003062 neural network model Methods 0.000 claims description 5
- 238000010586 diagram Methods 0.000 description 6
- 238000007726 management method Methods 0.000 description 4
- 238000010606 normalization Methods 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 238000010219 correlation analysis Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000009499 grossing Methods 0.000 description 3
- 230000015654 memory Effects 0.000 description 3
- 238000007405 data analysis Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003032 molecular docking Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000013138 pruning Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 230000026676 system process Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
- G08G1/0129—Traffic data processing for creating historical data or processing based on historical data
Abstract
The application provides a prediction method and a prediction device for road congestion conditions, wherein the prediction method comprises the steps of obtaining operation related data of unmanned vehicles arranged in a target area in a transportation process; determining a target feature for predicting road congestion conditions of a target area based on the operation-related data; and obtaining a congestion prediction result of the target area based on the target characteristics. Therefore, through designing a prediction model of the road congestion in the industrial park trained by the measured data and the simulation data, the external and internal relevant characteristics of the road congestion in the industrial park are extracted, the prediction model is trained by utilizing the characteristic data to obtain a congestion prediction result, and the accuracy of the prediction of the road congestion condition in the industrial park is improved.
Description
Technical Field
The present disclosure relates to the field of machine learning technologies, and in particular, to a method and an apparatus for predicting a road congestion condition.
Background
Along with the continuous development of unmanned vehicles technology, in unmanned vehicles transportation scheduling scene in industrial park, unmanned vehicles transportation scheduling's demand is more and more, and then, to solving the demand of the road jam condition in the industrial park also being increasing. In order to alleviate the condition of the congestion of the unmanned vehicles in the industrial park, warning signals are generally sent to the unmanned vehicles through a dispatching system, or the external vehicles are guided through setting traffic lights, and in addition, a method for improving dispatching efficiency through reference transportation path planning provided by the dispatching system is provided. However, since the circulation route of the unmanned vehicles in the park is simpler, the route selection is less, the transportation efficiency is lower, and the increase of the beat number is continuously slowed down and then reduced along with the increase of vehicles, the prediction of the congestion condition of the roads of the industrial park is of practical significance.
The existing map software is mainly used for predicting road traffic by means of taxi GPS positioning, personal mobile phone positioning, traffic department data docking and the like, and the prediction methods are mainly aimed at open roads, but are not effective for predicting roads of an industrial park, so that the accuracy of predicting road congestion conditions of the industrial park is lower.
Disclosure of Invention
In view of this, the present application aims to provide a method and a device for predicting road congestion, which are designed to train a prediction model of road congestion in an industrial park according to measured data and simulation data, extract relevant features of the outside world and the inside of a system of the road congestion in the industrial park, train the prediction model according to feature data to obtain a congestion prediction result, and improve the accuracy of prediction of the road congestion in the industrial park.
The embodiment of the application provides a method for predicting road congestion, which comprises the following steps:
acquiring operation related data of an unmanned vehicle arranged in a target area in the transportation process;
determining a target feature for predicting road congestion conditions of a target area based on the operation-related data;
and obtaining a congestion prediction result of the target area based on the target characteristics.
Further, the operation related data comprise actual measurement related data and preset related data of the unmanned vehicle arranged in the target area in the transportation process; the step of determining a target feature for predicting a road congestion condition of a target area based on the operation-related data includes:
preprocessing operation related data to obtain calibration operation related data;
normalizing the actual measurement related data in the calibration operation related data to obtain actual measurement operation characteristic data; performing feature extraction processing on the actually measured operation feature data to obtain actually measured operation related features used for representing external influences on road congestion of a target area;
carrying out coding processing on preset related data in the calibration operation related data to obtain scheduling task characteristic data; and carrying out feature extraction processing on the dispatching task feature data to obtain dispatching task related features used for representing internal influence of the vehicle transportation task on the road congestion of the target area.
Further, the actually measured operation characteristic data comprise traffic flow data, weather data and time sequence data in a target area; the scheduled task feature data includes vehicle scheduled task data and vehicle parameter data within a target area.
Further, the preprocessing includes at least one of: correcting the abnormal value in the operation related data, deleting the repeated value in the operation related data, and filling the missing value in the operation related data.
Further, the step of performing feature extraction processing on the actually measured operation feature data to obtain actually measured operation related features for characterizing an external influence on road congestion in a target area includes:
extracting traffic flow characteristics and historical congestion state characteristics from the traffic flow data through a time sequence processing model; the time sequence processing model is a cyclic neural network model which is obtained by machine learning training in advance by using a plurality of groups of operation related data of a target area with a sequence relation;
extracting weather state features from the weather data;
extracting time sequence features and holiday features from the time sequence data.
Further, the step of performing feature extraction processing on the feature data of the scheduling task to obtain a scheduling related feature for reflecting an internal influence of a vehicle transportation task on road congestion in a target area, includes:
Extracting vehicle dispatching task information characteristics from the vehicle dispatching task data through a congestion prediction model; the congestion prediction model is an integrated machine learning framework based on a decision tree, and is a gradient lifting prediction model obtained by training operation related data of a target area in advance;
and extracting vehicle parameter information features from the vehicle parameter data.
