Vehicle road state recognition and vehicle limit speed control method based on low orbit satellite technology

Abstract

The invention discloses a vehicle-road state recognition and vehicle top-speed driving control method based on a low-orbit satellite technology, and the method comprises the steps: obtaining and preprocessing a high-resolution SAR image, extracting a road boundary and a center line, and calculating the curvature of a road; estimating the real-time speed of the vehicle based on the Doppler frequency shift effect of multiple satellites, and transmitting the real-time speed to a target vehicle for vehicle speed control; evaluating the water accumulation state in front of the driving road surface according to the radar reflection wave frequency of the high-resolution SAR system, and planning the expected vehicle speed in combination with the road curvature; based on the difference value between the real-time vehicle speed and the expected vehicle speed, a control scheme is formulated, and the slip rate of each wheel is kept near the expected slip rate under the emergency acceleration / braking working condition of the vehicle through model-free self-adaptive control; the longitudinal tire force of each wheel is observed in real time based on a PID-SMO observer, the equivalent yawing moment is obtained, and auxiliary correction is conducted on the vehicle running track by means of the middle and rear wheel steering technology. According to the method, the road curvature, the road surface accumulated water and the snow surface state can be accurately recognized, the expected speed of the vehicle is dynamically adjusted to adapt to complex road conditions, and therefore the driving safety and efficiency are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle safety control, and particularly relates to a vehicle-road state recognition and vehicle limit speed control method based on low-orbit satellite technology. BACKGROUND

[0002] Accurate recognition of road curvature is the basis for optimizing traffic management and improving road safety. Traditional road curvature detection relies on vehicle-mounted LiDAR point cloud fitting (accuracy ±1.5%) or roadside camera visual recognition (night failure probability >40%), and its local perception characteristics result in a curvature estimation error of ±20% on urban expressway scenes.

[0003] Timely monitoring of road surface water is of great significance to ensure vehicle driving safety and improve emergency response capability. Existing road surface water detection mainly relies on vehicle-mounted camera spectral analysis (rain and fog weather false alarm rate >35%) or ground humidity sensor network (deployment density <1 / km), and its point-like perception mode is difficult to capture the sudden water spreading process (response delay >15 minutes). The SAR technology of low-orbit satellite can effectively identify and monitor the road surface water area under various weather conditions (such as rainy days, night, etc.). Water or ice and snow road surface will significantly reduce the friction coefficient of the road surface, increase the risk of vehicle skidding and water floating, and then may cause traffic accidents. By obtaining real-time water distribution data, traffic management departments and emergency rescue systems can timely adjust the driving speed of rescue vehicles to avoid vehicle instability caused by low adhesion road surface and ensure stable driving of vehicles on wet and slippery road surface.

[0004] Low-orbit satellite (Low Earth Orbit, LEO) has shown significant application potential in environmental monitoring, disaster warning and smart city construction in recent years due to its low orbit height (usually within a few hundred kilometers), high-resolution imaging capability and fast revisit frequency. Especially the high-resolution synthetic aperture radar (Synthetic Aperture Radar, SAR) technology, with its all-weather, all-time monitoring capability and excellent imaging accuracy, can play a key role in road curvature recognition, road surface water observation and vehicle longitudinal speed estimation. The comprehensive application of these technologies not only improves the efficiency of urban planning and traffic management, but also significantly enhances the driving safety and response speed of vehicles.

[0005] In summary, there is an urgent need for a method that can improve the accuracy of vehicle-road state recognition and the accuracy of vehicle limit speed control. SUMMARY

[0006] In order to solve the problems in the prior art, the application provides a vehicle-road state recognition and vehicle high-speed driving control method based on low-orbit satellite technology, which realizes accurate recognition of road curvature, road surface water and snow surface state by combining low-orbit satellite synthetic aperture radar (SAR) technology with image processing technology, and dynamically adjusts the expected speed of the vehicle to adapt to complex road conditions, thereby improving driving safety and efficiency.

