A vehicle engine oil quantity control method and system

Abstract

The application provides a vehicle engine oil quantity control method and system, which comprises the following steps: collecting target time sequence data, and creating a corresponding state parameter matrix in real time according to the target time sequence data; constructing a corresponding flow response prediction model according to the target time sequence data, and inputting the state parameter matrix into the flow response prediction model to correspondingly output a flow response prediction value generated in a future preset control period; constructing a corresponding two-dimensional compensation matrix in real time based on instruction transmission delay data and the flow response prediction value, and generating an adaptive reference flow instruction according to a target working condition of the vehicle engine; performing time period correction processing on the reference flow instruction through the two-dimensional compensation matrix to generate a corresponding target instruction sequence in real time, and converting the target instruction sequence into a corresponding target control instruction in real time, so as to correspondingly complete real-time output of fuel oil quantity according to the target control instruction. The application can effectively improve the control efficiency of fuel flow.

Description

Technical Field

[0001] This invention relates to the field of vehicle technology, and in particular to a method and system for controlling the oil quantity of a vehicle engine. Background Technology

[0002] With the advancement of technology and the rapid development of the times, people have made significant progress in the field of vehicle technology. Among them, precise control of fuel flow is the key to the safe operation and performance optimization of vehicle engines. As the requirements for engine control precision increase, traditional hydraulic control is gradually being replaced by fuel pumps. However, the complex driving conditions of vehicles pose challenges to the stability of engines.

[0003] Among them, the fuel pump inside the existing vehicle engine faces a number of uncertain problems in actual use. First, changes in vehicle vibration cause fuel pressure fluctuations; second, temperature fluctuations change fuel viscosity and density, affecting the flow inside the pump; and third, long-term operation leads to pump wear and dynamic decay of volumetric efficiency.

[0004] Furthermore, the aforementioned pressure and temperature fluctuations can cause deviations between the actual flow rate and the control command. At the same time, the decline in volumetric efficiency will delay the flow response, making it difficult for the fuel pump to achieve precise flow control across the entire flow range, thereby reducing the control efficiency of fuel flow. Summary of the Invention

[0005] Therefore, the purpose of this invention is to provide a vehicle engine fuel quantity control method and system to solve the problem that existing technologies generate commands and delay flow response during the actual control of fuel flow, resulting in the inability to accurately achieve flow control.

[0006] The first aspect of the present invention proposes: A method for controlling the oil quantity in a vehicle engine, wherein the method includes: During the actual operation of the vehicle engine, the fuel pressure, pipeline temperature, engine speed and target time-series data of the fuel pump output, the engine speed and the command transmission are collected in real time, and the corresponding state parameter matrix is ​​created in real time based on the target time-series data. A corresponding flow response prediction model is constructed based on the target time series data, and the state parameter matrix is ​​input into the flow response prediction model to output the corresponding flow response prediction value generated within the future preset control period. A corresponding two-dimensional compensation matrix is ​​constructed in real time based on the instruction transmission delay data and the predicted flow response value, and an adapted reference flow instruction is generated according to the target operating condition of the vehicle engine. The reference flow command is corrected in time periods by the two-dimensional compensation matrix to generate the corresponding target command sequence in real time, and the target command sequence is converted into the corresponding target control command in real time to complete the real-time output of fuel quantity according to the target control command.

[0007] The beneficial effects of this invention are as follows: By collecting target time-series data corresponding to the vehicle engine in real time, the current working state of the vehicle engine can be obtained, and a corresponding state parameter matrix can be generated. Based on this, an adapted flow response prediction model can be created in real time, and the corresponding flow response prediction value can be output in real time. Based on this, in order to improve the accuracy of the prediction, an adapted two-dimensional compensation matrix can also be created, and the current reference flow command can be corrected in real time, and the target control command can be generated. Based on this, the output of fuel flow can be completed objectively and accurately, thereby improving the control efficiency of fuel flow.

[0008] Furthermore, the step of inputting the state parameter matrix into the flow response prediction model to output the corresponding flow response prediction value generated within a future preset control period includes: The sample time series parameters generated at several consecutive sampling times are extracted from the state parameter matrix in real time, and several single-column matrices are constructed according to the parameter types contained in the sample time series parameters. A Fourier transform is performed on each of the single-column matrices to generate the corresponding frequency domain feature matrix in real time, and the amplitude data of the first 20 frequency components in each frequency feature matrix are extracted in real time. According to the preset splicing order, an 80-dimensional feature vector is generated in real time based on the amplitude data, and the feature vector is input into the preset flow response model to output the corresponding flow response prediction value.

[0009] Furthermore, the step of inputting the feature vector into the traffic response preset model to output the corresponding traffic response prediction value includes: The 80-dimensional feature vector is processed by a sliding window, and divided into 8 overlapping windows with each window consisting of 10 dimensions. The weighted average value of the elements in each overlapping window is then calculated. The weighted average of each overlapping window is multiplied by the feature vector to generate a corresponding enhanced feature vector. The convolutional attention module in the traffic response prediction model is then used to filter each enhanced feature vector to generate a corresponding target feature vector. The target feature vector is converted into corresponding time-series features in real time by using a preset GRU layer and tanh activation function, and the time-series features are mapped by a preset fully connected layer to generate the traffic response prediction value.

