A parameter and tire force real-time estimation method and system based on multi-model fusion

Abstract

The application discloses a parameter and tire force real-time estimation method and system based on multi-model fusion, relates to the vehicle parameter identification technical field, and comprises the following steps: establishing a three-state monorail vehicle dynamics model and a tire model according to vehicle parameters and dynamics parameters; performing formula transformation on the three-state monorail vehicle dynamics model to obtain formulas of longitudinal, lateral and yaw angle acceleration respectively; dividing parameters in the formulas into low-cost sensor collectable parameters and to-be-estimated parameters, and updating the formulas of longitudinal, lateral and yaw angle acceleration respectively; constructing a nonlinear optimization problem based on the updated formulas; performing parameter estimation based on the nonlinear optimization problem, vehicle parameters and dynamics parameters in a historical driving process; and performing real-time state estimation on tire force based on the parameter estimation result. The application can estimate complete dynamics parameters in real time through a low-cost sensor, and estimate the real-time state of the tire based on the complete dynamics parameters.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle parameter identification, in particular to a parameter and tire force real-time estimation method and system based on multi-model fusion. BACKGROUND

[0002] In the field of autonomous driving, obtaining accurate dynamic parameters is crucial for the design of vehicle motion controllers. However, the acquisition of dynamic parameters faces multiple challenges. First, many controller designs assume that dynamic parameters are known, but in actual scenarios, they cannot be directly measured by low-cost sensors and can only be estimated from sensor observations. Measuring dynamic parameters through specialized equipment usually requires sending the vehicle to a specific test site, which is complex, time-consuming and costly. In addition, many existing parameter estimation methods can only estimate partial parameters, while other parameters are assumed to be known, which limits the effective application of these methods. Furthermore, most studies ignore the time-varying characteristics of certain parameters in the dynamic parameters, which cannot estimate time-varying parameters in real time, affecting the accuracy of other parameter estimation and the consistency of motion control. SUMMARY

[0003] The purpose of the present application is to provide a parameter and tire force real-time estimation method and system based on multi-model fusion, which can realize real-time estimation of complete dynamic parameters through low-cost sensors and real-time state estimation of tire lateral, longitudinal and vertical forces based on dynamic parameter estimation.

[0004] To achieve the above purpose, the present application provides the following solutions:

[0005] In a first aspect, the present application provides a parameter and tire force real-time estimation method based on multi-model fusion, comprising:

[0006] Obtaining vehicle parameters and dynamic parameters during driving of a target vehicle; the vehicle parameters include body parameters and wheel parameters.

[0007] Based on the Newton-Euler law, a three-state single-track vehicle dynamics model fused with a wheel model and a simplified magic formula tire model considering load transfer effects are established respectively according to the vehicle parameters and dynamic parameters.

[0008] The three-state single-track vehicle dynamics model fused with the wheel model is transformed into formula expressions for longitudinal acceleration, lateral acceleration and yaw angle acceleration.

[0009] The parameters in the formula expressions of the longitudinal acceleration, the lateral acceleration and the yaw rate acceleration are divided into parameters that can be collected by low-cost sensors and parameters to be estimated, and the formula expressions of the longitudinal acceleration, the lateral acceleration and the yaw rate acceleration are updated respectively; the parameters to be estimated include a vehicle mass center longitudinal position, a vehicle mass, a vehicle inertia around a yaw axis and a wheel inertia.

[0010] Based on the updated formula expressions of the longitudinal acceleration, the lateral acceleration and the yaw rate acceleration, a nonlinear optimization problem based on the longitudinal acceleration, the lateral acceleration and the yaw rate acceleration is constructed.

[0011] According to the nonlinear optimization problem, the parameters to be estimated are estimated based on vehicle parameters of the target vehicle and dynamic parameters in a historical driving process, to obtain a parameter estimation result.

[0012] Based on the parameter estimation result, real-time state estimation is performed on tire forces, including tire longitudinal forces, tire vertical forces and tire lateral forces; the tire vertical forces are calculated based on a simplified magic formula tire model considering load transfer effects.

[0013] Optionally, a formula expression of a three-state single-track vehicle dynamics model fusing a wheel model is as follows:

[0014] .

