CN104821082A - Short-time traffic flow prediction method based on integrated evaluation - Google Patents
Short-time traffic flow prediction method based on integrated evaluation Download PDFInfo
- Publication number
- CN104821082A CN104821082A CN201510212889.4A CN201510212889A CN104821082A CN 104821082 A CN104821082 A CN 104821082A CN 201510212889 A CN201510212889 A CN 201510212889A CN 104821082 A CN104821082 A CN 104821082A
- Authority
- CN
- China
- Prior art keywords
- traffic flow
- data
- dttm
- evaluation index
- historical
- 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.)
- Granted
Links
- 238000011156 evaluation Methods 0.000 title claims abstract description 127
- 238000000034 method Methods 0.000 title claims abstract description 48
- 239000011159 matrix material Substances 0.000 claims abstract description 44
- 238000007781 pre-processing Methods 0.000 claims abstract description 25
- 238000004458 analytical method Methods 0.000 claims abstract description 11
- 238000011160 research Methods 0.000 claims abstract description 10
- 238000012216 screening Methods 0.000 claims abstract description 7
- 238000005070 sampling Methods 0.000 claims description 21
- 238000009825 accumulation Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 5
- 238000012163 sequencing technique Methods 0.000 claims description 4
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000000691 measurement method Methods 0.000 abstract description 6
- 230000008859 change Effects 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 8
- 230000008569 process Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000005291 chaos (dynamical) Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011985 exploratory data analysis Methods 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 238000007667 floating Methods 0.000 description 1
- 238000013277 forecasting method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000029305 taxis Effects 0.000 description 1
- 238000012731 temporal analysis Methods 0.000 description 1
- 238000000700 time series analysis Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- 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
Landscapes
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a short-time traffic flow prediction method based on integrated evaluation. Undetermined evaluation indexes are acquired via research and analysis and combination of actual traffic flow prediction scenes; historical data of the undetermined evaluation indexes are acquired so that a historical database is obtained; a historical data matrix based on the historical database is constructed via data preprocessing; main evaluation indexes are obtained via screening by calculating correlation coefficient of all the undetermined evaluation indexes and traffic flow, and respective contribution weight values of the main evaluation indexes are calculated; traffic flow prediction values under all the main evaluation indexes are obtained via combination of real-time traffic flow data and analysis on the traffic flow historical data in the historical database via an improved time sequence similarity measurement method; and traffic flow is predicted by adopting a linear weighed integrated evaluation method according to the obtained traffic flow prediction values under all the main evaluation indexes. Multidimensional time sequences are integrated by the method so that more accurate numerical value prediction results are obtained.
Description
Technical Field
The invention relates to a traffic flow prediction method, in particular to a short-time traffic passenger flow prediction method based on comprehensive evaluation.
Background
In recent years, with the continuous development of economy and the explosive increase of the quantity of various motor vehicles, the urban traffic problem becomes more serious. Under increasing traffic pressure, the concept of intelligent traffic arises. The intelligent transportation is the development direction of future transportation systems, and is a comprehensive transportation management system established by effectively integrating and applying advanced information technology, data communication transmission technology, electronic sensing technology, control technology, computer technology and the like to the whole ground transportation management system. And researching and analyzing historical traffic flow data by a data mining technology to obtain a related rule. The application of the obtained rule to the current scene and the prediction of the short-term traffic flow in the future are the premise and the basis of dynamic traffic guidance, and the optimal driving route can be provided for travelers by further applying the modern communication technology, the computer technology and the like due to the fact that the real-time traffic flow information with high precision exists, so that the purposes of smooth network and high-efficiency operation are achieved.
Traffic flow may be considered as a set of time series of values that change over time. Most of the current research on time series is limited to one-dimensional time series. In a complicated practical application scenario, the research target is often influenced by multiple factors, so that a large deviation is likely to occur if only one dimension is considered. Therefore a multi-dimensional time series must be introduced. The multi-dimensional time-series prediction is a prediction result of an analysis target obtained by comprehensively analyzing a series of observed values of a plurality of attributes acquired in time series. At present, research directions on multidimensional time sequences mainly comprise similarity-based mode mining research such as classification, clustering, correlation regularity and exploratory data analysis. The time series analysis model usually adopted comprises wavelet analysis, neural network, chaos theory, nonlinear prediction model supporting vector machine method and the like. Moreover, multi-dimensional time series prediction also focuses on the fact that a relatively macroscopic prediction analysis result and rule summarization are obtained on the basis of pattern mining of existing large amount of data, and an accurate numerical prediction result does not exist. The method cannot be applied to scenes with higher requirements on numerical prediction precision. The combined prediction model has strict requirements on data and is complex to implement, and is difficult to implement in an actual application scene.
Disclosure of Invention
The invention provides a short-term traffic flow prediction method based on comprehensive evaluation to solve the technical problems.
The technical scheme adopted by the invention is as follows: a short-term traffic flow prediction method based on comprehensive evaluation comprises the following steps:
s1, obtaining an evaluation index to be determined by combining research and analysis with an actual traffic flow prediction scene;
s2, acquiring historical data of the to-be-determined evaluation index obtained in the step S1 to obtain a historical database;
s3, constructing a historical data matrix based on a historical database through data preprocessing;
s4, calculating correlation coefficients of each undetermined evaluation index and traffic flow, converting the correlation coefficients into percentages, sequencing the correlation coefficients in descending order, accumulating the correlation coefficients in descending order until the accumulation result is larger than or equal to a first threshold value, stopping accumulation operation, eliminating the undetermined evaluation indexes with smaller correlation coefficients which are not accumulated, obtaining main evaluation indexes, and calculating the contribution weight of each main evaluation index;
s5, analyzing the traffic flow historical data in the historical database through the improved Euclidean distance by combining with the real-time traffic flow data to obtain a traffic flow predicted value under each main evaluation index; and predicting the traffic flow by adopting a linear weighted comprehensive evaluation method according to the obtained traffic flow predicted value under each main evaluation index.
