A precise control method for injection quantity of static leakage ejector based on physical information neural network

Abstract

The purpose of the present invention is to provide a method for precise control of the injection quantity of a static leak-free injector based on a physical information neural network, which belongs to the field of engines. First, a Transformer time series model is constructed to virtually predict the pressure fluctuations at the injector inlet, replacing the pressure sensor at the injector inlet. A high-pressure fuel pipe flow model is established, and the flow is calculated by measuring and predicting the pressure fluctuations at two locations and integrating the total amount of fuel. Then, a PINN model is constructed and trained based on the total amount of fuel flowing out of the high-pressure fuel pipe and the experimentally measured injection quantity data to achieve precise prediction of the injection quantity. Finally, the injection pulse width is corrected in real time by a segmented PID controller to achieve closed-loop control of the injection quantity. The present invention only requires the installation of one pressure sensor in the high-pressure fuel pipe, and uses virtual prediction to avoid the complexity and cost of multiple sensors. By combining physical constraints with data-driven, the injection quantity prediction accuracy and closed-loop control effect under complex working conditions are guaranteed.

Description

Technical Field

[0001] The present invention relates to a method for controlling an injection device, in particular to a method for controlling an injection quantity. Background Art

[0002] The fuel injection system is a crucial component in improving engine performance and reducing emissions. As a widely used technology, the injector, a core component of high-pressure fuel injection systems, directly impacts engine combustion efficiency and emissions through injection volume control. Precisely controlling the injection volume of injectors is of paramount importance, especially given the current push for energy conservation and emissions reduction policies.

[0003] Traditional fuel injection system control methods mostly rely on open-loop control or closed-loop control based on feedback variables such as engine speed. However, static leakage from the needle valve assembly introduces an uncontrolled variable, affecting the system's precise control of injection quantity. Therefore, most current injectors utilize internal structures with no static leakage. Despite the advantage of reduced leakage, injection quantity control still relies on traditional control algorithms or empirical models, which often fail to fully utilize the physical information in the injector. Furthermore, while purely data-driven methods can learn patterns from data, they struggle to interpret and ensure physical consistency, easily leading to overfitting and model instability. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for accurately controlling the injection amount of a static leak-free injector based on a physical information neural network, which can solve the problems of being unable to accurately predict the injection amount of the injector in real time online and accurately controlling the injection amount in a closed loop.

[0005] The object of the present invention is achieved like this:

[0006] The present invention provides a method for accurately controlling the injection amount of a static leak-free injector based on a physical information neural network, which is characterized by comprising the following steps:

[0007] (1) Install a clamp-on pressure sensor at the injector inlet and an embedded pressure sensor at the center of the high-pressure fuel pipe between the common rail and the injector. Measure the fuel pressure fluctuations of the clamp-on pressure sensor and the embedded pressure sensor under different injection conditions. After obtaining the pressure fluctuation data set, remove the clamp-on pressure sensor.

[0008] (2) A Transformer time series model is constructed with the pressure fluctuation at the center of the high-pressure fuel pipe as input and the pressure fluctuation at the injector inlet as output. The model is trained based on the pressure fluctuation dataset and performs virtual prediction of the pressure fluctuation during the actual operation of the high-pressure common rail system.

[0009] (3) A high-pressure fuel pipe flow model is established based on the transmission relationship between the center position of the high-pressure fuel pipe and the injector inlet pressure fluctuation. The pressure signal actually measured at the center position of the high-pressure fuel pipe and the injector inlet pressure fluctuation predicted in real time by the Transformer time series model are used as the input of the high-pressure fuel pipe flow model. The flow rate of the high-pressure fuel pipe is calculated, and the amount of fuel flowing out of the high-pressure fuel pipe is obtained by integration. This amount is taken as the sum of the injector injection amount and the return amount, and a data set of the amount of fuel flowing out of the high-pressure fuel pipe under different injection conditions is established.

[0010] (4) Based on the ratio of the total area of ​​the injection hole of the injector without static leakage to the area of ​​the OA hole, the proportion of the injection amount and the return amount under different injection working conditions is calculated to form a database; a physical information neural network PINN model is constructed with the fuel outflow amount of the high-pressure fuel pipe as input and the injection amount as output. The model adopts a multi-layer feedforward neural network as the framework, and the sum of the physical constraint loss and the regression loss is used as the total loss function of the network model. The data set of the fuel outflow amount of the high-pressure fuel pipe and the injection amount data under different injection working conditions measured by experiments are used for model training;

[0011] (5) Based on the deviation between the actual injection amount predicted in the current cycle and the target injection amount, the injection pulse width of the next cycle is corrected through the segmented PID to achieve closed-loop control of the injection amount.

[0012] The present invention may also include:

[0013] 1. The steps for building the Transformer time series model in step (2) are:

[0014] (21) Divide the data into time segments of the same length, and each time segment is a sample; let the input time series P pipe =[P pipe,1 ,P pipe,2 ,...,P pipe,T ], output time series P injector =[P injector,1 ,P injector,2 ,...,P injector,T ], where T represents the time step number, and the pressure fluctuation data is normalized;

[0015] (22) Create position encoding to make the model aware of the time step information. The position encoding vector PE is calculated as:

[0016]

[0017] Where i is the dimension index, d model is the embedding dimension;

[0018] (23) Establish the Transformer encoder layer, specifically:

[0019] Capturing the long-term temporal dependencies of input time series data through multi-head self-attention layers:

[0020]

[0021] Where Q, K, and V are the query matrix, key matrix, and value matrix corresponding to the input sequence, respectively. k It is the attention dimension;

[0022] A feedforward neural network is built after each attention layer, which is expressed as:

[0023] FFN(x)=max(0,XW1+b1)W2+b2

[0024] Where W and b are the weight and bias of the feedforward neural network, respectively. Subscripts 1 and 2 represent the first and second layers of the feedforward network, respectively.

[0025] On this basis, a residual connection is added after each layer, and the outputs of the self-attention layer and the feedforward neural network layer are expressed as:

[0026] Output=LayerNorm(X+Attention)

[0027] Output = LayerNorm(X+FFN)

[0028] (24) Construct the loss function of the model, specifically:

[0029]

[0030] Where, Characterize the input time series and predict the output injector inlet pressure fluctuation through the Transformer model; i is the actual value of the injector inlet pressure fluctuation. Finally, the model training is performed based on the pressure fluctuation dataset.

[0031] 2. The specific steps for constructing the high-pressure fuel pipe flow model in step (3) are as follows:

[0032] Establish the momentum balance equation in the high-pressure fuel pipe:

[0033]

[0034] Where ρ is the fuel density, u is the flow rate, P is the pressure, D is the pipe diameter, and γ is the wall shear stress;

[0035] The mass flow rate out of the high-pressure fuel pipe is obtained by integrating the distance L between the center of the high-pressure fuel pipe and the injector inlet and multiplying it by the pipe cross-sectional area A:

[0036]

[0037] Where ξ is a constant related to the speed of sound, and t is time;

[0038] The fuel mass of the high-pressure fuel pipe is expressed as:

[0039] M pipe-out =∫Q pipe-out dt.

[0040] 3. The loss function of the PINN model in step (4) is expressed as:

[0041] The physical constraint loss E1 is expressed as:

[0042]

[0043] Where, represents the injection amount predicted by the PINN model, κ is the ratio of the injection amount and the backflow amount recorded in the database;

[0044] The regression loss constraint E2 is expressed as:

[0045] The total loss function of the model is: E = λ1E1 + E2, where λ1 is the weight coefficient of the physical constraint loss.

[0046] 4. The specific step of the segmented PID control in step (5) is:

[0047] According to the deviation between the actual injection amount and the target injection amount, different PID control parameters are applied in different stages according to the deviation degree, as follows:

[0048]

[0049] Where ΔPW(k) is the injection pulse width correction value at the kth injection, e(k-1) is the injection quantity deviation at the k-1th injection, σ1, σ2, and σ3 are the respective deviation thresholds of different injection quantity deviation segments, is the correction coefficient of the first deviation segment, is the correction coefficient of the second deviation segment, is the correction coefficient of the third deviation segment;

[0050] The segmented PID outputs the control signal for the kth injection based on the calculated injection pulse width correction value:

[0051] PW(k)=PW(k-1)+ΔPW(k).

[0052] The advantages of the present invention are:

[0053] (1) The present invention organically combines physical models and data-driven methods, avoiding the error accumulation caused by traditional methods relying on empirical adjustments;

[0054] (2) The present invention only requires one pressure sensor to be installed on the high-pressure fuel pipe to virtually predict the pressure fluctuation at the injector inlet, effectively avoiding the need to add additional sensors in actual working conditions, thereby reducing complexity and cost;

[0055] (3) The present invention performs segmented PID injection pulse width correction according to the injection quantity deviation threshold, ensuring that the system can be adjusted quickly and enhancing the stability of injection quantity control. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 is a flow chart of the present invention;

[0057] Figure 2 This is a connection diagram of the test device. DETAILED DESCRIPTION

[0058] The present invention will be described in more detail below with reference to the accompanying drawings:

[0059] Combine Figure 1-2 The present invention provides a method for accurately controlling the injection amount of a static leak-free injector based on a physical information neural network, comprising the following steps:

[0060] Step S1, such as Figure 2 As shown, a clamp-type pressure sensor is installed at the injector inlet, and an embedded pressure sensor is installed at the center of the high-pressure fuel pipe between the common rail pipe and the injector. The fuel pressure fluctuations at the above two locations are measured under different injection conditions. After obtaining the pressure fluctuation data set, the clamp-type pressure sensor is removed.

[0061] In step S2, a Transformer time series model is constructed, using the pressure fluctuation at the center of the high-pressure fuel pipe as input and the pressure fluctuation at the injector inlet as output. The model is trained based on the pressure fluctuation dataset. The Transformer time series model is used to virtually predict the pressure fluctuation during actual operation of the high-pressure common rail system, replacing the clamp-on pressure sensor.

[0062] In step S3, a high-pressure fuel line flow model is established based on the transmission relationship between the center position of the high-pressure fuel line and the injector inlet pressure fluctuation according to the Euler partial differential equation of generalized mass and momentum. The pressure signal actually measured at the center position of the high-pressure fuel line and the injector inlet pressure fluctuation predicted in real time by the Transformer time series model are used as inputs of the high-pressure fuel line flow model. The flow rate of the high-pressure fuel line is calculated, and the outflowing fuel volume of the high-pressure fuel line is obtained by integration. This is used as the sum of the injector injection volume and the return volume, and a data set of the outflowing fuel volume of the high-pressure fuel line under different injection operating conditions is established.

[0063] In step S4, based on the ratio of the total nozzle area of ​​the static-leak-free injector to the OA hole area, the ratio of the injection volume to the return volume under different injection conditions is calculated to form a database. A PINN model is constructed, using the fuel flow rate out of the high-pressure fuel pipe as input and the injection volume as output. This model uses a multi-layer feedforward neural network framework and uses the sum of physical constraint loss and regression loss as the total loss function of the network model. The model is trained using a dataset of the fuel flow rate out of the high-pressure fuel pipe and experimentally measured injection volume data under different injection conditions. The resulting trained PINN model is capable of accurately predicting the injection volume.

[0064] Step S5: During the actual working process, based on the deviation between the actual injection amount predicted in the current cycle and the target injection amount, the injection pulse width of the next cycle is corrected through the segmented PID to achieve closed-loop control of the injection amount.

[0065] The specific steps for building the Transformer time series model in step S2 are:

[0066] Step (a) divides the data into time segments of the same length, and each time segment is a sample. Let the input time series P be pipe =[P pipe,1 ,P pipe,2 ,...,P pipe,T ], output time series P injector =[P injector,1 ,P injector,2 ,...,P injector,T ], where T represents the number of time steps. The pressure fluctuation data is normalized to improve the stability of the model.

[0067] Step (b) creates a positional encoding to enable the model to perceive the information of the time step. The positional encoding vector PE is calculated as:

[0068]

[0069] Where i is the dimension index, d model is the embedding dimension.

[0070] Step (c) builds the Transformer encoder layer, specifically:

[0071] Capturing the long-term temporal dependencies of input time series data through multi-head self-attention layers:

[0072]

[0073] Where Q, K, and V are the query matrix, key matrix, and value matrix corresponding to the input sequence, respectively. k It is the attention dimension.

[0074] A feedforward neural network is established after each attention layer to improve the nonlinear representation ability of the model, which is expressed as:

[0075] FFN(x)=max(0,XW1+b1)W2+b2 (3)

[0076] Where W and b are the weight and bias of the feedforward neural network, respectively. The subscripts 1 and 2 represent the first and second layers of the feedforward network, respectively.

[0077] On this basis, a residual connection is added after each layer to prevent the gradient from disappearing and improve the training stability. The outputs of the self-attention layer and the feedforward neural network layer are expressed as:

[0078] Output=LayerNorm(X+Attention) (4)

[0079] Output=LayerNorm(X+FFN) (5)

[0080] Step (d) constructs the model's loss function to evaluate the model's prediction error, specifically:

[0081]

[0082] Where, Characterize the input time series and predict the output injector inlet pressure fluctuation through the Transformer model; i is the actual value of the injector inlet pressure fluctuation. Finally, the model is trained based on the pressure fluctuation dataset.

[0083] The high-pressure fuel pipe flow model is constructed in step S3 as follows:

[0084] Establish the momentum balance equation in the high-pressure fuel pipe:

[0085]

[0086] Where ρ is the fuel density, u is the flow rate, P is the pressure, D is the pipe diameter, and γ is the wall shear stress.

[0087] The mass flow rate out of the high-pressure fuel pipe is obtained by integrating the distance L between the center of the high-pressure fuel pipe and the injector inlet and multiplying it by the pipe cross-sectional area A:

[0088]

[0089] Where ξ is a constant related to the speed of sound, and t is time.

[0090] The fuel mass of the high-pressure fuel pipe is expressed as:

[0091] M pipe-out =∫Q pipe-out dt (9)

[0092] The loss function of the PINN model in step S4 is expressed as:

[0093] The physical constraint loss E1 is expressed as:

[0094]

[0095] Where, represents the injection amount predicted by the PINN model, and κ is the ratio of the injection amount and the recirculation amount recorded in the database.

[0096] The regression loss constraint E2 is expressed as:

[0097] In summary, the total loss function of the model is: E = λ1E1 + E2 (12)

[0098] Where λ1 is the weight coefficient of the physical constraint loss.

[0099] The specific step of the segmented PID control in step S5 is:

[0100] According to the deviation between the actual injection amount and the target injection amount, different PID control parameters are applied in different stages according to the deviation degree. The specific formula is as follows:

[0101]

[0102] Where ΔPW(k) is the injection pulse width correction value at the kth injection, e(k-1) is the injection quantity deviation at the k-1th injection, σ1, σ2, and σ3 are the respective deviation thresholds of different injection quantity deviation segments, is the correction coefficient of the first deviation segment, is the correction coefficient of the second deviation segment, is the correction coefficient for the third deviation segment.

[0103] The segmented PID outputs the control signal for the kth injection based on the calculated injection pulse width correction value:

[0104] PW(k)=PW(k-1)+ΔPW(k) (14).

Claims

1. A method for precise control of the injection quantity of a static leak-free injector based on a physical information neural network, characterized by: The following steps are involved: (1) Install a clamp-on pressure sensor at the injector inlet and an embedded pressure sensor at the center of the high-pressure fuel pipe between the common rail and the injector. Measure the fuel pressure fluctuations of the clamp-on pressure sensor and the embedded pressure sensor under different injection conditions. After obtaining the pressure fluctuation data set, remove the clamp-on pressure sensor. (2) A Transformer time series model is constructed with the pressure fluctuation at the center of the high-pressure fuel pipe as input and the pressure fluctuation at the injector inlet as output. The model is trained based on the pressure fluctuation dataset and performs virtual prediction of the pressure fluctuation during the actual operation of the high-pressure common rail system. (3) A high-pressure fuel pipe flow model is established based on the transmission relationship between the center position of the high-pressure fuel pipe and the injector inlet pressure fluctuation. The pressure signal actually measured at the center position of the high-pressure fuel pipe and the injector inlet pressure fluctuation predicted in real time by the Transformer time series model are used as the input of the high-pressure fuel pipe flow model. The flow rate of the high-pressure fuel pipe is calculated, and the amount of fuel flowing out of the high-pressure fuel pipe is obtained by integration. This amount is taken as the sum of the injector injection amount and the return amount, and a data set of the amount of fuel flowing out of the high-pressure fuel pipe under different injection conditions is established. (4) Based on the ratio of the total area of ​​the injection hole of the injector without static leakage to the area of ​​the OA hole, the proportion of the injection amount and the return amount under different injection working conditions is calculated to form a database; a physical information neural network PINN model is constructed with the fuel outflow amount of the high-pressure fuel pipe as input and the injection amount as output. The model adopts a multi-layer feedforward neural network as the framework, and the sum of the physical constraint loss and the regression loss is used as the total loss function of the network model. The data set of the fuel outflow amount of the high-pressure fuel pipe and the injection amount data under different injection working conditions measured by experiments are used for model training; (5) Based on the deviation between the actual injection amount predicted in the current cycle and the target injection amount, the injection pulse width of the next cycle is corrected through the segmented PID to achieve closed-loop control of the injection amount.

2. The method for precise control of injection quantity of a static leak-free injector based on a physical information neural network according to claim 1, characterized in that: The steps for building the Transformer time series model in step (2) are: (21) Divide the data into time segments of the same length, and each time segment is a sample; let the input time series P pipe =[P pipe,1 ,P pipe,2 ,...,P pipe,T ], output time series P injector =[P injector,1 ,P injector,2 ,…,P injector,T ], where T represents the time step number, and the pressure fluctuation data is normalized; (22) Create position encoding to make the model aware of the time step information. The position encoding vector PE is calculated as: Where i is the dimension index, d model is the embedding dimension; (23) Establish the Transformer encoder layer, specifically: Capturing the long-term temporal dependencies of input time series data through multi-head self-attention layers: Where Q, K, and V are the query matrix, key matrix, and value matrix corresponding to the input sequence, respectively. k It is the attention dimension; A feedforward neural network is built after each attention layer, which is expressed as: FFN(x)=max(0,XW1+b1)W2+b2 Where W and b are the weight and bias of the feedforward neural network, respectively. Subscripts 1 and 2 represent the first and second layers of the feedforward network, respectively. On this basis, a residual connection is added after each layer, and the outputs of the self-attention layer and the feedforward neural network layer are expressed as: Output=LayerNorm(X+Attention) Output = LayerNorm(X+FFN) (24) Construct the loss function of the model, specifically: Where, Characterize the input time series and predict the output injector inlet pressure fluctuation through the Transformer model; i is the actual value of the injector inlet pressure fluctuation. Finally, the model training is performed based on the pressure fluctuation dataset.

3. The method for precise control of injection quantity of a static leak-free injector based on a physical information neural network according to claim 1, characterized in that: The specific steps for constructing the high-pressure fuel pipe flow model in step (3) are as follows: Establish the momentum balance equation in the high-pressure fuel pipe: Where ρ is the fuel density, u is the flow rate, P is the pressure, D is the pipe diameter, and γ is the wall shear stress; The mass flow rate out of the high-pressure fuel pipe is obtained by integrating the distance L between the center of the high-pressure fuel pipe and the injector inlet and multiplying it by the pipe cross-sectional area A: Where ξ is a constant related to the speed of sound, and t is time; The fuel mass of the high-pressure fuel pipe is expressed as: M pipe-out =∫Q pipe-out dt。 4. The method for precise control of injection quantity of a static leak-free injector based on a physical information neural network according to claim 1, characterized in that: The loss function of the PINN model in step (4) is expressed as: The physical constraint loss E1 is expressed as: Where, represents the injection amount predicted by the PINN model, κ is the ratio of the injection amount and the backflow amount recorded in the database; The regression loss constraint E2 is expressed as: The total loss function of the model is: E = λ1E1 + E2, where λ1 is the weight coefficient of the physical constraint loss.

5. The method for precise control of injection quantity of a static leak-free injector based on a physical information neural network according to claim 1, characterized in that: The specific step of the segmented PID control in step (5) is: According to the deviation between the actual injection amount and the target injection amount, different PID control parameters are applied in different stages according to the deviation degree, as follows: Where ΔPW(k) is the injection pulse width correction value at the kth injection, e(k-1) is the injection quantity deviation at the k-1th injection, σ1, σ2, and σ3 are the respective deviation thresholds of different injection quantity deviation segments, is the correction coefficient of the first deviation segment, is the correction coefficient of the second deviation segment, is the correction coefficient of the third deviation segment; The segmented PID outputs the control signal for the kth injection based on the calculated injection pulse width correction value: PW(k)=PW(k-1)+ΔPW(k).

Images