CN117828299A - Tire wear degree detection and calculation system - Google Patents
Tire wear degree detection and calculation system Download PDFInfo
- Publication number
- CN117828299A CN117828299A CN202410007683.7A CN202410007683A CN117828299A CN 117828299 A CN117828299 A CN 117828299A CN 202410007683 A CN202410007683 A CN 202410007683A CN 117828299 A CN117828299 A CN 117828299A
- Authority
- CN
- China
- Prior art keywords
- tire
- unit
- neural network
- data
- prediction model
- 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
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 57
- 238000004364 calculation method Methods 0.000 title claims description 8
- 238000012549 training Methods 0.000 claims abstract description 57
- 230000001133 acceleration Effects 0.000 claims abstract description 47
- 238000005457 optimization Methods 0.000 claims abstract description 32
- 238000010276 construction Methods 0.000 claims abstract description 26
- 238000000034 method Methods 0.000 claims abstract description 25
- 238000007781 pre-processing Methods 0.000 claims abstract description 18
- 238000000605 extraction Methods 0.000 claims abstract description 14
- 238000005299 abrasion Methods 0.000 claims abstract description 12
- 238000001914 filtration Methods 0.000 claims abstract description 12
- 238000012800 visualization Methods 0.000 claims abstract description 12
- 230000008569 process Effects 0.000 claims abstract description 10
- 239000002245 particle Substances 0.000 claims description 74
- 238000013528 artificial neural network Methods 0.000 claims description 71
- 239000013598 vector Substances 0.000 claims description 38
- 230000006870 function Effects 0.000 claims description 26
- 230000004913 activation Effects 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 6
- 238000012795 verification Methods 0.000 claims description 6
- 238000005516 engineering process Methods 0.000 claims description 5
- 238000010586 diagram Methods 0.000 claims description 4
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 230000002776 aggregation Effects 0.000 claims description 3
- 238000004220 aggregation Methods 0.000 claims description 3
- 238000002790 cross-validation Methods 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 3
- 230000003213 activating effect Effects 0.000 claims description 2
- 230000009286 beneficial effect Effects 0.000 abstract description 5
- 238000012544 monitoring process Methods 0.000 abstract description 5
- 230000009471 action Effects 0.000 description 3
- 230000000007 visual effect Effects 0.000 description 3
- 238000013461 design Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013016 damping Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Landscapes
- Tires In General (AREA)
Abstract
The invention discloses a tire wear degree detection computing system, which relates to the technical field of tire wear degree detection and comprises a parameter optimization unit, a detection model construction unit, a tire wear detection unit and a visualization unit, wherein a tire data acquisition unit is used for acquiring tire data, the tire data comprises tire internal temperature, tire pressure and triaxial acceleration in the tire operation process, the triaxial acceleration comprises X, Y and Z-axis acceleration, and a data preprocessing unit is used for receiving the tire data transmitted by the tire data acquisition unit and preprocessing the tire data through a filtering method. The invention can collect the data of the internal temperature of the tire, the tire pressure, the triaxial acceleration and the like, and realize the real-time monitoring of the abrasion degree of the tire by means of feature extraction and model training, thereby being beneficial to timely finding the abrasion condition of the tire and improving the running safety of the vehicle.
Description
Technical Field
The invention relates to the technical field of tire wear detection, in particular to a tire wear detection and calculation system.
Background
The automobile tire is a round rubber product arranged on an automobile wheel and is used for being in contact with the ground and providing traction force, supporting weight and damping effect, and the tire can support the weight of the automobile through air pressure and structural design, so that the gravity is uniformly transferred to the ground, and the stability and the operability of the automobile are ensured;
the traditional tire wear detection computing system may not be capable of collecting data such as the internal temperature of a tire, the tire pressure and the triaxial acceleration in real time, so that accuracy and timeliness of monitoring results are limited, the traditional tire wear detection computing system may have lower accuracy in aspects of feature extraction and model training, key features of the tire wear may not be accurately extracted or an accurate prediction model may not be established due to limitations of feature extraction methods and model design, accuracy of the detection results is affected, in the traditional tire wear detection computing system, parameter tuning often requires manual trial and error or empirical adjustment, time consumption of a tuning process may be complicated, optimal parameter configuration cannot be guaranteed, the traditional tire wear detection computing system may lack visual display functions, tire wear cannot be displayed to a driver in an intuitive manner, tire conditions cannot be known in time and corresponding measures are taken, the traditional tire wear detection computing system may not be capable of realizing real-time monitoring and early warning of four tire wear levels, the driver cannot master the tire conditions at any time, and therefore time of replacing or accident may be missed.
Disclosure of Invention
The present invention is directed to a tire wear detection and calculation system, which solves the above-mentioned problems in the prior art.
