CN113689081A - Automatic driving road test data quality determination method based on normal cloud model - Google Patents

Automatic driving road test data quality determination method based on normal cloud model Download PDF

Info

Publication number
CN113689081A
CN113689081A CN202110864149.4A CN202110864149A CN113689081A CN 113689081 A CN113689081 A CN 113689081A CN 202110864149 A CN202110864149 A CN 202110864149A CN 113689081 A CN113689081 A CN 113689081A
Authority
CN
China
Prior art keywords
cloud
test data
data
quality
evaluation index
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110864149.4A
Other languages
Chinese (zh)
Other versions
CN113689081B (en
Inventor
涂辉招
遇泽洋
李�浩
鹿畅
孙立军
郑叶明
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.)
Tongji University
Original Assignee
Tongji 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 Tongji University filed Critical Tongji University
Priority to CN202110864149.4A priority Critical patent/CN113689081B/en
Publication of CN113689081A publication Critical patent/CN113689081A/en
Application granted granted Critical
Publication of CN113689081B publication Critical patent/CN113689081B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06395—Quality analysis or management

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Development Economics (AREA)
  • Educational Administration (AREA)
  • Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Strategic Management (AREA)
  • Theoretical Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Game Theory and Decision Science (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Computational Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Tourism & Hospitality (AREA)
  • Pure & Applied Mathematics (AREA)
  • Algebra (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Traffic Control Systems (AREA)

Abstract

The invention relates to a method and a system for determining the quality of automatic driving road test data based on a normal cloud model, wherein the method comprises the following steps: determining the weight of each evaluation index for determining the quality of the test data of the automatic driving road by adopting an analytic hierarchy process; calculating a cloud model characteristic parameter value of each evaluation index of the test data by adopting a reverse cloud emitter algorithm; calculating a comprehensive cloud characteristic parameter value of the test data according to the weight of each evaluation index and the cloud model characteristic parameter value of each evaluation index of the test data; and determining the quality of the test data based on the normal cloud model according to the comprehensive cloud characteristic parameter value of the test data. The invention realizes the determination of the quality of the test data of the automatic driving road.

