Vehicle door anti-pinch detection method and system for vehicle and vehicle

Abstract

The invention provides a vehicle door anti-pinch detection method for a vehicle. The vehicle door anti-pinch detection method comprises the following steps: establishing a vehicle door pinch detection calculation model; echo signals in the door frame range of the vehicle door are collected in real time; calculating signal characteristics of the echo signal according to the echo signal; and substituting the signal characteristics into the vehicle door hand pinching detection calculation model, and determining a hand pinching detection result of the vehicle door. According to the technical scheme, the sensing capacity of the micro-motion target is higher, the characteristics of the micro-motion of the hand can be accurately recognized, whether the situation that the hand is pinched by the automobile door exists or not is judged in real time, the problem that in the prior art, the hand at the position of the automobile door frame cannot be detected is solved, the existing automobile function is expanded, and the automobile use safety is improved.

Description

Technical Field

[0001] The present invention relates to the field of vehicles, and particularly to a method and system for detecting door pinch prevention for vehicles, and a vehicle. Background Art

[0002] The existing door pinch prevention technical solutions mainly include foreign object detection technology before closing the door and door pinch prevention technology. For a hinged car door, the existing foreign object detection system uses a millimeter-wave radar and an angle sensor, and can only detect foreign objects at obvious positions of the door, and it is not easy to detect the situation where a human hand is on the door frame. For a sliding car door, the existing car door pinch prevention mainly adopts the method of motor current pinch prevention, which has a weak adaptability to different door mechanical structures, is prone to false pinch prevention operations, and changes with the increase of the motor operation time, affecting the reliability of pinch prevention. Summary of the Invention

[0003] In order to improve the detection accuracy of foreign objects at the vehicle door and improve the reliability of the pinch prevention detection result, the purpose of the present invention is to provide a method for detecting door pinch prevention for vehicles.

[0004] Specifically, the method for detecting door pinch prevention for vehicles in the present invention includes:

[0005] Establish a door pinching detection calculation model to determine the correspondence between signal characteristics and pinch detection results;

[0006] Collect echo signals in real time within the range of the door frame of the door;

[0007] Calculate the signal characteristics of the echo signals according to the echo signals;

[0008] Substitute the signal characteristics into the door pinching detection calculation model to determine the pinch detection result of the door.

[0009] Preferably, the establishment of the door pinching detection calculation model includes:

[0010] Collect signals: respectively and continuously collect the first echo signals and the second echo signals when there is a target object and when there is no target object within the range of the door frame;

[0011] Form training set data: calculate the first signal feature set {D10, A10, E i10} of the first echo signals, and set the label of the first signal feature set to 1; calculate the second signal feature set {D20, A20, E i20} of the second echo signals, and set the label of the second signal feature set to 0. The labeled first signal feature set and second signal feature set are training set data;

[0012] Training using machine learning methods: Divide the training set data into Z subsets, where (Z - 1) subsets are used as the training set to train an SVM (Support Vector Machine) classifier; the remaining 1 subset is used as the test set to perform classification testing on the SVM classifier;

[0013] Evaluate the performance of the door pinching detection calculation model according to the classification test results;

[0014] Determine the correspondence between the signal features and the pinching detection results. When the output result of the door pinching detection calculation model is 1, it is determined that there is a target object within the door frame range of the door; when the output result of the door pinching detection calculation model is 0, it is determined that there is no target object within the door frame range of the door.

[0015] Preferably, the evaluating the performance of the door pinching detection calculation model according to the classification test results includes:

[0016] Calculate the accuracy Ac, precision Pr, and recall Re of the door pinching detection calculation model;

[0017] Evaluate the performance of the door pinching detection calculation model according to the accuracy Ac, precision Pr, and recall Re.

[0018] Preferably, the third echo signal within the door frame range of the door is collected in real time, and according to the third echo signal, the third signal feature set {D30, A30, E i30} of the third echo signal is calculated.