In order to achieve the above purpose, the present invention provides the following technical solutions: the tire wear detection computing system comprises a parameter optimization unit, a detection model construction unit, a tire wear detection unit and a visualization unit, and further comprises a parameter optimization unit, a detection model construction unit and a visualization unit;
the tire data acquisition unit is used for acquiring tire data, wherein the tire data comprises the internal temperature of a tire, the tire pressure of the tire and the triaxial acceleration in the running process of the tire, the triaxial acceleration comprises X, Y and Z-axis acceleration, and the acquired tire data is transmitted to the data preprocessing unit;
the data preprocessing unit is used for receiving the tire data transmitted by the tire data acquisition unit, preprocessing the tire data through a filtering method, transmitting the triaxial acceleration in the preprocessed tire data to the feature extraction unit, and constructing the tire internal temperature and the tire pressure training set in the preprocessed tire data;
the feature extraction unit is used for receiving the preprocessed tire data transmitted by the data preprocessing unit, constructing a triaxial acceleration waveform chart according to triaxial acceleration and sampling points in the tire data, wherein the abscissa of the waveform chart represents the sampling points, the ordinate represents the triaxial acceleration, the width of a trough in a trough of a Z-axis waveform is extracted as a first feature parameter, the area of the trough in the trough of the Z-axis waveform is extracted as a second feature parameter, the ordinate of a crest of a Y-axis waveform is extracted as a third feature parameter, the distance between the crest and the trough of the Y-axis waveform is extracted as a fourth feature parameter, and transmitting the extracted first, second, third and fourth feature parameters to the training set construction unit;
the training set constructing unit is used for receiving the tire internal temperature and the first, second, third and fourth characteristic parameters transmitted by the tire pressure and characteristic extracting unit in the preprocessed tire data transmitted by the data preprocessing unit, constructing a training set vector according to the received data, adding check bits in the training set vector, and transmitting the constructed training set vector to the parameter optimizing unit.
Preferably, the parameter optimization unit receives the training set vector transmitted by the training set construction unit and the RBF neural network prediction model transmitted by the detection model construction unit, optimizes the training set vector based on the RBF neural network prediction model through a particle swarm optimization algorithm, and transmits the optimized training set vector to the detection model construction unit;
the detection model construction unit is used for constructing an RBF neural network prediction model, transmitting the constructed RBF neural network prediction model to the parameter optimization unit, receiving the optimized training set vector transmitted by the parameter optimization unit, activating the RBF neural network prediction model by using a Gaussian function, training the activated RBF neural network prediction model based on the optimized training set vector so as to obtain a trained RBF neural network prediction model, and transmitting the obtained trained RBF neural network prediction model to the tire wear detection unit, wherein the Gaussian function is specifically as follows:
wherein,representing a Gaussian function, c j Data center representing hidden layer->Representing the data width.
Preferably, the tire wear detection unit receives the trained RBF neural network prediction model transmitted by the detection model construction unit, invokes the RBF neural network prediction model trained by the real-time tire data input value, outputs the real-time wear degree of the tire through the trained RBF neural network prediction model, and transmits the output real-time wear degree of the tire to the visualization unit;
the visual unit receives the real-time wear degree of the tires transmitted by the tire wear detection unit, and displays the real-time wear degree of the four tires of the automobile in the intelligent user display terminal through the WEB technology.
Preferably, the tire data acquisition unit comprises a temperature and tire pressure acquisition module and a triaxial acceleration acquisition module;
the temperature and tire pressure acquisition module acquires the internal temperature and the tire pressure of the tire by installing a temperature and tire pressure integrated sensor on the inner wall of the tire;
the triaxial acceleration acquisition module acquires triaxial acceleration in the running process of the tire by installing an acceleration sensor on the inner wall of the tire.
Preferably, the filtering method includes the steps of:
a1, estimating the order p and the corresponding frequency u of fractional Fourier transform according to a two-dimensional peak searching method;
a2, carrying out fractional Fourier transform on the acquired signal X (t), wherein the fractional Fourier transform formula is as follows:
X P (u)＝Y P (u)+N P (u)
wherein X is p (u) represents the p-order fractional Fourier transform of the acquired signal, Y p (u) represents a p-order fractional Fourier transform of the normal signal y (t), N p (u) represents a p-order fractional fourier transform of the noise signal n (t);
a3, shielding the peak with the highest energy aggregation in the acquired signal X (t) by the following formula:
X p (u)＝X P (u)·C(u)＝Y P (u)·C(u)+N P (u)·C(u)
wherein C (u) represents a bandpass filter centered on u in the fractional fourier transform domain;
a4, carrying out p-order fractional Fourier transform on the fractional Fourier transform domain signal subjected to filtering denoising treatment to obtain a denoised acquisition signal X out (t)。
Preferably, the first characteristic parameter is used for describing the time length of the generated reverse acting force acting on the tire after the tire touches the ground, the second characteristic parameter is used for describing the energy of the generated reverse acting force acting on the tire after the tire touches the ground, the third characteristic parameter is used for describing the maximum difference value of acceleration after the tire touches the ground, and the fourth characteristic parameter is used for describing the time difference between two sections of impact acceleration peaks after the tire touches the ground.