Description

Automatic driving road test data quality determination method based on normal cloud model
Technical Field
The invention relates to the technical field of automatic driving tests, in particular to a method for determining the quality of test data of an automatic driving road based on a normal cloud model.
Background
The automatic driving technology enters a high-speed development stage, and mass data can be generated in the automatic driving road testing process. The massive road test data contains high potential value. However, the automatic driving road test data has the characteristics of multiple test main bodies, multiple test links and multiple acquisition devices, so that the quality of the data is to be improved, and the release of data value is limited. Therefore, in order to improve the quality of the test data of the automatic driving road, give full play to the data value and enable industrial development, a set of scientific and reasonable methods is needed to evaluate the quality of the test data of the automatic driving road.
Disclosure of Invention
The invention aims to provide an automatic driving road test data quality determination method based on a normal cloud model so as to determine the automatic driving road test data quality.
In order to achieve the purpose, the invention provides the following scheme:
the invention provides an automatic driving road test data quality determination method based on a normal cloud model, which comprises the following steps:
determining the weight of each evaluation index for determining the quality of the test data of the automatic driving road by adopting an analytic hierarchy process; the evaluation indexes comprise data correctness and qualification, data integrity, data uniqueness, data consistency and data effectiveness;
calculating a cloud model characteristic parameter value of each evaluation index of the test data by adopting a reverse cloud emitter algorithm;
calculating a comprehensive cloud characteristic parameter value of the test data according to the weight of each evaluation index and the cloud model characteristic parameter value of each evaluation index of the test data;
and determining the quality of the test data based on the normal cloud model according to the comprehensive cloud characteristic parameter value of the test data.
Optionally, the determining, by using an analytic hierarchy process, a weight of each evaluation index for determining the quality of the test data of the automated driving road specifically includes:
constructing an index evaluation system for determining the quality of the test data of the automatic driving road, and determining an initial weight judgment matrix of the index evaluation system;
using formulas
Figure BDA0003186747260000021
Calculating the consistency proportion of the weight judgment matrix;
wherein CR is a consistency ratio, CI is a consistency index,
Figure BDA0003186747260000022
λmaxjudging the maximum characteristic value of the matrix for the weight, wherein r is the dimension of the matrix for the weight judgment, and RI is a random consistency index;
judging whether the consistency ratio is less than or equal to 0.01 or not, and obtaining a judgment result;
if the judgment result shows no, the weight judgment matrix is updated, and the step of 'utilizing the formula' is returned
Figure BDA0003186747260000023
Calculating a consistency ratio of the weight judgment matrix;
if the judgment result shows yes, judging a matrix according to the weight and utilizing a formula
Figure BDA0003186747260000024
Figure BDA0003186747260000025
Calculating the weight of each evaluation index;
wherein, ω isiWeight representing the ith evaluation index, aijAnd (3) the (i, j) th element in the weight judgment matrix represents the risk importance degree of the i-th evaluation index relative to the j-th evaluation index.
Optionally, a reverse cloud emitter algorithm is adopted, and a cloud model characteristic parameter value of each evaluation index of the test data is calculated, which specifically includes:
respectively taking the test data of each vehicle as a subdata set to obtain K subdata sets;
respectively calculating the evaluation result of each subdata set under each evaluation index;
according to the evaluation result of each subdata set under each evaluation index, a reverse cloud emitter algorithm is adopted, and the cloud model characteristic parameter value of each evaluation index of the test data is calculated by using the following formula;
Figure BDA0003186747260000026
Figure BDA0003186747260000027
Figure BDA0003186747260000031
wherein E isxi(ii) an expectation of a cloud model that is the ith evaluation index; eniEntropy of the cloud model being the ith evaluation index; heiIs the hyper-entropy of the cloud model for the ith evaluation index; c. CkiIs the evaluation result of the kth sub-data set under the ith evaluation index.
Optionally, the calculating the evaluation result of each sub data set under each evaluation index respectively specifically includes:
using formulas
Figure BDA0003186747260000032
Calculating the data correctness of each subdata set;
wherein, c1For data correctness of the subdata set, Fc1_1(n)、Fc1_2(n)、Fc1_3(n) and Fc1_4(N) the correctness of the instantaneous vehicle speed, the correctness of the instantaneous acceleration, the conformity of the acceleration and the correctness of the driving mode which are recorded in the nth record are respectively, and N represents the number of records in the sub data set;
Figure BDA0003186747260000033
Figure BDA0003186747260000034
Figure BDA0003186747260000035
Figure BDA0003186747260000036
distance(latn，lonn，latn+1，lonn+1) Calculating a function for distance (lat)n，lonn) Latitude and longitude coordinates for the nth entry, (lat)n+1，lonn+1) The longitude and latitude coordinates of the n +1 record are recorded; v. ofnAnd vn+1The speed of the nth record and the speed of the (n + 1) th record respectively; a isnAcceleration recorded for the nth bar; t is tnAnd tn+1Respectively time of the nth record and the (n + 1) th record;
using formulas
Figure BDA0003186747260000041
Calculating the data integrity of each sub data set;
wherein, c2For data integrity of the subdata sets, Fc2_1(n) and Fc2_2(n) data value missing and data time continuity of the nth record respectively;
Figure BDA0003186747260000042
Figure BDA0003186747260000043
l is a time interval threshold;
using formulas
Figure BDA0003186747260000044
Calculating the data uniqueness of each subdata set;
wherein, c3For data uniqueness of the subdata set, Fc3_1(n) is the uniqueness of the time variable value of the nth record, and the repeat _ ratio is the index repetition rate of the subdata set;
Figure BDA0003186747260000045
Figure BDA0003186747260000046
using formulas
Figure BDA0003186747260000047
Calculating the data consistency of each subdata set;
wherein, c4For data consistency of a subdata set, Fc4_1(n) and Fc4_2(n) the format consistency and precision consistency of the nth record respectively;
Figure BDA0003186747260000048
Figure BDA0003186747260000049
using formulas
Figure BDA00031867472600000410
Calculating the data effectiveness of each sub data set;
wherein, c5For data validity of a subdata set, Fc5_1(n) and Fc5_2(n) are each independentlyThe data formats of the n records are valid and the longitude and latitude coordinate is valid;
Figure BDA0003186747260000051
Figure BDA0003186747260000052
w is a longitude and latitude collection in the measuring area.
Optionally, the calculating a comprehensive cloud characteristic parameter value of the test data according to the weight of each evaluation index and the cloud model characteristic parameter value of each evaluation index of the test data specifically includes:
calculating a comprehensive cloud characteristic parameter value of the test data by using the following formula according to the weight of each evaluation index and the cloud model characteristic parameter of each evaluation index of the test data;
Figure BDA0003186747260000053
Figure BDA0003186747260000054
Figure BDA0003186747260000055
