A vehicle slope estimation method based on a sliding mode observer and a static working condition fusion

Abstract

This invention relates to the field of vehicle control technology, specifically to a vehicle slope estimation method based on the fusion of sliding mode observer and static condition data. The method includes: S1: acquiring relevant vehicle state information; S2: determining the vehicle's operating condition; S3: selecting a slope estimation mode: if the vehicle is stationary, selecting the stationary slope mode; if the vehicle is in motion and meets effective excitation conditions, selecting the sliding mode observation slope mode; if the vehicle is in other operating conditions, selecting the hold mode; S4: in the stationary slope mode, calculating the slope estimate under the stationary condition; in the sliding mode observation slope mode, estimating the slope estimate using the sliding mode observer; in the hold mode, holding the slope estimate from the previous cycle; S5: generating a slope estimation result from the slope estimates output in the corresponding mode using a fusion algorithm to achieve corresponding vehicle control. This invention improves the efficiency, stability, and noise resistance of vehicle slope estimation.

Description

Technical Field

[0001] This invention relates to the field of vehicle control technology, specifically to a method for estimating vehicle slope based on the fusion of sliding mode observer and static conditions. Background Technology

[0002] With the development of automotive electronics technology, the intelligence level of vehicle chassis control and longitudinal control systems is increasing. Road gradient, as a key state variable in vehicle dynamics control, plays an indispensable role in many vehicle control functions. For example, functions such as Automatic Parking (AVH), Hill Start Assist (HHC), Parking Release Control, Brake Force Distribution, and Vehicle Longitudinal Stability Control all require real-time and accurate road gradient information.

[0003] Accurate, smooth, and real-time slope information is fundamental to ensuring the safe and effective operation of the aforementioned control functions. Especially under typical operating conditions such as vehicle parking, brake release, and hill starts, the accuracy and response speed of the slope estimation results directly affect the vehicle's safety, comfort, and control quality. Therefore, how to achieve high-precision and high-stability estimation of road slope under low-cost hardware conditions has always been a key research focus for those skilled in the art.

[0004] Currently, existing vehicle slope estimation methods can be mainly divided into two categories: one is based on the vehicle's longitudinal dynamics model, and the other is based on the vehicle's kinematic relationships.

[0005] Specifically:

[0006] (1) The estimation method based on the longitudinal dynamics model of the vehicle uses parameters such as the driving force, braking force, wind resistance, rolling resistance and vehicle mass to establish dynamic equations to inversely estimate the slope. On the one hand, this type of method is highly dependent on model parameters such as vehicle mass, driving force, braking force, wind resistance coefficient and rolling resistance coefficient. However, in actual vehicle operation, the vehicle load is often in dynamic change, and the input quantities such as driving force and braking force are difficult to obtain directly with high precision through sensors, requiring estimation, which leads to a large input error. On the other hand, this type of method usually involves complex mathematical calculations, and the algorithm has a large computational load, which increases the computational burden of the controller. Moreover, due to the dependence on the accuracy of dynamic model parameters and force signals, the estimation accuracy is prone to significant decrease when the vehicle load changes or the parameters are inaccurate.

[0007] (2) The method based on vehicle kinematics estimates the slope by using the difference between the longitudinal acceleration measured by the inertial measurement unit (IMU) and the actual longitudinal acceleration of the vehicle. However, this method mainly relies on the differential operation of the acceleration signal, which is very sensitive to noise and has weak anti-interference ability. Especially under conditions of uniform vehicle speed, low dynamic excitation, or stationary conditions, due to the lack of sufficient excitation signal, this method is prone to drastic fluctuations in the slope estimation results, or even failure to converge, resulting in insufficient stability under low excitation conditions and difficulty in meeting the real-time requirements of the control system.

[0008] In summary, existing vehicle slope estimation methods have significant drawbacks: while dynamic-based methods are theoretically rigorous, they are overly dependent on the accuracy of model parameters and force signals, lack robustness, and are computationally complex; whereas kinematic-based methods, although simple in structure, have poor noise resistance and poor stability under stationary or low-dynamic excitation conditions. Therefore, there is an urgent need for a vehicle slope estimation method that is independent of model parameters and force signals, possesses strong noise resistance, and can balance estimation stability and usability under low-excitation conditions to meet the growing engineering application requirements of braking-related control systems. Summary of the Invention

[0009] To address the shortcomings of the existing technologies, the technical problem this invention aims to solve is: how to provide a vehicle slope estimation method based on the fusion of sliding mode observer and static conditions, achieving high-precision and high-stability online estimation of road slope without relying on dynamic parameters such as vehicle mass, driving force, braking force, wind resistance, and rolling resistance, thereby better meeting the application requirements of slope signals for vehicle control functions such as Automatic Parking Hold (AVH), Hill Start Assist (HHC), Electronic Stability Program (ESP), and Electronic Parking Brake (EPB).

[0010] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0011] A vehicle slope estimation method based on the fusion of sliding mode observer and static conditions includes:

[0012] S1: Obtain relevant vehicle status information;

[0013] S2: Determine the vehicle's operating condition based on relevant vehicle status information; the vehicle's operating condition includes stationary condition and driving condition.

[0014] S3: Select the corresponding slope estimation mode based on the vehicle's operating conditions: if the vehicle is stationary, select the stationary slope mode; if the vehicle is in motion and meets the effective excitation conditions, select the sliding mode observation slope mode; if the vehicle is in other operating conditions, select the hold mode.

[0015] S4: Calculate the slope estimate based on the selected slope estimation mode: In static slope mode, calculate the slope estimate under static vehicle conditions based on the vehicle longitudinal kinematics model; in sliding mode observation slope mode, estimate the slope estimate using the sliding mode observer; in hold mode, hold the slope estimate from the previous cycle.