Further, the history database created in step S2 is: DB _ TABLE [ DTTM, F, r ]1,r2,…,rp]；
Wherein, DTTM represents the recording sampling time point, and F represents the traffic flow historical data corresponding to the sampling time point.
Further, the step S3 specifically includes: s31: preprocessing data; s32: constructing a historical data matrix;
the step S31 of preprocessing data specifically includes the following sub-steps:
s311: digitizing the symbol information, and converting the symbol information into digitized information;
s312: numerical deficiency and numerical valueError(s) inPreprocessing, namely, for a discontinuous traffic flow numerical sequence, filling data by adopting an average interpolation method; the data of each time interval is supplemented by adopting a method of equally dividing the data of the statistical sums of a plurality of time intervals according to the time intervals; for the significance ofError(s) inDeleting the data point and simultaneously adopting an average interpolation method for completion;
s313: data preprocessing, namely performing data processing on the value set of each to-be-evaluated index by adopting data segmentation processing;
the step S32: the specific steps for constructing the historical data matrix are as follows: a history data matrix is constructed based on the history database of step S2.
Further, the step S4 of selecting the index and calculating the weight specifically includes the following sub-steps:
s41: carrying out data standardization operation, and obtaining a standardized historical data matrix according to the standardized data;
s42: calculating correlation coefficients of evaluation indexes and traffic flow in a standardized historical data matrix, sorting the correlation coefficients, converting the correlation coefficients into percentages, sorting the correlation coefficients in a descending order, sequentially accumulating the correlation coefficients in a descending order until an accumulation result is greater than or equal to a first threshold value, stopping accumulation operation, and eliminating undetermined evaluation indexes with smaller correlation coefficients which are not accumulated, so as to obtain K main evaluation indexes;
s43: and calculating the contribution of each main evaluation index to the traffic flow.
Further, the step S5 specifically includes the following sub-steps:
obtaining the record number of the historical data in the time window according to the set length of the time window as T and the sampling frequency T of the historical data;
s52: constructing a traffic flow matrix, calculating a correlation coefficient matrix corresponding to the traffic flow matrix, and calculating the ratio of historical data in different time periods in a time window according to the obtained correlation coefficient matrix:
s53: calculating a traffic flow predicted value under each main evaluation index according to the improved Euclidean distance;
s54: and predicting the traffic flow by adopting a linear weighted comprehensive evaluation prediction method.
Further, the step S52 specifically includes the following sub-steps:
s521: constructing a traffic flow matrix, and assuming that the time sequence with the length of len of the traffic flow F is as follows: f. of1,f2,…fi,…,flenIf the condition len > n is satisfied, the following traffic flow matrix TF is constructed:
wherein, the 1 st column of the matrix represents the data of the predicted time, and the 2 nd to n +1 th columns represent n traffic flow historical data within the length of the time window;
s522: calculating a correlation coefficient matrix of the matrix TF:
s523: calculating the ratio alpha of each historical data in the time windowi′：
Further, the step S53 specifically includes the following sub-steps:
s531: obtaining the current main evaluation index rkCurrent main evaluation index rkWhen the value of (a) is Val, r in the history database is setkThe historical data with the value Val is extracted to form a new data set DTS:<DTTM,F>the number of records in the data set DTS is DTS _ count;
s532: if dts _ count is 0, this indicates the main evaluation index rkWhen the value of (A) is Val, there is no corresponding data in the history database, so let Fpredict_kGo to step S535 if it is 0, otherwise go to step S533;
wherein, Fpredict_kRepresenting a traffic flow predicted value corresponding to the kth main evaluation index;
s533: time of Day (DTTM)now-T) to DTTMnowThe time series of length m within a time period is denoted as F _ NOW for the data setEach DTTM in DTSaTime of Day (DTTM)a-T) to DTTMaThe time series with length m in the time period is recorded as F _ HISTORYaWhere a is 1,2, …, dts _ count;
wherein, DTTMnowRepresenting a predicted time point;
s534: in the main evaluation index rkNext, specific values for traffic flow are predicted using the following formula:
Fpredict_k＝Fkey_dttm
wherein, Fkey_dttmThe DTTM value in the historical database is a traffic flow value in the historical record corresponding to the key _ DTTM;
s535: and judging whether each main evaluation index completes the prediction of the traffic flow, if so, ending the judgment, otherwise, turning to the step S531 to obtain the next main evaluation index.
Further, the value of key _ dttm in step S534 should satisfy the following condition:
dist(F_NOW,F_HISTORYkey_dttm)
；
＝min{dist(F_NOW,F_HISTORYa)}，a＝1,2,…,dts_count
wherein dist (i) represents similarity calculation based on the improved Euclidean distance, and min {. cndot.) represents the minimum value in the set.
Further, the step S54 specifically includes:
s541: the accumulated contribution weight of each evaluation index is recorded as wtotalWhen the evaluation index r is determinedkWhen an effective prediction result is generated, accumulating the contribution weight of the effective prediction result to the traffic flow;
s542: w obtained from step S541totalPredicting the traffic flow F by adopting a prediction method of linear weighted comprehensive evaluationpredict。
The invention has the beneficial effects that: the short-term traffic flow prediction method based on comprehensive evaluation obtains an evaluation index to be determined by combining research and analysis with an actual traffic flow prediction scene; acquiring historical data of an index to be evaluated to obtain a historical database; through data preprocessing, a historical data matrix based on a historical database is constructed; screening out main evaluation indexes by calculating the correlation coefficient between each index to be evaluated and the traffic flow, and calculating the respective contribution weight of the main evaluation indexes; analyzing the traffic flow historical data in the historical database by an improved time series similarity measurement method in combination with real-time traffic flow data to obtain a traffic flow predicted value under each main evaluation index; according to the traffic flow predicted value under each main evaluation index, a linear weighting comprehensive evaluation method is adopted to predict the traffic flow.
