CN112182962A - Hybrid electric vehicle running speed prediction method - Google Patents

Hybrid electric vehicle running speed prediction method Download PDF

Info

Publication number
CN112182962A
CN112182962A CN202011015662.8A CN202011015662A CN112182962A CN 112182962 A CN112182962 A CN 112182962A CN 202011015662 A CN202011015662 A CN 202011015662A CN 112182962 A CN112182962 A CN 112182962A
Authority
CN
China
Prior art keywords
speed
time
working conditions
vehicle
cluster
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
Application number
CN202011015662.8A
Other languages
Chinese (zh)
Inventor
赵红
袁焕涛
马永志
崔翔宇
潘广纯
牟亮
仇俊政
李燕
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Qingdao University
Original Assignee
Qingdao University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Qingdao University filed Critical Qingdao University
Priority to CN202011015662.8A priority Critical patent/CN112182962A/en
Publication of CN112182962A publication Critical patent/CN112182962A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/10—Numerical modelling

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Probability & Statistics with Applications (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Computer Hardware Design (AREA)
  • Geometry (AREA)
  • Electric Propulsion And Braking For Vehicles (AREA)

Abstract

The invention discloses a method for predicting the running speed of a hybrid electric vehicle, which belongs to the technical field of vehicle real-time working condition prediction and comprises the following steps: s10, classifying the actual working conditions by using a K-means clustering analysis method; s20, constructing typical working conditions of different working conditions according to the classification result and by referring to the standard cycle working conditions; s30, aiming at typical working conditions, simulating a driving vehicle speed change rule by adopting a prediction model based on a Markov chain; s40, calculating a state transition probability matrix; and S50, based on the solution of the Markov chain prediction model and the state transition matrix, predicting the real-time running condition by using a matlab/markov toolbox. The method can predict the running speed of the automobile in a future period of time by adopting a Markov chain method under different types of working conditions, can improve the adaptivity and the accuracy of the control of the automobile energy management strategy under complex and changeable real-time working conditions, and can improve the performance of the control strategy when the automobile runs.

