CN117719519A - Vehicle running state estimation method - Google Patents
Vehicle running state estimation method Download PDFInfo
- Publication number
- CN117719519A CN117719519A CN202410126243.3A CN202410126243A CN117719519A CN 117719519 A CN117719519 A CN 117719519A CN 202410126243 A CN202410126243 A CN 202410126243A CN 117719519 A CN117719519 A CN 117719519A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- state estimation
- neural network
- state
- running state
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 70
- 238000013528 artificial neural network Methods 0.000 claims abstract description 50
- 238000001914 filtration Methods 0.000 claims abstract description 37
- 238000005259 measurement Methods 0.000 claims abstract description 15
- 230000003068 static effect Effects 0.000 claims abstract description 7
- 239000011159 matrix material Substances 0.000 claims description 30
- 238000004422 calculation algorithm Methods 0.000 claims description 21
- 230000006870 function Effects 0.000 claims description 21
- 230000001133 acceleration Effects 0.000 claims description 13
- 230000008569 process Effects 0.000 claims description 11
- 230000007704 transition Effects 0.000 claims description 11
- 238000004590 computer program Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000012546 transfer Methods 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 4
- 238000005070 sampling Methods 0.000 claims description 3
- 230000008859 change Effects 0.000 abstract description 6
- 238000012545 processing Methods 0.000 abstract description 6
- 230000000694 effects Effects 0.000 abstract description 3
- 238000004364 calculation method Methods 0.000 description 11
- 230000008901 benefit Effects 0.000 description 5
- 239000002245 particle Substances 0.000 description 5
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 241000288105 Grus Species 0.000 description 1
- 238000013398 bayesian method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Landscapes
- Traffic Control Systems (AREA)
Abstract
The invention discloses a vehicle running state estimation method, which comprises the following steps: acquiring vehicle operation data through a sensor; using the vehicle running data and the vehicle static parameters to establish a vehicle three-degree-of-freedom dynamics model comprising a system equation and an observation equation; and estimating the running state of the vehicle by a method combining extended Kalman filtering and a neural network based on the vehicle running data acquired by the vehicle three-degree-of-freedom dynamics model and the sensor, wherein the part for calculating the Kalman gain in the extended Kalman filtering is calculated by the neural network. The invention adopts the method of combining the extended Kalman filtering and the neural network to estimate the running state of the vehicle, obtains better nonlinear processing capacity and more robust effect, enhances the nonlinear processing capacity, the robustness and the adaptability of the extended Kalman filtering when being used for estimating the running state of the vehicle, and ensures that the extended Kalman filtering is better suitable for the system dynamics and the measurement noise which change in real time.
Description
Technical Field
The invention relates to the field of vehicle state estimation, in particular to a vehicle running state estimation method.
Background
The real-time accurate acquisition of automobile state and parameter information is a necessary condition for controlling an automobile, the running state quantity of the automobile comprises a centroid slip angle, longitudinal, lateral and vertical speeds, the angular speeds of the automobile in three directions and the like, the state parameters can be acquired by installing different vehicle-mounted sensors, but the problems of high sensor price and difficult installation exist, so that a design state observer is generally adopted to estimate the running state of the automobile, and currently, a Luenberger (Drabert) observer, a synovial observer, a robust observer, a fuzzy observer and the like based on a control idea, and Kalman Filtering (KF), extended Kalman Filtering (EKF), unscented Kalman Filtering (UKF), particle Filtering (PF) and the like based on a Bayesian method are mainly adopted,
At present, the development of the distributed driving electric automobile is still in an early stage, and in the control of the distributed driving automobile, the estimation of the centroid slip angle and the yaw rate is particularly important, so that the present method is mainly KF, EKF and UKF and variants thereof as main estimation objects of the present invention.
When the existing filtering method is used for formal state estimation, sensor data are susceptible to noise, drift or errors, and the traditional dynamics model is applied to describe a complex vehicle motion model, so that limitation can exist, adaptability to the vehicle running in a rapidly-changing environment is insufficient, estimation accuracy is reduced, and the complex estimation algorithm can influence real-time performance of estimation due to high calculation cost.