Further, the step of obtaining a congestion prediction result of the target area based on the target feature includes:
inputting the target characteristics into a congestion prediction model;
the congestion prediction model is based on an integrated machine learning framework of a decision tree, and the prediction probability of the road congestion of the target area is obtained by carrying out iterative prediction on the target characteristics;
and determining and outputting a congestion prediction result for representing the road congestion condition of the target area in a preset time interval by comparing the prediction probability with a preset congestion threshold value of the target area.
Further, the prediction method further includes:
acquiring simulated congestion characteristics in a dispatching system for managing unmanned vehicle transportation processes in a target area;
the congestion prediction model is based on the congestion prediction result, and the updated congestion prediction result is output by combining the simulated congestion characteristics; the updated congestion prediction result is used for improving the accuracy of the congestion prediction result in predicting the road congestion condition of the target area.
Further, the step of obtaining the simulated congestion feature of the unmanned vehicle for managing the target area in the dispatching system of the transportation process includes:
the dispatching system outputs simulated congestion characteristic data which influences road congestion conditions in a target area according to the operation related data;
and performing machine learning simulation on the simulated congestion feature data by a preset simulation prediction model in the scheduling system, and outputting simulated congestion features for representing simulated congestion conditions of the roads in the target area.
The embodiment of the application also provides a device for road congestion, and the predicting device comprises:
the data acquisition module is used for acquiring operation related data of the unmanned vehicle arranged in the target area in the transportation process;
the feature extraction module is used for determining target features for predicting road congestion conditions of a target area based on the operation related data;
and the congestion prediction module is used for obtaining a congestion prediction result of the target area based on the target characteristics.
Further, the feature extraction module is configured to, when determining, based on the operation-related data, a target feature for predicting a road congestion condition of a target area, the feature extraction module is configured to:
Preprocessing operation related data to obtain calibration operation related data;
normalizing the actual measurement related data in the calibration operation related data to obtain actual measurement operation characteristic data; performing feature extraction processing on the actually measured operation feature data to obtain actually measured operation related features used for representing external influences on road congestion of a target area;
carrying out coding processing on preset related data in the calibration operation related data to obtain scheduling task characteristic data; and carrying out feature extraction processing on the dispatching task feature data to obtain dispatching task related features used for representing internal influence of the vehicle transportation task on the road congestion of the target area.
Further, the feature extraction module is configured to perform at least one of the following when configured to pre-process the operation-related data: correcting the abnormal value in the operation related data, deleting the repeated value in the operation related data, and filling the missing value in the operation related data.
Further, when the feature extraction module is configured to perform feature extraction processing on the actually measured operation feature data to obtain a step of characterizing an actually measured operation related feature that has an external influence on road congestion in a target area, the feature extraction module is configured to:
Extracting traffic flow characteristics and historical congestion state characteristics from the traffic flow data through a time sequence processing model; the time sequence processing model is a cyclic neural network model which is obtained by machine learning training in advance by using a plurality of groups of operation related data of a target area with a sequence relation;
extracting weather state features from the weather data;
extracting time sequence features and holiday features from the time sequence data.
Further, when the feature extraction module is configured to perform feature extraction processing on the feature data of the scheduled task to obtain a scheduled task related feature for reflecting an internal influence of a vehicle transportation task on road congestion in a target area, the feature extraction module is configured to:
extracting vehicle dispatching task information characteristics from the vehicle dispatching task data through a congestion prediction model; the congestion prediction model is an integrated machine learning framework based on a decision tree, and is a gradient lifting prediction model obtained by training operation related data of a target area in advance;
and extracting vehicle parameter information features from the vehicle parameter data.
Further, when the congestion prediction module is used for obtaining a congestion prediction result of the target area based on the target feature, the congestion prediction module is used for:
Inputting the target characteristics into a congestion prediction model;
the congestion prediction model is based on an integrated machine learning framework of a decision tree, and the prediction probability of the road congestion of the target area is obtained by carrying out iterative prediction on the target characteristics;
and determining and outputting a congestion prediction result for representing the road congestion condition of the target area in a preset time interval by comparing the prediction probability with a preset congestion threshold value of the target area.
Further, the prediction device further comprises a simulated congestion module, wherein the simulated congestion module is used for:
a simulated congestion feature in a dispatch system for managing unmanned vehicle transportation in a target area is obtained.
Further, the congestion prediction module is configured to:
the congestion prediction model is based on the congestion prediction result, and the updated congestion prediction result is output by combining the simulated congestion characteristics; the updated congestion prediction result is used for improving the accuracy of the congestion prediction result in predicting the road congestion condition of the target area.
Further, in the step of acquiring the simulated congestion characteristics of the unmanned vehicle for managing the target area in the dispatching system of the transportation process, the simulated congestion module is configured to:
The dispatching system outputs simulated congestion characteristic data which influences road congestion conditions in a target area according to the operation related data;
and performing machine learning simulation on the simulated congestion feature data by a preset simulation prediction model in the scheduling system, and outputting simulated congestion features for representing simulated congestion conditions of the roads in the target area.
The embodiment of the application provides a method and a device for predicting road congestion, wherein the method for predicting road congestion comprises the following steps: acquiring operation related data of an unmanned vehicle arranged in a target area in the transportation process; determining a target feature for predicting road congestion conditions of a target area based on the operation-related data; and obtaining a congestion prediction result of the target area based on the target characteristics.
