Full-condition road slope estimation method, system, device and storage medium

Abstract

The application discloses a full-working-condition road slope estimation method, system and device and a storage medium, and relates to the technical field of vehicle engineering and intelligent transportation. According to the constructed vehicle longitudinal dynamics model, the system state equation and the system observation equation of the road slope estimation under the full working condition are determined, the state variables of the system state equation include the longitudinal speed and the road slope angle, then the system error compensation equation of the minimum model error criterion is determined according to the system observation equation, the volume Kalman filtering algorithm and the minimum model error criterion algorithm are combined, the state value is updated according to the compensation value output by the system state equation, the system observation equation and the system error compensation equation, the error compensation in the robust estimation process of the road slope under the full working condition is realized, and the road slope estimation has the advantages of strong real-time performance, high precision and strong robustness.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of vehicle engineering and intelligent transportation technology, and in particular to a full-working-condition road slope estimation method, system, device and storage medium. BACKGROUND

[0002] Road slope information is an important part of vehicle energy saving and emission reduction and active safety control. Vehicles travel on complex and changeable roads, and real-time, rapid and accurate estimation of road slope is of great practical significance for vehicle stability control, smooth gear shifting control and energy saving and emission reduction. For example, heavy vehicles travel in mountainous areas, and winding roads, continuous turns and slopes are important parts of the road. Real-time acquisition of road slope changes and use of the same as system input for real-time planning of vehicle speed and gear information can not only reduce engine fuel consumption, but also improve the driving experience of the vehicle.

[0003] In current research on road slope estimation, more attention is paid to longitudinal road slope estimation and road slope estimation during turning. The horizontal dynamics equation is often used to estimate the horizontal speed and then obtain the horizontal inertia force, and the magic tire model is used to estimate the horizontal tire force. The estimation of horizontal speed in the above research not only increases the complexity of horizontal road slope estimation, but also makes it difficult to ensure the accuracy of horizontal speed estimation. The error transfer greatly affects the accuracy of horizontal slope estimation. In the estimation of horizontal tire force, the center of mass of the heavy vehicle shifts after loading. The center of mass is an unknown quantity, and the center of mass of the heavy vehicle becomes a difficult point. Through analysis, it is found that the existing slope estimation algorithm based on horizontal dynamics equation has high complexity and poor estimation accuracy. When estimating the road slope, it is assumed that the sensor measurement value is not disturbed. In the actual transportation process of heavy vehicles, abnormal value interference of sensor measurement value is inevitable. When the sensor measurement value is abnormal, the above algorithm does not have robustness. In addition, the above scheme only focuses on single longitudinal slope estimation or horizontal slope estimation, and cannot realize road slope estimation in the full-working-condition scene of vehicle transportation process (which is roughly divided into longitudinal road, longitudinal road (including slope), curved road and curved road (including slope) and several coupled combinations). SUMMARY

[0004] The present application aims to at least solve one of the technical problems existing in the prior art. To this end, the present application provides a full-working-condition road slope estimation method, system, device and storage medium, which can improve the accuracy of road slope estimation in full-working-condition.

[0005] In one aspect, the present application embodiment provides a full-working-condition road slope estimation method, comprising the following steps:

[0006] Obtaining a vehicle longitudinal dynamics model;

[0007] determining a system state equation and a system observation equation of road slope estimation in all working conditions according to the vehicle longitudinal dynamics model, wherein a state variable of the system state equation comprises a longitudinal velocity and a road slope angle;

[0008] determining a system error compensation equation of minimum model error criterion according to the system observation equation;

[0009] updating a state value according to a compensation value output by the system state equation, the system observation equation and the system error compensation equation based on a combination of a cubature Kalman filtering algorithm and a minimum model error criterion algorithm.

[0010] According to some embodiments of the present application, the vehicle longitudinal dynamics model is represented as:

[0011]

[0012] wherein, represents a longitudinal acceleration, m represents a total mass of a heavy vehicle, T tq represents a driving torque, i g represents a transmission ratio, i0 represents a final transmission ratio, η represents a transmission efficiency, r e represents an effective rolling radius of a wheel, C d represents an air resistance coefficient, A represents a windward area, ρ represents an air density, g represents a gravitational acceleration, f represents a rolling resistance coefficient, and β represents a road slope angle.

