CN113076697B - Typical driving condition construction method, related device and computer storage medium - Google Patents
Typical driving condition construction method, related device and computer storage medium Download PDFInfo
- Publication number
- CN113076697B CN113076697B CN202110423665.3A CN202110423665A CN113076697B CN 113076697 B CN113076697 B CN 113076697B CN 202110423665 A CN202110423665 A CN 202110423665A CN 113076697 B CN113076697 B CN 113076697B
- Authority
- CN
- China
- Prior art keywords
- target
- target data
- typical driving
- working condition
- data group
- 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.)
- Active
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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N7/00—Computing arrangements based on specific mathematical models
- G06N7/01—Probabilistic graphical models, e.g. probabilistic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/08—Probabilistic or stochastic CAD
Abstract
The application provides a typical driving condition construction method, a related device and a computer storage medium, wherein the typical driving condition construction method comprises the following steps: firstly, acquiring a target market data set; then, preprocessing the data in the target market data set to obtain at least one target data group; finally, inputting the target data group into a construction model of typical driving conditions aiming at each target data group to obtain the typical driving conditions corresponding to the target data groups; the construction model of the typical driving condition is obtained by training a Markov-Monte Carlo model based on a non-uniform grid through the combination of at least one model and a training sample market and the original conditions corresponding to various types of indexes of the training sample market. Therefore, the aim of accurately constructing the typical driving condition is fulfilled.
Description
Technical Field
The present application relates to the field of computer technologies, and in particular, to a method for constructing typical driving conditions, a related device, and a computer storage medium.
Background
In the prior art, a markov-monte carlo simulation method is generally adopted, and based on a large amount of data of commercial vehicles on the market, typical driving conditions with minimum granularity such as a certain vehicle type application are extracted and constructed, so that input is provided for the subsequent test of tail gas pollutant emission of vehicles, fuel consumption, development technology for evaluating new vehicle types and the like.
However, in the prior art, in the process of constructing a typical driving condition by using a markov-monte carlo simulation method, the method for dividing the grid is uniform division, and information in the same state is encoded into the same state in the uniform grid in the actual application process, so that a large error is introduced, and the accuracy of a subsequently constructed model is influenced.
Disclosure of Invention
In view of this, the present application provides a method, a related apparatus, and a computer storage medium for constructing typical driving conditions accurately.
The application provides a method for constructing a typical driving condition in a first aspect, which comprises the following steps:
acquiring a target market data set;
preprocessing data in the target market data set to obtain at least one target data group;
inputting the target data set into a construction model of typical driving conditions aiming at each target data set to obtain the typical driving conditions corresponding to the target data set; the construction model of the typical driving condition is obtained by training a Markov-Monte Carlo model based on a non-uniform grid through the combination of at least one model and a training sample market and the original condition corresponding to each class index of the training sample market.
Optionally, the training method for constructing the model of the typical driving condition includes:
acquiring a target machine type and a target typical working condition length of a market segment; the target machine type is a machine type which needs to be extracted by a user; the target typical working condition length is a typical working condition length which needs to be extracted by a user;
automatically generating an encryption grid in a speed-acceleration distribution dense interval according to the speed cumulative function distribution or the acceleration cumulative function distribution, and performing sparse division in a speed-acceleration distribution sparse interval to obtain a speed-acceleration combined two-dimensional interval;
carrying out one-dimensional coding on the speed-acceleration combined two-dimensional interval according to a preset coding and decoding mode to obtain a one-dimensional state space; wherein each code in the one-dimensional state space represents a state;
establishing a corresponding relation between the time, the vehicle speed and the acceleration sequence and the state to obtain a time-state one-dimensional sequence;
determining a state transition relation according to the time-state one-dimensional sequence to obtain a state transition matrix;
simulating the state transition of the state transition matrix by adopting a Monte Carlo simulation mode to obtain at least one state sequence data meeting the target typical working condition length;
for each state sequence data, decoding the state sequence data according to the preset coding and decoding mode to obtain decoded data;
counting the representative indexes of the big data of the target model and the market segment in the training sample market for the decoded data, and generating at least one index result set for representative evaluation of the original working condition data of the target model and the market segment;
calculating deviation values of each item in the index result set and an item corresponding to an original working condition respectively aiming at each representative evaluation index result set, calculating the variance of all the deviation values, summing all the deviation absolute values and the variance values, and taking the decoded data of the target working condition length corresponding to the minimum item as the input original machine type and typical driving working conditions of the market segments;
carrying out fuel consumption economy test on an original working condition and a typical driving working condition under the same condition to obtain a test result of the typical driving working condition and a test result of the original working condition;
judging whether the test result of the typical driving condition and the test result of the original condition meet a preset error or not;
and if the test result of the typical driving condition and the test result of the original condition do not meet the preset error, optimizing the Markov-Monte Carlo model based on the non-uniform grid until the test result of the typical driving condition and the test result of the original condition meet the preset error.
Optionally, the preprocessing the data in the target market data set to obtain at least one target data group includes:
dividing data in the target market data set according to a preset category to obtain at least one data group;
for each data group, cleaning data in the data group to obtain a first type of target data group;
the step of inputting the target data group into a model for constructing typical driving conditions for each target data group to obtain the typical driving conditions corresponding to the target data group includes:
and inputting the first-class target data groups into a construction model of typical driving conditions for each first-class target data group to obtain the typical driving conditions corresponding to the first-class target data groups.
Optionally, after the data in the data group is cleaned for each data group to obtain a first-type target data group, the method further includes:
carrying out short-stroke division on the first type target data group to obtain a first type target data group with a short stroke;
removing abnormal values in the first-class target data group with the short stroke and cleaning abnormal data in the first-class target data group with the short stroke to obtain a cleaned first-class target data group;
extracting the characteristics of the cleaned first-class target data groups, and clustering all the cleaned first-class target data groups according to the characteristics to obtain second-class target data groups;
the step of inputting the target data set into a model for constructing typical driving conditions for each target data set to obtain the typical driving conditions corresponding to the target data set includes:
and inputting the second type target data group into a construction model of typical driving conditions aiming at each second type target data group to obtain the typical driving conditions corresponding to the second type target data group.
