CN107331161A - speed prediction method - Google Patents
speed prediction method Download PDFInfo
- Publication number
- CN107331161A CN107331161A CN201610272324.XA CN201610272324A CN107331161A CN 107331161 A CN107331161 A CN 107331161A CN 201610272324 A CN201610272324 A CN 201610272324A CN 107331161 A CN107331161 A CN 107331161A
- Authority
- CN
- China
- Prior art keywords
- speed
- different
- time points
- numerical value
- different time
- 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
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
- G08G1/052—Detecting movement of traffic to be counted or controlled with provision for determining speed or overspeed
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Traffic Control Systems (AREA)
Abstract
A kind of speed prediction method, comprises the following steps：Calculate a last-period forecast speed；Calculate a long-range forecasting speed；A hybrid predicting speed is calculated using the last-period forecast speed and the long-range forecasting speed；And the prediction speed in each section calculates the running time needed for whole piece path in passage path.The speed prediction method of the present invention, while having considered recent correlation and correlation at a specified future date, it is more accurate to predict.
Description
Technical field
The invention relates to a kind of speed prediction method.
Background technology
Existing traffic forecast mechanism includes that neural network is calculated, data are prospected, machine learning, statistical analysis and fuzzy
Algorithm (Fuzzy algorithm).At a specified future date or last-period forecast is used only in above-mentioned carried traffic forecast mechanism mostly.If only using
Last-period forecast, is assessed using only recent data, as predicted time increases, and the degree of accuracy will decline.If however, only with remote
Phase is predicted, is assessed using only data at a specified future date, when accident or construction occur in the recent period, it is impossible to which immediate reaction, the degree of accuracy also can
Decline.
The content of the invention
The present invention provides a kind of speed prediction method, using the vehicle speed data of last-period forecast and long-range forecasting, with reference to two kinds
Predict the outcome to estimate final speed, and the traveling needed for whole piece path is calculated by the prediction speed in each path Zhong Ge sections
Time.
One embodiment of the present invention provides a kind of speed prediction method, to calculate in a selected section when one predicts
The prediction speed of point, it is implemented by a processing unit, is comprised the following steps：Through the processing unit according to a recent speed number
Speed is predicted according to calculating one first；One second is calculated according to a vehicle speed data at a specified future date predict speed through the processing unit；And
The first prediction speed is multiplied by one first weight through the processing unit, the second prediction speed is multiplied by one second weight,
And said two devices superposition is obtained into a hybrid predicting speed, wherein, the recent vehicle speed data is that the pre- timing points take forward a spy
In the vehicle speed data of all vehicles for selecting sections of road in fixed time interval, the vehicle speed data at a specified future date is that the pre- timing points are special with one
Fixed cycle takes forward the vehicle speed data in all vehicles for selecting sections of road in an at least cycle time point.
In some embodiments, wherein the specific period is one day, one week, the one of January and more than 1 year.
In some embodiments, wherein first weight and second weight and be 1.
In some embodiments, wherein the first prediction speed is not by one the 3rd weight with one first statistics speed
With the summation of multiple products at time point, the multiple products of one the 4th weight and one first error amount in different time points are added
Obtained by summation, wherein, first error amount is defined as the difference of each time point actual vehicle speed and prediction speed, the first statistics car
Speed is in the selected section, in the average speed of all driving vehicles of each time point.
In some embodiments, wherein the 3rd weight has different numerical value at different time points, first statistics
Speed has different numerical value at different time points, and the 4th weight has different numerical value at different time points, first mistake
Difference has different numerical value at different time points.
In some embodiments, wherein the second prediction speed is not by one the 5th weight with one second statistics speed
With the summation of multiple products at time point, the multiple products of one the 6th weight and one second error amount in different time points are added
Obtained by summation, wherein, second error amount is defined as the difference of each time point actual vehicle speed and prediction speed, the second statistics car
Speed is in the selected section, in the average speed of all driving vehicles of each time point.