Preferably, the particle swarm optimization algorithm comprises the following steps:
b1, initializing the position and the speed of particles, and randomly generating an initial position and an initial speed;
b2, updating the position and the speed of the particles through a particle updating formula, wherein the particle updating formula is as follows:
v i+1 II＝ωv i II+c 1 ·r 1 ·(pbest II-present II)+c 2 ·r 2 ·(gbest II-present II)
X i+1 ＝X i +V i+1
wherein v is i+1 II represents a particle update, ω represents an inertial weight of the particle, v i II represents the velocity of the particles, c 1 And c 2 Represent learning factor, r 1 And r 2 Represents a random number, pbestII represents the current position of the particle, prentII represents a global extremum, gbest represents an individual extremum, X i+1 Representing the position of the particle at the ith iteration, V i+1 Representing the velocity of the particle at the i+1th iteration;
b3, setting a parameter optimization model based on particle updating, wherein the parameter optimization model is as follows:
wherein,represents the nth data in the xi characteristic parameter,>representing random data in the ζ th feature parameter,representing the original value corresponding to the nth data in the xi-th characteristic parameter,/for>Representing the minimum value in the xi-th feature parameter;
b4, setting a particle swarm layout fitness function, wherein the particle swarm layout fitness function is as follows:
wherein f represents a particle swarm layout fitness function,represents random data in the xi characteristic parameter, M represents adaptive adjustment constant, P j Represents the tire pressure, P, of the jth tire jmin Representing a tire pressure minimum value;
b5, calculating the particle fitness;
b6, calculating an individual optimal solution of each particle, and calculating a global optimal solution of the whole population based on the individual optimal solution of each particle;
b7, optimizing the particles, the speed and the position through a particle updating formula;
and B8, judging whether a termination condition is met, if so, ending the particle swarm optimization algorithm, otherwise, repeatedly executing the steps B5 to B8.
Preferably, in the step B5, calculating the particle fitness includes the steps of:
b501, dividing the random non-repeated training set vector into K parts;
b502, taking 1 part of the RBF neural network prediction model as a verification set, taking the remaining K-1 parts of the RBF neural network prediction model as training set vectors for the RBF neural network prediction model, obtaining a trained RBF neural network prediction model after training, testing the trained RBF neural network prediction model through the verification set, obtaining performance indexes of the trained RBF neural network prediction model, and repeating for K times continuously;
b503, calculating an average value of K groups of test indexes as estimation of the precision of the trained RBF neural network prediction model, and as a performance index of the trained RBF neural network prediction model under the current K-fold cross validation;
and B504, selecting the prediction precision of the trained RBF neural network prediction model as a performance evaluation index, and continuously repeating the steps B501 to B503 until the fitness calculation of all particles is completed.
Preferably, the RBF neural network prediction model adopts a three-layer feedforward neural network, the first layer is an input layer and consists of signal source nodes, the second layer is an hidden layer, an activation function of the hidden layer adopts a radial basis function, the third layer is an output layer, a linear activation function is adopted between the hidden layer and the output layer, the number of input vectors of the RBF neural network prediction model is 7, the number of hidden layers is h, and the number of output vectors is 1.
Compared with the prior art, the invention has the beneficial effects that:
1. the invention can collect data such as internal temperature of the tire, tire pressure, triaxial acceleration and the like of the tire, realize real-time monitoring of the abrasion degree of the tire through a mode of characteristic extraction and training of a model, thereby being beneficial to finding the abrasion condition of the tire in time, improving the running safety of the vehicle, optimizing a training set vector through a parameter optimizing unit by utilizing a particle swarm optimization algorithm, constructing an RBF neural network prediction model, improving the accuracy and stability of the abrasion degree detection of the tire through continuously optimizing parameters and constructing the model, displaying the abrasion degree of four tires of the automobile on a user intelligent display terminal in real time through a WEB technology by a visualization unit, so that a driver can intuitively know the abrasion condition of the tire, take corresponding measures in time, such as replacing the tire or maintaining the tire, prolonging the service life of the tire, monitoring the abrasion degree of the tire in real time, knowing the condition of the tire in time, and taking corresponding measures, thereby being beneficial to improving the running safety of the vehicle and reducing accident risks caused by the abrasion of the tire;
2. according to the invention, through collecting data such as the internal temperature of the tire, the tire pressure, the triaxial acceleration and the like, input parameters required by the tire wear detection can be provided, tire state information can be obtained in real time, an accurate data basis is provided for subsequent processing and analysis, the collected tire data is preprocessed through a filtering method, noise and abnormal values are removed, so that the data are more accurate and reliable, the effects of subsequent feature extraction and model training are improved, the complex tire data can be converted into representative feature vectors through extracting the feature parameters in the tire data such as the trough width, the trough area, the peak ordinate and the distance between the peaks and the troughs of a waveform diagram, so that the key features of the tire wear are extracted, effective input is provided for subsequent model training and prediction, the trained RBF neural network prediction model is received through a tire wear detection unit, the real-time tire data is predicted, and the real-time wear of the tire is output. The intelligent tire abrasion degree display device is beneficial to a driver to timely know the condition of the tire, corresponding measures are taken to ensure driving safety, and the real-time abrasion degree of the tire is visually displayed on the intelligent user display terminal by using a WEB technology through the visual unit.
Drawings
FIG. 1 is a schematic flow chart of an overall system according to an embodiment of the present invention;
FIG. 2 is a flowchart of a filtering method according to an embodiment of the present invention;
FIG. 3 is a flowchart of a particle swarm optimization algorithm according to an embodiment of the present invention;
FIG. 4 is a flowchart of step B5 according to an embodiment of the present invention;
fig. 5 is a block diagram of an internal module of the tire data collection unit according to an embodiment of the present invention.