Compared with the method for predicting the road congestion by the scheduling system sending warning signals to the unmanned vehicles through the scheduling system, the method for predicting the road congestion by the common map software provided by the scheduling system, the method for predicting the road congestion in the industrial park by designing the prediction model of the road congestion in the industrial park trained by actual measurement data and simulation data, extracting the external and internal relevant characteristics of the road congestion in the industrial park and utilizing the characteristic data to train the prediction model to obtain a congestion prediction result, and improving the accuracy of the prediction of the road congestion condition in the industrial park.
In order to make the above objects, features and advantages of the present application more comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered limiting the scope, and that other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a method for predicting road congestion according to an embodiment of the present application;
fig. 2 is a schematic structural diagram of a prediction system for road congestion according to an embodiment of the present application;
fig. 3 is a schematic structural diagram of a road congestion device according to an embodiment of the present disclosure;
fig. 4 is a second schematic structural diagram of a road congestion situation device according to an embodiment of the present disclosure.
Detailed Description
For the purposes of making the objects, technical solutions and advantages of the embodiments of the present application more clear, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is apparent that the described embodiments are only some embodiments of the present application, but not all embodiments. The components of the embodiments of the present application, which are generally described and illustrated in the figures herein, may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present application, as provided in the accompanying drawings, is not intended to limit the scope of the application, as claimed, but is merely representative of selected embodiments of the application. Based on the embodiments of the present application, every other embodiment that a person skilled in the art would obtain without making any inventive effort is within the scope of protection of the present application.
It has been found that, in order to alleviate the congestion of the unmanned vehicles in the industrial park, a warning signal is generally sent to the unmanned vehicles through a dispatching system, or the external vehicles are guided through setting traffic lights, and in addition, a method for improving dispatching efficiency by providing a reference transportation path plan through the dispatching system is provided. However, since the circulation route of the unmanned vehicles in the park is simpler, the route selection is less, the transportation efficiency is lower, and the increase of the beat number is continuously slowed down and then reduced along with the increase of vehicles, the prediction of the congestion condition of the roads of the industrial park is of practical significance.
Based on the above, the embodiment of the application provides a prediction method for road congestion, by designing a prediction model of road congestion in an industrial park trained by actual measurement data and simulation data, extracting the external and internal relevant characteristics of the road congestion of the industrial park, and training the prediction model by using characteristic data to obtain a congestion prediction result, thereby improving the accuracy of prediction for the road congestion of the industrial park.
Referring to fig. 1, fig. 1 is a flowchart of a method for predicting a road congestion situation according to an embodiment of the present application. As shown in fig. 1, the method for predicting road congestion provided in the embodiment of the present application includes:
S101, acquiring operation related data of an unmanned vehicle arranged in a target area in the transportation process.
Here, the target area includes, but is not limited to, a logistic park or an industrial park. The operation related data comprise actual measurement related data and preset related data of the unmanned vehicle arranged in the target area in the transportation process. The actually measured related data are collected data of objective conditions under the external environment of the target area, such as traffic flow, weather conditions, seasons, time, date, holidays and the like of the target area; the preset related data are internal controllable data about scheduled tasks and unmanned vehicles preset in a scheduled task management system set in a target area by a user, such as data of the number of unmanned vehicles on the way, the number of carrying tasks, the task priority, the electric quantity of the unmanned vehicles, the number of goods and the like.
In the step, sensors or cameras are arranged on roads in a target area, and traffic data including information of the number, speed, position and the like of vehicles are collected; collecting weather forecast data on the Internet, or additionally arranging a small weather station in a logistics park to obtain real-time weather data; the sensor of the unmanned vehicle is utilized to be connected with a dispatching task management system in real time so as to acquire relevant data of the unmanned vehicle, such as a radar, a camera, a laser radar and the like; in the scheduling task management system, the preset scheduling task related data in the system can be obtained.
S102, determining target characteristics for predicting road congestion conditions of a target area based on the operation related data.
The target features are to remove irrelevant features and redundant features through a feature selection method, obtain an optimal feature subset which is conducive to machine learning of a model, and have important significance in improving prediction performance and reducing system resources. Through analysis and understanding of the data, features with high correlation with the target variable can be selected, so that the dimension of the features can be reduced, the training efficiency of the model is improved, and the risk of overfitting of the model is reduced.
Here, suitable characteristics in the operation-related data, such as the number of vehicles, speed, weather conditions, etc., time-related characteristics, such as hours, dates, seasons, etc., and statistical characteristics in the operation-related data, such as mean values, standard deviations, etc., are selected, the historical data is analyzed by machine learning and data mining, and the characteristics most relevant to road congestion are screened and extracted by correlation analysis.
In one embodiment of the present application, the step of determining the target feature for predicting the road congestion condition of the target area in step S102 may include:
S1021, preprocessing the operation related data to obtain calibration operation related data.
In this step, since the data is often incomplete, unsmooth, and highly vulnerable to outliers during the data model training process, and because the data volume is too large and the data sets are often from a variety of data sources, low quality data will result in low quality mining training results, preprocessing the running-related data is a critical process prior to machine learning of the model.