In some embodiments, wherein the 5th weight has different numerical value at different time points, second statistics
Speed has different numerical value at different time points, and the 6th weight has different numerical value at different time points, second mistake
Difference has different numerical value at different time points.
One embodiment of the present invention provides a kind of speed prediction method, it is adaptable to calculate in a selected section in a prediction
The one of time point repeats the hybrid predicting speed of k time, and it is implemented by a processing unit, and wherein k is a positive integer, including following step
Suddenly：Through the processing unit according to a recent vehicle speed data, a last-period forecast speed for repeating k times, the wherein recent car are calculated
Fast data are that the pre- timing points take forward the vehicle speed data in all vehicles for selecting sections of road in a specific time zone；Through
The processing unit calculates a long-range forecasting speed for repeating k times according to a vehicle speed data at a specified future date, and the wherein vehicle speed data at a specified future date is
The pre- timing points take forward the speed in an at least cycle time point in all vehicles for selecting sections of road with a specific period
Data；The mixed of repetition k times is calculated using the last-period forecast speed of repetition k times and the long-range forecasting speed of repetition k times
Close prediction speed, the hybrid predicting speed of wherein repetition k times is the last-period forecast speed and repetition k times by repetition k times
Long-range forecasting speed be respectively multiplied by different weights and drawn.
In some embodiments, the hybrid predicting speed of wherein repetition k times is the last-period forecast car by repetition k times
Speed and the long-range forecasting speed of repetition k times are respectively multiplied by the step of different weights are drawn and included：Repetition k times it is recent pre-
Measuring car speed is multiplied by one first certain weights, and the long-range forecasting speed for adding repetition k times is multiplied by one second certain weights, draws this
Repeat the hybrid predicting speed of k time, wherein first certain weights are to be more than 0 by one and be less than 1 numerical value company to multiply k times, this
One certain weights and second certain weights and be 1.
In some embodiments, the last-period forecast speed of wherein repetition k times is by one the 3rd certain weights and a weight
Multiple k-i times last-period forecast speed adds one the 4th certain weights and one first in the summation of multiple products of different time points
Summation of the speed in multiple products of different time points is counted, one the 5th certain weights are added with one first prediction error value not
With the summation of multiple products at time point, add one the 6th certain weights and multiply with one first error amount in the multiple of different time points
Obtained by long-pending summation, wherein, first error amount is defined as the difference of each time point actual vehicle speed and prediction speed, and wherein i is
Positive integer, the first statistics speed is in the selected section, in the average speed of all driving vehicles of each time point.
In some embodiments, wherein the 3rd certain weights have different numerical value at different time points, the repetition
The last-period forecast speed of k-i times has different numerical value at different time points, and the 4th certain weights have at different time points
Different numerical value, the first statistics speed has different numerical value at different time points, and the 5th certain weights are when different
Between point have different numerical value, first prediction error value has different numerical value at different time points, and the 6th certain weights exist
There are different numerical value at different time points, and first error amount has different numerical value at different time points.
In some embodiments, the long-range forecasting speed of wherein repetition k times is by one the 7th certain weights and one the
Two statistics speeds add one the 8th certain weights with one second error amount in difference in the summation of multiple products of different time points
Obtained by the summation of multiple products at time point, wherein second error amount is defined as each time point actual vehicle speed and prediction speed
Difference, the second statistics speed is in the selected section, in the average speed of all driving vehicles of each time point.
In some embodiments, wherein the 7th certain weights have different numerical value at different time points, and this second
Statistics speed has different numerical value at different time points, and the 8th certain weights have different numerical value at different time points,
Second error amount has different numerical value at different time points.
Brief description of the drawings
Fig. 1 illustrates the flow chart of the speed prediction method according to first embodiment of the invention.
Fig. 2 illustrates the flow chart of the speed prediction method according to second embodiment of the invention.
Fig. 3 is illustrated calculates schematic diagram according to the last-period forecast speed of second embodiment of the invention.