[0010] Furthermore, the step of constructing the corresponding two-dimensional compensation matrix in real time based on the instruction transmission delay data and the traffic response prediction value includes: The instruction transmission delay data is frequency domain converted to extract the corresponding delay amplitude spectrum features in real time, and the traffic response prediction value is interpolated according to the time resolution of the preset control period to obtain the corresponding prediction value sequence. Using the frequency band division result obtained from the frequency domain transformation as the row dimension of the matrix and the time resolution node after interpolation as the column dimension of the matrix, a corresponding blank two-dimensional matrix is ​​constructed in real time, and the delay amplitude spectrum feature and the predicted value sequence are filled into the corresponding positions of the blank two-dimensional matrix to generate the corresponding intermediate two-dimensional matrix. The elements in the intermediate two-dimensional matrix are dynamically scaled based on the real-time collected fuel pressure fluctuation cycle to form the corresponding two-dimensional compensation matrix.

[0011] Furthermore, the step of generating an adapted reference flow command based on the target operating condition of the vehicle engine includes: The characteristic parameters generated by the vehicle engine during the start-up, cruising, and acceleration phases are analyzed, and the start-up, cruising, and acceleration phases are divided into several sub-phases according to their time proportions. The theoretical threshold of fuel flow corresponding to each sub-phase is extracted simultaneously. Using the division results of the sub-stages as matrix row identifiers and the engine speed gradient under the target operating condition as matrix column identifiers, a corresponding flow mapping matrix is ​​constructed, and the theoretical fuel flow threshold of each sub-stage is filled into each matrix unit of the flow mapping matrix. The maximum delay value corresponding to the instruction transmission delay data is collected in real time, and the maximum delay value is used to perform offset correction processing on each matrix unit. The matrix units after correction processing are then connected in a stage sequence to form the reference flow instruction.

[0012] Furthermore, the step of performing time-segmented correction processing on the baseline flow command using the two-dimensional compensation matrix to generate the corresponding target command sequence in real time includes: The baseline flow command is decomposed into several command segments, and the compensation coefficient corresponding to each command segment is extracted in real time from the two-dimensional compensation matrix. Each instruction segment and its corresponding compensation coefficient are multiplied point by point to obtain the preliminarily corrected instruction subsequence. The instruction subsequence is smoothed according to the rate of change of the vehicle engine speed, and all the smoothed subsequences are spliced ​​together in chronological order to form the target instruction sequence.

[0013] Furthermore, the step of smoothing the instruction sub-sequence according to the rate of change of the vehicle engine speed, and then concatenating all the smoothed sub-sequences in chronological order to form the target instruction sequence includes: The endpoint flow values ​​of two adjacent instruction subsequences and the corresponding engine speed change rate are collected in real time, and the number of sampling points of the transition segment between two adjacent instruction subsequences is determined in real time based on the engine speed change rate. The splicing coefficient of the corresponding transition segment is determined in real time based on the number of sampling points. Each instruction subsequence and each transition segment are connected according to the splicing coefficient to form a continuously changing target instruction sequence.

[0014] The second aspect of the present invention proposes: A vehicle engine oil quantity control system, wherein the system includes: The data acquisition module is used to collect the fuel pressure, pipeline temperature, engine speed and target time-series data of the fuel pump output in real time during the actual operation of the vehicle engine, and to create the corresponding state parameter matrix in real time based on the target time-series data. The construction module is used to construct a corresponding traffic response prediction model based on the target time series data, and input the state parameter matrix into the traffic response prediction model to output the corresponding traffic response prediction value generated in the future preset control period. The processing module is used to construct a corresponding two-dimensional compensation matrix in real time based on the instruction transmission delay data and the predicted flow response value, and to generate an adapted reference flow instruction according to the target operating condition of the vehicle engine. The correction module is used to perform time-segmented correction processing on the reference flow command through the two-dimensional compensation matrix to generate the corresponding target command sequence in real time, and to convert the target command sequence into the corresponding target control command in real time, so as to complete the real-time output of fuel quantity according to the target control command.

[0015] Furthermore, the building module is specifically used for: The sample time series parameters generated at several consecutive sampling times are extracted from the state parameter matrix in real time, and several single-column matrices are constructed according to the parameter types contained in the sample time series parameters. A Fourier transform is performed on each of the single-column matrices to generate the corresponding frequency domain feature matrix in real time, and the amplitude data of the first 20 frequency components in each frequency feature matrix are extracted in real time. According to the preset splicing order, an 80-dimensional feature vector is generated in real time based on the amplitude data, and the feature vector is input into the preset flow response model to output the corresponding flow response prediction value.

[0016] Furthermore, the building module is specifically used for: The 80-dimensional feature vector is processed by a sliding window, and divided into 8 overlapping windows with each window consisting of 10 dimensions. The weighted average value of the elements in each overlapping window is then calculated. The weighted average of each overlapping window is multiplied by the feature vector to generate a corresponding enhanced feature vector. The convolutional attention module in the traffic response prediction model is then used to filter each enhanced feature vector to generate a corresponding target feature vector. The target feature vector is converted into corresponding time-series features in real time by using a preset GRU layer and tanh activation function, and the time-series features are mapped by a preset fully connected layer to generate the traffic response prediction value.

