Intelligent anticipatory control method for vehicle stability based on integrated chassis

Abstract

The present invention relates to the field of data processing technology, and more specifically to an intelligent anticipatory control method for vehicle stability based on an integrated chassis. The method comprises: a lane line feature deep learning network based on vehicle parameters and road information; identifying the road features based on the network model; self-learning and self-correcting the features based on the network's predicted and identified features; predicting the road's radius of curvature based on the road features; and controlling the vehicle's motion stability based on the road's radius of curvature. By extracting vehicle parameters and road features, the method makes advance predictions of future travel routes, and uses feedback learning to correct the radius of curvature, thereby improving the accuracy of the radius of curvature and accurately controlling the vehicle's motion stability.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an intelligent anticipatory control method for vehicle stability based on an integrated chassis in the technical field of data processing. Background Art

[0002] With the continuous advancement of the automotive industry, consumers' requirements for vehicle performance are increasing. Modern cars not only need to have good power and comfort, but vehicle stability has also become a key indicator. Traditional chassis systems have gradually exposed their limitations when dealing with complex road conditions and driving conditions. For example, when encountering crosswinds or making emergency avoidance of obstacles while driving at high speeds, it is easy for the vehicle to lose control. The rapid development of automotive electronics technology has provided technical support for the emergence of intelligent chassis. However, during driving, especially in complex road conditions (such as slippery roads, rugged mountain roads) or emergency driving situations (such as emergency braking, sudden steering), it is easy for the vehicle to lose control. Summary of the Invention

[0003] The purpose of the present invention is to provide an intelligent anticipatory control method for extreme stability, and the technical solutions adopted are as follows:

[0004] In a first aspect, an embodiment of the present invention provides a vehicle stability intelligent anticipation control method based on an integrated chassis, the method comprising:

[0005] A lane feature deep learning network based on vehicle parameters and road information to identify road features;

[0006] Performing self-learning and self-correction of features based on the prediction features and recognition features of the lane feature deep learning network;

[0007] predicting a curvature radius of the road based on the road characteristics;

[0008] Based on the curvature radius of the road, the motion stability of the vehicle is controlled.

[0009] In a second aspect, a vehicle stability intelligent anticipation control system based on an integrated chassis is provided, the system comprising:

[0010] Recognition module, which is used to identify road features based on the lane line feature deep learning network of vehicle parameters and road information;

[0011] A self-learning module, configured to perform self-learning and self-correction of features based on the prediction features and recognition features of the lane feature deep learning network;

[0012] A prediction module, configured to predict a curvature radius of the road based on the road characteristics;

[0013] A control module is configured to control the vehicle's motion stability based on the curvature radius of the road. In a third aspect, a computer program product is provided, comprising: computer program code that, when executed on a computer, causes the computer to execute the method of the first aspect or any possible implementation of the first aspect.

[0014] In a fourth aspect, a computer-readable storage medium is provided, which stores a computer program code. When the computer program code runs on a computer, the computer executes the method in the first aspect or any possible implementation method described in the first aspect.

[0015] The present invention has the following beneficial effects: a lane line feature deep learning network based on vehicle parameters and road information identifies road features; and self-learning and self-correcting features are performed based on the predicted features and identified features of the lane line feature deep learning network. In this way, by performing lane line recognition on road information, multiple lanes on the road can be accurately predicted, so as to accurately implement self-learning and self-correction of the lane line feature deep learning network. Thereafter, based on the road features, the curvature radius of the road is predicted; and based on the curvature radius of the road, the motion stability of the vehicle is controlled. In this way, by extracting the vehicle parameters and road features, the future driving route is predicted in advance, and feedback learning is performed, so that the lane line feature deep learning network can be self-learned and self-corrected, which can improve the accuracy of the curvature radius and thus accurately control the motion stability of the vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0017] Figure 1 This is a schematic diagram of an implementation flow of an intelligent anticipatory control method for vehicle stability based on an integrated chassis provided by an embodiment of the present invention;

[0018] Figure 2 1 is a schematic diagram of an implementation framework of an intelligent anticipatory control method for vehicle stability based on an integrated chassis provided by an embodiment of the present invention;

[0019] Figure 3 1 is a schematic diagram of a curve for predicting vehicle motion stability provided by an embodiment of the present invention;

[0020] Figure 41 is a schematic diagram of a curve showing changes in vehicle stability with speed and adhesion provided by an embodiment of the present invention;

[0021] Figure 5 This is a framework diagram of a feedback system for vehicle stability pre-control provided by an embodiment of the present invention;

[0022] Figure 6 2 is another schematic diagram of an implementation framework of an intelligent anticipatory control method for vehicle stability based on an integrated chassis provided by an embodiment of the present invention;

[0023] Figure 7 1 is a schematic diagram of the structure of a vehicle stability intelligent anticipation control system based on an integrated chassis provided by an embodiment of the present invention;

[0024] Figure 8 It is a structural diagram of a computer device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0025] To further illustrate the technical means and effectiveness of the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effectiveness of the intelligent anticipatory control method for vehicle stability based on an integrated chassis proposed in accordance with the present invention. In the following description, references to different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0026] In the description of the embodiments of the present invention, unless otherwise specified, " / " means or, for example, A / B can mean A or B: "and / or" in the text is only a description of the association relationship of associated objects, indicating that there can be three relationships, for example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, in the description of the embodiments of the present invention, "multiple" refers to two or more than two.