[0013] According to some embodiments of the present application, the system state equation and the system observation equation of road slope estimation in all working conditions according to the vehicle longitudinal dynamics model comprise the following steps:

[0014] determining a time-varying parameter vehicle dynamics model according to the vehicle longitudinal dynamics model;

[0015] determining a state variable equation according to the time-varying parameter vehicle dynamics model;

[0016] determining a system state equation according to the state variable equation and a state error equation;

[0017] determining an observation variable equation according to the longitudinal velocity and the longitudinal acceleration;

[0018] determining a system observation equation according to the observation variable equation and an observation error equation.

[0019] According to some embodiments of the present application, the compensation value output by the system error compensation equation in the minimum model error criterion is determined by the following steps:

[0020] first-order Taylor expansion is performed on the system observation equation in the error compensation equation;

[0021] Partial derivatives of the error compensation equation after the first-order Taylor expansion are taken with respect to the system error, and a compensation value is determined based on a minimum value principle.

[0022] According to some embodiments of the present application, the combination of the cubature Kalman filter algorithm and the minimum model error criterion algorithm, and the updating of the state value according to the compensation value output by the system state equation, the system observation equation and the system error compensation equation comprises the following steps:

[0023] Time updating is performed on the compensation value output by the system state equation and the system error compensation equation to obtain a state prior estimate value and update a state estimation covariance matrix;

[0024] Measurement updating is performed on the updated state estimation covariance matrix, the system observation equation and the compensation value to obtain a measurement prediction value and calculate a measurement covariance matrix;

[0025] Innovation calculation and standardization are performed on the measurement prediction value and the measurement covariance matrix to obtain a detection function;

[0026] The detection function is segmented by using multiple quantile points of the chi-square distribution, and corresponding output state values are determined according to different segmentation results.

[0027] According to some embodiments of the present application, the determination of the corresponding output state values according to different segmentation results comprises the following steps:

[0028] For the detection function segment less than the first quantile point, the state prior estimate value is updated according to a cubature Kalman filter gain matrix to obtain an output state value;

[0029] For the detection function segment between the first quantile point and the second quantile point, the measurement noise in the measurement updating process is corrected to correct the measurement covariance matrix, and then the state prior estimate value is updated according to the cubature Kalman filter gain matrix to obtain an output state value;

[0030] For the detection function segment greater than the second quantile point, no measurement updating is performed, and the state prior estimate value is taken as the output state value.

[0031] According to some embodiments of the present application, the updating of the state prior estimate value according to the cubature Kalman filter gain matrix to obtain an output state value comprises the following steps:

[0032] The cubature Kalman filter gain matrix is multiplied by a prediction error in the measurement updating process to obtain a state correction value;

[0033] The state priori estimation value is added to the state correction value to obtain an output state value.

[0034] In another aspect, the embodiments of the present application also provide a full-condition road slope estimation system, comprising:

[0035] A first module is configured to acquire a vehicle longitudinal dynamics model;

[0036] A second module is configured to determine a system state equation and a system observation equation of road slope estimation in full condition according to the vehicle longitudinal dynamics model, wherein state variables of the system state equation include longitudinal velocity and road slope angle;

[0037] A third module is configured to determine a system error compensation equation of minimum model error criterion according to the system observation equation;

[0038] A fourth module is configured to update state values based on a combination of cubature Kalman filtering algorithm and minimum model error criterion algorithm according to compensation values output by the system state equation, the system observation equation and the system error compensation equation.

[0039] In another aspect, the embodiments of the present application also provide a full-condition road slope estimation device, comprising:

[0040] At least one processor;

[0041] At least one memory configured to store at least one program;

[0042] When the at least one program is executed by the at least one processor, the at least one processor is caused to implement the full-condition road slope estimation method as described above.

[0043] In another aspect, the embodiments of the present application also provide a computer readable storage medium, which stores computer executable instructions for causing a computer to execute the full-condition road slope estimation method as described above.

[0044] The technical solution of the present application has at least one of the following advantages or beneficial effects: according to a vehicle longitudinal dynamics model, a system state equation and a system observation equation of road slope estimation in full condition are determined, state variables of the system state equation include longitudinal velocity and road slope angle, then a system error compensation equation of minimum model error criterion is determined according to the system observation equation, based on a combination of cubature Kalman filtering algorithm and minimum model error criterion algorithm, state values are updated according to compensation values output by the system state equation, the system observation equation and the system error compensation equation, error compensation is implemented in a road slope robust estimation process in full condition, and the road slope estimation has strong real-time performance, high precision and strong robustness. BRIEF DESCRIPTION OF DRAWINGS

[0045] Figure 1 is a full working condition road slope estimation method flowchart provided by the embodiment of the application;

[0046] Figure 2 is a vehicle force analysis schematic diagram provided by the embodiment of the application;

[0047] Figure 3 is a road slope estimation process schematic diagram provided by the embodiment of the application;

[0048] Figure 4 is a full working condition road slope estimation device schematic diagram provided by the embodiment of the application. DETAILED DESCRIPTION

[0049] The embodiments of the present application are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar notations represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be understood as a limitation of the present application.

[0050] In the description of the present application, it should be understood that the orientation description, such as the orientation or position relationship indicated by up, down, left, right, etc. is based on the orientation or position relationship shown in the drawings, and is only for the convenience of describing the present application and simplifying the description, and therefore cannot be understood as a limitation of the present application. The indicated device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present application.

[0051] In the description of the present application, if there is a description of first, second, etc. only for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features or the sequence of indicated technical features.

[0052] The embodiment of the present application provides a full working condition road slope estimation method, the full working condition road slope estimation method in the embodiment of the present application can be applied to the processor of the vehicle, and the vehicle includes a CAN bus acquisition device, an accelerometer sensor and a processor. The CAN bus acquisition device and the accelerometer sensor are connected in sequence through a wired mode, the accelerometer sensor is installed at the empty mass center position of the heavy vehicle, and the accelerometer sensor is mainly used for collecting the longitudinal acceleration of the heavy vehicle and transmitting to the processor through the CAN bus acquisition device. The CAN bus acquisition device collects the engine torque through the CAN longitudinal acquisition.

[0053] Referring to Figure 1 , the full working condition road slope estimation method of the embodiment of the present application includes but is not limited to the following steps:

[0054] S110, acquire a vehicle longitudinal dynamics model;

[0055] S120, determine a system state equation and a system observation equation of road slope estimation in all working conditions according to the vehicle longitudinal dynamics model, wherein the state variable of the system state equation comprises a longitudinal velocity and a road slope angle;

[0056] S130, determine a system error compensation equation of a minimum model error criterion according to the system observation equation;

[0057] S140, update the state value according to the compensation value output by the system state equation, the system observation equation and the system error compensation equation based on the combination of the cubature Kalman filtering algorithm and the minimum model error criterion algorithm.

[0058] In step S110, refer to Figure 2 , the vehicle longitudinal dynamics model is determined by the following process:

[0059] When the heavy vehicle is in transportation, the longitudinal dynamics model of the heavy vehicle is constructed based on Newton's second law, wherein the longitudinal dynamics model is composed of driving force F force (F force =T tq i g i0η / r e ), air resistance rolling resistance F roll (F roll =mgsinβ) and slope resistance F β (F β =mgsinβ), specifically, the longitudinal dynamics model of the heavy vehicle is shown in formula (1):

[0060]

[0061] wherein, is the longitudinal acceleration, m is the total mass of the heavy vehicle, T tq is the driving torque, i g is the transmission ratio, i0 is the final transmission ratio, η is the transmission efficiency, r e is the effective rolling radius of the wheel, C d is the wind resistance coefficient, A is the windward area, ρ is the air density, g is the gravitational acceleration, f is the rolling resistance coefficient, and β is the road slope angle.

[0062] According to the road slope index of the highway route design specification, the road slope is generally small, so it is assumed that cosβ≈1 and sinβ≈tanβ≈β, and the longitudinal dynamics model of the heavy vehicle can be converted into formula (2) shown as follows:

[0063]

[0064] In the step S120, the system state equation and the system observation equation of the road slope estimation in all working conditions are determined according to the vehicle longitudinal dynamics model, and the specific steps include the following steps:

[0065] In step S210, a time-varying parameter vehicle dynamics model is determined according to the vehicle longitudinal dynamics model;

[0066] In step S220, a state variable equation is determined according to the time-varying parameter vehicle dynamics model;

[0067] In step S230, a system state equation is determined according to the state variable equation and the state error equation;

[0068] In step S240, an observation variable equation is determined according to the longitudinal velocity and the longitudinal acceleration;

[0069] In step S250, a system observation equation is determined according to the observation variable equation and the observation error equation.

[0070] Specifically, in order to accurately represent the state change of the heavy vehicle, a time-varying parameter vehicle dynamics model is established according to the vehicle longitudinal dynamics model, and the time-varying parameter vehicle dynamics model is shown in formula group (3):

[0071]

[0072] Wherein, F xij , F yij are the longitudinal force and the lateral force of the i-th and j-th tires of the vehicle, i=f, r, j=l, r, f represents the front wheel position, r represents the rear wheel position, l represents the left wheel position, and r represents the right wheel position; F roll is the rolling resistance, F aero is the air resistance, F β is the slope resistance, δ fl is the front wheel steering angle, v y is the lateral velocity of the heavy vehicle, and r z is the yaw rate of the heavy vehicle.

[0073] Considering that the heavy vehicle longitudinal dynamics model only focuses on the total driving force of the heavy vehicle tires and no longer focuses on the force of each tire, and by means of this idea, formula group (3) is integrated and combined to further obtain a time-varying parameter vehicle dynamics model as shown in formula (4):

[0074]

[0075] Wherein, ΔF represents the model error.

[0076] Considering the model error ΔF, the system state equation of the state observation system is shown in formula (5):

[0077]

[0078] The state variable in the system state equation is defined as x(k) = [v x , β1], f(x(k), u(k)) = [f1, f2] T , and specifically, the state variable equation f(x(k), u(k)) is expressed as shown in equation (6):

[0079]

[0080] The E s (k) in the system state equation represents a state error, which is composed of two parts of error and process noise, and the state error equation is shown in equation (7):

[0081] E s (k) = G(k)d(k) + w(k); (7)

[0082] where w(k) is the process noise, d(k) represents the tire error vector of the heavy vehicle under the full working condition road at time k, and G(k) is the state error matrix. The state error matrix G(k) is shown in equation (8):

[0083]

[0084] where δ f is the front wheel steering angle.

[0085] The tire error vector d(k) is shown in equation (9):

[0086] d(k) = [ΔF xfl , ΔF xfr , ΔF yfl , ΔF yfr , ΔF xrl , ΔF xrr ] T ; (9)

[0087] where ΔF xfl is the longitudinal tire force error component of the left side of the front axle of the heavy vehicle, ΔF xfr is the longitudinal tire force error component of the right side of the front axle of the heavy vehicle, ΔF yfl is the lateral tire force component of the left side of the front axle of the heavy vehicle, ΔF yfr is the lateral tire force component of the right side of the front axle of the heavy vehicle, ΔF xrl is the longitudinal tire force component of the left side of the rear axle of the heavy vehicle, and ΔF xrl is the longitudinal tire force component of the right side of the rear axle of the heavy vehicle.

[0088] The system observation equation of the state observation system is shown in equation (10):

[0089] z(k) = h(x(k), u(k)) + E O (k); (10)

[0090] In the system observation equation, the observation variable is defined as z(k) = [v x , a x ], a x represents the longitudinal acceleration, h(x(k), u(k)) = [h1, h2] T , and the observation variable equation h(x(k), u(k)) is shown in equation (11):

[0091]

[0092] In the system observation equation, E O (k) is the observation error, which is composed of error and measurement noise, and the observation error equation is shown in equation (12):

[0093] E o (k) = M(k)d(k) + v(k); (12)

[0094] where v(k) is the measurement noise, and M(k) is the measurement error matrix, and the measurement error matrix M(k) is shown in equation (13):

[0095]

[0096] Specifically, the system state equation is discretized by the forward Euler method, and the discretized vehicle dynamics model difference equation of the heavy vehicle under the full working condition road scene is shown in equation (14):

[0097]

[0098] In order to minimize the influence of system error on system observation of model uncertainty under full working condition road, the performance function of minimum model error compensator is defined, that is, the system error compensation equation is shown in equation (15):

[0099]

[0100] where J[d(k)] represents the performance function of system error at time k, represents the observation estimation value at time k, R represents the measurement noise, and Y is a semi-positive definite error weight matrix.

[0101] In step S130, the specific steps of calculating the compensation value based on the system error compensation equation of the minimum model error criterion include:

[0102] Step S310, first-order Taylor expansion is performed on the system observation equation in the performance function;

[0103] Step S320, partial derivative of the system error is calculated through the performance function after the first-order Taylor expansion, and the system error compensation equation is determined based on the minimum value principle.

[0104] Specifically, it can be known from the system observation equation that the observation equation needs the system state at time k, and since the prediction filter based on the minimum model error criterion does not know the state at time k, first-order Taylor expansion is performed on the observation equation in formula (15) to obtain formula group (16):

[0105]

[0106] wherein, is the Jacobian matrix of the system observation equation; L G (h) and L f (h) are the first-order Lie derivatives of the system observation matrix equation h with respect to the matrix G and the system state equation f respectively, and dt is the sampling time.

[0107] Partial derivative of d(k) is calculated through the performance function J[d(k)], and according to the minimum value principle, the minimum value is obtained at , so the calculation process of the uncertain term d(k) existing in the system under the condition of performance function minimization is shown in formula group (17):

[0108]

[0109] Based on the above formula, the compensation value d(k) of the system error of the vehicle under all working conditions can be calculated.

[0110] In step S140, based on the combination of the cubature Kalman filter algorithm and the minimum model error criterion algorithm, the state value is updated according to the compensation value output by the system state equation, the system observation equation and the system error compensation equation, which specifically includes the following steps:

[0111] Step S410, time update is performed according to the compensation value output by the system state equation and the system error compensation equation to obtain the state prior estimate value and update the state estimation covariance matrix;

[0112] Step S420, measurement update is performed according to the updated state estimation covariance matrix, the system observation equation and the compensation value to obtain the measurement prediction value and calculate the measurement covariance matrix;

[0113] Step S430, innovation calculation and standardization are performed according to the measurement prediction value and the measurement covariance matrix to obtain the detection function;

[0114] Step S440, the detection function is segmented by using the multiple sub-polygons of chi-square distribution, and the corresponding output state value is determined according to different segmented results.

[0115] Specifically, first, the state value, the process noise matrix, the measurement noise matrix, the state estimation covariance matrix, the tire error vector and the weight matrix are initialized.

[0116] Secondly, time update is performed according to the system state equation and the compensation value, and the time update process is as follows:

[0117] The state estimation covariance matrix at k-1 moment is known:

[0118]

[0119] The volume point is calculated according to the state value at k-1 moment, as shown in formula (19):

[0120]

[0121] Wherein, ξ i is the propagated volume point set, and the volume point set ξ i is as shown in formula (20):

[0122]

[0123] The propagated volume point is as shown in formula (21):

[0124]

[0125] The predicted value of the state variable at k moment, i.e. the state priori estimation value, is calculated, as shown in formula (22):

[0126]

[0127] The state estimation covariance matrix of the priori estimation is updated, as shown in formula (23):

[0128]

[0129] Wherein, Q k-1 is the process noise matrix.

[0130] Then, measurement update is performed according to the updated state estimation covariance matrix, the system observation equation and the compensation value, and the measurement update process is as follows:

[0131] The updated state estimation covariance matrix P k|k-1 is decomposed, as shown in formula (24):

[0132]

[0133] The volume point Xi,k,k-1 As formula (25):

[0134]

[0135] The propagation volume point is compensated by k-1 moment, as formula (26):

[0136] z i,k|k-1 = h(X i,k,k-1 , d k-1|k-1 ); (26)

[0137] The prediction value of the update system state equation, i.e. the measurement prediction value, is updated, as formula (27):

[0138]

[0139] The measurement covariance matrix at k moment is estimated, as formula (28):

[0140]

[0141] Wherein, R k is the measurement noise matrix.

[0142] The cross-covariance matrix is further calculated, as formula (29):

[0143]

[0144] The volume Kalman filter gain matrix is calculated, as formula (30):

[0145] K k = P xy,k|k-1 (P yy,k|k-1 ) -1 ; (30)

[0146] Then, the detection function is calculated according to the measurement prediction value and the measurement covariance matrix, and the innovation is calculated and standardized, as follows:

[0147] The innovation calculation error e k is as formula (31):

[0148]

[0149] The detection function λ k of the innovation normal distribution after standardization is calculated, as formula (32):

[0150]

[0151] Wherein, λ k ~ χ 2 (n) for χ 2The problem of degrees of freedom n is determined according to the observation variable, when the heavy vehicle runs in all working conditions, χ 2 The degrees of freedom are 2.

[0152] Finally, the detection function is segmented by using the multiple quantile points of the chi-square distribution, and the corresponding output state value is determined according to different segmentation results. The segmentation of the detection function is as follows:

[0153] The weight distribution result is as shown in formula group (33):

[0154]

[0155] Wherein, M1 and M2 are quantile points of the chi-square distribution, η k is a weight function.

[0156] Combining Figure 3 , the corresponding output state value is determined according to different weights, that is, different segmentation results, that is:

[0157] For the detection function segment less than the first quantile point, the output state value is obtained by updating the state prior estimation value according to the volume Kalman filter gain matrix;

[0158] For the detection function segment between the first quantile point and the second quantile point, the measurement noise in the measurement update process is corrected to correct the measurement covariance matrix, and then the output state value is obtained by updating the state prior estimation value according to the volume Kalman filter gain matrix;

[0159] For the detection function segment greater than the second quantile point, no measurement update is performed, and the state prior estimation value is taken as the output state value.

[0160] Specifically, when λ k <M1, that is, the weight result η k =1, it indicates that the measurement data in the measurement update process is effective, and the normal measurement update is continued, that is, the predicted output state value at k time and the updated state estimation covariance matrix P k|k are calculated through formulas (34) and (35) respectively:

[0161]

[0162]

[0163] Wherein, K k is the volume Kalman filter gain matrix calculated by formula (30), represents the prediction error of the measurement update process, The state priori estimation value calculated by the above process is denoted as, and the updated state estimation covariance matrix is used for the next state update.

[0164] When M1≤λ k <M2, the weight result It is indicated that the measurement data in the measurement update process is polluted, and the state priori estimation value obtained by the above process needs to be modified after modifying the measurement noise. According to the relationship between the measurement noise and the measurement covariance matrix and the gain, the modification of the noise covariance matrix is used to modify the gain, so as to realize the robustness of the algorithm. The specific process is as follows:

[0165] The measurement noise covariance matrix is decomposed, as shown in formula (36):

[0166]

[0167] The measurement noise covariance matrix is modified by the weight η k , as shown in formula (37):

[0168]

[0169] The measurement covariance matrix is corrected, as shown in formula (38):

[0170]

[0171] After modifying the measurement covariance matrix to modify the gain matrix of the robust cubature Kalman filter, the predicted output state value at time k is calculated according to formula (34) and (35) and the updated state estimation covariance matrix P k|k .

[0172] When λ k ≥M2, the weight result η k =0, indicating that the measurement data is invalid data, and the measurement update is not performed, only the time update stage data is updated and the iterative calculation is performed as shown in formula (39) and formula (40):

[0173]

[0174] P k|k =P k|k-1 ; (40)

[0175] After the robust cubature Kalman filter is combined with the minimum model error criterion algorithm to perform time update and measurement update, the output state variable estimation value is obtained, and the state variables in the error compensation equation and the system state equation are updated to continue the update calculation.

[0176] In the embodiment of the present application, a longitudinal dynamics model of a heavy vehicle is established, since lateral tire force and lateral inertia force are inevitably generated in the heavy vehicle in all working conditions, and the conventional method for obtaining the tire force of the vehicle is through a magic tire model, which often needs a plurality of parameters as input, and it is more difficult to accurately obtain the tire force of the heavy vehicle after the mass center is offset, and based on this, the model compensation algorithm combining the cubature Kalman filter is used to compensate the model error in real time and effectively suppress the abnormal error of the sensor, realize the robust estimation of the road slope in all working conditions, and has the advantages of strong real-time, high precision and strong robustness.

[0177] The embodiment of the present application also provides a road slope estimation system in all working conditions, comprising:

[0178] The first module is used for obtaining a longitudinal dynamics model of a vehicle;

[0179] The second module is used for determining a system state equation and a system observation equation of the road slope estimation in all working conditions according to the longitudinal dynamics model of the vehicle, wherein the state variables of the system state equation include the longitudinal velocity and the road slope angle;

[0180] The third module is used for determining a system error compensation equation of the minimum model error criterion according to the system observation equation;

[0181] The fourth module is used for updating the state value based on the compensation value output by the system state equation, the system observation equation and the system error compensation equation according to the combination of the cubature Kalman filter algorithm and the minimum model error criterion algorithm.

[0182] It can be understood that the contents in the above-mentioned road slope estimation method embodiment in all working conditions are all applicable to the present system embodiment, the present system embodiment specifically realizes the same functions as the above-mentioned road slope estimation method embodiment in all working conditions, and achieves the same beneficial effects as the above-mentioned road slope estimation method embodiment in all working conditions.

[0183] Reference Figure 4 , Figure 4 is a schematic diagram of a road slope estimation device in all working conditions provided by an embodiment of the present application. The road slope estimation device in all working conditions of the embodiment of the present application comprises one or more control processors and memories, Figure 4 In the embodiment, one control processor and one memory are taken as an example.

[0184] The control processor and the memory can be connected through a bus or other means, Figure 4 In the embodiment, connection through a bus is taken as an example.

[0185] The memory, as a non-transitory computer readable storage medium, can be used to store non-transitory software programs and non-transitory computer executable programs. In addition, the memory can include high-speed random access memory and can also include non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state memory device. In some embodiments, the memory can optionally include memory that is remotely disposed relative to the control processor, and these remote memories can be connected to the all-terrain road slope estimation device through a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0186] Those skilled in the art can understand that, Figure 4 The device structure shown in the above embodiments does not constitute a limitation on the all-terrain road slope estimation device, and can include more or fewer components than those shown, or combine certain components, or different component arrangements.

[0187] The non-transitory software programs and instructions required to implement the all-terrain road slope estimation method applied to the all-terrain road slope estimation device in the above embodiments are stored in the memory, and when executed by the control processor, the all-terrain road slope estimation method applied to the all-terrain road slope estimation device in the above embodiments is executed.

[0188] In addition, one embodiment of the present application also provides a computer readable storage medium storing computer executable instructions, which, when executed by one or more control processors, can cause the above-mentioned one or more control processors to execute the all-terrain road slope estimation method in the above method embodiments.

[0189] As will be appreciated by one of ordinary skill in the art, all or some of the steps, systems, and techniques disclosed herein can be embodied in software, firmware, hardware, and / or suitable combination thereof. Some or all of the physical components can be implemented in software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or in hardware, or in an integrated circuit, such as an application- specific integrated circuit. Such software can be distributed on computer readable media, which can comprise computer storage media (or non-transitory media), and communication media (or transitory media). As will be appreciated by one of ordinary skill in the art, the term computer storage media includes all physical and tangible computer storage media, such as a volatile and non-volatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by a computer. Further, as will be appreciated by one skilled in the art, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media.

[0190] The embodiments of the present application disclosed above are only used to explain the principle of the present application, but the present application is not limited to the above embodiments. Various changes can be made by those skilled in the art without departing from the spirit of the present application.

Claims

1. A full operating condition road slope estimation method, characterized by, The method comprises the following steps: acquiring a vehicle longitudinal dynamics model; determining a system state equation and a system observation equation of road slope estimation in all working conditions according to the vehicle longitudinal dynamics model, wherein a state variable of the system state equation comprises a longitudinal velocity and a road slope angle; determining a system error compensation equation of a minimum model error criterion according to the system observation equation; updating a state value according to a compensation value output by the system state equation, the system observation equation and the system error compensation equation based on a combination of a cubature Kalman filtering algorithm and a minimum model error criterion algorithm; the compensation value output by the system error compensation equation in the minimum model error criterion is determined by the following steps: performing a first-order Taylor expansion on the system observation equation in the error compensation equation; deriving a system error by the error compensation equation after the first-order Taylor expansion, and determining a compensation value based on a minimum value principle; the updating of the state value according to the compensation value output by the system state equation, the system observation equation and the system error compensation equation based on the combination of the cubature Kalman filtering algorithm and the minimum model error criterion algorithm comprises the following steps: performing time updating on the compensation value output by the system state equation and the system error compensation equation to obtain a state prior estimation value and update a state estimation covariance matrix; performing measurement updating on the state prior estimation value to obtain a measurement prediction value and calculate a measurement covariance matrix according to the updated state estimation covariance matrix, the system observation equation and the compensation value; performing innovation calculation and standardization on the measurement prediction value and the measurement covariance matrix to obtain a detection function; segmenting the detection function by using multiple quantile points of a chi-square distribution, updating the state prior estimation value to obtain an output state value according to a cubature Kalman filtering gain matrix for a detection function segment less than a first quantile point, modifying a measurement noise in a measurement updating process to modify the measurement covariance matrix, and then updating the state prior estimation value to obtain the output state value according to the cubature Kalman filtering gain matrix for a detection function segment between the first quantile point and a second quantile point, and not performing the measurement updating and taking the state prior estimation value as the output state value for a detection function segment greater than the second quantile point.

2. The all-condition road grade estimation method according to claim 1, characterized by, The vehicle longitudinal dynamics model is expressed as: ； wherein denotes the longitudinal acceleration, denotes the total mass of the heavy-duty vehicle, denotes the driving torque, denotes the transmission gear ratio, denotes the final drive ratio, denotes the transmission efficiency, denotes the effective rolling radius of the wheel, denotes the air resistance coefficient, denotes the wind- facing area, denotes the air density, denotes the gravitational acceleration, denotes the rolling resistance coefficient denotes the road slope angle.

3. The all-condition road grade estimation method according to claim 1, characterized by, the determining of the system state equation and the system observation equation of road slope estimation in all working conditions according to the vehicle longitudinal dynamics model comprises the following steps: determining a time-varying parameter vehicle dynamics model according to the vehicle longitudinal dynamics model; determining a state variable equation according to the time-varying parameter vehicle dynamics model; determining a system state equation according to the state variable equation and a state error equation; determining an observation variable equation according to the longitudinal velocity and the longitudinal acceleration; determining a system observation equation according to the observation variable equation and an observation error equation.

4. The all-condition road grade estimation method according to claim 1, characterized by, the updating of the state value according to the compensation value output by the system state equation, the system observation equation and the system error compensation equation based on the combination of the cubature Kalman filtering algorithm and the minimum model error criterion algorithm comprises the following steps: multiplying the cubature Kalman filtering gain matrix by a prediction error in a measurement updating process to obtain a state correction value; The state prior estimate value is added to the state correction value to obtain an output state value.

5. An all-condition road slope estimation system characterized by comprising: The method comprises the following steps: A first module is configured to acquire a vehicle longitudinal dynamics model; A second module is configured to determine a system state equation and a system observation equation of road slope estimation in all working conditions according to the vehicle longitudinal dynamics model, wherein a state variable of the system state equation comprises a longitudinal velocity and a road slope angle; A third module is configured to determine a system error compensation equation of a minimum model error criterion according to the system observation equation; A fourth module is configured to update a state value based on a compensation value output by the system state equation, the system observation equation and the system error compensation equation according to a cubature Kalman filter algorithm and a minimum model error criterion algorithm; The compensation value output by the system error compensation equation in the minimum model error criterion of the fourth module is determined by the following steps: First-order Taylor expansion is performed on the system observation equation in the error compensation equation; Partial derivatives of the system error are calculated based on the error compensation equation after the first-order Taylor expansion, and a compensation value is determined based on a minimum value principle; The fourth module is specifically configured to perform the following steps: Time updating is performed on the compensation value output by the system state equation and the system error compensation equation to obtain a state prior estimate value and update a state estimation covariance matrix; Measurement updating is performed on the state estimation covariance matrix after updating, the system observation equation and the compensation value to obtain a measurement prediction value and calculate a measurement covariance matrix; Novelty calculation and standardization are performed on the measurement prediction value and the measurement covariance matrix to obtain a detection function; The detection function is segmented by using multiple quantile points of a chi-square distribution, for a detection function segment less than a first quantile point, the state prior estimate value is updated based on a cubature Kalman filter gain matrix to obtain an output state value; for a detection function segment between the first quantile point and a second quantile point, measurement noise in the measurement updating process is modified to modify the measurement covariance matrix, and then the state prior estimate value is updated based on the cubature Kalman filter gain matrix to obtain an output state value; for a detection function segment greater than the second quantile point, no measurement updating is performed, and the state prior estimate value is taken as the output state value.

6. A full-condition road slope estimation device characterized by comprising: The method comprises the following steps: At least one processor; At least one memory is configured to store at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the all-working-condition road slope estimation method according to any one of claims 1 to 4.

7. A computer readable storage medium having stored therein a program which is executable by a processor, characterized in that, The program executable by the processor is used to implement the all-working-condition road slope estimation method according to any one of claims 1 to 4 when executed by the processor.

Images