In the figure: 1. a tire data acquisition unit; 101. a temperature and tire pressure acquisition module; 102. the triaxial acceleration acquisition module; 2. a data preprocessing unit; 3. a feature extraction unit; 4. a training set construction unit; 5. a parameter optimizing unit; 6. a detection model construction unit; 7. a tire wear detection unit; 8. and a visualization unit.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1-5, the present invention provides a technical solution: the tire wear detection computing system comprises a parameter optimization unit 5, a detection model construction unit 6, a tire wear detection unit 7 and a visualization unit 8, and further comprises a parameter optimization unit;
the tire data acquisition unit 1 is used for acquiring tire data, wherein the tire data comprise the internal temperature of a tire, the tire pressure of the tire and the triaxial acceleration in the running process of the tire, the triaxial acceleration comprises X, Y and Z-axis acceleration, and the acquired tire data are transmitted to the data preprocessing unit 2;
the data preprocessing unit 2 is used for receiving the tire data transmitted by the tire data acquisition unit 1, preprocessing the tire data through a filtering method, transmitting the triaxial acceleration in the preprocessed tire data to the feature extraction unit 3, and constructing the tire internal temperature and the tire pressure training set in the preprocessed tire data by the tire pressure training set construction unit 4;
the feature extraction unit 3, the feature extraction unit 3 receives the preprocessed tire data transmitted by the data preprocessing unit 2, constructs a waveform chart of triaxial acceleration according to triaxial acceleration and sampling points in the tire data, wherein an abscissa of the waveform chart represents the sampling points, an ordinate of the waveform chart represents the triaxial acceleration, a trough width in a trough of a Z-axis waveform is extracted as a first feature parameter, a trough area in the trough of the Z-axis waveform is extracted as a second feature parameter, an ordinate of a crest of a Y-axis waveform is extracted as a third feature parameter, a distance between the crest and the trough of the Y-axis waveform is extracted as a fourth feature parameter, and the extracted first, second, third and fourth feature parameters are transmitted to the training set construction unit 4;
the training set constructing unit 4, the training set constructing unit 4 receives the tire internal temperature and the first, second, third and fourth characteristic parameters transmitted by the tire pressure and characteristic extracting unit 3 in the preprocessed tire data transmitted by the data preprocessing unit 2, constructs a training set vector according to the received data, adds check bits in the training set vector, and transmits the constructed training set vector to the parameter optimizing unit 5.
The parameter optimization unit 5 receives the training set vector transmitted by the training set construction unit 4 and the RBF neural network prediction model transmitted by the detection model construction unit 6, optimizes the training set vector based on the RBF neural network prediction model through a particle swarm optimization algorithm, and transmits the optimized training set vector to the detection model construction unit 6;
the detection model construction unit 6 is configured to construct an RBF neural network prediction model, transmit the constructed RBF neural network prediction model to the parameter optimization unit 5, receive the optimized training set vector transmitted by the parameter optimization unit 5, activate the RBF neural network prediction model by using a gaussian function, train the activated RBF neural network prediction model based on the optimized training set vector, so as to obtain a trained RBF neural network prediction model, and transmit the obtained trained RBF neural network prediction model to the tire wear detection unit 7, where the gaussian function is specifically:
wherein,representing a Gaussian function, c j Data center representing hidden layer->Representing the data width;
the tire wear detection unit 7 receives the trained RBF neural network prediction model transmitted by the detection model construction unit 6, invokes the RBF neural network prediction model trained by the real-time tire data input value, outputs the real-time tire wear degree through the trained RBF neural network prediction model, and transmits the output real-time tire wear degree to the visualization unit 8;
the visualization unit 8 receives the real-time wear degrees of the tires transmitted by the tire wear detection unit 7, and displays the real-time wear degrees of the four tires of the automobile in the intelligent user display terminal through the WEB technology;
the tire data acquisition unit 1 comprises a temperature and tire pressure acquisition module 101 and a triaxial acceleration acquisition module 102;
the temperature and tire pressure acquisition module 101 acquires the internal temperature of the tire and the tire pressure by installing a temperature and tire pressure integrated sensor on the inner wall of the tire;
the triaxial acceleration acquisition module 102 acquires triaxial acceleration in the running process of the tire by installing an acceleration sensor on the inner wall of the tire;
the filtering method comprises the following steps:
a1, estimating the order p and the corresponding frequency u of fractional Fourier transform according to a two-dimensional peak searching method;
in this embodiment, a two-dimensional peak search method is a commonly used signal processing method, which is used to estimate the order and corresponding frequency of a signal;
a2, carrying out fractional Fourier transform on the acquired signal X (t), wherein the fractional Fourier transform formula is as follows:
X P (u)＝Y P (u)+N P (u)
wherein X is p (u) represents the p-order fractional Fourier transform of the acquired signal, Y p (u) represents a p-order fractional Fourier transform of the normal signal y (t), N p (u) represents a p-order fractional fourier transform of the noise signal n (t);
a3, shielding the peak with the highest energy aggregation in the acquired signal X (t) by the following formula:
X p (u)＝X P (u)·C(u)＝Y P (u)·C(u)+N P (u)·C(u)
wherein C (u) represents a bandpass filter centered on u in the fractional fourier transform domain;
a4, carrying out p-order fractional Fourier transform on the fractional Fourier transform domain signal subjected to filtering denoising treatment to obtain a denoised acquisition signal X out (t)；