Description

Hybrid electric vehicle running speed prediction method
Technical Field
The invention belongs to the technical field of automobile real-time working condition prediction, and particularly relates to a hybrid electric vehicle future speed state prediction method based on a Markov chain model.
Background
With the increasing automobile reserves in China, the problems of energy consumption resource shortage and exhaust emission pollution occur, and the automobile industry is forced to change to energy conservation and emission reduction. The hybrid electric vehicle is used as a new energy vehicle which is most widely applied at present by virtue of excellent overall vehicle performance, the fuel economy of the hybrid electric vehicle determines the economic driving capacity of the overall vehicle, and a reasonable control strategy is a key for improving the fuel economy. At present, the most applied control strategy is a rule-based control strategy, and the system is simple and easy to implement, but cannot ensure optimality. After that, a large number of scholars have made optimization-based energy control methods to minimize the cost function through mathematical analysis, but their huge calculation amount makes them unusable for real-time control. An equivalent fuel consumption minimization (ECMS) is an instantaneous optimal method, but an oil-electricity conversion system needs to be changed along with the change of a working state, and the optimality cannot be guaranteed. Model Predictive Control (MPC) is a very promising control technology in a control algorithm, can process a multi-input multi-output system, constraint processing and predict future information, and ensures the global optimum characteristic of the system in a prediction time domain, but the method needs future driving information, still faces the problems of large calculation amount and long time consumption, and influences the real-time performance and control effect of a control strategy when the information of the future working conditions of the automobile is slowly updated.
Therefore, a hybrid electric vehicle running speed prediction method capable of predicting the running speed of the vehicle in a future period of time so as to improve the adaptivity and the accuracy of the vehicle energy management strategy control under complex and variable real-time working conditions is urgently needed.
Disclosure of Invention
The invention aims to provide a hybrid electric vehicle running speed prediction method which can predict the running speed of an automobile in a period of time in the future so as to improve the adaptivity and the accuracy of the energy management strategy control of the automobile under complex and variable real-time working conditions, and the invention adopts the following technical scheme:
a hybrid electric vehicle driving speed prediction method comprises the following steps:
s10, classifying the actual working conditions by using a K-means clustering analysis method;
s20, constructing typical working conditions of different working conditions according to the classification result and by referring to the standard cycle working conditions;
s30, aiming at the typical working condition, simulating a driving vehicle speed change rule by adopting a prediction model based on a Markov chain;
s40, calculating a state transition probability matrix;
and S50, based on the solution of the Markov chain prediction model and the state transition matrix, predicting the real-time running condition by using a matlab/markov toolbox.
Further, in step S10, the following characteristic parameters are selected: the actual working conditions are classified according to the average speed, the maximum acceleration, the minimum deceleration, the idle speed time ratio, the acceleration time ratio, the deceleration time ratio, the uniform speed time ratio, the acceleration average value and the speed standard deviation.
Further, eliminating the correlation of the characteristic parameters by using a correlation analysis method, and selecting and replacing the characteristic parameters, wherein the selection steps are as follows:
s11, calculating the characteristic parameters, wherein the calculation formula is as follows:
Figure BDA0002698971230000021
s12, solving a characteristic parameter matrix by using a correlation coefficient method, wherein the formula of the correlation coefficient calculation is as follows:
Figure BDA0002698971230000022
in the formula (I), the compound is shown in the specification,
Figure BDA0002698971230000023
Figure BDA0002698971230000024
Figure BDA0002698971230000025
wherein r is a correlation coefficient, n is the total number of the working condition blocks, X represents different characteristic parameters,
Figure BDA0002698971230000026
and
Figure BDA0002698971230000027
respectively, represent the mean values of different characteristic parameters.
Further, the data classification is carried out by calculating the degree of affinity and sparseness among samples by utilizing K-means clustering analysis, so that the data in the same class have larger feature similarity, the difference among different classes is larger, and the clustering analysis process is as follows:
from all data points { x1，x2，…，xnIn the method, k points are randomly selected as the centers of primary clustering (k)1，k2，…，knCalculating the distances from the rest sample points to the selected k clustering centers, processing the distances from the rest sample points to the selected k clustering centers according to a minimum distance principle, and classifying each sample point by taking k categories as centers to form k clusters;
(1) selecting k points from n data points of the sample according to a minimum distance principle as the center of the initial clustering;
(2) classifying each sample to a nearest cluster center by using Euclidean distance, wherein the expression method of the Euclidean distance is shown as the following formula:
Figure BDA0002698971230000031
wherein x isiIs the ith sample point, kiIs the ith cluster center;
(3) carrying out iterative calculation again to update the clustering center of each cluster;
(4) repeating the steps (1) to (3) until the cluster center of each cluster is not changed.
Further, the effect of the working condition clustering analysis is represented by an index function F, wherein the larger the value of F is, the more remarkable the clustering effect is; the objective function F is expressed as:
Figure BDA0002698971230000032
wherein F is an objective function; d is the sum of the distances between the centers of various clusters; c is the average value of the sum of the distances from the sample point to the cluster centers of all types;
wherein D and C may be represented by the following formulae:
Figure BDA0002698971230000033
wherein k is the number of cluster center values, ciIs the i-th class center; c. CjIs a class j center; n isiThe number of samples in the ith clustering process; x is the number ofjIs the characteristic value of the jth sample.
Further, classifying all working conditions into results of K clusters according to K-means clustering analysis, planning time ratios of various clusters, and constructing the time occupied by the typical working conditions according to different working condition categories is represented by the following formula:
Figure BDA0002698971230000041
in the formula, tiThe time of the cluster i in the final working condition is taken; t is tcycleOccupying time for the working condition; t is tallThe total time of all actual data; t is ti，jThe time occupied by the working condition block j in the i-type cluster, njThe number of all working condition blocks belonging to the i-type cluster is determined;
and selecting representative working condition blocks by using a K-means clustering analysis and taking a clustering center as a standard, and combining the representative working condition blocks into the typical working condition.
Further, in step S30, assuming that the vehicle speed of the vehicle at each time is not related to the historical information and is determined only by the current information, it is considered that the change of the vehicle speed is a markov process, and the markov chain model is used to simulate the change rule of the vehicle speed, so as to predict the future vehicle speed under the typical condition.
Further, in step S40, a state transition probability matrix is calculated, and the traveling vehicle speed is discretized into finite numerical values by the nearest neighbor method:
vS∈{v1，v2，…，vN}
dividing the speed of the running process into 100 possible states, wherein the speed discrete interval takes 5Km/h, the running speed state number U is {1, 2, …, 25}, and the speed of the running vehicle is determined by the current speed state UiVehicle speed state U to the next momentjIs a state transition probability Pi，j；
Then P isi，j＝P(v(k+1)＝vj|v(k)＝vi)
In the formula, Pi，jIs the ith row and the jth column element of the state transition probability matrix and satisfies Pi，j≥0，
Figure BDA0002698971230000042
Pi，jThe value of (d) can be obtained by maximum likelihood estimation:
Figure BDA0002698971230000051
wherein i, j is 1, 2, …, N, Fi，jIndicating that the vehicle speed is from viTransfer to vjNumber of times of (F)iFor running vehicle speed from viSum of the number of transfers.
Further, by calculating the transition probability and the transition times from the current running vehicle speed to the next running vehicle speed, combining each state probability value to generate a Markov transition probability matrix P:
Figure BDA0002698971230000052
further, the predicted speed value in step S50 is solved according to the following equation:
v(k)＝[(Uk-1)+rk]d
wherein v (k) is the vehicle running speed at the time k; u shapekThe driving speed state at the moment k is obtained; d is the speed state division length, and the value is 5; r iskThe uniformly distributed random numbers generated for time k matlab.
The invention has the beneficial effects that:
the invention provides a method for predicting the future speed of a hybrid electric vehicle based on the Markov basic principle, which can predict the running speed of the vehicle in a future period of time by adopting a Markov chain method under different types of working conditions, can improve the adaptivity and accuracy of the control of an energy management strategy of the vehicle under complex and variable real-time working conditions, and improve the performance of the control strategy exerted when the vehicle runs. The Markov chain model forecasts the running speed and the contrast error of the actual typical working condition is small, and the following performance is good.
Drawings
FIG. 1 exemplary Condition construction flow
