CN-122009189-A - Automatic parking road bank identification method and related equipment

CN122009189ACN 122009189 ACN122009189 ACN 122009189ACN-122009189-A

Abstract

The invention discloses a road bank identification method for automatic parking and related equipment, relates to the technical field of automatic control and intelligent perception of vehicles, and mainly aims to solve the problem that the accuracy of the existing automatic parking road bank identification method is low. The method comprises the steps of triggering a physical verification decision process under the condition that the confidence coefficient of a first road bank height estimated value is lower than a preset confidence coefficient, wherein the first road bank height estimated value is obtained based on ultrasonic waves and/or visual sensors, the physical verification decision process is used for driving wheels to be in physical contact with a road bank, calculating the confidence coefficient of a second road bank height estimated value and the confidence coefficient of the second road bank height estimated value based on physical characteristics obtained after the wheels are in physical contact with the road bank, and respectively carrying out weighted calculation on the first road bank height estimated value and the second road bank height estimated value based on the confidence coefficient of the first road bank height estimated value and the confidence coefficient of the second road bank height estimated value so as to obtain a target road bank height estimated value.

Inventors

YANG GENG

Assignees

岚图汽车科技股份有限公司

Dates

Publication Date: 20260512
Application Date: 20260310

Claims (10)

1. The method for identifying the road bank for automatic parking is characterized by comprising the following steps of: Triggering a physical verification decision process under the condition that the confidence coefficient of the first road bank height estimation value is lower than a preset confidence coefficient, wherein the first road bank height estimation value is acquired based on ultrasonic waves and/or visual sensors, and the physical verification decision process is used for driving wheels to be in physical contact with the road bank; Calculating a second road bank height estimated value and a confidence coefficient of the second road bank height estimated value based on physical characteristics obtained after the wheels are in physical contact with the road bank; And respectively carrying out weighted calculation on the first road bank height estimated value and the second road bank height estimated value based on the confidence coefficient of the first road bank height estimated value and the confidence coefficient of the second road bank height estimated value so as to obtain a target road bank height estimated value.
2. The method of claim 1, wherein triggering the physical verification decision process if the confidence level of the first road ridge height estimate is below a preset confidence level comprises: Acquiring the first ridge height estimation value and the confidence coefficient of the first ridge height estimation value based on an ultrasonic wave and/or a visual sensor; Triggering a physical verification decision flow under the condition that the confidence coefficient of the first road bank height estimation value is lower than a preset confidence coefficient; And under the condition that the physical verification decision flow is triggered, controlling the vehicle to move close to the road bank at a preset speed so as to enable physical contact between the wheels and the road bank to occur.
3. The method of claim 2, wherein calculating the confidence of the second road bank height estimate and the second road bank height estimate based on the physical characteristics obtained after the wheel is in physical contact with the road bank comprises: Activating a vehicle sensor under the condition of triggering the physical verification decision flow so as to acquire an original time domain signal in the process that the vehicle moves at a preset speed to approach a road bank; Extracting time domain features and frequency domain features of the original time domain signals to form a physical feature vector group, wherein the time domain features are used for reflecting physical impact strength and motion states, and the frequency domain features are used for representing different physical modes; And inputting the physical feature vector group into a prediction model to output the second road bank height estimated value and the confidence degree of the second road bank height estimated value.
4. The method of claim 3, wherein the raw time domain signals include a body inertia measurement signal, a wheel speed pulse signal, and a drive system signal, The vehicle body inertia measurement signal is used for reflecting the linear acceleration and the angular velocity of the three-dimensional space of the vehicle, The wheel speed pulse signal is used for reflecting the rotation speed change of the wheels, The driving system signal is used for reflecting the output state of the driving system.
5. A method according to claim 3, wherein said extracting the time domain features and frequency domain features of the original time domain signal to form a set of physical feature vectors comprises: Extracting time domain features of the original time domain signal to determine each axial acceleration peak value, each axial angular velocity and a vehicle dynamic parameter, wherein, The axial acceleration peaks reflect the impact strength in all directions at the moment of physical contact of the wheel with the road ridge, Each axial angular velocity reflects the change of the vehicle body posture caused by the physical contact between the wheels and the road ridge, The vehicle dynamic parameters reflect initial motion conditions of the physical contact between the wheels and the road ridge, including the vehicle approach speed and the vehicle-road ridge included angle; Extracting frequency domain features of the original time domain signal to determine vertical vibration energy features and velocity fluctuation energy features, wherein, The vertical vibration energy characteristics reflect the overall lifting motion mode of the vehicle body, the transient impact of the tire and the edge of the road bank, and the high-frequency vibration of the suspension system caused by the physical contact between the wheel and the road bank, The speed fluctuation energy characteristics reflect the overall change trend of the vehicle speed before and after the wheels are in physical contact with the road ridge and the slip or torque fluctuation at the moment of the physical contact between the wheels and the road ridge.
6. The method of claim 3, wherein the step of, The input layer of the predictive model is used for inputting the set of physical feature vectors, The structure layer comprises an LSTM layer and an FCN layer, wherein the LSTM layer is used for modeling time sequence characteristics of the physical characteristic vector group in a time window before and after a physical contact event, and the FCN layer is used for fusing the time sequence characteristics output by the LSTM layer; The output layer is used for outputting the second road bank height estimated value and the confidence coefficient of the second road bank height estimated value.
7. The method of claim 1, wherein the weighting the first and second ridge height estimates based on the confidence level of the first and second ridge height estimates, respectively, to obtain a target ridge height estimate comprises: determining a first weight based on a square value of a confidence level of the first road bank height estimation value; Determining a second weight based on a square value of a confidence level of the second road ridge height estimation value; Determining a first weighted value based on a product of the first weight and the first ridge height estimate; Determining a second weighting value based on a product of the second weighting and the second road ridge height estimation value; obtaining a sum of weighted values added by the first weighted value and the second weighted value; acquiring the sum of the weights added by the first weight and the second weight; and determining the target road ridge height estimation value based on the ratio of the weighted value sum to the weighted value sum.
8. The utility model provides an automatic road bank recognition device parks which characterized in that still includes: the triggering unit is used for triggering a physical verification decision process under the condition that the confidence coefficient of the first road bank height estimated value is lower than a preset confidence coefficient, wherein the first road bank height estimated value is acquired based on ultrasonic waves and/or visual sensors, and the physical verification decision process is used for driving wheels to be in physical contact with the road bank; The calculating unit is used for calculating a second road ridge height estimated value and the confidence coefficient of the second road ridge height estimated value based on the physical characteristics obtained after the wheels are in physical contact with the road ridge; The obtaining unit is used for respectively carrying out weighted calculation on the first road bank height estimated value and the second road bank height estimated value based on the confidence coefficient of the first road bank height estimated value and the confidence coefficient of the second road bank height estimated value so as to obtain a target road bank height estimated value.
9. A computer readable storage medium, characterized in that the computer readable storage medium comprises a stored program, wherein the steps of the sill identifying method of automatic parking according to any of claims 1 to 7 are implemented when the program is executed by a processor.
10. An electronic device comprising at least one processor and at least one memory coupled to the processor, wherein the processor is configured to invoke program instructions in the memory to perform the steps of the method for identifying a threshold for automatic parking as claimed in any of claims 1 to 7.