An Extended Kalman Filtering (EKF) method can have the problem that an estimation result diverges under the condition of high nonlinearity, and solving a complex jacobian matrix can influence the real-time performance of state estimation; the Unscented Kalman Filtering (UKF) method has better performance under the condition of strong nonlinearity, but has the problems that the calculated amount is large, the posterior probability of the state variable does not meet Gaussian distribution, the precision is reduced, and the like; particle Filter (PF) methods require a large number of particles and samples to calculate the posterior probability density, and the calculation overhead of the estimation algorithm is large.
Disclosure of Invention
The invention provides a vehicle running state estimation method, which aims to solve the problems of insufficient adaptability and poor real-time performance of the existing vehicle running state estimation method.
The technical problems of the invention are solved by the following technical scheme:
a vehicle running state estimation method, comprising: acquiring vehicle operation data through a sensor; using the vehicle running data and the vehicle static parameters to establish a vehicle three-degree-of-freedom dynamics model comprising a system equation and an observation equation; and estimating the running state of the vehicle by a method combining extended Kalman filtering and a neural network based on the vehicle running data acquired by the vehicle three-degree-of-freedom dynamics model and the sensor, wherein the part for calculating the Kalman gain in the extended Kalman filtering is calculated by the neural network.
In some embodiments, the vehicle operation data includes a lateral vehicle speed of the vehicle, a vehicle centroid slip angle, a vehicle yaw rate, a vehicle front wheel corner, a vehicle yaw moment of inertia, a vehicle longitudinal acceleration, and a vehicle lateral acceleration; the vehicle static parameters include the total mass of the vehicle, the cornering stiffness of the front and rear axles of the vehicle, and the distance between the center of mass of the vehicle and the front and rear axles.
In some embodiments, the method for estimating the driving state of the vehicle by combining extended kalman filtering with the neural network specifically includes: (1) Initializing state estimation, covariance matrix, process noise covariance and measurement noise covariance of the extended Kalman filter; (2) Predicting the next state of the system according to the state transition equation; (3) Predicting a covariance matrix of the state estimation error according to the system model and the process noise; (4) Constructing a neural network for learning the Kalman gain, wherein the neural network approximates the optimal Kalman gain through training; (5) updating the state variables and calculating state variable estimated values; (6) updating a covariance matrix of the state estimation error; (7) The predicting and updating steps are repeated to continuously estimate the state of the system.
In some embodiments, in step (2), the next state of the system is predicted according to the following formula:
wherein,for the estimated value of the state variable at the next moment, x k-1 For the previous timeState variable, u k And f is a state transition equation.
In some embodiments, in step (3), a covariance matrix of state estimation errors is predicted according to the following formula:
Φ k ＝I+T s F
wherein,covariance matrix for state estimation error, Φ k For state transition matrix, Q k The method is characterized in that the method is a process noise covariance matrix, and I is an identity matrix with proper dimension; t (T) s Sampling time; f is a jacobian matrix of the state transfer function, and is obtained by taking the partial derivative of the state transfer function with respect to the independent variable.
In some embodiments, in step (4), the neural network uses a gating loop unit GRU to implicitly learn a second order statistical moment of the system, and uses an empirical MSE loss function and a random gradient descent algorithm to perform parameter optimization to achieve accurate estimation of the kalman gain.
In some embodiments, in step (4), the neural network comprises an input layer, an intermediate layer, and an output layer; the input layer is a full connection layer, and y is k Andas input, and maps it to a higher dimension, where the generated dimension is linearly proportional to the original input dimension, where y k For the observation at the current moment, +.>The state estimation value is the state estimation value of the previous moment; the middle layer is realized by GRU, which provides necessary capability for implicit learning of unknown second-order statistical moment to realize explicit learning of Kalman gain;the output layer is a fully connected layer that reconstructs the output to the dimension of the kalman gain.
