Four-stroke engine of agricultural tractor

Abstract

The invention relates to the technical field of engines, in particular to an agricultural tractor four-stroke engine which comprises a variable-compression-ratio engine, a variable-compression-ratio engine and a variable-compression-ratio engine, the compression ratio of the variable-compression-ratio engine is automatically adjusted according to the requirement of hydrogen combustion under different loads and rotating speeds, partial pre-combustion is conducted in an air cylinder in a staged combustion mode, and then more hydrogen is introduced step by step for combustion; the cooling system reduces the temperature generated in the hydrogen combustion process through gas cooling, liquid cooling and turbine cooling, a heat exchange device is designed near the injection system and the combustion chamber, and the overall temperature of the engine is reduced; according to the fuel injection system, the dual injection technology of direct injection and port injection is combined, and the fuel injection mode is optimized under different working conditions. By combining a combustion control strategy, self-adaptive control and predictive control and enhancing real-time fault detection, an intelligent control system of the hydrogen engine can adjust the combustion process more accurately, detonation is avoided, emission is reduced, and power output is optimized under different working conditions.

Description

Technical Field

[0001] The invention relates to the technical field of engines, in particular to a four-stroke engine for an agricultural tractor. Background Art

[0002] Traditional internal combustion engines use carbon-containing fossil fuels (mainly petroleum products) as their energy source, which inevitably produces a large amount of CO2 emissions. Hydrogen energy has many advantages over traditional fossil energy due to its own characteristics. For example, it has no carbon atoms, so it does not produce CO2 and other emissions during the combustion process;

[0003] There are two main types of applications for hydrogen as an automotive energy source, namely fuel cells and hydrogen engines. Fuel cells cannot be used on a large scale at this stage because they require high hydrogen purity and high thermal management technology. Hydrogen engines directly use hydrogen for combustion, which has obvious advantages over fuel cells in terms of hydrogen purity requirements. They can also be modified based on traditional gasoline engines, and the technology is relatively mature.

[0004] Although hydrogen engines have obvious technical advantages, there are also some problems. For example, hydrogen and air are pre-mixed before entering the combustion chamber for combustion. The abnormal combustion phenomenon of fast combustion rate and high combustion temperature can easily lead to serious problems such as knock. Due to the above abnormal combustion phenomenon, existing hydrogen engines need to burn under conditions where the mixed gas is relatively lean, which greatly limits the power performance and application prospects of hydrogen engines.

[0005] Therefore, a four-stroke engine for an agricultural tractor is proposed. Summary of the invention

[0006] The object of the present invention is to provide a four-stroke engine for an agricultural tractor to solve the problems raised in the above background technology.

[0007] To achieve the above object, the present invention provides the following technical solution: a four-stroke engine for an agricultural tractor, comprising:

[0008] The variable compression ratio engine automatically adjusts the compression ratio according to the demand for hydrogen combustion under different loads and speeds. It adopts a staged combustion method, first partially pre-combusting in the cylinder, and then gradually introducing more hydrogen for combustion;

[0009] The cooling system uses gas cooling, liquid cooling, and turbine cooling to reduce the temperature generated during hydrogen combustion. A heat exchange device is designed near the injection system and combustion chamber to reduce the overall temperature of the engine.

[0010] The fuel injection system combines dual injection technology of direct injection and port injection to optimize the fuel injection method under different working conditions. At low load, port injection technology is used to control the uniformity of the mixture; at high load, direct injection technology is used to accurately control the amount and timing of hydrogen injection to ensure the stability and efficiency of combustion. Hydrogen is injected in stages through multiple injection ports. A small amount of hydrogen is injected and ignited in the first stage, and hydrogen is injected in the subsequent stages to maintain a relatively stable combustion process. By controlling the time and amount of hydrogen injection, the combustion rate in the cylinder is avoided to be too fast and the temperature fluctuation is reduced. The variable injection time technology is used to adjust the timing of hydrogen injection under different working conditions to accurately match it with the intake and compression process, that is, under low load conditions, the injection time is delayed to reduce the combustion temperature, and under high load, the injection is carried out in advance to ensure sufficient hydrogen supply;

[0011] The intelligent control system monitors the combustion status of the engine in real time and automatically adjusts the injection timing, injection amount, and combustion chamber pressure parameters.

[0012] Preferably, the intelligent control system also includes combustion control strategy, dynamic parameter adjustment, adaptive control and real-time fault detection.