[0017] Furthermore, the processing module is specifically used for: The instruction transmission delay data is frequency domain converted to extract the corresponding delay amplitude spectrum features in real time, and the traffic response prediction value is interpolated according to the time resolution of the preset control period to obtain the corresponding prediction value sequence. Using the frequency band division result obtained from the frequency domain transformation as the row dimension of the matrix and the time resolution node after interpolation as the column dimension of the matrix, a corresponding blank two-dimensional matrix is ​​constructed in real time, and the delay amplitude spectrum feature and the predicted value sequence are filled into the corresponding positions of the blank two-dimensional matrix to generate the corresponding intermediate two-dimensional matrix. The elements in the intermediate two-dimensional matrix are dynamically scaled based on the real-time collected fuel pressure fluctuation cycle to form the corresponding two-dimensional compensation matrix.

[0018] Furthermore, the processing module is specifically used for: The characteristic parameters generated by the vehicle engine during the start-up, cruising, and acceleration phases are analyzed, and the start-up, cruising, and acceleration phases are divided into several sub-phases according to their time proportions. The theoretical threshold of fuel flow corresponding to each sub-phase is extracted simultaneously. Using the division results of the sub-stages as matrix row identifiers and the engine speed gradient under the target operating condition as matrix column identifiers, a corresponding flow mapping matrix is ​​constructed, and the theoretical fuel flow threshold of each sub-stage is filled into each matrix unit of the flow mapping matrix. The maximum delay value corresponding to the instruction transmission delay data is collected in real time, and the maximum delay value is used to perform offset correction processing on each matrix unit. The matrix units after correction processing are then connected in a stage sequence to form the reference flow instruction.

[0019] Furthermore, the correction module is specifically used for: The baseline flow command is decomposed into several command segments, and the compensation coefficient corresponding to each command segment is extracted in real time from the two-dimensional compensation matrix. Each instruction segment and its corresponding compensation coefficient are multiplied point by point to obtain the preliminarily corrected instruction subsequence. The instruction subsequence is smoothed according to the rate of change of the vehicle engine speed, and all the smoothed subsequences are spliced ​​together in chronological order to form the target instruction sequence.

[0020] Furthermore, the correction module is specifically used for: The endpoint flow values ​​of two adjacent instruction subsequences and the corresponding engine speed change rate are collected in real time, and the number of sampling points of the transition segment between two adjacent instruction subsequences is determined in real time based on the engine speed change rate. The splicing coefficient of the corresponding transition segment is determined in real time based on the number of sampling points. Each instruction subsequence and each transition segment are connected according to the splicing coefficient to form a continuously changing target instruction sequence.

[0021] The third aspect of the present invention proposes: A computer includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the vehicle engine fuel quantity control method as described above.

[0022] The fourth aspect of the present invention proposes: A readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the vehicle engine fuel quantity control method as described above.

[0023] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0024] Figure 1 A flowchart of a vehicle engine oil quantity control method provided in the first embodiment of the present invention; Figure 2 This is a structural block diagram of a vehicle engine oil quantity control system provided in the third embodiment of the present invention.

[0025] The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0026] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0027] It should be noted that when a component is said to be "fixed to" another component, it can be directly on the other component or there may be an intervening component. When a component is said to be "connected to" another component, it can be directly connected to the other component or there may be an intervening component. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0028] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0029] Please see Figure 1 The figure shows a vehicle engine fuel quantity control method provided in the first embodiment of the present invention. The vehicle engine fuel quantity control method provided in this embodiment can objectively and accurately control the fuel output of the vehicle engine, thereby improving the control efficiency of fuel flow.

[0030] Specifically, this embodiment provides: A method for controlling the oil quantity of a vehicle engine specifically includes the following steps: Step S10: During the actual operation of the vehicle engine, the fuel pressure, pipeline temperature, engine speed and target timing data of the command transmission output by the fuel pump are collected in real time, and the corresponding state parameter matrix is ​​created in real time based on the target timing data. It's important to note that during actual vehicle engine operation, four core parameters need to be collected in real time: fuel pump output pressure (reflecting fuel supply intensity), pipeline temperature (affecting fuel viscosity and flowability), engine speed (related to power demand), and target timing data for command transmission (recording the correlation between the sending and execution times of control commands). Based on this, a corresponding state parameter matrix can be constructed according to the time sequence of events. It should be noted that the rows and columns of this state parameter matrix correspond to different sampling times and parameter types, thus providing a data structure for subsequent analysis and processing.

[0031] Step S20: Construct a corresponding flow response prediction model based on the target time series data, and input the state parameter matrix into the flow response prediction model to output the corresponding flow response prediction value generated in the future preset control period. It should be noted that a flow response prediction model (which may be a machine learning model or a mechanistic model) is trained or constructed using target time-series data (including historical commands and corresponding flow responses). After inputting the state parameter matrix into the model, the model can output the predicted flow response value within a preset control period (e.g., 100ms), enabling the prediction of fuel flow change trends and providing a basis for adjusting the control strategy in advance. It should also be pointed out that after obtaining the required target time-series data through the above steps, the current target time-series data can be split into a validation set, a test set, and a training set according to a preset ratio. Based on this, and combined with an existing deep learning network, the corresponding model training can be performed to generate the aforementioned flow response prediction model for subsequent processing.