In some embodiments, in step (4), the neural network is trained offline using tagged data, wherein the dataset is sorted into N tracks of length T, noted asWherein:
wherein Y is i An observation sequence representing the ith track,representing the observed value, X, at the kth time in the ith track i Status data sequence representing the ith track, < >>A state value indicating the kth time in the ith track.
In some embodiments, in step (4), an empirical MSE loss function is employed for each cycle:
wherein l i (Θ) is the loss function of the ith trace, ψ Θ (. Cndot.) is the output of the neural network, Θ is a trainable parameter of the neural network, and γ is a regularization coefficient; randomly selecting M pieces from N pieces of training data by adopting a mini-batch random gradient descent algorithm, and calculating a total loss function
The invention also provides a computer readable storage medium storing a computer program, which is characterized in that the computer program, when executed by a processor, realizes the vehicle running state estimation method.
Compared with the prior art, the invention has the beneficial effects that:
according to the invention, the method of combining the extended Kalman filtering and the neural network is adopted to estimate the running state of the vehicle, and compared with the traditional Kalman filtering algorithm and the extended Kalman filtering algorithm, the method has the advantages that under the condition of not improving the calculation requirement, better nonlinear processing capacity and more robust effect are obtained; compared with the unscented Kalman filtering algorithm, the method is easier to use, has wide applicability and higher calculation efficiency; compared with particle filtering, the method has the advantage of low time and space complexity, and is more suitable for vehicle-mounted running state estimation with high real-time requirements. The invention also calculates the Kalman gain in the extended Kalman filtering by the neural network, enhances the nonlinear processing capacity, robustness and adaptability of the extended Kalman filtering when being used for estimating the running state of the automobile, and ensures that the extended Kalman filtering is better suitable for the system dynamics and measurement noise which change in real time. Aiming at the distributed driving automobile in a dynamic environment with real-time change, the system dynamic and measurement noise can change in real time, and the invention can track the changes and adjust the Kalman gain of the extended Kalman filter in real time so as to adapt to the changes of the system dynamic and measurement noise. In the case of sensor faults or inaccurate measurement data provided by the sensor, the abnormal conditions can be better processed through the method and the device, and the robustness of the system is improved.
In the preferred scheme, the second-order statistical moment of the system is implicitly learned by using a gate control circulation unit GRU through a neural network, and the parameter optimization is carried out by adopting an empirical MSE loss function and a random gradient descent algorithm, so that the accurate estimation of the Kalman gain is realized.
Other advantages of embodiments of the present invention are further described below.
Drawings
Fig. 1 is a flowchart of a vehicle running state estimation method in an embodiment of the invention.
FIG. 2 is a schematic diagram of a three degree of freedom dynamics model of a vehicle in an embodiment of the invention.
FIG. 3 is a block diagram of an extended Kalman filter neural network algorithm in an embodiment of the invention.
Detailed Description
The invention will be further described with reference to the following drawings in conjunction with the preferred embodiments. It should be noted that, in the case of no conflict, the embodiments and features in the embodiments may be combined with each other.
It should be noted that, in this embodiment, the terms of left, right, upper, lower, top, bottom, etc. are merely relative terms, or refer to the normal use state of the product, and should not be considered as limiting.
In order to solve the problems of insufficient adaptability and poor real-time performance of the existing vehicle running state estimation method, the embodiment of the invention provides a distributed driving vehicle running state estimation method based on an extended Kalman filter, the extended Kalman filter and a neural network are combined by embedding the neural network in the extended Kalman filter, an extended Kalman filter-neural network EKF-NN method is provided, the running state estimation of the vehicle is carried out in the combined mode, the Kalman gain calculation part in the extended Kalman filter is obtained through the neural network calculation, the original Kalman gain calculation is replaced, and the Kalman gain is obtained through the data driving method. The robustness and adaptability of the Extended Kalman Filter (EKF) in the estimation of the running state of the automobile are enhanced, so that the system dynamic and measurement noise which changes in real time are better adapted. Aiming at the distributed driving automobile in a dynamic environment with real-time change, the system dynamic and measurement noise can change in real time, and the extended Kalman filtering-neural network EKF-NN method provided by the embodiment of the invention can track the changes, optimize and adjust the Kalman gain of Extended Kalman Filtering (EKF) in real time so as to adapt to the changes of the system dynamic and measurement noise. And under the condition that the sensor fails or the sensor provides inaccurate measurement data, the abnormal conditions can be better processed through the extended Kalman filter-neural network EKF-NN method, and the robustness of the system is improved.