Drawings
Drawing (A) 1Is a process flow of the present inventionDrawing (A)。
Drawing (A) 2The data preprocessing flow of the inventionDrawing (A)。
Drawing (A) 3The screening of the evaluation indexes and the weight calculation process of the inventionDrawing (A)。
Drawing (A) 4Schematic of the time window for the present inventionDrawing (A)。
Drawing (A) 5The invention discloses a time series similarity measurement method based on improved Euclidean distance.
Drawing (A) 6Provide embodiments of the inventionOfIndication of number of records of historical data within time windowDrawing (A)。
Drawing (A) 7Is broken down for the detailed step of step S52Drawing (A)。
Drawing (A) 8Is broken down for the detailed step of step S53Drawing (A)。
Detailed Description
To facilitate understanding of the technical content of the present invention by those skilled in the art, the following is incorporatedDrawingsThe present disclosure is further explained.
As shown in the figure 1The method flow of the invention is shownDrawing (A)The short-term traffic flow prediction method based on comprehensive evaluation provided by the patentMethod ofThe method comprises the following five steps:
s1, selecting factor r influencing traffic flow by combining research analysis with actual traffic flow prediction scene1,r2,…rpAnd is used as an undetermined evaluation index;
s2, acquiring historical data of the to-be-determined evaluation index obtained in the step S1 to obtain a historical database;
s3, constructing a historical data matrix based on a historical database through data preprocessing;
s4, screening indexes and calculating weights to obtain the index r with great influence on traffic flow1,r2,…rKThe corresponding contribution weight is w1,w2,…,wK；
S5, considering each evaluation index rkAnalyzing the historical traffic flow data in the historical database by an improved time series similarity measurement method in combination with real-time traffic flow data to obtain the traffic flow prediction under the condition of a single evaluation indexMeasured value Fpredict_k(ii) a The method for predicting the future short-term traffic flow F by linear weighted comprehensive evaluationpredict。
The embodiment of the invention takes an airport as an example.
In the step S1, evaluation index selection may be performed by combining investigation and analysis with actual traffic flow prediction scenes, and p total influence factors affecting the traffic flow F are selected as undetermined evaluation indexes for traffic flow prediction, and are denoted as r1,r2,…,rp. The transport capacity of taxis in an airport is influenced by various factors, including flight information r1Weather information r2Different time periods r3The number of people riding a taxi r4Current traffic situation r of taxi5And the like.
Step S2 is to collect the historical data of the to-be-evaluated indicator obtained in step S1, and obtain a historical database as follows: DB _ TABLE [ DTTM, F, r ]1,r2,…,rp]；
Wherein, DTTM represents the recording sampling time point, and F represents the traffic flow historical data corresponding to the sampling time point.
The step S3 includes: s31: preprocessing data; s32: constructing a historical data matrix;
as shown in the figure 2As shown, the step S31 is specifically executed by the data preprocessing: the data preprocessing part takes the index r into account1,r2,…rpThe data form of (A) is various, including various data types such as symbols, numerical values and the like, which can not be directly calculated and are irregular or notError(s) inThe data of (2) may cause a large deviation of the calculation result. Therefore, corresponding data needs to be preprocessed, so that a better prediction result can be obtained, and meanwhile, the operation speed can be effectively improved.As shown in fig. 2The data preprocessing flow of the invention is shownDrawing (A)The specific main steps of data preprocessing are as follows:
s311: and digitizing the symbol information. Due to the evaluation index r1,r2,…rpMay contain symbolic information such as words, weather information having a large influence on the traffic flow, and the like. It is therefore desirable to digitize text-type data for ease of computation.
With weather information r2For example, the degree of influence it has on flight arrival will be the weather information r2The classification into 4 classes is carried out,as shown in Table 1As shown.
Watch (A) 1Weather information classification
| Weather type | Description of the invention |
| Weather type I (without influence) | Sunny and cloudy days |
| Class II weather (Weak influence) | Rain fall, small to medium rain, small to medium snow and small snow |
| Weather type III (moderate effect) | Medium snow, rain with snow, medium rain, medium to heavy rain, light fog and haze |
| Weather type IV (severe impact) | Heavy snow, floating dust, heavy to heavy rain, thunderstorm rain, rainstorm and fog |
Four types of weather are taken as basic types, and the weather can be obtainedAs shown in Table 2The shown weather change trend and the symbolic representation thereof, the time window for observing the weather change trend is set to be 2 hours in length.
Watch (A) 2Weather transformation trends and types thereofWatch (A)
| Trend of weather change | Type (B) |
| Ⅰ→Ⅰ | 1 |
| Ⅰ→Ⅱ | 2 |
| Ⅰ→Ⅲ | 3 |
| Ⅰ→Ⅳ | 4 |
| Ⅱ→Ⅰ | 5 |
| Ⅱ→Ⅱ | 6 |
| Ⅱ→Ⅲ | 7 |
| Ⅱ→Ⅳ | 8 |
| Ⅲ→Ⅰ | 9 |
| Ⅲ→Ⅱ | 10 |
| Ⅲ→Ⅲ | 11 |
| Ⅲ→Ⅳ | 12 |
| Ⅳ→Ⅰ | 13 |
| Ⅳ→Ⅱ | 14 |
| Ⅳ→Ⅲ | 15 |
| Ⅳ→Ⅳ | 16 |
S312: numerical deficiency and numerical valueError(s) inAnd (4) preprocessing. And for discontinuous traffic flow numerical value sequences, filling the data by adopting an average interpolation method. And for the data of the statistical sum of a plurality of time intervals, the data of each time interval is supplemented by adopting a method of equally dividing by time intervals. For the significance ofError(s) inThe data point is deleted and simultaneously filled up by using an average interpolation method.