[0007] The technical scheme adopted by the application is as follows:

[0008] A vehicle-road state recognition and vehicle high-speed driving control method based on low-orbit satellite technology, comprising the following steps:

[0009] S1: high-resolution SAR image acquisition and preprocessing, extracting road boundaries and center lines, and calculating road curvature;

[0010] S2: estimating the real-time speed of the vehicle based on the Doppler shift effect of multiple satellites and transmitting it to the target vehicle for speed control;

[0011] S3: evaluating the water accumulation state in front of the driving road surface according to the radar reflection wave frequency of the high-resolution SAR system, and planning the expected vehicle speed in combination with the road curvature;

[0012] S4: based on the difference between the real-time vehicle speed and the expected vehicle speed, formulating a control scheme, and using model-free adaptive control to keep the slip ratio of each wheel near the expected slip ratio under emergency acceleration / braking conditions;

[0013] S5: observing the longitudinal tire force of each wheel in real time based on the PID-SMO observer, obtaining the equivalent yaw moment, and using the rear wheel steering technology to assist in correcting the vehicle driving trajectory.

[0014] Further, the steps of S1 are as follows:

[0015] S1.1, converting the pixel position in the satellite image into the actual ground coordinate;

[0016] S1.2, extracting the road boundary and center line based on edge detection, feature extraction and curve fitting, calculating the road curvature, denoted as:

[0017]

[0018] wherein, are the first-order derivatives of the x coordinate and the y coordinate at time t, respectively, are the second-order derivatives of the x coordinate and the y coordinate at time t, respectively.

[0019] Furthermore, Hough transform is used for straight roads; for curved roads, polynomial fitting or spline curve fitting is used to extract the shape of the curve and obtain the road centerline.

[0020] Furthermore, the method for S2 to estimate the real-time speed of the vehicle is:

[0021] According to the Doppler frequency shift of multiple stars, the relationship with the relative velocity is obtained, which can be expressed as:

[0022]

[0023] The longitudinal velocity v of the vehicle is obtained by weighted averaging the observation data of N satellites. long :

[0024]

[0025] Since the distances between different satellites and vehicles are different, the relative speed measurement accuracy between satellites and vehicles is also different, so it is necessary to train the weight w i , so the longitudinal velocity of the vehicle is recorded as:

[0026]

[0027] Among them, v rel,i is the relative speed of the vehicle relative to the i-th satellite, λ e is the wavelength of the electromagnetic wave, θ i is the Doppler relative velocity detection angle of the i-th satellite.

[0028] Furthermore, combining SAR echo intensity, texture analysis, and spatiotemporal variations, the following formula is used to identify rain, snow, or waterlogged areas on the road:

[0029] Threshold = f(Γ mean ,Γ std ,σ texture )

[0030] Among them, Γ mean is the mean value of the echo intensity, Γ std is the standard deviation of the echo intensity, indicating the surface roughness, σ texture is the characteristic quantity of surface texture; based on this formula, different road surface types are distinguished by setting thresholds.

[0031] Furthermore, the expected vehicle speed is adjusted according to the observation results of the road water situation, and v xref Multiply by the safety factor υ for correction. When there is water on one side of the road, υ = 0.6; when there is water on both sides of the road, υ = 0.3.

[0032] Furthermore, according to the expected vehicle speed v xrefThe actual vehicle speed v x The difference between the two is used to execute the driving and braking control of the vehicle. When emergency acceleration / deceleration is required, the expected slip rate λ of each wheel of the vehicle is set. ref , and ensure that the slip rate of each wheel of the vehicle is close to the expected slip rate; and when the difference between the vehicle speed and the expected speed meets the set threshold, T ij =(v xref -v x )m / 6,T ij is the driving / braking torque of each wheel.

[0033] Furthermore, the expected slip ratio λ ref Take 0.15.

[0034] Furthermore, the expected vehicle speed v xref The actual vehicle speed v x The threshold of the difference between the values ​​is 5, which is recorded as |v xref -v x |≥5.