FIG. 2 exemplary operating conditions
FIG. 3 exemplary operating condition predicted transition probability map
FIG. 4 exemplary Condition vehicle speed prediction map
FIG. 5 characteristic parameters of the operating conditions
Detailed Description
Example 1
A hybrid electric vehicle driving speed prediction method comprises the following steps:
s10, classifying the actual working conditions by using a K-means clustering analysis method;
selecting the following characteristic parameters: average velocity (Km/h), maximum acceleration m/s2Minimum deceleration m/s2Idle time ratio (%), acceleration time ratio (%), deceleration time ratio (%), uniform speed time ratio (%), average acceleration value m/s2And classifying the actual working conditions by the speed standard deviation (Km/h).
Eliminating the correlation of the characteristic parameters by using a correlation analysis method, and selecting and replacing the characteristic parameters, wherein the selection steps are as follows:
s11, calculating the characteristic parameters, wherein the calculation formula is as follows:
Figure BDA0002698971230000061
s12, solving a characteristic parameter matrix by using a correlation coefficient method, wherein the formula of the correlation coefficient calculation is as follows:
Figure BDA0002698971230000062
in the formula (I), the compound is shown in the specification,
Figure BDA0002698971230000063
Figure BDA0002698971230000064
Figure BDA0002698971230000065
wherein r is a correlation coefficient, n is the total number of the working condition blocks, X represents different characteristic parameters,
Figure BDA0002698971230000066
and
Figure BDA0002698971230000067
respectively, represent the mean values of different characteristic parameters.
In this example, the average acceleration m/s is selected2Average speed (Km/h), idle time ratio (%) as representative characteristic parameters (shown in fig. 5).
The data classification is carried out by utilizing the K-means clustering analysis and calculating the degree of affinity and sparseness among samples, so that the data in the same class have larger characteristic similarity, the difference is larger among different classes, and the clustering analysis process is as follows:
from all data points { x1，x2，…，xnIn the method, k points are randomly selected as the centers of primary clustering (k)1，k2，…，knCalculating the distances from the rest sample points to the selected k clustering centers, processing the distances from the rest sample points to the selected k clustering centers according to a minimum distance principle, and classifying each sample point by taking k categories as centers to form k clusters;
(1) selecting k points from n data points of the sample according to a minimum distance principle as the center of the initial clustering;
(2) classifying each sample to a nearest cluster center by using Euclidean distance, wherein the expression method of the Euclidean distance is shown as the following formula:
Figure BDA0002698971230000071
wherein x isiIs the ith sample point, kiIs the ith cluster center;
(3) carrying out iterative calculation again to update the clustering center of each cluster;
(4) repeating the steps (1) to (3) until the cluster center of each cluster is not changed.
In this embodiment, the class of the cluster to which each sample point belongs is adjusted by continuously iterating and updating, and if the centers of two consecutive iterations do not change any more, it is indicated that the algorithm has converged, and the sample data has been reasonably classified into each cluster.
In this embodiment, the actual working conditions are divided into four different working conditions, namely congestion, urban area, suburban area, and express way.
S20, constructing typical working conditions of different working conditions according to the classification result and by referring to the standard cycle working conditions;
the effect of clustering analysis on four different working conditions is represented by an index function F, wherein the clustering effect is more obvious when the value of F is larger; the objective function F is expressed as:
Figure BDA0002698971230000081
wherein F is an objective function; d is the sum of the distances between the centers of various clusters; c is the average value of the sum of the distances from the sample point to the cluster centers of all types;
wherein D and C may be represented by the following formulae:
Figure BDA0002698971230000082
wherein k is the number of cluster center values, ciIs the i-th class center; c. CjIs a class j center; n isiThe number of samples in the ith clustering process; x is the number ofjIs the characteristic value of the jth sample.
The process and steps of typical working condition construction are given as shown in fig. 1, and the process comprises the steps of determining a driving route and a collecting method, collecting real vehicle data and processing the data, selecting characteristic parameters and performing cluster analysis and synthesizing typical working conditions. The method comprises the steps of determining a data acquisition route according to a modern intelligent traffic information system and an electronic map, acquiring data through experimental equipment, preprocessing the data to achieve the purpose of noise reduction, and then selecting proper characteristic parameters to analyze the data.
Classifying all working conditions into K clusters according to K-means clustering analysis, planning time ratios of various clusters, and expressing the time occupied by constructing typical working conditions by different working condition classes by the following formula:
Figure BDA0002698971230000083
in the formula, tiThe time of the cluster i in the final working condition is taken; t is tcycleOccupying time for the working condition; t is tallThe total time of all actual data; t is ti，jThe time occupied by the working condition block j in the i-type cluster, njThe number of all working condition blocks belonging to the i-type cluster is determined;
and selecting representative working condition blocks by using a K-means clustering analysis and taking a clustering center as a standard, and combining the representative working condition blocks into a typical working condition (as shown in figure 2).
S30, aiming at typical working conditions, simulating a driving vehicle speed change rule by adopting a prediction model based on a Markov chain;
the markov process is summarized as a process for a dynamic system to transfer from one state to another state, is independent of the historical state, is determined only by the current state, and is expressed by a mathematical formula as follows:
P{X(tn)≤xn|X(tn-1)＝xn-1，X(tn-2)＝xn-2，…，X(t1)＝x1}
＝P{X(tn)≤xn|X(tn-1)＝xn-1}
then some random process { X (T), T ∈ T } is a Markov process.
If the set of temporal parameters and the spatial state variables of this stochastic process are discrete, then what characterizes this is a markov chain, expressed in terms of mathematical formulas:
assume a random process { XtT ∈ T }, under the condition Xt＝xiI is 1, 2, …, n-1, and satisfies:
p{Xn＝xn|X1＝x1，X2＝x2，…，Xn-1＝xn-1}＝P{Xn＝xn|Xn-
then the random process is said to be { X }tAnd T ∈ T } is a Markov chain.
The speed of the vehicle at each moment is assumed to be unrelated to historical information and is determined only by current information, the change of the running speed of the vehicle is considered to be a Markov process, a Markov chain model is used for simulating the change rule of the running speed, and the future running speed is predicted under typical working conditions.
S40, calculating a state transition probability matrix;
calculating a state transition probability matrix, and dispersing the running speed into a limited numerical value by using a neighbor method:
vs∈{v1，v2，…，vN}
dividing the speed of the running process into 100 possible states, wherein the speed discrete interval takes 5Km/h, the running speed state number U is {1, 2, …, 25}, and the speed of the running vehicle is determined by the current speed state UiVehicle speed state U to the next momentjIs a state transition probability Pi，j；
Then P isi，j＝P(v(k+1)＝vj|v(k)＝vi)
In the formula, Pi，jIs the ith row and the jth column element of the state transition probability matrix and satisfies Pi，j≥0，
Figure BDA0002698971230000091
Pi，jThe value of (d) can be obtained by maximum likelihood estimation:
Figure BDA0002698971230000101
wherein i, j is 1, 2, …, N, Fi，jIndicating that the vehicle speed is from viTransfer to vjNumber of times of (F)iFor running vehicle speed from viSum of the number of transfers.
Fig. 3 is a vehicle speed transition probability diagram of a typical driving condition, which shows a vehicle speed state transition rule, and the vehicle speed transition probability three-dimensional histograms with step lengths of 2s and 5s are respectively from left to right, so that under the typical driving condition, along with the increase of the predicted step length, the state transition probability value of the vehicle driving speed gradually decreases, and the randomness of the transition probability increases, so that the future vehicle speed state of the vehicle under the real-time driving condition has great uncertainty.
Calculating the transition probability and the transition times from the current running vehicle speed to the next running vehicle speed, and combining the state probability values to generate a Markov transition probability matrix P:
Figure BDA0002698971230000102
s50, based on the solution of the Markov chain prediction model and the state transition matrix, predicting the real-time running condition by using a matlab/markov toolbox, and solving the prediction speed value according to the following formula:
v(k)＝[(Uk-1)+rk]d
wherein v (k) is the vehicle running speed at the time k; u shapekThe driving speed state at the moment k is obtained; d is the speed state division length, and the value is 5; r iskThe uniformly distributed random numbers generated for time k matlab.
And figure 4 is a typical working condition vehicle speed prediction graph, and the vehicle speed predicted by the Markov chain model at the future adjacent moment can accurately follow the actual vehicle speed by comparing the typical working condition vehicle speed prediction graph and analyzing, so that the effectiveness of the Markov chain prediction model is further verified.
The above description is only a preferred embodiment of the present invention, and is not intended to limit the technical scope of the present invention, so that any minor modifications, equivalent changes and modifications made to the above embodiment according to the technical spirit of the present invention are within the technical scope of the present invention.

Claims (10)

1. A hybrid electric vehicle running speed prediction method is characterized by comprising the following steps:
s10, classifying the actual working conditions by using a K-means clustering analysis method;
s20, constructing typical working conditions of different working conditions according to the classification result and by referring to the standard cycle working conditions;
s30, aiming at the typical working condition, simulating a driving vehicle speed change rule by adopting a prediction model based on a Markov chain;
s40, calculating a state transition probability matrix;
and S50, based on the solution of the Markov chain prediction model and the state transition matrix, predicting the real-time running condition by using a matlab/markov toolbox.
2. The method for predicting the traveling vehicle speed of a hybrid vehicle according to claim 1, wherein the following characteristic parameters are selected in step S10: the actual working conditions are classified according to the average speed, the maximum acceleration, the minimum deceleration, the idle speed time ratio, the acceleration time ratio, the deceleration time ratio, the uniform speed time ratio, the acceleration average value and the speed standard deviation.
3. The method for predicting the traveling speed of a hybrid vehicle according to claim 2, wherein the correlation of the characteristic parameters is eliminated by a correlation analysis method, and the representative characteristic parameters are selected and replaced by the following steps:
s11, calculating the characteristic parameters, wherein the calculation formula is as follows:
Figure FDA0002698971220000011
s12, solving a characteristic parameter matrix by using a correlation coefficient method, wherein the formula of the correlation coefficient calculation is as follows:
Figure FDA0002698971220000012
in the formula (I), the compound is shown in the specification,
Figure FDA0002698971220000013
Figure FDA0002698971220000014
Figure FDA0002698971220000015
wherein r is a correlation coefficient, n is the total number of the working condition blocks, X represents different characteristic parameters,
Figure FDA0002698971220000021
and
Figure FDA0002698971220000022
respectively, represent the mean values of different characteristic parameters.
4. The method for predicting the driving speed of the hybrid electric vehicle according to claim 3, wherein the data classification is performed by calculating the degree of affinity and sparseness among samples by using K-means cluster analysis, so that the data in the same class have larger feature similarity, and the difference between different classes is larger, and the cluster analysis process is as follows:
from all data points { x1，x2，…，xnIn the method, k points are randomly selected as the centers of primary clustering (k)1，k2，…，knCalculating the distances from the rest sample points to the selected k clustering centers, processing the distances from the rest sample points to the selected k clustering centers according to a minimum distance principle, and classifying each sample point by taking k categories as centers to form k clusters;
(1) selecting k points from n data points of the sample according to a minimum distance principle as the center of the initial clustering;
(2) classifying each sample to a nearest cluster center by using Euclidean distance, wherein the expression method of the Euclidean distance is shown as the following formula:
Figure FDA0002698971220000023
wherein x isiIs the ith sample point, kiIs the ith cluster center;
(3) carrying out iterative calculation again to update the clustering center of each cluster;
(4) repeating the steps (1) to (3) until the cluster center of each cluster is not changed.
5. The hybrid electric vehicle running speed prediction method according to claim 4, characterized in that the effect of the working condition cluster analysis is expressed by an index function F, wherein the larger the value of F is, the more remarkable the clustering effect is; the objective function F is expressed as:
Figure FDA0002698971220000024
wherein F is an objective function; d is the sum of the distances between the centers of various clusters; c is the average value of the sum of the distances from the sample point to the cluster centers of all types;
wherein D and C may be represented by the following formulae:
Figure FDA0002698971220000031
wherein k is the number of cluster center values, ciIs the i-th class center; c. CjIs a class j center; n isiThe number of samples in the ith clustering process; x is the number ofjIs the characteristic value of the jth sample.
6. The method according to claim 5, wherein the time taken for constructing the typical working conditions in different working condition categories is represented by the following formula:
Figure FDA0002698971220000032
in the formula, tiThe time of the cluster i in the final working condition is taken; t is tcycleOccupying time for the working condition; t is tallThe total time of all actual data; t is ti，jThe time occupied by the working condition block j in the i-type cluster, njThe number of all working condition blocks belonging to the i-type cluster is determined;
and selecting representative working condition blocks by using a K-means clustering analysis and taking a clustering center as a standard, and combining the representative working condition blocks into the typical working condition.
7. The method according to claim 6, wherein in step S30, assuming that the vehicle speed of the vehicle at each time is independent of the historical information and is determined only by the current information, the change of the vehicle speed is considered to be a markov process, and a markov chain model is used to simulate the change rule of the vehicle speed to predict the future vehicle speed under the typical condition.
8. The method of predicting the traveling vehicle speed of a hybrid vehicle according to claim 1, wherein in step S40, a state transition probability matrix is calculated, and the traveling vehicle speed is discretized into finite numerical values by a nearest neighbor method:
vs∈{v1，v2，…，vN}
dividing the speed of the running process into 100 possible states, wherein the speed discrete interval takes 5Km/h, the running speed state number U is {1, 2, …, 25}, and the speed of the running vehicle is determined by the current speed state UiVehicle speed state U to the next momentjIs a state transition probability Pi，j；
Then P isi，j＝P(v(k+1)＝vj|v(k)＝vi)
In the formula, Pi，jIs the ith row and the jth column element of the state transition probability matrix and satisfies Pi，j≥0，
Figure FDA0002698971220000041
j＝0，1，…；
Pi，jThe value of (d) can be obtained by maximum likelihood estimation:
Figure FDA0002698971220000042
wherein i, j is 1, 2, …, N, Fi，jIndicating that the vehicle speed is from viTransfer to vjNumber of times of (F)iFor running vehicle speed from viSum of the number of transfers.
9. The method according to claim 8, wherein transition probability and number of transitions from the current running vehicle speed to the next running vehicle speed are calculated, and each state probability value is combined to generate a markov transition probability matrix P:
Figure FDA0002698971220000043
10. the method for predicting the traveling vehicle speed of a hybrid vehicle according to claim 1, wherein the predicted speed value in step S50 is obtained according to the following equation:
v(k)＝[(Uk-1)+rk]d
wherein v (k) is the vehicle running speed at the time k; u shapekThe driving speed state at the moment k is obtained; d is the speed state division length, and the value is 5; r iskThe uniformly distributed random numbers generated for time k matlab.
CN202011015662.8A 2020-09-24 2020-09-24 Hybrid electric vehicle running speed prediction method Pending CN112182962A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011015662.8A CN112182962A (en) 2020-09-24 2020-09-24 Hybrid electric vehicle running speed prediction method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011015662.8A CN112182962A (en) 2020-09-24 2020-09-24 Hybrid electric vehicle running speed prediction method

Publications (1)

Publication Number Publication Date
CN112182962A true CN112182962A (en) 2021-01-05

Family

ID=73956603

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011015662.8A Pending CN112182962A (en) 2020-09-24 2020-09-24 Hybrid electric vehicle running speed prediction method

Country Status (1)

Country Link
CN (1) CN112182962A (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112757922A (en) * 2021-01-25 2021-05-07 武汉理工大学 Hybrid power energy management method and system for vehicle fuel cell
CN112800549A (en) * 2021-03-04 2021-05-14 山东大学 Automobile road spectrum synthesis method and system based on horizontal speed and vertical speed
CN113076697A (en) * 2021-04-20 2021-07-06 潍柴动力股份有限公司 Typical driving condition construction method, related device and computer storage medium
CN113221975A (en) * 2021-04-26 2021-08-06 中国科学技术大学先进技术研究院 Working condition construction method based on improved Markov analysis method and storage medium
CN113657484A (en) * 2021-08-13 2021-11-16 济南大学 Method for dividing and identifying typical working conditions of cement grate cooler
CN114965898A (en) * 2022-06-01 2022-08-30 北京市生态环境监测中心 Remote online monitoring method for heavy vehicle nitrogen oxide and carbon dioxide emission

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107516107A (en) * 2017-08-01 2017-12-26 北京理工大学 A kind of driving cycle classification Forecasting Methodology of motor vehicle driven by mixed power
CN108198425A (en) * 2018-02-10 2018-06-22 长安大学 A kind of construction method of Electric Vehicles Driving Cycle
CN108921200A (en) * 2018-06-11 2018-11-30 百度在线网络技术（北京）有限公司 Method, apparatus, equipment and medium for classifying to Driving Scene data
CN111047085A (en) * 2019-12-06 2020-04-21 北京理工大学 Hybrid vehicle working condition prediction method based on meta-learning

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107516107A (en) * 2017-08-01 2017-12-26 北京理工大学 A kind of driving cycle classification Forecasting Methodology of motor vehicle driven by mixed power
CN108198425A (en) * 2018-02-10 2018-06-22 长安大学 A kind of construction method of Electric Vehicles Driving Cycle
CN108921200A (en) * 2018-06-11 2018-11-30 百度在线网络技术（北京）有限公司 Method, apparatus, equipment and medium for classifying to Driving Scene data
US20190378035A1 (en) * 2018-06-11 2019-12-12 Baidu Online Network Technology (Beijing) Co., Ltd. Method, apparatus, device and medium for classifying driving scenario data
CN111047085A (en) * 2019-12-06 2020-04-21 北京理工大学 Hybrid vehicle working condition prediction method based on meta-learning