The vehicle running state estimation method provided by the embodiment of the invention is shown in fig. 1, and comprises the following steps:
1. acquiring vehicle operation data through a sensor, wherein the vehicle operation data comprises vehicle longitudinal acceleration and vehicle lateral acceleration;
2. using vehicle operation data and vehicle static parameters, a vehicle three degree of freedom dynamics model is built, comprising system equations and observation equations, which model, as shown in fig. 2, considers only longitudinal, lateral and yaw movements, where δ is the vehicle front wheel rotation angle, a x 、a y Longitudinal and lateral acceleration, covariance matrix P, process noise covariance Q and measured noise covariance R, v, respectively x Is the lateral speed of the vehicle. Wherein the vehicle operation data includes a lateral speed v of the vehicle x Vehicle and its control method
Vehicle centroid slip angle beta, vehicle yaw rateVehicle front wheel steering angle delta, vehicle yaw moment of inertia I z Longitudinal acceleration a of vehicle x Lateral acceleration a of vehicle y The method comprises the steps of carrying out a first treatment on the surface of the The vehicle static parameters comprise the total mass m of the vehicle and the cornering stiffness k of the front axle and the rear axle of the vehicle f /k r And the distance a/b of the vehicle centroid to the front and rear axles.
The three-degree-of-freedom dynamics model of the vehicle built in the step is as follows:
the system equation:
is the lateral acceleration of the vehicle; />The vehicle mass center slip angular velocity; />Yaw acceleration for the vehicle;
observation equation:
converting the above formula into a form of a state space:
for the state transition matrix, the observer has a state variable of +.>Observer input variable u= [ delta, a x ] T The observed variable is y=a y 。
Wherein: v x is the lateral speed of the vehicle, beta is the vehicle centroid slip angle,for yaw rate, k of the vehicle f 、k r Respectively the cornering stiffness of the front and rear axles of the vehicle, m is the total mass of the vehicle, delta is the front wheel turning angle of the vehicle, I z For yaw moment of inertia of the vehicle, a x 、a y The longitudinal acceleration and the lateral acceleration of the vehicle are respectively, and the distance between the mass center of the vehicle and the front axle and the rear axle are respectively a and b.
3. And estimating the running state of the vehicle by a method combining extended Kalman filtering and a neural network based on the vehicle running data acquired by the vehicle three-degree-of-freedom dynamics model and the sensor, wherein the part for calculating the Kalman gain in the extended Kalman filtering is calculated by the neural network. The embodiment of the invention establishes a vehicle running state estimation algorithm based on an EKF-NN (extended Kalman filter-neural network) method; the principle is as follows:
an extended kalman filter algorithm (EKF) is an extension of kalman filtering for processing nonlinear systems. Unlike standard kalman, the extended kalman filter algorithm (EKF) deals with nonlinearities by linearizing the nonlinear portion of the system model at each instant. The basic steps include initialization, time updates and metrology updates. The time update predicts the state through a state transition equation and then calculates the predicted covariance. The measurement update calculates a predicted measurement value by using an observation equation, adjusts state estimation and covariance according to a measurement update formula, and introduces a special compact neural network for Kalman gain calculation to enhance the nonlinearity capability and robustness of the EKF, as shown in FIG. 3, and carries out running state estimation of a vehicle by a method of combining extended Kalman filtering and the neural network, wherein the specific algorithm flow is as follows:
(1) Initializing:
initializing state estimation of extended Kalman filtering algorithm (EKF)A covariance matrix P, a process noise covariance Q, and a measurement noise covariance R.