Fig. 4 is illustrated calculates schematic diagram according to the long-range forecasting speed of second embodiment of the invention.
Embodiment
Multiple embodiments of the present invention, as clearly stated, the details in many practices will be disclosed with accompanying drawing below
It will be explained in the following description.It should be appreciated, however, that the details in these practices is not applied to limit the present invention.Also
It is to say, in some embodiments of the present invention, the details in these practices is non-essential.In addition, for the sake of simplifying accompanying drawing, one
A little known usual structures will be illustrated in the way of simply illustrating in the accompanying drawings with element.
The present invention provides a kind of speed prediction method, using the vehicle speed data of last-period forecast and long-range forecasting, with reference to two kinds
Predict the outcome to estimate final speed, and the traveling needed for whole piece path is calculated by the prediction speed in each path Zhong Ge sections
Time.
The speed prediction method that the present invention is provided, is divided into three phases：The collection stage, set up the model stage and prediction rank
Section.
At the stage of collection, speed prediction method of the invention will collect the speed historical data in city Nei Ge sections.
Then, perform and set up the model stage.In this stage, last-period forecast speed and long-range forecasting car must be first calculated respectively
Speed, in speed prediction method proposed by the invention, each section is all considered as independent individual, will utilize speed historical data
Come for each section set up self return rolling average integrate (Autoregressive Integrated Moving Average,
ARIMA), and using least squares method (Least Squares) each parameter in ARIMA models is asked for.
Fig. 1 illustrates the flow chart of the speed prediction method according to first embodiment of the invention, and this embodiment is applied to meter
Calculate in a selected section in the prediction speed of a pre- timing points, it is implemented by a processing unit.Fig. 1 is refer to, first, is passed through
The processing unit calculates one first according to a recent vehicle speed data and predicts speed (step S110), furthermore, through the processing unit
One second, which is calculated, according to a vehicle speed data at a specified future date predicts speed (step S120).Then, it is through the processing unit that this is first pre-
Measuring car speed is multiplied by one first weight, the second prediction speed is multiplied by into one second weight, and said two devices superposition acquirement one is mixed
Close prediction speed (step S130).Wherein the recent vehicle speed data takes forward selected at this in a specific time zone for the pre- timing points
The vehicle speed data of all vehicles of sections of road, the vehicle speed data at a specified future date is that the pre- timing points are taken forward at least with a specific period
In the vehicle speed data of all vehicles for selecting sections of road in one cycle time point.Wherein, the specific period be one day, one week,
January and more than 1 year first, first weight and second weight and be 1.
The above-described first prediction speed is in different time points by one the 3rd weight with one first statistics speed
The summation of multiple products, is added obtained by one the 4th weight and summation of one first error amount in multiple products of different time points,
Wherein, first error amount is defined as the difference of each time point actual vehicle speed and prediction speed, and the first statistics speed is at this
In selected section, in the average speed of all driving vehicles of each time point.Wherein the 3rd weight has not at different time points
Same numerical value, the first statistics speed has different numerical value at different time points, and the 4th weight has at different time points
Different numerical value, first error amount has different numerical value at different time points.
The above-described second prediction speed is in different time points by one the 5th weight with one second statistics speed
The summation of multiple products, is added obtained by one the 6th weight and summation of one second error amount in multiple products of different time points,
Wherein, second error amount is defined as the difference of each time point actual vehicle speed and prediction speed, and the second statistics speed is at this
In selected section, in the average speed of all driving vehicles of each time point.Wherein the 5th weight has not at different time points
Same numerical value, the second statistics speed has different numerical value at different time points, and the 6th weight has at different time points
Different numerical value, second error amount has different numerical value at different time points.