[0016] S5: The slope estimation value output in the corresponding mode is used to generate a slope estimation result through a fusion algorithm to achieve the corresponding vehicle control.

[0017] Preferably, in step S1, the relevant status information of the vehicle includes gear information, driving force information, vehicle speed, vehicle longitudinal acceleration, and braking status.

[0018] Preferably, in step S3, other operating conditions refer to all operating conditions other than the stationary operating condition and the driving operating condition that meets the effective excitation conditions.

[0019] Preferably, in step S3, the formula for the effective excitation condition is expressed as:

[0020]

[0021] In the formula: This represents the longitudinal acceleration measurement value of the vehicle obtained through the inertial measurement unit (IMU); This indicates the preset excitation threshold; This represents the longitudinal acceleration of the vehicle calculated from the vehicle speed signal from the vehicle's CAN bus. This indicates a preset acceleration difference threshold, used to determine whether the difference between the longitudinal acceleration measured by the inertial measurement unit and the longitudinal acceleration of the vehicle calculated from the vehicle speed signal via the vehicle CAN bus is significant, thereby ensuring that the vehicle is in an effective excitation state.

[0022] Preferably, in step S4, in the static slope mode, the slope estimate under static conditions is calculated based on the vehicle's longitudinal kinematics model through the following steps:

[0023] S401: Calculate the average longitudinal acceleration of the vehicle based on a preset time window;

[0024] The formula is expressed as:

[0025]

[0026] In the formula: This represents the average longitudinal acceleration of the vehicle within the time window. Represents the first time in the time window One longitudinal acceleration measurement value; Indicates the total number of longitudinal acceleration measurements within the time window;

[0027] S402: Based on the vehicle's longitudinal kinematics module, and combined with the vehicle's average longitudinal acceleration, the unfiltered slope estimate under stationary conditions is calculated using the following formula:

[0028]

[0029] In the formula: This represents the unfiltered slope estimate under static conditions. Represents gravitational acceleration;

[0030] S403: Perform a first-order low-pass filter on the unfiltered slope estimate under stationary conditions to obtain the final slope estimate under stationary conditions.

[0031] Preferably, in step S403, the unfiltered slope estimate under static conditions is obtained using the following formula. Perform first-order low-pass filtering:

[0032]

[0033] In the formula: This represents the estimated slope value when the vehicle is stationary at the current moment. This represents the unfiltered slope estimate at the current moment; Represents the filter coefficients under static operating conditions; This represents the filtered slope estimate from the previous time step.

[0034] Preferably, in step S4, under the sliding mode slope observation mode, the slope estimate is achieved using the sliding mode observer through the following steps:

[0035] S411: Structural slope estimation error:

[0036] Based on the vehicle's longitudinal kinematics, the vehicle's longitudinal acceleration satisfies the following formula:

[0037]

[0038] In the formula: This represents the longitudinal acceleration of the vehicle calculated from the vehicle speed signal from the vehicle's CAN bus. This represents the longitudinal acceleration measurement value of the vehicle obtained through the inertial measurement unit (IMU); Indicates the slope value; Represents gravitational acceleration;

[0039] Based on longitudinal kinematics, the slope value at the current moment Calculated using the following formula:

[0040]

[0041] In the formula: This represents the longitudinal acceleration measurement value of the vehicle obtained by the inertial measurement unit at the current moment; This represents the longitudinal acceleration of the vehicle calculated from the vehicle speed signal of the vehicle's CAN bus at the current moment.

[0042] Define the slope estimation error at the current moment. for:

[0043]

[0044] In the formula: This represents the estimated slope at the current moment;

[0045] S412: Design of a slope estimate update law based on the exponential reaching law:

[0046] Treating the slope estimate as a slow variable, that is, approximately satisfying ;

[0047] The derivative of the slope estimation error Rate of change of slope estimate Satisfy the following formula:

[0048]

[0049] Selecting the sliding surface The sliding mode observer is designed using an exponential reaching law; the reaching law is designed using a hyperbolic tangent function as follows:

[0050]

[0051] Right now:

[0052]

[0053] In the formula: Represents the linear gain coefficient; Indicates the nonlinear gain coefficient; Represents the hyperbolic tangent function;

[0054] Will Substitution The rate of change of the estimated slope at the current moment is obtained. for:

[0055]

[0056] The stability of the sliding mode observer is guaranteed by the following Lyapunov analysis:

[0057] Define Lyapunov functions Taking its derivative and substituting the update rate, we get:

[0058]

[0059] Only At that time, the slope estimation error asymptotically converged to zero, and the sliding mode observer stabilized;

[0060] S413: Obtain the slope estimate through discrete integration:

[0061] In the discretization implementation, the rate of change of the slope estimate over the sampling period is... Integrate to obtain the slope estimate of the current unlimited slope;

[0062] The formula is expressed as:

[0063]

[0064] Right now:

[0065]

[0066] In the formula: This represents the estimated slope value before the current amplitude limit is reached; This represents the estimated slope of the unlimited slope at the previous moment; Indicates the sampling period;

[0067] S414: Slope estimate for the current unlimited position. Angle limiting processing is performed to constrain the slope estimation range, and the slope estimate value estimated by the sliding mode observer is obtained;

[0068] The formula is expressed as:

[0069]

[0070] In the formula: This represents the slope estimate obtained by the sliding mode observer; , These represent the preset lower and upper limits of the slope, respectively.

[0071] Preferably, in step S4, in the hold mode, i.e., the non-static slope mode and the sliding mode observation slope mode, the slope estimate is frozen, and the slope estimate estimated by the sliding mode observer at the previous moment is output. .