[0015] wherein, is a distance between a vehicle mass center and a front axle, is a distance between the vehicle mass center and a rear axle, is a vehicle mass, is an inertia around a yaw axis, is a rear wheel lateral force, is a front wheel lateral force, is a yaw rate, is a yaw acceleration, is a mass center longitudinal velocity, is a mass center lateral velocity, is a front wheel rotation angle, is a rear wheel driving torque, is an inertia of the front wheel and the rear wheel, is a rear wheel rotation angle acceleration, is a radius of the front wheel and the rear wheel, is a mass center longitudinal acceleration, is a mass center lateral acceleration.

[0016] Optionally, a formula expression of the simplified magic formula tire model considering the load transfer effects is as follows:

[0017] .

[0018] .

[0019] wherein, is the vehicle center of mass height, is the gravitational acceleration, , , , , is a tire parameter, l is a distance.

[0020] Optionally, the three-state monorail vehicle dynamics model incorporating the wheel model is formula transformed to obtain formula expressions of longitudinal acceleration, lateral acceleration and yaw angle acceleration, specifically including:

[0021] The three equations in the three-state monorail vehicle dynamics model incorporating the wheel model are combined in pairs to form three groups of equation sets.

[0022] The three groups of equation sets are matrix transformed to obtain matrix forms of the three groups of equation sets.

[0023] The matrix forms of the three groups of equation sets are substituted into the three-state monorail vehicle dynamics model incorporating the wheel model to obtain formula expressions of longitudinal acceleration, lateral acceleration and yaw angle acceleration.

[0024] Optionally, the three groups of equation sets are specifically:

[0025] .

[0026] .

[0027] .

[0028] Optionally, the matrix forms of the three groups of equation sets are specifically:

[0029] .

[0030] .

[0031] .

[0032] Optionally, the formula expressions of longitudinal acceleration, lateral acceleration and yaw angle acceleration are specifically:

[0033] ; .

[0034] .

[0035] Optionally, the formula expression of the nonlinear optimization problem is:

[0036] .

[0037] wherein, is a predicted lateral acceleration at the time, is a predicted longitudinal acceleration at the time, is a predicted yaw angle acceleration at the time, is a lateral acceleration at the time, is a longitudinal acceleration at the time, is a yaw angle acceleration at the time; is a weight coefficient of the lateral acceleration cost term, is a weight coefficient of the longitudinal acceleration cost term, is a weight coefficient of the yaw angle acceleration cost term, is a length of a sliding window, k is a sampling point index corresponding to a time after a right boundary of the sliding window.

[0038] Optionally, based on the parameter estimation result, real-time state estimation of the tire force is performed, specifically including:

[0039] According to the formula , the tire longitudinal force is estimated; in the formula, is a rear wheel longitudinal force.

[0040] According to the formula , the vertical force is estimated; in the formula, is a front wheel vertical force, is a rear wheel vertical force.

[0041] According to the formula expression of the simplified magic formula tire model considering the load transfer effect, the front wheel lateral force and the rear wheel lateral force are estimated.

[0042] In a second aspect, the application provides a parameter and tire force real-time estimation method based on multi-model fusion, including:

[0043] a parameter acquisition module, configured to acquire vehicle parameters and dynamic parameters in a driving process of a target vehicle; the vehicle parameters include body parameters and wheel parameters.

[0044] A model establishing module is configured to establish a three-state monorail vehicle dynamics model fused with a wheel model and a simplified magic formula tire model considering load transfer effect based on Newton-Euler law according to the vehicle parameters and the dynamics parameters.

[0045] A formula transforming module is configured to transform the three-state monorail vehicle dynamics model fused with the wheel model to obtain formula expressions of longitudinal acceleration, lateral acceleration and yaw angle acceleration.

[0046] A formula updating module is configured to divide parameters in the formula expressions of the longitudinal acceleration, the lateral acceleration and the yaw angle acceleration into parameters collectable by low-cost sensors and to-be-estimated parameters, and to update the formula expressions of the longitudinal acceleration, the lateral acceleration and the yaw angle acceleration, respectively; the to-be-estimated parameters include a vehicle mass center longitudinal position, a vehicle mass, a vehicle inertia around a yaw axis and a wheel inertia.

[0047] An optimization problem constructing module is configured to construct a nonlinear optimization problem based on longitudinal acceleration, lateral acceleration and yaw angle acceleration based on the updated formula expressions of the longitudinal acceleration, the lateral acceleration and the yaw angle acceleration.

[0048] A parameter estimating module is configured to perform parameter estimation on the to-be-estimated parameters based on vehicle parameters of a target vehicle and dynamics parameters in a historical driving process according to the nonlinear optimization problem, to obtain a parameter estimation result.

[0049] A tire force estimating module is configured to perform real-time state estimation on tire forces based on the parameter estimation result; the tire forces include tire longitudinal forces, tire vertical forces and tire lateral forces; the tire vertical forces are calculated based on the simplified magic formula tire model considering load transfer effect.

[0050] According to the specific embodiments provided in the present application, the following technical effects are disclosed:

[0051] The application provides a parameter and tire force real-time estimation method and system based on multi-model fusion. BRIEF DESCRIPTION OF DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0053] Figure 1 A flowchart of a parameter and tire force real-time estimation method based on multi-model fusion provided by an embodiment of the present application.

[0054] Figure 2 A three-state monorail vehicle dynamics model provided by an embodiment of the present application

[0055] Figure 3 A wheel model provided by an embodiment of the present application.

[0056] Figure 4A vehicle longitudinal dynamics model schematic diagram provided by an embodiment of the present application.

[0057] Figure 5 A functional module schematic diagram of a parameter and tire force real-time system based on multi-model fusion provided by an embodiment of the present application. DETAILED DESCRIPTION

[0058] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0059] In order to make the above objectives, characteristics and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0060] Embodiment One

[0061] As shown in the figure, the present embodiment provides a parameter and tire force real-time estimation method based on multi-model fusion, comprising: Figure 1

[0062] Step 101: obtaining vehicle parameters and dynamics parameters in a driving process of a target vehicle; the vehicle parameters include body parameters and wheel parameters.

[0063] Step 102: based on Newton-Euler law, establishing a three-state single-track vehicle dynamics model fused with a wheel model and a simplified magic formula tire model considering load transfer effect, respectively, according to the vehicle parameters and the dynamics parameters.

[0064] Step 103: performing formula transformation on the three-state single-track vehicle dynamics model fused with the wheel model, to obtain formula expressions of longitudinal acceleration, lateral acceleration and yaw angle acceleration, respectively.

[0065] Step 104: dividing parameters in the formula expressions of the longitudinal acceleration, the lateral acceleration and the yaw angle acceleration into low-cost sensor collectable parameters and to-be-estimated parameters, and updating the formula expressions of the longitudinal acceleration, the lateral acceleration and the yaw angle acceleration, respectively; the to-be-estimated parameters include vehicle mass center longitudinal position, vehicle mass, vehicle inertia around yaw axis and wheel inertia.

[0066] Step 105: based on the updated formula expressions of the longitudinal acceleration, the lateral acceleration and the yaw angle acceleration, constructing a nonlinear optimization problem based on the longitudinal acceleration, the lateral acceleration and the yaw angle acceleration.

[0067] ​Step 106: according to the nonlinear optimization problem, based on the vehicle parameters of the target vehicle and the dynamic parameters in the historical driving process, parameter estimation is performed on the to-be-estimated parameters to obtain a parameter estimation result.

[0068] Step 107: based on the parameter estimation result, real-time state estimation is performed on the tire force; the tire force includes a tire longitudinal force, a tire vertical force and a tire lateral force; the tire vertical force is calculated based on a simplified magic formula tire model considering load transfer effect.

[0069] In some embodiments, when step 101 is performed, the following can be specifically performed:

[0070] Among the obtained parameters, the longitudinal position of the vehicle mass center, the vehicle mass, the moment of inertia of the vehicle around the yaw axis, the moment of inertia of the wheel, the tire parameters and the height of the vehicle mass center are unknown parameters, and the effective radii r of the front wheel and the rear wheel are known parameters, which are measured by a length measuring tool such as a tape measure.

[0071] Among the obtained parameters, the longitudinal position of the vehicle mass center, the vehicle mass, the moment of inertia of the vehicle around the yaw axis, the moment of inertia of the wheel, the tire parameters and the height of the vehicle mass center are unknown parameters, and the effective radii r of the front wheel and the rear wheel are known parameters, which are measured by a length measuring tool such as a tape measure.

[0072] In some embodiments, when step 102 is performed, the following can be specifically performed:

[0073] First, a three-state single-track vehicle dynamics model is established, as shown in Figure 2 In the modeling process, it is assumed that the rear wheel is driven and the front wheel is steered, so the front wheel longitudinal force =0. The force balance equations along the axis, axis and the moment balance equation around the yaw axis can be obtained:

[0074] (1).

[0075] (2).

[0076] (3).

[0077] Among them, is the distance between the vehicle mass center and the front axle, is the distance between the vehicle mass center and the rear axle, is the vehicle mass, is the moment of inertia around the yaw axis, is the rear wheel lateral force, is the front wheel lateral force, is the yaw angular velocity, is the yaw angular acceleration, is the mass center longitudinal velocity, is the mass center lateral velocity, is the front wheel rotation angle, is the longitudinal acceleration of the mass center, is the lateral acceleration of the mass center.

[0078] Then, the wheel model of the rear-wheel drive vehicle is established as shown in Figure 3

[0079] (4).

[0080] where, is the moment of inertia of the front and rear wheels, is the rear wheel rotation angle acceleration, is the effective radius of the front and rear wheels, is the rear wheel driving torque.

[0081] Substituting equation (4) into equation (2) gives:

[0082] (5).

[0083] By combining equations (1), (3) and (5), the three-state single-track vehicle dynamics model incorporating the wheel model is obtained as follows:

[0084] (6).

[0085] where, for the tire lateral force and , a simplified magic formula tire model considering the nonlinear characteristics of the tire is established as follows:

[0086] (7).

[0087] (8).

[0088] where, , , is the front wheel magic formula coefficient, , , is the rear wheel magic formula coefficient. is the front wheel side slip angle, is the rear wheel side slip angle. The magic formula coefficients and side slip angles can be further expressed as:

[0089] (9).

[0090] (10).

[0091] (11).

[0092] ​wherein, is the front wheel vertical force, is the rear wheel vertical force. , , , , is a tire parameter. To calculate the load transfer effect-considered and , a vehicle longitudinal dynamics model is established as shown in Figure 4 , and the moment balance around the axis at the contact points of the front and rear wheels with the ground can be obtained as follows:

[0093] (12).

[0094] wherein, is the vehicle center of mass height, is the gravitational acceleration.

[0095] Substituting equations (9)-(12) into equations (7)-(8) can obtain a simplified magic formula tire model considering the load transfer effect:

[0096] (13).

[0097] (14).

[0098] wherein, in some embodiments, when step 103 is performed, it can be specifically as follows:

[0099] The three-state single-track vehicle dynamics model incorporating the wheel model is formula transformed to obtain formula expressions of longitudinal acceleration, lateral acceleration, and yaw angle acceleration, specifically including:

[0100] The three equations in the three-state single-track vehicle dynamics model incorporating the wheel model are combined two by two to form three groups of equation sets.

[0101] The three groups of equation sets are matrix transformed to obtain matrix forms of the three groups of equation sets.

[0102] The matrix forms of the three groups of equation sets are substituted into the three-state single-track vehicle dynamics model incorporating the wheel model to obtain formula expressions of longitudinal acceleration, lateral acceleration, and yaw angle acceleration.

[0103] Specifically, based on the three-state single-track vehicle dynamics model incorporating the wheel model established in step 102, the three equations in equation (6) are combined two by two to form the following three groups of equation sets:

[0104] (15).

[0105] (16).

[0106] (17).

[0107] The formulae (15)-(17) are written in matrix form and arranged to obtain:

[0108] (18).

[0109] (19).

[0110] (20).

[0111] The formulae (1), (3) and (5) are written in matrix form and arranged to obtain:

[0112] (21).

[0113] (22).

[0114] (23).

[0115] The formula (18) is substituted into the formula (23), the formula (19) is substituted into the formula (22), and the formula (20) is substituted into the formula (21) to obtain the expressions of the longitudinal acceleration, the lateral acceleration and the yaw angle acceleration as follows:

[0116] (24).

[0117] (25).

[0118] (26).

[0119] In some embodiments, when the step 104 is performed, the following can be specifically performed:

[0120] The historical driving data that can be collected by the low-cost sensor is , , , , , , , , , and the state vector is defined as . The known parameter vector is defined as , and the parameter vector to be estimated is defined as .

[0121] Substituting the defined parameters to be estimated into equations (24) to (26) and rearranging, we get:

[0122] (27).

[0123] (28).

[0124] (29).

[0125] in, The lateral acceleration is predicted based on historical driving data. The longitudinal acceleration is predicted based on historical driving data. This is the yaw angle acceleration predicted based on historical driving data.

[0126] In some embodiments, when performing step 105, the specific steps may be as follows:

[0127] From equations (27) to (29), a nonlinear optimization problem based on the historical driving data of the current sliding window can be constructed:

[0128] (30).

[0129] in, For the first Predicted lateral acceleration at time [time] For the first Predicted longitudinal acceleration at time [time] For the first Predicted yaw angle acceleration at time [time] For the first Lateral acceleration at time t, For the first longitudinal acceleration at time t, For the first Yaw angle acceleration at time; It is the weighting coefficient of the lateral acceleration cost term. It is the weighting coefficient of the longitudinal acceleration cost term. It is the weighting coefficient of the yaw angle acceleration cost term. The length of the sliding window. k This is the index of the sampling point corresponding to the next moment after the right boundary of the sliding window. The time step interval covered by the sliding window itself is... There are a total of n sampling points.

[0130] In the nonlinear optimization problem constructed by equation (30), by selecting appropriate parameters to be estimated... To minimize the cost function based on the current sliding window This enables real-time estimation of the vehicle's longitudinal position of center of gravity, vehicle mass, vehicle moment of inertia about the yaw axis, and wheel moment of inertia.

[0131] In some embodiments, when performing step 106, the specific steps may be as follows:

[0132] We construct an optimization problem based on three equivalent cost terms of tire lateral force and solve it using historical driving data in the current sliding window to achieve real-time estimation of tire parameters and vehicle center of gravity height.

[0133] Based on the simplified magic formula tire model that takes into account the load transfer effect and the estimation results of the vehicle's longitudinal position of center of gravity, vehicle mass, vehicle rotational inertia about the yaw axis and wheel rotational inertia, the tire parameters and vehicle center of gravity height can be further estimated in real time.

[0134] From equations (18) to (20), three equivalent calculation formulas for the lateral force of the tire can be obtained:

[0135] (31).

[0136] (32).

[0137] (33).

[0138] in, and ( ) indicates the first The tire lateral force is calculated using these three methods. The state variables on the right-hand side of equations (31) to (33) can all be acquired using low-cost sensors, and the unknown parameters have already been estimated in step 104. Therefore, all three calculation methods can achieve a true estimate of the tire lateral force. From equations (13) to (14), the predicted value of the tire lateral force is:

[0139] (34).

[0140] (35).

[0141] in, The lateral force of the front wheels is predicted based on historical driving data. This represents the lateral force of the rear wheels predicted based on historical driving data. Furthermore, equations (34) to (35) can also be expressed in the following mathematical form:

[0142] (36).

[0143] (37).

[0144] wherein, is the parameter vector to be estimated in step 104. According to equations (31)-(37) and using the historical driving data in the current sliding window, three equivalent cost item optimization problems based on tire lateral force can be constructed as follows:

[0145] (38).

[0146] wherein, is the predicted front wheel lateral force at the time instant, is the predicted rear wheel lateral force at the time instant. is the front wheel lateral force calculated at the time instant by the th way, is the rear wheel lateral force calculated at the time instant by the th way. is the front wheel lateral force cost item in the th way, is the rear wheel lateral force cost item in the th way.

[0147] In the three equivalent cost item optimization problems based on tire lateral force constructed by equation (38), by selecting the appropriate to-be-estimated parameters to minimize the cost function in the current sliding window, real-time estimation of tire parameters and vehicle center of mass height is achieved.

[0148] wherein, in some embodiments, when step 107 is performed, the following can be specifically implemented:

[0149] Based on the established wheel model, vehicle longitudinal dynamics model and simplified magic formula tire model considering load transfer effect, and based on the real-time estimation results of parameters, real-time state estimation of tire longitudinal force, vertical force and lateral force can be performed.

[0150] The tire longitudinal force expression can be obtained from equation (4) as:

[0151] (39).

[0152] wherein, and can be directly measured by low-cost sensors, is a known parameter, is obtained by estimation in step 104. Therefore, real-time estimation of tire longitudinal force is achieved.

[0153] Formula (12) is a tire vertical force expression. Wherein, can be obtained by direct measurement of low-cost sensors, is a known parameter, , and are obtained by estimation in step 104. are obtained by estimation in step 104. Thus, the tire vertical force and are achieved in real time.

[0154] Formula (13)~Formula (14) are tire lateral force expressions. Wherein, , , , , can be obtained by direct measurement of low-cost sensors, is a known parameter, , and are obtained by estimation in step 104. , , , , , are obtained by estimation in step 104. Thus, the tire lateral force and are achieved in real time.

[0155] Embodiment two

[0156] As shown in Figure 5 , the embodiment provides a multi-model fusion-based parameter and tire force real-time estimation system, comprising:

[0157] A parameter acquisition module 501 is configured to acquire vehicle parameters of a target vehicle and dynamic parameters in a driving process; the vehicle parameters include body parameters and wheel parameters.

[0158] A model establishment module 502 is configured to establish a three-state single-track vehicle dynamics model of a fusion wheel model and a simplified magic formula tire model considering load transfer effect based on Newton-Euler law according to the vehicle parameters and the dynamic parameters.

[0159] A formula transformation module 503 is configured to perform formula transformation on the three-state single-track vehicle dynamics model of the fusion wheel model to obtain formula expressions of longitudinal acceleration, lateral acceleration and yaw angle acceleration, respectively.

[0160] The formula updating module 504 is configured to divide parameters in formulaic expressions of the longitudinal acceleration, the lateral acceleration and the yaw rate acceleration into parameters that can be collected by the low-cost sensor and parameters to be estimated, and update the formulaic expressions of the longitudinal acceleration, the lateral acceleration and the yaw rate acceleration respectively; the parameters to be estimated include the vehicle mass center longitudinal position, the vehicle mass, the vehicle inertia around the yaw axis and the wheel inertia.

[0161] The optimization problem constructing module 505 is configured to construct a nonlinear optimization problem based on the longitudinal acceleration, the lateral acceleration and the yaw rate acceleration based on the updated formulaic expressions of the longitudinal acceleration, the lateral acceleration and the yaw rate acceleration.

[0162] The parameter estimation module 506 is configured to perform parameter estimation on the parameters to be estimated based on the vehicle parameters of the target vehicle and the dynamic parameters in the historical driving process according to the nonlinear optimization problem, to obtain a parameter estimation result.

[0163] The tire force estimation module 507 is configured to perform real-time state estimation on the tire force based on the parameter estimation result; the tire force includes the tire longitudinal force, the tire vertical force and the tire lateral force; the tire vertical force is calculated based on a simplified magic formula tire model considering load transfer effect.

[0164] In summary, the present application has the following technical effects:

[0165] The present application provides a real-time estimation method for complete dynamic parameters and tire longitudinal, lateral and vertical forces based on multi-model fusion. Firstly, a three-state single-track vehicle dynamics model fused with a wheel model and a simplified magic formula tire model considering load transfer effect are established. Secondly, a nonlinear optimization problem based on the longitudinal acceleration, the lateral acceleration and the yaw rate acceleration is constructed, and is solved through historical driving data in a current sliding window, to realize real-time estimation on the vehicle mass center longitudinal position, the vehicle mass, the vehicle inertia around the yaw axis and the wheel inertia. Then, three equivalent cost item optimization problems based on the tire lateral force are constructed, and are solved through historical driving data in the current sliding window, to realize real-time estimation on the tire parameters and the vehicle mass center height. Finally, real-time state estimation is performed on the tire longitudinal force, the vertical force and the lateral force based on the parameter estimation result. This method can realize real-time estimation on complete dynamic parameters through only low-cost sensors, so as to ensure simplicity and efficiency of parameter acquisition, practical applicability, real-time performance, model precision and control robustness. Meanwhile, this method can perform real-time state estimation on the tire longitudinal force, the vertical force and the lateral force, to provide key state information for vehicle motion control, so as to be beneficial to accurate formulation of control strategies

[0166] Any technical features in the above embodiments can be combined, and for the sake of brevity, not all possible combinations are described above, however, it should be understood that the application encompasses all possible combinations of the technical features unless such a combination is not technically possible.

[0167] The principles and implementation manners of the present application are described herein by using specific examples, and the above embodiments are only used to help understand the method of the present application and its core idea; meanwhile, for those skilled in the art, the specific implementation manners and application range can be changed according to the idea of the present application. In conclusion, the content of the present specification should not be understood as a limitation of the present application.

Claims

1. A real-time parameter and tire force estimation method based on multi-model fusion, characterized in that, include: Obtain the vehicle parameters and dynamic parameters of the target vehicle during its driving process; The vehicle parameters include body parameters and wheel parameters; Based on the vehicle parameters and dynamic parameters, and using Newton-Euler's law, a three-state monorail vehicle dynamic model with integrated wheel model and a simplified magic formula tire model considering load transfer effect are established respectively. The three-state monorail vehicle dynamics model with integrated wheel model is transformed to obtain the formula expressions for longitudinal acceleration, lateral acceleration and yaw angle acceleration respectively. The parameters in the formulas for longitudinal acceleration, lateral acceleration, and yaw acceleration are divided into parameters that can be acquired by low-cost sensors and parameters to be estimated, and the formulas for longitudinal acceleration, lateral acceleration, and yaw acceleration are updated respectively. The parameters to be estimated include the longitudinal position of the vehicle's center of gravity, the vehicle's mass, the vehicle's moment of inertia about the yaw axis, and the wheel's moment of inertia. Based on the updated formulas for longitudinal acceleration, lateral acceleration, and yaw acceleration, a nonlinear optimization problem based on longitudinal acceleration, lateral acceleration, and yaw acceleration is constructed. Based on the nonlinear optimization problem, the parameters to be estimated are estimated based on the vehicle parameters of the target vehicle and the dynamic parameters during the historical driving process, and the parameter estimation results are obtained. Based on the parameter estimation results, the tire force is estimated in real time; the tire force includes the tire longitudinal force, the tire vertical force, and the tire lateral force. The tire vertical force is calculated based on a simplified magic formula tire model that takes into account load transfer effects; The formula for the three-state monorail vehicle dynamics model integrating the wheel model is as follows: ； in, This is the distance between the vehicle's center of gravity and the front axle. The distance between the vehicle's center of gravity and the rear axle. For vehicle quality, Let be the moment of inertia about the yaw axis. For the lateral force of the rear wheel, For the lateral force of the front wheel, Yaw angular velocity, Yaw acceleration, Let the longitudinal velocity of the center of mass be... Let the transverse velocity of the center of mass be... For the front wheel steering angle, For rear-wheel drive torque, Let be the moments of inertia of the front and rear wheels. The angular acceleration of the rear wheel. Let be the radii of the front and rear wheels. For the longitudinal acceleration of the center of mass, This is the lateral acceleration of the center of mass; The simplified magic formula for the tire model, considering load transfer effects, is expressed as follows: ； ； in, For the height of the vehicle's center of gravity, It is the acceleration due to gravity. , , , , For tire parameters, l For distance.

2. The real-time parameter and tire force estimation method based on multi-model fusion according to claim 1, characterized in that, The three-state monorail vehicle dynamics model incorporating the wheel model is transformed to obtain formula expressions for longitudinal acceleration, lateral acceleration, and yaw acceleration, specifically including: The three equations in the three-state monorail vehicle dynamics model that integrates the wheel model are combined in pairs to form three sets of equations. Perform matrix transformations on the three sets of equations to obtain the matrix form of the three sets of equations. Substituting the matrix form of the three sets of equations into the three-state monorail vehicle dynamics model with the fused wheel model, we obtain the formulas for longitudinal acceleration, lateral acceleration, and yaw acceleration.

3. The real-time parameter and tire force estimation method based on multi-model fusion according to claim 2, characterized in that, The three sets of equations constituted are as follows: ； ； 。 4. The real-time parameter and tire force estimation method based on multi-model fusion according to claim 3, characterized in that, The matrix form of the three sets of equations is as follows: ； ； 。 5. The real-time parameter and tire force estimation method based on multi-model fusion according to claim 4, characterized in that, The formulas for the longitudinal acceleration, lateral acceleration, and yaw acceleration are as follows: ； ； 。 6. The real-time parameter and tire force estimation method based on multi-model fusion according to claim 5, characterized in that, The formula for the nonlinear optimization problem is: ； in, For the first Predicted lateral acceleration at time [time] For the first Predicted longitudinal acceleration at time [time] For the first Predicted yaw angle acceleration at time [time] For the first Lateral acceleration at time t, For the first longitudinal acceleration at time t, For the first Yaw angle acceleration at time; It is the weighting coefficient of the lateral acceleration cost term. It is the weighting coefficient of the longitudinal acceleration cost term. It is the weighting coefficient of the yaw angle acceleration cost term. The length of the sliding window. k This is the index of the sampling point corresponding to the moment after the right boundary of the sliding window.

7. The real-time parameter and tire force estimation method based on multi-model fusion according to claim 6, characterized in that, Based on the parameter estimation results, real-time state estimation of tire forces is performed, specifically including: According to the formula Estimate the longitudinal force of the tire; where, This is the longitudinal force of the rear wheel; According to the formula Estimate the vertical force; where, It is the vertical force of the front wheel. It is the vertical force of the rear wheel; Based on the formula expression of the simplified magic formula tire model that considers load transfer effects, the lateral forces of the front and rear wheels are estimated.

8. A real-time parameter and tire force estimation system based on multi-model fusion, used to implement the real-time parameter and tire force estimation method based on multi-model fusion as described in any one of claims 1-7, characterized in that, include: The parameter acquisition module is used to acquire the vehicle parameters of the target vehicle and the dynamic parameters during the driving process. The vehicle parameters include body parameters and wheel parameters; The model building module is used to build a three-state monorail vehicle dynamic model that integrates the wheel model and a simplified magic formula tire model that considers the load transfer effect, based on the vehicle parameters and dynamic parameters and the Newton-Euler law. The formula transformation module is used to transform the three-state monorail vehicle dynamics model with integrated wheel model to obtain the formula expressions for longitudinal acceleration, lateral acceleration and yaw angle acceleration respectively. The formula update module is used to divide the parameters in the formula expressions of longitudinal acceleration, lateral acceleration, and yaw acceleration into parameters that can be acquired by low-cost sensors and parameters to be estimated, and update the formula expressions of longitudinal acceleration, lateral acceleration, and yaw acceleration respectively; the parameters to be estimated include the longitudinal position of the vehicle's center of gravity, the vehicle's mass, the vehicle's moment of inertia about the yaw axis, and the wheel's moment of inertia; The optimization problem construction module is used to construct nonlinear optimization problems based on the updated formulas for longitudinal acceleration, lateral acceleration, and yaw angle acceleration. The parameter estimation module is used to estimate the parameters to be estimated based on the vehicle parameters of the target vehicle and the dynamic parameters during the historical driving process, according to the nonlinear optimization problem, and to obtain the parameter estimation results. The tire force estimation module is used to perform real-time state estimation of tire force based on the parameter estimation results; the tire force includes tire longitudinal force, tire vertical force and tire lateral force; The tire vertical force is calculated based on a simplified magic formula tire model that takes into account load transfer effects.

Images