[0027] In the following, the terms "first" and "second" are used for descriptive purposes only and should not be understood to imply or suggest relative importance or implicitly indicate the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features.

[0028] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0029] The following is a detailed description of a specific solution of the vehicle stability intelligent anticipation control method based on an integrated chassis provided by the present invention in conjunction with the accompanying drawings. Figure 1, which shows a schematic diagram of an implementation flow of a vehicle stability intelligent anticipation control method based on an integrated chassis provided by one embodiment of the present invention, the method comprising:

[0030] 101, a lane feature deep learning network based on vehicle parameters and road information to identify road features.

[0031] Vehicle parameters include: vehicle length, width, height, wheelbase, track width; curb weight, loaded mass, and mass distribution; engine power and torque characteristics; transmission ratio range and shifting logic; and tire parameters such as tire size, type, tread, tire pressure, and rolling resistance coefficient. Road information includes movement trajectory, forward route data, lanes on the road, and intersections. In some possible implementations, road information is extracted from a continuous video stream of the vehicle traveling on the road. Sensor technology can accurately sense various vehicle state parameters, such as vehicle speed, wheel speed, yaw rate, and lateral acceleration. These parameters serve as the foundation for intelligent predictive control of vehicle stability based on the integrated chassis. By assessing the curvature of the road ahead, the vehicle maintains a stable state as it approaches the road ahead.

[0032] Here, lane-specific networks are formed based on road features. These networks include the multiple lanes of the road the vehicle is on. A convolutional residual network (CNN+ResNet) architecture is used to extract road features from road information, generating road features. These features are then fused to create a road feature space. Lanes are then classified and identified within this road feature space to generate the lane-specific networks.

[0033] In some possible implementations, first, feature extraction is performed on road information to obtain road features; second, feature fusion is performed on the road features to obtain a road feature space; finally, lane classification and recognition are performed on the road feature space to obtain the lane line feature deep learning network, thereby completing the self-learning and self-correction of the lane line feature deep learning network. Figure 2As shown, a video stream is input via the video input module 201. The lane feature extraction module 202 then extracts features from the video stream to obtain road features. The lane feature fusion module 203 then fuses the road features to obtain a lane feature space 204 and a lane feature space 205 at time (t-1). Lane feature space 204 is then used to form a lane differentiation network 206. In this way, edge detection operators such as Canny are first used to identify pixels with drastic intensity changes in the image as edges. The Hough transform is then used to cast votes on edge pixels in polar coordinates to determine line parameters, thereby detecting lane positions. Based on the fusion of road features, a new feature space, lane feature space 204, is formed. By training on a large number of images labeled with lane lines, the network can predict whether each pixel belongs to a lane line or background, thereby determining the lane line. Through image processing algorithms, machine learning models, and sensor fusion technologies, lanes can be accurately classified and identified, resulting in a precise lane differentiation network.

[0034] 102. Perform feature self-learning and self-correction based on the prediction features and recognition features of the lane feature deep learning network.

[0035] Here, a lane line feature deep learning network is used to determine the predicted curvature of the road online; and based on parameter information, the measured curvature of the road where the vehicle is located is determined online; finally, the predicted curvature is self-corrected through the measured curvature to complete the self-correction process of the lane line feature deep learning network.

[0036] 103. Predict the curvature radius of the road based on the road characteristics.

[0037] Here, lane prediction is performed on different lane networks to analyze the intersection information of the road, thereby constructing a detection network, and Kalman fusion is performed on the detection network to obtain the predicted curvature of the road. In some possible implementations, this can be achieved through the following process: first, lane prediction is performed on different lane networks of the road at multiple angles to determine the intersection information of the road; wherein the intersection information includes crossroads, T-junctions, etc. in the road. Secondly, based on the intersection information and the different lane networks, a detection network of the road is constructed; finally, the detection network is fused to obtain the predicted curvature of the road. Here, the formation of the detection network requires the integration of multiple detection technologies and equipment, such as cameras, radars, and lidars, to obtain traffic information of different lane networks and lane intersections from different angles and levels. Kalman fusion is performed on the detection network to finally obtain the predicted curvature of the road. As Figure 2As shown, lane prediction 207 is performed using the lane difference network 206 to output a lane intersection 208. A detection network 209 is formed using the lane intersection 208 and the lane difference network 206. Finally, the detection network is fused using a Kalman fusion module 210 to output a predicted curvature of the road.