Description

Automatic parking road bank identification method and related equipment Technical Field The invention relates to the technical field of automatic control and intelligent perception of vehicles, in particular to a road bank identification method for automatic parking and related equipment. Background In the "released parking space" scenario of automatic parking, the vehicle needs to autonomously cross the road bank, and the core premise is to accurately identify the height of the road bank so as to decide whether to pass and plan the control strategy. The existing automatic parking threshold identification method mainly relies on non-contact sensors such as ultrasonic radars and visual cameras to carry out indirect measurement and fusion estimation, but sound waves can generate serious scattering, the visual sensors are extremely easily affected by illumination conditions, and the technical routes have inherent defects which are difficult to overcome, so that the overall identification accuracy is low. Although the industry performs compensation optimization by adopting algorithms such as multi-sensor data fusion and complex geometric model calculation, the confidence of the output result still decreases remarkably under the adverse scene, and reliable basis cannot be provided for vehicle control. Therefore, the road bank height information with high precision and high confidence cannot be provided stably in all-weather and full-scene conditions, and the reliability and safety of the automatic parking function in a complex real environment are restricted. Disclosure of Invention In view of the above problems, the present invention provides a method for identifying a road bank for automatic parking and related equipment, and mainly aims to solve the problem of lower accuracy of the existing method for identifying a road bank for automatic parking. In order to solve at least one of the above technical problems, in a first aspect, the present invention provides a road bank identification method for automatic parking, the method comprising: Triggering a physical verification decision process under the condition that the confidence coefficient of the first road bank height estimation value is lower than a preset confidence coefficient, wherein the first road bank height estimation value is acquired based on ultrasonic waves and/or visual sensors, and the physical verification decision process is used for driving wheels to be in physical contact with the road bank; Calculating a second road bank height estimated value and a confidence coefficient of the second road bank height estimated value based on physical characteristics obtained after the wheels are in physical contact with the road bank; And respectively carrying out weighted calculation on the first road bank height estimated value and the second road bank height estimated value based on the confidence coefficient of the first road bank height estimated value and the confidence coefficient of the second road bank height estimated value so as to obtain a target road bank height estimated value. Optionally, triggering the physical verification decision process when the confidence level of the first road ridge height estimation value is lower than a preset confidence level includes: Acquiring the first ridge height estimation value and the confidence coefficient of the first ridge height estimation value based on an ultrasonic wave and/or a visual sensor; Triggering a physical verification decision flow under the condition that the confidence coefficient of the first road bank height estimation value is lower than a preset confidence coefficient; And under the condition that the physical verification decision flow is triggered, controlling the vehicle to move close to the road bank at a preset speed so as to enable physical contact between the wheels and the road bank to occur. Optionally, the calculating the second road bank height estimation value and the confidence coefficient of the second road bank height estimation value based on the physical feature acquired after the wheel is in physical contact with the road bank includes: Activating a vehicle sensor under the condition of triggering the physical verification decision flow so as to acquire an original time domain signal in the process that the vehicle moves at a preset speed to approach a road bank; Extracting time domain features and frequency domain features of the original time domain signals to form a physical feature vector group, wherein the time domain features are used for reflecting physical impact strength and motion states, and the frequency domain features are used for representing different physical modes; And inputting the physical feature vector group into a prediction model to output the second road bank height estimated value and the confidence degree of the second road bank height estimated value. Optionally, the raw time domain signals include body inertia measurement signals, whee