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
曹磊等: ""基于马尔可夫链的汽车行驶工况预测"", 《内燃机与动力装置》, vol. 34, no. 03, pages 13 - 17 *
詹森: ""基于工况与驾驶风格识别的混合动力汽车能量管理策略研究"", 《中国博士学位论文全文数据库工程科技Ⅱ辑》, vol. 2017, no. 3, pages 035 - 17 *

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112757922A (en) * 2021-01-25 2021-05-07 武汉理工大学 Hybrid power energy management method and system for vehicle fuel cell
CN112757922B (en) * 2021-01-25 2022-05-03 武汉理工大学 Hybrid power energy management method and system for vehicle fuel cell
CN112800549A (en) * 2021-03-04 2021-05-14 山东大学 Automobile road spectrum synthesis method and system based on horizontal speed and vertical speed
CN112800549B (en) * 2021-03-04 2022-05-27 山东大学 Automobile road spectrum synthesis method and system based on horizontal speed and vertical speed
CN113076697A (en) * 2021-04-20 2021-07-06 潍柴动力股份有限公司 Typical driving condition construction method, related device and computer storage medium
CN113221975A (en) * 2021-04-26 2021-08-06 中国科学技术大学先进技术研究院 Working condition construction method based on improved Markov analysis method and storage medium
CN113657484A (en) * 2021-08-13 2021-11-16 济南大学 Method for dividing and identifying typical working conditions of cement grate cooler
CN113657484B (en) * 2021-08-13 2024-02-09 济南大学 Method for dividing and identifying typical working conditions of cement grate cooler
CN114965898A (en) * 2022-06-01 2022-08-30 北京市生态环境监测中心 Remote online monitoring method for heavy vehicle nitrogen oxide and carbon dioxide emission

Similar Documents

Publication Publication Date Title
CN112182962A (en) Hybrid electric vehicle running speed prediction method
CN107862864B (en) Driving condition intelligent prediction estimation method based on driving habits and traffic road conditions
CN108806021B (en) Electric vehicle target road section energy consumption prediction method based on physical model and road characteristic parameters
CN113401143B (en) Individualized self-adaptive trajectory prediction method based on driving style and intention
CN106427589A (en) Electric car driving range estimation method based on prediction of working condition and fuzzy energy consumption
CN111815948B (en) Vehicle running condition prediction method based on condition characteristics
CN111292534A (en) Traffic state estimation method based on clustering and deep sequence learning
CN110979342B (en) Working condition information acquisition method for vehicle global energy management control
CN110728772A (en) Construction method for typical running condition of tramcar
CN114912195B (en) Aerodynamic sequence optimization method for commercial vehicle
CN110155073A (en) Driving behavior mode identification method and system based on driver's preference
CN113642768A (en) Vehicle running energy consumption prediction method based on working condition reconstruction
CN115935672A (en) Fuel cell automobile energy consumption calculation method fusing working condition prediction information
CN113479187B (en) Layered different-step-length energy management method for plug-in hybrid electric vehicle
CN113553350A (en) Traffic flow partition model for similar evolution mode clustering and dynamic time zone partitioning
CN110097757B (en) Intersection group critical path identification method based on depth-first search
CN112035536A (en) Electric automobile energy consumption prediction method considering dynamic road network traffic flow
CN115691140B (en) Analysis and prediction method for space-time distribution of automobile charging demand
Wang et al. Dynamic traffic prediction based on traffic flow mining
CN114463978B (en) Data monitoring method based on track traffic information processing terminal
Shenghua et al. Road traffic congestion prediction based on random forest and DBSCAN combined model
Topić et al. Synthesis and Validation of Multidimensional Driving Cycles
Topić et al. Analysis of City Bus Driving Cycle Features for the Purpose of Multidimensional Driving Cycle Synthesis
Feng et al. Traffic Flow Prediction of Urban Intersection Based on Environmental Impact Factors and Markov Chains
CN117901724B (en) Control method, system and equipment for thermal management system of pure electric vehicle

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