In a second aspect, the present application provides a device for constructing a typical driving condition, including:
the system comprises a first acquisition unit, a second acquisition unit and a third acquisition unit, wherein the first acquisition unit is used for acquiring a target market data set;
the preprocessing unit is used for preprocessing the data in the target market data set to obtain at least one target data group;
the input unit is used for inputting the target data groups into a construction model of typical driving conditions aiming at each target data group to obtain the typical driving conditions corresponding to the target data groups; the construction model of the typical driving condition is obtained by training a Markov-Monte Carlo model based on a non-uniform grid through the combination of at least one model and a training sample market and the original condition corresponding to each class index of the training sample market.
Optionally, the training unit for constructing a model of the typical driving condition includes:
the second acquisition unit is used for acquiring the target model and the target typical working condition length of the market segment; the target machine type is a machine type which needs to be extracted by a user; the target typical working condition length is a typical working condition length which needs to be extracted by a user;
the generating unit is used for automatically generating an encryption grid in a speed-acceleration distribution dense interval according to the speed cumulative function distribution or the acceleration cumulative function distribution, and performing sparse division in a speed-acceleration distribution sparse interval to obtain a speed-acceleration combined two-dimensional interval;
the coding unit is used for carrying out one-dimensional coding on the speed-acceleration combined two-dimensional interval according to a preset coding and decoding mode to obtain a one-dimensional state space; wherein each code in the one-dimensional state space represents a state;
the establishing unit is used for establishing a corresponding relation between the time, the vehicle speed and the acceleration sequence and the state to obtain a time-state one-dimensional sequence;
the transition matrix calculation unit is used for determining a state transition relation according to the time-state sequence to obtain a state transition matrix;
the simulation unit is used for simulating the state transition of the state transition matrix in a Monte Carlo simulation mode to obtain at least one state sequence data meeting the target typical working condition length;
the decoding unit is used for decoding the state sequence data according to the preset coding and decoding mode aiming at each state sequence data to obtain decoded data;
the statistical unit is used for counting the representative indexes of the big data of the subdivided market in the target model and the training sample market for the decoded data and generating at least one index result set for representative evaluation of the original working condition data of the target model and the subdivided market;
the determining unit is used for respectively calculating deviation values of each item in the index result set and an item corresponding to the original working condition aiming at each representative evaluation index result set, calculating the variance of all the deviation values, summing all the deviation absolute values and the variance values, and taking the decoded data of the target working condition length corresponding to the minimum item as the input data of the original model and the typical driving working condition of the subdivided market;
the test unit is used for carrying out fuel consumption economy test on an original working condition and a typical driving working condition under the same condition to obtain a test result of the typical driving working condition and a test result of the original working condition;
the judging unit is used for judging whether the test result of the typical driving working condition and the test result of the original working condition meet a preset error or not;
and the optimizing unit is used for optimizing the Markov-Monte Carlo model based on the non-uniform grid until the test result of the typical driving condition and the test result of the original working condition meet the preset error.
Optionally, the preprocessing unit includes:
the first dividing unit is used for dividing the data in the target market data set according to a preset category to obtain at least one data group;
the first cleaning unit is used for cleaning data in the data groups aiming at each data group to obtain a first type of target data group;
wherein the input unit is configured to:
and inputting the first-class target data groups into a construction model of typical driving conditions for each first-class target data group to obtain the typical driving conditions corresponding to the first-class target data groups.
Optionally, the device for constructing the typical driving condition further includes:
the second dividing unit is used for carrying out short-stroke division on the first-class target data group to obtain a first-class target data group with a short stroke;
the second cleaning unit is used for eliminating abnormal values in the first-class target data group of the short stroke and cleaning abnormal data in the first-class target data group of the short stroke to obtain a cleaned first-class target data group;
the clustering unit is used for extracting the characteristics of the cleaned first-class target data groups and clustering all the cleaned first-class target data groups according to the characteristics to obtain second-class target data groups;
wherein the input unit is configured to:
and inputting the second type target data groups into a construction model of typical driving conditions to obtain the typical driving conditions corresponding to the second type target data groups.
A third aspect of the present application provides an electronic device, comprising:
one or more processors;
a storage device having one or more programs stored thereon;
the one or more programs, when executed by the one or more processors, cause the one or more processors to implement a method of constructing typical driving conditions as described in any one of the first aspects.
A fourth aspect of the present application provides a computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method of constructing a typical driving situation according to any one of the first aspect.
According to the scheme, the construction method of the typical driving condition, the related device and the computer storage medium provided by the application comprise the following steps: firstly, acquiring a target market data set; then, preprocessing the data in the target market data set to obtain at least one target data group; finally, inputting the target data group into a construction model of typical driving conditions aiming at each target data group to obtain the typical driving conditions corresponding to the target data groups; the construction model of the typical driving condition is obtained by training a Markov-Monte Carlo model based on a non-uniform grid through the combination of at least one model and a training sample market and the original conditions corresponding to various types of indexes of the training sample market. Therefore, the aim of accurately constructing the typical driving condition is fulfilled.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
FIG. 1 is a schematic diagram of a speed-acceleration phase plane for an FTP driving condition according to an embodiment of the present application;
fig. 2 is a schematic diagram of a state encoding method according to an embodiment of the present application;
FIG. 3 is a graph illustrating velocity and acceleration profiles of a uniform grid according to an embodiment of the present disclosure;
fig. 4 is a distribution of speed and acceleration drawn by a certain heavy truck based on data of 10 ten thousand points of the highway condition in the market according to the embodiment of the present application;
FIG. 5 is a detailed flowchart of a method for constructing a typical driving condition according to an embodiment of the present disclosure;
FIG. 6 is a flow chart of a method for pre-processing data in a target market data set according to an embodiment of the present application;
FIG. 7 is a flow diagram of a method for pre-processing data in a target market data set according to another embodiment of the present application;
FIG. 8 is a flow chart of a method for training a model of typical driving conditions according to another embodiment of the present application;
FIG. 9 is a graph illustrating velocity and acceleration profiles for a uniform grid according to another embodiment of the present application;
FIG. 10 is a schematic illustration of an exemplary driving condition construction apparatus according to another embodiment of the present application;
fig. 11 is a schematic diagram of an electronic device implementing a method for constructing typical driving conditions according to another embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be noted that the terms "first", "second", and the like in this application are only used for distinguishing different apparatuses, modules or units, and are not used for limiting the order or interdependence of functions performed by these apparatuses, modules or units, and the terms "include", or any other variants thereof are intended to cover non-exclusive inclusions, so that a process, a method, an article or a device including a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such a process, a method, an article or a device. Without further limitation, an element defined by the phrase "comprising a … …" does not exclude the presence of another identical element in a process, method, article, or apparatus that comprises the element.