S313: and (4) preprocessing data. Due to arbitrary evaluation index rjJ is 1,2, …, and p has a large range of values, and if each value is considered as a class, the class is too many to affect the prediction efficiency. For the evaluation index rjIn other words, values of the same order of magnitude are assigned to the traffic flowF often has the same influence, so that the evaluation index r can be subjected to data segmentation processingjValue set of { v }je1,2, …, p; e 1,2, … s, where s represents the evaluation index rjThe number of values of (a). Evaluation index r after data preprocessingiThe value set of (a) is expressed as:
rj→{v* je|j＝1,2,…,p；e＝1,2,…s*}
wherein v is* jeIndicates the evaluation index rjValue after data preprocessing, s*Indicating the evaluation index r after data preprocessingjThe number of values taken.
Step S32: the specific steps for constructing the historical data matrix are as follows: from the data table of the history database obtained in step S2, the following is shown:
DB_TABLE＝[DTTM,F,r1,r2,…,rp]；
wherein, DTTM represents the recording sampling time point, and F represents the traffic flow historical data corresponding to the sampling time point. The historical data matrix constructed based on the historical data table is as follows:
MTX＝[F,r1,r2,…,rp]
the evaluation index screening and weight calculation in step S4 mainly determines and selects the evaluation index set forth in step S1.As shown in fig. 3As shown, by calculating the degree of correlation between each index and the traffic flow, an evaluation index having a large correlation is selected, and an index having a small correlation or no correlation is removed. The method comprises the following specific steps:
s41: a data normalization operation, an operation to normalize data according to the following equation:
wherein,andthe mean and variance of the g-th term in the historical data matrix are respectively expressed, and q is p +1, so that a normalized matrix MTX is obtained (x is x)ig *)n×q；
S42: calculating correlation coefficient u between each evaluation index of matrix MTX' and traffic flowFgAs follows:
sorting the correlation coefficients, converting the correlation numbers into percentages, sorting the correlation coefficients in a descending order, sequentially accumulating the correlation numbers in a descending order until the accumulation result is greater than or equal to a first threshold value, wherein the first threshold value is 80%, stopping accumulation operation, and eliminating undetermined evaluation indexes with smaller correlation numbers which are not accumulated, thereby obtaining K main evaluation indexes r1,r2,…rk,…,rK；
S43: calculating the contribution weight w of each of the main evaluation indexes to the traffic flowkThe formula is as follows:
wherein u isFkRepresenting the kth primary traffic flow correlation coefficient.
In the process of traffic flow prediction under the condition of only considering a single evaluation index in step S5, the most similar traffic flow data is searched from the historical database as a predicted value by adopting a time series similarity measurement method. When the change rule of the traffic flow along with the time is examined, the traffic flow can be considered as a time sequence with the change of the value along with the time. Due to the continuity of the traffic flow, the traffic flow data of the past time period can affect the traffic flow of the current time period,as shown in fig. 4Shown is a time window schematicDrawing (A). Considering that the influence degrees of different historical periods on the current traffic flow are different, the invention adopts the improved Euclidean distance as the method design flow of the time series similarity measurementAs shown in fig. 5As shown. The detailed design metric method comprises the following steps:
s51: setting the length of the time window, and assuming that the length of the time window is T (min), when the sampling frequency of the historical data is t (min), the number of records of the historical data in the time window is as follows:
wherein, the current prediction time point is recorded as tpredictLet the current nearest sampling time point be tsample，Δt＝tpredict-tsample， Indicating rounding.
S52: constructing a traffic flow matrix, calculating a correlation coefficient matrix corresponding to the traffic flow matrix, and calculating the ratio alpha of historical data of different time periods in a time window according to the obtained correlation coefficient matrixi′：
S53: calculating the traffic flow F under each main evaluation indexpredict_k；
S54: forecasting method for forecasting traffic flow F by adopting linear weighted comprehensive evaluationpredict。
The step S51 specifically includes: setting the length of the time window as T (min), assuming that the sampling frequency of the historical data is t (min), and the number of records of the historical data in the available time window is as follows:
recording the current prediction time point as tpredictLet the current nearest sampling time point be tsampleThen, the number of records of the history data in the time window can be represented as:
where, t ispredict-tsample，Indicating rounding.