Here, in the implementation process, the process of preprocessing the operation-related data is completed through the data smoothing process and the data interpolation process.
Specifically, the data smoothing process is to correct the abnormal value in the operation related data and delete the repeated value in the operation related data, so as to achieve the technical effect of data smoothing. And carrying out average weighting processing on the data values before and after the time series data so as to correct the abnormal value in the data and delete the repeated value, thereby obtaining the smooth data at the current moment.
In particular, the data interpolation process is to fill up missing values in the operation related data, and the time series data of the operation related data contains a large amount of information.
S1022, carrying out normalization processing on actual measurement related data in the calibration operation related data to obtain actual measurement operation characteristic data; and carrying out feature extraction processing on the actually measured operation feature data to obtain actually measured operation related features used for representing external influence on road congestion of a target area.
In the step, normalization processing is performed on the actually measured related data, so that the problem that the accuracy of a result is affected due to the fact that data analysis and modeling are complicated aiming at data with different indexes and different magnitudes in data processing and data analysis is solved, the magnitudes of all indexes are unified through normalization processing of the data, and meanwhile, the data characteristics are scaled to a reasonable range. Here, in the implementation process of the present application, the data in the original range is rescaled by the maximum and minimum normalization method, so that all values are in the range of 0 and 1.
The measured operation characteristic data comprises traffic flow data, weather data and time sequence data in a target area, wherein the three data are the classified measured operation characteristic data through correlation analysis of calibration operation related data. The measured operational characteristic data is traffic flow data and weather data in the target area represented in time series, and includes objective time series data in the target area.
Specifically, for example, the traffic flow data may include information such as the degree of traffic flow of an unmanned vehicle, the speed and the position of the unmanned vehicle, etc. on a certain road in the target area; the weather data may include temperature, weather, season, etc. information of the current target area; the time series data may include information of a current date, time of day, solar terms, and whether it is holidays, etc.
In one embodiment of the present application, in step S1022, the step of performing feature extraction processing on the measured operation feature data to obtain a measured operation related feature for characterizing an external impact on road congestion in a target area, includes:
s10221, extracting traffic flow characteristics and historical congestion state characteristics from the traffic flow data through a time sequence processing model; the time sequence processing model is a cyclic neural network model which is obtained by machine learning training through the operation related data of a plurality of groups of target areas with sequence relations in advance.
Here, the time series processing model is an LSTM (Long-short Time Memory) model, which is a variant of a Recurrent Neural Network (RNN) adapted to process time series data, which is capable of capturing Long-term dependency and has a memory unit to process time series information in a time series, and the LSTM model is excellent in a scenario of processing complex nonlinear relationships and Long-term memory, is affected by a plurality of factors due to congestion conditions, and has complex time series dependency, and is very practical in solving the above problems.
In the step, characteristics of the unmanned vehicle flow of each time period in the target area of the unmanned vehicle are analyzed and found out through a time sequence processing model, for example, characteristics that the unmanned vehicle flow is large in the early peak time period, and roads in the upper week or the previous day target area are very congested and the like are obtained. According to the characteristics, the traffic flow characteristics and the historical congestion state characteristics are extracted from the traffic flow data.
S10222, extracting weather state features from the weather data.
In the step, the influence of weather on road congestion generated by unmanned vehicle transportation in a target area is analyzed from weather data through a time sequence processing model, for example, in rainy days, in order to avoid damage of the unmanned vehicles, unmanned vehicles are put in a small amount in a park, and then the congestion degree of the target area is small. According to the characteristics, the weather state characteristics are extracted from the weather data.
S10223, extracting time sequence features and holiday features from the time sequence data.
In the step, the time sequence processing model is used for analyzing the congestion condition of a target area in various time periods or holidays, and finding out the influence relationship of time on road congestion, for example, the quantity of transportation tasks on weekends is large, so that the number of unmanned vehicles put in is increased, and the road congestion of the target area is caused; the transportation task of legal holidays is also difficult, and the congestion situation of the road in the target area is more prominent. According to the above features, the time series features and holiday features are extracted from the time series data.
S1023, carrying out coding processing on preset related data in the calibration operation related data to obtain scheduling task feature data; and carrying out feature extraction processing on the dispatching task feature data to obtain dispatching task related features used for representing internal influence of the vehicle transportation task on the road congestion of the target area.
It should be noted that the purpose of the encoding process of the preset related data in the calibration operation related data is to convert the target variable that is not numerical into numerical for use in the machine learning model.
In the step, for the preset related data, the data of different vehicle dispatching tasks are subjected to natural number coding, the time is divided into one time according to each half hour in the characteristic time period, so that the time period information is coded into 1-48, the characteristic and vehicle parameter data in the dispatching tasks are subjected to sequential coding, and after all the data are subjected to coding processing, all the data are quantized and then spliced.
The dispatching task characteristic data comprise vehicle dispatching task data and vehicle parameter data in a target area. The two data are the characteristic data of the dispatching task which are classified by the correlation analysis of the calibration operation related data. The scheduling task feature data is related data in the target area to schedule tasks according to the unmanned vehicles in the scheduling management system.