The first characteristic parameter is used for describing the time length of the generated reverse acting force acting on the tire after the tire touches the ground, the second characteristic parameter is used for describing the energy of the generated reverse acting force acting on the tire after the tire touches the ground, the third characteristic parameter is used for describing the maximum difference value of acceleration after the tire touches the ground, and the fourth characteristic parameter is used for describing the time difference between two sections of impact acceleration peaks after the tire touches the ground;
the particle swarm optimization algorithm comprises the following steps:
b1, initializing the position and the speed of particles, and randomly generating an initial position and an initial speed;
b2, updating the position and the speed of the particles through a particle updating formula, wherein the particle updating formula is as follows:
v i+1 II＝ωv i II+c 1 ·r 1 ·(pbest II-present II)+c 2 ·r 2 ·(gbest II-present II)
X i+1 ＝X i +V i+1
wherein v is i+1 II represents a particle update, ω represents an inertial weight of the particle, v i II represents the velocity of the particles, c 1 And c 2 Represent learning factor, r 1 And r 2 Represents a random number, pbestII represents the current position of the particle, prentII represents a global extremum, gbest represents an individual extremum, X i+1 Representing the position of the particle at the ith iteration, V i+1 Representing the velocity of the particle at the i+1th iteration;
b3, setting a parameter optimization model based on particle updating, wherein the parameter optimization model is as follows:
wherein,represents the nth data in the xi characteristic parameter,>represents the xiRandom data in the characteristic parameters of the device,representing the original value corresponding to the nth data in the xi-th characteristic parameter,/for>Representing the minimum value in the xi-th feature parameter;
and B4, setting a particle swarm layout fitness function, wherein the particle swarm layout fitness function is as follows:
wherein f represents a particle swarm layout fitness function,represents random data in the xi characteristic parameter, M represents adaptive adjustment constant, P j Represents the tire pressure, P, of the jth tire jmin Representing a tire pressure minimum value;
b5, calculating the particle fitness;
b6, calculating an individual optimal solution of each particle, and calculating a global optimal solution of the whole population based on the individual optimal solution of each particle;
b7, optimizing the particles, the speed and the position through a particle updating formula;
b8, judging whether the termination condition is met, if so, ending the particle swarm optimization algorithm, otherwise, repeatedly executing the steps B5 to B8;
in step B5, calculating the particle fitness comprises the steps of:
b501, dividing the random non-repeated training set vector into K parts;
b502, taking 1 part of the RBF neural network prediction model as a verification set, taking the remaining K-1 parts of the RBF neural network prediction model as training set vectors for the RBF neural network prediction model, obtaining a trained RBF neural network prediction model after training, testing the trained RBF neural network prediction model through the verification set, obtaining performance indexes of the trained RBF neural network prediction model, and repeating for K times continuously;
b503, calculating an average value of K groups of test indexes as estimation of the precision of the trained RBF neural network prediction model, and as a performance index of the trained RBF neural network prediction model under the current K-fold cross validation;
b504, selecting the prediction precision of the trained RBF neural network prediction model as a performance evaluation index, and continuously repeating the steps B501 to B503 until the fitness calculation of all particles is completed;
the RBF neural network prediction model adopts a three-layer feedforward neural network, the first layer is an input layer and consists of signal source nodes, the second layer is an hidden layer, the activation function of the hidden layer adopts a radial basis function, the third layer is an output layer, a linear activation function is adopted between the hidden layer and the output layer, the number of input vectors of the RBF neural network prediction model is 7, the number of hidden layers is h, and the number of output vectors is 1.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (9)
1. The utility model provides a tire wear degree detects computing system, includes parameter optimization unit (5), detects model construction unit (6), tire wear detection unit (7) and visualization unit (8), its characterized in that:
the tire data acquisition unit (1) is used for acquiring tire data, wherein the tire data comprise tire internal temperature, tire pressure and triaxial acceleration in the running process of the tire, the triaxial acceleration comprises X, Y and Z-axis acceleration, and the acquired tire data are transmitted to the data preprocessing unit (2);
the data preprocessing unit (2), the data preprocessing unit (2) receives the tire data transmitted by the tire data acquisition unit (1), preprocesses the tire data through a filtering method, transmits the triaxial acceleration in the preprocessed tire data to the feature extraction unit (3), and constructs the tire internal temperature and the tire pressure training set in the preprocessed tire data (4);
the feature extraction unit (3), the feature extraction unit (3) receives the preprocessed tire data transmitted by the data preprocessing unit (2), constructs a waveform diagram of triaxial acceleration according to triaxial acceleration and sampling points in the tire data, wherein the abscissa of the waveform diagram represents the sampling points, the ordinate represents the triaxial acceleration, the trough width in the trough of the Z-axis waveform is extracted as a first feature parameter, the trough area in the trough of the Z-axis waveform is extracted as a second feature parameter, the ordinate of the crest of the Y-axis waveform is extracted as a third feature parameter, the distance between the crest and the trough of the Y-axis waveform is extracted as a fourth feature parameter, and the extracted first, second, third and fourth feature parameters are transmitted to the training set construction unit (4);
the training set constructing unit (4), the training set constructing unit (4) receives the tire internal temperature and the first, second, third and fourth characteristic parameters transmitted by the tire pressure and characteristic extracting unit (3) in the preprocessed tire data transmitted by the data preprocessing unit (2), constructs a training set vector according to the received data, adds check bits in the training set vector, and transmits the constructed training set vector to the parameter optimizing unit (5).