The prior art method for constructing typical driving conditions generally includes: and acquiring speed-time sequence data, and performing operations such as abnormal value elimination, data cleaning and preprocessing according to the type of the vehicle (such as a passenger vehicle, a commercial vehicle, an electric vehicle and the like).
It will be appreciated that the speed-time series data for the driving regime of the vehicle may also be expressed as a relative planar motion of speed and acceleration states, for example for an FTP driving regime, which may be illustrated in fig. 1.
According to the Markov analysis of the driving condition of the automobile, the state transition relationship of the driving condition of the automobile can be expressed by using the speed-acceleration state (v, a). The state division of the driving condition of the automobile can be realized by state coding.
Specifically, for the speed-time sequence, setting the value interval of the speed as [ v1, v2], and the value interval of the acceleration as [ a1, a2], the speed-time sequence may be divided into m × n value intervals, each interval is a set, that is, a state, and all the states form the state interval of the driving condition, which may be specifically shown in a schematic diagram of a state encoding method in fig. 3; and after one-dimensional coding is carried out on the coded speed-acceleration combined two-dimensional interval, simulation of state transition is carried out by using Monte Carlo simulation, so that at least one state sequence data is obtained, and an interface is carried out according to the coding rule to obtain a decoded speed-time sequence. And carrying out statistics on idle speed ratio on the coded data to obtain a representative working condition which is closest to the original speed-time sequence, namely constructing to obtain a typical driving working condition.
Referring to fig. 2 and the above, for the velocity-acceleration joint two-dimensional interval, assuming that (v, a) is a state in the running process of the vehicle, the grid where the state (v, a) is located can be found through grid division. It should be noted that the key of the mesh partition method lies in the selection of the mesh size (step size), and if the partition is too large, the precision of the subsequent decoding step is reduced; if the partitioning is too small, it results in reduced algorithm efficiency. Therefore, boundary values of the velocity and the acceleration in the two-dimensional space are obtained from the velocity and the acceleration distributions of the reference condition, a certain step is defined, and the velocity and the acceleration distributions are gridded according to the boundary values, so that the velocity and the acceleration distributions of the uniform grid shown in fig. 3 are obtained.
Since the mesh designed by the mesh division method is a uniform mesh, as can be seen from the area a and the area b in fig. 3, if the uniform mesh method is adopted, the number of the speed and acceleration information contained in the same mesh of the area a and the area b is different under the condition of the same step length, even if the resolution ratio is set to be higher, the speed and acceleration information in the area a and the area b are still obviously different, but the uniform mesh encodes the information in the same state into the same state, so that a large error is introduced, and the accuracy of the model is affected.
Therefore, the even grid mode is adopted, the adaptability is good only in the working condition that the speed and acceleration are distributed evenly, but in the practical application scene, the speed and acceleration distribution changes along with the change of the model of the vehicle and the market segment, most of the model and the market segment of the vehicle have own unique working condition, the distribution condition of the speed and the acceleration, and the condition that the speed and the acceleration are distributed evenly in each speed interval is few. As shown in fig. 4, the distribution of speed and acceleration is plotted for a certain card based on the data of 10 ten thousand points of the highway condition in the market, and the inapplicability of the uniform grid method is particularly prominent in such a distribution of speed and acceleration. The main performance is as follows: under a uniform grid, c, d and e respectively represent a state, the difference of the points falling into c, d and e is large, the probability difference of the states corresponding to c, d and e is large, when a state transition matrix is simulated, the probability that d and e are simulated and reproduced is large due to high probability, but the probability that c state is simulated and reproduced is very small due to low occurrence probability, at this time, the simulation result is always transferred in high-speed areas such as d, e and the vicinity thereof, and the probability of two processes of accelerating ignition to high speed and decelerating from high speed to flameout is very small. Therefore, the 1800s or 3600s working condition extraction result is very difficult to cover the complete process of ignition-acceleration-high-speed stable operation-deceleration-flameout, in short, the uniform grid method is not suitable for the working condition extraction with the unique driving characteristics of the market subdivision.
In addition, when a large amount of data of millions or even tens of millions of levels or a situation with high real-time requirement is met, if the grid design is fine due to the fact that the step length is reduced for increasing the precision, the exponential level of the state is increased, the uniform grid method polls grids one by one no matter whether speed and acceleration data exist in the grids, the polling can increase the expenditure of a computer memory, the running time of a model is prolonged, and the efficiency of the algorithm is reduced.
Finally, although the idle ratio under typical conditions has a great influence on driving conditions, the idle ratio cannot be used as the only selection criterion; in actual driving, idle ratio, acceleration and deceleration ratio, average speed, maximum and minimum speed acceleration, speed acceleration distribution, particularly fuel consumption performance and the like are all important factors for evaluating a driving condition result.
Therefore, the method for constructing the typical driving condition provided by the embodiment of the application, as shown in fig. 5, specifically includes the following steps:
s501, acquiring a target market data set.
S502, preprocessing the data in the target market data set to obtain at least one target data group.
Optionally, in another embodiment of the present application, an implementation manner of step S502, as shown in fig. 6, includes:
s601, dividing data in the target market data set according to preset categories to obtain at least one data group.