Specifically, for example, the vehicle dispatch task data may include the number of unmanned vehicles in operation, the number of handling tasks, task priority, cargo weight, and the like; the vehicle parameter data may include data such as vehicle set-up power and vehicle speed.
In one embodiment of the present application, the step of performing feature extraction processing on the dispatch task feature data in step S1023 to obtain a dispatch task related feature for reflecting an internal influence of a vehicle transportation task on a road congestion in a target area, includes:
s10231, extracting vehicle dispatching task information features from the vehicle dispatching task data through a congestion prediction model; the congestion prediction model is an integrated machine learning framework based on a decision tree, and is a gradient lifting prediction model obtained by training operation related data of a target area in advance.
Here, the congestion prediction model is a XGBoost (Extreme Gradient Boost) model, which is a gradient lifting tree model suitable for processing structured data and feature engineering, and is more suitable for processing abundant traffic data and related features, and is essentially a method based on a tree structure and combined with ensemble learning, and the basic tree structure is a classification regression tree, similar to a local weighted linear regression algorithm.
In the step, structural features of unmanned vehicles in a target area are analyzed and found out through a congestion prediction model from the vehicle dispatching task data, for example, the number of unmanned vehicles in work is excessive, the number of carrying tasks is excessive, road congestion is caused, and the congestion influence caused by low task priority of some long distances is avoided. According to the characteristics, the vehicle dispatching task information characteristics are extracted from the vehicle dispatching task data.
S10232, extracting vehicle parameter information features from the vehicle parameter data.
In the step, structural features of the unmanned vehicle in a target area are analyzed and found out through a congestion prediction model from the vehicle parameter data, for example, the congestion influence caused by too low speed due to too large load cargo weight of the unmanned vehicle and low electric quantity of the unmanned vehicle is obtained. According to the above feature, the vehicle parameter information feature is extracted from the vehicle parameter data.
S103, obtaining a congestion prediction result of the target area based on the target characteristics.
In the step, the congestion prediction result of the target area is obtained through machine learning iterative simulation of the model, and the congestion probability of a future period is compared with a preset congestion threshold value to generate the congestion prediction result.
In one embodiment of the present application, the step of obtaining the congestion prediction result of the target area in step S103 includes:
s1031, inputting the target characteristics into a congestion prediction model.
In the step, the traffic flow characteristics, the historical congestion state characteristics, the weather state characteristics, the time sequence characteristics, the holiday characteristics, the vehicle scheduling task information characteristics, the vehicle parameter information characteristics and other target characteristics obtained through the steps are input into the congestion prediction model, and are used for iterative operation simulation of the model to obtain a congestion prediction result of a target area.
S1032, the congestion prediction model is based on an integrated machine learning framework of a decision tree, and the prediction probability of the road congestion of the target area is obtained through iterative prediction of the target features.
In the step, firstly, a CART tree is used as a base learner to generate a first tree, then, pruning is carried out on the tree to construct an XGBoost model, then, a negative gradient value of the output of the last CART tree in a loss function is calculated to be used as an approximate value of the loss of the round, the gradient value is used as a training data label value of the next CART tree, a new CART tree is generated through iteration until the model converges, and finally, a cross verification method is adopted to obtain the prediction probability of road congestion about a target area in the XGBoost model.
Here, the predicted probability of the road congestion in the target area is obtained as a probability value converted into a percentage, the probability value is unequal from 1% to 99%, and the probability value takes the decimal point next bit when being output.
S1033, determining and outputting a congestion prediction result for representing the road congestion condition of the target area in a preset time interval through comparing the prediction probability with a preset congestion threshold value of the target area.
In the step, according to the congestion effect generated by the congestion probability, the congestion situation is divided into four situations of 'smooth', 'comparatively congested' and 'congested', and a congestion probability threshold corresponding to the congestion situation is set, for example, when the predicted probability of the road congestion in the target area is 1% to 25%, the output result is 'smooth'; when the predicted probability of the road congestion of the target area is 26-50%, outputting a result of 'smooth comparison'; when the prediction probability of the road congestion of the target area is 51-75%, outputting a result of 'comparative congestion'; when the predicted probability of the road congestion in the target area is 76% to 99%, the output result is "congestion".
Optionally, the method for predicting the road congestion situation includes steps S104 and S105 in addition to the steps S101 to S103, specifically, steps S104 and S105 are used to describe a method for outputting an updated congestion prediction result by simulating the characteristic simulation of congestion data, which is beneficial to improving the accuracy of predicting the road congestion situation of the target area by using the congestion prediction result.
Here, specific steps of steps S101 to S103 are described above, and are not described here again.
S104, obtaining simulated congestion characteristics in a dispatching system for managing unmanned vehicle transportation processes in a target area.
It should be noted that, for the machine learning model, the quality and the quantity of the data volume are required to be very high, here we innovatively add the simulation data of the dispatching system, so that the data volume is greatly increased, then the accuracy of the simulation prediction is continuously improved along with the time, and the accuracy of the road congestion prediction is effectively improved.
In one embodiment of the present application, the step of obtaining a simulated congestion feature in a dispatch system for managing unmanned vehicle transportation in a target area in step S104 includes:
s1041, the scheduling system outputs simulated congestion feature data which influences road congestion conditions in a target area according to the operation related data.