2. The tire wear detection computing system of claim 1, wherein: the parameter optimization unit (5) receives the training set vector transmitted by the training set construction unit (4) and the RBF neural network prediction model transmitted by the detection model construction unit (6), optimizes the training set vector based on the RBF neural network prediction model through a particle swarm optimization algorithm, and transmits the optimized training set vector to the detection model construction unit (6);
the detection model construction unit (6) is used for constructing an RBF neural network prediction model, transmitting the constructed RBF neural network prediction model to the parameter optimization unit (5), receiving the optimized training set vector transmitted by the parameter optimization unit (5), activating the RBF neural network prediction model by using a Gaussian function, training the activated RBF neural network prediction model based on the optimized training set vector so as to obtain a trained RBF neural network prediction model, and transmitting the obtained trained RBF neural network prediction model to the tire abrasion detection unit (7).
3. A tire wear detection and calculation system according to claim 2, wherein: the tire wear detection unit (7) receives the trained RBF neural network prediction model transmitted by the detection model construction unit (6), invokes the RBF neural network prediction model trained by the real-time tire data input value, outputs the real-time tire wear degree through the trained RBF neural network prediction model, and transmits the output real-time tire wear degree to the visualization unit (8);
the visualization unit (8) receives the real-time wear degrees of the tires transmitted by the tire wear detection unit (7), and displays the real-time wear degrees of the four tires of the automobile in the intelligent user display terminal through the WEB technology.
4. The tire wear detection computing system of claim 1, wherein: the tire data acquisition unit (1) comprises a temperature and tire pressure acquisition module (101) and a triaxial acceleration acquisition module (102);
the temperature and tire pressure acquisition module (101) acquires the internal temperature and tire pressure of the tire by installing a temperature and tire pressure integrated sensor on the inner wall of the tire;
the triaxial acceleration acquisition module (102) acquires triaxial acceleration in the running process of the tire by installing an acceleration sensor on the inner wall of the tire.
5. The tire wear detection computing system of claim 1, wherein: the filtering method comprises the following steps:
a1, estimating the order p and the corresponding frequency u of fractional Fourier transform according to a two-dimensional peak searching method;
a2, carrying out fractional Fourier transform on the acquired signal X (t), wherein the fractional Fourier transform formula is as follows:
X P (u)＝Y P (u)+N P (u)
wherein X is p (u) represents the p-order fractional Fourier transform of the acquired signal, Y p (u) represents a p-order fractional Fourier transform of the normal signal y (t), N p (u) represents a p-order fractional fourier transform of the noise signal n (t);
a3, shielding the peak with the highest energy aggregation in the acquired signal X (t) by the following formula: x'. p (u)＝X P (u)·C(u)＝Y P (u)·C(u)+N P (u)·C(u)
Wherein C (u) represents a bandpass filter centered on u in the fractional fourier transform domain;
a4, carrying out p-order fractional Fourier transform on the fractional Fourier transform domain signal subjected to filtering denoising treatment to obtain a denoised acquisition signal X out (t)。
6. The tire wear detection computing system of claim 1, wherein: the first characteristic parameter is used for describing the time length of the generated reverse acting force acting on the tire after the tire touches the ground, the second characteristic parameter is used for describing the energy of the generated reverse acting force acting on the tire after the tire touches the ground, the third characteristic parameter is used for describing the maximum difference value of acceleration after the tire touches the ground, and the fourth characteristic parameter is used for describing the time difference between two sections of impact acceleration peaks after the tire touches the ground.
