An adaptive hydrogen injection control method for a hydrogen energy heat engine for a vehicle

Abstract

The present invention discloses an adaptive hydrogen injection control method for a vehicle hydrogen energy heat engine. The method collects the gas pressure in the intake pipeline at all times to form and divide a data set, constructs a detection model and trains to obtain a training model, adjusts the operating parameters of a pressure reducing valve and a boost pump through the training model, controls their operation through the adjusted operating parameters, senses the system injection conditions in real time, and realizes adaptive detection and control to a certain extent by optimizing the training model and the detection model, thus solving the problem that there is no method for sensing the real-time conditions of the injection system to adaptively realize a reasonable hydrogen injection control method at this stage.

Description

Technical Field

[0001] The invention relates to the technical field of engine improvement, and in particular to an adaptive hydrogen injection control method for a vehicle hydrogen energy heat engine. Background Art

[0002] As a carbon-free clean energy, hydrogen only produces water when burned, which can achieve near-zero emissions of carbon and pollutants. The use of hydrogen fuel engines can promote the process of carbon neutrality. At present, hydrogen fuel engines at home and abroad are in the research stage. Most of them use a hydrogen supply system with airway injection, which is characterized by low hydrogen supply pressure, usually 5 bar, corresponding to high hydrogen bottle utilization rate and long cruising range; the disadvantages are low power density, low thermal efficiency, and high risk of backfire. The main reason is that the hydrogen density is small, and airway injection occupies a large amount of intake volume, which makes the charging efficiency low, resulting in low power density and low efficiency. In addition, the hydrogen density in the intake duct is high. Once there is a hot spot or the gas in the cylinder flows back to the intake duct during the intake stage, it will cause the hydrogen in the intake duct to burn or cause backfire explosion, which will cause serious consequences of structural damage.

[0003] At present, in order to ensure the safety of hydrogen flow in the intake duct, there are generally two overall solutions. One is to detect abnormal conditions in the intake duct and replace the corresponding components when abnormalities are detected in the intake duct. The other is to consider controlling the opening and closing degree and timing of the pressure reducing valve through a program to control the hydrogen injection state to ensure the safe operation of the injection system. At present, there is no control method for the injection system to sense real-time conditions and adaptively achieve reasonable hydrogen injection. Summary of the invention

[0004] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0005] In view of the above-mentioned problems existing in the operation of the existing hydrogen energy thermal engine injection system, the present invention is proposed.

[0006] Therefore, the technical problem solved by the present invention is to solve the problem that there is no method for sensing real-time conditions of the injection system to adaptively realize reasonable hydrogen injection control at this stage.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions: a method for adaptive hydrogen injection control of a hydrogen energy heat engine for a vehicle, the control method is controlled by a hydrogen injection control system, the hydrogen injection control system comprises a hydrogen cylinder, an air intake pipeline, a pressure sensor, a pressure reducing valve, a gas booster pump, an internal combustion engine and an MCU computing control unit; wherein the hydrogen cylinder and the internal combustion engine are connected and operated through the air intake pipeline, the pressure reducing valve is provided at the connection between the hydrogen cylinder and the air intake pipeline, the hydrogen cylinder is connected to an external gas supply device, the gas booster pump is provided at the connection port, the pressure sensor and the MCU computing control unit are connected by electrical signals and are integrated on the inner wall of the air intake pipeline, the pressure reducing valve and the gas booster pump are connected by signals to the MCU computing control unit and are controlled and operated by the MCU computing control unit;

[0008] The control method specifically includes the following steps: S1: the pressure sensor monitors the gas pressure in the intake pipeline in real time, and sequentially stores the gas pressure values ​​at each moment to form a data set; S2: the data set is transmitted to the MCU computing control unit for division, and a detection model is constructed; S3: the MCU computing control unit trains the detection model to obtain a training model; S4: the operating parameters of the pressure reducing valve and the boost pump are adjusted through the training model; S5: the MCU computing control unit controls the operation of the pressure reducing valve and the boost pump through the adjusted operating parameters.

[0009] As a preferred solution of the adaptive hydrogen injection control method for a vehicle hydrogen energy heat engine described in the present invention, a nano-promotion layer is arranged on the inner wall of the intake pipeline, and the nano-promotion layer includes a nano-powder promoter, a binder, an expansion coefficient regulator and a mixing aid.

[0010] As a preferred solution of the adaptive hydrogen injection control method for a vehicle hydrogen energy heat engine described in the present invention, the time is an interval of 0.1 ms.

[0011] As a preferred scheme of the adaptive hydrogen injection control method for a vehicle hydrogen energy heat engine described in the present invention, wherein: dividing the data set and constructing the detection model include: dividing the data set according to a ratio of 6:2:2, and naming the divided data sets as a training set, a validation set, and a test set respectively; training the training set, the validation set, and the test set to obtain the detection model.

[0012] As a preferred scheme of the adaptive hydrogen injection control method for a vehicle hydrogen energy heat engine described in the present invention, wherein: training and obtaining the detection model specifically includes: using the YOLOV3 method to detect each time point in the training set whose pressure variation range exceeds the threshold compared with the previous time point; counting the detected time points and their pressure information to form the training set after detection; using cross entropy as the loss function to train the training set, the validation set and the test set after detection to obtain the detection model.

[0013] As a preferred solution of the adaptive hydrogen injection control method for a vehicle hydrogen energy heat engine described in the present invention, the threshold value of the pressure variation range is 0.1N to 0.3N.

[0014] As a preferred solution of the adaptive hydrogen injection control method for a vehicle hydrogen energy heat engine described in the present invention, wherein: training the detection model, obtaining the training model includes configuring relevant training parameters using the source code in gluon-cn; training the detection model again in combination with code running; and obtaining the training model.

[0015] As a preferred solution of the adaptive hydrogen injection control method for a vehicle hydrogen energy heat engine described in the present invention, it includes: using the test set to test whether the training model reaches relevant indicators, and judging whether the training model is in a converged state; if the loss decrease value and the accuracy increase value of the training model during training are significantly reduced, the training model converges; if the test accuracy of the training model on the test set reaches more than 90%, the condition is met; if the test accuracy of the training model on the test set cannot reach more than 90%, continue to expand the data set, and adjust the relevant parameters to train again until the condition is met.

[0016] Beneficial effects of the present invention: The present invention provides an adaptive hydrogen injection control method for a hydrogen energy heat engine for a vehicle, which collects the gas pressure in the intake pipeline at all times, forms a data set and divides it, constructs a detection model and trains to obtain a training model, adjusts the operating parameters of the pressure reducing valve and the boost pump through the training model, controls their operation through the adjusted operating parameters, senses the system injection conditions in real time, and realizes adaptive detection and control to a certain extent by optimizing the training model and the detection model, thereby solving the problem that there is currently no method for sensing the real-time conditions of the injection system to adaptively realize a reasonable hydrogen injection control method. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. Among them:

[0018] Figure 1 This is an overall method flow chart of the adaptive hydrogen injection control method for a vehicle hydrogen energy heat engine provided by the present invention.

[0019] Figure 2 This is a component connection diagram of the adaptive hydrogen injection control system for a hydrogen energy heat engine for a vehicle provided by the present invention.

[0020] Figure 3 A flow chart of the method for training and obtaining a detection model provided by the present invention. DETAILED DESCRIPTION

[0021] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0022] At present, in order to ensure the safety of hydrogen flow in the intake duct, there are generally two overall solutions. One is to detect abnormal conditions in the intake duct and replace the corresponding components when abnormalities are detected in the intake duct. The other is to consider controlling the opening and closing degree and timing of the pressure reducing valve through a program to control the hydrogen injection state to ensure the safe operation of the injection system. At present, there is no control method for the injection system to sense real-time conditions and adaptively achieve reasonable hydrogen injection.

[0023] Therefore, please refer to Figure 1 and Figure 2 The present invention provides an adaptive hydrogen injection control method for a hydrogen energy heat engine for a vehicle, the control method is controlled by a hydrogen injection control system, the hydrogen injection control system includes a hydrogen cylinder, an air intake pipeline, a pressure sensor, a pressure reducing valve, a gas booster pump, an internal combustion engine and an MCU computing control unit;

[0024] The hydrogen cylinder and the internal combustion engine are connected and operated through an intake pipeline, a pressure reducing valve is provided at the connection point between the hydrogen cylinder and the intake pipeline, the hydrogen cylinder is connected to an external gas supply device, a gas booster pump is provided at the connection port, the pressure sensor and the MCU computing control unit are connected by electrical signals and are integrated and arranged on the inner wall of the intake pipeline, the pressure reducing valve and the gas booster pump are connected by signals with the MCU computing control unit and are controlled and operated by the MCU computing control unit;

[0025] The control method specifically includes the following steps:

[0026] S1: The pressure sensor monitors the gas pressure in the air intake line in real time and sequentially stores the gas pressure values ​​at each moment to form a data set;

[0027] The time interval is 0.1ms.

[0028] S2: The data set is transferred to the MCU computing control unit for division and the detection model is constructed;

[0029] Furthermore, dividing the data set and building the detection model specifically include:

[0030] The data set is divided into two parts according to the ratio of 6:2:2, and the divided data sets are named training set, validation set, and test set respectively;

[0031] Train the training set, validation set, and test set to obtain the detection model.

[0032] For further information, see Figure 3 , training and obtaining the detection model specifically includes:

[0033] S2.1: Use the YOLOV3 method to detect each time point in the training set where the pressure change range exceeds the threshold compared with the previous time point;

[0034] Among them, the threshold value of the pressure variation range is 0.1N~0.3N.

[0035] S2.2: Count the detected time points and their pressure information to form a training set after detection;

[0036] S2.3: Use cross entropy as the loss function to train the training set, validation set and test set after detection to obtain the detection model.

[0037] It should be noted that the YOLOV3 method - the target detection method and loss function are all creative applications of existing technologies. As for the technology itself, no unnecessary elaboration will be given here.

[0038] S3: The MCU calculation control unit trains the detection model to obtain the training model;

[0039] Further, the detection model is trained, and the training model is obtained including:

[0040] Use the source code in gluon-cn to configure relevant training parameters;

[0041] Combined with the code run, train the detection model again;

[0042] Get the trained model.

[0043] Furthermore, it also includes:

[0044] Use the test set to test whether the training model has reached relevant indicators and determine whether the training model is in a convergence state;

[0045] If the loss decrease value and accuracy increase value of the training model are significantly reduced during training, the training model converges;

[0046] If the test accuracy of the training model on the test set reaches more than 90%, the condition is met;

[0047] If the test accuracy of the training model on the test set cannot reach more than 90%, continue to expand the data set and adjust the relevant parameters for training again until the conditions are met.

[0048] It should be noted that gluon-cn is used to configure relevant training parameters, which is a creative application of existing technologies. As for the technology itself, no unnecessary elaboration is made here.

[0049] S4: Adjust the operating parameters of the pressure reducing valve and the booster pump through the training model;

[0050] It should be noted that: when the training model is in a non-convergent state, the pressure reducing valve is controlled to be fully closed, and no requirements are made for the booster pump; when the training model is in a convergent state, when the test accuracy reaches more than 90%, the device operates normally. When the test accuracy does not reach more than 90%, the data set continues to be expanded. At the same time, the pressure reducing valve is opened and the booster pump is closed. The degree of opening of the pressure reducing valve is the calculated accuracy ratio. For example, if the test accuracy is 70%, the pressure reducing valve is opened to 70% of the valve body closing degree, and the booster pump is closed at the same time.

[0051] S5: The MCU calculation control unit controls the operation of the pressure reducing valve and the boosting pump through the adjusted operating parameters.

[0052] Additionally, a nano-promotion layer is provided on the inner peripheral wall of the air intake line, and the nano-promotion layer includes a nano-powder promoter, a binder, an expansion coefficient regulator and a mixing aid.

[0053] The ratio of nano powder accelerator: binder: expansion coefficient regulator: mixing aid (calculated by weight ratio) is 80:6:12:20.

[0054] Among them, the nano powder promoter includes 30% nano silicon carbide powder, 10% nano copper oxide, 5% zirconium oxide powder, 45% chromium oxide powder and 10% silicon powder, wherein the particle size of the nano silicon carbide powder is 100-200nm; the particle size of the nano copper oxide is 200-300nm; the particle size of the zirconium oxide powder and the chromium oxide powder can be 20μm-500μm; the particle size of the silicon powder is less than 4.5μm, specifically 2.0μm-4.5μm.

[0055] The binder is aluminum dihydrogen phosphate; the expansion coefficient regulator is cordierite powder and cobalt chromate spinel powder, and the weight ratio of the cordierite powder to the cobalt chromate spinel powder is 1:1.2.

[0056] The mixing aids are dispersants and defoamers, and the ratio of dispersants to defoamers is 4:6. The defoamer is a mixture of one or more of mineral oil, silicone or modified paraffin, and the dispersant is one or more of sodium hexametaphosphate, sodium dodecylbenzene sulfonate or anionic polymer salt. The dispersant and defoamer can effectively ensure the mixing of the above-mentioned nano powder promoter.

[0057] The present invention adopts nano silicon carbide powder and nano copper oxide powder to realize the promotion and supercharging mixing of gas fuel in the intake pipeline through the promotion hole, wherein silicon carbide, zirconium oxide powder and chromium oxide powder constitute the first layer of supercharging promotion, nano copper oxide and silicon powder constitute the second layer of supercharging promotion, and the two layers of supercharging promotion are superimposed on each other, and the supercharging promotion specifically irradiates the gas fuel in the intake pipeline, and the energy is diffused in the form of electromagnetic waves, and the nano promotion layer removes other outer shells on the side relative to the promotion channel, and the inner side is a silver-plated fiber structure, which effectively gathers energy on the side of the nano promotion layer relative to the intake pipeline, thereby increasing the activity rate of gas molecules as a whole, improving the activity of the gas fuel, and making it stably and quickly burn and release energy.

[0058] The present invention provides an adaptive hydrogen injection control method for a vehicle hydrogen energy heat engine, which collects the gas pressure in the intake pipeline at all times, forms a data set and divides it, constructs a detection model and trains to obtain a training model, adjusts the operating parameters of a pressure reducing valve and a boost pump through the training model, controls their operation through the adjusted operating parameters, senses the system injection conditions in real time, and realizes adaptive detection and control to a certain extent by optimizing the training model and the detection model, thereby solving the problem that there is no method for sensing the real-time conditions of the injection system to adaptively realize a reasonable hydrogen injection control method at this stage.

[0059] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for adaptive hydrogen injection control of a hydrogen energy heat engine for a vehicle, characterized in that: The control method is controlled by a hydrogen injection control system, which includes a hydrogen cylinder, an air intake pipeline, a pressure sensor, a pressure reducing valve, a gas booster pump, an internal combustion engine and an MCU computing control unit; The hydrogen cylinder and the internal combustion engine are connected and operated through the air intake pipeline, the pressure reducing valve is arranged at the connection point between the hydrogen cylinder and the air intake pipeline, the hydrogen cylinder is connected to an external gas supply device, and the gas booster pump is arranged at its connection port, the pressure sensor and the MCU computing control unit are connected by electrical signals and are integrated on the inner wall of the air intake pipeline, the pressure reducing valve and the gas booster pump are connected by signals with the MCU computing control unit and are controlled and operated by the MCU computing control unit; The control method specifically comprises the following steps: S1: The pressure sensor monitors the gas pressure in the air intake line in real time and sequentially stores the gas pressure values ​​at each moment to form a data set; S2: transmitting the data set to the MCU computing control unit for division and building a detection model; S3: The MCU calculation control unit trains the detection model to obtain a training model; S4: adjusting the operating parameters of the pressure reducing valve and the boosting pump through the training model; S5: the MCU calculation control unit controls the operation of the pressure reducing valve and the boosting pump according to the adjusted operating parameters; Dividing the data set and constructing the detection model includes: The data set is divided into a training set, a validation set, and a test set according to a ratio of 6:2:

2. The training set, the validation set, and the test set are trained to obtain the detection model.

2. The method for adaptive hydrogen injection control of a hydrogen energy heat engine for a vehicle according to claim 1, characterized in that: A nanometer promotion layer is arranged on the inner peripheral wall of the air intake pipeline, and the nanometer promotion layer comprises a nanometer powder promoter, a binder, an expansion coefficient regulator and a mixing auxiliary agent.

3. The method for adaptive hydrogen injection control of a hydrogen energy heat engine for a vehicle according to claim 2, characterized in that: The time instants are at intervals of 0.1 ms.

4. The method for adaptive hydrogen injection control of a hydrogen energy heat engine for a vehicle according to claim 1, characterized in that: Training and obtaining the detection model specifically includes: Using the YOLOV3 method to detect each time point in the training set where the pressure variation range exceeds the threshold compared with the previous time point; Counting the detected time points and their pressure information to form the training set after detection; The cross entropy is used as a loss function to train the training set, the validation set and the test set after detection to obtain the detection model.

5. The method for adaptive hydrogen injection control of a hydrogen energy heat engine for a vehicle according to claim 4, characterized in that: The threshold of the pressure variation range is 0.1N to 0.3N.

6. The method for adaptive hydrogen injection control of a hydrogen energy heat engine for a vehicle according to claim 5, characterized in that: Training the detection model, obtaining the training model comprises: Use the source code in gluon-cn to configure relevant training parameters; Combined with the code running, the detection model is trained again; The training model is obtained.

7. The method for adaptive hydrogen injection control of a hydrogen energy heat engine for a vehicle according to claim 6, characterized in that: Also includes, Using the test set to test whether the training model reaches relevant indicators, and to determine whether the training model is in a convergence state; If the loss decrease value and the accuracy increase value of the training model during training are significantly reduced, the training model converges; If the test accuracy of the training model on the test set reaches 90% or more, the condition is met; If the test accuracy of the training model on the test set cannot reach more than 90%, continue to expand the data set and adjust relevant parameters to train again until the conditions are met.

Images