Fig. 2 illustrates the flow chart of the speed prediction method according to second embodiment of the invention, and this embodiment is applied to meter
Calculate in a selected section in the one of the pre- timing points hybrid predicting speeds for repeating k times, it is implemented by above-mentioned processing unit,
Wherein k is a positive integer.Fig. 2 is refer to, first, through the processing unit according to recent vehicle speed data, one is calculated and repeats k times
Last-period forecast speed, wherein (step S210), the recent vehicle speed data are that the pre- timing points are taken forward in a specific time zone at this
The vehicle speed data of all vehicles of selected sections of road.Then, through the processing unit according to vehicle speed data at a specified future date, a weight is calculated
Multiple k times long-range forecasting speed (step 220), the wherein vehicle speed data at a specified future date are that the pre- timing points are taken forward with a specific period
In the vehicle speed data of all vehicles for selecting sections of road in an at least cycle time point.Next, utilizing the near of repetition k times
Phase predicts the long-range forecasting speed of speed and repetition k times to calculate the hybrid predicting speed (step 230) of repetition k times, its
In the hybrid predicting speed of repetition k times be last-period forecast speed by repetition k times and the long-range forecasting speed of repetition k times
Different weights are respectively multiplied by be drawn.Wherein, step S230 includes, and it is special that the last-period forecast speed that the repetition is k times is multiplied by one first
Determine weight, the long-range forecasting speed for adding repetition k times is multiplied by one second certain weights, draw the hybrid predicting of repetition k times
Speed, wherein first certain weights be by one be more than 0 be less than 1 numerical value even multiply k time, first certain weights and this second
Certain weights and for 1.
The last-period forecast speed that the above-described repetition is k times is to repeat k-i times near by one the 3rd certain weights and one
Phase predicts summation of the speed in multiple products of different time points, adds one the 4th certain weights with one first statistics speed not
With the summation of multiple products at time point, one the 5th certain weights are added with one first prediction error value in many of different time points
The summation of individual product, adds the summation institute of one the 6th certain weights and one first error amount in multiple products of different time points
, wherein, first error amount is defined as the difference of each time point actual vehicle speed and prediction speed, and wherein i is positive integer, and this
One statistics speed is in the selected section, in the average speed of all driving vehicles of each time point.Wherein the 3rd specific weights
Focusing on different time points has different numerical value, and the last-period forecast speed that the repetition is k-i times has different at different time points
Numerical value, the 4th certain weights have different numerical value at different time points, and the first statistics speed has at different time points
Different numerical value, the 5th certain weights have different numerical value at different time points, and first prediction error value is different
There are different numerical value at time point, and the 6th certain weights have different numerical value at different time points, and first error amount is not
There are different numerical value at same time point.
The long-range forecasting speed that the above-described repetition is k times is to be existed by one the 7th certain weights with one second statistics speed
The summation of multiple products of different time points, adds one the 8th certain weights with one second error amount in the multiple of different time points
Obtained by the summation of product, wherein second error amount is defined as the difference of each time point actual vehicle speed and prediction speed, and this second
Statistics speed is in the selected section, in the average speed of all driving vehicles of each time point.Wherein the 7th certain weights
There are different numerical value at different time points, the second statistics speed there are different numerical value at different time points, and the 8th is special
Determining weight has different numerical value at different time points, and second error amount has different numerical value at different time points.
Then, the speed prediction method of the present invention is illustrated using actual example, Fig. 3 is illustrated according to the second embodiment party of the invention
The last-period forecast speed of formula calculates schematic diagram, and Fig. 4, which illustrates to be calculated according to the long-range forecasting speed of second embodiment of the invention, to be shown
It is intended to.First, speed prediction method of the invention asks for last-period forecast speedChronomere can be set as five minutes, or
Other times unit is set according to user's demand.In time t prediction speedIt can be drawn by below equation (1), its
In, Vt-iRefer to time t-i actual vehicle speed, and εt-iRefer to the error in time t-i actual speed and prediction speed.It is intended to predict
Last-period forecast speed in time tFirst to Vt-iPlus weight φi, then to εt-iPlus weight θi.Then, p is added up
Secondary φiVt-iAnd the θ of q timesiεt-i.As for weight parameter φiAnd θi, permeable historical statistical data calculates.Historical statistics
Predetermined speed and actual speed in data are all known, calculate historical statistical data using formula (1), can obtain multigroup
(φi,θi).Then optimal (φ is obtained using least squares methodi,θi)。
However, with the increase of predicted time, the degree of accuracy can decline.Below equation (2) repeats pre- to define recent speed
Survey k times, in time t+k last-period forecast speedIn formula (2), wherein, φi、φi+1、θiAnd θi+1Join for weight
Number, permeable historical statistical data is calculated.Last-period forecast speed during for time t+k-i, Vt+k-1-iWhen referring to
Between t+k-1-i actual vehicle speed,The predicated error of speed and actual vehicle speed, ε are predicted during for time t+k-it+k-1-iFor when
Between t+k-1-i when, prediction speed and actual vehicle speed error.
Then, long-range forecasting speed is asked forUnit interval is one day, or sets other times according to user's demand
Unit.In time t prediction speedIt can be drawn by below equation (3).Wherein, Vt-i×hIt is time t-i × h reality
Speed, and εt-i×hRefer to the error in time t-i × h actual speeds and predetermined speed.It is intended to predict that the long term in time t is pre-
Measuring car speedFirst to Vt-i×hPlus weight λi, then to εt-i×hPlus weight mui.Then, the λ of m times is added upiVt-i×hAnd n times
μiεt-i×h.As for weight parameter λiAnd μi, permeable historical statistical data draws.Predetermined speed and reality in historical statistical data
Border speed is all known, calculates historical data using formula (3), can obtain multigroup (λi,μi), reuse least square
Method obtains optimal (λi,μi)。
Below equation (4) definition speed at a specified future date repeats prediction k times, in time t+k long-range forecasting speedIn public affairs
In formula (4), λiAnd μiFor weight parameter, permeable historical statistical data is calculated.Vt+k-i×hIt is time t+k-i × h actual speed
Degree, εt+k-i×hDuring for time t+k-i × h, the error of prediction speed and actual vehicle speed.
After recent and long-range forecasting speed is all calculated, hybrid predicting speed is obtained followed by hybrid predicting resultIt is determined
Justice in below equation (5),Weight parameter, its value between 0 to 1, andIt can switch over time.In order to allow
There should not be too many, so quantity was limited in one day, so using below equation (6), t quantity is maintained one day,
AsFor example, 12 points of today, 12 points 5 minutes and 12 points 10 minutesValue is all different.But every other day 12 points, 12 points 5 minutes
And 12 points 10 minutesValue and today are the same.
The hybrid predicting speed of k times is repeatedIt is then to use following formula (7), as predicted time is more and more long, Its proportion can be fewer and fewer after company multiplies k times,Ratio regular meeting it is more and more many, k reach one face
After dividing value x, long-range forecasting speed need only be calculated, such as shown in formula (8).It is currently selected and x values is selected
WhenWeight parameterDuring less than 0.1, the k is critical value x.When k is less than x, the hybrid predicting of k times is repeated
SpeedCalculate as shown in below equation (7), when k is more than or equal to x, the hybrid predicting speed of k times is repeatedMeter
Calculate as shown in below equation (8).
In the present invention, using weight parameterTo adjust last-period forecast speed and the ratio shared by long-range forecasting speed
Example.Weight parameterCalculation be through calculate historical statistical data draw multigroup weight parameterAnd drawn using least squares method optimal
One paths are segmented into multiple section r1, r2, r3 ... rz, its distance respectively d1, d2, d3 ... dz, Yu Shi
Between t predetermined speedSo section r1 running time π1Calculation such as formula (9), and section ri
Running time πiCalculation such as formula (10).Finally, the running time T in whole piece path, during by the traveling in each section
Between add up, such as shown in formula (11).
In forecast period, time and path according to input, the ARIMA moulds that the model stage set up are set up using above-mentioned
Type, you can obtain last-period forecast speed, the long-range forecasting speed in each section in outbound path.Then, predicted using above-mentioned formula (7)
Final speed.Then, the prediction speed and distance according to each section, and predicted path running time.