[0032] Step S30: Based on the instruction transmission delay data and the predicted flow response value, a corresponding two-dimensional compensation matrix is ​​constructed in real time, and an adapted reference flow instruction is generated according to the target operating condition of the vehicle engine. It should be noted that in practical applications, various commands incur transmission delays during execution. Therefore, to improve control accuracy, this invention needs to avoid the impact of transmission delays. To this end, this invention monitors the transmission delay data of any commands that have already occurred in real time, i.e., the time difference between the issuance and execution of the control command. Specifically, this invention combines the aforementioned flow response prediction value and constructs a corresponding two-dimensional compensation matrix in real time. The dimension of this two-dimensional compensation matrix can correlate the delay time with the prediction error, thereby quantifying the compensation requirements under different scenarios. Simultaneously, based on the engine's target operating condition (such as start-up, cruising, and acceleration), an adapted baseline flow command (theoretical flow value that meets the power requirements of the current operating condition) is generated as a basic reference for control, facilitating subsequent processing.

[0033] Step S40: The reference flow command is corrected in time periods using the two-dimensional compensation matrix to generate a corresponding target command sequence in real time, and the target command sequence is converted into a corresponding target control command in real time to complete the real-time output of fuel quantity according to the target control command.

[0034] It should be noted that after obtaining the required two-dimensional compensation matrix and reference flow command through the above steps, the reference flow command is immediately corrected in stages using the two-dimensional compensation matrix (for different delay and prediction deviation scenarios), generating a dynamically adjusted target command sequence. This sequence is converted into target control commands (such as voltage / frequency signals) executable by the fuel pump, ultimately achieving real-time and accurate fuel flow output, ensuring efficient and stable engine operation. This method ultimately enables objective and accurate control of fuel output within the vehicle engine, thereby improving fuel control efficiency.

[0035] Second Embodiment Furthermore, the step of inputting the state parameter matrix into the flow response prediction model to output the corresponding flow response prediction value generated within a future preset control period includes: The sample time series parameters generated at several consecutive sampling times are extracted from the state parameter matrix in real time, and several single-column matrices are constructed according to the parameter types contained in the sample time series parameters. A Fourier transform is performed on each of the single-column matrices to generate the corresponding frequency domain feature matrix in real time, and the amplitude data of the first 20 frequency components in each frequency feature matrix are extracted in real time. According to the preset splicing order, an 80-dimensional feature vector is generated in real time based on the amplitude data, and the feature vector is input into the preset flow response model to output the corresponding flow response prediction value.

[0036] It should be noted that parameters from a series of consecutive sampling times (e.g., the last 50 times) are extracted from the state parameter matrix and split into several single-column matrices according to parameter type (fuel pressure, temperature, etc.), with each column corresponding to the time-series change of a single parameter. This allows for independent feature analysis of the parameters, avoiding interference from different parameter types. Based on this, a Fourier transform is performed on each single-column matrix to convert the time-domain signal into a frequency-domain feature matrix (reflecting the frequency distribution of parameter changes). The amplitude data of the first 20 frequency components (the main frequencies with the most concentrated energy) are extracted, high-frequency noise is filtered out, and key fluctuation features (such as the periodic pulsation of fuel pressure) are retained. Finally, all amplitude data are concatenated in a preset order (e.g., pressure, temperature, speed, time-series data) to form an 80-dimensional feature vector (4 types of parameters × 20 frequency components). This vector is input into the flow response prediction model, which learns the mapping relationship between features and flow response to output an accurate flow response prediction value. The above method can form a preliminary prediction value for subsequent processing.

[0037] Furthermore, the step of inputting the feature vector into the traffic response preset model to output the corresponding traffic response prediction value includes: The 80-dimensional feature vector is processed by a sliding window, and divided into 8 overlapping windows with each window consisting of 10 dimensions. The weighted average value of the elements in each overlapping window is then calculated. The weighted average of each overlapping window is multiplied by the feature vector to generate a corresponding enhanced feature vector. The convolutional attention module in the traffic response prediction model is then used to filter each enhanced feature vector to generate a corresponding target feature vector. The target feature vector is converted into corresponding time-series features in real time by using a preset GRU layer and tanh activation function, and the time-series features are mapped by a preset fully connected layer to generate the traffic response prediction value.

[0038] It should be noted that a sliding window process is applied to the 80-dimensional feature vector, dividing it into 10-dimensional windows, resulting in 8 overlapping windows (e.g., window 1: dimensions 1-10, window 2: dimensions 5-15, etc.). A weighted average of the elements within each window is calculated (weights can be assigned according to feature importance) to enhance the representativeness of local features. The weighted average of each window is multiplied by the original feature vector to generate an enhanced feature vector (strengthening the influence of key features). A convolutional attention module in the model (focusing on important features and suppressing irrelevant information) filters out target feature vectors strongly correlated with traffic response, improving the model's predictive accuracy. Based on this, a pre-defined GRU layer (gated recurrent unit, adept at handling time-series data) and a tanh activation function (mapping feature values ​​to the [-1,1] interval, enhancing non-linear expression) are used to convert the target feature vector into time-series features (capturing the dynamic correlation of parameters over time). Finally, a fully connected layer maps the time-series features to specific traffic response prediction values, completing the transformation from features to prediction results. The above method can objectively and accurately generate a preliminary traffic response prediction value, thus providing a basis for subsequent analysis and processing.

[0039] Furthermore, the step of constructing the corresponding two-dimensional compensation matrix in real time based on the instruction transmission delay data and the traffic response prediction value includes: The instruction transmission delay data is frequency domain converted to extract the corresponding delay amplitude spectrum features in real time, and the traffic response prediction value is interpolated according to the time resolution of the preset control period to obtain the corresponding prediction value sequence. Using the frequency band division result obtained from the frequency domain transformation as the row dimension of the matrix and the time resolution node after interpolation as the column dimension of the matrix, a corresponding blank two-dimensional matrix is ​​constructed in real time, and the delay amplitude spectrum feature and the predicted value sequence are filled into the corresponding positions of the blank two-dimensional matrix to generate the corresponding intermediate two-dimensional matrix. The elements in the intermediate two-dimensional matrix are dynamically scaled based on the real-time collected fuel pressure fluctuation cycle to form the corresponding two-dimensional compensation matrix.

[0040] It should be noted that, in order to objectively and accurately generate the two-dimensional compensation matrix for subsequent compensation, the following steps are taken: First, the command transmission delay data undergoes frequency domain transformation (e.g., Fourier transform) to extract the delay amplitude spectrum features (reflecting the frequency distribution characteristics of the delay). Second, the flow response prediction values ​​are interpolated according to a preset control cycle time resolution (e.g., 10ms / step) to obtain a continuous sequence of prediction values, ensuring precision in the time dimension. Third, a blank two-dimensional matrix is ​​constructed using the frequency band division results of the frequency domain transformation (e.g., 0-10Hz, 10-20Hz, etc.) as the matrix row dimension and the interpolated time resolution nodes as the column dimension. Fourth, the delay amplitude spectrum features (row direction) and the prediction value sequence (column direction) are filled into the corresponding positions to form an intermediate two-dimensional matrix, initially linking the delay characteristics with the predicted flow. Fifth, the fuel pressure fluctuation cycle (reflecting the inherent dynamic characteristics of the fuel system) is collected in real time. Based on this, the elements of the intermediate two-dimensional matrix are dynamically scaled, for example, increasing the compensation coefficient during periods of severe pressure fluctuation and decreasing it during stable periods, ultimately forming a two-dimensional compensation matrix that can adapt to the dynamic changes of the system. The above method can effectively complete the matrix processing of data, which facilitates subsequent processing.

[0041] Furthermore, the step of generating an adapted reference flow command based on the target operating condition of the vehicle engine includes: The characteristic parameters generated by the vehicle engine during the start-up, cruising, and acceleration phases are analyzed, and the start-up, cruising, and acceleration phases are divided into several sub-phases according to their time proportions. The theoretical threshold of fuel flow corresponding to each sub-phase is extracted simultaneously. Using the division results of the sub-stages as matrix row identifiers and the engine speed gradient under the target operating condition as matrix column identifiers, a corresponding flow mapping matrix is ​​constructed, and the theoretical fuel flow threshold of each sub-stage is filled into each matrix unit of the flow mapping matrix. The maximum delay value corresponding to the instruction transmission delay data is collected in real time, and the maximum delay value is used to perform offset correction processing on each matrix unit. The matrix units after correction processing are then connected in a stage sequence to form the reference flow instruction.

[0042] It's important to note that this involves analyzing characteristic parameters (such as speed range and power demand) of the engine during the starting (low speed, high fuel quantity), cruising (stable speed, economical fuel quantity), and acceleration (rapid speed increase, fuel surge) phases. Each phase is divided into several sub-phases based on time proportion (e.g., the starting phase can be divided into ignition, warm-up, and stable start sub-phases), and the theoretical fuel flow threshold (minimum / maximum fuel quantity to meet the requirements of that phase) is extracted for each sub-phase. Based on this, a flow mapping matrix is ​​constructed using the sub-phase division results as row identifiers and the engine speed gradient (speed change rate) under the target operating condition as column identifiers. The theoretical fuel flow thresholds for each sub-phase are filled into the corresponding matrix cells, forming a three-dimensional correlation between operating condition, speed, and flow rate, enabling rapid querying of flow benchmarks under different scenarios. Finally, the maximum delay value (the delay under the worst-case scenario) is extracted from the command transmission delay data to correct the offset of the matrix cells (e.g., issuing commands in advance to offset the delay effect). The corrected matrix cells are then concatenated according to the time sequence of the sub-phases to form continuous benchmark flow commands, ensuring that commands are synchronized with changes in operating conditions. The above method can initially form a baseline flow command for subsequent control, which will facilitate subsequent processing.