[0035] Furthermore, the PID observer is combined with the SMO observer to construct the PID-SMO observer, and the observer sliding mode surface S is defined as follows:

[0036]

[0037] Among them, ω ij is the actual tire speed, is the estimated value of the corresponding tire speed, which is obtained by iteratively calculating the following formula:

[0038]

[0039] Among them, k represents the current time step, τ represents the unit step length, Represents the observed value of the tire longitudinal force, J T is the tire moment of inertia, R Te is the effective radius of the tire, T bij is the driving torque of each wheel, R Te is the effective radius of the tire; the sliding mode observer is designed as follows:

[0040]

[0041] Among them, k S is the sliding mode gain, f(S) is the PID error function related to the sliding surface, which can be expressed as:

[0042]

[0043] Among them, K P , K I and K D Both are gain coefficients.

[0044] The Lyapunov function is defined as follows:

[0045]

[0046] Further derivation of the above equation gives:

[0047]

[0048] Since F xij is an unknown bounded input, there exists a sufficiently large normal number σ satisfying the following conditions:

[0049]

[0050] Then can be expressed as:

[0051]

[0052] Assume satisfies the following conditions:

[0053]

[0054] Where sgn(·) is the sign function, and satisfies the convergence condition Combining the above, we get:

[0055]

[0056] The actual observation value of F xij is:

[0057]

[0058] According to the estimation results, the additional yaw moment ΔM r produced by the difference in braking force is calculated as:

[0059]

[0060] The front wheel steering angle is controlled to track the driver's previewed desired path, and the steering yaw moment ΔM s produced by the middle and rear axle steering is calculated according to the following equation: r And ΔM s is compensated, and the steering yaw moment ΔM s is denoted as:

[0061]

[0062] Where k f , k m , and k r represent the equivalent cornering stiffness of the front, middle, and rear axles, respectively, and tf m r respectively represent the wheel track of the front, middle and rear axles, I z represents the vehicle yaw moment of inertia, a, b, c respectively represent the distance from the mass center of the front, middle and rear axles to the mass center of the vehicle, δ f , δ m , δ r respectively represent the rotation angle of the front, middle and rear axles; the equivalent rotation angles of the middle and rear axles present a fixed proportional relationship, δ m =(b / c) δ r . Make ΔM r =-ΔM s , that is, the corresponding equivalent rotation angles of the middle and rear axles are obtained.

[0063] The beneficial effects of the present application are:

[0064] (1) Through low-orbit satellite synthetic aperture radar (SAR) technology combined with image processing technology, accurate identification of road curvature, road surface water and snow surface state is realized, and the expected speed of the vehicle is dynamically adjusted to adapt to complex road conditions, thereby improving driving safety and efficiency.

[0065] (2) Using multi-satellite Doppler frequency shift technology, the vehicle longitudinal speed is estimated in real time, and through model-free adaptive control method (MFAC), the wheel slip rate is ensured to be within the expected range, realizing accurate driving and braking control in high dynamic environment.

[0066] (3) Based on the tire longitudinal force observer and the middle and rear axle auxiliary steering control strategy, the yaw moment caused by the difference in braking force is compensated, the vehicle yaw stability is optimized, and efficient steering on curves and straight roads is realized in combination with Ackerman steering constraints, significantly improving the handling and stability of the vehicle. BRIEF DESCRIPTION OF DRAWINGS

[0067] Figure 1 A vehicle road state recognition and vehicle limit speed control method based on low-orbit satellite technology. DETAILED DESCRIPTION

[0068] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0069] ​​The application provides a vehicle-road state recognition and vehicle limit speed control method based on low-orbit satellite technology, and the imaging technology of a high-resolution synthetic aperture radar based on a low-orbit satellite is used to recognize the curvature of a road, the strength of a radar reflected wave is used to identify the road surface water condition, and the Doppler frequency shift effect of a constellation of multiple satellites is used to estimate the driving speed of a target vehicle. Furthermore, a model-free adaptive control method is used to track the expected vehicle speed and ensure that the wheel slip rate is kept near the expected slip rate. Finally, the control of auxiliary steering of the middle and rear axles is completed through the estimation of the longitudinal force of each wheel.

[0070] A vehicle-road state recognition and vehicle limit speed control method based on low-orbit satellite technology. The application mainly includes the following steps:

[0071] Step 1: High-resolution SAR image acquisition and preprocessing, extracting the road boundary and center line, and calculating the road curvature.

[0072] In step 1, it is assumed that there is a conversion relationship between the pixel position (x pixel ,y pixel ) in the satellite image and the actual ground coordinate (x,y), and an affine transformation model is used for correction:

[0073]

[0074] Where a1, a2, a3, a4 are the coefficients of the transformation matrix, and b1, b2 are the translation vectors.

[0075] When extracting the road boundary and center line, image processing techniques are usually involved, including edge detection, feature extraction, and curve fitting. Here, Hough Transform is used to extract the straight line part of the road (if there is a curve, a curved model is applied):

[0076] (1) For the Hough transform of a straight line, assuming that a straight line in the image is ρ=xcosθ+ysinθ, the Hough transform formula is:

[0077] H(ρ,θ)=∫∫δ(ρ-xcosθ-ysinθ)dxdy

[0078] (2) For a curve, the shape of the curve can be accurately extracted by polynomial fitting or spline curve fitting. If a quadratic polynomial is used to fit the road center line, the equation of the center line is:

[0079] y(x)=p2x 2 +p1x+p0

[0080] where p2, p1, p0 are the fitting coefficients, θ is the angle between the specific direction and the x-axis, and δ is the Dirac delta function. Assuming that the road centerline is represented by the parametric equation r(t) = (x(t), y(t)) in the plane rectangular coordinate system, the curvature κ is an important parameter to describe the degree of road bending. The curvature can be calculated by the following formula:

[0081]

[0082] The global speed planning is based on the trajectory, and the speed of the preview trajectory is planned based on the extracted friction circle method. Therefore, the longitudinal acceleration a x and the lateral acceleration a y must satisfy the following conditions:

[0083]

[0084] where μ is the road adhesion coefficient, μ = 0.85, and σ = 0.9 is the safety factor.

[0085] v xref The required speed starts from the path s n = 50 m in front of the starting point. The preview interval s n+1 - s n = 2 m,

[0086]

[0087] where t(s n ) is the preview interval time, and a x (s n ) must satisfy the following constraint conditions:

[0088]

[0089] where k(s n ) is the road curvature.

[0090] Step 2: Based on the Doppler shift effect of multiple stars, the real-time speed of the vehicle is accurately estimated and transmitted to the target vehicle for speed control.

[0091] Assuming that there are N satellites observing a vehicle on the ground, the vehicle is driving along the road on the ground, and its speed can be represented as,

[0092]

[0093] where v long is the longitudinal speed of the vehicle (in the direction of the road), and v lat is the lateral speed. The relative motion between the satellite and the vehicle will produce a Doppler shift, and based on these frequency shifts, the speed of the vehicle can be calculated. The following assumptions are made, let is the position vector of the i-th satellite; let is the position vector of the vehicle; is the relative position vector between the i-th satellite and the vehicle; is the unit vector pointing from the satellite to the vehicle, i.e.,

[0094]

[0095] When the satellite emits electromagnetic waves and receives the signal reflected back from the ground vehicle, due to the relative motion between the vehicle and the satellite, the frequency of the echo signal is shifted, which is called the Doppler effect. The Doppler frequency shift Δf i can be expressed by the following formula:

[0096]

[0097] where: is the relative velocity of the vehicle with respect to the i-th satellite, λ e is the wavelength of the electromagnetic wave. Then the relative velocity v rel,i between the vehicle and the satellite is given by the following formula:

[0098]

[0099] where, and represent the unit vector components in the longitudinal and transverse directions from the vehicle to the satellite, respectively. Each satellite can obtain the relationship with the relative velocity by receiving the echo signal and measuring the frequency shift Δf i :

[0100]

[0101] By weighted averaging of the observation data of N satellites, the longitudinal velocity v long of the vehicle is obtained:

[0102]

[0103] Since the distances between different satellites and the vehicle are different, the relative velocity measurement accuracy between the satellite and the vehicle is also different. Therefore, in actual calculation, the weighted average method is used, in which the measurement results of satellites with closer distances are given higher weights w i .

[0104]

[0105] Step 3, according to the radar reflection wave frequency of the high-resolution SAR system, the water accumulation state in front of the driving road surface is evaluated, and the expected vehicle speed is planned in combination with the road curvature.

[0106] SAR images are generated based on the interaction of microwave signals with ground objects. Different ground surfaces (such as water, snow, dry pavement, etc.) have different reflection, scattering, and absorption characteristics of microwave signals, which can be used to identify different ground conditions. The SAR received echo signal is the electromagnetic wave reflected from the target surface. Assuming that the radar emits an electromagnetic wave E(t) as a high-frequency electromagnetic signal, it can be represented in complex form as:

[0107]

[0108] where E0 is the amplitude of the signal, ω is the frequency of the signal, t is time, and φ0 is the initial phase. The intensity of the radar reflected echo I(t) can be represented as:

[0109]

[0110] The reflection and scattering characteristics of the ground surface to the radar wave are determined by its physical properties (such as surface roughness, humidity, temperature, dielectric constant, etc.). Water area: The water surface is usually flat and has strong specular reflectivity, and the echo intensity of the radar wave is high. Snow area: Snow is a relatively rough surface, and radar waves will be scattered, with complex echo signal intensity and scattering angle. Dry pavement: The dry pavement surface is relatively flat, but rougher than the water surface, so its reflection characteristics and echo intensity are between snow and water. These factors directly affect the intensity and phase of the radar echo,

[0111] Γ = K x σ(θ) x cos(θ)

[0112] where Γ is the received echo intensity (i.e., the reflection coefficient), K is a system constant, σ(θ) is the relationship between the radar cross section (RCS) and the observation angle θ, indicating the scattering ability of the ground feature to the radar wave. θ is the radar wave incidence angle. In rain, snow, and water areas, the radar cross section (RCS) is related to the ground characteristics, and the RCS of the water area is usually high, while the RCS of the snow or wet pavement is low. In practical applications, combined with SAR echo intensity, texture analysis, and spatiotemporal changes, the following formula can be used to identify the rain, snow, or water area on the road surface:

[0113] Threshold = f(Γ mean ,Γ std ,σ texture )

[0114] where: Γ mean is the mean of the echo intensity. Γ std is the standard deviation of the echo intensity, indicating the surface roughness. σ textureis a feature of surface texture, usually extracted by texture analysis algorithm. With this formula, different road types (water, snow, dry ground, etc.) are distinguished by setting a reasonable threshold. At the same time, according to the observation results of the road water, the adjustment of the expected vehicle speed is carried out, and v xref is multiplied by the safety factor υ to be modified, when there is water on one side of the road υ = 0.6, when there is water on both sides of the road υ = 0.3.

[0115] Step 4, based on the difference between real-time vehicle speed and expected vehicle speed, a control scheme is developed, using model-free adaptive control to keep the slip ratio of each wheel near 0.15 in emergency acceleration / braking conditions. According to the expected vehicle speed v xref and the actual vehicle speed v x , the driving and braking control of the vehicle is executed, when emergency acceleration / deceleration is needed, the expected slip ratio λ ref of each wheel of the vehicle is set to 0.15, and the slip ratio of each wheel of the vehicle is ensured to be near the expected slip ratio, that is:

[0116] λ ref = 0.15, if |v xref -v x | ≥ 5

[0117] When the vehicle speed is far from the expected speed, T ij = (v xref -v x )m / 6, T ij is the driving / braking torque of each wheel. Further, the control method of the slip ratio of each wheel of the vehicle in emergency acceleration / deceleration is introduced, for the single-input single-output nonlinear system shown in each wheel driving / braking, its discrete form can be expressed as follows,

[0118]

[0119] where ζ1 and ζ2 are positive integers related to the iteration step number. At the same time, the system satisfies the following theorem assumptions outside the finite time point:

[0120] (1) The partial derivative of f(…) with respect to the (ζ1+2)th variable is continuous.

[0121] (2) It satisfies the generalized Lipschitz condition. That is, there is a positive number ξ, for any time step k1 ≠ k2 (and ≥ 0) satisfies the following inequality,

[0122] |λ ij (k1+1)-λ ij (k2+1)| ≤ ξ|T ij (k1)-T ij (k2)|, s.t. T ij (k1) ≠ Tij (k2)

[0123] Then, ΔT ij (k)≠0, the MFAC based on the tight format dynamic linearization can be expressed as the following mathematical model:

[0124] λ ij (k+1) = λ ij (k) + φ c (k)ΔT ij (k)

[0125] where the time-varying parameter vector is the pseudo partial derivative of the system. Based on the above steps, the nonlinear system shown in the above equation can be realized as an equivalent dynamic linearization representation. Because of the complexity of the system model, it is difficult to obtain the accurate true value of φ c , but the estimated value of the pseudo partial derivative can be obtained according to the input and output data of the controlled object in the iteration process. For this purpose, the corresponding criterion function is as follows,

[0126]

[0127] where κ1>0 is a weight factor. Taking the extreme value of the above equation with respect to , we get,

[0128]

[0129] where η∈(0,1] is a step factor, which can improve the flexibility of the algorithm. In order to enhance the tracking ability of the pseudo partial derivative estimation algorithm to the time-varying parameters, the algorithm reset mechanism is introduced,

[0130]

[0131] In order to balance the control accuracy and the control cost, the control input criterion function is further obtained,

[0132]

[0133] where κ2>0 is a weight factor. Further, dJ[T ij (k)] / ΔT ij (k) = 0 can be obtained,

[0134]

[0135] Where, the step factor p e (0, 1] is introduced to make the control algorithm more general. In the design of MFAC algorithm, the selection of step factor η is based on the response time and robustness requirements of the system. Through simulation analysis, η = 0.01 is selected to ensure that the system can still converge quickly under disturbance. The selection of weight factors κ1 = 0.05 and κ2 = 0.1 is based on the balance between control accuracy and control cost. Through the comparison of simulation results of different settings, it is determined that this parameter combination can most effectively suppress system oscillation and improve control accuracy.

[0136] Step 5, based on the PID-SMO observer, the longitudinal tire force of each wheel is observed in real time, the equivalent yaw moment is obtained, and the vehicle driving trajectory is assisted to correct by virtue of the rear wheel steering technology.

[0137] The PID observer has the characteristics of simplicity, reliability and easy implementation. However, in the process of emergency braking control of nonlinear system, it is difficult to adapt to various operating conditions on open road due to its fixed parameters. The SMO observer has the advantage of strong robustness, but it still needs to establish the dynamic model of the observation system in the design process, so uncertainty always exists. The combination of the two observers can further improve their ability to handle uncertainty factors. The observer sliding mode surface S is defined as follows,

[0138]

[0139] Where, is the estimated value corresponding to the tire speed, ω ij is the actual value of the tire speed, K P , K I and K D are gain coefficients, which can be obtained by iteration as follows,

[0140]

[0141] Where, k represents the current time step, τ represents the unit step, represents the tire longitudinal force observation value, J T is the tire moment of inertia, R Te is the effective radius of the tire. The design of the sliding mode observer is as follows,

[0142]

[0143] Where, k S is the sliding mode gain, and f(S) is the PID error function related to the sliding mode surface,

[0144]

[0145] The Lyapunov function is defined as follows,

[0146]

[0147] Taking the derivative of the above equation, we have

[0148]

[0149] F xij is an unknown bounded input, there exists a sufficiently large normal number σ to satisfy the following conditions,

[0150]

[0151] Then can be expressed as

[0152]

[0153] To this end, we assume satisfy the following conditions,

[0154]

[0155] where sgn(·) is the sign function. To satisfy the convergence condition of , we can get

[0156]

[0157] Then the actual observation value of F xij is

[0158]

[0159] In the design of the PID-SMO observer, the PID parameters K P = 1.5, K I = 0.1, and K D = 0.05 are set based on the Ziegler-Nichols tuning method, and are fine-tuned in combination with the actual simulation results to optimize the response speed and stability of the observer. The selection of the sliding mode gain k S = 0.2 is based on a compromise between the anti-interference ability and the chattering degree of the system. Through a series of simulation tests, the combination of these parameters is ultimately determined to exhibit good tracking accuracy and robustness under various working conditions. According to the estimation results of the tire longitudinal force observer, the additional yaw moment ΔM r caused by the braking force difference can be calculated as

[0160]

[0161] The front wheel steering angle is controlled to track the driver's previewed desired path, and the steering yaw moment ΔMs and compensating for ΔM r ,

[0162]

[0163] where k f ,k m ,k r represent the equivalent cornering stiffness of the front, middle and rear axles, t f ,t m ,t r represent the track of the front, middle and rear axles, I z represents the yaw moment of inertia of the vehicle, a, b, c represent the distance from the center of mass of the vehicle to the center of mass of the front, middle and rear axles, δ f ,δ m ,δ r represent the steering angle of the front, middle and rear axles. The steering angle of the front axle is a known quantity controlled by the driver, the control target of the middle and rear axle auxiliary steering control strategy is to reduce the steady-state lateral error caused by the difference in braking force, and the steering control is subject to Ackerman steering constraints. To this end, the equivalent steering angles of the middle and rear axles present a fixed proportional relationship, δ m =(b / c)δ r . By making ΔM r =-ΔM s , the corresponding equivalent steering angles of the middle and rear axles can be obtained

[0164] The above examples are only used to illustrate the design idea and characteristics of the present application, and the purpose is to enable those skilled in the art to understand the content of the present application and implement it, and the protection scope of the present application is not limited to the above examples. Therefore, any equivalent changes or modifications made according to the principles and design ideas disclosed by the present application are within the protection scope of the present application.

Claims

1. A vehicle-road state recognition and vehicle extreme speed driving control method based on low-orbit satellite technology, characterized in that: The steps include: S1: Acquisition and preprocessing of high-resolution SAR images, extraction of road boundaries and centerlines, and calculation of road curvature; S2: Estimate the real-time vehicle speed based on the Doppler frequency shift effect of multiple satellites and transmit it to the target vehicle for speed control; S3: Evaluate the water status ahead of the road based on the radar reflection wave frequency of the high-resolution SAR system and plan the desired vehicle speed based on the road curvature; S4: Based on the difference between the real-time vehicle speed and the desired speed, a control scheme is developed to maintain the slip ratio of each wheel near the desired slip ratio during emergency acceleration / braking conditions using model-free adaptive control. S5: Based on the PID-SMO observer, the longitudinal tire force of each wheel is observed in real time to obtain the equivalent yaw moment, and the mid- and rear-wheel steering technology is used to assist in correcting the vehicle's driving trajectory.

2. The method for vehicle-road state recognition and vehicle extreme speed driving control based on low-orbit satellite technology according to claim 1 is characterized in that: The steps of S1 are as follows: S1.

1. Convert the pixel positions in the satellite image to the actual ground coordinates; S1.

2. Extract the road boundary and centerline based on edge detection, feature extraction, and curve fitting, and calculate the road curvature, which is expressed as: in, are the first-order derivatives of the x-coordinate and y-coordinate at time t, are the second-order derivatives of the x-coordinate and y-coordinate at time t respectively.

3. The method for vehicle-road state recognition and vehicle extreme speed driving control based on low-orbit satellite technology according to claim 2 is characterized in that: For straight roads, Hough transform is used; for curved roads, polynomial fitting or spline curve fitting is used to extract the shape of the curve and obtain the road centerline.

4. The method for vehicle-road state recognition and vehicle extreme speed driving control based on low-orbit satellite technology according to claim 1 is characterized in that: The method for S2 to estimate the real-time speed of the vehicle is: According to the Doppler frequency shift of multiple stars, the relationship with the relative velocity is obtained, which can be expressed as: The longitudinal velocity v of the vehicle is obtained by weighted averaging the observation data of N satellites. long : Since the distances between different satellites and vehicles are different, the relative speed measurement accuracy between satellites and vehicles is also different, so it is necessary to train the weight w i , so the longitudinal velocity of the vehicle is recorded as: Among them, v rel,i is the relative speed of the vehicle relative to the i-th satellite, λ e is the wavelength of the electromagnetic wave, θ i is the Doppler relative velocity detection angle of the i-th satellite.