[0038] 104 . Control the motion stability of the vehicle based on the curvature radius of the road.

[0039] Here, the vehicle parameters of the vehicle are estimated through the vehicle motion model to calculate the radius of the vehicle's upcoming turn and the driving path, so as to accurately calculate the measured curvature of the road where the vehicle is located by combining the motion status of other vehicles around the vehicle.

[0040] In some possible implementations, this can be achieved through the following process: first, based on the vehicle parameters, predict the turning radius and driving path of the vehicle in the next bend of the road; wherein the next bend of the vehicle on the road is the first bend the vehicle is about to enter. The turning radius and driving path of the next bend can be predicted using the vehicle parameters using a vehicle motion model. Figure 2 As shown, vehicle parameters 21 of the vehicle are input into a vehicle motion model 22 to output a vehicle turning motion estimate 23 , ie, a turning radius and a driving path of the vehicle in the next bend of the road.

[0041] In some possible implementations, the real-time driving speed of the vehicle and the real-time friction coefficient of the road on the vehicle are obtained from the vehicle parameters; then, based on the real-time driving speed and the real-time friction coefficient, the turning radius and driving path of the vehicle in the next bend of the road are determined. Here, by inputting the real-time driving speed, the real-time friction coefficient, the speed limit of the vehicle into the vehicle motion model in combination with the predicted curvature, the turning radius and driving path of the vehicle in the next bend of the road can be calculated. Figure 3 As shown, the current speed V of the vehicle curr , the predicted curvature C of the detected future road futr , the friction coefficient μ of the road where the vehicle is detected futr and the vehicle's current driving force Calculate the vehicle's speed limit V on the road lim and driving force limitations The turning radius and driving path of the vehicle in the next curve of the road are calculated by using the speed limit and driving force limit. Figure 3 In, F cf represents the centripetal force, σ str Indicates the turn signal, ω calc represents the yaw rate, β calcIndicates the vehicle's side slip angle. First, C futr Obtain ω through a calculation module calc (calculated yaw rate), and μ futr After a calculation module, β calc (vehicle sideslip angle), ω calc =V curr ·C futr ω calc , β calc and their predicted values ​​ω pred and β pred By calculating, predicting and arbitrating / coordinating multiple parameters, the system ensures the stability and safety of the vehicle under different driving conditions. The system performs a series of calculations and predictions based on parameters such as the current vehicle speed, vehicle status, friction coefficient and driving force / braking force, and outputs control instructions through the arbitration / coordination module to maintain the vehicle's stable driving. Figure 4 In the figure, curve 402 shows how the vehicle's stability changes when the vehicle's speed v remains constant and the adhesion μ gradually increases, based on curve 401. Curve 403 shows how the vehicle's stability changes when the vehicle's speed v remains constant and the adhesion μ gradually decreases, based on curve 401. Curve 404 shows how the vehicle's stability changes when the vehicle's speed v remains constant and the adhesion μ gradually increases, based on curve 401. Curve 405 shows how the vehicle's stability changes when the vehicle's speed v gradually decreases, based on curve 401, while the adhesion μ remains constant.

[0042] Secondly, on the road, the motion status of other vehicles whose distance to the vehicle is less than a preset distance threshold is obtained; here, the preset distance threshold can be a custom setting, and the other vehicles whose distance to the vehicle is less than the preset distance threshold are the vehicles around the vehicle. The motion status of other vehicles includes: the motion trajectory, motion speed, turning tendency, etc. of other vehicles. After extracting the length, width, height, wheelbase, track width, curb weight, load mass and mass distribution of the vehicle body, the power and torque characteristics of the engine, the transmission ratio range and shifting logic of the transmission, tire parameters such as tire size, type, pattern, tire pressure and rolling resistance coefficient, etc., these comprehensive and accurate vehicle parameters are obtained, and the vehicle parameters are deeply integrated into the design process of the vehicle motion model to obtain the vehicle turning motion estimation.

[0043] Again, based on the motion state of the other vehicles, the turning radius and the driving path are adjusted respectively to obtain the adjusted turning radius and the adjusted driving path; here, the turning radius and the driving path of the vehicle in the upcoming curve are calculated in advance based on the vehicle motion model, and at the same time, according to the motion state of the surrounding vehicles, the turning strategy and motion estimation are adjusted, that is, the turning radius and the driving path are adjusted to obtain the adjusted turning radius and the adjusted driving path, thereby ensuring that the vehicle can complete the turning action safely and accurately in a complex traffic environment.

