Method for regulating and controlling injection consistency of injectors of all cylinders of engine by considering transient supply-injection disturbance

Abstract

The invention aims to provide a method for regulating and controlling injection consistency of injectors of all cylinders of an engine by considering transient supply-injection disturbance, and belongs to the field of engines. The method comprises the following steps: firstly, acquiring data of a rail pressure curve, an injector inlet pressure curve, a driving current curve and an injection rate curve at different supply moments and injection moments on a high-pressure fuel injection system experiment table, and constructing a data set; then, based on the data set, a Transform time sequence model with the rail pressure, the pressure and the current curve as input and the injection rate as output is established, the injection rate is predicted online, and the injection amount is calculated through time integration. And then, feedback regulation and control are conducted on the injection pulse widths at different supply and injection moments through a programmable logic controller, and accurate control over the single-time injection amount and the multiple-time injection amount of each cylinder is achieved. According to the method, supply-injection disturbance under the transient working condition is effectively dealt with, the consistency of the injection amount of each cylinder is improved, and the combustion efficiency and the operation stability of the engine are improved.

Description

Technical Field

[0001] The present invention relates to an engine control method, in particular to an injector control method. Background Art

[0002] With the continuous improvement of performance requirements for modern engines, especially in the context of increasingly stringent emission regulations, precise control of engine fuel injection systems has become a key technology to improve engine performance and reduce emissions. High-pressure fuel injection systems have been widely used due to their good injection pressure control capabilities and injection flexibility. However, during the actual operation of the engine, due to the influence of multiple factors such as transient operating conditions, supply pressure fluctuations, and injection pressure fluctuations, there are certain differences in the injection amount and injection pattern of each cylinder injector. This injection inconsistency will directly affect the combustion uniformity of the engine, and then have an adverse effect on the engine's power, economy, and emission performance.

[0003] At present, the research on injection quantity control mainly focuses on injector control under static and steady-state conditions, while relatively few studies are conducted on injector injection consistency under transient disturbances. Traditional control methods are difficult to achieve accurate matching of the injection quantity of each cylinder injector under dynamic conditions. Summary of the invention

[0004] The purpose of the present invention is to provide a method for controlling the injection consistency of injectors in each cylinder of an engine taking into account transient supply-injection disturbances, which can improve the injection consistency of each cylinder of the engine and reduce the injection deviation caused by transient supply-injection disturbances, so as to achieve goals such as efficient combustion and low emissions of the engine under different operating conditions.

[0005] The object of the present invention is achieved in that:

[0006] The present invention provides a method for controlling the injection consistency of injectors of each cylinder of an engine taking into account transient supply-injection disturbances, which is characterized by comprising the following steps:

[0007] (1) On a high-pressure fuel injection system test bench, obtain experimental data of the rail pressure curve of each cycle, the injector inlet pressure curve of each cylinder, the drive current curve, and the injection rate curve at different supply and injection times, and establish a data set;

[0008] (2) A Transformer time series model is constructed with rail pressure curve, injector inlet pressure curve and drive current curve as input and injection rate as output. The model is trained based on the data set so that the model can predict the injection rate online and output the injection amount by time integration at the same time.

[0009] (3) The Transformer model is embedded in a programmable logic controller, and the injection pulse width feedback control logic at different supply and injection times is programmed in the controller so that the control parameters of the injection strategy of each cylinder are optimized according to the deviation between the actual injection amount of each cylinder and the target injection amount;

[0010] (4) During the operation of the engine, the rail pressure curve of each cycle, the injector inlet pressure curve of each cylinder, and the drive current curve are collected in real time. The constructed Transformer time series model is used to online predict the injection amount of each cylinder, and the deviation between the actual injection amount and the target injection amount of each cylinder is calculated. The injection amount of each cylinder is controlled online in a closed loop based on the programmable logic controller.

[0011] The present invention may also include:

[0012] 1. The different supply timings and injection timings in step (1) include:

[0013] Three situations under the single injection strategy: the supply time is before the injection time, the supply time is after the injection time and before the end of injection, and the supply time is after the injection time and after the end of injection;

[0014] There are three situations under the multiple injection strategy: the supply time is before the first injection time, the supply time is after the first injection time and before the end of the last injection, and the supply time is after the first injection time and after the end of the last injection.