[0019] Preferably, calculating the signal feature set {Da0, Aa0, E ia0} of the echo signal, where a takes values of 1, 2, and 3 respectively, includes:

[0020] Perform two-dimensional FFT transformation on the echo signal:

[0021]

[0022] where x(n, m) is the echo signal, N is the number of FFT points in the range dimension, and M is the number of FFT points in the velocity dimension;

[0023] Perform constant false alarm rate processing on the two-dimensional range-Doppler matrix X(k, l) to obtain the range bin rbin and velocity bin vbin where the target object is located; calculate the range information Da0 of the target object according to formula (2) as the first feature of the signal feature set;

[0024] Da0 = rbin * c / 2B Formula (2)

[0025] Among them, c is the speed of light and B is the radar bandwidth;

[0026] Perform an FFT transformation on the two-dimensional range-Doppler matrix X(k, l) in the antenna dimension:

[0027]

[0028] Among them, S is the number of FFT points in the antenna dimension;

[0029] Calculate the angle information Aa0 of the target object, which is the second feature of the signal feature set;

[0030] Aa0 = asin(f w *λ / d)*180 / π Formula (4)

[0031] λ = c / f

[0032] d = λ / 2

[0033] Among them, f w is the frequency corresponding to the maximum value of X(k, l, v), c is the speed of light, and f is the radar carrier frequency;

[0034] Perform time-frequency analysis on the data of ten range cells near the range cell, calculate the short-time Fourier transform, and obtain the micro-Doppler spectrum data:

[0035]

[0036] Among them, x(m) is the echo signal, ω(m) is the window function, and i is the time index;

[0037] Continue to window the echo signal along the time of the echo signal, obtain the situation of the signal frequency changing with time, and convert the first echo signal into a Doppler time-signal matrix;

[0038] According to the two-dimensional Doppler time-signal matrix, obtain the instantaneous frequency corresponding to each range cell, and form a ten-dimensional feature with the signal energy values corresponding to the instantaneous frequencies, which is the third feature E of the signal feature set ia0 .

[0039] Preferably, bring the third signal feature into the door pinching detection calculation model. When the output result of the door pinching detection calculation model is 1, it is determined that there is a target object within the door frame range. When the output result of the door pinching detection calculation model is 0, it is determined that there is no target object within the door frame range.

[0040] Preferably, the echo signal is the echo signal of a millimeter-wave radar signal.

[0041] On the other hand, the present invention provides a door anti-pinch detection system for a vehicle, which applies the door anti-pinch detection method for a vehicle described in any one of the above.

[0042] Preferably, it includes a millimeter-wave radar detection module, a communication module, and an alarm module;

[0043] The millimeter-wave radar detection module includes a millimeter-wave radar and a calculation unit. The millimeter-wave radar is used to transmit millimeter-wave radar signals and collect echo signals within the range of the door frame of the door in real time. The calculation unit is used to calculate the signal characteristics of the echo signals according to the echo signals, and bring the signal characteristics into the door pinching detection calculation model to determine the pinching detection result of the door;

[0044] The communication module is used to send the pinching detection result to the alarm module. When there is a target object within the range of the door frame of the door, the alarm module is used to issue an alarm.

[0045] The present invention also discloses a vehicle equipped with the door anti-pinch detection system for a vehicle described in any one of the above.

[0046] After adopting the above technical solution, compared with the prior art, the technical solution of the present invention can make full use of the relatively wide Doppler bandwidth of the millimeter-wave radar, which has stronger perception ability for micro-moving targets and can more accurately identify the small movements of the hand. The characteristics are used to extract the characteristics of the radar echo signal, and the support vector machine (SVM) is used to train the extracted characteristics to obtain the door pinching detection algorithm model. Using this model, it can judge in real time whether there is a situation of the door pinching a hand, and can report the detection result to the user in time, so that the user can adjust the driving state and vehicle state in time. It solves the problem that the human hand at the position of the car door frame cannot be detected in the prior art, expands the existing functions of the car, and improves the safety of car use. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a flowchart of the door anti-pinch detection method for a vehicle according to an embodiment of the present invention;

[0048] Figure 2 It is a flowchart of another door anti-pinch detection method for a vehicle according to an embodiment of the present invention;

[0049] Figure 3 For Figure 2 The flowchart of establishing the door pinching detection calculation model in the shown embodiment;

[0050] Figure 4 It is a flowchart of another door anti-pinch detection method for a vehicle according to an embodiment of the present invention;

[0051] Figure 5It is a schematic diagram of a door anti-pinch detection system for a vehicle according to an embodiment of the present invention;

[0052] Figure 6 is Figure 5 a schematic installation diagram of the millimeter-wave radar in the illustrated embodiment. Specific embodiments

[0053] The advantages of the present invention will be further elaborated below in conjunction with the accompanying drawings and specific embodiments.