As shown in the figure 6Shown, 1) the predicted point is a half-hour point, i.e. 8: 00. 8: 30. 9:00, 9:30, 10:00, the sampling time point is a half point, namely 8: 00. 8: 30. 9:00, 9:30, 10:00, then t is the predicted time point according to the presentpredict9:30, the current nearest sampling time point is tsample9:00, get Δ t ═ tpredict-tsample＝30min<T＝1For 30min, then
2) The prediction point is a half-hour point, namely 8: 00. 8: 30. 9:00, 9:30 and 10:00, and the sampling time points are 5 minutes apart from each other at half the point, namely 7: 55. 8: 25. 8:55, 9:25, 9:55, then t is the current predicted time pointpredict9:30, the current nearest sampling time point is tsample9:25, yielding Δ t ═ tpredict-tsample＝5min<T is 130min, then
3) The prediction point is a half-hour point, namely 8: 00. 8: 30. 9:00, 9:30 and 10:00, and the sampling time points are half-point differences of 15 minutes, namely 7: 45. 8: 15. 8:45, 9:15, 9:45, then t is the current predicted time pointpredict9:30, the current nearest sampling time point is tsample9:15, get Δ t ═ tpredict-tsample＝15min<T is 130min, then
As shown in the figure 7As shown, the step S52 specifically includes: when the change rule of the traffic flow along with the time is examined, the traffic flow can be considered as a time sequence with the change of the value along with the time. The method for measuring similarity of time series by using improved Euclidean distance is adopted, and records most similar to the current scene are searched from a historical database and used as predicted values. Wherein the modified euclidean distance is expressed as follows:
wherein P and Q each represent a length ofmTime series of (1), pi′，qi′Representing the sequence value, alpha, of the corresponding time pointi′ Watch (A)Shows the proportion occupied in the time sequence and satisfies As shown in fig. 7The method specifically comprises the following steps:
s521: suppose that the time series of the length len of the traffic flow F is: f. of1,f2,…fi,…,flenAnd satisfies the condition len > n. The following traffic flow matrix TF is constructed:
where the first column of the matrix is the data representing the predicted time, then n，The column then represents n within the time window·And (4) traffic flow historical data.
S522: and calculating a correlation coefficient matrix.
Calculating the correlation coefficient matrix of the matrix TF:
s523: calculating the ratio alpha of each historical data in the time windowi′：
Calculating similarity according to improved Euclidean distance formula
Wherein P represents a time series of length m, Q represents a time series of length m, Pi′，qi′Representing the order of the corresponding time pointsColumn value, αi′Represents the proportion of each historical data in the time series and satisfies
The step S53 specifically includes: suppose the predicted time point is DTTMnowCorresponding to the evaluation index rkThe value of (1) is Val.As shown in fig. 8The method comprises the following specific steps:
s531: obtaining the current main evaluation index rkCurrent main evaluation index rkWhen the current value of (a) is Val, r in the history database is setkThe historical data with the value Val is extracted to form a new data set DTS:<DTTM,F>the number of records in the data set DTS is DTS _ count.
S532: if dts _ count indicates the main evaluation index rkWhen the value of (A) is Val, there is no corresponding data in the history database, so let Fpredict_kGo to step S535 if it is 0, otherwise go to step S533.
S533: time of Day (DTTM)now-T) to DTTMnowThe time series of length m within the time period is denoted as F _ NOW, DTTM for each of the data sets DTSaTime of Day (DTTM)a-T) to DTTMaThe time series with length m in the time period is recorded as F _ HISTORYaWhere a is 1,2, …, dts _ count;
s534: in the main evaluation index rkNext, specific values for traffic flow are predicted using the following formula:
Fpredict_k＝Fkey_dttm
wherein, Fkey_dttmAnd the value of the traffic flow in the history record corresponding to the DTTM value key _ DTTM in the history database is shown. The value of key _ dttm should satisfy the following condition:
the DTTM value in the data set DTS is key _ DTTM value which satisfies the following conditions:
dist(F_NOW,F_HISTORYkey_dttm)
。
＝min{dist(F_NOW,F_HISTORYa)}，a＝1,2,…,dts_count
that is, only the evaluation index r is consideredkIn all the history records, the history data corresponding to the sampling time key _ dttm is most similar to the current scene. Thus with Fkey_dttmAs a value ofExpedition principalIndex to be evaluated rkThe traffic flow of time is Fpredict_k。
S535: judging whether each main evaluation index completes the prediction of the traffic flow, if so, ending the judgment, otherwise, turning to the step S531, and executing k to be k + 1;
where K denotes the kth main evaluation index, and K is 1,2, …, K.
The step S54 specifically includes the following sub-steps:
s541: the accumulated contribution weight of each evaluation index is recorded as wtotalWhen the evaluation index r is determinedkWhen an effective prediction result is generated, the contribution weight value of the effective prediction result to the traffic flow is accumulated to wtotalAs shown in the following formula:
wtotal＝wtotal+wk；
s542: w obtained from step S541totalPredicting the traffic flow F by adopting a prediction method of linear weighted comprehensive evaluationpredictThe calculation formula is as follows:
wherein, wkIs the main evaluation index rkFor traffic flow FpredictDegree of contribution of (1), wtotalIs the cumulative contribution of each of the main evaluation indexes, Fpredict_kIs only atExpedition principalIndex to be evaluated rkThe traffic flow prediction result obtained on the premise of (1).
The short-term traffic flow prediction method based on comprehensive evaluation obtains an evaluation index to be determined by combining research and analysis with an actual traffic flow prediction scene; acquiring historical data of an index to be evaluated to obtain a historical database; through data preprocessing, a historical data matrix based on a historical database is constructed; screening out main evaluation indexes by calculating correlation coefficients of each undetermined evaluation index and a traffic flow, and calculating respective contribution weights of the main evaluation indexes; analyzing the traffic flow historical data in the historical database by an improved time series similarity measurement method in combination with real-time traffic flow data to obtain a traffic flow predicted value under each main evaluation index; according to the traffic flow predicted value under each main evaluation index, a linear weighting comprehensive evaluation method is adopted to predict the traffic flow.
It will be appreciated by those of ordinary skill in the art that the embodiments described herein are intended to assist the reader in understanding the principles of the invention and are to be construed as being without limitation to such specifically recited embodiments and examples. Various modifications and alterations to this invention will become apparent to those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the scope of the claims of the present invention.