In the step, in order to solve the problem of poor prediction effect when the data amount is not abundant in the initial stage, a scheduling system is used for simulating simulated congestion feature data influencing the road congestion condition in a target area according to traffic flow data, weather data and time sequence data in actual measurement operation feature data and vehicle scheduling task data and vehicle parameter data in scheduling task feature data by combining the congestion condition feature.
S1042, carrying out machine learning simulation on the simulated congestion feature data by a preset simulation prediction model in the dispatching system, and outputting simulated congestion features for representing simulated congestion conditions of the roads in the target area.
In the step, the preset simulation prediction model is a simulation prediction model trained through machine learning according to the simulated congestion feature data, and the preset simulation prediction model is integrated in the scheduling system. And carrying out machine learning simulation processing on the simulated congestion feature data by a preset simulation prediction model, and outputting simulated congestion features for representing the simulated congestion condition of the road in the target area.
S105, the congestion prediction model is based on the congestion prediction result, and the updated congestion prediction result is output by combining the simulated congestion characteristics; the updated congestion prediction result is used for improving the accuracy of the congestion prediction result in predicting the road congestion condition of the target area.
In the step, the XGBoost model outputs updated congestion prediction results for improving the accuracy of the congestion prediction results on the road congestion condition prediction of the target area through iterative training based on the congestion prediction results obtained through the output of the steps and combined with the simulated congestion characteristics simulated in the step S104.
Specifically, referring to fig. 2, fig. 2 is a schematic structural diagram of a prediction system for road congestion according to an embodiment of the present application. As shown in fig. 2, the prediction system of the road congestion situation represents a process that the time sequence processing model, the congestion prediction model and the scheduling system process the target characteristics and perform iterative simulation to obtain a road congestion prediction result with higher accuracy: firstly, the time sequence model normalizes the extracted traffic flow characteristics, historical congestion characteristics, weather state characteristics, time sequence characteristics and holiday characteristics and iteratively simulates and outputs the traffic flow characteristics, the historical congestion characteristics, the weather state characteristics, the time sequence characteristics and the holiday characteristics to a congestion prediction model; then, the information features of the vehicle dispatching task and the information features of the vehicle parameters are output to a congestion prediction model through coding; then, the data characteristics are simulated by a dispatching system to obtain simulated congestion characteristics and output the simulated congestion characteristics to a congestion prediction model; and finally, outputting a road congestion prediction result by iterative simulation according to the input target characteristics by the congestion prediction model.
According to the method for predicting the road congestion condition, operation related data of unmanned vehicles arranged in a target area in a transportation process are obtained; determining a target feature for predicting road congestion conditions of a target area based on the operation-related data; and obtaining a congestion prediction result of the target area based on the target characteristics. Therefore, through designing a prediction model of the road congestion in the industrial park trained by the measured data and the simulation data, the external and internal relevant characteristics of the road congestion in the industrial park are extracted, the prediction model is trained by utilizing the characteristic data to obtain a congestion prediction result, and the accuracy of the prediction of the road congestion condition in the industrial park is improved.
Referring to fig. 3 and fig. 4, fig. 3 is a schematic structural diagram of a road congestion prediction apparatus according to an embodiment of the present application, and fig. 4 is a schematic structural diagram of a road congestion prediction apparatus according to an embodiment of the present application. As shown in fig. 3, the prediction apparatus 300 includes:
a data acquisition module 310, configured to acquire operation-related data for an unmanned vehicle disposed in a target area during transportation;
a feature extraction module 320, configured to determine, based on the operation-related data, a target feature for predicting a road congestion condition of a target area;
and the congestion prediction module 330 is configured to obtain a congestion prediction result of the target area based on the target feature.
Further, the feature extraction module 320 is configured to, when determining, based on the operation-related data, a target feature for predicting a road congestion situation of a target area, the feature extraction module 320 is configured to:
preprocessing operation related data to obtain calibration operation related data;
normalizing the actual measurement related data in the calibration operation related data to obtain actual measurement operation characteristic data; performing feature extraction processing on the actually measured operation feature data to obtain actually measured operation related features used for representing external influences on road congestion of a target area;
Carrying out coding processing on preset related data in the calibration operation related data to obtain scheduling task characteristic data; and carrying out feature extraction processing on the dispatching task feature data to obtain dispatching task related features used for representing internal influence of the vehicle transportation task on the road congestion of the target area.
Further, the feature extraction module 320 is configured to perform at least one of the following when the feature extraction module 320 is configured to pre-process the operation-related data: correcting the abnormal value in the operation related data, deleting the repeated value in the operation related data, and filling the missing value in the operation related data.
Further, when the feature extraction module 320 is configured to perform feature extraction processing on the measured operation feature data to obtain a step of characterizing a measured operation related feature that has an external influence on the road congestion of the target area, the feature extraction module 320 is configured to:
extracting traffic flow characteristics and historical congestion state characteristics from the traffic flow data through a time sequence processing model; the time sequence processing model is a cyclic neural network model which is obtained by machine learning training in advance by using a plurality of groups of operation related data of a target area with a sequence relation;
Extracting weather state features from the weather data;
extracting time sequence features and holiday features from the time sequence data.