wherein E isxTo test the expectations of the synthetic cloud of data, EnEntropy of the synthetic cloud to test data, HeHyper-entropy, omega, of a synthetic cloud for testing dataiThe weight of the i-th evaluation index is represented.
Optionally, the determining, according to the comprehensive cloud characteristic parameter value of the test data, the quality of the test data based on the normal cloud model specifically includes:
taking the expectation of the comprehensive cloud of the test data as an expectation, and taking the super entropy of the comprehensive cloud of the test data as a standard deviation to generate a first normal random number;
taking the entropy of the comprehensive cloud of the test data as an expectation, and taking the first normal random number as a standard deviation to generate a second normal random number;
taking the expectation of the p-th quality level cloud as the expectation, and taking the super-entropy of the p-th quality level cloud as the standard deviation, and generating a third normal random number;
according to the second normal random number and the third normal random number, utilizing a formula
Figure BDA0003186747260000056
Figure BDA0003186747260000057
Calculating a similarity evaluation result of the comprehensive cloud of the test data and the p-th quality grade cloud;
wherein ξpmM-th similarity evaluation result, x, of the synthetic cloud representing the test data and the p-th quality class cloudimDenotes a second normal random number, EnpIndicating that the desire of the pth quality level cloud is desired,
Figure BDA0003186747260000061
represents a third normal random number;
increasing the value of M by 1, and returning to the step of generating a first normal random number by taking the expectation of the comprehensive cloud of the test data as the expectation and the super-entropy of the comprehensive cloud of the test data as the standard deviation until M similarity evaluation results of the comprehensive cloud of the test data and the p-th quality grade cloud are generated;
according to the M similarity evaluation results of the comprehensive cloud of the test data and the p-th quality grade cloud, a formula xi is usedp＝∑ξpmCalculating the comprehensive similarity of the comprehensive cloud of the test data and the p-th quality grade cloud; wherein ξpRepresenting the comprehensive similarity of the comprehensive cloud of the test data and the p-th quality grade cloud;
increasing the value of p by 1, initializing the value of m, and returning to the step of generating a first normal random number by taking the expectation of the comprehensive cloud of the test data as the expectation and the super-entropy of the comprehensive cloud of the test data as the standard deviation until the comprehensive similarity of the comprehensive cloud of the test data and each quality grade cloud is obtained;
and determining the quality grade of the test data according to the comprehensive similarity between the comprehensive cloud of the test data and each quality grade cloud.
Optionally, the determining, according to the comprehensive cloud characteristic parameter value of the test data, the quality of the test data based on the normal cloud model further includes:
calculating the cloud characteristic parameters of each quality grade cloud by using the following formula;
Figure BDA0003186747260000062
Figure BDA0003186747260000063
Figure BDA0003186747260000064
wherein E isxp(ii) a desire for a pth quality class cloud; enpEntropy of the p-th quality class cloud; hepThe super entropy of the p-th quality level cloud; top ispTaking an upper bound for the evaluation of the p-th grade; btmpAnd taking a lower bound for the evaluation of the p-th grade.
An autonomous driving road test data quality determination system based on a normal cloud model, the system comprising:
the weight determination module is used for determining the weight of each evaluation index for determining the quality of the test data of the automatic driving road by adopting an analytic hierarchy process; the evaluation indexes comprise data correctness and qualification, data integrity, data uniqueness, data consistency and data effectiveness;
the cloud model characteristic parameter value calculation module is used for calculating the cloud model characteristic parameter value of each evaluation index of the test data by adopting a reverse cloud emitter algorithm;
the comprehensive cloud characteristic parameter value calculation module is used for calculating a comprehensive cloud characteristic parameter value of the test data according to the weight of each evaluation index and the cloud model characteristic parameter value of each evaluation index of the test data;
and the quality determination module is used for determining the quality of the test data based on the normal cloud model according to the comprehensive cloud characteristic parameter value of the test data.
Optionally, the weight determining module specifically includes:
the initialization submodule is used for constructing an index evaluation system for determining the quality of the test data of the automatic driving road and determining an initial weight judgment matrix of the index evaluation system;
a consistency ratio calculation submodule for using a formula
Figure BDA0003186747260000071
Calculating the consistency proportion of the weight judgment matrix;
wherein CR is a consistency ratio, CI is a consistency index,
Figure BDA0003186747260000072
λmaxjudging the maximum characteristic value of the matrix for the weight, wherein r is the dimension of the matrix for the weight judgment, and RI is a random consistency index;
the consistency judgment submodule is used for judging whether the consistency ratio is less than or equal to 0.01 or not to obtain a judgment result;
a weight judgment matrix updating submodule for updating the weight judgment matrix if the judgment result shows no, and returning to the step of using the formula
Figure BDA0003186747260000073
Calculating a consistency ratio of the weight judgment matrix;
a weight calculation submodule for judging a matrix according to the weight and using a formula if the judgment result shows yes
Figure BDA0003186747260000074
Calculating the weight of each evaluation index;
wherein, ω isiWeight representing the ith evaluation index, aijAnd (3) the (i, j) th element in the weight judgment matrix represents the risk importance degree of the i-th evaluation index relative to the j-th evaluation index.
Optionally, the cloud model characteristic parameter value calculating module specifically includes:
the sub data set segmentation sub module is used for taking the test data of each vehicle as a sub data set to obtain n sub data sets;
the evaluation result calculation submodule is used for calculating the evaluation result of each subdata set under each evaluation index;
the cloud model characteristic parameter value calculation submodule is used for calculating the cloud model characteristic parameter value of each evaluation index of the test data by adopting a reverse cloud emitter algorithm and utilizing the following formula according to the evaluation result of each subdata set under each evaluation index;
Figure BDA0003186747260000081
Figure BDA0003186747260000082
Figure BDA0003186747260000083
wherein E isxi(ii) an expectation of a cloud model that is the ith evaluation index; eniEntropy of the cloud model being the ith evaluation index; heiIs the hyper-entropy of the cloud model for the ith evaluation index; c. CkiIs the evaluation result of the kth sub-data set under the ith evaluation index.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the invention discloses a method for determining the quality of automatic driving road test data based on a normal cloud model, which comprises the following steps: determining the weight of each evaluation index for determining the quality of the automatic driving road test data by adopting an analytic hierarchy process; calculating a cloud model characteristic parameter value of each evaluation index of the test data by adopting a reverse cloud emitter algorithm; calculating a comprehensive cloud characteristic parameter value of the test data according to the weight of each evaluation index and the cloud model characteristic parameter value of each evaluation index of the test data; and determining the quality of the test data based on the normal cloud model according to the comprehensive cloud characteristic parameter value of the test data. The invention realizes the determination of the quality of the test data of the automatic driving road.
The invention provides data quality evaluation indexes, specific judgment criteria and a calculation method aiming at the specific characteristics of the test data of the automatic driving road, maps the quality evaluation scores to specific quality grades based on a normal cloud model method, can realize the reasonability and accuracy of evaluation results, and has the advantages of originality, scientificity, practicability and the like.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without inventive labor.
FIG. 1 is a flow chart of a method for determining the quality of data of an automatic driving road test based on a normal cloud model according to the present invention;
FIG. 2 is a schematic diagram of a method for determining the quality of data of an automatic driving road test based on a normal cloud model according to the present invention;
fig. 3 is a point cloud distribution diagram for drawing comprehensive evaluation cloud and cloud of each data quality grade provided by the invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be described clearly and completely with reference to the accompanying 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.
The invention aims to provide an automatic driving road test data quality determination method based on a normal cloud model so as to determine the automatic driving road test data quality.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
The invention provides an automatic driving road test data quality determination method based on a normal cloud model, which comprises the following steps:
step 101, determining the weight of each evaluation index for determining the quality of the test data of the automatic driving road by adopting an analytic hierarchy process; the evaluation indexes comprise data correctness and qualification, data integrity, data uniqueness, data consistency and data validity.
Step 101 includes the following specific indicators and criteria:
data correctness c1The method mainly comprises the following judgment criteria:
according to longitude and latitude coordinates of two data before and after the time sequence, the calculated instantaneous vehicle speed is in a reasonable range, and the judgment function is as follows:
Figure BDA0003186747260000101
in the formula: fc1_1(n) is a judgment function of the nth data under the criterion, the value of 1 represents that the quality problem does not exist, the value of 0 represents that the quality problem exists, and n is a data number; distance (lat)n，lonn，latn+1，lonn+1) Is a distance calculation functionNumber, longitude and latitude coordinate (lat) according to the nth recordn，lonn) And (2) latitude and longitude coordinates (lat) of the (n + 1) th recordn+1，lonn+1) Calculating to obtain the distance between the two points; t is tnTime, t, representing the nth datan+1Indicating the time of the (n + 1) th piece of data.
According to the vehicle speed record values of the two data before and after the time sequence, the calculated instantaneous acceleration is in a reasonable range, and the judgment function is as follows:
Figure BDA0003186747260000102
in the formula: fc1_2(n) is a judgment function of the nth data under the criterion, the value of 1 represents that the quality problem does not exist, the value of 0 represents that the quality problem exists, and n is a data number;
Figure BDA0003186747260000103
a vehicle speed record value representing the nth data,
Figure BDA0003186747260000104
a vehicle speed record value representing the (n + 1) th data; t is tnTime, t, of the nth datan+1Indicating the time of the (n + 1) th piece of data.
According to the vehicle speed record values of two data before and after the time sequence, the calculated instantaneous acceleration needs to be matched with the record value of the acceleration, and the judgment function is as follows:
Figure BDA0003186747260000105
in the formula: fc1_3(n) is a judgment function of the nth data under the criterion, the value of 1 represents that the quality problem does not exist, the value of 0 represents that the quality problem exists, and n is a data number;
Figure BDA0003186747260000106
vehicle speed meter for representing nth dataThe value is recorded in the data storage device,
Figure BDA0003186747260000107
a vehicle speed record value representing the (n + 1) th data;
Figure BDA0003186747260000108
the recorded value of the acceleration of the automobile representing the nth data,
Figure BDA0003186747260000111
a recorded value of the automobile acceleration representing the (n + 1) th data; t is tnTime, t, representing the nth datan+1Indicating the time of the (n + 1) th piece of data.
The record of the driving mode is correct, and the judgment function is as follows:
Figure BDA0003186747260000112
in the formula: fc1_4(n) is a judgment function of the nth data under the criterion, the value of 1 represents that the quality problem does not exist, the value of 0 represents that the quality problem exists, and n is a data number; and judging whether the driving mode record is the same as the actual condition or not, and adopting an automatic driving road test driving mode distinguishing method driven based on actual measurement data.
Combining the above criteria, data correctness c1The evaluation method (2) is as follows:
Figure BDA0003186747260000113
in the formula: n is the data number, and m is the total number of data in the data set.
Data integrity c2The method mainly comprises the following judgment criteria:
the data value is not missing, and the judgment function is as follows:
Figure BDA0003186747260000114
in the formula: fc2_1And (n) is a judgment function of the nth data under the criterion, the value of 1 represents that the quality problem does not exist, the value of 0 represents that the quality problem exists, and n is a data number.
The data time is kept continuous, and the judgment function is as follows;
Figure BDA0003186747260000115
in the formula: fc2_2(n) is a judgment function of the nth data under the criterion, the value of 1 represents that the quality problem does not exist, the value of 0 represents that the quality problem exists, and n is a data number; t is tnTime, t, representing the nth datanTime, k, of the n +1 th data1Are time intervals specified in the data description.
Combining the above criteria, data integrity c2The evaluation method (2) is as follows:
Figure BDA0003186747260000116
in the formula: n is the data number, and m is the total number of data in the data set.
Data uniqueness c3The method mainly comprises the following judgment criteria:
the same time variable value does not exist, and the judgment function is as follows:
Figure BDA0003186747260000121
in the formula: fc3_1And (n) is a judgment function of the nth data under the criterion, the value of 1 represents that the quality problem does not exist, the value of 0 represents that the quality problem exists, and n is a data number.
The evaluation method of the statistical indexes without the same name comprises the following steps:
Figure BDA0003186747260000122
in the formula: repeat _ ratio is the index repetition rate, and the repetition statistical index is calculated only once.
Combining the above criteria, data uniqueness c3The evaluation method (2) is as follows:
Figure BDA0003186747260000123
in the formula: n is the data number, and m is the total number of data in the data set.
Data consistency c4The method mainly comprises the following judgment criteria:
whether the data expression formats (fraction, decimal and percentage) are consistent or not is judged according to the following function:
Figure BDA0003186747260000124
in the formula: fc4_1And (n) is a judgment function of the nth data under the criterion, the value of 1 represents that the quality problem does not exist, the value of 0 represents that the quality problem exists, and n is a data number.
Whether the data precision (several decimal places) is consistent or not is judged according to the following function:
Figure BDA0003186747260000125
in the formula: fc4_2And (n) is a judgment function of the nth data under the criterion, the value of 1 represents that the quality problem does not exist, the value of 0 represents that the quality problem exists, and n is a data number.
Integrating the above criteria, data consistency c4The evaluation method (2) is as follows:
Figure BDA0003186747260000126
in the formula: n is the data number, and m is the total number of data in the data set.
Data validity c5The method mainly comprises the following judgment criteria:
the data format is valid (e.g. no text is included in the number), and the judgment function is as follows:
Figure BDA0003186747260000131
in the formula: fc5_1And (n) is a judgment function of the nth data under the criterion, the value of 1 represents that the quality problem does not exist, the value of 0 represents that the quality problem exists, and n is a data number.
The longitude and latitude coordinates of the test vehicle are in the measuring area range, and the judgment function is as follows:
Figure BDA0003186747260000132
in the formula: fc5_2(n) is a judgment function of the nth data under the criterion, the value of 1 represents that the quality problem does not exist, the value of 0 represents that the quality problem exists, and n is a data number; latnDimension, lon, representing nth piece of datanRepresents the longitude of the nth piece of data; w represents a longitude and latitude collection within the survey area.
Combining the above criteria, data validity c5The evaluation method (2) is as follows:
Figure BDA0003186747260000133
in the formula: n is the data number, and m is the total number of data in the data set.
Step 102, determining the weight of each evaluation index for determining the quality of the test data of the automatic driving road by using an analytic hierarchy process, specifically comprising:
constructing an index evaluation system for determining the quality of the test data of the automatic driving road, and determining an initial weight judgment matrix of the index evaluation system;
determining the weight of each data quality evaluation index by expert method,
and establishing a paired judgment matrix by adopting a 1-9 scale method according to the importance degree of each data quality evaluation index in a pairwise comparison mode, and performing pairwise comparison by expert scoring. 1 to 9 scale method, as shown in Table 1, wherein aijIndicating how important i is relative to j.
Table 1 data quality evaluation index importance table
Figure BDA0003186747260000134
Figure BDA0003186747260000141
All the comparison results can be represented by a comparison matrix A:
Figure BDA0003186747260000142
note: wherein a isijRepresenting the risk importance of i relative to j, j when compared to i may be represented by the inverse of the scale value of i to j comparison.
Using formulas
Figure BDA0003186747260000143
Calculating the consistency proportion of the weight judgment matrix; wherein CR is a consistency ratio, CI is a consistency index,
Figure BDA0003186747260000144
λmaxjudging the maximum characteristic value of the matrix for the weight, wherein r is the dimension of the matrix for the weight judgment, and RI is a random consistency index; judging whether the consistency ratio is less than or equal to 0.01 or not, and obtaining a judgment result; if the judgment result shows no, updating the weight judgment matrixReturning to the step of using the formula
Figure BDA0003186747260000145
And calculating the consistency ratio of the weight judgment matrix. For example when the weight decision matrix is shown in table 2,
Figure BDA0003186747260000146
Figure BDA0003186747260000147
RI is a random consistency index and can be obtained by looking up a table.
Table 2 data quality evaluation index weight judgment matrix
Figure BDA0003186747260000148
Figure BDA0003186747260000151
If the judgment result shows yes, judging a matrix according to the weight and utilizing a formula
Figure BDA0003186747260000152
Figure BDA0003186747260000153
Calculating the weight of each evaluation index; wherein, ω isiWeight representing the ith evaluation index, aijAnd (3) the (i, j) th element in the weight judgment matrix represents the risk importance degree of the i-th evaluation index relative to the j-th evaluation index. For example, a weight determination matrix as shown in table 2. That is, CR is less than or equal to 0.10, indicating that the consistency of the decision matrix is acceptable.
And 102, calculating a cloud model characteristic parameter value of each evaluation index of the test data by adopting a reverse cloud emitter algorithm.
1. The test data set is cut. The autopilot roadway test data set used in an embodiment may be divided into 33 sub-data sets according to the test vehicle cut data.
2. And for each subdata set, calculating the evaluation result of each data quality evaluation index according to the judgment criterion of each evaluation index in the step one, as shown in table 3.
TABLE 3 data quality evaluation results for each data subset
Figure BDA0003186747260000154
Figure BDA0003186747260000161
Figure BDA0003186747260000171
3. Calculating the cloud model characteristic parameter value E of each evaluation index by adopting a reverse cloud emitter algorithmxi、 EniAnd HeiThe calculation method is as follows:
Figure BDA0003186747260000172
Figure BDA0003186747260000173
Figure BDA0003186747260000174
in the formula: exiIs the expectation of the ith evaluation index cloud model; eniIs the entropy of the ith evaluation index cloud model; heiThe entropy of the ith evaluation index cloud model is obtained; c. CkiIs the kth dataAnd (5) collecting the evaluation results under the ith evaluation index.
And 103, calculating a comprehensive cloud characteristic parameter value of the test data according to the weight of each evaluation index and the cloud model characteristic parameter value of each evaluation index of the test data.
Generating a comprehensive evaluation cloud, and calculating characteristic parameter values T (E) of the comprehensive evaluation cloudx，En，He) The calculation method is as follows:
Figure BDA0003186747260000175
Figure BDA0003186747260000176
Figure BDA0003186747260000177
in the formula: exiIs the expectation of the ith evaluation index cloud model; eniIs the entropy of the ith evaluation index cloud model; heiThe entropy of the ith evaluation index cloud model is obtained; omegaiIs the weight of the i-th evaluation index.
The calculation results of the cloud model characteristic parameter values and the comprehensive evaluation cloud characteristic parameter values of each index are shown in table 4 below.
Table 4 cloud model characteristic parameter values of each index and comprehensive evaluation cloud
E E H
Integrity of 94.03 3.21 1.02
Accuracy of measurement 68.52 7.17 3.52
Uniqueness of 75.20 1.61 0.58
Consistency 90.72 2.43 0.66
Effectiveness of 73.04 14.74 7.01
Comprehensive evaluation cloud T 81.60 7.16 3.13
And step 104, determining the quality of the test data based on the normal cloud model according to the comprehensive cloud characteristic parameter value of the test data.
1. General textureThe quantity grades are divided into five grades of 'good', 'medium', 'poor' and 'poor', the upper limit and the lower limit of the grade are graded, the suggested values of a plurality of experts are inquired, and finally the average value is taken to determine. Finally, the upper bound of the p-th grade is determined to be toppThe lower bound is btmpThe calculation method is as follows:
Figure BDA0003186747260000181
Figure BDA0003186747260000182
in the formula: top ispTaking an upper bound for the evaluation of the p-th grade; btmpTaking a lower bound for the evaluation of the p-th grade; top ispiTaking an upper bound for the evaluation of the p grade suggested by the ith expert; btmpiAnd (4) taking a lower bound for the evaluation of the p grade suggested by the ith expert.
2. Calculating cloud characteristic parameters T of each quality gradep(Exp，Enp，Hep) In which Exp、EnpAnd HepThe calculation method of (2) is as follows:
Figure BDA0003186747260000183
Figure BDA0003186747260000184
Figure BDA0003186747260000191
in the formula: exp(ii) a desire for a pth quality class cloud; enpEntropy of the p-th quality class cloud; hepThe super entropy of the p-th quality level cloud; top ispTaking an upper bound for the evaluation of the p-th grade;btmpand taking a lower bound for the evaluation of the p-th grade.
The data quality class cloud characteristic parameter value calculation results are shown in table 5.
TABLE 5 data quality class cloud characteristic parameter value calculation results
topp btmp E E H
Good taste
100 85 92.5 2.50 0.015
Is preferably used 85 65 75 3.33 0.02
In 65 40 52 4.17 0.025
Is poor 40 15 27.5 4.17 0.025
Difference (D) 15 0 7.5 2.50 0.015
3. And calculating the similarity of the comprehensive evaluation cloud and each grade cloud, and determining the quality grade of the test data of the automatic driving road according to the maximum similarity.
Computing data quality comprehensive evaluation cloud T (E)x，En，He) And each data quality grade cloud Tp(ENp，Enp，Hep) The calculation method of the similarity is as follows:
with EnTo expect, HeGenerating a normal random number E 'for the standard deviation'n；
Then with ExTo expect, E'nGenerating a normal random number x for the standard deviationi；
With EnpTo expectation, HepGenerating a normal random number for the standard deviation
Figure BDA0003186747260000192
Computing
Figure BDA0003186747260000193
Repeating the steps to generate N similarity evaluation results;
calculating the integrated similarity xip＝∑ξpi/N。
The results of the similarity calculation between the comprehensive evaluation cloud and each data quality level cloud are shown in table 6:
table 6 comprehensive evaluation of the similarity between the cloud and each data quality class cloud
Data quality class Good taste Is preferably used In Is poor Difference (D)
Degree of similarity 0.09733 0.272443 0.004889 1.80×10-5 3.54×10-33
And drawing a point cloud distribution diagram of the comprehensive evaluation cloud and the cloud of each data quality grade, as shown in fig. 3.
And comprehensively considering the point cloud distribution diagram and the similarity calculation structure, and judging the data quality grade of the automatic driving road test data set to be a 'better' grade.
The invention also provides an automatic driving road test data quality determination system based on the normal cloud model, which comprises the following steps:
the weight determination module is used for determining the weight of each evaluation index for determining the quality of the test data of the automatic driving road by adopting an analytic hierarchy process; the evaluation indexes comprise data correctness and qualification, data integrity, data uniqueness, data consistency and data effectiveness.
The weight determining module specifically includes: the initialization submodule is used for constructing an index evaluation system for determining the quality of the test data of the automatic driving road and determining an initial weight judgment matrix of the index evaluation system; a consistency ratio calculation submodule for using a formula
Figure BDA0003186747260000202
Calculating the consistency ratio of the weight judgment matrix; wherein CR is a consistency ratio, CI is a consistency index,
Figure BDA0003186747260000203
λmaxjudging the maximum characteristic value of the matrix for the weight, wherein r is the dimension of the matrix for the weight judgment, and RI is a random consistency index; the consistency judgment submodule is used for judging whether the consistency ratio is less than or equal to 0.01 or not to obtain a judgment result; a weight judgment matrix updating submodule for updating the weight judgment matrix if the judgment result shows no, and returning to the step of using the formula
Figure BDA0003186747260000204
Calculating a consistency ratio of the weight judgment matrix; a weight calculation submodule for judging a matrix according to the weight and using a formula if the judgment result shows yes
Figure BDA0003186747260000205
Calculating the weight of each evaluation index; wherein, ω isiWeight representing the ith evaluation index, aijAnd (3) the (i, j) th element in the weight judgment matrix represents the risk importance degree of the i-th evaluation index relative to the j-th evaluation index.
And the cloud model characteristic parameter value calculation module is used for calculating the cloud model characteristic parameter value of each evaluation index of the test data by adopting a reverse cloud emitter algorithm.
The cloud model characteristic parameter value calculation module specifically includes: the sub data set segmentation sub module is used for taking the test data of each vehicle as a sub data set to obtain n sub data sets; the evaluation result calculation submodule is used for calculating the evaluation result of each subdata set under each evaluation index respectively; the cloud model characteristic parameter value calculation submodule is used for calculating the cloud model characteristic parameter value of each evaluation index of the test data by adopting a reverse cloud emitter algorithm and utilizing the following formula according to the evaluation result of each subdata set under each evaluation index;
Figure BDA0003186747260000211
Figure BDA0003186747260000212
Figure BDA0003186747260000213
wherein E isxi(ii) an expectation of a cloud model that is the ith evaluation index; eniEntropy of the cloud model being the ith evaluation index; heiIs the hyper-entropy of the cloud model for the ith evaluation index; c. CkiIs the evaluation result of the kth sub-data set under the ith evaluation index.
The comprehensive cloud characteristic parameter value calculation module is used for calculating a comprehensive cloud characteristic parameter value of the test data according to the weight of each evaluation index and the cloud model characteristic parameter value of each evaluation index of the test data;
and the quality determination module is used for determining the quality of the test data based on the normal cloud model according to the comprehensive cloud characteristic parameter value of the test data.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the invention discloses a method for determining the quality of automatic driving road test data based on a normal cloud model, which comprises the following steps: determining the weight of each evaluation index for determining the quality of the automatic driving road test data by adopting an analytic hierarchy process; calculating a cloud model characteristic parameter value of each evaluation index of the test data by adopting a reverse cloud emitter algorithm; calculating a comprehensive cloud characteristic parameter value of the test data according to the weight of each evaluation index and the cloud model characteristic parameter value of each evaluation index of the test data; and determining the quality of the test data based on the normal cloud model according to the comprehensive cloud characteristic parameter value of the test data. The invention realizes the determination of the quality of the test data of the automatic driving road.
The invention provides data quality evaluation indexes, specific judgment criteria and a calculation method aiming at the specific characteristics of the test data of the automatic driving road, maps the quality evaluation scores to specific quality grades based on a normal cloud model method, can realize the reasonability and accuracy of evaluation results, and has the advantages of originality, scientificity, practicability and the like.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.

Claims (10)

1. A method for determining the quality of automatic driving road test data based on a normal cloud model is characterized by comprising the following steps:
determining the weight of each evaluation index for determining the quality of the test data of the automatic driving road by adopting an analytic hierarchy process; the evaluation indexes comprise data correctness and qualification, data integrity, data uniqueness, data consistency and data effectiveness;
calculating a cloud model characteristic parameter value of each evaluation index of the test data by adopting a reverse cloud emitter algorithm;
calculating a comprehensive cloud characteristic parameter value of the test data according to the weight of each evaluation index and the cloud model characteristic parameter value of each evaluation index of the test data;
and determining the quality of the test data based on the normal cloud model according to the comprehensive cloud characteristic parameter value of the test data.
2. The method for determining the quality of the automatic driving road test data based on the normal cloud model according to claim 1, wherein the determining the weight of each evaluation index for determining the quality of the automatic driving road test data by using the analytic hierarchy process specifically comprises:
constructing an index evaluation system for determining the quality of the test data of the automatic driving road, and determining an initial weight judgment matrix of the index evaluation system;
using formulas
Figure FDA0003186747250000011
Calculating the consistency proportion of the weight judgment matrix;
wherein CR is a consistency ratio, CI is a consistency index,
Figure FDA0003186747250000012
λmaxdetermining maximum of matrix for weightThe characteristic value, r is the dimension of the weight judgment matrix, and RI is the random consistency index;
judging whether the consistency ratio is less than or equal to 0.01 or not, and obtaining a judgment result;
if the judgment result shows no, the weight judgment matrix is updated, and the step of 'utilizing the formula' is returned
Figure FDA0003186747250000013
Calculating a consistency ratio of the weight judgment matrix;
if the judgment result shows yes, judging a matrix according to the weight and utilizing a formula
Figure FDA0003186747250000014
Figure FDA0003186747250000015
Calculating the weight of each evaluation index;
wherein, ω isiWeight representing the ith evaluation index, aijAnd (3) the (i, j) th element in the weight judgment matrix represents the risk importance degree of the i-th evaluation index relative to the j-th evaluation index.
3. The method for determining the quality of the automatic driving road test data based on the normal cloud model according to claim 1, wherein a reverse cloud emitter algorithm is adopted to calculate the cloud model characteristic parameter value of each evaluation index of the test data, and specifically comprises:
respectively taking the test data of each vehicle as a subdata set to obtain K subdata sets;
respectively calculating the evaluation result of each subdata set under each evaluation index;
according to the evaluation result of each subdata set under each evaluation index, a reverse cloud emitter algorithm is adopted, and the cloud model characteristic parameter value of each evaluation index of the test data is calculated by using the following formula:
Figure FDA0003186747250000021
Figure FDA0003186747250000022
Figure FDA0003186747250000023
wherein E isxi(ii) an expectation of a cloud model that is the ith evaluation index; eniEntropy of the cloud model being the ith evaluation index; heiIs the hyper-entropy of the cloud model for the ith evaluation index; c. CkiIs the evaluation result of the kth sub-data set under the ith evaluation index.
4. The method for determining the quality of the automatic driving road test data based on the normal cloud model according to claim 3, wherein the calculating the evaluation result of each sub data set under each evaluation index respectively specifically comprises:
using formulas
Figure FDA0003186747250000024
Calculating the data correctness of each subdata set;
wherein, c1For data correctness of the subdata set, Fc1_1(n)、Fc1_2(n)、Fc1_3(n) and Fc1_4(N) the correctness of the instantaneous vehicle speed, the correctness of the instantaneous acceleration, the conformity of the acceleration and the correctness of the driving mode which are recorded in the nth record are respectively, and N represents the number of records in the sub data set;
Figure FDA0003186747250000031
Figure FDA0003186747250000032
Figure FDA0003186747250000033
Figure FDA0003186747250000034
distance(latn,lonn,latn+1,lonn+1) Calculating a function for distance (lat)n,lonn) Latitude and longitude coordinates for the nth record, (lat)n+1,lonn+1) The longitude and latitude coordinates of the n +1 record are recorded; v. ofnAnd vn+1The speed of the nth record and the speed of the (n + 1) th record respectively; a isnAcceleration recorded for the nth bar; t is tnAnd tn+1Respectively time of the nth record and the (n + 1) th record;
using formulas
Figure FDA0003186747250000035
Calculating the data integrity of each sub data set;
wherein, c2For data integrity of the subdata sets, Fc2_1(n) and Fc2_2(n) data value missing and data time continuity of the nth record respectively;
Figure FDA0003186747250000036
Figure FDA0003186747250000037
l is a time interval threshold;
using formulas
Figure FDA0003186747250000038
Calculating the data uniqueness of each subdata set;
wherein, c3For data uniqueness of the subdata set, Fc3_1(n) is the uniqueness of the time variable value of the nth record, and the repeat _ ratio is the index repetition rate of the subdata set;
Figure FDA0003186747250000041
Figure FDA0003186747250000042
using formulas
Figure FDA0003186747250000043
Calculating the data consistency of each subdata set;
wherein, c4For data consistency of a subdata set, Fc4_1(n) and Fc4_2(n) the format consistency and precision consistency of the nth record respectively;
Figure FDA0003186747250000044
Figure FDA0003186747250000045
using formulas
Figure FDA0003186747250000046
Calculating the data effectiveness of each sub data set;
wherein, c5For data validity of a subdata set, Fc5_1(n) and Fc5_2(n) the data format validity and the longitude and latitude coordinate validity of the nth record are respectively recorded;
Figure FDA0003186747250000047
Figure FDA0003186747250000048
w is a longitude and latitude collection in the measuring area.
5. The method for determining the quality of the normal cloud model-based automatic driving road test data according to claim 3, wherein the calculating of the comprehensive cloud characteristic parameter value of the test data according to the weight of each evaluation index and the cloud model characteristic parameter value of each evaluation index of the test data specifically comprises:
according to the weight of each evaluation index and the cloud model characteristic parameter value of each evaluation index of the test data, calculating a comprehensive cloud characteristic parameter value of the test data by using the following formula:
Figure FDA0003186747250000049
Figure FDA00031867472500000410
Figure FDA0003186747250000051
wherein E isxTo test the expectations of the synthetic cloud of data, EnEntropy of the synthetic cloud to test data, HeHyper-entropy, omega, of a synthetic cloud for testing dataiThe weight of the i-th evaluation index is represented.
6. The method for determining the quality of the automatic driving road test data based on the normal cloud model according to claim 3, wherein the determining the quality of the test data based on the normal cloud model according to the comprehensive cloud characteristic parameter value of the test data specifically comprises:
taking the expectation of the comprehensive cloud of the test data as an expectation, and taking the super entropy of the comprehensive cloud of the test data as a standard deviation to generate a first normal random number;
taking the entropy of the comprehensive cloud of the test data as an expectation, and taking the first normal random number as a standard deviation to generate a second normal random number;
taking the expectation of the p-th quality level cloud as the expectation, and taking the super-entropy of the p-th quality level cloud as the standard deviation, and generating a third normal random number;
according to the second normal random number and the third normal random number, utilizing a formula
Figure FDA0003186747250000052
Figure FDA0003186747250000053
Calculating a similarity evaluation result of the comprehensive cloud of the test data and the p-th quality grade cloud; wherein ξpmM-th similarity evaluation result, x, of the synthetic cloud representing the test data and the p-th quality class cloudimDenotes a second normal random number, EnpIndicating that the desire of the pth quality level cloud is desired,
Figure FDA0003186747250000054
represents a third normal random number;
increasing the value of M by 1, returning to the step of generating a first normal random number by taking the expectation of the comprehensive cloud of the test data as the expectation and the super-entropy of the comprehensive cloud of the test data as the standard deviation until M similarity evaluation results of the comprehensive cloud of the test data and the p-th quality grade cloud are generated;
according to the M similarity evaluation results of the comprehensive cloud of the test data and the p-th quality grade cloud, a formula xi is utilizedp＝∑ξpmand/M, calculating the synthesis of the comprehensive cloud of the test data and the p quality grade cloudSimilarity; wherein ξpRepresenting the comprehensive similarity of the comprehensive cloud of the test data and the p-th quality grade cloud;
increasing the value of p by 1, initializing the value of m, and returning to the step of generating a first normal random number by taking the expectation of the comprehensive cloud of the test data as the expectation and the super-entropy of the comprehensive cloud of the test data as the standard deviation until the comprehensive similarity of the comprehensive cloud of the test data and each quality grade cloud is obtained;
and determining the quality grade of the test data according to the comprehensive similarity between the comprehensive cloud of the test data and each quality grade cloud.
7. The method for determining the quality of the automatic driving road test data based on the normal cloud model according to claim 3, wherein the method for determining the quality of the test data based on the normal cloud model according to the comprehensive cloud characteristic parameter value of the test data further comprises:
calculating the cloud characteristic parameters of each quality grade cloud by using the following formula:
Figure FDA0003186747250000061
Figure FDA0003186747250000062
Figure FDA0003186747250000063
wherein E isxp(ii) a desire for a pth quality class cloud; enpEntropy of the p-th quality class cloud; hepThe super entropy of the p-th quality level cloud; top ispTaking an upper bound for the evaluation of the p-th grade; btmpAnd taking a lower bound for the evaluation of the p-th grade.
8. An automatic driving road test data quality determination system based on a normal cloud model, characterized in that the system comprises:
the weight determination module is used for determining the weight of each evaluation index for determining the quality of the test data of the automatic driving road by adopting an analytic hierarchy process; the evaluation indexes comprise data correctness and qualification, data integrity, data uniqueness, data consistency and data effectiveness;
the cloud model characteristic parameter value calculation module is used for calculating the cloud model characteristic parameter value of each evaluation index of the test data by adopting a reverse cloud emitter algorithm;
the comprehensive cloud characteristic parameter value calculation module is used for calculating a comprehensive cloud characteristic parameter value of the test data according to the weight of each evaluation index and the cloud model characteristic parameter value of each evaluation index of the test data;
and the quality determination module is used for determining the quality of the test data based on the normal cloud model according to the comprehensive cloud characteristic parameter value of the test data.
9. The system for determining the quality of the automatic driving road test data based on the normal cloud model according to claim 8, wherein the weight determining module specifically comprises:
the initialization submodule is used for constructing an index evaluation system for determining the quality of the test data of the automatic driving road and determining an initial weight judgment matrix of the index evaluation system;
a consistency ratio calculation submodule for using a formula
Figure FDA0003186747250000071
Calculating the consistency proportion of the weight judgment matrix;
wherein CR is a consistency ratio, CI is a consistency index,
Figure FDA0003186747250000072
λmaxjudging the maximum characteristic value of the matrix for the weight, wherein r is the dimension of the matrix for the weight judgment, and RI is a random consistency index;
the consistency judgment submodule is used for judging whether the consistency ratio is less than or equal to 0.01 or not to obtain a judgment result;
a weight judgment matrix updating submodule for updating the weight judgment matrix if the judgment result shows no, and returning to the step of using the formula
Figure FDA0003186747250000073
Calculating a consistency ratio of the weight judgment matrix;
a weight calculation submodule for judging a matrix according to the weight and using a formula if the judgment result shows yes
Figure FDA0003186747250000074
Calculating the weight of each evaluation index;
wherein, ω isiWeight representing the ith evaluation index, aijAnd (3) the (i, j) th element in the weight judgment matrix represents the risk importance degree of the i-th evaluation index relative to the j-th evaluation index.
10. The system for determining the quality of the automatic driving road test data based on the normal cloud model according to claim 9, wherein the cloud model characteristic parameter value calculation module specifically includes:
the sub data set segmentation sub module is used for taking the test data of each vehicle as a sub data set to obtain n sub data sets;
the evaluation result calculation submodule is used for calculating the evaluation result of each subdata set under each evaluation index respectively;
the cloud model characteristic parameter value calculation submodule is used for calculating the cloud model characteristic parameter value of each evaluation index of the test data by adopting a reverse cloud emitter algorithm and utilizing the following formula according to the evaluation result of each subdata set under each evaluation index:
Figure FDA0003186747250000075
Figure FDA0003186747250000081
Figure FDA0003186747250000082
wherein E isxi(ii) an expectation of a cloud model that is the ith evaluation index; eniEntropy of the cloud model being the ith evaluation index; heiIs the hyper-entropy of the cloud model for the ith evaluation index; c. CkiIs the evaluation result of the kth sub-data set under the ith evaluation index.
CN202110864149.4A 2021-07-29 2021-07-29 Automatic driving road test data quality determination method based on normal cloud model Active CN113689081B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110864149.4A CN113689081B (en) 2021-07-29 2021-07-29 Automatic driving road test data quality determination method based on normal cloud model

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110864149.4A CN113689081B (en) 2021-07-29 2021-07-29 Automatic driving road test data quality determination method based on normal cloud model

Publications (2)

Publication Number Publication Date
CN113689081A true CN113689081A (en) 2021-11-23
CN113689081B CN113689081B (en) 2022-06-10

Family

ID=78578215

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110864149.4A Active CN113689081B (en) 2021-07-29 2021-07-29 Automatic driving road test data quality determination method based on normal cloud model

Country Status (1)

Country Link
CN (1) CN113689081B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114996800A (en) * 2022-05-05 2022-09-02 中铁第四勘察设计院集团有限公司 BIM software-based urban road safety assessment method and system

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105426656A (en) * 2015-10-29 2016-03-23 浙江群力电气有限公司 Error calculating method and system
WO2017008180A1 (en) * 2015-07-16 2017-01-19 广东产品质量监督检验研究院 Photovoltaic module failure risk determination method
CN108229813A (en) * 2017-12-29 2018-06-29 北京交通大学 Industry integrated countermeasures system evaluation method based on objectivity cloud matter-element
CN110909983A (en) * 2019-10-25 2020-03-24 浙江工业大学 Multidimensional assessment method for electric energy quality of active power distribution network
CN112396252A (en) * 2020-12-04 2021-02-23 国网江苏省电力有限公司南京供电分公司 Method for acquiring construction success evaluation values of double-creation park of power internet of things
CN112966381A (en) * 2021-03-10 2021-06-15 贵州大学 Power transformer comprehensive state evaluation method based on evidence cloud matter element model

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2017008180A1 (en) * 2015-07-16 2017-01-19 广东产品质量监督检验研究院 Photovoltaic module failure risk determination method
CN105426656A (en) * 2015-10-29 2016-03-23 浙江群力电气有限公司 Error calculating method and system
CN108229813A (en) * 2017-12-29 2018-06-29 北京交通大学 Industry integrated countermeasures system evaluation method based on objectivity cloud matter-element
CN110909983A (en) * 2019-10-25 2020-03-24 浙江工业大学 Multidimensional assessment method for electric energy quality of active power distribution network
CN112396252A (en) * 2020-12-04 2021-02-23 国网江苏省电力有限公司南京供电分公司 Method for acquiring construction success evaluation values of double-creation park of power internet of things
CN112966381A (en) * 2021-03-10 2021-06-15 贵州大学 Power transformer comprehensive state evaluation method based on evidence cloud matter element model

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
王光昊: "基于 AHP 的通抗装备试验任务数据质量评价指标体系构建与分析", 《工程与试验》 *
王光昊: "基于 AHP 的通抗装备试验任务数据质量评价指标体系构建与分析", 《工程与试验》, vol. 61, no. 1, 31 March 2021 (2021-03-31), pages 100 - 103 *
颜宏文: "基于云模型的电网统计数据质量评估方法研究", 《计算机应用与软件》 *
颜宏文: "基于云模型的电网统计数据质量评估方法研究", 《计算机应用与软件》, vol. 31, no. 12, 31 December 2014 (2014-12-31), pages 75 - 77 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114996800A (en) * 2022-05-05 2022-09-02 中铁第四勘察设计院集团有限公司 BIM software-based urban road safety assessment method and system

Also Published As

Publication number Publication date
CN113689081B (en) 2022-06-10

Similar Documents

Publication Publication Date Title
Johannesson et al. Guide to load analysis for durability in vehicle engineering
US11709979B1 (en) Bridge damage identification method considering uncertainty
CN108091133A (en) A kind of safe driving of vehicle behavior analysis method based on huge traffic data analysis
CN104866578A (en) Hybrid filling method for incomplete data
US20070282775A1 (en) Method for ordinal ranking
CN106649709B (en) Vehicle stopping behavior pattern prediction and evaluation method based on data mining
CN109635010B (en) User characteristic and characteristic factor extraction and query method and system
CN109886724B (en) Robust resident travel track identification method
CN106126613A (en) One composition of digressing from the subject determines method and device
CN107016400A (en) The assessment of training set
CN104463548A (en) Carriage quantitative selection method influenced by multiple factors
CN113689081B (en) Automatic driving road test data quality determination method based on normal cloud model
CN105677878A (en) Method and system for vehicle information multi-dimensional display based on BI system
Poot Meta-analysis of previous empirical research findings
CN114580828A (en) Safety evaluation method for numerical control machine during transportation process
CN106780258A (en) A kind of method for building up and device of minor crime decision tree
CN117057190A (en) Temperature rise prediction method for brake disc of railway vehicle
Algur et al. Educational data mining: RT and RF classification models for higher education professional courses
CN110020686A (en) A kind of road surface method for detecting abnormality based on intelligent perception sensing data
CN112329283A (en) Ballastless track CA mortar adjustment layer damage identification method and system
CN109063313B (en) Train traction energy consumption calculation method based on machine learning
CN114169426A (en) Beidou position data-based highway traffic potential safety hazard investigation method
CN113808393A (en) Method for eliminating influence of hybrid control object
Tan et al. Predicting dropout from online education based on neural networks
CN117333067B (en) Intelligent physical education data management method and system

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