7. The tire wear detection computing system of claim 1, wherein: the particle swarm optimization algorithm comprises the following steps:
b1, initializing the position and the speed of particles, and randomly generating an initial position and an initial speed;
b2, updating the position and the speed of the particles through a particle updating formula, wherein the particle updating formula is as follows:
v i+1 II＝ωv i II+c 1 ·r 1 ·(pbestII-presentII)+c 2 ·r 2 ·(gbestII-presentII)
X i+1 ＝X i +V i+1
wherein v is i+1 H represents the update of the particles, ω represents the inertial weight of the particles, v i H represents the velocity of the particles, c 1 And c 1 Represent learning factor, r 1 And r 2 Represents a random number, pbestII represents the current position of the particle, prentII represents a global extremum, gbest represents an individual extremum, X i+1 Representing the position of the particle at the ith iteration, V i+1 Representing the velocity of the particle at the i+1th iteration;
b3, setting a parameter optimization model based on particle updating, wherein the parameter optimization model is as follows:
wherein,represents the nth data in the xi characteristic parameter,>represents random data in the xi-th characteristic parameter,/, and>representing the original value corresponding to the nth data in the xi-th characteristic parameter,/for>Representing the minimum value in the xi-th feature parameter;
b4, setting a particle swarm layout fitness function, wherein the particle swarm layout fitness function is as follows:
wherein f represents a particle swarm layout fitness function,represents random data in the xi characteristic parameter, M represents adaptive adjustment constant, P j Represents the tire pressure, P, of the jth tire jmin Representing a tire pressure minimum value;
b5, calculating the particle fitness;
b6, calculating an individual optimal solution of each particle, and calculating a global optimal solution of the whole population based on the individual optimal solution of each particle;
b7, optimizing the particles, the speed and the position through a particle updating formula;
and B8, judging whether a termination condition is met, if so, ending the particle swarm optimization algorithm, otherwise, repeatedly executing the steps B5 to B8.
8. The tire wear detection computing system of claim 1, wherein: in the step B5, calculating the particle fitness includes the steps of:
b501, dividing the random non-repeated training set vector into K parts;
b502, taking 1 part of the RBF neural network prediction model as a verification set, taking the remaining K-1 parts of the RBF neural network prediction model as training set vectors for the RBF neural network prediction model, obtaining a trained RBF neural network prediction model after training, testing the trained RBF neural network prediction model through the verification set, obtaining performance indexes of the trained RBF neural network prediction model, and repeating for K times continuously;
b503, calculating an average value of K groups of test indexes as estimation of the precision of the trained RBF neural network prediction model, and as a performance index of the trained RBF neural network prediction model under the current K-fold cross validation;
and B504, selecting the prediction precision of the trained RBF neural network prediction model as a performance evaluation index, and continuously repeating the steps B501 to B503 until the fitness calculation of all particles is completed.
9. The tire wear detection computing system of claim 1, wherein: the RBF neural network prediction model adopts a three-layer feedforward neural network, the first layer is an input layer and consists of signal source nodes, the second layer is an hidden layer, an activation function of the hidden layer adopts a radial basis function, the third layer is an output layer, a linear activation function is adopted between the hidden layer and the output layer, the number of input vectors of the RBF neural network prediction model is 7, the number of hidden layers is h, and the number of output vectors is 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410007683.7A CN117828299B (en) | 2024-01-03 | 2024-01-03 | Tire wear degree detection and calculation system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410007683.7A CN117828299B (en) | 2024-01-03 | 2024-01-03 | Tire wear degree detection and calculation system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117828299A true CN117828299A (en) | 2024-04-05 |
| CN117828299B CN117828299B (en) | 2024-06-07 |
Family
ID=90520616
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410007683.7A Active CN117828299B (en) | 2024-01-03 | 2024-01-03 | Tire wear degree detection and calculation system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117828299B (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2015022320A (en) * | 2013-07-16 | 2015-02-02 | 横浜ゴム株式会社 | Data processing method and structure design method |
| CN109033730A (en) * | 2018-09-30 | 2018-12-18 | 北京工业大学 | A kind of tool wear prediction technique based on improved particle swarm optimization algorithm |
| US20190171187A1 (en) * | 2016-05-09 | 2019-06-06 | StrongForce IoT Portfolio 2016, LLC | Methods and systems for the industrial internet of things |
| CN109918792A (en) * | 2019-03-09 | 2019-06-21 | 闽南理工学院 | A kind of dynamics simulation system and method for computer based unbalance amount of tire |
| CN110198585A (en) * | 2019-06-25 | 2019-09-03 | 佛山职业技术学院 | A kind of automobile combination lamps and lanterns device and its application |
| CN110829485A (en) * | 2019-10-10 | 2020-02-21 | 国网湖南综合能源服务有限公司 | LCL filter parameter and control parameter global optimization design method, system and medium based on particle swarm optimization |
| CN112829526A (en) * | 2021-02-09 | 2021-05-25 | 上海交通大学 | Tire wear degree monitoring method and system |