[0013] Preferably, the combustion control strategy includes combustion rate and combustion temperature control. In a hydrogen engine, the combustion rate Q is closely related to the change of temperature T;

[0014] The formula of burning rate Q and temperature T is:

[0015]

[0016] in: is the heat output power of the combustion process, η is the combustion efficiency, α is the calorific value constant of hydrogen, V f is the gas flow rate in the combustion chamber, p is the gas pressure, T is the instantaneous temperature, and T0 is the ambient temperature;

[0017] The dynamic parameter adjustment includes a dynamic adjustment strategy:

[0018] The temperature change in the cylinder is monitored by real-time temperature feedback temperature sensors and thermocouples. When the temperature rises to the threshold, that is, exceeds 1200°C, measures are taken to adjust the injection timing or injection amount to reduce the generation of NOx caused by excessive temperature.

[0019] Use PID control to optimize injection quantity and timing:

[0020]

[0021] Where u(t) is the control signal, i.e. the adjustment of the injection amount or injection timing, e(t) = T set -T(t) is the error, Tset is the set temperature, T(t) is the real-time temperature.

[0022] Preferably, the adaptive control and predictive control automatically adjust the control parameters to adapt to system changes by continuously estimating and updating the system model. In the hydrogen engine, a strategy based on model reference adaptive control is adopted, and the target system is:

[0023]

[0024] y(t)=Cx(t)+v(t)y(t)=Cx(t)+v(t)

[0025] Where: x(t) is the system state variable, i.e., temperature, pressure, and combustion rate; v(t) is the control input, i.e., injection amount and injection timing; w(t) and v(t) are external disturbances, i.e., load changes and fuel characteristic changes. By comparing the system output y(t) with the output y of the reference model ref (t), and make control adjustments based on the error:

[0026]

[0027] in: is the estimate of the control parameters, Γ is the adaptive gain matrix, and the update speed of the control system, e(t) = y(t) - y ref (t) is the output error.

[0028] Preferably, the predictive control

[0029] Model predictive control is a method that predicts future states and optimizes control inputs based on a system dynamic model. Through future predictions, i.e., temperature and pressure in the next few time steps, the control strategy is adjusted in real time according to certain optimization criteria;

[0030] The dynamic model of the system is:

[0031] x(t)+Bu(t)x(t+1)=A x(t)+B u(t)

[0032] The goal of MPC is to minimize the following cost function:

[0033]

[0034] Among them: (x(t+k) is the state prediction at the future time, x ref (t+k) is the reference state, i.e. the desired temperature and pressure, Q and R are weight matrices, controlling the cost of state deviation and control input, and u(t+k) is the control input at the future time, i.e. the injection amount and injection timing;

[0035] By solving this optimization problem, we can obtain the optimal control input sequence u * (t),u * (t+1),…, and apply the optimal input at each moment.

[0036] Preferably, the real-time fault detection includes a fault detection model. When the fuel injection quantity control fails, the temperature fluctuation of the system will exceed the expected range, and the temperature error threshold is set:

[0037] ΔT=|T(t)-T ref (t)|>ΔT threshold

[0038] When the temperature error exceeds the threshold, the system will determine it as a fault and trigger emergency control measures, that is, adjusting the injection amount or injection timing;

[0039] Fault-tolerant control strategy:

[0040] After a fault is detected, the system's operating mode is adjusted through fault-tolerant control technology. That is, when a problem occurs in the injection system, the stable operation of the engine is ensured by adjusting the injection timing or using a backup injection system.

[0041] Preferably, the fault detection model uses Kalman filtering or particle filtering technology, combined with multiple sensor data, namely pressure, temperature, vibration, flow, to dynamically estimate the system state, thereby improving the reliability and anti-interference ability of fault detection;

[0042] The basic form of Kalman filtering is:

[0043] xk∣k=xk∣k-1+Kk(zk-Hkxk∣k-1)

[0044] Where: xk|k is the estimated system state, Kk is the Kalman gain, which determines the degree of dependence on the measurement value, zk is the sensor measurement value, and Hk is the measurement matrix;

[0045] Model-based fault detection:

[0046] Use the physical model of the engine, i.e. the thermodynamic model and the combustion model, to establish the expected normal working behavior and compare it with the measurement results of the actual system. When the difference between the output of the system and the expected value of the model exceeds the set threshold, it can be judged as a fault;

[0047] The state of the engine is described by the following dynamic equations:

[0048]

[0049] y(t)=h(x(t),t)

[0050] Where x(t) is the system state, u(t) is the control input, y(t) is the observed output, f(·) and h(·) are the system dynamic equation and output equation, respectively;

[0051] By comparing with the actual measured value, the residual is calculated:

[0052] r(t)=y measured When the residual (t)-h(x(t),t) exceeds the threshold, it can be determined as a fault and the system enters the fault protection mode.

[0053] Preferably, the fault-tolerant control strategy optimization includes control switching based on fault mode, dynamically switching different control strategies according to fault detection results, switching to a more conservative injection strategy when a fuel injection system fault is detected; temporarily lowering the compression ratio or delaying the ignition timing when it is found that the temperature is too high to cause potential knocking, and the control switching rules are defined by a discrete state machine, where each state represents a working mode;

[0054] Normal mode: the fuel injection system, ignition timing, and compression ratio parameters are controlled according to normal rules;

[0055] Fuel injection system failure mode: adjust the injection strategy according to the remaining available number of injectors or injection flow to ensure that the combustion process is as smooth as possible;

[0056] Overtemperature mode: adjust the intake system, improve cooling effect or adjust ignition timing to avoid knock;

[0057] Fault reconstruction and state observer:

[0058] When a sensor or actuator fails, the Luenberger observer is used to estimate the system state and perform control.

[0059] Assume the state equation of the system is:

[0060] Observer design: construct a state observer to calculate the state of the system given known inputs and outputs;

[0061]

[0062] Where: L is the observer gain, is the observed state;

[0063] Through state reconstruction, when a sensor fails, the system infers the normal operating state and continues to operate stably;

[0064] Fault-tolerant optimization and optimal control:

[0065] When a fault occurs, the control strategy is optimized through the optimal control method, and the linear quadratic regulator is used to calculate the optimal control input after the fault occurs;

[0066] The goal of the LQR control problem is to minimize the following cost function:

[0067]

[0068] Where: Q and R are weighted matrices of state and control input, x(t) is the system state, and u(t) is the control input;

[0069] When a fault is detected, the state constraints of the system will change, and the LQR controller will recalculate the optimal control input u(t) so that the system can optimize its performance as much as possible under the fault condition.

[0070] Compared with the prior art, the present invention has the following beneficial effects:

[0071] By combining combustion control strategies, adaptive control and predictive control, and enhancing real-time fault detection, the intelligent control system of hydrogen engines can more accurately regulate the combustion process, avoid knock, reduce emissions, and optimize power output under different operating conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0073] Figure 1 is a block diagram of the engine system of the present invention;

[0074] Figure 2 It is a structural view of the four-stroke engine of the present invention. DETAILED DESCRIPTION

[0075] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0076] See also Figure 1 to Figure 2 , the present invention provides a technical solution:

[0077] A four-stroke engine for an agricultural tractor, comprising:

[0078] The variable compression ratio engine automatically adjusts the compression ratio according to the demand for hydrogen combustion under different loads and speeds. It adopts a staged combustion method, first partially pre-combusting in the cylinder, and then gradually introducing more hydrogen for combustion;

[0079] The cooling system uses gas cooling, liquid cooling, and turbine cooling to reduce the temperature generated during hydrogen combustion. A heat exchange device is designed near the injection system and combustion chamber to reduce the overall temperature of the engine.

[0080] The fuel injection system combines dual injection technology of direct injection and port injection to optimize the fuel injection method under different working conditions. At low load, port injection technology is used to control the uniformity of the mixture; at high load, direct injection technology is used to accurately control the amount and timing of hydrogen injection to ensure the stability and efficiency of combustion. Hydrogen is injected in stages through multiple injection ports. A small amount of hydrogen is injected and ignited in the first stage, and hydrogen is injected in the subsequent stages to maintain a relatively stable combustion process. By controlling the time and amount of hydrogen injection, the combustion rate in the cylinder is avoided to be too fast and the temperature fluctuation is reduced. The variable injection time technology is used to adjust the timing of hydrogen injection under different working conditions to accurately match it with the intake and compression process, that is, under low load conditions, the injection time is delayed to reduce the combustion temperature, and under high load, the injection is carried out in advance to ensure sufficient hydrogen supply;

[0081] The intelligent control system monitors the combustion status of the engine in real time and automatically adjusts parameters such as injection timing, injection amount, and combustion chamber pressure. Based on real-time data, the intelligent control system can respond to different working conditions more flexibly, avoid knocking, and optimize combustion efficiency.

[0082] The intelligent control system also includes combustion control strategy, dynamic parameter adjustment, adaptive control and real-time fault detection.