Further, when the feature extraction module 320 is configured to perform feature extraction processing on the feature data of the scheduled task to obtain a scheduled task related feature for reflecting an internal effect of a vehicle transportation task on road congestion in a target area, the feature extraction module 320 is configured to:
extracting vehicle dispatching task information characteristics from the vehicle dispatching task data through a congestion prediction model; the congestion prediction model is an integrated machine learning framework based on a decision tree, and is a gradient lifting prediction model obtained by training operation related data of a target area in advance;
and extracting vehicle parameter information features from the vehicle parameter data.
Further, in the step of obtaining the congestion prediction result of the target area based on the target feature, the congestion prediction module 330 is configured to:
inputting the target characteristics into a congestion prediction model;
the congestion prediction model is based on an integrated machine learning framework of a decision tree, and the prediction probability of the road congestion of the target area is obtained by carrying out iterative prediction on the target characteristics;
And determining and outputting a congestion prediction result for representing the road congestion condition of the target area in a preset time interval by comparing the prediction probability with a preset congestion threshold value of the target area.
Further, as shown in fig. 4, the prediction apparatus 300 further includes a simulated congestion module 340, where the simulated congestion module 340 is configured to:
a simulated congestion feature in a dispatch system for managing unmanned vehicle transportation in a target area is obtained.
Further, the congestion prediction module 330 is configured to:
the congestion prediction model is based on the congestion prediction result, and the updated congestion prediction result is output by combining the simulated congestion characteristics; the updated congestion prediction result is used for improving the accuracy of the congestion prediction result in predicting the road congestion condition of the target area.
Further, in the step of obtaining the simulated congestion characteristics of the unmanned vehicle for managing the target area in the dispatch system of the transportation process, the simulated congestion module 340 is configured to:
the dispatching system outputs simulated congestion characteristic data which influences road congestion conditions in a target area according to the operation related data;
And performing machine learning simulation on the simulated congestion feature data by a preset simulation prediction model in the scheduling system, and outputting simulated congestion features for representing simulated congestion conditions of the roads in the target area.
The prediction device for road congestion conditions obtains operation related data of unmanned vehicles in a transportation process aiming at a target area; determining a target feature for predicting road congestion conditions of a target area based on the operation-related data; and obtaining a congestion prediction result of the target area based on the target characteristics. Therefore, through designing a prediction model of the road congestion in the industrial park trained by the measured data and the simulation data, the external and internal relevant characteristics of the road congestion in the industrial park are extracted, the prediction model is trained by utilizing the characteristic data to obtain a congestion prediction result, and the accuracy of the prediction of the road congestion condition in the industrial park is improved.
It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, and are not repeated herein.
In the several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods may be implemented in other manners. The above-described apparatus embodiments are merely illustrative, for example, the division of the units is merely a logical function division, and there may be other manners of division in actual implementation, and for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some communication interface, device or unit indirect coupling or communication connection, which may be in electrical, mechanical or other form.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a non-volatile computer readable storage medium executable by a processor. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
Finally, it should be noted that: the foregoing examples are merely specific embodiments of the present application, and are not intended to limit the scope of the present application, but the present application is not limited thereto, and those skilled in the art will appreciate that while the foregoing examples are described in detail, the present application is not limited thereto. Any person skilled in the art may modify or easily conceive of the technical solution described in the foregoing embodiments, or make equivalent substitutions for some of the technical features within the technical scope of the disclosure of the present application; such modifications, changes or substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present application, and are intended to be included in the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A method for predicting road congestion, the method comprising:
acquiring operation related data of an unmanned vehicle arranged in a target area in the transportation process;
determining a target feature for predicting road congestion conditions of a target area based on the operation-related data;
and obtaining a congestion prediction result of the target area based on the target characteristics.
2. The method according to claim 1, wherein the operation-related data includes measured-related data and preset-related data of an unmanned vehicle disposed in a target area during transportation; the step of determining a target feature for predicting a road congestion condition of a target area based on the operation-related data includes:
preprocessing operation related data to obtain calibration operation related data;
normalizing the actual measurement related data in the calibration operation related data to obtain actual measurement operation characteristic data; performing feature extraction processing on the actually measured operation feature data to obtain actually measured operation related features used for representing external influences on road congestion of a target area;
carrying out coding processing on preset related data in the calibration operation related data to obtain scheduling task characteristic data; and carrying out feature extraction processing on the dispatching task feature data to obtain dispatching task related features used for representing internal influence of the vehicle transportation task on the road congestion of the target area.
3. The method of claim 2, wherein the measured operating characteristic data includes traffic flow data, weather data, and time series data within a target area; the scheduled task feature data includes vehicle scheduled task data and vehicle parameter data within a target area.
4. The method of claim 2, wherein the preprocessing comprises at least one of: correcting the abnormal value in the operation related data, deleting the repeated value in the operation related data, and filling the missing value in the operation related data.