Claims (8)
1. A short-term traffic flow prediction method based on comprehensive evaluation is characterized by comprising the following steps:
s1, obtaining an evaluation index to be determined by combining research and analysis with an actual traffic flow prediction scene;
s2, acquiring historical data of the to-be-determined evaluation index obtained in the step S1 to obtain a historical database;
s3, constructing a historical data matrix based on a historical database through data preprocessing;
s4, calculating correlation coefficients of each undetermined evaluation index and traffic flow, converting the correlation coefficients into percentages, sequencing the correlation coefficients in descending order, accumulating the correlation coefficients in descending order until the accumulation result is larger than or equal to a first threshold value, stopping accumulation operation, eliminating the undetermined evaluation indexes with smaller correlation coefficients which are not accumulated, obtaining main evaluation indexes, and calculating the contribution weight of each main evaluation index;
s5, analyzing the traffic flow historical data in the historical database through the improved Euclidean distance by combining with the real-time traffic flow data to obtain a traffic flow predicted value under each main evaluation index; and predicting the traffic flow by adopting a linear weighted comprehensive evaluation method according to the obtained traffic flow predicted value under each main evaluation index.
2. The short-term traffic flow prediction method based on comprehensive evaluation according to claim 1, wherein the historical database created in step S2 is: DB _ TABLE [ DTTM, F, r ]1,r2,…,rp]；
Wherein, DTTM represents the recording sampling time point, and F represents the traffic flow historical data corresponding to the sampling time point.
3. The short-term traffic flow prediction method based on comprehensive evaluation according to claim 1, wherein the step S3 specifically includes: s31: preprocessing data; s32: constructing a historical data matrix;
the step S31 of preprocessing data specifically includes the following sub-steps:
s311: digitizing the symbol information, and converting the symbol information into digitized information;
s312: numerical value missing and numerical value error preprocessing, namely, for a discontinuous traffic flow numerical value sequence, filling data by adopting an average interpolation method; the data of each time interval is supplemented by adopting a method of equally dividing the data of the statistical sums of a plurality of time intervals according to the time intervals; deleting the data point for the obviously wrong data, and simultaneously adopting an average interpolation method for completing;
s313: data preprocessing, namely performing data processing on the value set of each to-be-evaluated index by adopting data segmentation processing;
the step S32: the specific steps for constructing the historical data matrix are as follows: a history data matrix is constructed based on the history database of step S2.
4. The short-term traffic flow prediction method based on comprehensive evaluation according to claim 1, wherein the step S4 of index screening and weight calculation specifically comprises the following substeps:
s41: carrying out data standardization operation, and obtaining a standardized historical data matrix according to the standardized data;
s42: calculating correlation coefficients of all evaluation indexes and traffic flow in a standardized historical data matrix, sequencing the correlation coefficients, converting the correlation coefficients into percentages, sequencing the correlation coefficients from large to small, sequentially accumulating the correlation coefficients from large to small until an accumulation result is larger than or equal to a first threshold value, stopping accumulation operation, and eliminating undetermined evaluation indexes with smaller correlation coefficients which are not accumulated, thereby obtaining K main evaluation indexes;
s43: and calculating the contribution of each main evaluation index to the traffic flow.
5. The short-term traffic flow prediction method based on comprehensive evaluation according to claim 1, wherein the step S5 specifically includes the following substeps:
s51: obtaining the record number of the historical data in the time window according to the set length of the time window as T and the sampling frequency T of the historical data;
s52: constructing a traffic flow matrix, calculating a correlation coefficient matrix corresponding to the traffic flow matrix, and calculating the ratio of historical data in different time periods in a time window according to the obtained correlation coefficient matrix;
s53: calculating a traffic flow predicted value under each main evaluation index according to the improved Euclidean distance;
s54: and predicting the traffic flow by adopting a linear weighted comprehensive evaluation prediction method.
6. The short-term traffic flow prediction method based on comprehensive evaluation according to claim 5, wherein the step S53 specifically comprises the following substeps:
s531: obtaining the current main evaluation index rkCurrent main evaluation index rkWhen the value of (a) is Val, r in the history database is setkThe historical data with the value Val is extracted to form a new data set DTS:<DTTM,F>the number of records in the data set DTS is DTS _ count;
s532: if dts _ count is 0, this indicates the main evaluation index rkWhen the value of (A) is Val, there is no corresponding data in the history database, so let Fpredict_kGo to step S535 if it is 0, otherwise go to step S533;
wherein, Fpredict_kRepresenting a traffic flow predicted value corresponding to the kth main evaluation index;
s533: time of Day (DTTM)now-T) to DTTMnowThe time series of length m within the time period is denoted as F _ NOW, DTTM for each of the data sets DTSaTime of Day (DTTM)a-T) to DTTMaThe time series with length m in the time period is recorded as F _ HISTORYaWhere a is 1,2, …, dts _ count;
wherein, DTTMnowRepresenting a predicted time point;
s534: in the main evaluation index rkNext, specific values for traffic flow are predicted using the following formula:
Fpredict_k＝Fkey_dttm
wherein, Fkey_dttmThe DTTM value in the historical database is a traffic flow value in the historical record corresponding to the key _ DTTM;
s535: and judging whether each main evaluation index completes the prediction of the traffic flow, if so, ending the judgment, otherwise, turning to the step S531 to obtain the next main evaluation index.
7. The short-term traffic flow prediction method based on comprehensive evaluation according to claim 6, wherein the value of key _ dttm in the step S534 should satisfy the following condition:
dist(F_NOW,F_HISTORYkey_dttm)；
＝min{dist(F_NOW,F_HISTORYa)}，a＝1,2,…,dts_count
where dist (i) indicates similarity calculation based on the improved euclidean distance, and min { i } indicates the minimum value in the set.
8. The short-term traffic flow prediction method based on comprehensive evaluation according to claim 5, wherein the step S54 is specifically:
s541: the accumulated contribution weight of each evaluation index is recorded as wtotalWhen the evaluation index r is determinedkWhen an effective prediction result is generated, accumulating the contribution weight of the effective prediction result to the traffic flow;
s542: w obtained from step S541totalPredicting the traffic flow F by adopting a prediction method of linear weighted comprehensive evaluationpredict。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510212889.4A CN104821082B (en) | 2015-04-29 | 2015-04-29 | Short-time traffic flow prediction method based on integrated evaluation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510212889.4A CN104821082B (en) | 2015-04-29 | 2015-04-29 | Short-time traffic flow prediction method based on integrated evaluation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104821082A true CN104821082A (en) | 2015-08-05 |
| CN104821082B CN104821082B (en) | 2017-01-18 |
Family
ID=53731362
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510212889.4A Expired - Fee Related CN104821082B (en) | 2015-04-29 | 2015-04-29 | Short-time traffic flow prediction method based on integrated evaluation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104821082B (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105118294A (en) * | 2015-09-25 | 2015-12-02 | 山东易构软件技术股份有限公司 | Short-time traffic flow prediction method based on state pattern |
| CN105957329A (en) * | 2016-06-29 | 2016-09-21 | 肖锐 | Intelligentized information system for highway |
| CN106951976A (en) * | 2016-10-12 | 2017-07-14 | 华南理工大学 | A kind of bus passenger flow Forecasting Methodology based on pattern classification |
| CN107066831A (en) * | 2017-05-19 | 2017-08-18 | 君晟合众（北京）科技有限公司 | A kind of regional complex environmental assessment techniques, apparatus and system |
| CN107085942A (en) * | 2017-06-26 | 2017-08-22 | 广东工业大学 | A kind of traffic flow forecasting method based on wolf pack algorithm, apparatus and system |
| CN107610470A (en) * | 2017-10-31 | 2018-01-19 | 迈锐数据（北京）有限公司 | A kind of traffic congestion evaluation method and device |
| CN107844848A (en) * | 2016-09-20 | 2018-03-27 | 中国移动通信集团湖北有限公司 | A kind of region flow of the people Forecasting Methodology and system |
| CN108053646A (en) * | 2017-11-22 | 2018-05-18 | 华中科技大学 | Traffic characteristic acquisition methods, Forecasting Methodology and system based on time-sensitive feature |
| CN108090241A (en) * | 2016-11-23 | 2018-05-29 | 财团法人工业技术研究院 | trend variable identification method and system of continuous process |
| CN108922183A (en) * | 2018-07-17 | 2018-11-30 | 重庆大学 | A kind of highway short term traffic forecasting method of space time correlation degree in conjunction with SVR |
| CN109241077A (en) * | 2018-08-30 | 2019-01-18 | 东北大学 | Production target variation tendency visual query system and method based on similitude |
| CN109559512A (en) * | 2018-12-05 | 2019-04-02 | 北京掌行通信息技术有限公司 | A kind of regional traffic flow prediction technique and device |
| CN111862592A (en) * | 2020-05-27 | 2020-10-30 | 浙江工业大学 | Traffic flow prediction method based on RGCN |
| CN111862583A (en) * | 2019-04-24 | 2020-10-30 | 北京嘀嘀无限科技发展有限公司 | Traffic flow prediction method and device |
| CN117437786A (en) * | 2023-12-21 | 2024-01-23 | 速度科技股份有限公司 | Real-time traffic flow prediction method based on artificial intelligence for traffic network |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110517479B (en) * | 2018-05-22 | 2020-11-03 | 杭州海康威视系统技术有限公司 | Urban road traffic prediction method and device and electronic equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101694744A (en) * | 2009-10-28 | 2010-04-14 | 北京交通大学 | Method and system for evaluating road emergency evacuation capacity and method and system for grading road emergency evacuation capacity |
| JP2010272142A (en) * | 2010-08-27 | 2010-12-02 | Nippon Telegr & Teleph Corp <Ntt> | Method and device for estimating traveling time, traveling time estimation program mounted with the method, medium storing the program, method and device for predicting traveling time, traveling time prediction program mounted with the method, and medium storing the program |
| CN102693633A (en) * | 2012-06-07 | 2012-09-26 | 浙江大学 | Short-term traffic flow weighted combination prediction method |
| WO2014003321A1 (en) * | 2012-06-27 | 2014-01-03 | 명지대학교 산학협력단 | Start point-based traffic allocation method using shortest path |
-
2015
- 2015-04-29 CN CN201510212889.4A patent/CN104821082B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101694744A (en) * | 2009-10-28 | 2010-04-14 | 北京交通大学 | Method and system for evaluating road emergency evacuation capacity and method and system for grading road emergency evacuation capacity |
| JP2010272142A (en) * | 2010-08-27 | 2010-12-02 | Nippon Telegr & Teleph Corp <Ntt> | Method and device for estimating traveling time, traveling time estimation program mounted with the method, medium storing the program, method and device for predicting traveling time, traveling time prediction program mounted with the method, and medium storing the program |
| CN102693633A (en) * | 2012-06-07 | 2012-09-26 | 浙江大学 | Short-term traffic flow weighted combination prediction method |
| WO2014003321A1 (en) * | 2012-06-27 | 2014-01-03 | 명지대학교 산학협력단 | Start point-based traffic allocation method using shortest path |
Non-Patent Citations (2)
| Title |
|---|
| 刘敬贤等: "船舶交通流组合预测方法研究", 《中国航海》 * |
| 张一: "基于卡尔曼滤波的短时交通流量预测模型研究", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》 * |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105118294A (en) * | 2015-09-25 | 2015-12-02 | 山东易构软件技术股份有限公司 | Short-time traffic flow prediction method based on state pattern |