The preset category includes, but is not limited to, an Idle (Idle) condition, an Acceleration (Acceleration) condition, a Deceleration (cancellation) condition, a constant speed (Cruise) condition, an Average speed (Average speed), an Average running speed (Average running speed), an Average Acceleration (Average Acceleration) of an Acceleration section, an Average Deceleration (Average Deceleration) of a Deceleration section, an Acceleration ratio (Acceleration ratio), a Deceleration ratio (cancellation ratio), a constant speed ratio (Cruise ratio), and an Idle ratio (Idle ratio).
Wherein the idle working condition represents that the absolute value of the acceleration a in the running process of the vehicle is less than 0.15m/s 2 I.e., -0.15m/s 2 <a<0.15m/s 2 And the running speed v of the vehicle<Working condition of 0.5 km/h; the acceleration condition represents that the acceleration a of the vehicle in the running process is more than or equal to 0.15m/s 2 The working condition of (1); the deceleration working condition represents that the acceleration a of the vehicle in the running process is less than or equal to-0.15 m/s 2 The working condition of (1); the constant speed working condition means that the absolute value of the acceleration a in the running process of the vehicle is less than 0.15m/s 2 I.e., -0.15m/s 2 <a<0.15m/s 2 And the running speed v of the vehicle is more than or equal to 0.5 km/h; the average speed is the average value of the speeds of all working conditions in the running process of the vehicle; the running average speed represents the vehicle driving mileage divided by the total duration of the non-idle operating point; the average acceleration of the acceleration section represents the average acceleration of all acceleration working condition points in the running process of the vehicle; the deceleration segment average deceleration represents the average deceleration of all deceleration operating points during the running process of the vehicle; the acceleration ratio represents the ratio of the acceleration working condition point to the total working condition duration in the running process of the vehicle; the deceleration proportion represents the proportion of deceleration working condition points in the running process of the vehicle to the total working condition duration; the constant speed proportion represents the proportion of the constant speed working condition point in the running process of the vehicle to the total working condition duration; the idle ratio represents the ratio of the idle operating point to the total operating time during which the vehicle is running.
S602, for each data group, cleaning the data in the data group to obtain a first-class target data group.
It should be noted that the data in the data set is cleaned, including but not limited to cleaning the driving cycle with idle ratio greater than 90%, and cleaning the driving cycle with non-zero vehicle speed at the beginning or end. Wherein the vehicle is electrically powered from one ignition on to off for one driving cycle.
Optionally, in another embodiment of the present application, an implementation manner of step S502, as shown in fig. 7, includes:
s701, dividing data in the target market data set according to preset categories to obtain at least one data group.
S702, for each data group, cleaning the data in the data group to obtain a first-class target data group.
It should be noted that, the execution manners of step S701 and step S702 may refer to step S601 and step S602, which are not described herein again.
And S703, carrying out short-stroke division on the first-class target data group to obtain a first-class target data group with a short stroke.
One driving cycle may include a short trip, and N is a positive integer. The short trip refers to one driving cycle from one vehicle speed of 0 to the next vehicle speed of 0.
S704, removing abnormal values in the first-class target data group with the short stroke, and cleaning abnormal data in the first-class target data group with the short stroke to obtain the cleaned first-class target data group.
For example: most of the speed data are 0-120 km/h, but if a speed value of 170km/h appears, the speed of 170km/h can be judged to be an abnormal value, and the abnormal value is removed.
S705, extracting the characteristics of the cleaned first-class target data groups, and clustering all the cleaned first-class target data groups according to the characteristics to obtain second-class target data groups.
The characteristics may be divided into, but not limited to, short stroke parameter characteristics, speed parameter characteristics, and acceleration parameter characteristics, which are not limited herein.
Short-trip parameter characteristics may include, but are not limited to, time, distance, idle time, acceleration time, deceleration time, uniform speed time, idle duty cycle, acceleration duty cycle, deceleration duty cycle, and the like; the speed parameter characteristics may include, but are not limited to, maximum speed, average speed, standard deviation of speed, etc.; the acceleration parameter may be, but is not limited to, maximum acceleration, average acceleration, number of accelerations, maximum deceleration, average deceleration, number of accelerations, acceleration standard deviation, etc., and is not limited herein.
It should be noted that the clustering manner may be, but is not limited to, kmeans clustering, som, gaussian clustering, hierarchical clustering, unsupervised clustering, artificially labeled according to business knowledge and then supervised clustering, and the like, and the manner is very diverse and mature, and is not limited herein.
S503, inputting the target data group into a construction model of typical driving conditions aiming at each target data group to obtain the typical driving conditions corresponding to the target data group.
The construction model of the typical driving condition is obtained by training a Markov-Monte Carlo model based on a non-uniform grid by at least one model, a training sample subdivision market and original conditions corresponding to various types of indexes of the training sample subdivision market.
It should be noted that the target data set may be the first type target data set in the foregoing embodiment, or may be the second type target data set in the foregoing embodiment, which is not limited herein. However, it is understood that the second type of target data set is more finely divided than the first type of target data set, and thus, the corresponding typical driving condition is more accurately constructed.
Optionally, in another embodiment of the present application, an implementation manner of the training method for constructing a model of a typical driving condition, as shown in fig. 8, specifically includes the following steps:
s801, obtaining a target model and a target typical working condition length of a market segment.
The target machine type is a machine type which needs to be extracted by a user; the target typical working condition length is the typical working condition length which needs to be extracted by a user.
S802, according to the cumulative function distribution of the speed or the cumulative function distribution of the acceleration, an encryption grid is automatically generated in a speed-acceleration distribution dense interval, and sparse division is carried out in a speed-acceleration distribution sparse interval, so that a speed-acceleration combined two-dimensional interval is obtained.
Referring to fig. 9, in the velocity-acceleration combined two-dimensional interval generated according to the method, it can be seen that the grid densities of the two regions a and b are not consistent, and the grid density corresponding to a is obviously smaller than that corresponding to b, so that the probability of the state corresponding to the sparse position is improved compared with that of the uniform grid, the probability of the state corresponding to the dense position is reduced compared with that of the uniform grid, the probability difference between different states is reduced to a certain extent, and further, when the state of the dense region is transferred to the state having a transfer relationship with the sparse region, the probability of transferring to the sparse region is increased. This solves the problem of the prior art that the complete process of ignition-acceleration-high speed steady operation-deceleration-flameout cannot be covered when the typical condition extraction with the speed and acceleration profile shown in fig. 4 is encountered.