[0072] Preferably, in step S5, when different slope estimation modes are switched, in order to avoid abrupt changes in slope output due to differences in the internal states of different estimators, a unified gradual fusion strategy is adopted to smooth the transition of the slope estimation results; the slope estimation modes include static slope mode, sliding mode observation slope mode, and hold mode;

[0073] The unified gradient blending strategy specifically includes the following steps:

[0074] S501: Mode Switching Detection and Initial Slope Recording

[0075] When a switch in the current slope estimation mode is detected, record the initial slope estimate value at the moment of the switch. , as the starting baseline value for fusion;

[0076] S502: Calculate the blending weights for linear gradients:

[0077] During the fusion phase following mode switching, the gradual fusion weights are calculated based on the fusion duration. :

[0078]

[0079] In the formula: Indicates the fusion weights; The value of the timer is reset to zero the instant a change in operating condition is detected, and the value is incremented by the sampling period in each calculation cycle. ; Indicates the preset fusion duration; Indicates the system sampling period;

[0080] in:

[0081] (1) When When =0, The system maintains the original output mode;

[0082] (2) When = hour, The system then completes the switch to output in the target mode;

[0083] S503: Based on the initial slope estimate before mode switch Compared with the slope estimate under the current target model and fusion weight Calculate the estimated fusion slope at the current moment;

[0084] The formula is expressed as:

[0085]

[0086] In the formula: This represents the estimated fusion slope at the current moment; This represents the slope estimate under the current target model; The initial slope estimate before mode switching;

[0087] in, Depending on the current target mode being switched, the corresponding modes are as follows:

[0088] (1) Slope estimate under static conditions;

[0089] (2) Slope estimate from the sliding mode observer;

[0090] (3) The slope estimate of the previous cycle under the maintenance mode;

[0091] S504: Estimated slope at the current moment Post-processing and calibration correction are performed to improve the smoothness, dynamic constraint capability, and system consistency of the slope estimation results, so as to obtain the final output slope estimation results.

[0092] Preferably, in step S504, the post-processing and calibration correction of the fused slope estimate at the current moment are achieved through the following steps:

[0093] S5041: First-order low-pass filtering:

[0094] The estimated fusion slope at the current moment A first-order low-pass filter is performed to suppress high-frequency noise and residual jitter, and the filtered slope estimate is obtained.

[0095] The formula is expressed as:

[0096]

[0097] In the formula: This represents the filtered slope estimate at the current time. Indicates the filter coefficients; This represents the filtered output value at the previous moment;

[0098] S5042: Calculation of Slope Change Rate Constraints:

[0099] The maximum allowable change in a single period is calculated based on the preset maximum slope change rate.

[0100] The formula is expressed as:

[0101]

[0102] In the formula: Indicates the maximum permissible change in a single period; Indicates the sampling period; The preset maximum rate of change of slope;

[0103] S5043: Slope Limitation Constraints:

[0104] Based on the maximum permissible change amount in step S5042 The filtered slope estimate for the current time. Perform amplitude limiting correction to obtain the slope output value after amplitude limiting;

[0105] The formula is expressed as:

[0106]

[0107] In the formula: This represents the slope output value after the current amplitude limit; This represents the slope output value after the previous time step was limited; This represents the filtered slope estimate at the current time.

[0108] S5044: Calibration and Correction Process:

[0109] Slope output value after amplitude limiting Calibration and correction are performed to compensate for inherent system biases, resulting in the final slope estimation output.

[0110] The formula is expressed as:

[0111]

[0112] In the formula: This indicates the final slope estimation result output; This represents the slope output value after the current amplitude limit; This represents the slope gain correction coefficient, used to compensate for signal amplitude deviation; This represents the slope zero-bias correction coefficient, used to compensate for inherent offset errors.

[0113] The vehicle slope estimation method based on the fusion of sliding mode observer and static conditions in this invention has the following advantages compared with the prior art:

[0114] In the static slope mode, this invention directly and quickly calculates the slope estimate using the longitudinal acceleration signal measured by the vehicle's inertial measurement unit (IMU), which meets the control requirements of the parking and starting preparation stages and ensures that the slope is quickly ready before starting, thereby improving the efficiency of vehicle slope estimation.

[0115] This invention, during vehicle operation (i.e., in sliding mode observation of slope), constructs a sliding mode observation update law based on the error between the measured longitudinal acceleration value and the estimated actual longitudinal acceleration value. It designs an effective excitation determination mechanism to estimate the slope online under driving conditions. Furthermore, it obtains the slope estimation output through post-processing techniques such as low-pass filtering, rate-of-change limiting, and angle limiting, thereby improving the stability and noise resistance of vehicle slope estimation. Simultaneously, under low-excitation conditions, this invention switches to a hold mode, maintaining the slope estimation value from the previous cycle. This hold strategy effectively suppresses noise and abnormal fluctuations, thereby improving the stability and engineering usability of the slope output. In addition, during the switching process between stationary and driving conditions (i.e., mode switching), this invention uses a fixed-duration linear weighting method to smoothly fuse the slope estimates from the stationary slope mode and the sliding mode observation slope mode through a fusion strategy, avoiding jumps and shocks caused by mode switching, thus further improving the stability of vehicle slope estimation. In addition, in order to compensate for the systematic deviations caused by IMU (Inertial Measurement Unit) installation deviations, sensor zero bias, and algorithm model errors, this invention also performs error compensation on the slope estimation output through calibration.

[0116] In summary, this invention can achieve high-precision and high-stability online estimation of road slope without relying on dynamic parameters such as vehicle mass, driving force, braking force, wind resistance, and rolling resistance. This better meets the application requirements of vehicle control functions such as Automatic Parking Hold (AVH), Hill Start Assist (HHC), Electronic Stability Program (ESP), and Electronic Parking Brake (EPB) for slope signals. Attached Figure Description

[0117] To make the objectives, technical solutions, and advantages of the invention clearer, the invention will now be described in further detail with reference to the accompanying drawings, wherein:

[0118] Figure 1 This is a logic block diagram of a vehicle slope estimation method based on the fusion of sliding mode observer and static conditions.

[0119] Figure 2 To compare the slope estimation results when the vehicle transitions from a standstill to a constant speed.

[0120] Figure 3 The effect of slope estimation under continuous start-stop conditions with varying slope. Detailed Implementation

[0121] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but only to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0122] The following detailed explanation illustrates the specific implementation methods:

[0123] Example:

[0124] Road gradient is a critical state variable in vehicle chassis control and longitudinal control, especially in functions such as automatic parking, hill start assist, parking release control, brake force distribution, and vehicle longitudinal stability control, all of which require accurate, smooth, and real-time gradient information. Particularly under typical operating conditions such as vehicle parking, brake release, and hill start, the accuracy and response speed of gradient estimation directly affect the vehicle's safety, comfort, and control quality.

[0125] The applicant found that existing methods for estimating vehicle gradient can be mainly divided into the following two categories.

[0126] (1) Method based on vehicle longitudinal dynamics model

[0127] This type of method estimates the gradient based on the longitudinal dynamics of the vehicle, and its typical expression is:

[0128]

[0129] In the formula: Longitudinal force (the sum of driving force and braking force). Vehicle quality, Longitudinal acceleration, air density, drag coefficient, Windward area, Longitudinal speed, Gravitational acceleration, Slope value, Rolling resistance coefficient.

[0130] Common methods in this category include Kalman filtering, extended Kalman filtering, unscented Kalman filtering, and recursive least squares. While these methods can achieve online slope estimation to some extent, they generally suffer from the following problems: (1) they are highly dependent on parameters such as vehicle mass, driving force, braking force, wind resistance, and rolling resistance; (2) the estimation accuracy tends to decrease when the vehicle load changes; (3) it is difficult to obtain high-precision input quantities such as driving force and braking force in actual vehicles; and (4) the computational load of these methods is relatively large, which is not conducive to the real-time deployment of resource-constrained controllers. In summary, the common defects of these methods are: they are highly dependent on the accuracy of dynamic model parameters and force signals, their robustness is greatly affected by parameter errors, and they involve a large amount of computation.

[0131] (2) Methods based on vehicle kinematics

[0132] This type of method typically estimates the slope using the difference between the longitudinal acceleration measured by the inertial measurement unit (IMU) and the actual longitudinal acceleration of the vehicle. A typical expression for this is:

[0133]

[0134] In the formula: This represents the longitudinal acceleration of the vehicle measured by the inertial measurement unit. This represents longitudinal acceleration, estimated from the vehicle speed signal; It represents the acceleration due to gravity.

[0135] This method does not directly rely on vehicle mass parameters and is relatively simple in form, but it still has the following shortcomings: under the condition of uniform vehicle speed or low dynamic excitation, it is easy to cause fluctuations in the slope estimation results and has weak resistance to noise interference.

[0136] The applicant found that existing vehicle slope estimation methods mainly have the following problems: (1) the dynamic method is too dependent on the accuracy of model parameters and force signals, and lacks robustness; (2) although the kinetic method has a simple structure, it has weak noise resistance and insufficient stability under low excitation conditions. Therefore, it is necessary to design a vehicle slope estimation method that does not depend on model parameters and force signals, has strong noise resistance, and can take into account the estimation stability and availability under low excitation conditions, so as to meet the engineering application requirements of braking-related control systems.

[0137] In summary, this embodiment discloses a vehicle slope estimation method based on the fusion of sliding mode observer and static conditions.

[0138] like Figure 1 As shown, a vehicle slope estimation method based on the fusion of sliding mode observer and static conditions includes:

[0139] S1: Obtain relevant vehicle status information;

[0140] S2: Determine the vehicle's operating condition based on relevant vehicle status information; the vehicle's operating condition includes stationary condition and driving condition.

[0141] S3: Select the corresponding slope estimation mode based on the vehicle's operating conditions: if the vehicle is stationary, select the stationary slope mode; if the vehicle is in motion and meets the effective excitation conditions, select the sliding mode observation slope mode; if the vehicle is in other operating conditions, select the hold mode.

[0142] In this embodiment, other operating conditions refer to all operating conditions other than the stationary operating condition and the driving operating condition that meets the effective excitation conditions.

[0143] In this embodiment, the formula for the effective excitation condition is expressed as:

[0144]

[0145] In the formula: This indicates the measured value of the vehicle's longitudinal acceleration. This indicates the preset excitation threshold; This represents the longitudinal acceleration of the vehicle estimated from the change in vehicle speed. This indicates a preset acceleration difference threshold, used to determine whether the difference between the longitudinal acceleration measurement value obtained by the inertial measurement unit (IMU) and the longitudinal acceleration estimated by the vehicle speed is significant, thereby ensuring that the vehicle is in an effective excitation state.

[0146] S4: Calculate the slope estimate based on the selected slope estimation mode: In static slope mode, calculate the slope estimate under static vehicle conditions based on the vehicle longitudinal kinematics model; in sliding mode observation slope mode, estimate the slope estimate using the sliding mode observer; in hold mode, hold the slope estimate from the previous cycle.

[0147] S5: The slope estimation values ​​output in the corresponding mode are used to generate a slope estimation result through a fusion algorithm to achieve the corresponding vehicle control. The slope estimation result will be sent to the upper-level vehicle control system (such as AVH, HHC, etc.) to achieve the corresponding vehicle control.

[0148] To better illustrate the technical solution of the present invention, this embodiment will be described in more detail through the following parts.

[0149] I. Vehicle Operating Conditions

[0150] In the specific implementation process, the relevant status information of the vehicle includes gear information, driving force information, vehicle speed, vehicle longitudinal acceleration, and braking status.