(2) State prediction: predicting the next state of the system according to the state transition equation; the next state of the system is predicted in particular according to the following formula:
wherein,for the estimated value of the state variable at the next moment, x k-1 As a state variable at the previous time, u k And f is a state transition equation.
(3) Error covariance prediction: predicting a covariance matrix of the state estimation error according to the system model and the process noise; specifically, a covariance matrix of the state estimation error is predicted according to the following formula:
Φ k ＝I+T s F
in the method, in the process of the invention,covariance matrix for state estimation error, Φ k For state transition matrix, Q k The method is characterized in that the method is a process noise covariance matrix, and I is an identity matrix with proper dimension; t (T) s Sampling time; f is a Jacobian matrix of the state transfer function, obtained by taking the partial derivative of the state transfer function with respect to the independent variable, i.e.>
(4) Obtaining Kalman filtering gain K by using neural network k : specifically, a neural network for learning the Kalman gain is constructed, and the neural network approximates the optimal Kalman gain through training; more specifically, embodiments of the present invention contemplate a neural network learning Kalman gain, although the Kalman gain is not dependent on the current observations y k Nor is it a previous estimateCan still be derived from the system equations and the observation equations in the three-degree-of-freedom dynamics model of the vehicle k Andthe trained portion through the neural network may be used as a kalman gain.
The calculation of the kalman gain in EKF involves a second order statistical moment, so the neural network employs gated loop units (GRUs) with internal memory elements, and in embodiments of the present invention, the neural network uses the second order statistical moment of the gated loop units GRU implicitly learning system, and performs parameter optimization using an empirical MSE loss function and a random gradient descent algorithm to achieve accurate estimation of the kalman gain. The neural network in the embodiment of the invention mainly comprises 3 main layers, namely an input layer, an intermediate layer and an output layer. The input layer is a full connection layer, which will y k Andas input, and maps it to a higher dimension, where the generated dimension is linearly proportional to the original input dimension, where y k For the observation at the current moment, +.>Is the state estimate at the previous time. The middle layer is implemented by a GRU (gated loop unit) which provides the necessary capability for implicit learning of unknown second order statistical moments to enable explicit learning of the kalman gain. The output layer is also implemented by a fully connected layer, which functions to reconstruct the output to the dimension of the kalman gain.
Specifically, the embodiment of the invention trains the neural network offline by using the labeled data, sorts the data set into N tracks with the length of T, and marks the tracks asWherein:
wherein Y is i An observation sequence representing the ith track,representing the observed value, X, at the kth time in the ith track i Status data sequence representing the ith track, < >>A state value indicating the kth time in the ith track.
The invention employs an empirical MSE (mean square error) loss function for each cycle:
wherein l i (Θ) is the loss function of the ith trace, ψ Θ (. Cndot.) is the output of the neural network, Θ is a parameter that the neural network can train, and γ is a regularization coefficient.
Randomly selecting M pieces from N training data by adopting mini-batch random gradient descent algorithm, and calculating total loss function
(5) Updating the state variables, i.e. calculating state variable estimates:
h represents the observation function.
(6) Updating a state estimation error covariance matrix:
H k jacobian moment representing an observation functionThe array of which is arranged in a row,is the predicted value of time k.
(7) The predicting and updating steps are repeated to continuously estimate the state of the system.
The embodiment of the invention also provides a computer readable storage medium which stores a computer program, and when the computer program is executed by a processor, the vehicle running state estimation method is realized.
In the embodiment of the invention, an estimation method combining the extended Kalman filtering and the neural network is adopted, and compared with the traditional Kalman filtering algorithm and the extended Kalman filtering algorithm, the method has the advantages that under the condition of not improving the calculation requirement, better nonlinear processing capacity and more robust effect are obtained; compared with the unscented Kalman filtering algorithm, the vehicle running state estimation method provided by the embodiment of the invention is easier to use, has wide applicability and higher calculation efficiency; compared with particle filtering, the vehicle running state estimation method provided by the embodiment of the invention has the advantages of low time and space complexity, and is more suitable for vehicle running state estimation with high real-time requirement.