And S803, carrying out one-dimensional coding on the speed-acceleration combined two-dimensional interval according to a preset coding and decoding mode to obtain a one-dimensional state space.
Where each code in the one-dimensional state space represents a state.
S804, establishing a corresponding relation between the time, the vehicle speed and the acceleration sequence and the state to obtain a time-state one-dimensional sequence.
S805, determining a state transition relation according to the time-state one-dimensional sequence to obtain a state transition matrix.
S806, simulating state transition of the state transition matrix in a Monte Carlo simulation mode to obtain at least one state sequence data meeting the target typical working condition length.
And S807, decoding the state sequence data according to a preset coding and decoding mode aiming at each state sequence data to obtain decoded data.
And S808, counting the representative indexes of the big data of the market segment in the target model and the training sample market for the decoded data, and generating at least one index result set for representative evaluation of the original working condition data of the target model and the market segment.
And S809, aiming at each representative evaluation index result set, respectively calculating deviation values of each item in the index result set and an item corresponding to the original working condition, calculating the variance of all the deviation values, summing all the deviation absolute values and the variance values, and taking the decoded data of the target working condition length corresponding to the minimum item as the input original machine type and typical driving working conditions of the market segments.
And S810, carrying out an oil consumption economy test on the original working condition and the typical driving working condition under the same condition to obtain a test result of the typical driving working condition and a test result of the original working condition.
S811, judging whether the test result of the typical driving condition and the test result of the original condition meet a preset error.
The preset error is set by a technician or the like according to the actual situation, and may be changed, and is not limited herein.
Specifically, if it is determined that the test result of the typical driving condition and the test result of the original driving condition do not satisfy the preset error, the step S812 is executed; and if the test result of the typical driving condition and the test result of the original condition meet the preset error S813.
And S812, optimizing the Markov-Monte Carlo model based on the non-uniform grids.
S813, taking the Markov-Monte Carlo model based on the non-uniform grid as a construction model of a typical driving condition.
According to the scheme, the construction method of the typical driving condition comprises the following steps: firstly, acquiring a target market data set; then, preprocessing the data in the target market data set to obtain at least one target data group; finally, inputting the target data group into a construction model of typical driving conditions for each target data group to obtain the typical driving conditions corresponding to the target data groups; the construction model of the typical driving condition is obtained by training a Markov-Monte Carlo model based on a non-uniform grid by at least one model, a training sample subdivision market and original conditions corresponding to various types of indexes of the training sample subdivision market. Therefore, the aim of accurately constructing the typical driving condition is fulfilled.
Another embodiment of the present application provides a device for constructing typical driving conditions, as shown in fig. 10, specifically including:
a first obtaining unit 1001 is configured to obtain a target market data set.
The preprocessing unit 1002 is configured to preprocess data in the target market data set to obtain at least one target data group.
And the input unit 1003 is configured to input the target data set into the model for constructing the typical driving condition for each target data set, so as to obtain the typical driving condition corresponding to the target data set.
The construction model of the typical driving condition is obtained by training a Markov-Monte Carlo model based on a non-uniform grid by at least one model, a training sample subdivision market and original conditions corresponding to various types of indexes of the training sample subdivision market.
For a specific working process of the unit disclosed in the above embodiment of the present application, reference may be made to the content of the corresponding method embodiment, as shown in fig. 5, which is not described herein again.
Optionally, in another embodiment of the present application, an implementation manner of the preprocessing unit 1002 includes:
the first dividing unit is used for dividing the data in the target market data set according to a preset category to obtain at least one data group.
And the first cleaning unit is used for cleaning the data in the data groups aiming at each data group to obtain a first-class target data group.
Wherein, the input unit 1003 is configured to:
and inputting the first type of target data group into a construction model of the typical driving condition aiming at each first type of target data group to obtain the typical driving condition corresponding to the first type of target data group.
For a specific working process of the unit disclosed in the above embodiment of the present application, reference may be made to the content of the corresponding method embodiment, as shown in fig. 6, which is not described herein again.
Optionally, in another embodiment of the present application, an implementation manner of the preprocessing unit 1002 further includes:
and the second dividing unit is used for carrying out short-stroke division on the first-class target data group to obtain a short-stroke first-class target data group.
And the second cleaning unit is used for eliminating abnormal values in the first-class target data group with the short stroke and cleaning abnormal data in the first-class target data group with the short stroke to obtain the cleaned first-class target data group.
And the clustering unit is used for extracting the characteristics of the cleaned first-class target data groups and clustering all the cleaned first-class target data groups according to the characteristics to obtain second-class target data groups.
Wherein, the input unit 1003 is configured to:
and inputting the second type target data group into a construction model of the typical driving working conditions aiming at each second type target data group to obtain the typical driving working conditions corresponding to the second type target data groups.
For a specific working process of the unit disclosed in the above embodiment of the present application, reference may be made to the content of the corresponding method embodiment, as shown in fig. 7, which is not described herein again.
Optionally, in another embodiment of the present application, an implementation of the training unit for modeling of typical driving conditions includes:
and the second acquisition unit is used for acquiring the target model and the target typical working condition length of the market segment.
The target machine type is a machine type which needs to be extracted by a user; the target typical working condition length is the typical working condition length which needs to be extracted by a user.
And the generating unit is used for automatically generating an encryption grid in a speed-acceleration distribution dense interval according to the speed cumulative function distribution or the acceleration cumulative function distribution, and performing sparse division in a speed-acceleration distribution sparse interval to obtain a speed-acceleration combined two-dimensional interval.
And the coding unit is used for carrying out one-dimensional coding on the speed-acceleration combined two-dimensional interval according to a preset coding and decoding mode to obtain a one-dimensional state space.
Where each code in the one-dimensional state space represents a state.