| CN105957329A (en) * | 2016-06-29 | 2016-09-21 | 肖锐 | Intelligentized information system for highway |
| CN105957329B (en) * | 2016-06-29 | 2019-04-19 | 芜湖达成储运有限公司 | A kind of highway information intelligence system |
| CN107844848A (en) * | 2016-09-20 | 2018-03-27 | 中国移动通信集团湖北有限公司 | A kind of region flow of the people Forecasting Methodology and system |
| CN107844848B (en) * | 2016-09-20 | 2020-12-29 | 中国移动通信集团湖北有限公司 | Regional pedestrian flow prediction method and system |
| CN106951976A (en) * | 2016-10-12 | 2017-07-14 | 华南理工大学 | A kind of bus passenger flow Forecasting Methodology based on pattern classification |
| CN108090241A (en) * | 2016-11-23 | 2018-05-29 | 财团法人工业技术研究院 | trend variable identification method and system of continuous process |
| CN107066831A (en) * | 2017-05-19 | 2017-08-18 | 君晟合众（北京）科技有限公司 | A kind of regional complex environmental assessment techniques, apparatus and system |
| CN107085942A (en) * | 2017-06-26 | 2017-08-22 | 广东工业大学 | A kind of traffic flow forecasting method based on wolf pack algorithm, apparatus and system |
| CN107610470A (en) * | 2017-10-31 | 2018-01-19 | 迈锐数据（北京）有限公司 | A kind of traffic congestion evaluation method and device |
| CN108053646A (en) * | 2017-11-22 | 2018-05-18 | 华中科技大学 | Traffic characteristic acquisition methods, Forecasting Methodology and system based on time-sensitive feature |
| CN108922183A (en) * | 2018-07-17 | 2018-11-30 | 重庆大学 | A kind of highway short term traffic forecasting method of space time correlation degree in conjunction with SVR |
| CN109241077A (en) * | 2018-08-30 | 2019-01-18 | 东北大学 | Production target variation tendency visual query system and method based on similitude |
| CN109559512A (en) * | 2018-12-05 | 2019-04-02 | 北京掌行通信息技术有限公司 | A kind of regional traffic flow prediction technique and device |
| CN111862583A (en) * | 2019-04-24 | 2020-10-30 | 北京嘀嘀无限科技发展有限公司 | Traffic flow prediction method and device |
| CN111862583B (en) * | 2019-04-24 | 2021-12-03 | 北京嘀嘀无限科技发展有限公司 | Traffic flow prediction method and device |
| CN111862592A (en) * | 2020-05-27 | 2020-10-30 | 浙江工业大学 | Traffic flow prediction method based on RGCN |
| CN117437786A (en) * | 2023-12-21 | 2024-01-23 | 速度科技股份有限公司 | Real-time traffic flow prediction method based on artificial intelligence for traffic network |
| CN117437786B (en) * | 2023-12-21 | 2024-02-27 | 速度科技股份有限公司 | Real-time traffic flow prediction method based on artificial intelligence for traffic network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104821082B (en) | 2017-01-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104821082B (en) | Short-time traffic flow prediction method based on integrated evaluation | |
| CN107610469B (en) | Day-dimension area traffic index prediction method considering multi-factor influence | |
| CN110018670B (en) | Industrial process abnormal working condition prediction method based on dynamic association rule mining | |
| CN111882869B (en) | Deep learning traffic flow prediction method considering adverse weather | |
| CN102081846A (en) | Expressway charge data track matching based traffic state recognition method | |
| CN108417033A (en) | Expressway traffic accident analysis prediction technique based on multi-dimensional factors | |
| CN104090974B (en) | The Application of Data Mining of the follow-up water of extension reservoir and system | |
| CN104809877A (en) | Expressway site traffic state estimation method based on feature parameter weighted GEFCM algorithm | |
| CN113159374B (en) | Data-driven urban traffic flow rate mode identification and real-time prediction early warning method | |
| CN112508237B (en) | Rain type region division method based on data analysis and real-time rain type prediction method | |
| CN104376384A (en) | Typhoon day maximum daily load prediction system based on power big data analysis | |
| CN115907822A (en) | Load characteristic index relevance mining method considering region and economic influence | |
| CN113836808A (en) | PM2.5 deep learning prediction method based on heavy pollution feature constraint | |
| CN111582603B (en) | Intelligent early warning method for coal and gas outburst based on multi-source information fusion | |
| CN111145535B (en) | Travel time reliability distribution prediction method under complex scene | |
| CN117251722A (en) | Intelligent traffic management system based on big data | |
| CN116739376A (en) | Highway pavement preventive maintenance decision method based on data mining | |
| CN108053646B (en) | Traffic characteristic obtaining method, traffic characteristic prediction method and traffic characteristic prediction system based on time sensitive characteristics | |
| CN117892165B (en) | Low-temperature disaster agricultural influence prediction method based on disaster analysis | |
| CN111680888A (en) | Method for determining road network capacity based on RFID data | |
| CN117370759A (en) | Gas consumption prediction method based on artificial intelligence | |
| CN113268867B (en) | Glacier debris flow area early warning method based on hydrothermal combination | |
| CN106779241B (en) | Short-term passenger flow prediction method for rail transit | |
| Fior et al. | Estimating the incidence of adverse weather effects on road traffic safety using time series embeddings | |
| CN112633592A (en) | Vehicle frequent-running route calculation method and system based on machine learning clustering algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170118 Termination date: 20190429 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |