An anti-interference adaptive new energy vehicle electronic hydraulic brake system control method

Abstract

The application discloses a new energy vehicle electronic hydraulic brake system control method with anti-interference self-adaptation, comprising the following steps: establishing a nonlinear model between a master cylinder piston position and a brake pressure; adopting a recursive least square method with a variable forgetting factor to update coefficients of a third-order polynomial in real time online; converting a desired target brake pressure into a target piston position trajectory, and designing a multi-constrained model predictive controller; adopting an improved extended LuGre friction model to describe a friction behavior inside the system, and designing an extended Kalman filter to identify friction parameters online in real time, and then feeding the identification results to the model predictive controller; based on the nonlinear model, the model predictive controller and the extended Kalman filter, using motor rotation angle, current and piston position information, estimating the master cylinder pressure in real time, and finally realizing closed-loop pressure control. The application can improve the robustness and smoothness of control.

Description

Technical Field

[0001] This invention relates to the field of new energy vehicle technology, and more specifically to a disturbance-resistant and adaptive control method for an electro-hydraulic braking system for new energy vehicles. Background Technology

[0002] As a core actuator for realizing the intelligence and functional safety of new energy vehicles, the control performance of the electro-hydraulic braking system (EHB) directly affects the accuracy, smoothness, and reliability of the braking process. Existing systems mostly design controllers based on linear or fixed parameter models, which are insufficient to fully address the strong nonlinearity, time-varying parameters, and complex disturbances that exist during the braking process.

[0003] Specifically, the mapping relationship between master cylinder pressure and piston displacement drifts significantly with brake pad wear and hydraulic oil temperature changes; the nonlinear friction effect inside the system, especially its asymmetric and pre-slip characteristics, severely restricts the position tracking accuracy; in addition, the physical constraints of the actuator and the cost of sensors are also challenges that cannot be ignored in engineering practice.

[0004] Although advanced methods such as model predictive control have shown advantages in handling constraints, their performance is highly dependent on the accuracy of the model. In the absence of online adaptive and disturbance compensation mechanisms, the robustness and smoothness of control need to be improved. Summary of the Invention

[0005] In view of this, the present invention provides a disturbance-resistant and adaptive control method for an electronic hydraulic braking system of new energy vehicles to improve the robustness and smoothness of control.

[0006] A disturbance-resistant and adaptive control method for an electro-hydraulic braking system of a new energy vehicle, comprising: Step S1: Establish a nonlinear model between the master cylinder piston position and the braking pressure. This nonlinear model uses a third-order polynomial as the basic model structure and introduces an adaptive compensation factor related to temperature and brake pad wear to capture the time-varying nonlinearity of the system's static characteristics. Step S2: Based on the nonlinear model established in step S1, the coefficients of the third-order polynomial are updated online in real time using the recursive least squares method with a variable forgetting factor, so that the nonlinear model can dynamically track the drift of system characteristics caused by component aging and temperature changes. Step S3: Based on the updated nonlinear model in step S2, the desired target braking pressure is converted into the target piston position trajectory. Based on the integrated dynamic model of the system including motor dynamics and hydraulic dynamics, a multi-constraint model predictive controller is designed. The model predictive controller is used to explicitly handle the hard constraints of motor torque and piston position during the optimization process, and introduces a pressure change rate penalty term to ensure control smoothness. Step S4: An improved extended LuGre friction model is used to describe the friction behavior inside the system, and an extended Kalman filter is designed to identify the friction parameters online in real time. The identification results are then fed forward to the model predictive controller in step S3 to achieve active compensation for friction disturbances. Step S5: Based on the nonlinear model, model predictive controller, and extended Kalman filter, the master cylinder pressure is estimated in real time using the motor rotation angle, current, and piston position information. The master cylinder pressure is used as a feedback input to the model predictive controller and, on the other hand, replaces the physical pressure sensor signal for model error calculation using the recursive least squares method in step S2, thereby achieving closed-loop pressure control.

[0007] The anti-disturbance adaptive control method for the electro-hydraulic braking system of new energy vehicles provided by the present invention has the following beneficial effects: (1) This invention introduces a nonlinear model with environmental compensation capability and uses a recursive least squares method with a variable forgetting factor to update the coefficients of the third-order polynomial in the nonlinear model in real time. This enables the control system to actively sense and respond to the drift of model parameters caused by temperature fluctuations, component wear, etc., and realizes accurate online identification and adaptive tracking of the dynamic characteristics of the system.

[0008] (2) The model predictive controller (MPC) designed in this invention not only explicitly handles the physical constraints of the actuator to ensure the safe and stable operation of the system, but also innovatively introduces the pressure change rate as a key performance indicator in the cost function as an optimization objective, achieving high-performance optimization control under multi-objective constraints and ensuring control smoothness.

[0009] (3) Active observation and compensation for complex nonlinear frictional disturbances are achieved. This invention uses an improved extended LuGre frictional model to describe the internal frictional behavior of the system, and combines an extended Kalman filter to accurately identify the internal frictional state and key parameters that are difficult to measure directly online. The identification results are integrated into the predictive control framework of MPC in a feedforward form, realizing "predictive" compensation for frictional force, rather than the traditional passive suppression, thereby significantly reducing the impact of nonlinear friction on control accuracy and improving the robustness of control.

[0010] (4) A system state reconstruction capability based on soft sensing was constructed, and the system architecture was optimized. Based on the aforementioned high-precision model (nonlinear model, model predictive controller), a reliable estimation of the key variable (master cylinder pressure) was achieved by fusing low-cost and easily measurable signals through an extended Kalman filter. This invention can reduce the dependence on high-cost physical sensors without sacrificing performance, which not only improves the economy of the system, but also enhances the overall reliability by reducing potential failure points. Attached Figure Description

[0011] Figure 1 A flowchart illustrating the disturbance-resistant adaptive control method for an electronic hydraulic braking system of a new energy vehicle provided in an embodiment of the present invention.

[0012] Figure 2 This is a comparison chart of the hydraulic pressure tracking performance of the method of the present invention and the traditional PID following method under all operating conditions.

[0013] Figure 3 This is a comparison chart of the hydraulic pressure tracking error between the method of the present invention and the conventional PID control.

[0014] Figure 4 This diagram illustrates the convergence process of parameter estimation error in the recursive least squares method with a variable forgetting factor in this invention. Detailed Implementation

[0015] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain embodiments of the present invention, and should not be construed as limiting the present invention.

[0016] Please see Figure 1 The disturbance-resistant adaptive control method for an electronic hydraulic braking system of a new energy vehicle provided by the present invention includes steps S1 to S5: Step S1: Establish a nonlinear model between the master cylinder piston position and the braking pressure. This nonlinear model uses a third-order polynomial as the basic model structure and introduces an adaptive compensation factor related to temperature and brake pad wear to capture the time-varying nonlinearity of the system's static characteristics.

[0017] The expression for the nonlinear model is:

[0018]

[0019] in, for Master cylinder pressure at any given time; for The piston position at any given moment; , , , The coefficients of the third-order polynomial to be identified; For temperature and the degree of brake pad wear Related adaptive compensation factors; , This is the compensation coefficient; For reference temperature; For reference to the degree of wear.

[0020] temperature The degree of brake pad wear can be obtained from a temperature sensor mounted on the master cylinder or brake caliper. The number of braking events and energy can be accumulated through algorithms or indirectly estimated based on model observations.

[0021] Step S2: Based on the nonlinear model established in step S1, the coefficients of the third-order polynomial are updated online in real time using the recursive least squares method with a variable forgetting factor, so that the nonlinear model can dynamically track the drift of system characteristics caused by component aging and temperature changes.

[0022] Specifically, the polynomial coefficients in step S1 are updated in real time using the recursive least squares method with a variable forgetting factor (VFF-RLS). , , , To adapt to dynamic changes in the system; the forgetting factor is no longer a constant, but is dynamically adjusted according to the model fitting error.

[0023] Specifically, the expression for the variable forgetting factor is:

[0024]

[0025]

[0026] in, for The forgetting factor of time; This represents the maximum value of the forgetting factor. This represents the minimum value of the forgetting factor. The attenuation coefficient; for The time estimation error is specifically... The deviation between the pressure sensor's measured value and the model's estimated value at any given time; For regression vectors; Indicates transpose; express The parameter vector estimated at time step, i.e. Estimated value , , , for The coefficients of the third-order polynomial at time t.

[0027] When estimation error When it gets bigger, Automatically reduce the weight of new data, assigning higher weights to them and accelerating the algorithm's learning speed; when the system stabilizes... Increase the size to enhance the algorithm's ability to withstand disturbances.

[0028] Step S3: Based on the updated nonlinear model in step S2, the desired target braking pressure is converted into the target piston position trajectory. Based on the integrated dynamic model of the system including motor dynamics and hydraulic dynamics, a multi-constraint model predictive controller is designed. The model predictive controller is used to explicitly handle the hard constraints of motor torque and piston position during the optimization process, and introduces a pressure change rate penalty term to ensure control smoothness.

[0029] Among them, the cost function of the Model Predictive Controller (MPC) for:

[0030] in, For prediction in the time domain; To control the time domain; For at any time Predicted future moments Piston position; For the future Piston position reference trajectory; For a moment Calculated and acting on future moments In this embodiment, the control increment is specifically the control increment of the motor torque. Directly optimizing the control quantity in MPC helps to ensure control smoothness. For at any time Predicting future moments Main cylinder pressure; For at any time Predicting future moments Main cylinder pressure; This is the weight matrix of the error in state tracking; The weight matrix is ​​used to control the increment; The weight matrix for the rate of pressure change explicitly penalizes the rate of pressure change, thus introducing a penalty term for the rate of pressure change into the cost function of the MPC controller.

[0031] In this embodiment, the constraints of the model prediction controller are:

[0032]

[0033]

[0034] in, For a moment Calculated and acting on future moments In this embodiment, the control quantity is specifically the control quantity of the motor torque; , In this embodiment, the minimum and maximum values ​​of the control quantity are specifically the minimum and maximum values ​​of the motor torque control quantity. , In this embodiment, the minimum and maximum values ​​of the increment are specifically the minimum and maximum values ​​of the motor torque control increment. , These represent the minimum and maximum values ​​of the piston position.

[0035] The aforementioned constraints ensure that the motor does not overload and the piston does not overtravel, guaranteeing the physical safety of the system. MPC, through online optimization, naturally incorporates these constraints, which is difficult to achieve with traditional control methods. Therefore, the MPC in step S3 is the core actuator of the entire control strategy. It fully utilizes the accurate model and state information provided by the preceding steps to achieve smooth and accurate pressure tracking control under strict physical constraints.

[0036] Step S4: An improved extended LuGre friction model is used to describe the friction behavior inside the system, and an extended Kalman filter is designed to identify the friction parameters online in real time. The identification results are then fed forward to the model predictive controller in step S3 to achieve active compensation for friction disturbances.

[0037] Nonlinear friction within the system remains the primary source of disturbance affecting control accuracy. To further improve performance, this invention incorporates friction identification and compensation. Specifically, an improved extended LuGre friction model is employed, and an extended Kalman filter is designed to identify and compensate for friction parameters online.

[0038] The dynamic equations of the improved extended LuGre friction model are as follows:

[0039]

[0040]

[0041]

[0042] in, This is the total frictional force; This is the bristle stiffness coefficient; This represents the average deformation of the bristles; The rate of change of mane deformation over time; This refers to the bristle damping coefficient; It is the coefficient of viscous friction; The relative velocity of the pistons; This is the Stribeck function, used to describe the static properties of friction as a function of velocity; Coulomb friction; It is static friction; Stribeck characteristic velocity; The asymmetric friction compensation coefficient is updated through online identification. It is a function used to describe the asymmetry of frictional forces during the reciprocating motion of a piston.

[0043] In this embodiment, the design of the extended Kalman filter for online real-time identification of friction parameters specifically includes: Define the extended state vector of the extended Kalman filter. for:

[0044] Nonlinear state equations for:

[0045] in, For control input; For process noise; The state prediction formula and covariance prediction formula in the prediction step are as follows:

[0046]

[0047] in, For at any time Prior state estimation; It is a nonlinear state transition function; For at any time Posterior state estimation; For at any time Control input; For at any time The prior estimate of the covariance matrix; For at any time The prior estimate of the covariance matrix; Is it a nonlinear function in Jacobian matrix at the location; The process noise covariance matrix; The Kalman gain calculation, state update formula, and covariance update formula in the update step are as follows:

[0048]

[0049]

[0050] in, For at any time Kalman gain; For at any time The Jacobian matrix of the observation matrix; To measure the noise covariance matrix; For at any time The actual measured value; This is a predicted value for the observed quantity; Indicates measurement residuals; For at any time Posterior state estimation; For at any time The posterior estimated covariance matrix; It is the identity matrix; Finally, the friction parameters (including) are analyzed using an extended Kalman filter. , , , , , , The friction parameters identified online by the extended Kalman filter are used to calculate the currently estimated friction force in real time. This estimated friction force is then input as a measurable feedforward perturbation into the MPC in step S3.

[0051] Step S5: Based on the nonlinear model, model predictive controller, and extended Kalman filter, the master cylinder pressure is estimated in real time using the motor rotation angle, current, and piston position information. The master cylinder pressure is used as a feedback input to the model predictive controller and, on the other hand, replaces the physical pressure sensor signal for model error calculation using the recursive least squares method in step S2, thereby achieving closed-loop pressure control.

[0052] The accuracy of master cylinder pressure estimation heavily depends on model precision. In this embodiment, steps S1 and S2 provide accurate pressure-position relationships, and step S4 provides a precise friction model, which together ensure the reliability of the state-space model. Furthermore, in the VFF-RLS step S2, if a physical sensor is used, its measurements are used for parameter updates; however, in this embodiment, a soft sensor is used, allowing the estimated values ​​to participate in or assist parameter updates under certain conditions, forming a complete closed loop. Ultimately, during normal operation, the system can achieve high-precision pressure control without relying on expensive physical pressure sensors, using only the motor's own sensors and advanced algorithms, significantly improving the system's economy and reliability.

[0053] To verify the effectiveness of the present invention, relevant simulation tests were conducted in this embodiment.

[0054] Figure 2 This chart compares the performance of the method of this invention with that of the traditional PID follower method under full operating conditions for hydraulic pressure tracking. It simulates a complete braking cycle including gradual increase, holding, step-by-step pressurization, high-frequency regulation, and pressure release. The target hydraulic pressure varies within the range of 0-0.7 MPa. The performance of the multi-constraint model predictive controller of this invention is compared between the traditional PID follower and the method of this invention. Figure 2 As can be seen, the method of the present invention can closely track the target pressure curve throughout the braking process, while the traditional PID control exhibits obvious tracking delay and overshoot in the rapid pressure change and high-frequency adjustment stages. Especially under high-frequency pulsating conditions (6.5-8.5s), the present invention can effectively suppress pressure fluctuations and maintain stable pressure output.

[0055] Figure 3 This is a comparison chart of the hydraulic pressure tracking error between the method of this invention and traditional PID control, quantitatively demonstrating the tracking accuracy of the two control strategies under different operating conditions; from Figure 3 As can be seen, the tracking error of the present invention (i.e. Figure 3 The MPC error in this invention is always controlled within ±20 kPa, and the error fluctuation is significantly smaller than that of traditional PID error. Especially in the slow increase phase (0-2.5s) and high-frequency adjustment phase (6.5-8.5s) of rapid pressure change, the peak error of this invention is reduced by about 60% compared with PID control, which verifies the superiority of online parameter identification and multi-constraint MPC control in this invention in dealing with the nonlinear and time-varying characteristics of the system.

[0056] Figure 4 This diagram illustrates the convergence process of the parameter estimation error using the recursive least squares method with a variable forgetting factor in this invention, demonstrating the system's online identification capability of the pressure-position nonlinear mapping model parameters during operation; from Figure 4 As can be seen, the parameter estimation error converges rapidly after system startup, decreasing from the initial 10 within 2 seconds. 2 The magnitude dropped to 10 0 The magnitude of the error was measured, and the system maintained a stable small error state throughout the simulation process. This verifies that the update mechanism proposed in step S2 of this invention can effectively track the drift of system characteristics caused by component aging, temperature changes, etc., and provides an accurate system model for the MPC controller. It is a key technical support for achieving high-precision pressure tracking.

[0057] In summary, the disturbance-resistant adaptive control method for the electro-hydraulic braking system of new energy vehicles according to the above embodiments has the following beneficial effects: (1) This invention introduces a nonlinear model with environmental compensation capability and uses a recursive least squares method with a variable forgetting factor to update the coefficients of the third-order polynomial in the nonlinear model in real time. This enables the control system to actively sense and respond to the drift of model parameters caused by temperature fluctuations, component wear, etc., and realizes accurate online identification and adaptive tracking of the dynamic characteristics of the system.

[0058] (2) The model predictive controller (MPC) designed in this invention not only explicitly handles the physical constraints of the actuator to ensure the safe and stable operation of the system, but also innovatively introduces the pressure change rate as a key performance indicator in the cost function as an optimization objective, achieving high-performance optimization control under multi-objective constraints and ensuring control smoothness.

[0059] (3) Active observation and compensation for complex nonlinear frictional disturbances are achieved. This invention uses an improved extended LuGre frictional model to describe the internal frictional behavior of the system, and combines an extended Kalman filter to accurately identify the internal frictional state and key parameters that are difficult to measure directly online. The identification results are integrated into the predictive control framework of MPC in a feedforward form, realizing "predictive" compensation for frictional force, rather than the traditional passive suppression, thereby significantly reducing the impact of nonlinear friction on control accuracy and improving the robustness of control.

[0060] (4) A system state reconstruction capability based on soft sensing was constructed, and the system architecture was optimized. Based on the aforementioned high-precision model (nonlinear model, model predictive controller), a reliable estimation of the key variable (master cylinder pressure) was achieved by fusing low-cost and easily measurable signals through an extended Kalman filter. This invention can reduce the dependence on high-cost physical sensors without sacrificing performance, which not only improves the economy of the system, but also enhances the overall reliability by reducing potential failure points.

[0061] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A disturbance-resistant and adaptive control method for an electro-hydraulic braking system of a new energy vehicle, characterized in that, include: Step S1: Establish a nonlinear model between the master cylinder piston position and the braking pressure. This nonlinear model uses a third-order polynomial as the basic model structure and introduces an adaptive compensation factor related to temperature and brake pad wear to capture the time-varying nonlinearity of the system's static characteristics. Step S2: Based on the nonlinear model established in step S1, the coefficients of the third-order polynomial are updated online in real time using the recursive least squares method with a variable forgetting factor, so that the nonlinear model can dynamically track the drift of system characteristics caused by component aging and temperature changes. Step S3: Based on the updated nonlinear model in step S2, the desired target braking pressure is converted into the target piston position trajectory. Based on the integrated dynamic model of the system including motor dynamics and hydraulic dynamics, a multi-constraint model predictive controller is designed. The model predictive controller is used to explicitly handle the hard constraints of motor torque and piston position during the optimization process, and introduces a pressure change rate penalty term to ensure control smoothness. Step S4: An improved extended LuGre friction model is used to describe the friction behavior inside the system, and an extended Kalman filter is designed to identify the friction parameters online in real time. The identification results are then fed forward to the model predictive controller in step S3 to achieve active compensation for friction disturbances. Step S5: Based on the nonlinear model, model predictive controller, and extended Kalman filter, the master cylinder pressure is estimated in real time using the motor rotation angle, current, and piston position information. The master cylinder pressure is used as a feedback input to the model predictive controller and also replaces the physical pressure sensor signal for model error calculation using the recursive least squares method in step S2, thereby achieving closed-loop pressure control. In step S1, the expression for the nonlinear model is: in, for Master cylinder pressure at any given time; for The piston position at any given moment; , , , The coefficients of the third-order polynomial to be identified; For temperature and the degree of brake pad wear Related adaptive compensation factors; , The compensation coefficient; For reference temperature; For reference to the degree of wear.

2. The disturbance-resistant adaptive control method for an electronic hydraulic braking system of a new energy vehicle according to claim 1, characterized in that, In step S2, the expression for the variable forgetting factor is: in, for The forgetting factor of time; This represents the maximum value of the forgetting factor. This represents the minimum value of the forgetting factor. The attenuation coefficient; for The time estimation error; For regression vectors; Indicates transpose; express The parameter vector estimated at time step.

3. The disturbance-resistant adaptive control method for an electronic hydraulic braking system of a new energy vehicle according to claim 2, characterized in that, The cost function of the model predictive controller for: in, For prediction in the time domain; To control the time domain; For at any time Predicted future moments Piston position; For the future Piston position reference trajectory; For a moment Calculated and acting on future moments Control increment; For at any time Predicting future moments Main cylinder pressure; For at any time Predicting future moments Main cylinder pressure; This is the weight matrix of the error in state tracking; The weight matrix is ​​used to control the increment; This is the weight matrix for the rate of change of pressure.

4. The anti-disturbance adaptive control method for an electronic hydraulic braking system of a new energy vehicle according to claim 3, characterized in that, In step S3, the constraints of the model predictive controller are: in, For a moment Calculated and acting on future moments The control quantity, , To control the minimum and maximum values ​​of the quantity, , To control the minimum and maximum values ​​of the increment, , These represent the minimum and maximum values ​​of the piston position.

5. The disturbance-resistant adaptive control method for an electronic hydraulic braking system of a new energy vehicle according to claim 4, characterized in that, In step S4, the dynamic equations of the improved extended LuGre friction model are: in, This is the total frictional force; This is the bristle stiffness coefficient; This represents the average deformation of the bristles; The rate of change of mane deformation over time; This refers to the bristle damping coefficient; It is the coefficient of viscous friction; The relative velocity of the pistons; This is the Stribeck function, used to describe the static properties of friction as a function of velocity; Coulomb friction; It is static friction; Stribeck characteristic velocity; The asymmetric friction compensation coefficient is updated through online identification. It is a function used to describe the asymmetry of frictional forces during the reciprocating motion of a piston.

6. The disturbance-resistant adaptive control method for an electronic hydraulic braking system of a new energy vehicle according to claim 5, characterized in that, In step S4, the design of the extended Kalman filter for online real-time identification of friction parameters specifically includes: Define the extended state vector of the extended Kalman filter. for: Nonlinear state equations for: in, For control input; For process noise; The state prediction formula and covariance prediction formula in the prediction step are as follows: in, For at any time Prior state estimation; It is a nonlinear state transition function; For at any time Posterior state estimation; For at any time Control input; For at any time The prior estimate of the covariance matrix; For at any time The prior estimate of the covariance matrix; Is it a nonlinear function in Jacobian matrix at the location; The process noise covariance matrix; The Kalman gain calculation, state update formula, and covariance update formula in the update step are as follows: in, For at any time Kalman gain; For at any time The Jacobian matrix of the observation matrix; To measure the noise covariance matrix; For at any time The actual measured value; This is a predicted value for the observed quantity; Indicates measurement residuals; For at any time Posterior state estimation; For at any time The posterior estimated covariance matrix; It is the identity matrix; Finally, the friction parameters are identified online in real time using an extended Kalman filter.

Images