[0043] Furthermore, the step of performing time-segmented correction processing on the baseline flow command using the two-dimensional compensation matrix to generate the corresponding target command sequence in real time includes: The baseline flow command is decomposed into several command segments, and the compensation coefficient corresponding to each command segment is extracted in real time from the two-dimensional compensation matrix. Each instruction segment and its corresponding compensation coefficient are multiplied point by point to obtain the preliminarily corrected instruction subsequence. The instruction subsequence is smoothed according to the rate of change of the vehicle engine speed, and all the smoothed subsequences are spliced ​​together in chronological order to form the target instruction sequence.

[0044] It should be noted that the baseline flow rate command is decomposed into several command segments based on time or operating condition changes (e.g., one segment per sub-stage). Compensation coefficients corresponding to each command segment are extracted from the two-dimensional compensation matrix (based on the delay scenario and prediction deviation of the segment) for targeted correction. Based on this, each command segment is multiplied point-by-point by its corresponding compensation coefficient (e.g., if the baseline command is 100 L / h and the compensation coefficient is 1.05, the corrected value is 105 L / h), resulting in a preliminarily corrected command subsequence, eliminating errors caused by delay and prediction deviation. Finally, the command subsequence is smoothed according to the engine speed change rate (e.g., a smoother fuel transition is needed when the engine speed changes drastically) (using linear interpolation or exponential transition) to avoid sudden fuel quantity fluctuations impacting the engine. All smoothed subsequences are then concatenated in chronological order to form a continuous and stable target command sequence. This method allows for objective and accurate correction of the command sequence, facilitating subsequent processing.

[0045] Furthermore, the step of smoothing the instruction sub-sequence according to the rate of change of the vehicle engine speed, and then concatenating all the smoothed sub-sequences in chronological order to form the target instruction sequence includes: The endpoint flow values ​​of two adjacent instruction subsequences and the corresponding engine speed change rate are collected in real time, and the number of sampling points of the transition segment between two adjacent instruction subsequences is determined in real time based on the engine speed change rate. The splicing coefficient of the corresponding transition segment is determined in real time based on the number of sampling points. Each instruction subsequence and each transition segment are connected according to the splicing coefficient to form a continuously changing target instruction sequence.

[0046] It should be noted that, in order to objectively and effectively complete the splicing process of each instruction subsequence, the endpoint flow values ​​of two adjacent instruction subsequences are collected (e.g., the endpoint of the previous sequence is 100 L / h, and the starting point of the next sequence is 150 L / h) and the corresponding engine speed change rate at that moment. The larger the speed change rate, the longer the transition section needs to be (e.g., 20 sampling points for high speed changes and 5 for low speed changes) to ensure smoothness. Based on this, the splicing coefficient is determined according to the number of sampling points (e.g., using a linearly increasing coefficient from 0 to 1). The endpoint value of the previous subsequence decreases according to the coefficient, and the starting point value of the next subsequence increases according to the coefficient, realizing a gradual transition between the two. Finally, the splicing coefficient is used to connect the instruction subsequence and the transition section (e.g., the flow rate at the i-th transition point = the endpoint of the previous sequence × (1 - coefficient i) + the starting point of the next sequence × coefficient i), ultimately forming a target instruction sequence with no abrupt changes and continuous changes, ensuring the stability of engine operation. The above methods can objectively and accurately control the amount of fuel inside the vehicle engine, thereby improving fuel control efficiency.

[0047] Please see Figure 2 The third embodiment of the present invention provides: A vehicle engine oil quantity control system, wherein the system includes: The data acquisition module is used to collect the fuel pressure, pipeline temperature, engine speed and target time-series data of the fuel pump output in real time during the actual operation of the vehicle engine, and to create the corresponding state parameter matrix in real time based on the target time-series data. The construction module is used to construct a corresponding traffic response prediction model based on the target time series data, and input the state parameter matrix into the traffic response prediction model to output the corresponding traffic response prediction value generated in the future preset control period. The processing module is used to construct a corresponding two-dimensional compensation matrix in real time based on the instruction transmission delay data and the predicted flow response value, and to generate an adapted reference flow instruction according to the target operating condition of the vehicle engine. The correction module is used to perform time-segmented correction processing on the reference flow command through the two-dimensional compensation matrix to generate the corresponding target command sequence in real time, and to convert the target command sequence into the corresponding target control command in real time, so as to complete the real-time output of fuel quantity according to the target control command.

[0048] Furthermore, the building module is specifically used for: The sample time series parameters generated at several consecutive sampling times are extracted from the state parameter matrix in real time, and several single-column matrices are constructed according to the parameter types contained in the sample time series parameters. A Fourier transform is performed on each of the single-column matrices to generate the corresponding frequency domain feature matrix in real time, and the amplitude data of the first 20 frequency components in each frequency feature matrix are extracted in real time. According to the preset splicing order, an 80-dimensional feature vector is generated in real time based on the amplitude data, and the feature vector is input into the preset flow response model to output the corresponding flow response prediction value.

[0049] Furthermore, the building module is specifically used for: The 80-dimensional feature vector is processed by a sliding window, and divided into 8 overlapping windows with each window consisting of 10 dimensions. The weighted average value of the elements in each overlapping window is then calculated. The weighted average of each overlapping window is multiplied by the feature vector to generate a corresponding enhanced feature vector. The convolutional attention module in the traffic response prediction model is then used to filter each enhanced feature vector to generate a corresponding target feature vector. The target feature vector is converted into corresponding time-series features in real time by using a preset GRU layer and tanh activation function, and the time-series features are mapped by a preset fully connected layer to generate the traffic response prediction value.