[0054] Here, exemplary embodiments will be described in detail, and examples thereof are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0055] The terms used in the present disclosure are for the purpose of describing specific embodiments only and are not intended to limit the present disclosure. The singular forms "a", "the", and "said" used in the present disclosure and the appended claims are also intended to include the plural forms unless the context clearly dictates otherwise. It should also be understood that the term "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.

[0056] In the description of the present invention, unless otherwise specified and defined, it should be noted that the terms "installed", "connected", and "coupled" should be understood in a broad sense. For example, they can be mechanical connections or electrical connections, or the communication inside two components. They can be directly connected or indirectly connected through an intermediate medium. For those of ordinary skill in the art, the specific meanings of the above terms can be understood according to specific circumstances.

[0057] An embodiment of the present invention provides a method for detecting door anti-pinch for a vehicle, as Figure 1 shown, including:

[0058] Establishing a calculation model for detecting door pinching;

[0059] Real-time collecting echo signals within the door frame range of the door;

[0060] Calculating signal characteristics of the echo signals according to the echo signals;

[0061] Bringing the signal characteristics into the calculation model for detecting door pinching to determine the detection result of door pinching.

[0062] In this embodiment, an in-vehicle-door pinching detection calculation model is established to detect the echo signal within the range of the vehicle door frame, so as to verify whether there is an object within the door frame. This can effectively avoid the situation in the prior art where the vehicle door frame cannot be detected, and at the same time avoid the method of preventing pinching through the motor current, improving the reliability of the anti-pinching detection result.

[0063] Figure 3 The following is a flowchart for establishing an in-vehicle-door pinching detection calculation model in another embodiment of the present invention, including:

[0064] Signal acquisition: Continuously acquire the first echo signal and the second echo signal respectively in the case where there is a target object and there is no target object within the range of the door frame. The target object can be various objects that may appear within the vehicle door frame, such as a human hand, etc.

[0065] Forming training set data: Calculate the first signal feature set {D10, A10, E i10} of the first echo signal, and set the label of the first signal feature set to 1; calculate the second signal feature set {D20, A20, E i20} of the second echo signal, and set the label of the second signal feature set to 0. The labeled first signal feature set and second signal feature set are the training set data.

[0066] Training using machine learning methods: Divide the training set data into 10 subsets, where 9 subsets are used as the training set to train the SVM classifier; the remaining 1 subset is the test set to perform classification testing on the SVM classifier. As is well known in the art, SVM (Support Vector Machine) is a machine learning method for classification and regression. By finding the optimal hyperplane, it can complete the task of classifying data. SVM aims to maximize the margin. After mapping the original data to a high-dimensional space through the kernel function, it finds the maximum margin hyperplane, thus being able to effectively solve the problems of linearly separable and linearly inseparable data. In this embodiment, the training set data formed by the signal feature sets of the first echo signal and the second echo signal is sent to the SVM classifier for training. During the training process, the ten-fold cross-validation method is adopted.

[0067] According to the classification test results, evaluate the performance of the in-vehicle-door pinching detection calculation model. The evaluation results can provide a reference for selecting the parameters of the SVM classifier. Common evaluation parameters in the art include accuracy Ac, precision Pr, and recall Re, etc. Among them, the accuracy Ac of the model is:

[0068] Ac = (TP + TN) / (TP + FP + TN + FN)

[0069] The precision Pr of the model is:

[0070] Pr = TP / (TP + FP)

[0071] The recall rate Re is as follows:

[0072] Re = TP / (TP + FN)

[0073] Wherein, TP indicates that there is a target object at the car door frame and the model judges that there is a target object; TN indicates that there is no target object at the door frame and the model judges that there is no target object; FP indicates that there is a target object at the door frame but the model judges that there is no target object; FN indicates that there is no target object within the door frame range but is judged to be a situation where there is a target object.