The foregoing is a further detailed description of the invention in connection with the preferred embodiments, and it is not intended that the invention be limited to the specific embodiments described. It will be apparent to those skilled in the art that several equivalent substitutions and obvious modifications can be made without departing from the spirit of the invention, and the same should be considered to be within the scope of the invention.
Claims (10)
1. A vehicle running state estimation method, characterized by comprising:
acquiring vehicle operation data through a sensor;
using the vehicle running data and the vehicle static parameters to establish a vehicle three-degree-of-freedom dynamics model comprising a system equation and an observation equation;
and estimating the running state of the vehicle by a method combining extended Kalman filtering and a neural network based on the vehicle running data acquired by the vehicle three-degree-of-freedom dynamics model and the sensor, wherein the part for calculating the Kalman gain in the extended Kalman filtering is calculated by the neural network.
2. The vehicle running state estimation method according to claim 1, wherein the vehicle running data includes a lateral vehicle speed of the vehicle, a vehicle centroid slip angle, a vehicle yaw rate, a vehicle front wheel turning angle, a vehicle yaw moment of inertia, a vehicle longitudinal acceleration, and a vehicle lateral acceleration; the vehicle static parameters include the total mass of the vehicle, the cornering stiffness of the front and rear axles of the vehicle, and the distance between the center of mass of the vehicle and the front and rear axles.
3. The vehicle running state estimation method according to claim 1 or 2, characterized in that the vehicle running state estimation by combining the extended kalman filter with the neural network specifically includes:
(1) Initializing state estimation, covariance matrix, process noise covariance and measurement noise covariance of the extended Kalman filter;
(2) Predicting the next state of the system according to the state transition equation;
(3) Predicting a covariance matrix of the state estimation error according to the system model and the process noise;
(4) Constructing a neural network for learning the Kalman gain, wherein the neural network approximates the optimal Kalman gain through training;
(5) Updating the state variable, and calculating a state variable estimated value;
(6) Updating a covariance matrix of the state estimation error;
(7) The predicting and updating steps are repeated to continuously estimate the state of the system.
4. A vehicle running state estimation method according to claim 3, wherein in step (2), the next state of the system is predicted according to the following formula:
wherein,for the estimated value of the state variable at the next moment, x k-1 As a state variable at the previous time, u k And f is a state transition equation.
5. The vehicle running state estimation method according to claim 3, wherein in step (3), a covariance matrix of the state estimation error is predicted according to the following formula:
Φ k ＝I+T s F
wherein,covariance matrix for state estimation error, Φ k For state transition matrix, Q k The method is characterized in that the method is a process noise covariance matrix, and I is an identity matrix with proper dimension; t (T) s Sampling time; f is a jacobian matrix of the state transfer function, and is obtained by taking the partial derivative of the state transfer function with respect to the independent variable.
6. The vehicle driving state estimation method according to any one of claims 3 to 5, wherein in the step (4), the neural network uses a second-order statistical moment of a gate-control loop unit GRU implicit learning system, and uses an empirical MSE loss function and a random gradient descent algorithm to perform parameter optimization so as to achieve accurate estimation of a kalman gain.
7. The vehicle running state estimation method according to claim 6The method is characterized in that in the step (4), the neural network comprises an input layer, an intermediate layer and an output layer; the input layer is a full connection layer, and y is k Andas input, and maps it to a higher dimension, where the generated dimension is linearly proportional to the original input dimension, where y k For the observation at the current moment, +.>The state estimation value is the state estimation value of the previous moment; the middle layer is realized by GRU, which provides necessary capability for implicit learning of unknown second-order statistical moment to realize explicit learning of Kalman gain; the output layer is a fully connected layer that reconstructs the output to the dimension of the kalman gain.
8. The vehicle running state estimation method according to any one of claims 3 to 7, wherein in step (4), the neural network is trained offline using tagged data, wherein the data set is sorted into N tracks of length T, noted asWherein:
wherein Y is i An observation sequence representing the ith track,representing the observed value, X, at the kth time in the ith track i Status data sequence representing the ith track, < >>A state value indicating the kth time in the ith track.
