A dynamic event-triggered based vehicle state adaptive estimation method

Abstract

This invention discloses a vehicle state estimation method based on dynamic event triggering. The method first designs an output estimator based on dynamic event triggering to reduce sampling of the actual output and obtain a continuous output estimation signal. Then, it designs an adaptive observer based on the triggered sampling signal to estimate the vehicle state and avoid Zeno behavior. When applied to road vehicles, this method effectively estimates the vehicle state and avoids Zeno behavior.

Description

Technical Field

[0001] This invention relates to an adaptive method, specifically to a vehicle state adaptive estimation method based on dynamic event triggering. Background Technology

[0002] In recent years, with the rapid development of the automotive industry and the continuous advancement of automation technology, vehicle systems have been increasingly widely used in transportation, logistics, and intelligent driving. Especially against the backdrop of the rapid development of new energy vehicles and intelligent connected vehicles, the structure and control strategies of vehicle systems are becoming increasingly complex, and the requirements for vehicle operational safety, stability, and reliability are constantly increasing. In actual operation, vehicle systems are often affected by various factors such as changes in road conditions, external disturbances, and model uncertainties. These factors may cause changes in vehicle dynamic characteristics, thereby affecting the system's control performance. Therefore, accurate and real-time state monitoring and parameter identification of vehicle systems are of great significance for ensuring safe vehicle operation.

[0003] Furthermore, with the rapid development of communication technology and its deep integration with industrial systems, networked control has gradually become an important means of promoting industrial system development. In a networked control framework, data exchange is typically conducted through network channels based on periodic time triggering. However, with the expansion of system scale and the frequent transmission of redundant data packets, network congestion problems are becoming increasingly prominent, severely impacting system performance. Therefore, it is urgent to introduce efficient communication strategies into networked control systems to reduce data transmission frequency and conserve limited network resources. Existing research largely integrates static event-triggered strategies, which have limited effectiveness in reducing network load. Therefore, improving existing event-triggered strategies to further optimize bandwidth utilization has become one of the key research directions. Summary of the Invention

[0004] The purpose of this invention is to propose an adaptive vehicle state estimation method based on dynamic event triggering, which can effectively estimate the state of road vehicles and avoid Zeno behavior.

[0005] The specific technical solution of the present invention is as follows: A vehicle state adaptive estimation method based on dynamic event triggering, comprising the following steps:

[0006] Design an event-driven dynamic event-triggered output estimator to reduce sampling of the true output and obtain a continuous output estimation signal. The specific steps are as follows:

[0007] The relationship between the front and rear wheel slip angles (δ1, δ2) and the sideslip angle δ at the center of gravity of the vehicle is as follows:

[0008]

[0009] Where α is the steering wheel angle, ψ is the yaw rate, V is the vehicle speed, and l1 and l2 are the distances from the front and rear axles to the center of mass, respectively.

[0010] Lateral forces F of front and rear tires Y1 and F Y2 Represented as:

[0011]

[0012] Where μ is the relaxation length, D1 and D2 are the front and rear wheel lateral stiffness, and s is a complex frequency domain variable;

[0013] Lateral acceleration γ t It is given by the following formula:

[0014]

[0015] The equation of motion is:

[0016]

[0017] Where M is the vehicle weight, I z It is the moment of inertia;

[0018] Next, the continuous-time vehicle model described above is transformed into the following continuous linear time-varying system model:

[0019]

[0020] in,

[0021] For the aforementioned continuous vehicle model, in order to reduce the sampling of the actual output and obtain continuous output estimation signals, an output estimator based on dynamic event triggering is designed. First, the estimator is reinitialized through triggering; the dynamic event triggering mechanism is constructed as follows:

[0022]

[0023] Where k is the sequence number of the triggering time, e w Let y(t) = w(t) - y(t) represent the trigger error of the output, where w(t) is the predicted estimate of the output, and t is the trigger error of the output. k t represents the time of the k-th trigger. k+1 Indicates the time of the next trigger, t > t k Dynamic threshold at time Updated by the following formula:

[0024]

[0025] in, and They are The upper and lower bounds of y(t) are given, where y(t) is the actual output. This is the estimated output, where α1 and β1 are S... x The upper and lower bounds of (t), ε is the convergence rate of the observer, and μ x It is S x The attenuation factor of (t), where h is a predefined positive scalar. This represents the upper bound of the squared norm of the output matrix C(t);

[0026] When not triggered, the continuous output estimation signal will be obtained through the following mechanism:

[0027]

[0028] in, An estimated value representing the vehicle's condition. Indicates estimated values ​​of vehicle parameters;

[0029] When the sensor signal is determined to be trigger-required, the estimator will be reinitialized:

[0030] w(t k )=y(t k )

[0031] Design an adaptive observer based on triggered sampling signals to estimate the vehicle's state and avoid Zeno behavior. The specific steps are as follows:

[0032] First, construct the following adaptive estimator:

[0033]

[0034] in, Indicates state estimation; Indicates parameter estimation; S x (t) represents the time-varying state estimation gain; S θ γ(t) represents the time-varying parameter estimation gain; D(t) represents the coefficient matrix of the unknown parameters; μ x μ θ These are the parameters of the estimator to be designed;

[0035] Select e w (t)=w(t)-y(t), η(t)=e x (t)-γ(t)e θ (t), we can obtain the following error system:

[0036]

[0037]

[0038]

[0039] This estimator can simultaneously estimate unknown states and avoid Zeno behavior. The proof is as follows:

[0040] D001: Selecting candidate Lyapunov functions: Differentiate and substitute Integration:

[0041]

[0042] D002: Next, through η(t) = e x (t)-γ(t)e θ (t) replace e x (t), D001 scaled down to:

[0043]

[0044] D003: Based on the objective and α1I n ≤S x (t)≤β1I n α2I n ≤S θ (t)≤β2I n We further scale D002 using Young's inequality:

[0045]

[0046] In the formula, It is the upper bound of the norm square of matrix γ(t);

[0047] Next, we further scale D003 using Young's inequality with parameters:

[0048]

[0049] In the formula,

[0050] D005: Next, the design parameter μ is required. x and μ θ Each condition is satisfied and Make the set exist;

[0051] D006: Combining event triggering mechanisms: Results obtained:

[0052]

[0053] D007: Then, due to sensor accuracy limitations, it is required that ||e y (t)|| 2 ≥ρ, combined with D006, the triggering time interval t of the proposed triggering mechanism can be obtained. k+1 -t k There is a lower bound:

[0054]

[0055] In the formula, ρ, Λ1 and Λ2 are all positive constants. Attached Figure Description

[0056] Figure 1 This is a system structure diagram for an embodiment;

[0057] Figures 2 to 5 An example diagram showing the estimated vehicle state x using the method proposed in this invention;

[0058] Figure 6 This is an example of an estimation diagram of vehicle parameter θ using the method proposed in this invention;

[0059] Figure 7 This is a trigger diagram for event triggering using the method proposed in this invention, as shown in the embodiment. Detailed Implementation

[0060] The present invention will be further illustrated below with reference to specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. After reading the present invention, any modifications of the present invention in various equivalent forms by those skilled in the art will fall within the scope defined by the appended claims.

[0061] like Figure 1 As shown, a vehicle state adaptive estimation method based on dynamic event triggering includes the following steps:

[0062] Step 1: Set initial values ​​for each parameter;

[0063] Step 2: Obtain the state estimation gain S x (t);

[0064] Step 3: Obtain the parameter estimation gain S θ (t) and γ(t);

[0065] Step 4: Sample and obtain the vehicle system output y(t);

[0066] Step 5: Update the threshold parameters

[0067] Step 6: Use threshold parameters With error e w (t), e y Verify the event triggering condition and update the estimated output w(t);

[0068] Step 7: Use the adaptive estimator to output the state estimate of the vehicle system. and parameter estimates

[0069] Step 8: Update error e w (t), e y (t), e θ (t) and e x (t);

[0070] Step 9: Repeat steps 2, 3, 4, 5, 6, 7, and 8 until the simulation time ends.

[0071] An embodiment of the present invention is described below:

[0072] Consider the system parameters shown in the table below: symbol value symbol value symbol value symbol value μ 0.5 <![CDATA[l1]]> 1.12 V 20+2sin(0.1t) <![CDATA[l2]]> 1.565 <![CDATA[I z ]]> 2500 M 1450 α 0.01(sin(t / 4)+sin(t / 12))

[0073] Figure 1 This is a system structure diagram of an embodiment of the present invention; Figures 2 to 5 An estimation diagram of vehicle state x using the method proposed in this invention; Figure 6 This is an estimation diagram of the vehicle parameter θ using the method proposed in this invention; Figure 7 The diagram illustrates the triggering effect of the dynamic event triggering mechanism designed in this invention; from Figures 2 to 6 As can be seen, the method proposed in this invention successfully estimates the system state and parameters simultaneously; from Figure 7 It can be seen that the proposed event triggering mechanism reduces the number of signal transmissions and lowers network bandwidth usage.

[0074] References

[0075] [1]Song, C., Wang, H., Tian, ​​Y., & Zheng, G. (2021). Event-triggered observerdesign for output-sampled systems. Nonlinear Analysis: Hybrid Systems, 43, 101112.

[0076] [2]Besancon,G.,de León-Morales,J.,&Huerta-Guevara,O.(2006).Onadaptive observers for state affine systems.International journal of Control,79(06),581-591.

[0077] [3]Zhang,Q.(2002).Adaptive observer for multiple-input-multiple-output(MIMO)linear time-varying systems.IEEE transactions onautomaticcontrol,47(3),525-529.

Claims

1. A vehicle state adaptive estimation method based on dynamic event triggering, characterized in that, Includes the following steps: Design an output estimator based on dynamic event triggering to reduce sampling of the true output and obtain a continuous output estimation signal. The specific steps are as follows: The relationship between the front and rear wheel slip angles (δ1, δ2) and the sideslip angle δ at the center of gravity of the vehicle is as follows: Where α is the steering wheel angle, ψ is the yaw rate, V is the vehicle speed, and l1 and l2 are the distances from the front and rear axles to the center of mass, respectively. Lateral forces F of front and rear tires Y1 and F Y2 Represented as: Where μ is the relaxation length, D1 and D2 are the front and rear wheel lateral stiffness, and s is a complex frequency domain variable; Lateral acceleration γ t It is given by the following formula: The equation of motion is: Where M is the vehicle weight, I z It is the moment of inertia; Next, the continuous-time vehicle model described above is transformed into the following continuous linear time-varying system model: in, For the aforementioned continuous vehicle model, in order to reduce the sampling of the actual output and obtain continuous output estimation signals, an output estimator based on dynamic event triggering is designed. First, the estimator is reinitialized through triggering. The dynamic event triggering mechanism is constructed as follows: Where k is the sequence number of the triggering time, e w Let y(t) = w(t) - y(t) represent the trigger error of the output, where w(t) is the predicted estimate of the output, and t is the trigger error of the output. k t represents the time of the k-th trigger. k+1 Indicates the time of the next trigger, t > t k Dynamic threshold at time Updated by the following formula: in, and They are The upper and lower bounds of y(t) are given, where y(t) is the actual output. This is the estimated output, where α1 and β1 are S... x The upper and lower bounds of (t), ε is the convergence rate of the observer, and μ x It is S x The attenuation factor of (t), where h is a predefined positive scalar. This represents the upper bound of the squared norm of the output matrix C(t); When not triggered, the continuous output estimation signal will be obtained through the following mechanism: in, An estimated value representing the vehicle's condition. Indicates estimated values ​​of vehicle parameters; When the sensor signal is determined to be trigger-required, the estimator will be reinitialized: w(t k )=y(t k ) Design an adaptive observer based on triggered sampling signals to estimate the vehicle's state and avoid Zeno behavior. The specific steps are as follows: First, construct the following adaptive estimator: in, Indicates state estimation; Indicates parameter estimation; S x (t) represents the time-varying state estimation gain; S θ γ(t) represents the time-varying parameter estimation gain; D(t) represents the coefficient matrix of the unknown parameters; μ x μ θ These are the parameters of the estimator to be designed; Select e w (t)=w(t)-y(t), η(t)=e x (t)-γ(t)e θ (t), we can obtain the following error system: This estimator can simultaneously estimate unknown states and avoid Zeno behavior.

Images