As the magnitude of traffic flow is grown up year by year, how to carry out traffic forecast to accurate and effective rate and disappeared with saving time and the energy
Consumption is particularly significant, and good non-intersection speed and the Forecasting Methodology of route time can not only help driver's planning traveling road
Line, can also help traffic congestion of releiving, and improve traffic.Disclosed speed prediction method, includes at a specified future date pre-
Survey stabilization and average advantage and last-period forecast can react the characteristic of temporary variations, while having considered recent correlation and long term
Correlation, will be more accurate compared to traditional speed prediction method.
Although the present invention is disclosed above with numerous embodiments, so it is not limited to the present invention, any to be familiar with this
Those skilled in the art, without departing from the spirit and scope of the present invention, when can be used for a variety of modifications and variations, therefore the protection model of the present invention
Enclose to work as and be defined depending on the scope of which is defined in the appended claims.
Claims (13)
1. a kind of speed prediction method, it is characterised in that to calculate the prediction speed in a selected section in a pre- timing points,
It is implemented by a processing unit, is comprised the following steps：
(A) calculate one first according to a recent vehicle speed data through the processing unit and predict speed；
(B) calculate one second according to a vehicle speed data at a specified future date through the processing unit and predict speed；And
(C) the first prediction speed is multiplied by one first weight through the processing unit, the second prediction speed is multiplied by one the
Two weights, and said two devices superposition is obtained into a hybrid predicting speed；
Wherein the recent vehicle speed data is that the pre- timing points take forward all cars in the selected sections of road in a specific time zone
Vehicle speed data, the vehicle speed data at a specified future date be the pre- timing points taken forward in an at least cycle time point at this with a specific period
The vehicle speed data of all vehicles of selected sections of road.
2. speed prediction method according to claim 1, it is characterised in that the specific period is one day, one week, January and 1 year
Above one.
3. speed prediction method according to claim 1, it is characterised in that first weight and second weight and be 1.
4. speed prediction method according to claim 1, it is characterised in that the first prediction speed is by one the 3rd weight and one
First statistics speed adds one the 4th weight and one first error amount when different in the summation of multiple products of different time points
Between obtained by the summation of multiple products put, wherein, first error amount is defined as each time point actual vehicle speed and prediction speed
Difference, the first statistics speed is in the selected section, in the average speed of all driving vehicles of each time point.
5. speed prediction method according to claim 4, it is characterised in that the 3rd weight has different at different time points
Numerical value, the first statistics speed has different numerical value at different time points, and the 4th weight has difference at different time points
Numerical value, first error amount has different numerical value at different time points.
6. speed prediction method according to claim 1, it is characterised in that the second prediction speed is by one the 5th weight and one
Second statistics speed adds one the 6th weight and one second error amount when different in the summation of multiple products of different time points
Between obtained by the summation of multiple products put, wherein, second error amount is defined as each time point actual vehicle speed and prediction speed
Difference, the second statistics speed is in the selected section, in the average speed of all driving vehicles of each time point.
7. speed prediction method according to claim 6, it is characterised in that the 5th weight has different at different time points
Numerical value, the second statistics speed has different numerical value at different time points, and the 6th weight has difference at different time points
Numerical value, second error amount has different numerical value at different time points.
8. a kind of speed prediction method, it is characterised in that suitable for calculating the repetition k in a selected section in a pre- timing points
Secondary hybrid predicting speed, it is implemented by a processing unit, and wherein k is a positive integer, including：
Through the processing unit according to a recent vehicle speed data, a last-period forecast speed for repeating k times, the wherein recent car are calculated
Fast data are that the pre- timing points take forward the vehicle speed data in all vehicles for selecting sections of road in a specific time zone；
Through the processing unit according to a vehicle speed data at a specified future date, a long-range forecasting speed for repeating k times, the wherein car at a specified future date are calculated
Fast data are that the pre- timing points take forward all cars for selecting sections of road at this in an at least cycle time point with a specific period
Vehicle speed data；And
The mixing of repetition k times is calculated using the last-period forecast speed of repetition k times and the long-range forecasting speed of repetition k times
Predict speed, the hybrid predicting speed of wherein repetition k times is the last-period forecast speed and repetition k times by repetition k times
Long-range forecasting speed is respectively multiplied by different weights and drawn.