And the establishing unit is used for establishing a corresponding relation between the time, the vehicle speed and the acceleration sequence and the state to obtain a time-state one-dimensional sequence.
And the transition matrix calculation unit is used for determining the state transition relation according to the time-state sequence to obtain the state transition matrix.
And the simulation unit is used for simulating the state transition of the state transition matrix in a Monte Carlo simulation mode to obtain at least one state sequence data meeting the target typical working condition length.
And the decoding unit is used for decoding the state sequence data according to a preset coding and decoding mode aiming at each state sequence data to obtain decoded data.
And the statistical unit is used for counting the representative indexes of the big data of the market segment in the target machine type and the training sample market for the decoded data and generating at least one index result set for representative evaluation of the original working condition data of the target machine type and the market segment.
And the determining unit is used for respectively calculating deviation values of each item in the index result set and the item corresponding to the original working condition aiming at each representative evaluation index result set, calculating the variance of all the deviation values, summing all the deviation absolute values and the variance values, and taking the decoded data of the target working condition length corresponding to the minimum item as the input original model and the typical driving condition of the market segmentation.
And the test unit is used for carrying out fuel consumption economy test on the original working condition and the typical driving working condition under the same condition to obtain a test result of the typical driving working condition and a test result of the original working condition.
And the judging unit is used for judging whether the testing result of the typical driving working condition and the testing result of the original working condition meet the preset error or not.
And the optimization unit is used for optimizing the Markov-Monte Carlo model based on the non-uniform grid until the test result of the typical driving condition and the test result of the original working condition meet the preset error if the judgment unit judges that the preset error does not meet the preset error between the test result of the typical driving condition and the test result of the original working condition.
For a specific working process of the unit disclosed in the above embodiment of the present application, reference may be made to the content of the corresponding method embodiment, as shown in fig. 8, which is not described herein again.
According to the scheme, the construction device for the typical driving condition comprises the following components: first, the acquisition unit 1001 acquires a target market data set; then, the preprocessing unit 1002 preprocesses data in the target market data set to obtain at least one target data group; finally, the input unit 1003 inputs the target data set into a model for constructing typical driving conditions for each target data set, so as to obtain the typical driving conditions corresponding to the target data set; the construction model of the typical driving condition is obtained by training a Markov-Monte Carlo model based on a non-uniform grid through at least one model, a training sample market and original conditions corresponding to various types of indexes of the training sample market. Therefore, the aim of accurately constructing the typical driving condition is fulfilled.
Another embodiment of the present application provides an electronic device, as shown in fig. 11, including:
one or more processors 1101.
The one or more programs, when executed by the one or more processors 1101, cause the one or more processors 1101 to implement a method of constructing typical driving conditions as described in any one of the above embodiments.
Another embodiment of the present application provides a computer storage medium, on which a computer program is stored, wherein the computer program, when executed by a processor, implements the method for constructing a typical driving condition as described in any one of the above embodiments.
In the above embodiments disclosed in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The apparatus and method embodiments described above are illustrative only, as the flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
In addition, functional modules in the embodiments of the present disclosure may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part. The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present disclosure may be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a live broadcast device, or a network device) to execute all or part of the steps of the method according to the embodiments of the present disclosure. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
Those skilled in the art can make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (8)
1. A method of constructing a typical driving condition, comprising:
acquiring a target market data set;
preprocessing data in the target market data set to obtain at least one target data group;
inputting the target data set into a construction model of typical driving conditions aiming at each target data set to obtain the typical driving conditions corresponding to the target data set; the construction model of the typical driving condition is obtained by training a Markov-Monte Carlo model based on a non-uniform grid through the combination of at least one model and a training sample market and the original condition corresponding to each class index of the training sample market;
the training method for the construction model of the typical driving condition comprises the following steps:
acquiring a target machine type and a target typical working condition length of a market segment; the target machine type is a machine type which needs to be extracted by a user; the target typical working condition length is a typical working condition length which needs to be extracted by a user;
automatically generating an encryption grid in a speed-acceleration distribution dense interval according to the speed cumulative function distribution or the acceleration cumulative function distribution, and performing sparse division in a speed-acceleration distribution sparse interval to obtain a speed-acceleration combined two-dimensional interval;
carrying out one-dimensional coding on the speed-acceleration combined two-dimensional interval according to a preset coding and decoding mode to obtain a one-dimensional state space; wherein each code in said one-dimensional state space represents a state;
establishing a corresponding relation between the time, the vehicle speed and the acceleration sequence and the state to obtain a time-state one-dimensional sequence;
determining a state transition relation according to the time-state one-dimensional sequence to obtain a state transition matrix;
simulating the state transition of the state transition matrix by adopting a Monte Carlo simulation mode to obtain at least one state sequence data meeting the target typical working condition length;
for each state sequence data, decoding the state sequence data according to the preset coding and decoding mode to obtain decoded data;
counting the representative indexes of the target machine type and the market segmentation big data of the decoded data to generate at least one index result set for representative evaluation of the original working condition data of the target machine type and the market segmentation;
calculating deviation values of each item in the index result set and an item corresponding to an original working condition respectively aiming at each representative evaluation index result set, calculating the variance of all the deviation values, summing all the deviation absolute values and the variance values, and taking the decoded data of the target working condition length corresponding to the minimum item as the input original machine type and typical driving working conditions of the market segments;
carrying out fuel consumption economy test on an original working condition and a typical driving working condition under the same condition to obtain a test result of the typical driving working condition and a test result of the original working condition;
judging whether the test result of the typical driving condition and the test result of the original condition meet a preset error or not;
and if the test result of the typical driving condition and the test result of the original condition do not meet the preset error, optimizing the Markov-Monte Carlo model based on the non-uniform grid until the test result of the typical driving condition and the test result of the original condition meet the preset error.
2. The construction method according to claim 1, wherein the preprocessing the data in the target market data set to obtain at least one target data group comprises:
dividing data in the target market data set according to a preset category to obtain at least one data group;
for each data group, cleaning data in the data group to obtain a first type of target data group;
the step of inputting the target data group into a model for constructing typical driving conditions for each target data group to obtain the typical driving conditions corresponding to the target data group includes:
and inputting the first-class target data groups into a construction model of typical driving conditions for each first-class target data group to obtain the typical driving conditions corresponding to the first-class target data groups.
3. The building method according to claim 2, wherein after the cleaning of the data in the data group for each data group to obtain the first type target data group, the method further comprises:
carrying out short-stroke division on the first type target data group to obtain a first type target data group with a short stroke;
removing abnormal values in the first-class target data group with the short stroke and cleaning abnormal data in the first-class target data group with the short stroke to obtain a cleaned first-class target data group;
extracting the characteristics of the cleaned first-class target data groups, and clustering all the cleaned first-class target data groups according to the characteristics to obtain second-class target data groups;
the step of inputting the target data set into a model for constructing typical driving conditions for each target data set to obtain the typical driving conditions corresponding to the target data set includes:
and inputting the second type target data groups into a construction model of typical driving conditions to obtain the typical driving conditions corresponding to the second type target data groups.
4. A device for constructing typical driving conditions, comprising:
the system comprises a first acquisition unit, a second acquisition unit and a third acquisition unit, wherein the first acquisition unit is used for acquiring a target market data set;
the preprocessing unit is used for preprocessing the data in the target market data set to obtain at least one target data group;
the input unit is used for inputting the target data set into a construction model of typical driving conditions aiming at each target data set to obtain the typical driving conditions corresponding to the target data set; the construction model of the typical driving condition is obtained by training a Markov-Monte Carlo model based on a non-uniform grid through the combination of at least one model and a training sample market and the original condition corresponding to each class index of the training sample market;
wherein, the training unit of the construction model of the typical driving condition comprises:
the second acquisition unit is used for acquiring the target model and the target typical working condition length of the market segment; the target machine type is a machine type which needs to be extracted by a user; the target typical working condition length is a typical working condition length which needs to be extracted by a user;
the generating unit is used for automatically generating an encryption grid in a speed-acceleration distribution dense interval according to the speed cumulative function distribution or the acceleration cumulative function distribution, and performing sparse division in a speed-acceleration distribution sparse interval to obtain a speed-acceleration combined two-dimensional interval;
the coding unit is used for carrying out one-dimensional coding on the speed-acceleration combined two-dimensional interval according to a preset coding and decoding mode to obtain a one-dimensional state space; wherein each code in the one-dimensional state space represents a state;
the establishing unit is used for establishing a corresponding relation between the time, the vehicle speed and the acceleration sequence and the state to obtain a time-state one-dimensional sequence;
the transition matrix calculation unit is used for determining a state transition relation according to the time-state sequence to obtain a state transition matrix;
the simulation unit is used for simulating the state transition of the state transition matrix in a Monte Carlo simulation mode to obtain at least one state sequence data meeting the target typical working condition length;
the decoding unit is used for decoding the state sequence data according to the preset coding and decoding mode aiming at each state sequence data to obtain decoded data;
the statistical unit is used for counting the representative indexes of the big data of the subdivided market in the target model and the training sample market for the decoded data and generating at least one index result set for representative evaluation of the original working condition data of the target model and the subdivided market;
the determining unit is used for respectively calculating deviation values of each item in the index result set and an item corresponding to the original working condition aiming at each representative evaluation index result set, calculating the variance of all the deviation values, summing all the deviation absolute values and the variance values, and taking the decoded data of the target working condition length corresponding to the minimum item as the input typical driving working conditions of the original machine type and the subdivided market;
the test unit is used for carrying out fuel consumption economy test on an original working condition and a typical driving working condition under the same condition to obtain a test result of the typical driving working condition and a test result of the original working condition;
the judging unit is used for judging whether the test result of the typical driving working condition and the test result of the original working condition meet a preset error or not;
and the optimizing unit is used for optimizing the Markov-Monte Carlo model based on the non-uniform grid until the test result of the typical driving condition and the test result of the original working condition meet the preset error if the judging unit judges that the test result of the typical driving condition and the test result of the original working condition do not meet the preset error.
5. The building apparatus according to claim 4, wherein the preprocessing unit comprises:
the first dividing unit is used for dividing data in the target market data set according to a preset category to obtain at least one data group;
the first cleaning unit is used for cleaning data in the data groups aiming at each data group to obtain a first type of target data group;
wherein the input unit is configured to:
and inputting the first-class target data groups into a construction model of typical driving conditions for each first-class target data group to obtain the typical driving conditions corresponding to the first-class target data groups.
6. The build device of claim 5, further comprising:
the second dividing unit is used for carrying out short-stroke division on the first-class target data group to obtain a first-class target data group with a short stroke;
the second cleaning unit is used for eliminating abnormal values in the first-class target data group of the short stroke and cleaning abnormal data in the first-class target data group of the short stroke to obtain a cleaned first-class target data group;
the clustering unit is used for extracting the characteristics of the cleaned first-class target data groups and clustering all the cleaned first-class target data groups according to the characteristics to obtain second-class target data groups;
wherein the input unit is configured to:
and inputting the second type target data groups into a construction model of typical driving conditions to obtain the typical driving conditions corresponding to the second type target data groups.
7. An electronic device, comprising:
one or more processors;
a storage device having one or more programs stored thereon;
the one or more programs, when executed by the one or more processors, cause the one or more processors to implement a method of constructing typical driving conditions as recited in any one of claims 1 to 3.
