A high-pressure common rail fuel injection system injection law prediction method based on rail pressure fluctuation model

Abstract

The application discloses a kind of high-pressure common rail fuel injection system injection law prediction method based on rail pressure fluctuation model, belong to diesel engine fuel injection system technical field, the method of the present application includes the following steps:1) establish rail pressure fluctuation model;2) according to the rail pressure fluctuation model in step 1) establish the discretization state space model;3) rail pressure fluctuation optimal estimation based on Kalman filter;4) using rail pressure drop optimal estimation result to calculate injection law.The present application simultaneously considers the rail pressure drop process caused by injection, and the rail pressure rise process caused by oil supply, establishes rail pressure fluctuation model, which simultaneously considers the influence of injection and oil supply process on pressure fluctuation, can more comprehensively and accurately reflect the pressure fluctuation law, on this basis, designs rail pressure observer based on Kalman filtering, realizes the real-time observation calculation of high-pressure common rail system injection rate and injection volume using the observed rail pressure drop process.

Description

Technical Field

[0001] This invention belongs to the technical field of diesel engine fuel systems, specifically relating to a real-time prediction method for injection patterns applicable to high-pressure common rail fuel systems. Background Technology

[0002] Currently, the fuel injection quantity control of high-pressure common rail systems in diesel engines is based on an open-loop control mode using a calibrated MAP (Modular Mapping) diagram. Furthermore, real-time measurement of injection information is impossible during actual operation. Due to factors such as changes in system operating conditions and degradation of structural parameters, it is often difficult to guarantee the consistency and reliability of cyclic injection performance. If injection information could be monitored in real-time during actual diesel engine operation, allowing for closed-loop adjustment and correction of the injection pattern, the accuracy of injection control could be significantly improved.

[0003] The instantaneous pressure drop caused by fuel injection directly reflects the information of the injection process. Current research on fuel injection prediction based on fuel pressure signals mainly involves establishing dynamic mathematical models and numerically solving them to obtain the injection quantity. These methods are essentially open-loop calculations, and due to modeling errors, improper initial conditions, noise interference, and other factors, the calculated injection quantity results have significant errors. To address this limitation, the inventors have applied for a Chinese patent, "A Method for Predicting Injection Quantity of a High-Pressure Common Rail System Based on a Closed-Loop Observer," which introduces the concept of closed-loop feedback correction to achieve real-time observation and closed-loop correction of the cyclic injection quantity (Chinese Patent: ZL202010577723.3, which has been granted). Furthermore, when the rail pressure varies significantly, the model parameters change with the operating conditions, and the system exhibits significant nonlinearity. The inventors have optimized the observation process using a Kalman filter algorithm, which allows for continuous iterative and rolling optimization of the estimated values ​​of state variables and feedback gain (Chinese Patent: CN2022112886412). However, the above method only considers the impact of the injection process on pressure fluctuations, that is, it only models the situation where there is no fuel supply during the injection process. This method is suitable for situations where the injection process and the fuel supply process do not overlap. When the fuel supply process and the injection process overlap, the rail pressure drop caused by the injection will be affected by the fuel supply process, and the injection observation model established above is no longer applicable. Summary of the Invention

[0004] To address the aforementioned problems, this invention proposes a method for predicting the injection pattern of a high-pressure common rail fuel system based on an optimized rail pressure fluctuation model. This method considers both the rail pressure drop caused by injection and the rail pressure rise caused by fuel supply, establishing a rail pressure fluctuation model. This model simultaneously considers the influence of injection and fuel supply processes on pressure fluctuations, providing a more comprehensive and accurate reflection of pressure fluctuation patterns. Based on this model, a Kalman filter-based rail pressure observer is designed. Utilizing the observed rail pressure drop process, real-time observation and calculation of the injection rate and quantity of the high-pressure common rail system are achieved. Note: In this invention, (·) represents the first derivative, (··) represents the second derivative, and (^) represents the estimated value.

[0005] The technical solution adopted in this invention is as follows:

[0006] A method for predicting the injection pattern of a high-pressure common rail fuel system based on an optimized rail pressure fluctuation model, comprising the following steps:

[0007] 1) Establish a rail pressure fluctuation model;

[0008] 2) Establish a discretized state-space model based on the rail pressure fluctuation model in step 1);

[0009] 3) Optimal estimation of rail pressure fluctuation based on Kalman filter;

[0010] 4) Calculate the injection pattern using the optimal estimation result of rail pressure drop;

[0011] Step 1) establishes the rail pressure fluctuation model as follows:

[0012] The rail pressure change p is the difference between the steady-state value p and the rail pressure change p. ss The subsequent instantaneous rail pressure fluctuation, consisting of the superposition of the upward and downward processes, is expressed as:

[0013] p(t) = p down (t)+p up (t) (1)

[0014] Among them, the rail pressure descent phase model:

[0015]

[0016] proportionality coefficient K down (p ss )=K(p ss )·α(p ss ), its value is equal to the ratio of the rail pressure drop during the injection process to the corresponding injection pulse width; K is the proportionality coefficient, the value of which depends on the target rail pressure; τ down The time constant of the fuel injection rate output response;

[0017]

[0018] C loss This is the fuel loss coefficient, and its value is Q. inj With Q loss The ratio of the steady-state value p ss The target rail pressure under current operating conditions is given by V0, where V0 is the initial volume of the common rail pipe, and ΔV(p) is the initial volume of the common rail pipe. ss ) represents the common rail volume compensation amount, and the injection timing signal u inj (t) is the pulse width modulation signal

[0019]

[0020] The period T is the injection interval, and the injection start time t is the injection start time. start The time when the rail pressure begins to decrease is determined by the time when it ends. end Determined by the duration of fuel injection;

[0021] Rail pressure rise phase model:

[0022]

[0023] p up (t) represents the increase in rail pressure, τ up The inertial time constant of the output response during the rail pressure rise process, and the proportionality coefficient K. up (p ss This represents the ratio of the increase in rail pressure to the corresponding fuel supply.

[0024]

[0025] u pump (t) represents the step input for fuel supply, where the amplitude of the step input signal is the reference fuel supply for the current operating condition, and the step input is applied at time t. step It can be determined by the moment when the rail pressure begins to rise;

[0026] The discretized state-space model in step 2) is as follows:

[0027]

[0028] In the formula, k represents the number of sampling points. C d =C=[1 0 1]; Output y is the instantaneous rail pressure fluctuation p after subtracting the steady-state value; Input signal u=[u inj u pump ] T ; Includes rail pressure drop p down Rail pressure drop rate Rail pressure rise p up Three variables are used as state variables;

[0029] Step 3) The steps for optimal estimation of rail pressure fluctuation based on Kalman filter are as follows:

[0030] Considering model uncertainty w u Given the measurement noise v(k) and the discrete system state-space model, it can be represented as:

[0031]

[0032] Among them, the input process noise w u (k) is [w inj (k)w pump (k)] T ;w u v(k) and v(k) are assumed to be uncorrelated zero-mean Gaussian white noise, with their covariance matrices Q and R, respectively.

[0033] Because in model (19) B d The process noise covariance matrix Q changes with the rail pressure conditions. To adapt to these changes, the process noise covariance matrix Q can be optimized into a time-varying matrix.

[0034] Q(k) = B d (k)E[w u (k)w u (k) T B d (k) T (6)

[0035] Setting initial values and P(0) + After setting Q and R,

[0036] ① Calculate the prior estimate

[0037]

[0038] ② Calculate the covariance matrix P(k) of the prior estimation error. - :

[0039] P(k) - =A d (k-1)P(k-1) + A d (k-1) T +Q(k-1) (8)

[0040] ③According to P(k) - Calculate the Kalman feedback gain K(k):

[0041] K(k)=P(k) - C d (k) T [Cd (k)P(k) - C d (k) T +R(k)] -1 (9)

[0042] ④ At time k, the measured value y(k) and the prior estimated output are compared. The difference serves as feedback to correct the prior estimate, thus obtaining the posterior estimate.

[0043]

[0044] ⑤ Calculate the posterior covariance matrix P(k) + :

[0045] P(k) + =(IK(k)C d (k))P(k) - (IK(k)C d (k)) T +K(k)R(k)K(k) T (11)

[0046] By iterating through steps ① to ⑤, the system state, error covariance matrix, and Kalman filter gain are continuously updated, so that the posterior estimate approaches the true value, that is, the posterior error approaches 0, thus completing the optimal estimation of the state variables of the high-pressure common rail system.

[0047] Step 4) The steps for calculating the injection pattern using the optimal estimation result of rail pressure drop are as follows:

[0048] Using the estimated state variables of the high-pressure common rail system in step 3) Calculate the injection rate according to equation (4):

[0049]

[0050] In the formula, C loss This is the fuel loss coefficient; under a certain rail pressure p1, this coefficient is related to the fuel leakage rate V. leak Oil return volume V re and fuel injection quantity V inj related:

[0051]

[0052] Using experimental or simulation methods, measure V under a set rail pressure p1. leak V re and V inj The data is used to obtain C according to equation (14). loss (p1), and by changing the rail pressure, C is obtained under different set rail pressures.loss Applying least squares fitting to C over a wide range of derailment pressures loss (p ss );

[0053] Will The fuel injection quantity is obtained by summing the values ​​during the fuel injection phase.

[0054]

[0055] The advantages of this invention are:

[0056] 1. Based on the influence of the injection and fuel supply processes on rail pressure changes, transfer functions were established between injection and the rail pressure drop phase, and between fuel supply and the rail pressure rise phase. The rail pressure drop amount p was then selected. down Rail pressure drop rate Rail pressure rise p up Three variables are used to construct an observable state-space model, which is applicable to the observation of rail pressure fluctuations when there is overlap between fuel supply and injection.

[0057] 2. Taking into account model uncertainties and measurement noise, an optimal estimation method for fuel injection patterns based on the Kalman filter algorithm is proposed. Utilizing the recursive principle of the Kalman filter, the feedback gain matrix is ​​updated in real time, and the process noise covariance matrix Q is optimized to adapt to a wide range of rail pressure variations. This yields the optimal estimation result for the rail pressure drop process caused by fuel injection, thereby enabling accurate online real-time observation of fuel injection information. Attached Figure Description

[0058] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] Figure 1 Schematic diagram of a fuel injection pattern observation method based on rail pressure fluctuations

[0060] Figure 2 Schematic diagram of the Kalman filter process for rail pressure fluctuation

[0061] Figure 3 Fuel injection timing input signal

[0062] Figure 4 fuel supply step input signal

[0063] Figure 5 Measured rail pressure and observed pressure when fuel injection supply does not overlap

[0064] Figure 6 Measured and observed injection rates when fuel injection supply does not overlap

[0065] Figure 7 Measured and observed fuel injection quantity when fuel injection supply does not overlap

[0066] Figure 8 Measured rail pressure and observed pressure during fuel injection supply overlap

[0067] Figure 9 Measured and observed injection rates when fuel injection supply overlaps

[0068] Figure 10 Measured and observed fuel injection quantity when fuel injection supply overlaps Detailed Implementation

[0069] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other implementations obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0070] Figure 1 This invention provides a schematic diagram of a method for observing the injection pattern of a high-pressure common rail system based on optimal estimation of rail pressure fluctuations. The method involves measuring the rail pressure signal in real time using a pressure sensor on the common rail pipe; inputting this signal into a designed rail pressure Kalman filter to obtain the optimal estimation result for the rail pressure drop phase caused by injection; calculating the injection rate using the rail pressure drop estimation result; and integrating the injection rate during the injection phase to obtain the predicted injection quantity.

[0071] Figure 2 This diagram illustrates the design process of a Kalman filter for rail pressure fluctuations. First, based on the impact of the fuel injection and fuel supply processes on rail pressure fluctuations, dynamic models are established for the periods of fuel injection and rail pressure decrease, and fuel supply and rail pressure increase. State variables are selected to construct a state-space model of rail pressure fluctuations, which is then discretized. Based on this, considering model uncertainties and measurement noise, a Kalman filter for rail pressure fluctuations is designed. Using a recursive principle, the difference between the real-time measured rail pressure signal from the rail pressure sensor and the estimated value is used as feedback for correction, continuously updating the system state and covariance matrix to achieve optimal estimation of the rail pressure decrease process.

[0072] The following sections will provide detailed explanations.

[0073] Step 1: Establish a rail pressure fluctuation model

[0074] Assuming a uniform fuel pressure distribution within the common rail, the common rail fuel continuity equation can be expressed as:

[0075]

[0076] In the formula, p is the value minus the steady-state value p ss The subsequent instantaneous rail pressure fluctuation, steady-state value p ss The target rail pressure under current operating conditions; Q pump For fuel supply rate; Q inj Q represents the fuel injection rate. loss denoted as fuel loss rate; E is fuel bulk modulus; V is common rail control volume.

[0077] Ignoring fuel temperature changes during operation, E is related to rail pressure conditions, and the empirical formula is:

[0078] E = 1.2 × 10 4 (1+0.001p ss (2)

[0079] The common rail is subjected to high-pressure fuel, and V changes with the rail pressure. Let V consist of the initial volume V0 of the common rail and its compensation ΔV, that is:

[0080] V = V0 + △V(p) ss (3)

[0081] During fuel injection, the ejected fuel causes a momentary drop in pressure within the common rail; during fuel supply, the high-pressure fuel pump supplies high-pressure fuel to the common rail, causing a momentary rise in rail pressure. To decouple the effects of the injection and fuel supply processes on rail pressure fluctuations, the rail pressure p can be decomposed into a rail pressure decrease process p0. down With the rail pressure rise process p up The relationships between the injection process and rail pressure drop, and between the fuel supply process and rail pressure rise, were established respectively.

[0082] Step 1.1: Establish a model between the fuel injection process and the rail pressure drop phase.

[0083] The injection rate Q can be simplified from equation (1). inj Rail pressure drop p down Mathematical model between them:

[0084]

[0085] In the formula, C loss This is the fuel loss coefficient, and its value is Q. inj With Q loss The ratio. According to equations (2) to (4), α(p) ss The value of ) is related to the target rail pressure condition, and can be approximated as a constant under a certain rail pressure condition.

[0086] Clearly, there is an integral relationship between the injection rate and the rail pressure drop. Since the injection rate signal cannot be obtained in real time during actual operation, and the injection rate can be determined by the injection timing and injection pulse width, this invention uses the injection timing signal u. inj (t) is the input to establish a model for the rail pressure drop phase.

[0087] Fuel injection timing signal u inj (t) is a pulse width modulation signal, where the period T is the injection interval, and the duty cycle is the ratio of the injection duration to the injection interval. The injection start time t start The time when the rail pressure begins to decrease is determined by the time when it ends. end Determined by the injection duration. This signal is as follows: Figure 3 As shown, its expression can be described as

[0088]

[0089] Considering the fuel injection timing signal u inj (t) and injection rate Q inj There is a certain inertia between (t), and the dynamic model between the two can be expressed as:

[0090]

[0091] In the formula, K is the proportionality coefficient, the value of which depends on the target rail pressure; τ down Let τ be the time constant of the fuel injection rate output response. Since the response at the beginning and end of the fuel injection rate phases is extremely fast, the inertial time constant is similar under different operating conditions. down It can be considered a constant.

[0092] Substituting equation (5) into equation (6), we obtain u inj (t) is the input, p down (t) represents the output rail pressure drop stage model:

[0093]

[0094] In the formula, the proportionality coefficient K down (p ss )=K(p ss )·α(p ss Its value is equal to the ratio of the rail pressure drop during the injection process to the corresponding injection pulse width.

[0095] Step 1.2: Establish a model between the oil supply process and the rail pressure rise stage.

[0096] The oil supply rate Q can be obtained from equation (1). pump With the rise of rail pressure p up Mathematical model between them:

[0097]

[0098] However, the fuel supply rate cannot be measured in real time, and due to the unclear duration of the fuel supply phase, a timing signal cannot be constructed. The reference fuel supply quantity V for the current operating condition can only be obtained by referring to a pre-calibrated MAP of fuel supply quantity under various operating conditions (different speeds, rail pressures, and fuel metering valve openings). pump The relationship between fuel supply quantity and fuel supply rate is integral. Therefore, this invention constructs a fuel supply quantity step signal u. pump (t) serves as the input during the rail pressure rise phase. The amplitude of this step input signal is the current operating condition reference fuel supply, and the time t at which this step input is applied is... step It can be determined based on the timing of the rail pressure rise, such as Figure 4 As shown.

[0099] Considering the inertia during the pressure rise process, the rail pressure rise p up With step input u of fuel supply pump The dynamic model between (t) is represented as:

[0100]

[0101] In the formula, the proportionality coefficient K up τ is the ratio of rail pressure increase to corresponding fuel supply; up It is the inertial time constant of the output response during the rail pressure rise process, and its value can be regarded as a constant under different operating conditions.

[0102] Although the models for the injection and fuel supply processes are established separately, they are equally applicable when they occur simultaneously. The rail pressure change p is a superposition of the rising and falling processes, expressed as:

[0103] p(t) = p down (t)+p up (t) (10)

[0104] Step 1.3: Identify model coefficients

[0105] The rail pressure drop phase model includes two undetermined coefficients: K down With τ down Among them, K down This is the ratio of the rail pressure drop to the corresponding injection pulse width for a single injection under the current operating condition, and it is related to the rail pressure condition. This invention focuses on a high-pressure common rail system with four injectors, a dual-plunger, double-acting camshaft-driven high-pressure pump. Taking a rail pressure of 1200 bar and a camshaft speed of 1000 r / min as an example, the rail pressure drop for different pulse widths is obtained, and K is calculated. down As shown in Table 1.

[0106] Table 1 Rail pressure drop parameters and K down

[0107]

[0108]

[0109] K under different fuel injection pulse widths down Similar, therefore, under the current rail pressure condition, K down The average value is taken as -44768. Repeating this step when setting rail pressure variations yields model coefficients for a wide range of rail pressure changes, adapting to calculations under different operating conditions. This system fits K under different operating conditions. down The expression is:

[0110] K down (p ss )=-7769-29.51·p ss (11)

[0111] Based on the characteristics of a first-order system, the time constant τ down τ represents the time it takes for the injection rate to reach 0.632 times its steady-state value from 0. Under different operating conditions, the injection rate curve shows extremely rapid rise and fall phases. down Take a constant value of 0.00015.

[0112] The model for the rail pressure rise phase also includes two undetermined coefficients: K up With τ up According to equation (8), K up The expression can be written as:

[0113]

[0114] In the formula, based on the common rail pipe structural parameters, V0 = 29061 mm 3 Compensation amount △V(p) ss )=5.048·p ss .

[0115] Under step input conditions, the time constant τ up Let τ be the time it takes for the output response to rise from 0 to 0.632 times the steady-state amplitude under different operating conditions. up The average value is 0.0019.

[0116] Step 2: Construct the rail pressure state-space model and discretize it.

[0117] A state-space model of rail pressure fluctuation is established based on the transfer function model described above. The rail pressure drop amount p is selected. down Rail pressure drop rate p down Rail pressure rise p up The three variables are used as state variables, that is

[0118]

[0119] Based on equations (7), (9), and (10), the state-space model is obtained as follows:

[0120]

[0121] In the formula, C = [1 0 1]. Output y is the instantaneous rail pressure change p after subtracting the steady-state value; input signal u = [u inj u pump ] T .

[0122] Determine the observability of the system. The observability matrix Lo of model (13) is calculated as follows:

[0123]

[0124] The Lo rank is full, indicating that the system is considerable and suitable for designing a rail-voltage drop Kalman filter.

[0125] Before designing the Kalman filter observer, the state-space model of the continuous system needs to be discretized. With a sampling step size of Δt, the state variable x(t) at time t... k The derivative at time t can be approximated as:

[0126]

[0127] The above formula can be transformed into:

[0128]

[0129] In the formula,

[0130] The sampling time t in equation (16) k The number of sampling points k is used to represent the total number of sampling points, i.e.:

[0131] x(k)=A d x(k-1)+B d u(k-1) (17)

[0132] The discrete state-space model can then be expressed as:

[0133]

[0134] In the formula, C d =C=

[101] .

[0135] Step 3: Optimal estimation of rail voltage drop based on Kalman filter

[0136] Considering model uncertainty w uGiven the measurement noise v(k) and the system state-space model, it can be represented as:

[0137]

[0138] Among them, the input process noise w u (k) is [w inj (k)w pump (k)] T w u v(k) and v(k) are assumed to be uncorrelated zero-mean Gaussian white noise, with their covariance matrices Q and R, respectively.

[0139] According to the Kalman filter algorithm, the estimated state variable at time k is defined as... It is further divided into prior estimates. Posterior estimate The algorithm consists of two stages: time update and measurement update. In the time update stage, the system model is used to calculate the prior estimate of the state. In the measurement update stage, the error between the rail pressure measurement value and the prior estimate is used for feedback correction, and the posterior estimate is calculated.

[0140] The performance of a Kalman filter is determined by the noise covariance matrices Q and R. The measurement noise covariance matrix R depends on the degree of filtering of the measurement signal. Since B in model (19) d The process noise covariance matrix Q changes with the rail pressure conditions. To adapt to these changes, the process noise covariance matrix Q can be optimized into a time-varying matrix.

[0141] Q(k) = B d (k)E[w u (k)w u (k) T B d (k) T (20)

[0142] Setting initial values and P(0) + After selecting appropriate Q and R, the system state, error covariance matrix and Kalman filter gain can be continuously updated according to equations (21) to (25) to make the posterior estimate approach the true value, that is, the posterior error approaches 0, and the optimal estimation of the state variables of the high-pressure common rail system can be completed.

[0143] Time update phase:

[0144] ① Calculate the prior estimate using model (18)

[0145]

[0146] ② Calculate the covariance matrix P(k) of the prior estimation error. - :

[0147] P(k) - =A d (k-1)P(k-1) + A d (k-1) T +Q(k-1) (22)

[0148] ③According to P(k) - Calculate the Kalman feedback gain K(k):

[0149] K(k)=P(k) - C d (k) T [C d (k)P(k) - C d (k) T +R(k)] -1 (twenty three)

[0150] ④ At time k, the measured value y(k) and the prior estimated output are compared. The difference serves as feedback to correct the prior estimate, thus obtaining the posterior estimate.

[0151]

[0152] ⑤ Calculate the posterior covariance matrix P(k) + :

[0153] P(k) + =(IK(k)C d (k))P(k) - (IK(k)C d (k)) T +K(k)R(k)K(k) T (25)

[0154] At time k, the posterior estimate is obtained according to equation (24). In and This is the optimal estimate of the rail pressure drop and the rail pressure drop rate, output as follows: This is the result of rail voltage filtering.

[0155] Step 4: Calculate the injection pattern using the optimal estimation result of rail pressure drop.

[0156] get Then, the injection rate can be calculated according to equation (4):

[0157]

[0158] In the formula, C loss This is the fuel loss coefficient. Under a certain set rail pressure p1, this coefficient is related to the fuel leakage rate V. leak Oil return volume V re and fuel injection quantity V inj related:

[0159]

[0160] Using experimental or simulation methods, measure V under a set rail pressure p1. leak V re and V inj The data is used to obtain C according to equation (27). loss (p1), and by changing the rail pressure, C is obtained under different set rail pressures. loss Applying least squares fitting to C over a wide range of derailment pressures loss (p ss In this example, the fitted C loss (p ss )for:

[0161] C loss (p ss ) = 0.2115 + 5.85 × 10 -6 ·p ss (28)

[0162] Will The fuel injection quantity is obtained by summing the values ​​during the fuel injection phase.

[0163]

[0164] To verify the filtering effect and observation accuracy of the proposed observation method under different operating conditions, a simulation study was conducted using the method proposed in this invention under the operating conditions of 1200 bar rail pressure and 1.2 ms injection pulse width. Figures 5 to 7 The observation results of common rail pressure, injection rate, and injection quantity under the condition that the fuel injection supply does not overlap. Figures 8 to 10 The observed results for rail pressure, injection rate, and injection quantity are shown when the ratio of injection times to fuel supply times per cycle is 6:4. It can be seen that rapid tracking of rail pressure and injection rate can be achieved under both operating conditions. The actual values ​​of the single injection quantity are compared, and the errors are shown in Table 2 below.

[0165] Table 2. Observation error of fuel injection quantity under different operating conditions

[0166] Operating conditions Maximum error Minimum error average error Fuel injection and fuel supply do not overlap 4.62％ 1.06％ 2.62％ Fuel injection supply overlap 4.86％ 0.20％ 2.55％ .

Claims

1. A method for predicting the injection pattern of a high-pressure common rail fuel system based on a rail pressure fluctuation model, characterized in that... The method includes the following steps: 1) Simultaneously consider the rail pressure drop process caused by fuel injection and the rail pressure rise process caused by fuel supply, and establish a rail pressure fluctuation model. This model simultaneously considers the influence of fuel injection and fuel supply processes on pressure fluctuation. 2) Establish a discretized state-space model based on the rail pressure fluctuation model in step 1); 3) Optimal estimation of rail pressure fluctuation based on Kalman filter; 4) Calculate the injection pattern using the optimal estimation result of rail pressure drop; Step 1) establishes the rail pressure fluctuation model as follows: The rail pressure change p is the difference between the steady-state value p and the rail pressure change p. ss The subsequent instantaneous rail pressure fluctuation, consisting of the superposition of the upward and downward processes, is expressed as: （1） Among them, the rail pressure descent phase model: （2） proportionality coefficient Its value is equal to the ratio of the rail pressure drop during the injection process to the corresponding injection pulse width; K is a proportionality coefficient, the value of which depends on the target rail pressure; τ down The time constant of the fuel injection rate output response; C loss This is the fuel loss coefficient, and its value is Q. inj With Q loss The ratio of the steady-state value p ss The target rail pressure under current operating conditions is given by V0, where V0 is the initial volume of the common rail pipe, and ΔV(p) is the initial volume of the common rail pipe. ss ) represents the common rail volume compensation amount, and the injection timing signal u inj (t) is the pulse width modulation signal The period T is the injection interval, and the injection start time t is the injection start time. start The time when the rail pressure begins to decrease is determined by the time when it ends. end Determined by the duration of fuel injection; Rail pressure rise phase model: （3） p up (t) represents the increase in rail pressure, τ up The inertial time constant of the output response during the rail pressure rise process, and the proportionality coefficient K. up (p ss This represents the ratio of the increase in rail pressure to the corresponding fuel supply. u pump (t) represents the step input for fuel supply, where the amplitude of the step input signal is the reference fuel supply for the current operating condition, and the step input is applied at time t. step It can be determined by the moment when the rail pressure begins to rise; The discretized state-space model in step 2) is as follows: （4） In the formula, k represents the number of sampling points. , , Output y is the instantaneous rail pressure fluctuation p after subtracting the steady-state value; input signal ; Includes rail pressure drop p down Rail pressure drop rate Rail pressure rise p up Three variables are used as state variables; Step 3) The steps for optimal estimation of rail pressure fluctuation based on Kalman filter are as follows: Considering model uncertainty w u Given the measurement noise v(k) and the discrete system state-space model, it can be represented as: （5） Among them, the input process noise w u (k) is [w inj (k) w pump (k)] T ;w u v(k) and v(k) are assumed to be uncorrelated zero-mean Gaussian white noise, with their covariance matrices Q and R, respectively. Because in model (19) B d The process noise covariance matrix Q changes with the rail pressure conditions. To adapt to these changes, the process noise covariance matrix Q can be optimized into a time-varying matrix. （6） Setting initial values and After setting Q and R, ① Calculate the prior estimate : （7） ② Calculate the covariance matrix of the prior estimation error. : （8） ③According to Calculate Kalman feedback gain : （9） ④ At time k, the measured value y(k) and the prior estimated output are compared. The difference serves as feedback to correct the prior estimate, thus obtaining the posterior estimate. （10） ⑤ Calculate the posterior covariance matrix : （11） By iterating through steps ① to ⑤, the system state, error covariance matrix, and Kalman filter gain are continuously updated, so that the posterior estimate approaches the true value, that is, the posterior error approaches 0, thus completing the optimal estimation of the state variables of the high-pressure common rail system. Step 4) The steps for calculating the injection pattern using the optimal estimation result of rail pressure drop are as follows: Using the estimated state variables of the high-pressure common rail system in step 3) The injection rate is calculated according to equation (4): （12） In the formula, C loss This is the fuel loss coefficient; under a certain rail pressure p1, this coefficient is related to the fuel leakage rate V. leak Oil return volume V re and fuel injection quantity V inj related: （13） Using experimental or simulation methods, measure V under a set rail pressure p1. leak V re and V inj The data is used to obtain C according to equation (13). loss (p1), and by changing the rail pressure, C is obtained under different set rail pressures. loss Applying least squares fitting to C over a wide range of derailment pressures loss (p ss ); Will The fuel injection quantity is obtained by summing the values ​​during the fuel injection phase. : （14）。

Images