[0044] Finally, the measured curvature of the road on which the vehicle is located is determined based on the adjusted turning radius and the adjusted driving path. Here, the adjusted turning radius and the adjusted driving path are processed using a Kalman fusion approach to calculate the measured curvature of the road on which the vehicle is located online. The lane feature deep learning network is then self-corrected using the measured and predicted curvatures to determine the curvature radius.

[0045] Here, the curvature difference between the measured curvature and the predicted curvature is calculated and fed back to the vehicle feature extraction module of the network model. Figure 2 As shown, the vehicle feature extraction module extracts road features in real time as the vehicle travels, and re-determines the predicted curvature based on the feedback curvature difference until the curvature difference between the measured curvature and the predicted curvature is small, then stops feedback to obtain the curvature radius.

[0046] In some possible implementations, step 102 may be implemented by the following steps:

[0047] In the first step, a curvature difference between a predicted curvature corresponding to the predicted feature and a measured curvature corresponding to the identified feature is determined.

[0048] Here, the measured curvature and the predicted curvature are subtracted to obtain the curvature difference.

[0049] In the second step, based on the curvature difference, the lane line feature deep learning network is self-learned and self-corrected.

[0050] like Figure 2As shown in the figure, road feature detection is based on feedback comparing the measured curvature derived from vehicle parameters with the predicted curvature derived from predictions of road information. The vehicle's speed sensor provides speed information. Combined with the vehicle's turning radius and speed, the basic physics of circular motion is used to more accurately calculate an approximate value of the road curvature corresponding to the vehicle's current trajectory. Based on the feedback from the road curvature comparison, the road feature extraction module enters a new phase, constructing a new road feature space. This new road feature space will continue to evolve and improve over time and with the continuous accumulation of vehicle driving data. Each feedback correction will enable the road feature detection model to gain a deeper understanding of road features in different scenarios. Based on the new road feature space, lane predictions and lane intersection predictions are performed, forming a detection network with different lane networks. Starting from the geometric characteristics of the lanes, through in-depth analysis of the changing patterns of lane curvature in continuous road sections, and using advanced time series prediction models such as long-short-term memory networks, it is possible to accurately infer the direction and curvature trend of lanes within a certain distance in the future. Based on the road topology information in the new road feature space, combined with traffic flow data and historical traffic patterns, a probability-based statistical method is used to predict traffic conditions at lane intersections. The lane predictions and lane intersection predictions based on the new road feature space are integrated with this to form a comprehensive and efficient detection network. Then, through Kalman fusion of the detection network, a more accurate curvature radius can be obtained.

[0051] In some possible implementations, first, based on the predicted curvature and the current speed of the vehicle, a predicted yaw rate and a predicted vehicle sideslip angle of the vehicle are determined; and based on the measured curvature and the current speed of the vehicle, a measured yaw rate and a measured vehicle sideslip angle of the vehicle are determined; that is, Figure 3 As shown, C futr The measured yaw angular velocity ω is obtained through a calculation module calc , while μ futr The measured vehicle side slip angle β is obtained through a calculation module calc ,ω calc =V curr ·C futr ω calc , β calc and the corresponding predicted yaw rate ω pred And the predicted vehicle sideslip angle β pred Make a comparison.

[0052] Thereafter, the yaw rate difference between the predicted yaw rate and the measured yaw rate, as well as the slip angle difference between the predicted vehicle slip angle and the measured vehicle slip angle, are determined respectively. Based on the yaw rate difference, the slip angle difference, and the curvature difference, the lane-specific network of the road is re-identified to re-determine the predicted curvature of the road until the curvature difference is less than a preset difference threshold, thereby completing the self-learning and self-correction of the lane line feature deep learning network. Here, the yaw rate difference and the slip angle difference are fed back to the road feature extraction module to extract road features in real time, thereby constructing a new lane-specific network, re-identifying the lane line feature deep learning network of the road to re-determine the predicted curvature of the road until the curvature difference is less than a preset difference threshold, thereby completing the self-learning and self-correction of the lane line feature deep learning network.

[0053] In some possible implementations, feedback on the intersection information of the road is received; the yaw rate difference, the sideslip angle difference, the curvature difference, and the intersection information are input into a road feature extraction module of a network model, thereby extracting the road features in real time to obtain real-time road features. Based on the real-time road features, the lane diversity network of the road is re-identified to obtain the latest lane diversity network. A detection network is constructed using the latest lane diversity network, and Kalman fusion is performed on the detection network to re-determine the predicted curvature of the road until the curvature difference is less than a preset difference threshold. In this manner, online learning, advance prediction, online evaluation learning, and updating of road feature information are performed through a local feedback network to obtain the predicted curvature.