[0015] 2. The construction method of the Transformer time series model in step (2) is:

[0016] (21) Divide the data into time segments of the same length, and each time segment is used as a sample; let the input time series Output time series Y inject =[Y inject,1 ,Y inject,2 ,...,Y inject,T ], where T represents the number of time steps, and the input time series and output time series are normalized respectively;

[0017] (22) Create position encoding to make the model aware of the information of the time step. The position encoding vector PE is:

[0018]

[0019] Where i is the dimension index, d model is the embedding dimension;

[0020] (23) Establish the Transformer encoder layer:

[0021] Capture the long-term temporal dependencies of input time series data through multi-head self-attention layers:

[0022]

[0023] 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;

[0024] After each attention layer, build a feed-forward neural network:

[0025] FFN(x)=max(0,xW1+b1)W2+b2

[0026] In the formula, x represents the input of each layer, W and b are the weight and bias of the feedforward neural network, and the subscripts 1 and 2 represent the first and second layers in the feedforward network, respectively;

[0027] 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:

[0028] Output=LayerNorm(x+Attention)

[0029] Output = LayerNorm (x + FFN)

[0030] (24) Construct the loss function of the model:

[0031]

[0032] In the formula, Characterize the input time series and predict the output of the injector inlet pressure fluctuation through the Transformer model; Y is the actual value of the injector inlet pressure fluctuation, and N represents the number of samples in the data set;

[0033] (25) For the six situations under single injection strategy and multiple injection strategy, model training is performed based on the data set to obtain six sets of Transformer model parameters to meet the prediction requirements in different situations.

[0034] 3. The injection pulse width feedback control logic in step (3) is:

[0035] For a single injection, the injection pulse width is corrected according to the deviation between the actual injection amount and the target injection amount predicted online by the Transformer model in the current cycle:

[0036] PW(k)=k p ·e(k-1)+k i ·∫e(k)dk

[0037] Where PW(k) is the injection pulse width at the kth injection, e(k-1) is the injection amount deviation at the k-1th injection, and k p , k i is the correction factor;

[0038] For multiple injections, the injection pulse width of each injection stage is corrected according to the deviation between the actual injection amount and the target injection amount predicted online by the Transformer model in the current cycle:

[0039] PW m (k) = k p ·e m (k-1)+k i ·∫e m (k)dk

[0040] Wherein, the subscript m represents the injection stage.

[0041] 4. The predicted injection amount of each cylinder in step (4) is specifically:

[0042] According to the engine supply timing and the injection timing of each cylinder, the Transformer model parameters suitable for each cylinder are used to predict the injection rate of each cylinder online based on the model, and the injection amount is obtained by integration; for a single injection, only the injection rate is predicted, and the injection amount is obtained by integration; for multiple injections, the injection rate of each injection stage is predicted, and the injection amount of each injection stage is obtained by integration.

[0043] The advantages of the present invention are as follows: the present invention can predict the single injection rate and multiple injection rates of each cylinder and the corresponding injection amount under different supply-injection disturbance conditions in real time online; not only can the injection amount of each cylinder under a single injection be regulated, but also the injection amount of each injection stage under multiple injections can be regulated, thereby significantly improving the injection consistency of the injectors of each cylinder. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a flow chart of the present invention;

[0045] Figure 2 Schematic diagram of the experiment for jet consistency control. DETAILED DESCRIPTION

[0046] The present invention is described in more detail below with reference to the accompanying drawings:

[0047] Combination Figure 1-2 The present invention provides a method for controlling the injection consistency of each cylinder injector of an engine considering transient supply-injection disturbance, comprising the following steps:

[0048] Step S1, on a high-pressure fuel injection system test bench, obtaining experimental data of a rail pressure curve of each cycle, an injector inlet pressure curve of each cylinder, a driving current curve, and an injection rate curve at different supply times and injection times, and establishing a data set;

[0049] Step S2, constructing a Transformer time series model with rail pressure curve, injector inlet pressure curve and driving current curve as input and injection rate as output, and training based on the data set so that the model can predict the injection rate online and output the injection amount by time integration at the same time;

[0050] Step S3, embedding the Transformer model into a programmable logic controller, and programming the injection pulse width feedback control logic at different supply timings and injection timings in the controller, so that the control parameters of the injection strategy of each cylinder can be optimized according to the deviation between the actual injection amount of each cylinder and the target injection amount;

[0051] Step S4, such as Figure 2 As shown in the figure, during the actual engine operation, the rail pressure curve of each cycle, the injector inlet pressure curve of each cylinder, and the driving current curve are collected in real time. The constructed Transformer time series model is used to online predict the injection amount of each cylinder, and the deviation between the actual injection amount and the target injection amount of each cylinder is calculated. The injection amount of each cylinder is controlled online in a closed loop based on the programmable logic controller to improve the injection consistency of each cylinder.

[0052] The different supply timings and injection timings in step S1 include:

[0053] Three situations under the single injection strategy: the supply time is before the injection time, the supply time is after the injection time and before the end of injection, and the supply time is after the injection time and after the end of injection;

[0054] There are three situations under the multiple injection strategy: the supply time is before the first injection time, the supply time is after the first injection time and before the end of the last injection, and the supply time is after the first injection time and after the end of the last injection.

[0055] The specific steps of building the Transformer time series model in step S2 are:

[0056] Step (a) divides the data into time segments of the same length, and each time segment is used as a sample. Suppose the input time series is Output time series Y inject =[Y inject,1 ,Y inject,2 ,…,Y inject,T ], where T represents the number of time steps. The input time series and the output time series are normalized separately.

[0057] Step (b) creates a position encoding to make the model aware of the time step information. The position encoding vector PE is calculated as:

[0058]

[0059] Where i is the dimension index, d model is the embedding dimension.

[0060] Step (c) builds the Transformer encoder layer, specifically:

[0061] Capture the long-term temporal dependencies of input time series data through multi-head self-attention layers:

[0062]

[0063] 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.

[0064] A feedforward neural network is established after each attention layer to improve the nonlinear representation ability of the model, which is expressed as:

[0065] FFN(x)=max(0,xW1+b1)W2+b2 (3)

[0066] Where x represents the input of each layer, W and b are the weight and bias of the feedforward neural network, and the subscripts 1 and 2 represent the first and second layers in the feedforward network, respectively.

[0067] 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:

[0068] Output=LayerNorm(x+Attention) (4)

[0069] Output=LayerNorm(x+FFN) (5)

[0070] Step (d) constructs the loss function of the model to evaluate the prediction error of the model, specifically:

[0071]

[0072] In the formula, Characterize the input time series and predict the output injector inlet pressure fluctuation through the Transformer model; Y is the actual value of the injector inlet pressure fluctuation, and N represents the number of samples in the data set.

[0073] In step (e), for six situations under the single injection strategy and the multiple injection strategy, model training is performed based on the data set to obtain six sets of Transformer model parameters to meet the prediction requirements under different situations.

[0074] The injection pulse width feedback control logic in step S3 is specifically as follows:

[0075] For a single injection, the injection pulse width is corrected according to the deviation between the actual injection amount and the target injection amount predicted online by the Transformer model in the current cycle:

[0076] PW(k)=k p ·e(k-1)+k i ·∫e(k)dk (7)

[0077] Where PW(k) is the injection pulse width at the kth injection, e(k-1) is the injection amount deviation at the k-1th injection, and k p , k i is the correction factor.

[0078] For multiple injections, the injection pulse width of each injection stage is corrected according to the deviation between the actual injection amount and the target injection amount predicted online by the Transformer model in the current cycle:

[0079] PW m (k) = k p ·e m (k-1)+k i ·∫e m (k)dk (8)

[0080] Wherein, the subscript m represents the injection stage.

[0081] The injection amount of each cylinder predicted in step S4 is specifically:

[0082] According to the engine supply time and the injection time of each cylinder, the Transformer model parameters suitable for each cylinder are used to predict the injection rate of each cylinder online based on the model, and the injection amount is obtained by integration. Among them, for a single injection, only the injection rate is predicted, and the injection amount is obtained by integration; for multiple injections, the injection rate of each injection stage is predicted, and the injection amount of each injection stage is obtained by integration.

Claims

1. A method for controlling the injection consistency of injectors in each cylinder of an engine considering transient supply-injection disturbances, characterized in that: The following steps are involved: (1) On a high-pressure fuel injection system test bench, obtain experimental data of the rail pressure curve of each cycle, the injector inlet pressure curve of each cylinder, the drive current curve, and the injection rate curve at different supply and injection times, and establish a data set; (2) A Transformer time series model is constructed with rail pressure curve, injector inlet pressure curve and drive current curve as input and injection rate as output. The model is trained based on the data set so that the model can predict the injection rate online and output the injection amount by time integration at the same time. (3) The Transformer model is embedded in a programmable logic controller, and the injection pulse width feedback control logic at different supply and injection times is programmed in the controller so that the control parameters of the injection strategy of each cylinder are optimized according to the deviation between the actual injection amount of each cylinder and the target injection amount; (4) During the operation of the engine, the rail pressure curve of each cycle, the injector inlet pressure curve of each cylinder, and the drive current curve are collected in real time. The constructed Transformer time series model is used to online predict the injection amount of each cylinder, and the deviation between the actual injection amount and the target injection amount of each cylinder is calculated. The injection amount of each cylinder is controlled online in a closed loop based on the programmable logic controller.

2. The method for controlling the injection consistency of each cylinder injector of an engine considering transient supply-injection disturbance according to claim 1, characterized in that: The different supply timings and injection timings in step (1) include: Three situations under the single injection strategy: the supply time is before the injection time, the supply time is after the injection time and before the end of injection, and the supply time is after the injection time and after the end of injection; There are three situations under the multiple injection strategy: the supply time is before the first injection time, the supply time is after the first injection time and before the end of the last injection, and the supply time is after the first injection time and after the end of the last injection.

3. The method for controlling the injection consistency of each cylinder injector of an engine considering transient supply-injection disturbance according to claim 1, characterized in that: The construction method of the Transformer time series model in step (2) is: (21) Divide the data into time segments of the same length, and each time segment is used as a sample; let the input time series Output time series Y inject =[Y inject,1 ,Y inject,2 ,...,Y inject,T ], where T represents the number of time steps, and the input time series and output time series are normalized respectively; (22) Create position encoding to make the model aware of the information of the time step. The position encoding vector PE is: Where i is the dimension index, d model is the embedding dimension; (23) Establish the Transformer encoder layer: Capture 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; After each attention layer, build a feed-forward neural network: FFN(x)=max(0,xW1+b1)W2+b2 In the formula, x represents the input of each layer, W and b are the weight and bias of the feedforward neural network, and the subscripts 1 and 2 represent the first and second layers in the feedforward network, respectively; 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: In the formula, Characterize the input time series and predict the output of the injector inlet pressure fluctuation through the Transformer model; Y is the actual value of the injector inlet pressure fluctuation, and N represents the number of samples in the data set; (25) For the six situations under single injection strategy and multiple injection strategy, model training is performed based on the data set to obtain six sets of Transformer model parameters to meet the prediction requirements in different situations.

4. The method for controlling the injection consistency of each cylinder injector of an engine considering transient supply-injection disturbance according to claim 1, characterized in that: The injection pulse width feedback control logic in step (3) is: For a single injection, the injection pulse width is corrected according to the deviation between the actual injection amount and the target injection amount predicted online by the Transformer model in the current cycle: PW(k)=k p ·e(k-1)+k i ·∫e(k)dk Where PW(k) is the injection pulse width at the kth injection, e(k-1) is the injection amount deviation at the k-1th injection, and k p , k i is the correction factor; For multiple injections, the injection pulse width of each injection stage is corrected according to the deviation between the actual injection amount and the target injection amount predicted online by the Transformer model in the current cycle: PW m (k)=k p ·e m (k-1)+k i ·∫e m (k)dk Wherein, the subscript m represents the injection stage.

5. The method for controlling the injection consistency of each cylinder injector of an engine considering transient supply-injection disturbance according to claim 1, characterized in that: The predicted injection amount of each cylinder in step (4) is specifically: According to the engine supply timing and the injection timing of each cylinder, the Transformer model parameters suitable for each cylinder are used to predict the injection rate of each cylinder online based on the model, and the injection amount is obtained by integration; for a single injection, only the injection rate is predicted, and the injection amount is obtained by integration; For multiple injections, the injection rate of each injection phase is predicted and the injection quantity of each injection phase is obtained by integration.

Images