9. speed prediction method according to claim 8, it is characterised in that the hybrid predicting speed that the repetition is k times is heavy by this
The last-period forecast speed of multiple k times and the long-range forecasting speed of repetition k times are respectively multiplied by the step of different weights are drawn and included：
The last-period forecast speed that the repetition is k times is multiplied by one first certain weights, and the long-range forecasting speed for adding repetition k times is multiplied by
One second certain weights, draw the hybrid predicting speed of repetition k times, wherein first certain weights be will one to be more than 0 small
Numerical value in 1 even multiplies k time, first certain weights and second certain weights and be 1.
10. speed prediction method according to claim 9, it is characterised in that the last-period forecast speed that the repetition is k times is by one
Three certain weights with one repeat k-i times last-period forecast speed different time points multiple products summation, add one the 4th
Certain weights, in the summation of multiple products of different time points, add one the 5th certain weights and one the with one first statistics speed
One prediction error value adds one the 6th certain weights with one first error amount or not the summation of multiple products of different time points
With the summation gained of multiple products at time point, wherein, first error amount is defined as each time point actual vehicle speed and pre- measuring car
The difference of speed, wherein i is positive integer, and the first statistics speed is in the selected section, in all driving vehicles of each time point
Average speed.
11. speed prediction method according to claim 10, it is characterised in that the 3rd certain weights have at different time points
Different numerical value, the last-period forecast speed that the repetition is k-i times has different numerical value, the 4th certain weights at different time points
There are different numerical value at different time points, the first statistics speed there are different numerical value at different time points, and the 5th is special
Determining weight has different numerical value at different time points, and first prediction error value has different numerical value at different time points,
6th certain weights have different numerical value at different time points, and first error amount has different numbers at different time points
Value.
12. speed prediction method according to claim 9, it is characterised in that the long-range forecasting speed that the repetition is k times is by one
Seven certain weights, in the summation of multiple products of different time points, add one the 8th certain weights and one with one second statistics speed
Obtained by summation of second error amount in multiple products of different time points, wherein second error amount is defined as each time point reality
The difference of speed and prediction speed, the second statistics speed is in the selected section, in all driving vehicles of each time point
Average speed.
13. speed prediction method according to claim 12, it is characterised in that the 7th certain weights have at different time points
Different numerical value, the second statistics speed has different numerical value at different time points, and the 8th certain weights are when different
Between point have different numerical value, second error amount has different numerical value at different time points.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610272324.XA CN107331161A (en) | 2016-04-28 | 2016-04-28 | speed prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610272324.XA CN107331161A (en) | 2016-04-28 | 2016-04-28 | speed prediction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107331161A true CN107331161A (en) | 2017-11-07 |
Family
ID=60193016
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610272324.XA Pending CN107331161A (en) | 2016-04-28 | 2016-04-28 | speed prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107331161A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111071259A (en) * | 2019-12-27 | 2020-04-28 | 清华大学 | Vehicle speed prediction method, vehicle speed prediction device, vehicle control device, and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101625796A (en) * | 2008-07-10 | 2010-01-13 | 现代自动车株式会社 | Estimation method of traffic information |