9. The vehicle running state estimation method according to claim 8, wherein in step (4), an empirical MSE loss function is employed for each cycle:
wherein l i (Θ) is the loss function of the ith trace, ψ Θ (. Cndot.) is the output of the neural network, Θ is a trainable parameter of the neural network, and γ is a regularization coefficient;
randomly selecting M pieces from N pieces of training data by adopting a mini-batch random gradient descent algorithm, and calculating a total loss function
10. A computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the vehicle running state estimation method according to any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410126243.3A CN117719519A (en) | 2024-01-29 | 2024-01-29 | Vehicle running state estimation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410126243.3A CN117719519A (en) | 2024-01-29 | 2024-01-29 | Vehicle running state estimation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117719519A true CN117719519A (en) | 2024-03-19 |
Family
ID=90203767
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410126243.3A Pending CN117719519A (en) | 2024-01-29 | 2024-01-29 | Vehicle running state estimation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117719519A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117932234A (en) * | 2024-03-25 | 2024-04-26 | 苏州观瑞汽车技术有限公司 | Data processing method and system for manufacturing brake calibration table |
-
2024
- 2024-01-29 CN CN202410126243.3A patent/CN117719519A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117932234A (en) * | 2024-03-25 | 2024-04-26 | 苏州观瑞汽车技术有限公司 | Data processing method and system for manufacturing brake calibration table |
| CN117932234B (en) * | 2024-03-25 | 2024-06-07 | 苏州观瑞汽车技术有限公司 | Data processing method and system for manufacturing brake calibration table |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109885883B (en) | Unmanned vehicle transverse motion control method based on GK clustering algorithm model prediction | |
| Singh et al. | Literature review and fundamental approaches for vehicle and tire state estimation | |
| CN112758097B (en) | State prediction and estimation method for unmanned vehicle | |
| CN117719519A (en) | Vehicle running state estimation method | |
| CN109606378A (en) | Vehicle running state estimation method towards non-Gaussian noise environment | |
| CN111159642B (en) | Online track prediction method based on particle filtering | |
| CN113002549A (en) | Vehicle state estimation method, device, equipment and storage medium | |
| CN113806861B (en) | Method for estimating vehicle mass and road gradient of two-step structure | |
| CN108931233B (en) | Road side slope value detection method and device | |
| CN111959516B (en) | Method for jointly estimating vehicle state and road adhesion coefficient | |
| CN115406446A (en) | Multi-axis special vehicle state estimation method based on neural network and unscented Kalman filtering | |
| CN112298354A (en) | State estimation method for steering wheel and front wheel corner of steering system of unmanned automobile | |
| CN113447021A (en) | MEMS inertial navigation system positioning enhancement method based on LSTM neural network model | |
| CN110687907A (en) | Intelligent automobile decoupling prediction controller based on model dynamic reconstruction and control method thereof | |
| CN112270039A (en) | Distributed asynchronous fusion-based nonlinear state estimation method for drive-by-wire chassis vehicle | |
| CN111942399A (en) | Vehicle speed estimation method and system based on unscented Kalman filtering | |
| CN113771857A (en) | Longitudinal speed estimation method and system for vehicle control | |
| Sieberg et al. | Hybrid state estimation combining artificial neural network and physical model | |
| CN116674571A (en) | Real-time estimation method for automobile quality and gradient based on data confidence factor | |
| CN114154231A (en) | Vehicle driving state estimation system and method based on deep reinforcement learning | |
| CN114590264A (en) | Pavement adhesion coefficient estimation method based on deep integration network adaptive Kalman filtering | |
| Yuhao | Estimation of Vehicle Status and Parameters Based on Nonlinear Kalman Filtering | |
| Da Lio et al. | Robust and sample-efficient estimation of vehicle lateral velocity using neural networks with explainable structure informed by kinematic principles | |
| CN111703429A (en) | Method for estimating longitudinal speed of wheel hub motor driven vehicle | |
| Gräber et al. | Application of Neural Networks to External Parameter Estimation for Nonlinear Vehicle Models |
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 |