[0050] Furthermore, the processing module is specifically used for: The instruction transmission delay data is frequency domain converted to extract the corresponding delay amplitude spectrum features in real time, and the traffic response prediction value is interpolated according to the time resolution of the preset control period to obtain the corresponding prediction value sequence. Using the frequency band division result obtained from the frequency domain transformation as the row dimension of the matrix and the time resolution node after interpolation as the column dimension of the matrix, a corresponding blank two-dimensional matrix is ​​constructed in real time, and the delay amplitude spectrum feature and the predicted value sequence are filled into the corresponding positions of the blank two-dimensional matrix to generate the corresponding intermediate two-dimensional matrix. The elements in the intermediate two-dimensional matrix are dynamically scaled based on the real-time collected fuel pressure fluctuation cycle to form the corresponding two-dimensional compensation matrix.

[0051] Furthermore, the processing module is specifically used for: The characteristic parameters generated by the vehicle engine during the start-up, cruising, and acceleration phases are analyzed, and the start-up, cruising, and acceleration phases are divided into several sub-phases according to their time proportions. The theoretical threshold of fuel flow corresponding to each sub-phase is extracted simultaneously. Using the division results of the sub-stages as matrix row identifiers and the engine speed gradient under the target operating condition as matrix column identifiers, a corresponding flow mapping matrix is ​​constructed, and the theoretical fuel flow threshold of each sub-stage is filled into each matrix unit of the flow mapping matrix. The maximum delay value corresponding to the instruction transmission delay data is collected in real time, and the maximum delay value is used to perform offset correction processing on each matrix unit. The matrix units after correction processing are then connected in a stage sequence to form the reference flow instruction.

[0052] Furthermore, the correction module is specifically used for: The baseline flow command is decomposed into several command segments, and the compensation coefficient corresponding to each command segment is extracted in real time from the two-dimensional compensation matrix. Each instruction segment and its corresponding compensation coefficient are multiplied point by point to obtain the preliminarily corrected instruction subsequence. The instruction subsequence is smoothed according to the rate of change of the vehicle engine speed, and all the smoothed subsequences are spliced ​​together in chronological order to form the target instruction sequence.

[0053] Furthermore, the correction module is specifically used for: The endpoint flow values ​​of two adjacent instruction subsequences and the corresponding engine speed change rate are collected in real time, and the number of sampling points of the transition segment between two adjacent instruction subsequences is determined in real time based on the engine speed change rate. The splicing coefficient of the corresponding transition segment is determined in real time based on the number of sampling points. Each instruction subsequence and each transition segment are connected according to the splicing coefficient to form a continuously changing target instruction sequence.

[0054] The fourth embodiment of the present invention provides a computer, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the vehicle engine oil quantity control method as described above.

[0055] The fifth embodiment of the present invention provides a readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the vehicle engine oil quantity control method as described above.

[0056] In summary, the vehicle engine fuel quantity control method and system provided in the above embodiments of the present invention can objectively and accurately control the fuel output of the vehicle engine, thereby improving fuel control efficiency.

[0057] It should be noted that the above modules can be functional modules or program modules, and can be implemented through software or hardware. For modules implemented through hardware, the above modules can reside in the same processor; or the above modules can be located in different processors in any combination.

[0058] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0059] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0060] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0061] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0062] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. A method for controlling the oil quantity of a vehicle engine, characterized in that, The method includes: During the actual operation of the vehicle engine, the fuel pressure, pipeline temperature, engine speed and target time-series data of the fuel pump output, the engine speed and the command transmission are collected in real time, and the corresponding state parameter matrix is ​​created in real time based on the target time-series data. A corresponding flow response prediction model is constructed based on the target time series data, and the state parameter matrix is ​​input into the flow response prediction model to output the corresponding flow response prediction value generated within the future preset control period. A corresponding two-dimensional compensation matrix is ​​constructed in real time based on the instruction transmission delay data and the predicted flow response value, and an adapted reference flow instruction is generated according to the target operating condition of the vehicle engine. The reference flow command is corrected in time periods by the two-dimensional compensation matrix to generate the corresponding target command sequence in real time, and the target command sequence is converted into the corresponding target control command in real time to complete the real-time output of fuel quantity according to the target control command. The step of constructing the corresponding two-dimensional compensation matrix in real time based on the instruction transmission delay data and the traffic response prediction value includes: The instruction transmission delay data is frequency domain converted to extract the corresponding delay amplitude spectrum features in real time, and the traffic response prediction value is interpolated according to the time resolution of the preset control period to obtain the corresponding prediction value sequence. Using the frequency band division result obtained from the frequency domain transformation as the row dimension of the matrix and the time resolution node after interpolation as the column dimension of the matrix, a corresponding blank two-dimensional matrix is ​​constructed in real time, and the delay amplitude spectrum feature and the predicted value sequence are filled into the corresponding positions of the blank two-dimensional matrix to generate the corresponding intermediate two-dimensional matrix. The elements in the intermediate two-dimensional matrix are dynamically scaled based on the real-time collected fuel pressure fluctuation cycle to form the corresponding two-dimensional compensation matrix.