[0074] Through the above cross - validation method, the performance of the car door pinching detection calculation model can be improved, and a car door pinching detection calculation model with accurate judgment can be obtained.

[0075] Figure 2 The following is a flowchart of the car door anti - pinch detection method using the above - established car door pinching detection calculation model. It includes:

[0076] Establish a car door pinching detection calculation model;

[0077] Collect the third echo signal within the door frame range of the car door in real - time, and calculate the third signal feature set {D30, A30, E i30} of the third echo signal according to the third echo signal;

[0078] Bring the third signal feature into the car door pinching detection calculation model. When the output result of the car door pinching detection calculation model is 1, it is determined that there is a target object within the door frame range of the car door; when the output result of the car door pinching detection calculation model is 0, it is determined that there is no target object within the door frame range of the car door.

[0079] In the present invention, the signal feature set of the echo signal is an important basis for establishing the model and judging whether there is a target object. The following is the calculation process for calculating the signal feature set of the echo signal:

[0080] Calculate the signal feature set {Da0, Aa0, E ia0}, where a takes values of 1, 2, and 3 respectively. That is, the signal feature sets corresponding to the first echo signal, the second echo signal, and the third echo signal collected in real - time within the door frame range of the car door involved in establishing the car door pinching detection calculation model are all calculated using the following calculation steps.

[0081] Specifically, it includes: performing two - dimensional FFT transformation on the echo signal:

[0082]

[0083] Wherein, x(n, m) is the echo signal, N is the number of FFT points in the range dimension, and M is the number of FFT points in the velocity dimension;

[0084] Perform a constant false alarm rate (CFAR) process on the two-dimensional range-Doppler matrix X(k, l) to obtain the range bin rbin and velocity bin vbin where the target object is located; calculate the range information Da0 of the target object according to formula (2), which is the first feature of the signal feature set;

[0085] Da0 = rbin * c / 2B Formula (2)

[0086] Wherein, c is the speed of light and B is the radar bandwidth;

[0087] Perform an FFT transform on the two-dimensional range-Doppler matrix X(k, l) in the antenna dimension:

[0088]

[0089] Wherein, S is the number of FFT points in the antenna dimension;

[0090] Calculate the angle information Aa0 of the target object, which is the second feature of the signal feature set;

[0091] Aa0 = asin(f w * λ / d) * 180 / π Formula (4)

[0092]

[0093] d = λ / 2

[0094] Wherein, f w is the frequency corresponding to the maximum value of X(k, l, v), c is the speed of light, and f is the radar carrier frequency;

[0095] Perform time-frequency analysis on the data of ten range bins near the range bin, calculate the short-time Fourier transform, and obtain the micro-Doppler spectrum data:

[0096]

[0097] Wherein, x(m) is the echo signal, ω(m) is the window function, and i is the time index;

[0098] Continue to window the echo signal along the time of the echo signal to obtain the variation of the signal frequency with time, and convert the first echo signal into a Doppler time-signal matrix;

[0099] According to the two-dimensional Doppler time-signal matrix, the instantaneous frequency corresponding to each range cell is obtained, and the signal energy values corresponding to the instantaneous frequencies are combined into ten-dimensional features, which are the third feature E of the signal feature set. ia0 .

[0100] In another preferred embodiment, as Figure 4 shown, in this embodiment, after the vehicle is powered on, it is first determined whether the vehicle door and the pinch detection function of the vehicle are turned on. When it is determined that there is a target object (i.e., the human hand in this embodiment) within the door frame range of the vehicle door, the alarm mode is immediately activated, and it is further detected whether the alarm mode is manually turned off. If not, the determination result that there is a target object within the door frame range of the vehicle door is sent to the actuator to adjust the current state of the vehicle. The alarm mode is any common mode in the art, including but not limited to the flashing of a warning light, the sounding of a warning tone, etc., to remind the passengers in the vehicle of the abnormality and adjust the vehicle state accordingly. The present invention does not make any restrictions here.