| CN113761649A (en) * | 2021-08-05 | 2021-12-07 | 江苏大学 | Intelligent automobile tire eccentric wear prediction method based on one-dimensional convolutional neural network |
| CN114219360A (en) * | 2021-12-29 | 2022-03-22 | 北京淇瑀信息科技有限公司 | Monitoring safety prediction method and system based on model optimization |
| CN114880806A (en) * | 2022-05-25 | 2022-08-09 | 国网天津市电力公司 | New energy automobile sales prediction model parameter optimization method based on particle swarm optimization |
| CN114964776A (en) * | 2021-02-23 | 2022-08-30 | 华东交通大学 | Wheel set bearing fault diagnosis method based on MSE and PSO-SVM |
| CN116563104A (en) * | 2023-04-20 | 2023-08-08 | 佛山职业技术学院 | Image registration method and image stitching method based on particle swarm optimization |
-
2024
- 2024-01-03 CN CN202410007683.7A patent/CN117828299B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2015022320A (en) * | 2013-07-16 | 2015-02-02 | 横浜ゴム株式会社 | Data processing method and structure design method |
| US20190171187A1 (en) * | 2016-05-09 | 2019-06-06 | StrongForce IoT Portfolio 2016, LLC | Methods and systems for the industrial internet of things |
| CN109033730A (en) * | 2018-09-30 | 2018-12-18 | 北京工业大学 | A kind of tool wear prediction technique based on improved particle swarm optimization algorithm |
| CN109918792A (en) * | 2019-03-09 | 2019-06-21 | 闽南理工学院 | A kind of dynamics simulation system and method for computer based unbalance amount of tire |
| CN110198585A (en) * | 2019-06-25 | 2019-09-03 | 佛山职业技术学院 | A kind of automobile combination lamps and lanterns device and its application |
| CN110829485A (en) * | 2019-10-10 | 2020-02-21 | 国网湖南综合能源服务有限公司 | LCL filter parameter and control parameter global optimization design method, system and medium based on particle swarm optimization |
| CN112829526A (en) * | 2021-02-09 | 2021-05-25 | 上海交通大学 | Tire wear degree monitoring method and system |
| CN114964776A (en) * | 2021-02-23 | 2022-08-30 | 华东交通大学 | Wheel set bearing fault diagnosis method based on MSE and PSO-SVM |
| CN113761649A (en) * | 2021-08-05 | 2021-12-07 | 江苏大学 | Intelligent automobile tire eccentric wear prediction method based on one-dimensional convolutional neural network |
| CN114219360A (en) * | 2021-12-29 | 2022-03-22 | 北京淇瑀信息科技有限公司 | Monitoring safety prediction method and system based on model optimization |
| CN114880806A (en) * | 2022-05-25 | 2022-08-09 | 国网天津市电力公司 | New energy automobile sales prediction model parameter optimization method based on particle swarm optimization |
| CN116563104A (en) * | 2023-04-20 | 2023-08-08 | 佛山职业技术学院 | Image registration method and image stitching method based on particle swarm optimization |
Non-Patent Citations (3)
| Title |
|---|
| MEIQI WANG等: "Research and application of neural network for tread wear prediction and optimization", 《MECHANICAL SYSTEMS AND SIGNAL PROCESSING》, vol. 162, 31 May 2021 (2021-05-31), pages 1 - 26 * |
| 王高阳 等: "基于FRFT-RBF网络的船舶电网故障定位方法", 《武汉船舶职业技术学院学报》, vol. 22, no. 06, 25 December 2023 (2023-12-25), pages 77 - 84 * |
| 许斌: "基于粒子群算法的水库灌区输水灌溉管道布局优化方法", 《水利技术监督》, no. 12, 25 December 2023 (2023-12-25), pages 106 - 108 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117828299B (en) | 2024-06-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108108766B (en) | Driving behavior identification method and system based on multi-sensor data fusion | |
| CN111507233B (en) | Multi-mode information fusion intelligent vehicle pavement type identification method | |
| Luo et al. | A dual-tree complex wavelet enhanced convolutional LSTM neural network for structural health monitoring of automotive suspension | |
| Xu et al. | Tire slip angle estimation based on the intelligent tire technology | |
| CN105139070A (en) | Fatigue driving evaluation method based on artificial nerve network and evidence theory | |
| CN113188807B (en) | Automatic abs result judging algorithm | |
| CN113761649B (en) | Intelligent automobile tire eccentric wear prediction method based on one-dimensional convolutional neural network | |
| US20230012186A1 (en) | System and method for vibroacoustic diagnostic and condition monitoring a system using neural networks | |
| CN104008644A (en) | Urban road traffic noise measurement method based on gradient descent | |
| CN104807640A (en) | Method of detecting life state of spatial rolling bearing based on vibration-sensitive time-frequency characteristics | |
| Aboah et al. | Smartphone-based pavement roughness estimation using deep learning with entity embedding | |
| CN104699980A (en) | Method for realizing IRI value soft measurement based on acceleration sensor | |
| CN115758289B (en) | Rail wave mill identification method based on multitask learning neural network | |
| CN114739671A (en) | Bearing fault diagnosis method based on improved generalized S transformation | |
| US20220018906A1 (en) | Predicting An Outcome Associated With A Driver Of A vehicle | |
| CN106448059A (en) | Wrist strap instrument based driver fatigue detection method | |
| CN117828299B (en) | Tire wear degree detection and calculation system | |
| Liu et al. | An interpretable multiplication-convolution network for equipment intelligent edge diagnosis | |
| Singh | Development of an intelligent tire based tire-vehicle state estimator for application to global chassis control | |
| Singh et al. | Application of machine learning & deep learning techniques in the context of use cases relevant for the tire industry | |
| Yang et al. | An intelligent tyre system for road condition perception | |
| CN113836476B (en) | Pavement evenness measuring method based on smart phone | |
| Menegazzo et al. | Vehicular perception and proprioception based on inertial sensing: a systematic review | |
| Beilfuss et al. | Real-Time Classification of Road Type and Condition in Passenger Vehicles | |
| JP7112909B2 (en) | Tire type discrimination method and tire type discrimination device |
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 |