2. The vehicle engine oil quantity control method according to claim 1, characterized in that: The step of inputting the state parameter matrix into the flow response prediction model to output the corresponding flow response prediction value generated within a future preset control period includes: The sample time series parameters generated at several consecutive sampling times are extracted from the state parameter matrix in real time, and several single-column matrices are constructed according to the parameter types contained in the sample time series parameters. Perform a Fourier transform on each of the single-column matrices to generate the corresponding frequency feature matrix in real time, and extract the amplitude data of the first 20 frequency components in each frequency feature matrix in real time. According to the preset splicing order, an 80-dimensional feature vector is generated in real time based on the amplitude data, and the feature vector is input into the flow response prediction model to output the corresponding flow response prediction value.

3. The vehicle engine oil quantity control method according to claim 2, characterized in that: The step of inputting the feature vector into the traffic response prediction model to output the traffic response prediction value includes: The 80-dimensional feature vector is processed by a sliding window, and divided into 8 overlapping windows with each window consisting of 10 dimensions. The weighted average value of the elements in each overlapping window is then calculated. The weighted average of each overlapping window is multiplied by the feature vector to generate a corresponding enhanced feature vector. The convolutional attention module in the traffic response prediction model is then used to filter each enhanced feature vector to generate a corresponding target feature vector. The target feature vector is converted into corresponding time-series features in real time by using a preset GRU layer and tanh activation function, and the time-series features are mapped by a preset fully connected layer to generate the traffic response prediction value.

4. The vehicle engine oil quantity control method according to claim 1, characterized in that: The step of generating a suitable reference flow command based on the target operating condition of the vehicle engine includes: The characteristic parameters generated by the vehicle engine during the start-up, cruising, and acceleration phases are analyzed, and the start-up, cruising, and acceleration phases are divided into several sub-phases according to their time proportions. The theoretical threshold of fuel flow corresponding to each sub-phase is extracted simultaneously. Using the division results of the sub-stages as matrix row identifiers and the engine speed gradient under the target operating condition as matrix column identifiers, a corresponding flow mapping matrix is ​​constructed, and the theoretical fuel flow threshold of each sub-stage is filled into each matrix unit of the flow mapping matrix. The maximum delay value corresponding to the instruction transmission delay data is collected in real time, and the maximum delay value is used to perform offset correction processing on each matrix unit. The matrix units after correction processing are then connected in a stage sequence to form the reference flow instruction.

5. The vehicle engine oil quantity control method according to claim 1, characterized in that: The step of performing time-segmented correction processing on the baseline flow command using the two-dimensional compensation matrix to generate the corresponding target command sequence in real time includes: The baseline flow command is decomposed into several command segments, and the compensation coefficient corresponding to each command segment is extracted in real time from the two-dimensional compensation matrix. Each instruction segment and its corresponding compensation coefficient are multiplied point by point to obtain the preliminarily corrected instruction subsequence. The instruction subsequence is smoothed according to the rate of change of the vehicle engine speed, and all the smoothed subsequences are spliced ​​together in chronological order to form the target instruction sequence.

6. The vehicle engine oil quantity control method according to claim 5, characterized in that: The step of smoothing the instruction subsequence according to the rate of change of the vehicle engine speed, and then concatenating all the smoothed subsequences in chronological order to form the target instruction sequence includes: The endpoint flow values ​​of two adjacent instruction subsequences and the corresponding engine speed change rate are collected in real time, and the number of sampling points of the transition segment between two adjacent instruction subsequences is determined in real time based on the engine speed change rate. The splicing coefficient of the corresponding transition segment is determined in real time based on the number of sampling points. Each instruction subsequence and each transition segment are connected according to the splicing coefficient to form a continuously changing target instruction sequence.

7. A vehicle engine oil quantity control system, characterized in that, For implementing the vehicle engine oil quantity control method as described in any one of claims 1 to 6, the system comprises: The data acquisition module is used to collect the fuel pressure, pipeline temperature, engine speed and target time-series data of the fuel pump output in real time during the actual operation of the vehicle engine, and to create the corresponding state parameter matrix in real time based on the target time-series data. The construction module is used to construct a corresponding traffic response prediction model based on the target time series data, and input the state parameter matrix into the traffic response prediction model to output the corresponding traffic response prediction value generated in the future preset control period. The processing module is used to construct a corresponding two-dimensional compensation matrix in real time based on the instruction transmission delay data and the predicted flow response value, and to generate an adapted reference flow instruction according to the target operating condition of the vehicle engine. The correction module is used to perform time-segmented correction processing on the reference flow command through the two-dimensional compensation matrix to generate the corresponding target command sequence in real time, and to convert the target command sequence into the corresponding target control command in real time, so as to complete the real-time output of fuel quantity according to the target control command.

8. A computer comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the vehicle engine oil quantity control method as described in any one of claims 1 to 6.

9. A readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the vehicle engine oil quantity control method as described in any one of claims 1 to 6.

Images