5. The method for vehicle-road state recognition and vehicle extreme speed driving control based on low-orbit satellite technology according to claim 1 is characterized in that: Combining SAR echo intensity, texture analysis, and spatiotemporal variations, the following formula is used to identify rain, snow, or waterlogged areas on the road: Threshold=f(Γ mean ,C std ,s texture ) Among them, Γ mean is the mean value of the echo intensity, Γ std is the standard deviation of the echo intensity, indicating the surface roughness, σ texture is the characteristic quantity of surface texture; based on this formula, different road surface types are distinguished by setting thresholds.

6. The method for vehicle-road state recognition and vehicle extreme speed driving control based on low-orbit satellite technology according to claim 1 is characterized in that: According to the observation results of road waterlogging, the expected vehicle speed is adjusted. xref Multiply by the safety factor υ for correction. When there is water on one side of the road, υ = 0.6; when there is water on both sides of the road, υ = 0.

3.

7. The method for vehicle-road state recognition and vehicle extreme speed driving control based on low-orbit satellite technology according to claim 1 is characterized in that: The expected speed v xref The actual vehicle speed v x The difference between the two is used to execute the driving and braking control of the vehicle. When emergency acceleration / deceleration is required, the expected slip rate λ of each wheel of the vehicle is set. ref , and ensure that the slip rate of each wheel of the vehicle is close to the expected slip rate; and when the difference between the vehicle speed and the expected speed meets the set threshold, T ij =(v xref -v x )m / 6,T ij is the driving / braking torque of each wheel.

8. The method for vehicle-road state recognition and vehicle extreme speed driving control based on low-orbit satellite technology according to claim 7 is characterized in that: Desired slip ratio λ ref Take 0.

15.

9. The method for vehicle-road state recognition and vehicle extreme speed driving control based on low-orbit satellite technology according to claim 7 is characterized in that: Expected vehicle speed v xref The actual vehicle speed v x The threshold of the difference between the values ​​is 5, which is recorded as |v xref -v x |≥5.

10. The method for vehicle-road state recognition and vehicle extreme speed driving control based on low-orbit satellite technology according to claim 1, characterized in that: The PID observer is combined with the SMO observer to construct the PID-SMO observer, and the observer sliding mode surface S is defined as follows: Among them, ω ij is the actual tire speed, is the estimated value of the corresponding tire speed, which is obtained by iteratively calculating the following formula: Among them, k represents the current time step, τ represents the unit step length, Represents the observed value of the tire longitudinal force, J T is the tire moment of inertia, R Te is the effective radius of the tire, T bij is the driving torque of each wheel, R Te is the effective radius of the tire; the sliding mode observer is designed as follows: Among them, k S is the sliding mode gain, f(S) is the PID error function related to the sliding surface, which can be expressed as: Among them, K P , K I and K D are gain coefficients; The Lyapunov function is defined as follows: Further deriving the above formula yields: Because F xij is an unknown bounded input quantity, then there exists a sufficiently large positive constant σ that satisfies the following conditions: but It can be expressed as; assumed The following conditions are met: Among them, sgn(·) is the sign function, which satisfies The convergence conditions of , combined with the above, are: Then F xij The actual observed value is: Calculate the additional yaw moment ΔM generated by the braking force difference based on the estimation results r : The front wheel steering angle is controlled by the driver's preview of the desired path. The steering yaw moment ΔM generated by the steering of the center and rear axles is obtained according to the following formula: s , and ΔM r Compensation is performed on the yaw moment ΔM s Denoted as: Among them, k f ,k m ,k r Indicates the equivalent cornering stiffness of the tires on the front, middle and rear axles, t f ,t m ,t r Respectively represent the wheelbase of the front, middle and rear axles, I z represents the vehicle's yaw moment of inertia, a, b, c represent the distances from the center of mass of the front, middle, and rear axles to the center of mass of the vehicle, respectively, and δ f ,δ m ,δ r Represents the turning angles of the front, middle, and rear axles respectively; the equivalent turning angles of the middle and rear axles present a fixed proportional relationship, δ m =(b / c)δ r ; make ΔM r =-ΔM s That is to obtain the corresponding equivalent turning angle of the middle and rear axles.

Images