[0151] Based on the relevant vehicle status information, the vehicle's operating conditions are divided as follows:

[0152] (1) Ldle: Initial idle state;

[0153] (2) Stop detected: Detected a transitional state in which the vehicle may stop (stationary detection);

[0154] (3) Stop state: static operating condition;

[0155] (4) Non-stopstate: Driving condition (non-stationary condition).

[0156] II. Static Slope Mode (Mode 1)

[0157] In the specific implementation process, under the static slope mode, the slope estimate under static conditions is calculated based on the vehicle's longitudinal kinematic relationship model through the following steps:

[0158] In static slope mode, the following steps are taken to calculate the slope estimate under static vehicle conditions based on the vehicle's longitudinal kinematics model:

[0159] S401: Calculate the average longitudinal acceleration of the vehicle based on a preset time window;

[0160] The formula is expressed as:

[0161]

[0162] In the formula: This represents the average longitudinal acceleration of the vehicle within the time window. Represents the first time in the time window One longitudinal acceleration measurement value; This indicates the total number of longitudinal acceleration measurements within the time window (i.e., the number of sampling points).

[0163] S402: Based on the vehicle's longitudinal kinematics module, and combined with the vehicle's average longitudinal acceleration, the unfiltered slope estimate under stationary conditions is calculated using the following formula:

[0164]

[0165] In the formula: This represents the unfiltered slope estimate under static conditions.

[0166] S403: To suppress transient fluctuations in the direct calculation results, a first-order low-pass filter is applied to the unfiltered slope estimate under static conditions to obtain the final slope estimate under static conditions.

[0167] Specifically, the unfiltered slope estimate under static conditions is obtained using the following formula. Perform first-order low-pass filtering:

[0168]

[0169] In the formula: This represents the estimated slope value when the vehicle is stationary at the current moment. This represents the unfiltered slope estimate at the current moment; Represents the filter coefficients under static operating conditions; This represents the filtered slope estimate from the previous time step.

[0170] Through the above-described working logic, this invention can quickly obtain slope information when the vehicle is stationary, thereby meeting the control preparation requirements for parking and starting on a slope.

[0171] III. Sliding model for slope observation (Mode 2)

[0172] In the specific implementation process, under the sliding mode slope observation mode, the slope estimate is achieved using the sliding mode observer through the following steps:

[0173] S411: Structural slope estimation error:

[0174] Based on the vehicle's longitudinal kinematics, the vehicle's longitudinal acceleration satisfies the following formula:

[0175]

[0176] In the formula: This represents the longitudinal acceleration of the vehicle calculated from the vehicle speed signal from the vehicle's CAN bus. This indicates the measured value of the vehicle's longitudinal acceleration. Indicates the slope value; Represents gravitational acceleration;

[0177] Based on the vehicle's longitudinal kinematics, the slope value at the current moment Calculated using the following formula:

[0178]

[0179] In the formula: This represents the longitudinal acceleration measurement value of the vehicle obtained by the inertial measurement unit at the current moment; This represents the longitudinal acceleration of the vehicle calculated from the vehicle speed signal of the vehicle's CAN bus at the current moment.

[0180] Define the slope estimation error at the current moment. for:

[0181]

[0182] In the formula: This represents the estimated slope at the current moment;

[0183] S412: Design of a slope estimate update law based on the exponential reaching law:

[0184] Considering that the actual road slope changes slowly over a short timescale, the slope estimate can be regarded as a slow variable, that is, it approximately satisfies... ;

[0185] Under the above assumptions, the derivative of the slope estimation error Rate of change of slope estimate Satisfy the following formula:

[0186]

[0187] Selecting the sliding surface And an exponential approach rate is used to design a sliding mode observer;

[0188] To reduce chattering, a hyperbolic tangent function is used, and the approach rate is designed as follows:

[0189]

[0190] Right now:

[0191]

[0192] In the formula: Represents the linear gain coefficient; Indicates the nonlinear gain coefficient; Represents the hyperbolic tangent function;

[0193] Will Substitution The rate of change of the estimated slope at the current moment is obtained. for:

[0194]

[0195] The stability of the sliding mode observer is guaranteed by the following Lyapunov analysis:

[0196] Define Lyapunov functions Taking its derivative and substituting the update rate, we get:

[0197]

[0198] Only At that time, the slope estimation error asymptotically converged to zero, and the sliding mode observer stabilized;

[0199] S413: Obtain the slope estimate through discrete integration:

[0200] In the discretization implementation, the rate of change of the slope estimate over the sampling period is... Integrate to obtain the slope estimate of the current unlimited slope;

[0201] The formula is expressed as:

[0202]

[0203] Right now:

[0204]

[0205] In the formula: This represents the estimated slope value before the current amplitude limit is reached; This represents the estimated slope of the unlimited slope at the previous moment; Indicates the sampling period;

[0206] S414: Slope estimate for the current unlimited position. Angle limiting processing is performed to constrain the slope estimation range, and the slope estimate value estimated by the sliding mode observer is obtained;

[0207] The formula is expressed as:

[0208]

[0209] In the formula: This represents the slope estimate obtained by the sliding mode observer; , These represent the preset lower and upper limits of the slope, respectively.

[0210] IV. Hold Mode (Mode 0)

[0211] In hold mode, i.e., non-static slope mode and sliding mode observation slope mode, the slope estimate is frozen, and the slope estimate estimated by the sliding observer at the previous time step is output. The slope output maintains the slope estimate from the sliding mode observer in the previous cycle, which can suppress noise fluctuations under low excitation.

[0212] To avoid noise dominating the estimation results under low-excitation conditions, this invention establishes an effective excitation determination and freezing mechanism: if the current longitudinal acceleration measurement value of the vehicle... satisfy Then output the slope estimate obtained by the sliding mode observer at the current time. Otherwise, output the slope estimate from the sliding mode observer at the previous time step. .

[0213] The principle behind the effective excitation determination and freezing mechanism of this invention is as follows: under conditions where the vehicle is stationary, moving at a constant speed, or experiencing low excitation, The effective information for medium slopes is relatively weak, while Noise, wheel speed differential error, and vehicle vibration account for a significant proportion of interference. If the slope estimate is continuously updated based on the residual during this stage, the output is easily driven by noise and fluctuates, reducing the stability and reliability of the slope result. Furthermore, when the vehicle transitions from deceleration to near stop, from creeping to a standstill, or is in parking mode, the actual road slope usually remains constant or changes very little for a short period. Before entering these conditions, the system has typically obtained a reliable slope estimate during a phase with effective longitudinal excitation. Therefore, for higher-level functions such as automatic parking, hill start assist, electronic parking brake, and electronic stability control, retaining the most recently obtained effective slope estimate is sufficient to meet control requirements; there is no need to continuously refresh the slope output during low-excitation phases.

[0214] Based on this, the present invention employs a freezing mechanism when the effective excitation conditions are not met, keeping the slope estimate of the previous moment unchanged, thereby improving the stability and engineering applicability of the slope output under low dynamic conditions without affecting the use of the vehicle control function.

[0215] V. Fusion Algorithm

[0216] In the specific implementation process, when different slope estimation modes are switched, in order to avoid abrupt changes in slope output due to differences in the internal states of different estimators, a unified gradual fusion strategy is adopted to smooth the transition of slope estimation results; the slope estimation modes include static slope mode, sliding mode observation slope mode and hold mode.

[0217] The unified gradient blending strategy specifically includes the following steps:

[0218] S501: Mode Switching Detection and Initial Slope Recording

[0219] When a switch in the current slope estimation mode is detected, record the initial slope estimate value at the moment of the switch. , as the starting baseline value for fusion;

[0220] S502: Calculate the blending weights for linear gradients:

[0221] During the fusion phase following mode switching, the gradual fusion weights are calculated based on the fusion duration. :

[0222]

[0223] In the formula: Indicates the fusion weights; The value of the timer is reset to zero the instant a change in operating condition is detected, and the value is incremented by the sampling period in each calculation cycle. ; Indicates the preset fusion duration; Indicates the system sampling period;

[0224] in:

[0225] (1) When When =0, The system maintains the original output mode;

[0226] (2) When = hour, The system then completes the switch to output in the target mode;

[0227] S503: Based on the initial slope estimate before mode switch Compared with the slope estimate under the current target model and fusion weight Calculate the estimated fusion slope at the current moment;

[0228] The formula is expressed as:

[0229]

[0230] In the formula: This represents the estimated fusion slope at the current moment; This represents the slope estimate under the current target model; The initial slope estimate before mode switching;

[0231] in, Depending on the current target mode being switched, the corresponding modes are as follows:

[0232] (1) Slope estimate under static conditions;

[0233] (2) Slope estimate from the sliding mode observer;

[0234] (3) The slope estimate of the previous cycle under the maintenance mode;

[0235] S504: Estimated slope at the current moment Post-processing and calibration correction are performed to improve the smoothness, dynamic constraint capability, and system consistency of the slope estimation results, resulting in the final output slope estimation result. The fusion process is completed within a preset time, achieving a smooth transition without jumps or shocks from static slope to sliding mode observed slope.

[0236] VI. Post-processing and Calibration

[0237] To meet the requirements of the vehicle control system for the smoothness and availability of the slope input signal, this invention performs post-processing on the slope estimation results.

[0238] Specifically, the post-processing and calibration correction of the fused slope estimate at the current moment are achieved through the following steps:

[0239] S5041: First-order low-pass filtering:

[0240] The estimated fusion slope at the current moment A first-order low-pass filter is performed to suppress high-frequency noise and residual jitter, and the filtered slope estimate is obtained.

[0241] The formula is expressed as:

[0242]

[0243] In the formula: This represents the filtered slope estimate at the current time. Indicates the filter coefficients; This represents the filtered output value at the previous moment;

[0244] S5042: Calculation of Slope Change Rate Constraints:

[0245] The maximum allowable change in a single period is calculated based on the preset maximum slope change rate.

[0246] The formula is expressed as:

[0247]

[0248] In the formula: Indicates the maximum permissible change in a single period; Indicates the sampling period; The preset maximum rate of change of slope;

[0249] S5043: Slope Limitation Constraints:

[0250] Based on the maximum permissible change amount in step S5042 The filtered slope estimate for the current time. Perform amplitude limiting correction to obtain the slope output value after amplitude limiting;

[0251] The formula is expressed as:

[0252]

[0253] In the formula: This represents the slope output value after the current amplitude limit; This represents the slope output value after the previous time step was limited; This represents the filtered slope estimate at the current time.

[0254] S5044: Calibration and Correction Process:

[0255] Slope output value after amplitude limiting Calibration and correction are performed to compensate for inherent system biases, resulting in the final slope estimation output.

[0256] The formula is expressed as:

[0257]

[0258] In the formula: This indicates the final slope estimation result output; This represents the slope output value after the current amplitude limit; This represents the slope gain correction coefficient, which is related to the magnitude of the acceleration acquired by the IMU and is used to compensate for signal amplitude deviation. This represents the slope zero-bias correction coefficient, used to compensate for inherent errors such as sensor zero bias and installation deviation.

[0259] VII. Experimental Instructions

[0260] To better illustrate the technical solution of the present invention, the following experiment is disclosed in this embodiment.

[0261] Figure 2 The paper presents a comparison of the slope estimation performance of the present invention under the condition of a vehicle transitioning from a standstill to constant speed. The test scenario is set as follows: the vehicle starts on a 4% slope and then maintains a constant speed. During braking, the excitation is insufficient, and the sliding mode algorithm estimation value is not updated.

[0262] Figure 3 The effect of slope estimation under the condition of continuous start-stop operation with varying slope is shown. This condition simulates the process of a vehicle starting from a standstill on a 3.2% slope, driving to a stop on a 4% slope, and then starting again on a 4% slope.

[0263] T0-T1 (3.2% ramp stationary): The fusion algorithm output converges quickly to 3.2%.

[0264] T1-T2 (gradient change while driving): The gradient smoothly follows the actual change, transitioning from 3% to 4% without overshoot.

[0265] T2-T3 (Braking to a standstill): Insufficient excitation during braking, maintaining the previous value of 3.8%.

[0266] After T3 (4% ramp restart): stationary recognition reset, fusion algorithm output 4.0%.

[0267] The present invention has the following beneficial effects:

[0268] In the static slope mode, this invention directly and quickly calculates the slope estimate using the longitudinal acceleration signal measured by the vehicle's inertial measurement unit (IMU), which meets the control requirements of the parking and starting preparation stages and ensures that the slope is quickly ready before starting, thereby improving the efficiency of vehicle slope estimation.

[0269] This invention, during vehicle operation (i.e., in sliding mode observation of slope), constructs a sliding mode observation update law based on the error between the measured longitudinal acceleration value and the estimated actual longitudinal acceleration value. It designs an effective excitation determination mechanism to estimate the slope online under driving conditions. Furthermore, it obtains the slope estimation output through post-processing techniques such as low-pass filtering, rate-of-change limiting, and angle limiting, thereby improving the stability and noise resistance of vehicle slope estimation. Simultaneously, under low-excitation conditions, this invention switches to a hold mode, maintaining the slope estimation value from the previous cycle. This hold strategy effectively suppresses noise and abnormal fluctuations, thereby improving the stability and engineering usability of the slope output. In addition, during the switching process between stationary and driving conditions (i.e., mode switching), this invention uses a fixed-duration linear weighting method to smoothly fuse the slope estimates from the stationary slope mode and the sliding mode observation slope mode through a fusion strategy, avoiding jumps and shocks caused by mode switching, thus further improving the stability of vehicle slope estimation. In addition, in order to compensate for the systematic deviations caused by IMU (Inertial Measurement Unit) installation deviations, sensor zero bias, and algorithm model errors, this invention also performs error compensation on the slope estimation output through calibration.

[0270] In summary, this invention can achieve high-precision and high-stability online estimation of road slope without relying on dynamic parameters such as vehicle mass, driving force, braking force, wind resistance, and rolling resistance. This better meets the application requirements of vehicle control functions such as Automatic Parking Hold (AVH), Hill Start Assist (HHC), Electronic Stability Program (ESP), and Electronic Parking Brake (EPB) for slope signals.

[0271] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit the technical solutions. Those skilled in the art should understand that any modifications or equivalent substitutions to the technical solutions of the present invention without departing from the spirit and scope of the present invention should be covered within the scope of the claims of the present invention.

Claims

1. A method for estimating vehicle slope based on the fusion of sliding mode observer and static conditions, characterized in that, include: S1: Obtain relevant vehicle status information; S2: Determine the vehicle's operating condition based on relevant vehicle status information; The operating conditions of a vehicle include stationary conditions and driving conditions; S3: Select the corresponding slope estimation mode based on the vehicle's operating conditions: if the vehicle is stationary, select the stationary slope mode; if the vehicle is in motion and meets the effective excitation conditions, select the sliding mode observation slope mode; if the vehicle is in other operating conditions, select the hold mode. S4: Calculate the slope estimate based on the selected slope estimation mode: In the static slope mode, calculate the slope estimate under static vehicle conditions based on the vehicle longitudinal kinematics relationship model; In sliding mode, the slope estimate is estimated using a sliding mode observer; in hold mode, the slope estimate from the previous period is held. S5: The slope estimation value output in the corresponding mode is used to generate a slope estimation result through a fusion algorithm to achieve the corresponding vehicle control.

2. The vehicle slope estimation method based on the fusion of sliding mode observer and static condition as described in claim 1, characterized in that: In step S1, the relevant status information of the vehicle includes gear information, driving force information, vehicle speed, vehicle longitudinal acceleration, and braking status.

3. The vehicle slope estimation method based on the fusion of sliding mode observer and static condition as described in claim 1, characterized in that: In step S3, other operating conditions refer to all operating conditions other than the stationary operating condition and the driving operating condition that meets the effective excitation conditions.

4. The vehicle slope estimation method based on the fusion of sliding mode observer and static condition as described in claim 1, characterized in that: In step S3, the formula for the effective excitation condition is expressed as: In the formula: This represents the measured value of the vehicle's longitudinal acceleration. This indicates the preset excitation threshold; This indicates the vehicle's longitudinal acceleration; This indicates the preset acceleration difference threshold.

5. The vehicle slope estimation method based on the fusion of sliding mode observer and static condition as described in claim 1, characterized in that: In step S4, under static slope mode, the slope estimate under static conditions is calculated based on the vehicle's longitudinal kinematics model through the following steps: S401: Calculate the average longitudinal acceleration of the vehicle based on a preset time window; The formula is expressed as: In the formula: This represents the average longitudinal acceleration of the vehicle within the time window. Represents the first time in the time window One longitudinal acceleration measurement value; Indicates the total number of longitudinal acceleration measurements within the time window; S402: Based on the vehicle's longitudinal kinematics module, and combined with the vehicle's average longitudinal acceleration, the unfiltered slope estimate under stationary conditions is calculated using the following formula: In the formula: This represents the unfiltered slope estimate under static conditions. Represents gravitational acceleration; S403: Perform a first-order low-pass filter on the unfiltered slope estimate under stationary conditions to obtain the final slope estimate under stationary conditions.

6. The vehicle slope estimation method based on the fusion of sliding mode observer and static condition as described in claim 5, characterized in that: In step S403, the unfiltered slope estimate under static conditions is obtained using the following formula. Perform first-order low-pass filtering: In the formula: This represents the estimated slope value when the vehicle is stationary at the current moment. This represents the unfiltered slope estimate at the current moment; Represents the filter coefficients under static operating conditions; This represents the filtered slope estimate from the previous time step.

7. The vehicle slope estimation method based on the fusion of sliding mode observer and static condition as described in claim 1, characterized in that: In step S4, under the sliding mode slope observation mode, the slope estimate is obtained using the sliding mode observer through the following steps: S411: Structural slope estimation error: Based on the vehicle's longitudinal kinematics, the vehicle's longitudinal acceleration satisfies the following formula: In the formula: This indicates the vehicle's longitudinal acceleration; This represents the measured value of the vehicle's longitudinal acceleration. Indicates the slope value; Represents gravitational acceleration; Based on the vehicle's longitudinal kinematics, the slope value at the current moment Calculated using the following formula: In the formula: This represents the measured value of the vehicle's longitudinal acceleration. This indicates the vehicle's longitudinal acceleration; Define the slope estimation error at the current moment. for: In the formula: This represents the estimated slope at the current moment; S412: Design of a slope estimate update law based on the exponential reaching law: Treating the slope estimate as a slow variable, that is, approximately satisfying ; The derivative of the slope estimation error Rate of change of slope estimate Satisfy the following formula: Selecting the sliding surface And an exponential approach rate is used to design a sliding mode observer; Using the hyperbolic tangent function, the approach rate is designed as follows: Right now: In the formula: Represents the linear gain coefficient; Indicates the nonlinear gain coefficient; Represents the hyperbolic tangent function; Will Substitution The rate of change of the estimated slope at the current moment is obtained. for: S413: Obtain the slope estimate through discrete integration: In the discretization implementation, the rate of change of the slope estimate over the sampling period is... Integrate to obtain the slope estimate of the current unlimited slope; The formula is expressed as: Right now: In the formula: This represents the estimated slope at the current moment before the slope is limited; This represents the estimated slope of the unlimited slope at the previous moment; Indicates the sampling period; S414: Slope estimate for the current unlimited position. Angle limiting processing is performed to obtain the slope estimate value estimated by the sliding mode observer; The formula is expressed as: In the formula: This represents the slope estimate obtained by the sliding mode observer; , These represent the preset lower and upper limits of the slope, respectively.

8. The vehicle slope estimation method based on the fusion of sliding mode observer and static condition as described in claim 1, characterized in that: In step S4, in hold mode, the slope estimate is frozen, and the slope estimate estimated by the sliding mode observer at the previous time step is output. .

9. The vehicle slope estimation method based on the fusion of sliding mode observer and static condition as described in claim 1, characterized in that: In step S5, when different slope estimation modes are switched, a unified gradual fusion strategy is used to smooth the transition of the slope estimation results. The unified gradient blending strategy specifically includes the following steps: S501: Mode Switching Detection and Initial Slope Recording When a switch in the current slope estimation mode is detected, record the initial slope estimate value at the moment of the switch. , as the starting baseline value for fusion; S502: Calculate the blending weights for linear gradients: During the fusion phase following mode switching, the gradual fusion weights are calculated based on the fusion duration. : In the formula: Indicates the fusion weights; The value of the timer is reset to zero the instant a change in operating condition is detected, and the value is incremented by the sampling period in each calculation cycle. ; Indicates the preset fusion duration; Indicates the system sampling period; in: (1) When When =0, ; (2) When = hour, ; S503: Based on the initial slope estimate before mode switch Compared with the slope estimate under the current target model and fusion weight Calculate the estimated fusion slope at the current moment; The formula is expressed as: In the formula: This represents the estimated fusion slope at the current moment; This represents the slope estimate under the current target model; The initial slope estimate before mode switching; S504: Estimated slope at the current moment Post-processing and calibration correction are performed to obtain the final slope estimation result.

10. The vehicle slope estimation method based on the fusion of sliding mode observer and static condition as described in claim 9, characterized in that: In step S504, the post-processing and calibration correction of the fused slope estimate at the current moment are achieved through the following steps: S5041: First-order low-pass filtering: The estimated fusion slope at the current moment Perform a first-order low-pass filter to obtain the filtered slope estimate. The formula is expressed as: In the formula: This represents the filtered slope estimate at the current time. Indicates the filter coefficients; This represents the filtered output value at the previous moment; S5042: Calculation of Slope Change Rate Constraints: The maximum allowable change in a single period is calculated based on the preset maximum slope change rate. The formula is expressed as: In the formula: Indicates the maximum permissible change in a single period; Indicates the sampling period; The preset maximum rate of change of slope; S5043: Slope Limitation Constraints: Based on the maximum permissible change amount in step S5042 The filtered slope estimate for the current time. Perform amplitude limiting correction to obtain the slope output value after amplitude limiting; The formula is expressed as: In the formula: This represents the slope output value after the current amplitude limit; This represents the slope output value after the previous time step was limited; This represents the filtered slope estimate at the current time. S5044: Calibration and Correction Process: Slope output value after amplitude limiting The calibration and correction are performed to obtain the final slope estimation result. The formula is expressed as: In the formula: This indicates the final slope estimation result output; This represents the slope output value after the current amplitude limit; This represents the slope gain correction coefficient, used to compensate for signal amplitude deviation; This represents the slope zero-bias correction coefficient, used to compensate for inherent offset errors.

Images