8. A computer storage medium, characterized in that a computer program is stored thereon, wherein the computer program, when being executed by a processor, implements the method of constructing typical driving situations according to any one of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110423665.3A CN113076697B (en) | 2021-04-20 | 2021-04-20 | Typical driving condition construction method, related device and computer storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110423665.3A CN113076697B (en) | 2021-04-20 | 2021-04-20 | Typical driving condition construction method, related device and computer storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113076697A CN113076697A (en) | 2021-07-06 |
| CN113076697B true CN113076697B (en) | 2022-11-29 |
Family
ID=76618335
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110423665.3A Active CN113076697B (en) | 2021-04-20 | 2021-04-20 | Typical driving condition construction method, related device and computer storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113076697B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113297685B (en) * | 2021-07-27 | 2021-11-02 | 中汽研（天津）汽车工程研究院有限公司 | Vehicle operation condition mode identification method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102881060A (en) * | 2012-10-17 | 2013-01-16 | 潍柴动力股份有限公司 | Method and system for acquiring vehicle typical working condition |
| CN106777625A (en) * | 2016-12-02 | 2017-05-31 | 潍柴动力股份有限公司 | Evaluation method and system that driving behavior influences on oil consumption |
| CN111775925A (en) * | 2020-06-09 | 2020-10-16 | 同济大学 | Working mode decision method and device for power split hybrid electric vehicle |
Family Cites Families (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107464007A (en) * | 2016-06-02 | 2017-12-12 | 南京理工大学 | Continuous time Probabilistic Load Flow Forecasting Methodology based on Markov theory and pro rate principle |
| CN107274086A (en) * | 2017-06-09 | 2017-10-20 | 东莞市毅豪电子科技有限公司 | A kind of gridding governance information approach based on hidden Markov model |
| CN108128309A (en) * | 2017-09-01 | 2018-06-08 | 特百佳动力科技有限公司 | A kind of method that vehicle working condition is predicted in real time |
| CN108198425A (en) * | 2018-02-10 | 2018-06-22 | 长安大学 | A kind of construction method of Electric Vehicles Driving Cycle |
| CN108596208B (en) * | 2018-03-21 | 2020-12-29 | 上海交通大学 | Vehicle driving cycle construction method for full-working-condition road |
| US11126199B2 (en) * | 2018-04-16 | 2021-09-21 | Baidu Usa Llc | Learning based speed planner for autonomous driving vehicles |
| CN108806021B (en) * | 2018-06-12 | 2020-11-06 | 重庆大学 | Electric vehicle target road section energy consumption prediction method based on physical model and road characteristic parameters |
| CN108802845B (en) * | 2018-07-16 | 2019-06-28 | 燕山大学 | A kind of indoor occupant occupation rate estimation method based on infrared sensor array |
| CN109409672A (en) * | 2018-09-25 | 2019-03-01 | 深圳市元征科技股份有限公司 | A kind of auto repair technician classifies grading modeling method and device |
| CN109409008B (en) * | 2018-11-23 | 2022-06-17 | 哈尔滨理工大学 | Construction method for typical speed-time running condition of urban rail vehicle line |
| CN110728772A (en) * | 2019-10-15 | 2020-01-24 | 西南交通大学 | Construction method for typical running condition of tramcar |
| CN111008505B (en) * | 2019-11-18 | 2023-05-23 | 西华大学 | Urban ramp driving condition construction method and application |
| CN111126819B (en) * | 2019-12-18 | 2022-08-16 | 南京理工大学 | Intelligent analysis method for urban driving condition |
| CN111666715B (en) * | 2020-06-05 | 2023-04-25 | 北京理工大学 | Electric automobile energy consumption prediction method and system |
| CN112182962A (en) * | 2020-09-24 | 2021-01-05 | 青岛大学 | Hybrid electric vehicle running speed prediction method |
-
2021
- 2021-04-20 CN CN202110423665.3A patent/CN113076697B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102881060A (en) * | 2012-10-17 | 2013-01-16 | 潍柴动力股份有限公司 | Method and system for acquiring vehicle typical working condition |
| CN106777625A (en) * | 2016-12-02 | 2017-05-31 | 潍柴动力股份有限公司 | Evaluation method and system that driving behavior influences on oil consumption |
| CN111775925A (en) * | 2020-06-09 | 2020-10-16 | 同济大学 | Working mode decision method and device for power split hybrid electric vehicle |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113076697A (en) | 2021-07-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chen et al. | BLOCK-DBSCAN: Fast clustering for large scale data | |
| CN113076697B (en) | Typical driving condition construction method, related device and computer storage medium | |
| CN109410588B (en) | Traffic accident evolution analysis method based on traffic big data | |
| CN110827326B (en) | Method, device, equipment and storage medium for generating simulation man-vehicle conflict scene model | |
| CN113221975A (en) | Working condition construction method based on improved Markov analysis method and storage medium | |
| Le Rhun et al. | A stochastic data-based traffic model applied to vehicles energy consumption estimation | |
| CN114971156A (en) | Driving style analysis method based on big data dangerous scene | |
| Zhang et al. | Spatiotemporal interaction pattern recognition and risk evolution analysis during lane changes | |
| Naing et al. | Dynamic data-driven microscopic traffic simulation using jointly trained physics-guided long short-term memory | |
| Markley et al. | Joint parsimonious modeling and model order selection for multivariate Gaussian mixtures | |
| CN116438434A (en) | Computer-aided method and apparatus for predicting vehicle speed based on probability | |
| Barr et al. | A real-time data-driven method for battery health prognostics in electric vehicle use | |
| Xu et al. | Traversing the local polytopes of relu neural networks | |
| CN115563020B (en) | Method and system for generating danger test scene, electronic device and storage medium | |
| CN114842645B (en) | Road network speed situation rule extraction method based on k-means | |
| CN115080450B (en) | Automatic driving test data generation method and system, electronic device and storage medium | |
| Guo et al. | Application of PCA-K-means++ combination model to construction of light vehicle driving conditions in intelligent traffic | |
| Zhang et al. | Improved driver clustering framework by considering the variability of driving behaviors across traffic operation conditions | |
| CN110148298B (en) | Private car regular travel behavior discovery method based on motor vehicle electronic identification data | |
| CN115761599A (en) | Video anomaly detection method and system | |
| CN115761551A (en) | Automatic driving simulation scene generation method, device, equipment and medium | |
| CN105043786B (en) | A kind of vehicle operational mode Markov chain design method containing road grade | |
| Abdelatif et al. | Vehicular-cloud simulation framework for predicting traffic flow data | |
| CN111538752A (en) | Vehicle use identification method and system based on Internet of vehicles and storage medium | |
| CN116680517B (en) | Method and device for determining failure probability in automatic driving simulation test |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |