CN114818120A - Road surface unevenness grade identification method and system considering vehicle speed - Google Patents

Road surface unevenness grade identification method and system considering vehicle speed Download PDF

Info

Publication number
CN114818120A
CN114818120A CN202210358079.XA CN202210358079A CN114818120A CN 114818120 A CN114818120 A CN 114818120A CN 202210358079 A CN202210358079 A CN 202210358079A CN 114818120 A CN114818120 A CN 114818120A
Authority
CN
China
Prior art keywords
imf
module
time domain
road surface
vehicle speed
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202210358079.XA
Other languages
Chinese (zh)
Inventor
巩明德
陈文彬
陈浩
唐家豪
冀承杨
于志伟
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.)
Yanshan University
Original Assignee
Yanshan 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 Yanshan University filed Critical Yanshan University
Priority to CN202210358079.XA priority Critical patent/CN114818120A/en
Publication of CN114818120A publication Critical patent/CN114818120A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G06F30/15—Vehicle, aircraft or watercraft design
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B21/00—Measuring arrangements or details thereof, where the measuring technique is not covered by the other groups of this subclass, unspecified or not relevant
    • G01B21/30—Measuring arrangements or details thereof, where the measuring technique is not covered by the other groups of this subclass, unspecified or not relevant for measuring roughness or irregularity of surfaces
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • 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
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/08—Feature extraction
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/12—Classification; Matching

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Geometry (AREA)
  • Artificial Intelligence (AREA)
  • Data Mining & Analysis (AREA)
  • Software Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computing Systems (AREA)
  • Computer Hardware Design (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Mathematical Physics (AREA)
  • Medical Informatics (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Automation & Control Theory (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Vehicle Body Suspensions (AREA)

Abstract

The invention provides a method and a system for recognizing road surface unevenness grade considering vehicle speed, wherein the method comprises the following steps: s1, simulating random road surfaces with N unevenness grades; s2, dividing the vehicle speed into n levels; s3, taking V j ～V j+1 Within the range ofSimulating on random road surface at any speed; s4, performing empirical mode decomposition on the time domain signal of the suspension dynamic stroke and the time domain signal of the vehicle body vertical acceleration; s5, extracting the characteristics of all signals; s6, obtaining a characteristic sample; s7, obtaining n characteristic sample groups through circulation; s8, training n neural networks by using n characteristic sample groups; s9, according to the vehicle speed interval of the actual vehicle speed signal v, directly executing S8 after the suspension dynamic stroke signal and the vehicle body vertical acceleration signal pass through S4 and S5; and S10, giving the road surface unevenness grade. The invention considers the vehicle speed which is a factor obviously influencing the identification of the road surface unevenness grade, and improves the accuracy of the identification of the road surface grade on the basis of not adding a sensor.

Description

Road surface unevenness grade identification method and system considering vehicle speed
Technical Field
The invention relates to the field of neural networks, in particular to a method and a system for recognizing road surface unevenness grade by considering vehicle speed.
Background
Road surface roughness is a main source of excitation of vehicle running motion, and has important influence on riding comfort, riding safety, vehicle maneuverability and vehicle dynamic load, and a road roughness model is widely researched and applied to research of vehicle dynamics and the like. The vehicle suspension system has the advantages that impact force caused by road unevenness in the driving process and given to a vehicle body at the bottom of a bed is buffered, rapid and accurate road unevenness grade recognition is carried out on the premise that the driving comfort and driving safety of the vehicle are improved through accurate control of active suspension, and the accurate recognition of the road unevenness grade has important significance on improving the control systematicness of the active suspension and on the reality and authenticity of dynamic simulation of the vehicle.
In the aspect of road surface unevenness grade identification, at present, a large amount of theoretical research and experimental verification are carried out by experts at home and abroad, the power spectral density of the road surface is usually used for describing the statistical characteristics of the road surface unevenness, and the road surface unevenness is usually divided into 8 grades from A to H. Most of signals used by the existing road surface grade identification method are vehicle body acceleration, wheel dynamic load and suspension dynamic travel signals, and few people consider the influence of vehicle speed on the road surface unevenness grade identification result, and the dynamic responses of different speeds through the same road surface are obviously different, so that the identification result of the road surface unevenness grade is obviously influenced; few studies considering the speed factor do not show how to reasonably process the speed signal or directly input the speed signal and other signals into the neural network, and the recognition speed is slow and the accuracy is low.
For the identification of the road surface unevenness grade, the prior art mainly adopts the following two modes for identification: 1) the direct measurement method, such as Road profile measurement using the two degrees of Road response-type measurement published by O kavianiour of the university of iran science and technology, is mostly used for experimental Road surfaces, and the Road surface irregularity grade of a fixed experimental Road surface is obtained through instruments, but the instrument debugging is usually complicated and is not suitable for vehicles needing to obtain the Road surface irregularity grade in real time. 2) The method is characterized in that road surface unevenness information is obtained through a transfer function by utilizing vehicle dynamic response, for example, a paper "road surface identification research based on vehicle vibration acceleration response" of the howling of Beijing university of science and engineering proposes that signal analysis is carried out on wheel vibration speed response from correlation, frequency and time domain respectively, and then a road surface identification algorithm is proposed. However, the method does not consider the key factor of speed and influences the identification accuracy.
Therefore, a scheme for quickly and accurately identifying the road surface unevenness grade on the basis of considering the speed is needed at present.
Disclosure of Invention
In order to overcome the defects of the prior art, the influence of speed on the identification of the road surface unevenness grade is considered on the basis of not improving the prior equipment, and the road surface unevenness grade is identified by utilizing a neural network.
The invention provides a road surface unevenness grade identification method considering a vehicle speed, which comprises the following steps of:
step 1: dividing the road surface unevenness into N grades according to the road surface space power spectrum, and respectively simulating N random road surfaces, wherein the N random road surfaces respectively correspond to the road surface unevenness of the N grades, the simulation time of the N random road surfaces is set to t seconds, and the sampling period is t 1 Second;
step 2: initializing vehicle speed segment data and a cycle count j value;
the vehicle speed is driven to travel 0-V max The test vehicle is divided into n-level vehicle speeds which are respectively V 1 ～V 2 ，V 2 ～V 3 ，V 3 ～V 4 ，…，V i ～V i+1 ，…，V n ～V n+1 I ═ 1, 2, 3, …, n; wherein V max Is the maximum value of vehicle speed, V 1 ＝0，
Figure BDA0003582702430000021
Figure BDA0003582702430000022
V n+1 ＝V max ；
The cycle count j value is initialized to 1, i.e., j equals 1;
and step 3: get V j ～V j+1 Simulating any speed W in the range on the N random roads to obtain suspension dynamic travel and vehicle body acceleration signal data;
during simulation, firstly, a system dynamics model is established according to the existing vehicle dynamics, then simulation is carried out on N random roads by using speed W according to the system dynamics model, each random road is simulated to obtain a time domain signal of the suspension dynamic stroke and a time domain signal of the vertical acceleration of the vehicle body, and the speed W can be in V in the simulation process j ～V j+1 The range is changed randomly;
and 4, step 4: recording a time domain signal of the suspension dynamic stroke obtained from the same random road surface as a first time domain signal A 1 The time domain signal of the vertical acceleration of the vehicle body is recorded as a second time domain signal A 2 And carrying out Empirical Mode Decomposition (EMD) on the suspension dynamic travel data to obtain an intrinsic mode component IMF 11 、IMF 12 、…、IMF 1h 、…、IMF 1m M is a positive integer, h is 1, 2, 3, …, m; performing Empirical Mode Decomposition (EMD) on the vehicle body vertical acceleration data, and performing IMF (intrinsic mode component) 21 、IMF 22 、…、IMF 2h 、…、IMF 2m ；
And 5: carrying out feature extraction on the time domain signal and the content modal component;
for the first time domain signal A 1 A second time domain signal A 2 Intrinsic mode component IMF 11 、IMF 12 、…、IMF 1h 、…、IMF 1m 、IMF 21 、IMF 22 、…、IMF 2q 、…、IMF 2m Total of said 2(m +1) signals every t 2 Performing feature extraction once every second, and simultaneously extracting p features each time, wherein p is a positive integer;
step 6: obtaining a characteristic sample number I in a time domain, wherein each sample comprises L sample data;
and 7: j is j +1, j is a positive integer, j is judged to be less than or equal to n, the step 4 is returned, and if not, the step 9 is executed;
simulation in different speed ranges is carried out on N random roads in a circulating mode, and because of N-level vehicle speeds, N groups of feature sample groups containing I feature samples are obtained to prepare for offline training of a neural network;
and 8: inputting n characteristic sample groups into n neural networks for training, wherein each neural network corresponds to a first-level vehicle speed interval to obtain a road surface unevenness grade, and simultaneously obtaining n trained neural networks;
the first neural network classification, the second neural network classification, …, the ith neural network classification … and the nth neural network classification are recorded as corresponding vehicle speed V 1 ～V 2 ，V 2 ～V 3 ，V 3 ～V 4 ，…，V i ～V i+1 ，…，V n ～V n+1 ；
And step 9: acquiring a vehicle speed signal v, a suspension dynamic travel signal d and a vehicle body vertical acceleration signal a through a vehicle sensor, extracting characteristic values of the signals d and a through the steps 4 and 5 according to a vehicle speed interval where the vehicle speed signal v is located, and then directly executing the step 8 to enter the neural network classification of the corresponding vehicle speed interval, namely:
if V i ≤v<V i+1 After the signals d and a are subjected to feature extraction in the steps 4 and 5, the signals enter the ith neural network classification;
wherein the suspension moving stroke signal d and the vehicle body droopThe acceleration signals a respectively correspond to the first time domain signals A 1 And a second time domain signal A 2 ；
Step 10: the ith neural network gives the grade of the road surface unevenness.
Preferably, in the step 5, the feature samples extracted for the time-domain signal and the content modal component specifically include:
extracting the first time domain signal A 1 Obtaining a characteristic sample a through characteristic extraction 11 、a 12 、…a 1p (ii) a Sequentially aiming at connotation modal components IMF 11 、IMF 12 、…、IMF 1h 、…、IMF 1m Using the selected features to perform feature extraction to obtain a feature sample, IMF 1h Obtaining a characteristic sample a through characteristic extraction (h+1)1 、a (h+1)2 、…a (h+1)p (ii) a Such as: IMF 11 Obtaining a characteristic sample a through characteristic extraction 21 、a 22 、…a 2p ；IMF 1m Obtaining a characteristic sample a through characteristic extraction (m+1)1 、a (m+1)2 、…a (m+1)p (ii) a The time domain signal A 2 、IMF 21 、IMF 22 、…、IMF 2q 、…、IMF 2m Also after the above feature extraction process, the second time domain signal A 2 Obtaining a characteristic sample b through characteristic extraction 11 、b 12 、…b 1p (ii) a Sequentially aiming at connotation modal components IMF 21 、IMF 22 、…、IMF 2h 、…、IMF 2m Performing feature extraction to obtain a feature sample, IMF 2q Obtaining a characteristic sample b through characteristic extraction (h+1)1 、b (h+1)2 、…b (h+1)p 。
Preferably, in step 6:
characteristic sample passes through every t for 2(m +1) signals 2 Performing characteristic value extraction once every second, and selecting the characteristic number as p, so that the characteristic sample number I is 2 x (m +1) x p;
because t is used for simulating the N random pavements, the simulation time of the N random pavements is set to t seconds, and the sampling period is t 1 Second, every t for 2(m +1) signals 2 One characteristic value extraction is carried out in secondThus each sample contains
Figure BDA0003582702430000041
Individual sample data.
Preferably, the step 1 specifically comprises:
dividing the road surface unevenness into 8 grades, respectively using ISO Level A-ISO Level H to correspond to road surface unevenness grades 1-8, respectively simulating 8 random road surfaces, respectively corresponding the 8 grades of road surface unevenness to the 8 random road surfaces, setting the simulation time of the 8 random road surfaces to t 1000 seconds, and setting the sampling period to t 1 0.001 second.
Preferably, every t in the step 5 is 2 Feature extraction is performed once in seconds, t 2 0.1 second.
Preferably, in step 9, the vehicle speed signal v is obtained by a speed sensor, the suspension stroke signal d is obtained by a displacement sensor, and the vehicle body vertical acceleration signal a is obtained by an acceleration sensor.
Preferably, the step 2 specifically comprises: and dividing the test vehicle with the vehicle speed of 0-180km/h into 18-level vehicle speed grades.
The invention also discloses a road surface unevenness grade identification system considering the vehicle speed, which comprises the following modules: the simulation system comprises a simulation road surface generation module, an initialization module, a signal simulation module, an empirical mode decomposition module, a characteristic acquisition module, a training data module, a neural network module and a vehicle sensor module; wherein, simulation road surface generation module, the initialization module is all with data transmission to signal simulation module, signal simulation module generates time domain signal and sends to empirical mode decomposition module, empirical mode decomposition module sends time domain signal and connotation modal component to the characteristic acquisition module again, the characteristic acquisition module carries out the source according to time domain signal after the characteristic extraction, send training data module or neural network module, training data module preserves the training sample, and train the neural network module, vehicle sensor module gathers the real-time signal of vehicle and sends to empirical mode decomposition module, explain each module in detail below:
a simulation road surface generation module for generating a simulation road surface,the method is used for simulating N random pavements with different grades according to the pavement space power spectrum, the simulation time of the N random pavements is set to t seconds, and the sampling period is t 1 Second;
the vehicle speed control device comprises an initialization module and a control module, wherein the initialization module is used for dividing a vehicle speed into n continuous vehicle speed sections which are not intersected with each other and initializing a cycle count j value;
the signal simulation module is used for simulating any speed in each vehicle speed subsection range on the N random road surfaces according to a system dynamics model to obtain a time domain signal of a suspension dynamic stroke and a time domain signal of a vehicle body vertical acceleration; the system dynamics model is established according to the existing vehicle dynamics, and the speed can be randomly changed in a segmentation range in the simulation process;
the empirical mode decomposition module is used for performing empirical mode decomposition on the time domain signal of the suspension dynamic stroke and the time domain signal of the vehicle body vertical acceleration to obtain m corresponding connotative modal components; recording a time domain signal of the suspension dynamic stroke obtained from the same random road surface as a first time domain signal A 1 The time domain signal of the vertical acceleration of the vehicle body is recorded as a second time domain signal A 2 And carrying out Empirical Mode Decomposition (EMD) on the suspension dynamic travel data to obtain an intrinsic mode component IMF 11 、IMF 12 、…、IMF 1h 、…、IMF 1m M is a positive integer, h is 1, 2, 3, …, m; performing Empirical Mode Decomposition (EMD) on the vehicle body vertical acceleration data, and performing IMF (intrinsic mode component) 21 、IMF 22 、…、IMF 2h 、…、IMF 2m ；
The characteristic acquisition module is used for extracting the characteristics of the time domain signal of the suspension dynamic stroke, the time domain signal of the vehicle body vertical acceleration and the connotative modal components of the time domain signal and the time domain signal of the vehicle body vertical acceleration; the 2(m +1) signals are arranged every t 2 Extracting characteristic values once every second, wherein the characteristic values comprise a mean value, a variance, a square root amplitude value, a root mean square value, a peak value, skewness or kurtosis, and selecting p characteristic values from the characteristic values, wherein p is a positive integer;
a training data module for storing the feature samples extracted by the feature acquisition module to obtain the number of the feature samples I ═ 2 × (m +)1) X p, each sample containing
Figure BDA0003582702430000051
The characteristic samples in the same vehicle speed section are put in the same characteristic sample group, and n characteristic sample groups are total;
the neural network module is used for giving the road surface unevenness grade, the neural network module comprises a first neural network classification, a second neural network classification, …, an ith neural network classification … and an nth neural network classification, the first neural network classification, the second neural network classification, the ith neural network classification and the nth neural network classification respectively correspond to n vehicle speed subsections, and the n neural network classifications are respectively trained by using feature sample groups of the corresponding vehicle speed subsections;
the vehicle sensor module comprises a speed sensor, a displacement sensor and a speed sensor, and is respectively used for obtaining a vehicle speed signal v, a suspension dynamic travel signal d and a vehicle body vertical acceleration signal a; the suspension dynamic travel signal d and the vehicle body vertical acceleration signal a respectively correspond to the first time domain signal A 1 And a second time domain signal A 2 ；
Preferably, after the neural network module is trained, the neural network module, the empirical mode decomposition module and the feature acquisition module are only mounted on the vehicle body, and the vehicle sensor module uses a sensor in the existing vehicle.
Compared with the prior art, the invention has the following beneficial effects:
(1) the invention utilizes the sensor in the prior active suspension system to identify the road surface, and the sensor is not added, so the system is simple and reliable.
(2) The vehicle speed which is a factor obviously influencing the identification of the road surface unevenness grade is considered, and the identification accuracy is improved.
(3) The identification method of the invention has high precision and reliable work.
Drawings
FIG. 1 is a flow chart of the steps of a method for identifying a road surface irregularity level in consideration of a vehicle speed according to the present invention;
FIG. 2 is a block schematic diagram of a vehicle speed-based road irregularity grade identification system according to the present invention;
FIG. 3 is a schematic diagram of the location of sensors in the vehicle sensor module of the present invention;
fig. 4 is a schematic view of an embodiment of a method for identifying a road unevenness grade in consideration of a vehicle speed according to the present invention.
Detailed Description
In order to better understand the technical solution of the present invention, the following detailed description is made with reference to the accompanying drawings and examples. In the drawings, like reference numbers can indicate functionally identical or similar elements. While the various aspects of the embodiments are presented in drawings, the drawings are not necessarily drawn to scale unless specifically indicated.
The invention discloses a road surface unevenness grade identification method considering a vehicle speed, which comprises the following specific steps as shown in figure 1:
step 1: according to a spatial power spectrum of a road surface, the road surface unevenness is generally divided into 8 levels and expressed by using ISO Level N, wherein N is a character from A to H, and the A to H correspond to the levels of the road surface unevenness of 1 to 8 respectively. The invention respectively simulates 8 random road surfaces, and the 8 random road surfaces respectively correspond to 8 grades of road surface unevenness. The simulation time of 8 random pavements is set as t seconds, and the sampling period is t 1 Second (t) 1 ＜＜t)。
Step 2: and initializing vehicle speed segment data and a cycle count j value.
The vehicle speed is driven to travel 0-V max The test vehicle is divided into n-level vehicle speeds which are respectively V 1 ～V 2 ，V 2 ～V 3 ，V 3 ～V 4 ，…，V i ～V i+1 ，…，V n ～V n+1 I ═ 1, 2, 3, …, n; wherein V max Is the maximum value of vehicle speed, V 1 ＝0，
Figure BDA0003582702430000061
Figure BDA0003582702430000062
V n+1 ＝V max 。
The cycle count j value is initialized to 1, i.e., j equals 1;
and step 3: get V j ～V j+1 And simulating any speed W in the range on 8 random roads to obtain suspension dynamic stroke and vehicle body acceleration signal data.
During simulation, firstly, a system dynamics model is established according to the existing vehicle dynamics, then, simulation is carried out on 8 random roads by using speed W according to the system dynamics model, each random road is simulated respectively to obtain a time domain signal of a suspension dynamic stroke and a time domain signal of a vehicle body vertical acceleration, and the speed W can be measured at V in the simulation process j ～V j+1 The range was varied freely.
And 4, step 4: recording a time domain signal of the suspension dynamic stroke obtained from the same random road surface as a first time domain signal A 1 The time domain signal of the vertical acceleration of the vehicle body is recorded as a second time domain signal A 2 Performing Empirical Mode Decomposition (EMD) on the suspension dynamic travel data to obtain an intrinsic mode component IMF 11 、IMF 12 、…、IMF 1h 、…、IMF 1m M is a positive integer, h is 1, 2, 3, …, m; empirical Mode Decomposition (EMD) of vehicle body vertical acceleration data and intrinsic mode component IMF 21 、IMF 22 、…、IMF 2h 、…、IMF 2m 。
And 5: and performing feature extraction on the time domain signals and the content modal components.
For the first time domain signal A 1 A second time domain signal A 2 IMF (intrinsic mode function) of content modal component 11 、IMF 12 、…、IMF 1h 、…、IMF 1m 、IMF 21 、IMF 22 、…、IMF 2q 、…、IMF 2m Total 2(m +1) signals every t 2 Second (t) 1 ＜＜t 2 T) carrying out one-time characteristic extraction. The features comprise a mean value, a variance, a square root amplitude, a root mean square value, a peak value, skewness, kurtosis and the like, and p features are selected from the features, wherein p is a positive integer. First time domain signal A 1 Obtaining a characteristic sample a through characteristic extraction 11 、a 12 、…a 1p (ii) a Sequentially aligning content modal components IMF 11 、IMF 12 、…、IMF 1h 、…、IMF 1m Using the selected features to perform feature extraction to obtain a feature sample, IMF 1h Obtaining a characteristic sample a through characteristic extraction (h+1)1 、a (h+1)2 、…a (h+1)p (ii) a Such as: IMF 11 Obtaining a characteristic sample a through characteristic extraction 21 、a 22 、…a 2p ；IMF 1m Obtaining a characteristic sample a through characteristic extraction (m+1)1 、a (m+1)2 、…a (m+1)p . Time domain signal A 2 、IMF 21 、IMF 22 、…、IMF 2q 、…、IMF 2m Also after the above feature extraction process, the second time domain signal A 2 Obtaining a characteristic sample b through characteristic extraction 11 、b 12 、…b 1p (ii) a Sequentially aiming at connotation modal components IMF 21 、IMF 22 、…、IMF 2h 、…、IMF 2m Performing feature extraction to obtain a feature sample, IMF 2h Obtaining a characteristic sample b through characteristic extraction (h+1)1 、b (h+1)2 、…b (h+1)p (ii) a Such as: from IMF 21 Obtaining a characteristic sample b through characteristic extraction 21 、b 22 、…b 2p (ii) a From IMF 2m Obtaining a characteristic sample b through characteristic extraction (m+1)1 、b (m+1)2 、…b (m+1)p 。
Step 6: and obtaining a characteristic sample number I in the time domain, wherein each sample comprises L sample data.
Characteristic sample passes through every t for 2(m +1) signals 2 Performing characteristic value extraction once every second, and selecting the characteristic number as p, so that the characteristic sample number I is 2 x (m +1) x p;
because t is simulated on 8 random road surfaces, the simulation time of the 8 random road surfaces is set to be t seconds, and the sampling period is t 1 Second, every t for 2(m +1) signals 2 One eigenvalue extraction per second, so each sample contains
Figure BDA0003582702430000071
Individual sample data.
And 7: j equals j +1, j is a positive integer, j is judged to be less than or equal to n, the step 4 is returned, and otherwise, the step 9 is executed.
Simulation in different speed ranges is carried out on 8 random roads in a circulating mode, n groups of feature sample groups containing I feature samples are obtained, and preparation is made for offline training of the neural network.
And 8: inputting the n characteristic sample groups into n neural networks for training, wherein each neural network corresponds to a first-level vehicle speed interval to obtain the road surface unevenness grade, and simultaneously obtaining n trained neural networks.
The first neural network classification, the second neural network classification, …, the ith neural network classification … and the nth neural network classification are recorded as corresponding vehicle speed V 1 ～V 2 ，V 2 ～V 3 ，V 3 ～V 4 ，…，V i ～V i+1 ，…，V n ～V n+1 。
And step 9: acquiring a vehicle speed signal v, a suspension dynamic travel signal d and a vehicle body vertical acceleration signal a through a vehicle sensor, extracting characteristic values of the signals d and a through steps 4 and 5 according to a vehicle speed interval in which the vehicle speed signal v is located, and then directly executing step 8 to enter corresponding neural network classification; wherein the suspension dynamic stroke signal d and the vehicle body vertical acceleration signal a respectively correspond to the first time domain signal A 1 And a second time domain signal A 2 。
The vehicle speed signal v is obtained by a speed sensor 3, the suspension dynamic stroke signal d is obtained by a displacement sensor 1, and the vehicle body vertical acceleration signal a is obtained by an acceleration sensor 2.
If V i ≤v<V i+1 And (5) performing feature extraction on the signals d and a in the ith neural network classification.
Step 10: the ith neural network gives the grade of the road surface unevenness.
In the method, the steps 1-8 belong to off-line identification, and the steps 9-10 belong to on-line identification.
In order to implement the method, the invention also provides a road surface unevenness grade identification system considering the vehicle speed, as shown in fig. 2, comprising the following modules: in order to realize the method, the invention also provides a road surface unevenness grade identification system considering the vehicle speed, which comprises the following modules: the simulation system comprises a simulation road surface generation module 1, an initialization module 2, a signal simulation module 3, an empirical mode decomposition module 4, a characteristic acquisition module 5, a training data module 6, a neural network module 7 and a vehicle sensor module 8; the simulated pavement generating module 1 and the initializing module 2 both send data to the signal simulation module 3, the signal simulation module 3 generates a time domain signal and sends the time domain signal to the empirical mode decomposition module 4, the empirical mode decomposition module 4 sends the time domain signal and the content modal component to the feature acquisition module 5, the feature acquisition module 5 sends the time domain signal to the training data module 6 or the neural network module 7 according to the source of the time domain signal after feature extraction, the training data module 6 stores a training sample and trains the neural network module 7, and the vehicle sensor module 8 acquires a real-time signal of a vehicle and sends the real-time signal to the empirical mode decomposition module 4. The following describes each module in detail:
the simulation road surface generation module 1 is used for simulating 8 random road surfaces with different grades according to a road surface space power spectrum, the simulation time of the 8 random road surfaces is set to t seconds, and the sampling period is t 1 And second.
And the initialization module 2 is used for dividing the vehicle speed into n continuous vehicle speed sections which are not intersected with each other and initializing a cycle count j value.
And the signal simulation module 3 is used for simulating any speed in each vehicle speed subsection range on 8 random roads according to the system dynamics model to obtain a time domain signal of the suspension dynamic stroke and a time domain signal of the vehicle body vertical acceleration. The system dynamics model is established according to the existing vehicle dynamics, and the speed can be randomly changed in a segmentation range in the simulation process.
And the empirical mode decomposition module 4 is used for performing empirical mode decomposition on the time domain signal of the suspension dynamic stroke and the time domain signal of the vehicle body vertical acceleration to obtain m corresponding content modal components respectively. Recording a time domain signal of the suspension dynamic stroke obtained from the same random road surface as a first time domain signal A 1 Vertical acceleration of vehicle bodyIs denoted as a second time-domain signal a 2 Performing Empirical Mode Decomposition (EMD) on the suspension dynamic travel data to obtain an intrinsic mode component IMF 11 、IMF 12 、…、IMF 1h 、…、IMF 1m M is a positive integer, h is 1, 2, 3, …, m; empirical Mode Decomposition (EMD) of vehicle body vertical acceleration data and intrinsic mode component IMF 21 、IMF 22 、…、IMF 2h 、…、IMF 2m 。
And the characteristic acquisition module 5 is used for extracting the characteristics of the time domain signal of the suspension dynamic stroke, the time domain signal of the vehicle body vertical acceleration and the content modal components of the time domain signal and the vehicle body vertical acceleration. 2(m +1) signals every t 2 And extracting characteristic values once every second, wherein the characteristic values comprise a mean value, a variance, a square root amplitude value, a root mean square value, a peak value, skewness, kurtosis and the like, and p characteristic values are selected from the characteristic values, and are positive integers.
A training data module 6, configured to store the feature samples extracted by the feature acquisition module, and obtain a total number of feature samples I ═ 2 × (m +1) × p, where each sample includes
Figure BDA0003582702430000091
And (4) putting the feature samples in the same vehicle speed section into the same feature sample group by using the sample data, wherein n feature sample groups are in total.
The neural network module 7 is used for giving the road surface unevenness grade, the neural network module comprises a first neural network classification, a second neural network classification, …, an ith neural network classification … and an nth neural network classification, the first neural network classification, the second neural network classification, the …, the ith neural network classification and the nth neural network classification respectively correspond to n vehicle speed subsections, and the n neural network classifications are respectively trained by using feature sample groups of the corresponding vehicle speed subsections.
The vehicle sensor module 8 comprises a speed sensor 83, a displacement sensor 81 and a speed sensor 82, which are respectively used for obtaining a vehicle speed signal v, a suspension dynamic travel signal d and a vehicle body vertical acceleration signal a. This module is a collection of sensors in existing vehicles and does not require additional installation. The suspension dynamic travel signal d and the vehicle body vertical acceleration signal a respectively correspond to the first time domain signal A 1 And a second time domain signal A 2 。
In the present embodiment, all modules may be installed on the vehicle body, or after the neural network module is trained, the neural network module, the empirical mode decomposition module, the feature acquisition module, the training data module, and the judgment module are installed on the vehicle body, and the vehicle sensor module uses a sensor in an existing vehicle, so as to recognize the grade of the road surface.
For better understanding of the present invention, the specific implementation steps of a road surface irregularity grade identification method considering the vehicle speed are illustrated as follows:
step 1: according to the spatial power spectrum of the road surface, the unevenness of the road surface is generally divided into 8 levels of A-H, and 8-section random road surfaces of ISOLEvel A, ISO Level B, ISO Level C … and ISO Level H are simulated. The simulation time under the unevenness grade of each section of road surface is set as t 1000 seconds, and the sampling period is t 1 0.001 second.
And 2, step: as shown in figure 3, the vehicle with a speedometer of 0-180km/h is selected for detailed description in the example, and as shown in figure 4, 18-grade vehicle speeds of the test vehicle with the vehicle speed of 0-180km/h are respectively graded as 0-10km/h, 10-20km/h, 20-30km/h, … … and 170-180 km/h.
The initialization loop count j has a value of j-1.
And step 3: and any value in the j-th level vehicle speed range is taken to be obtained by simulation on the A-H level road surface, and the suspension dynamic stroke and the vehicle body acceleration of the road surface corresponding to the vehicle speed of 8 levels are obtained.
And 4, step 4: as shown in fig. 3. Firstly, a time domain signal of the suspension motion stroke is recorded as a first time domain signal A 1 The time domain signal of the acceleration of the vehicle body is recorded as a second time domain signal A 2 Performing Empirical Mode Decomposition (EMD) on the suspension dynamic travel data to obtain an intrinsic mode component IMF 11 、IMF 12 、IMF 13 、…、IMF 15 (ii) a Empirical Mode Decomposition (EMD) of vehicle body vertical acceleration data and intrinsic mode component IMF 21 、IMF 22 、IMF 23 、…、IMF 25 。
And 5: extracting the characteristics of time domain signals and content modal components, and performing feature extraction on 12 signals every t 2 The eigenvalue extraction is performed once for 0.1 second, and 3 eigenvalues of root mean square, root mean square amplitude and root mean square value are selected in this example. First time domain signal A 1 Obtaining a characteristic sample a through characteristic extraction 11 、a 12 、a 13 ；IMF 11 Extracting characteristic value to obtain characteristic sample a 21 、a 22 、a 23 ；IMF 12 Extracting characteristic value to obtain characteristic sample a 31 、a 32 、a 33 ；…；IMF 15 Extracting characteristic value to obtain characteristic sample a 61 、a 62 、a 63 (ii) a Time domain signal A 2 、IMF 21 、IMF 22 、IMF 23 、IMF 24 、IMF 25 Also through the above-mentioned feature value extraction process.
Step 6: a total of 36 feature samples were obtained, each sample containing 800000 sample data.
And 7: j is j +1, j is a positive integer, j is judged to be less than or equal to 18, if so, the step 4 is returned, and if not, the step 9 is executed.
The steps are repeated for 3-5 times within the speed range of 10-20km/h, 20-30km/h, … …, 170-180km/h through circulation, and the total time is 18 times, so as to obtain 18 characteristic sample groups containing 36 characteristic samples, wherein each characteristic sample contains 800000 sample data. As shown in fig. 3, the above data processing provides for offline training of the neural network.
And 8: inputting 18 characteristic sample groups into 18 neural networks for training to obtain the road surface unevenness grade, simultaneously obtaining 18 trained neural networks, recording the neural networks as a first neural network classification, a second neural network classification, … and an 18 th neural network classification, and respectively corresponding to the vehicle speed V 1 ～V 2 、V 2 ～V 3 、…、V 18 ～V 19 。
And step 9: as shown in fig. 4, a suspension moving stroke signal d of the dual-displacement sensor (1), an acceleration sensor (2) obtains a vehicle body vertical acceleration signal a, a speed sensor (3) obtains a vehicle speed signal v, and the vehicle speed signal v enters different neural networks according to the vehicle speed signal v, specifically:
if v is more than or equal to 0 and less than 10km/s, the signals d and a enter a first neural network for classification after feature extraction in the steps 4 and 5;
if v is more than or equal to 0 and less than 20km/s, the signals d and a enter a second neural network for classification after feature extraction in the steps 4 and 5;
if v is more than or equal to 0 and less than 30km/s, the signals d and a enter a third neural network for classification after feature extraction in the steps 4 and 5;
……
if v is more than or equal to 0 and less than 170km/s, the signals d and a enter a 17 th neural network for classification after feature extraction in the steps 4 and 5;
if v is more than or equal to 0 and less than 180km/s, the signals d and a enter the 18 th neural network for classification after feature extraction in the steps 4 and 5.
Step 10: the road surface irregularity grade is output through step 9.
Finally, it should be noted that: the above-mentioned embodiments are only used for illustrating the technical solution of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.

Claims (9)

1. A road surface unevenness grade identification method considering a vehicle speed is characterized in that: which comprises the following steps:
step 1: dividing the road surface unevenness into N grades according to the road surface space power spectrum, respectively simulating N random road surfaces, wherein the N random road surfaces respectively correspond to the road surface unevenness of the N grades, the simulation time of the N random road surfaces is set to t seconds, and the sampling period is t seconds 1 Second;
step 2: initializing vehicle speed segment data and a cycle count j value;
the vehicle speed is driven to travel 0-V max Test (2)The vehicle is divided into n grades of vehicle speeds which are respectively V 1 ～V 2 ，V 2 ～V 3 ，V 3 ～V 4 ，…，V i ～V i+1 ，…，V n ～V n+1 I ═ 1, 2, 3, …, n; wherein V max Is the maximum value of vehicle speed, V 1 ＝0，
Figure FDA0003582702420000011
…3，
Figure FDA0003582702420000012
…，
Figure FDA0003582702420000013
V n+1 ＝V max ；
The cycle count j value is initialized to 1, i.e., j equals 1;
and step 3: get V j ～V j+1 Simulating any speed W in the range on the N random roads to obtain suspension dynamic travel and vehicle body acceleration signal data;
during simulation, firstly, a system dynamics model is established according to the existing vehicle dynamics, then simulation is carried out on N random roads by using speed W according to the system dynamics model, each random road is simulated to obtain a time domain signal of the suspension dynamic stroke and a time domain signal of the vertical acceleration of the vehicle body, and the speed W can be in V in the simulation process j ～V j+1 The range is changed randomly;
and 4, step 4: recording a time domain signal of the suspension dynamic stroke obtained from the same random road surface as a first time domain signal A 1 The time domain signal of the vertical acceleration of the vehicle body is recorded as a second time domain signal A 2 And carrying out Empirical Mode Decomposition (EMD) on the suspension dynamic travel data to obtain an intrinsic mode component IMF 11 、IMF 12 、…、IMF 1h 、…、IMF 1m M is a positive integer, h is 1, 2, 3, …, m; empirical Mode Decomposition (EMD) of the vertical acceleration data of the vehicle body and IMF (intrinsic mode function) of the content mode component 21 、IMF 22 、…、IMF 2h 、…、IMF 2m ；
And 5: carrying out feature extraction on the time domain signal and the content modal component;
for the first time domain signal A 1 A second time domain signal A 2 Intrinsic mode component IMF 11 、IMF 12 、…、IMF 1h 、…、IMF 1m 、IMF 21 、IMF 22 、…、IMF 2q 、…、IMF 2m Total of said 2(m +1) signals every t 2 Performing feature extraction once every second, and simultaneously extracting p features each time, wherein p is a positive integer;
step 6: obtaining a characteristic sample number I in a time domain, wherein each sample comprises L sample data;
and 7: j is j +1, j is a positive integer, j is judged to be less than or equal to n, the step 4 is returned, and if not, the step 9 is executed;
simulation in different speed ranges is performed on N random road surfaces in a circulating mode, and because N levels of vehicle speeds exist, N groups of feature sample groups containing I feature samples are obtained to prepare for offline training of a neural network;
and 8: inputting n characteristic sample groups into n neural networks for training, wherein each neural network corresponds to a first-level vehicle speed interval to obtain a road surface unevenness grade, and simultaneously obtaining n trained neural networks;
the first neural network classification, the second neural network classification, …, the ith neural network classification … and the nth neural network classification are recorded as corresponding vehicle speed V 1 ～V 2 ，V 2 ～V 3 ，V 3 ～V 4 ，…，V i ～V i+1 ，…，V n ～V n+1 ；
And step 9: acquiring a vehicle speed signal v, a suspension dynamic travel signal d and a vehicle body vertical acceleration signal a through a vehicle sensor, extracting characteristic values of the signals d and a through the steps 4 and 5 according to a vehicle speed interval where the vehicle speed signal v is located, and then directly executing the step 8 to enter the neural network classification of the corresponding vehicle speed interval, namely:
if V i ≤v<V i+1 After the signals d and a are subjected to feature extraction in the steps 4 and 5, the signals enter the ith neural network classification;
wherein, the suspension frame dynamic travel signal d and the vehicle body vertical acceleration signal a respectively correspond to the first time domain signal A 1 And a second time domain signal A 2 ；
Step 10: the ith neural network gives the grade of the road surface unevenness.
2. The method of identifying a level of unevenness of a road surface in consideration of a vehicle speed according to claim 1, characterized in that: in the step 5, the feature samples extracted for the time domain signal and the content modal component are specifically:
extracting the first time domain signal A 1 Obtaining a characteristic sample a through characteristic extraction 11 、a 12 、…a 1p (ii) a Sequentially aiming at connotation modal components IMF 11 、IMF 12 、…、IMF 1h 、…、IMF 1m Using the selected features to perform feature extraction to obtain a feature sample, IMF 1h Obtaining a characteristic sample a through characteristic extraction (h+1)1 、a (h+1)2 、…a (h+1)p (ii) a Such as: IMF 11 Obtaining a characteristic sample a through characteristic extraction 21 、a 22 、…a 2p ；IMF 1m Obtaining a characteristic sample a through characteristic extraction (m+1)1 、a (m+1)2 、…a (m+1)p (ii) a The time domain signal A 2 、IMF 21 、IMF 22 、…、IMF 2q 、…、IMF 2m Also after the above feature extraction process, the second time domain signal A 2 Obtaining a characteristic sample b through characteristic extraction 11 、b 12 、…b 1p (ii) a Sequentially aiming at connotation modal components IMF 21 、IMF 22 、…、IMF 2h 、…、IMF 2m Performing feature extraction to obtain a feature sample, IMF 2q Obtaining a characteristic sample b through characteristic extraction (h+1)1 、b (h+1)2 、…b (h+1)p 。
3. The method of identifying a level of unevenness of a road surface in consideration of a vehicle speed according to claim 1, characterized in that: in the step 6:
characteristic sample passes through every t for 2(m +1) signals 2 Performing characteristic value extraction once every second, and selecting the characteristic number as p, so that the characteristic sample number I is 2 x (m +1) x p;
because t is used for simulating the N random pavements, the simulation time of the N random pavements is set to t seconds, and the sampling period is t 1 Second, every t for 2(m +1) signals 2 One eigenvalue extraction per second, so each sample contains
Figure FDA0003582702420000021
Individual sample data.
4. The method of identifying a level of unevenness of a road surface in consideration of a vehicle speed according to claim 1, characterized in that: the step 1 specifically comprises the following steps:
dividing the road surface unevenness into 8 grades, respectively using ISO Level A-ISO Level H to correspond to road surface unevenness grades 1-8, respectively simulating 8 random road surfaces, respectively corresponding the 8 grades of road surface unevenness to the 8 random road surfaces, setting the simulation time of the 8 random road surfaces to t 1000 seconds, and setting the sampling period to t 1 0.001 second.
5. The method of identifying a level of unevenness of a road surface in consideration of a vehicle speed according to claim 1, characterized in that: every t in the step 5 2 Feature extraction is performed once in seconds, t 2 0.1 second.
6. The method of identifying a level of unevenness of a road surface in consideration of a vehicle speed according to claim 1, characterized in that: in the step 9, the vehicle speed signal v is obtained through a speed sensor, the suspension dynamic travel signal d is obtained through a displacement sensor, and the vehicle body vertical acceleration signal a is obtained through an acceleration sensor.
7. The method of identifying a level of unevenness of a road surface in consideration of a vehicle speed according to claim 1, characterized in that: the step 2 specifically comprises the following steps: and dividing the test vehicle with the vehicle speed of 0-180km/h into 18-level vehicle speed grades.
8. A road surface unevenness grade recognition system considering a vehicle speed is characterized in that: it includes the following modules: the simulation system comprises a simulation road surface generation module, an initialization module, a signal simulation module, an empirical mode decomposition module, a characteristic acquisition module, a training data module, a neural network module and a vehicle sensor module; wherein, simulation road surface generation module, the initialization module is all with data transmission to signal simulation module, signal simulation module generates time domain signal and sends to empirical mode decomposition module, empirical mode decomposition module sends time domain signal and connotation modal component to the characteristic acquisition module again, the characteristic acquisition module carries out the source according to time domain signal after the characteristic extraction, send training data module or neural network module, training data module preserves the training sample, and train the neural network module, vehicle sensor module gathers the real-time signal of vehicle and sends to empirical mode decomposition module, explain each module in detail below:
the simulation road surface generation module is used for simulating N random road surfaces with different grades according to the road surface space power spectrum, the simulation time of the N random road surfaces is set to t seconds, and the sampling period is t 1 Second;
the vehicle speed control device comprises an initialization module and a control module, wherein the initialization module is used for dividing a vehicle speed into n continuous vehicle speed sections which are not intersected with each other and initializing a cycle count j value;
the signal simulation module is used for simulating any speed in each vehicle speed subsection range on the N random road surfaces according to a system dynamics model to obtain a time domain signal of a suspension dynamic stroke and a time domain signal of a vehicle body vertical acceleration; the system dynamics model is established according to the existing vehicle dynamics, and the speed can be randomly changed in a segmentation range in the simulation process;
the empirical mode decomposition module is used for performing empirical mode decomposition on the time domain signal of the suspension dynamic stroke and the time domain signal of the vehicle body vertical acceleration to obtain m corresponding connotative modal components; recording a time domain signal of the suspension dynamic stroke obtained from the same random road surface as a first time domain signal A 1 The time domain signal of the vertical acceleration of the vehicle body is recorded as a second time domain signal A 2 The suspension framePerforming Empirical Mode Decomposition (EMD) on the dynamic travel data to obtain an intrinsic mode component IMF 11 、IMF 12 、…、IMF 1h 、…、IMF 1m M is a positive integer, h is 1, 2, 3, …, m; performing Empirical Mode Decomposition (EMD) on the vehicle body vertical acceleration data, and performing IMF (intrinsic mode component) 21 、IMF 22 、…、IMF 2h 、…、IMF 2m ；
The characteristic acquisition module is used for extracting the characteristics of the time domain signal of the suspension dynamic stroke, the time domain signal of the vehicle body vertical acceleration and the connotative modal components of the time domain signal and the time domain signal of the vehicle body vertical acceleration; the 2(m +1) signals are arranged every t 2 Extracting characteristic values once every second, wherein the characteristic values comprise a mean value, a variance, a square root amplitude value, a root mean square value, a peak value, skewness or kurtosis, and selecting p characteristic values from the characteristic values, wherein p is a positive integer;
a training data module for storing the feature samples extracted by the feature acquisition module to obtain a total number of feature samples, I being 2 × (m +1) × p, each sample including
Figure FDA0003582702420000041
The characteristic samples in the same vehicle speed section are put in the same characteristic sample group, and n characteristic sample groups are total;
the neural network module is used for giving the road surface unevenness grade, the neural network module comprises a first neural network classification, a second neural network classification, …, an ith neural network classification … and an nth neural network classification, the first neural network classification, the second neural network classification, the ith neural network classification and the nth neural network classification respectively correspond to n vehicle speed subsections, and the n neural network classifications are respectively trained by using feature sample groups of the corresponding vehicle speed subsections;
the vehicle sensor module comprises a speed sensor, a displacement sensor and a speed sensor, and is respectively used for obtaining a vehicle speed signal v, a suspension dynamic travel signal d and a vehicle body vertical acceleration signal a; the suspension dynamic travel signal d and the vehicle body vertical acceleration signal a respectively correspond to the first time domain signal A 1 And a second time domain signal A 2 。
9. The system for recognizing the level of unevenness of a road surface in consideration of a vehicle speed according to claim 8, wherein: after the simulation road surface generation module, the initialization module, the signal simulation module, the empirical mode decomposition module, the feature acquisition module, the training data module and the judgment module are used for training the neural network module, only the neural network module, the empirical mode decomposition module and the feature acquisition module are installed on a vehicle body, and the vehicle sensor module uses a sensor in the existing vehicle.
CN202210358079.XA 2022-04-06 2022-04-06 Road surface unevenness grade identification method and system considering vehicle speed Pending CN114818120A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210358079.XA CN114818120A (en) 2022-04-06 2022-04-06 Road surface unevenness grade identification method and system considering vehicle speed

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210358079.XA CN114818120A (en) 2022-04-06 2022-04-06 Road surface unevenness grade identification method and system considering vehicle speed

Publications (1)

Publication Number Publication Date
CN114818120A true CN114818120A (en) 2022-07-29

Family

ID=82534148

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210358079.XA Pending CN114818120A (en) 2022-04-06 2022-04-06 Road surface unevenness grade identification method and system considering vehicle speed

Country Status (1)

Country Link
CN (1) CN114818120A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116881768A (en) * 2023-06-21 2023-10-13 清华大学 Data processing method, device, computer equipment and storage medium

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116881768A (en) * 2023-06-21 2023-10-13 清华大学 Data processing method, device, computer equipment and storage medium
CN116881768B (en) * 2023-06-21 2024-03-12 清华大学 Data processing method, device, computer equipment and storage medium

Similar Documents

Publication Publication Date Title
CN111507233B (en) Multi-mode information fusion intelligent vehicle pavement type identification method
CN104164829B (en) Detection method of road-surface evenness and intelligent information of road surface real-time monitoring system based on mobile terminal
CN106092600B (en) A kind of pavement identification method for strengthening road for proving ground
CN104802737A (en) Mobile phone based vehicle abnormality driving behavior detection method
CN106021789B (en) Railway vehicle suspension system Fault Classification and system based on fuzzy intelligence
CN103674578A (en) Detection method for high-speed train operation dynamics performance state
CN104833534A (en) Train running fault diagnosis device based on multi-source information fusion, and method
CN108839657B (en) A method of online recognition road roughness information is responded based on automobile vibration
CN104792937A (en) Bridge head bump detection evaluation method based on vehicle-mounted gravitational acceleration sensor
Gorges et al. Road classification for two-wheeled vehicles
CN114818120A (en) Road surface unevenness grade identification method and system considering vehicle speed
CN110806324A (en) Wheel polygon abrasion detection method based on rail displacement and data acquisition equipment
CN114997218A (en) Recognition and detection method for polygonal abrasion of wheels of railway vehicle
CN106710242A (en) Method for recognizing vehicle quantity of motorcade based on dynamic strain of bridge
CN104182995B (en) Highway roadside landscape color evaluation method based on driving fatigue
CN107512143A (en) Tire pressure calculates method, apparatus and the application on bicycle
Hurtado et al. Unsupervised learning-based framework for indirect structural health monitoring using adversarial autoencoder
CN107357983A (en) A kind of bridge mobile vehicle Load Identification Methods based on compressed sensing
CN108717548A (en) A kind of increased Activity recognition model update method of facing sensing device dynamic and system
CN116881819A (en) Stay cable working state monitoring method based on isolated forest
CN115101147A (en) Road bearing capacity monitoring method
CN106872005B (en) Method based on bridge dynamic strain identification fleet&#39;s bicycle car weight
CN117312962A (en) Long-service elevator guide rail fault diagnosis method based on transfer learning and data driving
CN114323512B (en) Heavy-load vehicle identification method and system
Zhu et al. A comfort quantification method based on semi-supervised learning for automated vehicle at lane change scenarios

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