[0101] It should be noted that the echo signal in the present invention is the echo signal of the signal emitted by the vehicle-mounted millimeter-wave radar. The technical solutions in the above embodiments can be implemented without additionally installing other electronic components.

[0102] In another embodiment of the present invention, a door anti-pinch detection system for a vehicle is provided, which uses the door anti-pinch detection method for a vehicle in the previous embodiment.

[0103] In another preferred embodiment, as Figure 5 shown, the door anti-pinch detection system for a vehicle specifically includes a millimeter-wave radar detection module, a communication module, and an alarm module;

[0104] The millimeter-wave radar detection module includes a millimeter-wave radar and a calculation unit. The millimeter-wave radar is used to emit millimeter-wave radar signals and real-time collect echo signals within the door frame range of the vehicle door. The calculation unit is used to calculate the signal features of the echo signals according to the echo signals, and bring the signal features into the door pinch detection calculation model to determine the pinch detection result of the vehicle door. In addition, the millimeter-wave radar detection module in this embodiment can detect in real time whether there is a target object within the door frame range of the vehicle door. In practical applications, this millimeter-wave radar detection module can be applied to other scenarios, such as detecting whether there are abnormal objects around the vehicle, not limited to detecting the door position only. Only the training method of the detection calculation model needs to be adjusted accordingly. The present invention does not make further restrictions here.

[0105] The communication module is a vehicle Ethernet or a CAN / LIN network, which is used to send the pinch detection result to the alarm module. When there is a target object within the door frame range of the vehicle door, the alarm module is used to issue an alarm.

[0106] In this embodiment, the door anti-pinch detection system further includes a human-machine interaction module to achieve information interaction with the user, including a control interface and an alarm processing interface, through which alarm information can be obtained. For example, an alarm flag can be displayed on the display screen, and the user can also click a button on the screen to turn off the alarm reminder, etc.

[0107] In addition, Figure 6 FIG. is a schematic diagram of the installation positions of the millimeter-wave radar on a swing car door and a sliding car door in this embodiment. In this embodiment, the millimeter-wave radar is installed at the B-pillar position of the vehicle, which can cover the door frame to the greatest extent. In other embodiments, the installation position of the millimeter-wave radar can be determined according to the actual design requirements of the vehicle, and the present invention does not make specific limitations here.

[0108] Another embodiment of the present invention provides a vehicle equipped with the door anti-pinch detection system in the above embodiment, and the technical features thereof will not be elaborated here.

[0109] It should be noted that the embodiments of the present invention have good implementability and are not any form of limitation to the present invention. Any person skilled in the art may use the disclosed technical content to change or modify it into an equivalent effective embodiment. However, as long as it does not depart from the technical content of the present invention, any modification, equivalent change or modification made to the above embodiments according to the technical essence of the present invention still falls within the scope of the technical solution of the present invention.

Claims

1. A method for detecting door pinch protection for a vehicle, characterized in that, it includes: Establish a door pinch detection calculation model to determine the correspondence between signal characteristics and pinch detection results; Collect echo signals within the door frame range of the door in real time; Calculate the signal characteristics of the echo signals according to the echo signals; Substitute the signal characteristics into the door pinch detection calculation model to determine the pinch detection result of the door.

2. The method for detecting door pinch protection for a vehicle according to claim 1, characterized in that, The establishment of the door pinch detection calculation model includes: Collect signals: respectively and continuously collect the first echo signal and the second echo signal when there is a target object and when there is no target object within the door frame range; Form training set data: Calculate the first signal feature set {D10, A10, E i10} of the first echo signal, and set the label of the first signal feature set to 1; Calculate the second signal feature set {D20, A20, E i20} of the second echo signal, and set the label of the second signal feature set to 0. The labeled first signal feature set and second signal feature set are the training set data; Use machine learning methods for training: divide the training set data into Z subsets, where (Z - 1) subsets are used as the training set to train the SVM classifier; the remaining 1 subset is the test set to perform classification testing on the SVM classifier; Evaluate the performance of the door pinch detection calculation model according to the classification test results; Determine the correspondence between signal characteristics and pinch detection results. When the output result of the door pinch detection calculation model is 1, it is determined that there is a target object within the door frame range of the door. When the output result of the door pinch detection calculation model is 0, it is determined that there is no target object within the door frame range of the door.

3. The method for detecting door pinch protection for a vehicle according to claim 2, characterized in that, The evaluation of the performance of the door pinch detection calculation model according to the classification test results includes: Calculate the accuracy Ac, precision Pr, and recall Re of the door pinch detection calculation model; Evaluate the performance of the door pinch detection calculation model according to the accuracy Ac, precision Pr, and recall Re.

4. The method for detecting door pinch protection for a vehicle according to claim 2, characterized in that, Real-time collect the third echo signal within the range of the door frame of the vehicle door, and calculate a third signal feature set {D30, A30, E of the third echo signal according to the third echo signal i30}.

5. The method for detecting door pinch protection for a vehicle according to claim 4, characterized in that, Calculate the signal feature set {Da0, Aa0, E of the echo signal ia0}, where a takes values of 1, 2, and 3 respectively, including: Perform two-dimensional FFT transformation on the echo signals: where x(n, m) is the echo signal, N is the number of FFT points in the distance dimension, and M is the number of FFT points in the velocity dimension; Perform constant false alarm processing on the two-dimensional range-Doppler matrix X(k, l) to obtain the range bin rbin and velocity bin vbin where the target object is located; calculate the distance information Da0 of the target object according to formula (2) as the first feature of the signal feature set; Da0 = rbin * c / 2B Formula (2) where c is the speed of light and B is the radar bandwidth; Perform FFT transformation on the two-dimensional range-Doppler matrix X(k, l) according to the antenna dimension: where S is the number of FFT points in the antenna dimension; Calculate the angle information Aa0 of the target object as the second feature of the signal feature set; Aa0 = asin(f w *λ / d)*180 / π Formula (4) λ = c / f d = λ / 2 where f w is the frequency corresponding to the maximum value of X(k, l, v), c is the speed of light, and f is the radar carrier frequency; Perform time-frequency analysis on the data of ten range bins near the range bin to calculate the short-time Fourier transform and obtain the micro-Doppler spectrum data: where x(m) is the echo signal, ω(m) is the window function, and i is the time index; Continue windowing the echo signal along the time of the echo signal to obtain the variation of the signal frequency with time, and convert the first echo signal into a Doppler time-signal matrix; According to the two-dimensional Doppler time-signal matrix, the instantaneous frequency corresponding to each of the distance cells is obtained, and the signal energy values corresponding to the instantaneous frequencies are combined to form ten-dimensional features, which are the third feature E of the signal feature set ia0 .

6. The method for detecting door pinch prevention for a vehicle according to claim 5, wherein, Substitute the third signal feature into the door pinch detection calculation model. When the output result of the door pinch detection calculation model is 1, it is determined that there is a target object within the door frame range of the door. When the output result of the door pinch detection calculation model is 0, it is determined that there is no target object within the door frame range of the door.

7. The method for detecting door pinch prevention for a vehicle according to any one of claims 1-6, wherein, The echo signal is the echo signal of a millimeter-wave radar signal.

8. A door pinch prevention detection system for a vehicle, wherein, Apply the method for detecting door pinch prevention for a vehicle according to any one of claims 1-7.

9. The door pinch prevention detection system for a vehicle according to claim 8, wherein, It includes a millimeter-wave radar detection module, a communication module and an alarm module; The millimeter-wave radar detection module includes a millimeter-wave radar and a calculation unit. The millimeter-wave radar is used to transmit millimeter-wave radar signals and real-time collect echo signals within the door frame range of the door. The calculation unit is used to calculate the signal features of the echo signals according to the echo signals, and substitute the signal features into the door pinch detection calculation model to determine the door pinch detection result; The communication module is used to send the pinch detection result to the alarm module. When there is a target object within the door frame range of the door, the alarm module is used to issue an alarm.

10. A vehicle, wherein, Install the door pinch prevention detection system for a vehicle according to claim 8 or 9.

Images