[0054] In some possible implementations, the predicted curvature corresponding to when the curvature difference is less than the preset difference threshold is determined as the curvature radius.

[0055] Here, the preset difference threshold can be custom-set. Specifically, when the curvature difference is small, the difference between the predicted curvature and the calculated measured curvature is small, indicating that the network model's prediction accuracy is high. Therefore, the predicted curvature corresponding to the curvature difference being less than the preset difference threshold is determined as the curvature radius, so that the vehicle's driving force and speed can be further controlled using this curvature radius, thereby ensuring stable vehicle motion. In this way, by controlling the vehicle's speed within the speed limit using the curvature radius, the vehicle's motion stability can be ensured.

[0056] In some possible implementations, first, a speed limit of the vehicle on the road is obtained based on the curvature radius.

[0057] Here, after the curvature radius is obtained, the speed limit corresponding to the curvature radius can be found through the corresponding relationship table between the curvature radius and the speed limit.

[0058] Next, a speed difference between the speed limit and the current speed of the vehicle is determined.

[0059] Here, after calculating the speed difference between the speed limit and the current speed of the vehicle, the difference is further normalized to obtain a normalized value of the speed difference.

[0060] Next, the speed difference is subjected to feedforward processing to obtain driving information of the vehicle.

[0061] Here, the driving information includes: the current driving force and the maximum driving force of the vehicle. Figure 5 As shown, the current speed limit Input feedforward system to output real-time friction of the vehicle on the road By dividing the centripetal distance Input feedback control to output the vehicle's current driving force and will and Fusion and centripetal force

[0062] Finally, the motion stability of the vehicle is controlled based on the driving information and the real-time friction coefficient of the road on which the vehicle is located.

[0063] Here, the current speed of the vehicle and the speed limit of the vehicle on the road are calculated by the driving force and real-time friction coefficient of the vehicle in the driving information, so that the real-time speed of the vehicle is controlled within the speed limit. Figure 5 As shown, in the feedback system of vehicle speed control, the curvature radius and the current speed of the vehicle are used to calculate the current speed limit of the vehicle on the road. For example, the current speed limit is Where D represents the normalized value of the speed difference. In this way, by calculating the vehicle's driving information and the real-time friction coefficient, the vehicle's real-time speed can be controlled within a safe range, thereby controlling the vehicle's motion stability. Figure 5 As shown, the difference between the current speed of the vehicle and the speed limit is calculated, normalized, and then passed through the feedforward and feedback control modules to finally obtain a control output to ensure that the vehicle speed is within a safe range.

[0064] The vehicle stability intelligent anticipation control method based on the integrated chassis provided by the embodiment of the present invention can be achieved by Figure 6The framework diagram shown in FIG. 1 is implemented as follows. The framework includes: a road ahead information detection module 601, a road time limit prediction module 602, a vehicle dynamic stability prediction module 603, a vehicle dynamic stability pre-control module 604, a vehicle dynamic state extraction module 605, a road time limit estimation module 606, a vehicle dynamic stability forward control module 607, a vehicle dynamic inspection module 608, a vehicle dynamic stability feedback control module 609, an arbitration / coordination module 610, a vehicle steering control module 611, a vehicle privacy control module 612, and a vehicle braking control module 613. A real-time video stream of a moving vehicle is input into the road ahead information detection module 601 to extract road features and then into the road time limit prediction module 602 to predict the time limit of the road the vehicle is on. The vehicle dynamic stability prediction module 603 then predicts the predicted curvature and speed limit of the road the vehicle is on. The vehicle dynamic stability pre-control module 604 then controls the vehicle's stability motion based on these parameters. Simultaneously, the vehicle's motion state is analyzed by the vehicle dynamic state extraction module 605 and input into the road time limit estimation module 606, enabling the vehicle dynamic stability forward control module 607 to calculate the measured curvature. The calculated measured curvature is fed back to the arbitration / coordination module 610 via the vehicle dynamic inspection module 608 and the vehicle dynamic stability feedback control module 609, enabling the arbitration / coordination module 610 to more stably control the vehicle steering control module 611, the vehicle privacy control module 612, and the vehicle braking control module 613. In this way, through autonomous learning and online learning mechanisms, the vehicle perception system predicts the road conditions ahead and enables proactive adjustments to the chassis system. By extracting vehicle parameters and road features, the future travel route is predicted in advance, and feedback learning is used to correct the predicted road curvature. The intelligent chassis's anticipatory control can predict road curvature in advance, assisting the vehicle's stability control and the driver's driving.