5. The method of claim 3, wherein the step of performing feature extraction processing on the measured operation feature data to obtain a measured operation-related feature for characterizing an external impact on road congestion in the target area comprises:
extracting traffic flow characteristics and historical congestion state characteristics from the traffic flow data through a time sequence processing model; the time sequence processing model is a cyclic neural network model which is obtained by machine learning training in advance by using a plurality of groups of operation related data of a target area with a sequence relation;
Extracting weather state features from the weather data;
extracting time sequence features and holiday features from the time sequence data.
6. A method according to claim 3, wherein the step of performing feature extraction processing on the scheduled task feature data to obtain scheduled task related features for reflecting an internal effect of a vehicle transportation task on road congestion in a target area includes:
extracting vehicle dispatching task information characteristics from the vehicle dispatching task data through a congestion prediction model; the congestion prediction model is an integrated machine learning framework based on a decision tree, and is a gradient lifting prediction model obtained by training operation related data of a target area in advance;
and extracting vehicle parameter information features from the vehicle parameter data.
7. The method of claim 1, wherein the step of obtaining a congestion prediction result for a target area based on the target feature comprises:
inputting the target characteristics into a congestion prediction model;
the congestion prediction model is based on an integrated machine learning framework of a decision tree, and the prediction probability of the road congestion of the target area is obtained by carrying out iterative prediction on the target characteristics;
And determining and outputting a congestion prediction result for representing the road congestion condition of the target area in a preset time interval by comparing the prediction probability with a preset congestion threshold value of the target area.
8. The method of claim 1, wherein the predictive method further comprises:
acquiring simulated congestion characteristics in a dispatching system for managing unmanned vehicle transportation processes in a target area;
the congestion prediction model is based on the congestion prediction result, and the updated congestion prediction result is output by combining the simulated congestion characteristics; the updated congestion prediction result is used for improving the accuracy of the congestion prediction result in predicting the road congestion condition of the target area.
9. The method of claim 8, wherein the step of obtaining simulated congestion characteristics of the unmanned vehicle for managing the target area in the dispatch system of the transportation process comprises:
the dispatching system outputs simulated congestion characteristic data which influences road congestion conditions in a target area according to the operation related data;
and performing machine learning simulation on the simulated congestion feature data by a preset simulation prediction model in the scheduling system, and outputting simulated congestion features for representing simulated congestion conditions of the roads in the target area.
10. A prediction apparatus for road congestion conditions, the prediction apparatus comprising:
the data acquisition module is used for acquiring operation related data of the unmanned vehicle arranged in the target area in the transportation process;
the feature extraction module is used for determining target features for predicting road congestion conditions of a target area based on the operation related data;
and the congestion prediction module is used for obtaining a congestion prediction result of the target area based on the target characteristics.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311506953.0A CN117391257A (en) | 2023-11-13 | 2023-11-13 | Road congestion condition prediction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311506953.0A CN117391257A (en) | 2023-11-13 | 2023-11-13 | Road congestion condition prediction method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117391257A true CN117391257A (en) | 2024-01-12 |
Family
ID=89439150
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311506953.0A Pending CN117391257A (en) | 2023-11-13 | 2023-11-13 | Road congestion condition prediction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117391257A (en) |
-
2023
- 2023-11-13 CN CN202311506953.0A patent/CN117391257A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113962364B (en) | Multi-factor power load prediction method based on deep learning | |
| CN115578015B (en) | Sewage treatment whole process supervision method, system and storage medium based on Internet of things | |
| US20150317589A1 (en) | Forecasting system using machine learning and ensemble methods | |
| CN105678398A (en) | Power load forecasting method based on big data technology, and research and application system based on method | |
| CN110874550A (en) | Data processing method, device, equipment and system | |
| CN111127246A (en) | Intelligent prediction method for transmission line engineering cost | |
| CN112285807B (en) | Meteorological information prediction method and device | |
| Song et al. | A match‐then‐predict method for daily traffic flow forecasting based on group method of data handling | |
| CN111445694A (en) | Festival and holiday traffic scheduling method and device based on traffic flow prediction | |
| CN110555551B (en) | Air quality big data management method and system for smart city | |
| CN114664091A (en) | Early warning method and system based on holiday traffic prediction algorithm | |
| CN111008726B (en) | Class picture conversion method in power load prediction | |
| CN111768030A (en) | Bank transportation distribution line planning method and device, equipment and medium | |
| CN111242395B (en) | Method and device for constructing prediction model for OD (origin-destination) data | |
| CN111967675A (en) | Photovoltaic power generation amount prediction method and prediction device | |
| CN114970926A (en) | Model training method, enterprise operation risk prediction method and device | |
| CN116092294A (en) | Expressway traffic flow prediction method based on Transformer and graph meaning network | |
| CN115545331A (en) | Control strategy prediction method and device, equipment and storage medium | |
| CN113706291A (en) | Fraud risk prediction method, device, equipment and storage medium | |
| CN115796434A (en) | Management and control method and device for power distribution network, electronic equipment and storage medium | |
| CN115034430A (en) | Carbon emission prediction method, device, terminal and storage medium | |
| CN110751317A (en) | Power load prediction system and prediction method | |
| CN117113159A (en) | Deep learning-based power consumer side load classification method and system | |
| CN116663746A (en) | Power load prediction method and device, computer equipment and storage medium | |
| CN116663742A (en) | Regional capacity prediction method based on multi-factor and model fusion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |