Engine boost pressure control method based on nonlinear algebraic filtering and ascending and descending order observation

Abstract

The invention belongs to the technical field of internal combustion engine electric control, and discloses an engine supercharging pressure control method based on nonlinear algebraic filtering and step-up and step-down observation, which comprises the following steps: step 1, an electric supercharger supercharging pressure dynamic model outputs supercharging pressure according to an eTurbo target rotating speed at the current moment; the nonlinear algebraic filter module calculates an original tracking error of the boost pressure according to the target boost pressure and the boost pressure measurement value, and then performs normalization, scaling and reconstruction to obtain a smoothed tracking error sum; step 3, increasing and decreasing the order of the expansion state observer to input and output boost pressure change rates and estimated values of first-order derivatives and second-order derivatives of the boost pressure change rates; and step 4, calculating the eTurbo target rotating speed at the next moment through an active-disturbance-rejection control law. According to the method, accurate estimation of rapid dynamic disturbance and intelligent suppression of measurement noise are realized, and the dynamic response, the anti-interference capability and the steady-state stability of eTurbo control are effectively improved.

Description

Technical Field

[0001] This invention belongs to the field of internal combustion engine electronic control technology, specifically relating to an engine boost pressure control method based on nonlinear algebraic filtering and rising and falling order observation, and in particular an intelligent control method combining nonlinear algebraic filtering and rising and falling order extended state observer, which is suitable for improving the dynamic response, disturbance rejection capability and steady-state stability of the boost system. Background Technology

[0002] The evolution of internal combustion engine technology is converging towards higher efficiency and deeper electrification. Against this backdrop, the electric turbocharger (eTurbo) has emerged as a transformative technology. By coaxially integrating a high-speed electric motor into a traditional turbocharger, it achieves rapid boost pressure establishment and independent control, effectively solving the inherent "turbo lag" problem of traditional exhaust gas turbochargers. This significantly improves the engine's low-speed torque output and transient response capabilities, making it a key component of modern high-efficiency internal combustion engines and hybrid power systems. Existing technologies have focused on improving the dynamic performance of turbocharging systems through system architecture optimization. For example, the patent application CN202110722029.0, "Diesel Engine Constant Speed ​​Rapid Torque Boosting System and Control Method," employs an electrically assisted turbo dual-boost system combined with variable valve timing and fuel injection coordinated control to achieve rapid torque boosting. This scheme demonstrates the potential of electrified turbocharging and multi-system synergy in improving engine transient performance.

[0003] However, the high-performance potential of the eTurbo system is highly dependent on the precision of its electronic control system. The system faces a series of intertwined and formidable challenges in its engineering implementation, placing near-stringent demands on the robustness, response speed, and steady-state accuracy of the control algorithm.

[0004] The eTurbo control system operates in a highly dynamic and uncertain environment. It faces a variety of complex disturbances: periodic exhaust pulses from the engine cause severe, high-frequency fluctuations in turbine input torque; the actions of other actuators in the intake system (such as the throttle valve) can trigger sudden changes in intake flow; the high-speed rotor system exhibits complex nonlinear friction effects (Coulomb friction, viscous friction, etc.); system rotational inertia and motor parameters (such as flux linkage and resistance) drift with changes in temperature and operating point; and the boost pressure sensor signal used for feedback control is inevitably mixed with high-frequency measurement noise. These intense dynamic disturbances, coupled with high-frequency noise, severely interfere with the controller's perception and judgment of the system's true state, directly affecting the accuracy of observation and the steady-state performance of closed-loop control.

[0005] To suppress measurement noise, linear low-pass filters are commonly used in engineering to preprocess the raw sensor signals. For example, the "Optimization Control Method for Target Booster Pressure" (application number CN202411442689.3) uses a first-order low-pass filter to process the EGR rate characteristic coefficient and dynamically optimizes the target boost pressure, aiming to improve the stability and EGR response accuracy of the boost closed-loop control. While this method can smooth the signal to some extent, its inherent defects are amplified in systems like eTurbo, which have extremely high dynamic performance requirements: linear filters inevitably introduce non-negligible phase lag in the passband while filtering out noise. This lag causes the filtered signal to fail to accurately reflect the real-time state of the system, resulting in delays in state observation and feedback control actions based on this signal. To reduce the impact of phase lag on dynamics, the cutoff frequency of the filter needs to be increased, but this weakens its noise suppression capability; conversely, lowering the cutoff frequency to obtain a cleaner signal introduces even greater phase lag, impairing the system's fast response capability and stability margin. This contradiction is difficult to reconcile within the traditional linear filtering framework.

[0006] Active disturbance rejection control (ADRC) and its core—the extended state observer (ESO)—have gained attention in industrial control and have been attempted to be applied to power system control due to their ability to uniformly estimate and compensate for the system's "total disturbance." However, directly applying traditional ESOs to eTurbo boost pressure control reveals the following key shortcomings: Most existing studies, in order to simplify the design, model the intake manifold pressure dynamics as a first-order system. However, the physical essence of eTurbo's process from motor torque command to manifold pressure output is a series dynamic process: motor torque affects rotor speed (first-order inertia), speed changes affect intake flow through the compressor, and ultimately change manifold pressure (first-order volumetric effect). Therefore, its dominant dynamic is actually a second-order system. ESO designs based on first-order models are structurally mismatched with the physical essence of the object, fundamentally limiting the controller's performance potential; standard linear ESOs are usually based on the premise that "the rate of change of the total disturbance is approximately zero ( The assumption is based on the fact that the ESO (Electronic Stability Optimizer) is too simplistic for rapidly changing disturbances caused by exhaust pulses and other factors in eTurbo. Based on this assumption, the ESO exhibits significant lag in tracking such dynamic disturbances, leading to untimely compensation and severely limiting the system's dynamic anti-interference capability. The ESO corrects for the error between the output measured value and the estimated value. When the measured signal contains noise, the high-gain (high-bandwidth) ESO amplifies these noise components equally in its disturbance estimation channel, resulting in an inflated estimated total disturbance value. This generates severe high-frequency fluctuations. These fluctuations are directly transmitted to the control input (such as the target speed or torque command of the motor) through the control law, causing high-frequency chattering of the actuator. This not only increases energy consumption but also exacerbates mechanical wear and deteriorates noise, vibration, and harshness (NVH) performance.

[0007] The aforementioned issues reveal a core contradiction in eTurbo's high-performance control: improving the system's dynamic response and disturbance rejection capability requires increasing the bandwidth of the observer and controller, but this amplifies noise, leading to deterioration of steady-state performance; conversely, employing strong filtering or low-bandwidth strategies to ensure steady-state stability sacrifices dynamic performance. Traditional control architectures struggle to fundamentally resolve this contradiction. Therefore, developing a novel observation and filtering collaborative system capable of comprehensively addressing these issues has significant theoretical and engineering value for unlocking the full performance potential of eTurbo and promoting the development of advanced internal combustion engine control technology. Summary of the Invention

[0008] The purpose of this invention is to address the technical deficiencies in the existing technology by providing an engine boost pressure control method based on nonlinear algebraic filtering and order rise and fall observation.

[0009] The technical solution adopted to achieve the purpose of this invention is:

[0010] An engine boost pressure control method based on nonlinear algebraic filtering and order rise / fall observation includes the following steps:

[0011] Step 1: The dynamic model of electric supercharger boost pressure is based on the current eTurbo target speed. Output boost pressure The model includes the rate of change of boost pressure. ;

[0012] Step 2, the nonlinear algebraic filter module is based on the target boost pressure. Compared with the boost pressure measurement value The original tracking error for calculating boost pressure Then, normalization is performed to obtain the normalization error. Non-linear scaling function right Scaling and reconstruction are performed to obtain the smoothed tracking error. ,in accordance with and Calculate the smoothed boost pressure ;

[0013] Step 3, the ascending and descending order extended state observer with As input, output the rate of change of boost pressure and its estimated first and second derivatives. , and ;

[0014] Step 4, based on , , And the estimated system parameters are used to calculate the target eTurbo speed at the next moment using an active disturbance rejection control law. This feedback is then sent to the speed control layer to form a closed loop, achieving boost pressure. Target boost pressure Tracking.

[0015] In the above technical solution, the dynamic model of the boost pressure of the electric booster in step 1 is as follows:

[0016]

[0017] Let be the boost pressure at time t. Let be the first derivative of the pressurization pressure at time t. Let be the second derivative of the pressurization pressure at time t. Let t be the target eTurbo rotational speed. These are the dynamic coefficients of the damping term and the dynamic coefficients of the restoring force term, respectively. To control the gain, For constant perturbation, Let be the rate of change of boost pressure at time t.

[0018] In the above technical solution, in step 2, the original tracking error The calculation formula is:

[0019]

[0020] in, For discrete time series labeling, for The original tracking error at time step for The target boost pressure at any given time for The measured value of the boost pressure at any given moment;

[0021] Normalized error The calculation formula is:

[0022] ;

[0023] in The preset dead zone threshold, , for Normalization error at time step;

[0024] Preferred, , The standard deviation of noise is measured for pressure sensors.

[0025] In the above technical solution, in step 2, the nonlinear scaling function The expression is:

[0026]

[0027] Smoothed tracking error The calculation formula is:

[0028]

[0029] for Smoothed tracking error at different times , for Normalization error at time step;

[0030] Smoothed boost pressure .

[0031] In the above technical solution, in step 3, the state update equation of the ascending-descending expansion state observer is:

[0032]

[0033]

[0034] ;

[0035] , , They are respectively The intermediate state variables of the observer at time t. , , They are respectively The intermediate state variables of the observer at time t. , , These are the coefficient matrices of the state update equation. , , These are intermediate variables, The sampling period of the digital control system. For the observer bandwidth, preferably, , The expected bandwidth of the closed-loop system;

[0036] The output equation of the ascending-descending-level extended state observer is:

[0037] ;

[0038] , and for The rate of change of boost pressure at time t and the estimated values ​​of its first and second derivatives.

[0039] In the above technical solution, the active disturbance rejection control law in step 4 takes the following form:

[0040]

[0041] for Smoothed tracking error at different times , For controller gain, For proportional gain, the preferred option is... , , The expected bandwidth of the closed-loop system. For the damping ratio, Let be the estimated value of the first derivative of the boost pressure at time t. These are the parameters in the dynamic model of the boost pressure of the electric supercharger. Real-time estimates, for The estimated rate of change of boost pressure at time t.

[0042] In another aspect of the present invention, a control device for an electric supercharger includes a processor and a memory; the memory is used to store program code and transmit the program code to the processor.

[0043] The processor is used to execute the engine boost pressure control method based on nonlinear algebraic filtering and order rise and fall observation according to the instructions in the program code.

[0044] Another aspect of the present invention includes a computer-readable storage medium storing computer-executable instructions, which, when executed, are used to implement the engine boost pressure control method based on nonlinear algebraic filtering and order rise and fall observation.

[0045] In another aspect of the present invention, an electric supercharger includes an electric supercharger body and the aforementioned control device.

[0046] In another aspect of the invention, there is an engine comprising the aforementioned electric supercharger.

[0047] Compared with the prior art, the beneficial effects of the present invention are:

[0048] 1. Precise model matching and dynamic disturbance capture: The second-order system model built on the physical essence of eTurbo enables more accurate matching between the controller structure and the dynamic characteristics of the object. ARO-ESO-2 greatly improves the tracking speed and compensation timeliness of rapidly changing disturbances (such as exhaust pulses) by jointly estimating the disturbance and its derivative, fundamentally overcoming the disturbance estimation lag problem of traditional second-order ESOs.

[0049] 2. Intelligent Context-Aware Noise Suppression: The nonlinear algebraic filter overcomes the contradiction between phase lag and noise suppression inherent in traditional linear filters. Its ability to adaptively adjust gain based on error magnitude allows the system to maintain a fast response under large transient errors and powerfully filter out noise under small steady-state errors, achieving a balance between dynamic performance and steady-state stability.

[0050] 3. Synergistic Optimization of Observation and Control: The smoothed pressure signal serves as input to the ascending-descending order dynamic disturbance (ESO), allowing the ESO to employ a higher bandwidth to enhance dynamic disturbance estimation capabilities without excessive concern about noise amplification. Simultaneously, the smoothed error is used for control law calculation, significantly reducing high-frequency fluctuations in control commands. This synergy creates a virtuous cycle of "filtering-observation-control."

[0051] 4. Superior comprehensive control performance: Through simulation and bench testing, the control system of this invention, compared with the traditional PID controller, can reduce the boost pressure overshoot from more than 4.5% to almost zero, shorten the settling time by more than 17%, reduce steady-state pressure fluctuation by about 62%, and smooth the control input, effectively improving the system's NVH performance and actuator life.

[0052] 5. Clear engineering architecture and good feasibility: This invention integrates nonlinear filtering and order raising / lowering observers in a modular manner, with a clear structure and well-defined physical meanings of the parameters (such as dead zones). ,bandwidth This provides a system tuning guide. Its discretized form is easy to implement in embedded controllers (such as ECUs) and has great potential for engineering applications. Attached Figure Description

[0053] Figure 1 The diagram shown is an overall block diagram of the present invention.

[0054] Figure 2 The figure shown is a comparison of the boost pressure following effects of second-order ESO and rising-falling-order ESO. Detailed Implementation

[0055] The present invention will be further described in detail below with reference to specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0056] Example 1

[0057] An engine boost pressure control method based on nonlinear algebraic filtering and order rise / fall observation includes the following steps:

[0058] Step 1: Construct a dynamic model of the electric supercharger boost pressure in the supercharging system. The control input of the model is the current eTurbo target speed. The controlled output is boost pressure. The model includes the rate of change of boost pressure. .

[0059] The dynamic model of the electric supercharger's boost pressure is a second-order linearized model, which includes unmodeled dynamic disturbances and external interferences. The main energy path of the eTurbo system is: motor torque → rotor acceleration → speed change → compressor work change → intake flow rate change → manifold pressure change. There are two main inertial / energy storage components: rotor mechanical dynamics and intake manifold volume dynamics.

[0060] Combining the two inertial elements mentioned above, and ignoring higher-order coupling details, we establish a system from... To the actual manifold pressure output The dominant dynamic model between them. It considers the system's inherent damping, possible natural pressure leakage / attenuation, and reduces all unmodeled factors (nonlinearity, coupling, internal and external disturbances) to the rate of change of boost pressure. The following dynamic model of the boost pressure of the electric supercharger (second-order linearized model) is established as the basis for the controller design:

[0061]

[0062] In the formula: The boost pressure (controlled output) at time t. Let be the first derivative of the boost pressure at time t. Let be the second derivative of the boost pressure at time t. The target eTurbo speed (control input) at time t. These are the dynamic coefficients of the damping term and the dynamic coefficients of the restoring force term, typically... , (Related to the volume effect), indicating that the open-loop system is stable. To control the gain, a positive number indicates that an increase in rotational speed ultimately leads to an increase in pressure. For modelable constant disturbances or biases (such as the effect of average exhaust back pressure). The rate of change of boost pressure at time t includes: unmodeled dynamics (compressor nonlinearity, coupling effects), external disturbances (exhaust pulse, sudden changes in intake flow), and parameters. The deviation between the actual value and the nominal value.

[0063] This model accurately describes the second-order inertial nature of the eTurbo pressure control loop and reduces all complex uncertainties to a single total disturbance. This laid the foundation for the subsequent design of active disturbance rejection controllers that do not rely on accurate models.

[0064] Step 2, the nonlinear algebraic filter module is based on the target boost pressure. Compared with the boost pressure measurement value The original tracking error for calculating boost pressure Then, normalization is performed to obtain the normalization error. Non-linear scaling function right Scaling and reconstruction are performed to obtain the smoothed tracking error. .

[0065] Step 2.1: The nonlinear algebraic filter module obtains the target boost pressure for the current control cycle. Compared with the boost pressure measurement value Calculate the original tracking error The nonlinear algebraic filter module is used for... The core of context-aware preprocessing lies in dynamically adjusting the filtering strength based on the magnitude of the tracking error.

[0066] The original tracking error The calculation formula is:

[0067] ;

[0068] in, The discrete time series is labeled to indicate the current control period. for The original tracking error at any given moment. A smooth reference trajectory generated from the boost pressure target transition planning layer. for The target boost pressure at any given time Measurement signal from manifold pressure sensor, for The measured value of the boost pressure at any given time.

[0069] Step 2.2, will Normalization:

[0070] ;

[0071] in The preset dead zone threshold, , for Normalization error at time step.

[0072] Preferred, Based on the standard deviation of noise measured by the pressure sensor It is determined that its tuning relationship is as follows:

[0073] .

[0074] Step 2.3, Construct the non-linear scaling function And based on a non-linear scaling function For normalization error Scaling and reconstruction are performed to obtain the smoothed tracking error. Then, the smoothed boost pressure signal is calculated. .

[0075] The nonlinear scaling function Designed to be based on normalized error The input-output relationship of an odd-symmetric polynomial function is defined as follows:

[0076] ;

[0077] This function satisfies odd symmetry. Boundary continuity Boundary derivative continuity And it has attenuation gain at the origin. .

[0078] Using the nonlinear scaling function For the original tracking error The rules for scaling and refactoring are as follows:

[0079] ;

[0080] Right now, for The normalized error at time, when When the absolute value is less than or equal to 1 (small error range), the error is non-linearly scaled. At this point, the scaling gain is 0.25 at the origin, achieving strong smoothing. When the absolute value of the error signal is greater than 1 (large error range), the error signal is basically unscaled, the gain is about 1, and the fast response is maintained.

[0081] The smoothed boost pressure is:

[0082] ;

[0083] Step 3: Construct an ascending and descending order extended state observer for the fifth-order system, in order to The input is an estimate of the rate of change of the boost pressure. ;

[0084] The smoothed boost pressure signal As the primary input to the expansion state observer, it replaces the original noisy boost pressure measurement. This allows the extended state observer to maintain high bandwidth (for fast tracking of disturbances) while being protected from high-frequency measurement noise. The smoothed tracking error... It is directly input into the feedback loop of the active disturbance rejection control law to calculate the control quantity.

[0085] Design of the ascending-descending extended state observer (ARO-ESO-2):

[0086] To estimate the rate of change of boost pressure in real time and accurately To overcome the lag in estimating fast disturbances by traditional observers, this invention designs a rising-falling-order extended state observer. Its core idea is to first increase the order of the description of the dynamic characteristics of the disturbance (rising the order), and then avoid direct differentiation of the output signal through mathematical transformation (falling the order).

[0087] Abandoning the assumption of "the rate of change of disturbance is approximately zero" in traditional extended state observers, a model that is more in line with engineering reality, namely "the disturbance has a continuous derivative," is adopted instead. The rate of change of boost pressure is... The first derivative of the rate of change of boost pressure The second derivative of the rate of change of boost pressure All are considered as new state variables, and it is assumed that the third derivative of the perturbation changes very slowly (can be considered zero) during the sampling period. Thus, the original second-order system is extended to the following fifth-order system:

[0088]

[0089] To eliminate the output signal in the observer equation Differential term and To avoid noise amplification, a set of intermediate variables is introduced. Perform the transformation. As an intermediate state variable of the observer, through rigorous mathematical derivation, it is obtained that depends only on the smoothed boost pressure signal. and control quantity The discrete-time observer implementation.

[0090]

[0091] , , They are respectively The intermediate state variables of the observer at time t. , , intermediate variables, for The boost pressure after smoothing at any moment.

[0092] The observer's state update equation is as follows:

[0093] in, , , They are respectively Intermediate state variables of the observer at each time step; The observer bandwidth is denoted by , and the core adjustable parameter ... This refers to the sampling period of the digital control system.

[0094] The observer bandwidth With the expected bandwidth of the closed-loop system The tuning relationship satisfies:

[0095] .

[0096] The output equation for the estimated values ​​of the disturbance and its derivative is as follows (in each control cycle). calculate):

[0097]

[0098] In the formula, , , They are respectively Estimates of the rate of change of boost pressure at time t, estimates of the first derivative of the rate of change of boost pressure, and estimates of the second derivative of the rate of change of boost pressure.

[0099] Step 4, based on , , And the estimated system parameters are used to calculate the target eTurbo speed at the next moment using an active disturbance rejection control law. This feedback is then sent to the speed control layer to form a closed loop, achieving boost pressure. Target boost pressure Tracking.

[0100] The boost pressure control layer employs an active disturbance rejection control algorithm based on a second-order system model, utilizing the estimated boost pressure change rate provided by ARO-ESO-2. Based on the estimated system parameters, an active disturbance rejection control law is designed to provide real-time, proactive feedforward compensation for the total disturbance, thereby dynamically decoupling the original complex and uncertain second-order system into a simple dual integrator system.

[0101] The active disturbance rejection control law takes the following form:

[0102]

[0103] For controller gain, For proportional gain, it can be configured according to classical control theory, for example, by setting the desired bandwidth of the closed-loop system. Damping ratio , The value range is 0.7-2, so we can take... , ; Let be the estimated value of the first derivative of the boost pressure at time t. Parameters in the dynamic model of boost pressure for electric superchargers Real-time estimates can be provided by offline calibration or online parameter learning algorithms (such as recursive least squares with a forgetting factor). For the output of ARO-ESO-2 The estimated rate of change of boost pressure at time t.

[0104] Mechanism of the control law: This control law is a composite structure integrating feedback, feedforward, and dynamic compensation. The first three terms in the numerator constitute the PD control and reference feedforward for the nominal dynamics of the system; the fourth term... Compensation for constant bias; Item 5 At its core, it directly and actively offsets the rate of change of boost pressure observed in real time by ARO-ESO-2. The entire molecule is subtracted from these compensation terms and then divided by the estimated control gain. This ultimately generates the target speed command. Theoretical analysis shows that when the disturbance estimation is accurate, the original system can be compensated to... ,in This provides a new input consisting of error feedback and feedforward, thereby enabling linearization and decoupled control of complex objects.

[0105] The method proposed in this invention designs a controller based on an accurate second-order physical model of the eTurbo system. It achieves fast, differential-free estimation of the total disturbance and its higher-order derivatives through an innovative ascending-descending extended state observer and algebraic filtering, and utilizes an active disturbance rejection control law for real-time dynamic compensation. Figure 2As shown, compared with the second-order ESO, the ARO-ESO-2 of this invention has the advantages of fast disturbance estimation, strong anti-interference ability and high control accuracy. Through simulation and bench verification, it can reduce the pressure overshoot from more than 4.5% to near zero compared with the traditional PID control, shorten the adjustment time by more than 17%, and significantly suppress steady-state fluctuations, demonstrating excellent comprehensive performance and engineering practicality.

[0106] Example 2

[0107] A control device for an electric supercharger includes a processor and a memory; the memory is used to store program code and transmit the program code to the processor.

[0108] The processor is used to execute the engine boost pressure control method based on nonlinear algebraic filtering and order rise and fall observation as described in Example 1, according to the instructions in the program code.

[0109] Example 3

[0110] A computer-readable storage medium storing computer-executable instructions, which, when executed, implement the engine boost pressure control method based on nonlinear algebraic filtering and order rise and fall observation as described in Embodiment 1.

[0111] Example 4

[0112] An electric supercharger includes an electric supercharger body and the control device described in Embodiment 2.

[0113] Example 5

[0114] An engine comprising the electric supercharger described in Example 4.

[0115] The above description is only a preferred embodiment of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for controlling engine boost pressure based on nonlinear algebraic filtering and order rise / fall observation, characterized in that, Includes the following steps: Step 1: The dynamic model of electric supercharger boost pressure is based on the current eTurbo target speed. Output boost pressure The model includes the rate of change of boost pressure. ; Step 2, the nonlinear algebraic filter module is based on the target boost pressure. Compared with the boost pressure measurement value The original tracking error for calculating boost pressure Then, normalization is performed to obtain the normalization error. Non-linear scaling function right Scaling and reconstruction are performed to obtain the smoothed tracking error. ,in accordance with and Calculate the smoothed boost pressure ; Step 3, the ascending and descending order extended state observer with As input, output the rate of change of boost pressure and its estimated first and second derivatives. , and ; Step 4, based on , , And the estimated system parameters are used to calculate the target eTurbo speed at the next moment using an active disturbance rejection control law. This feedback is then sent to the speed control layer to form a closed loop, achieving boost pressure. Target boost pressure Tracking.

2. The engine boost pressure control method based on nonlinear algebraic filtering and order rise / fall observation as described in claim 1, characterized in that, In step 1, the dynamic model of the boost pressure of the electric booster is as follows: ； Let be the boost pressure at time t. Let be the first derivative of the pressurization pressure at time t. Let be the second derivative of the pressurization pressure at time t. Let t be the target eTurbo rotational speed. These are the dynamic coefficients of the damping term and the dynamic coefficients of the restoring force term, respectively. To control the gain, For constant perturbation, Let be the rate of change of boost pressure at time t.

3. The engine boost pressure control method based on nonlinear algebraic filtering and order rise / fall observation as described in claim 1, characterized in that, In step 2, the original tracking error The calculation formula is: ； in, For discrete time series labeling, for The original tracking error at time step for The target boost pressure at any given time for The measured value of the boost pressure at any given moment; Normalized error The calculation formula is: ； in The preset dead zone threshold, , for Normalization error at time step; Preferred, , The standard deviation of noise is measured for pressure sensors.

4. The engine boost pressure control method based on nonlinear algebraic filtering and order rise / fall observation as described in claim 3, characterized in that, In step 2, the nonlinear scaling function The expression is: ； Smoothed tracking error The calculation formula is: ； for Smoothed tracking error at different times , for Normalization error at time step; Smoothed boost pressure .

5. The engine boost pressure control method based on nonlinear algebraic filtering and order rise / fall observation as described in claim 1, characterized in that, In step 3, the state update equation for the ascending-descending expansion state observer is: ； ； ； , , They are respectively The intermediate state variables of the observer at time t. , , They are respectively The intermediate state variables of the observer at time t. , , These are the coefficient matrices of the state update equation. , , These are intermediate variables, The sampling period of the digital control system. For the observer bandwidth, preferably, , The expected bandwidth of the closed-loop system; The output equation of the ascending-descending-level extended state observer is: ； , and for The rate of change of boost pressure at time t and the estimated values ​​of its first and second derivatives.

6. The engine boost pressure control method based on nonlinear algebraic filtering and order rise / fall observation as described in claim 1, characterized in that, In step 4, the active disturbance rejection control law takes the following form: ； for Smoothed tracking error at different times , For controller gain, For proportional gain, the preferred option is... , , The expected bandwidth of the closed-loop system. For the damping ratio, Let be the estimated value of the first derivative of the boost pressure at time t. These are the parameters in the dynamic model of the boost pressure of the electric supercharger. Real-time estimates, for The estimated rate of change of boost pressure at time t.

7. A control device for an electric booster, characterized in that, It includes a processor and a memory; the memory is used to store program code and transfer the program code to the processor; The processor is used to execute the engine boost pressure control method based on nonlinear algebraic filtering and order rise and fall observation as described in any one of claims 1 to 6 according to the instructions in the program code.

8. A computer-readable storage medium, characterized in that, The system stores computer-executable instructions, which, when executed, are used to implement the engine boost pressure control method based on nonlinear algebraic filtering and order rise and fall observation as described in any one of claims 1 to 6.

9. An electric booster, characterized in that, It includes an electric supercharger body and the control device as described in claim 7.

10. An engine, characterized in that, Includes the electric supercharger as described in claim 8.

Images