| JP2011221960A (en) * | 2010-04-14 | 2011-11-04 | Sumitomo Electric Ind Ltd | Starting/end point traffic volume calculation device, traffic simulator and starting/end point traffic volume calculation method |
| CN103383819A (en) * | 2013-07-19 | 2013-11-06 | 新疆交通科学研究院 | Driver cognitive characteristic based predicting and calculating system for running speeds of vehicles on desert roads |
| CN104464304A (en) * | 2014-12-25 | 2015-03-25 | 北京航空航天大学 | Urban road vehicle running speed forecasting method based on road network characteristics |
| WO2016036024A1 (en) * | 2014-09-01 | 2016-03-10 | 한국교통연구원 | Traffic link speed prediction method and apparatus for same |
-
2016
- 2016-04-28 CN CN201610272324.XA patent/CN107331161A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101625796A (en) * | 2008-07-10 | 2010-01-13 | 现代自动车株式会社 | Estimation method of traffic information |
| JP2011221960A (en) * | 2010-04-14 | 2011-11-04 | Sumitomo Electric Ind Ltd | Starting/end point traffic volume calculation device, traffic simulator and starting/end point traffic volume calculation method |
| CN103383819A (en) * | 2013-07-19 | 2013-11-06 | 新疆交通科学研究院 | Driver cognitive characteristic based predicting and calculating system for running speeds of vehicles on desert roads |
| WO2016036024A1 (en) * | 2014-09-01 | 2016-03-10 | 한국교통연구원 | Traffic link speed prediction method and apparatus for same |
| CN104464304A (en) * | 2014-12-25 | 2015-03-25 | 北京航空航天大学 | Urban road vehicle running speed forecasting method based on road network characteristics |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111071259A (en) * | 2019-12-27 | 2020-04-28 | 清华大学 | Vehicle speed prediction method, vehicle speed prediction device, vehicle control device, and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102044149B (en) | City bus operation coordinating method and device based on time variant passenger flows | |
| Li et al. | Connected vehicles based traffic signal timing optimization | |
| Guo et al. | Real-time short-term traffic speed level forecasting and uncertainty quantification using layered Kalman filters | |
| CN105023433B (en) | A kind of traffic abnormal events of expressway coverage predictor method | |
| CN104183119B (en) | Based on the anti-arithmetic for real-time traffic flow distribution forecasting method pushed away of section OD | |
| Bhatta et al. | Errors in variables in multinomial choice modeling: A simulation study applied to a multinomial logit model of travel mode choice | |
| CN105513414A (en) | Parking-lot parking space predication and recommendation method based on real-time traffic query and cloud model | |
| Osorio et al. | Simulation-based optimization: achieving computational efficiency through the use of multiple simulators | |
| CN103116808A (en) | Method of real-timely predicting short time traffic flow of express way | |
| CN107346460A (en) | Following operating mode Forecasting Methodology based on the lower front truck operation information of intelligent network contact system | |
| Yao et al. | Hybrid methods for short-term traffic flow prediction based on ARIMA-GARCH model and wavelet neural network | |
| JPH08503317A (en) | Traffic parameter prediction method | |
| CN104157132A (en) | Self-adaptive dynamic optimization method for bus dispatching timetable | |
| CN103903430A (en) | Dynamic fusion type travel time predicting method with multi-source and isomorphic data adopted | |
| CN104183134A (en) | Expressway short-time traffic flow forecast method based on intelligent car type classification | |
| CN103279669A (en) | Method and system for simulating calculation of transport capacity of urban rail transit network | |
| CN104781863A (en) | Method for predicting future travel time on link | |
| López et al. | Predictive model for energy consumption of battery electric vehicle with consideration of self-uncertainty route factors | |
| Mittal et al. | Network flow relations and travel time reliability in a connected environment | |
| CN107730893A (en) | A kind of shared bus website passenger flow forecasting for multidimensional characteristic of being gone on a journey based on passenger | |
| CN103745602B (en) | A kind of traffic flow forecasting method average based on sliding window | |
| Kedia et al. | Fuzzy logic approach in mode choice modelling for education trips: A case study of Indian metropolitan city | |
| TWI623920B (en) | Speed prediction method | |
| CN105185107A (en) | GPS-based traffic running tendency prediction method | |
| CN105575105B (en) | Dispatching method and equipment for the vehicles |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20171107 |