[0065] In an embodiment of the present invention, a lane line feature deep learning network based on vehicle parameters and road information identifies road features; and based on the predicted features and identified features of the lane line feature deep learning network, self-learning and self-correction of features are performed. In this way, by performing lane line recognition on road information, multiple lanes in the road can be accurately predicted, so as to accurately implement self-learning and self-correction of the lane line feature deep learning network. Thereafter, based on the road features, the curvature radius of the road is predicted; and based on the curvature radius of the road, the motion stability of the vehicle is controlled. In this way, by extracting the vehicle parameters and road features, the future driving route is predicted in advance, and feedback learning is performed, so that the lane line feature deep learning network can be self-learned and self-corrected, which can improve the accuracy of the curvature radius and thus accurately control the motion stability of the vehicle.

[0066] The embodiment of the present invention provides a vehicle stability intelligent anticipation control system based on an integrated chassis, please refer to Figure 7 , which shows a schematic diagram of the structure of a vehicle stability intelligent anticipation control system based on an integrated chassis provided by one embodiment of the present invention. The system 700 includes:

[0067] Identification module 701, for identifying road features based on a lane feature deep learning network of vehicle parameters and road information;

[0068] A self-learning module 702 is configured to perform self-learning and self-correction of features based on the prediction features and recognition features of the lane feature deep learning network;

[0069] A prediction module 703 is configured to predict the curvature radius of the road based on the road characteristics;

[0070] The control module 704 is configured to control the motion stability of the vehicle based on the curvature radius of the road.

[0071] In some possible implementations, the self-learning module 702 is further used to determine the curvature difference between the predicted curvature corresponding to the predicted feature and the measured curvature corresponding to the identified feature; based on the curvature difference, the lane line feature deep learning network is self-learned and self-corrected.

[0072] In some possible implementations, the self-learning module 702 is further used to determine the predicted yaw rate and the predicted vehicle sideslip angle of the vehicle based on the predicted curvature of the road and the current speed of the vehicle; determine the measured yaw rate and the measured vehicle sideslip angle of the vehicle based on the measured curvature and the current speed of the vehicle; determine the yaw rate difference between the predicted yaw rate and the measured yaw rate, and the sideslip angle difference between the predicted vehicle sideslip angle and the measured vehicle sideslip angle, respectively; and re-identify the different lane networks of the road based on the yaw rate difference, the sideslip angle difference, and the curvature difference to re-determine the predicted curvature of the road, until the curvature difference is less than a preset difference threshold, so as to complete the self-learning and self-correction of the lane line feature deep learning network.

[0073] In some possible implementations, the self-learning module 702 is further configured to receive feedback from a deep learning network of lane line features of the road; extract the road features of the road in real time based on the yaw angular velocity difference, the sideslip angle difference, the curvature difference, and the intersection information to obtain real-time road features; and re-identify the deep learning network of lane line features of the road based on the real-time road features to re-determine the predicted curvature of the road until the curvature difference is less than a preset difference threshold.

[0074] In some possible implementations, the self-learning module 702 is also used to extract features from the road information to obtain road features; perform feature fusion on the road features to obtain a road feature space; perform lane classification and identification on the road feature space to obtain the lane line feature deep learning network, so as to complete self-learning and self-correction of the lane line feature deep learning network.

[0075] In some possible implementations, the self-learning module 702 is also used to predict lanes of the road at multiple angles based on different lane networks to determine the intersection information of the road; construct a detection network for the road based on the intersection information and the different lane networks; and fuse the detection networks to obtain the predicted curvature of the road.

[0076] In some possible implementations, the self-learning module 702 is further used to determine the turning radius and driving path of the vehicle in the next bend of the road based on the vehicle parameters; obtain the motion status of other vehicles on the road whose distance from the vehicle is less than a preset distance threshold; based on the motion status of the other vehicles, adjust the turning radius and driving path respectively to obtain an adjusted turning radius and an adjusted driving path; and determine the measured curvature of the road on which the vehicle is located based on the adjusted turning radius and the adjusted driving path.

[0077] In some possible implementations, the self-learning module 702 is also used to obtain the real-time driving speed of the vehicle and the real-time friction coefficient of the road on the vehicle from the vehicle parameters; and based on the real-time driving speed and the real-time friction coefficient, determine the turning radius and driving path of the vehicle in the next bend of the road.

[0078] In some possible implementations, the control module 704 is further used to obtain the speed limit of the vehicle on the road based on the curvature radius of the road; determine the speed difference between the speed limit and the current speed of the vehicle; perform feed-forward processing on the speed difference to obtain driving information of the vehicle; and control the motion stability of the vehicle in anticipation based on the driving information and the real-time friction coefficient of the road on which the vehicle is located.

[0079] Optionally, the transmission medium can be a wired link (for example, but not limited to, coaxial cable, optical fiber and digital subscriber line (DSL)) or a wireless link (for example, but not limited to, wireless Fidelity (WIFI), Bluetooth and mobile device network). It should be noted that the system provided in the above embodiment is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the computer device can be divided into different functional modules to complete all or part of the functions described above. In addition, the method embodiments provided in the above embodiments belong to the same concept. The specific implementation process is detailed in the method embodiments, which will not be repeated here.

[0080] Figure 8 FIG. 1 is a schematic diagram of the structure of a computer device provided by an embodiment of the present invention. For example, Figure 8 As shown, the computer device 800 includes: a memory 801, a processor 802, and a computer program 803 stored in the memory 801 and running on the processor 802, wherein when the processor 802 executes the computer program 803, the computer device can execute any one of the vehicle stability intelligent anticipation control methods based on the integrated chassis introduced above.

[0081] In addition, an embodiment of the present invention also protects a system, which may include a memory and a processor, wherein the memory stores an executable program code, and the processor is used to call and execute the executable program code to perform an intelligent anticipatory control method for vehicle stability based on an integrated chassis provided by an embodiment of the present invention. This embodiment can divide the system into functional modules according to the above method example. For example, it can correspond to each functional module, or two or more functions can be integrated into one processing module. The above integrated module can be implemented in the form of hardware. It should be noted that the division of modules in this embodiment is schematic, which is only a logical function division. There may be other division methods in actual implementation. It should be noted that all relevant contents of each step involved in the above method embodiment can be referred to the functional description of the corresponding functional module, and will not be repeated here.

[0082] It should be understood that the system provided in this embodiment is used to execute the above-mentioned intelligent anticipatory control method for vehicle stability based on an integrated chassis, and therefore can achieve the same effect as the above-mentioned implementation method. In the case of an integrated unit, the system may include a processing module and a storage module. Specifically, when the system is applied to a device, the processing module can be used to control and manage the actions of the device. The storage module can be used to support the device in executing mutual program codes, etc. Specifically, the processing module can be a processor or a controller, which can implement or execute the various exemplary logic boxes, modules and circuits described in conjunction with the disclosure of the present invention. The processor can also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of digital signal processing (DSP) and a microprocessor, etc., and the storage module can be a memory.

[0083] Furthermore, the system provided by the embodiments of the present invention may be specifically a chip, component, or module. The chip may include a connected processor and memory; the memory is used to store instructions. When the processor calls and executes the instructions, the chip can execute the intelligent anticipatory control method for vehicle stability based on an integrated chassis provided in the above embodiment. This embodiment also provides a computer-readable storage medium storing computer program code. When the computer program code is executed on a computer, the computer executes the above-mentioned method steps to implement the intelligent anticipatory control method for vehicle stability based on an integrated chassis provided in the above embodiment.

[0084] This embodiment also provides a computer program product. When the computer program product is executed on a computer, it causes the computer to execute the above-mentioned steps to implement the vehicle stability intelligent anticipation control method based on an integrated chassis provided in the above embodiment. The system, computer-readable storage medium, computer program product, or chip provided in this embodiment are all used to execute the corresponding method provided above. Therefore, the beneficial effects achieved by the system can refer to the beneficial effects of the corresponding method provided above and will not be repeated here. Through the description of the above embodiments, those skilled in the art will understand that for the convenience and brevity of description, only the division of the above-mentioned functional modules is used as an example. In actual application, the above-mentioned functions can be distributed to different functional modules as needed, that is, the internal structure of the system can be divided into different functional modules to complete all or part of the functions described above. In the embodiments provided by the present invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the division of modules or units is only a logical functional division. In actual implementation, other division methods can be used. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not implemented. On the other hand, the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, system or unit, which may be electrical, mechanical or other forms.

[0085] It should be noted that the above-mentioned order of the embodiments of the present invention is for description only and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are also possible or may be advantageous. The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other. Each embodiment focuses on the differences from other embodiments. The above content is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or replacements within the technical scope disclosed by the present invention, which should be covered within the scope of protection of the present invention.

Claims

1. The vehicle stability intelligent anticipation control method based on the integrated chassis is characterized by: The vehicle stability intelligent anticipation control method based on the integrated chassis includes: A lane feature deep learning network based on vehicle parameters and road information to identify road features; Determining a curvature difference between a predicted curvature corresponding to a predicted feature of the lane feature deep learning network and a measured curvature corresponding to the identified feature; determining a predicted yaw rate and a predicted vehicle slip angle of the vehicle based on the predicted curvature of the road and the current speed of the vehicle; determining a measured yaw rate of the vehicle and a measured vehicle sideslip angle based on the measured curvature and a current speed of the vehicle; determining a yaw rate difference between the predicted yaw rate and the measured yaw rate, and a sideslip angle difference between the predicted vehicle sideslip angle and the measured vehicle sideslip angle, respectively; Re-identifying different lane networks of the road based on the yaw rate difference, the sideslip angle difference, and the curvature difference to re-determine a predicted curvature of the road until the curvature difference is less than a preset difference threshold, thereby completing self-learning and self-correction of the lane feature deep learning network; predicting a curvature radius of the road based on the road characteristics; Based on the curvature radius of the road, the motion stability of the vehicle is controlled.

2. The vehicle stability intelligent anticipation control method based on the integrated chassis according to claim 1 is characterized in that: The method of re-identifying different lane networks of the road based on the yaw rate difference, the sideslip angle difference, and the curvature difference to re-determine the predicted curvature of the road until the curvature difference is less than a preset difference threshold, thereby completing self-learning and self-correction of features of the lane feature deep learning network, includes: A lane feature deep learning network receiving feedback of the road; Extracting road features of the road in real time based on the yaw rate difference, the sideslip angle difference, the curvature difference, and intersection information to obtain real-time road features; A lane feature deep learning network of the road is re-identified based on the real-time road feature to re-determine the predicted curvature of the road until the curvature difference is less than a preset difference threshold.

3. The vehicle stability intelligent anticipation control method based on integrated chassis according to claim 1, characterized in that: The method further comprises: performing feature extraction on the road information to obtain road features; Performing feature fusion on the road features to obtain a road feature space; Lane classification and identification are performed on the road feature space to obtain the lane line feature deep learning network, so as to complete self-learning and self-correction of the lane line feature deep learning network.

4. The vehicle stability intelligent anticipation control method based on integrated chassis according to claim 1, characterized in that: The method further comprises: Performing lane prediction on different lane networks of the road at multiple angles to determine intersection information of the road; Constructing a detection network of the road based on the intersection information and the lane difference network; The detection network is fused to obtain the predicted curvature of the road.

5. The vehicle stability intelligent anticipation control method based on integrated chassis according to claim 1 is characterized in that: The method further comprises: determining a turning radius and a driving path of the vehicle at a next curve on the road based on the vehicle parameters; On the road, obtaining the motion status of other vehicles whose distance from the vehicle is less than a preset distance threshold; adjusting the turning radius and the driving path respectively based on the motion state of the other vehicles to obtain an adjusted turning radius and an adjusted driving path; A measured curvature of a road on which the vehicle is located is determined based on the adjusted turning radius and the adjusted driving path.

6. The vehicle stability intelligent anticipation control method based on integrated chassis according to claim 5 is characterized in that: The determining, based on the vehicle parameters, a turning radius and a driving path of the vehicle at a next curve on the road, comprises: Among the vehicle parameters, the real-time driving speed of the vehicle and the real-time friction coefficient of the road on the vehicle are obtained; Based on the real-time driving speed and the real-time friction coefficient, a turning radius and a driving path of the vehicle in a next curve of the road are determined.

7. The vehicle stability intelligent anticipation control method based on integrated chassis according to claim 1, characterized in that: The controlling the motion stability of the vehicle based on the curvature radius of the road includes: obtaining a speed limit of the vehicle on the road based on a curvature radius of the road; determining a speed difference between the speed limit and a current speed of the vehicle; performing feedforward processing on the speed difference to obtain driving information of the vehicle; Based on the driving information and a real-time friction coefficient of the road on which the vehicle is traveling, the motion stability of the vehicle is expected to be controlled.

8. The vehicle stability intelligent anticipation control system based on the integrated chassis is characterized by: The system comprises: Recognition module, which is used to identify road features based on the lane line feature deep learning network of vehicle parameters and road information; a self-learning module for determining a curvature difference between a predicted curvature corresponding to a predicted feature of the lane feature deep learning network and a measured curvature corresponding to an identified feature; determining a predicted yaw rate and a predicted vehicle sideslip angle of the vehicle based on the predicted curvature of the road and the current speed of the vehicle; determining a measured yaw rate and a measured vehicle sideslip angle of the vehicle based on the measured curvature and the current speed of the vehicle; respectively determining a yaw rate difference between the predicted yaw rate and the measured yaw rate, and a sideslip angle difference between the predicted vehicle sideslip angle and the measured vehicle sideslip angle; and re-identifying different lane networks of the road based on the yaw rate difference, the sideslip angle difference, and the curvature difference to re-determine the predicted curvature of the road, until the curvature difference is less than a preset difference threshold, thereby completing self-learning and self-correction of the lane feature deep learning network; A prediction module, configured to predict a curvature radius of the road based on the road characteristics; A control module is configured to control the motion stability of the vehicle based on the curvature radius of the road.

Images