[0083] The combustion control strategy includes combustion rate and combustion temperature control. In a hydrogen engine, the combustion rate Q is closely related to the change of temperature T;

[0084] The formula of burning rate Q and temperature T is:

[0085]

[0086] in: is the heat output power of the combustion process, η is the combustion efficiency, α is the calorific value constant of hydrogen, V f is the gas flow rate in the combustion chamber, p is the gas pressure, T is the instantaneous temperature, and T0 is the ambient temperature;

[0087] The dynamic parameter adjustment includes a dynamic adjustment strategy:

[0088] The temperature change in the cylinder is monitored by real-time temperature feedback temperature sensors and thermocouples. When the temperature rises to the threshold, that is, exceeds 1200°C, measures are taken to adjust the injection timing or injection amount to reduce the generation of NOx caused by excessive temperature.

[0089] Use PID control to optimize injection quantity and timing:

[0090]

[0091] Where u(t) is the control signal, i.e. the adjustment of the injection amount or injection timing, e(t) = T set -T(t) is the error, T set is the set temperature, T(t) is the real-time temperature.

[0092] The adaptive control and predictive control automatically adjust the control parameters to adapt to system changes by continuously estimating and updating the system model. In the hydrogen engine, the strategy of model reference adaptive control is adopted, and the target system is:

[0093]

[0094] Where: x(t) is the system state variable, i.e., temperature, pressure, and combustion rate; v(t) is the control input, i.e., injection amount and injection timing; w(t) and v(t) are external disturbances, i.e., load changes and fuel characteristic changes. By comparing the system output y(t) with the output y of the reference model ref (t), and make control adjustments based on the error:

[0095]

[0096] in: is the estimate of the control parameters, Γ is the adaptive gain matrix, and the update speed of the control system, e(t) = y(t) - y ref (t) is the output error.

[0097] The predictive control:

[0098] Model predictive control is a method that predicts future states and optimizes control inputs based on a system dynamic model. Through future predictions, i.e., temperature and pressure in the next few time steps, the control strategy is adjusted in real time according to certain optimization criteria;

[0099] The dynamic model of the system is:

[0100] x(t)+Bu(t)x(t+1)=A x(t)+B u(t)

[0101] The goal of MPC is to minimize the following cost function:

[0102]

[0103] Among them: (x(t+k) is the state prediction at the future time, x ref (t+k) is the reference state, i.e. the desired temperature and pressure, Q and R are weight matrices, controlling the cost of state deviation and control input, and u(t+k) is the control input at the future moment, i.e. the injection amount and injection timing.

[0104] By solving this optimization problem, we can obtain the optimal control input sequence u * (t),u * (t+1),…, and apply the optimal input at each moment.

[0105] The real-time fault detection and fault-tolerant control is because hydrogen engines work under high temperature and high pressure environment, and may have faults (such as nozzle blockage, temperature sensor failure, etc.), which will affect the combustion process. Therefore, it is very important to design a real-time fault detection system.

[0106] Fault Detection Model:

[0107] Assume that after a component in the engine fails, the system output will change abnormally. The fault can be detected by monitoring the thermal state of the engine (such as temperature, pressure). For example, when the injection quantity control fails, the temperature fluctuation of the system will exceed the expected range. Set the temperature error threshold:

[0108] ΔT=|T(t)-T ref (t)|>ΔT threshold

[0109] When the temperature error exceeds the threshold, the system will determine it as a fault and trigger emergency control measures (such as adjusting the injection amount or injection timing).

[0110] Fault-tolerant control strategy:

[0111] After a fault is detected, the system's operating mode can be adjusted through fault-tolerant control technology. For example, when a problem occurs in the injection system, the engine can be kept stable by adjusting the injection timing or using a backup injection system.

[0112] The fault detection model adopts Kalman filtering or particle filtering technology, combined with multiple sensor data (such as pressure, temperature, vibration, flow, etc.), to dynamically estimate the system state, thereby improving the reliability and anti-interference ability of fault detection.

[0113] The basic form of Kalman filtering is:

[0114] xk∣k=xk∣k-1+Kk(zk-Hkxk∣k-1)

[0115] Where: xk|k is the estimated system state, Kk is the Kalman gain, which determines the degree of dependence on the measurement value, zk is the sensor measurement value, and Hk is the measurement matrix.

[0116] Model-based fault detection:

[0117] Use the physical model of the engine (such as thermodynamic model, combustion model, etc.) to establish the expected normal working behavior and compare it with the measurement results of the actual system. When the difference between the output of the system and the expected value of the model exceeds the set threshold, it can be judged as a fault.

[0118] Assume that the state of the engine is described by the following dynamic equation:

[0119]

[0120] y(t)=h(x(t),t)

[0121] Among them, x(t) is the system state, u(t) is the control input, y(t) is the observed output, f(·) and h(·) are the system dynamic equation and output equation, respectively.

[0122] By comparing with the actual measured value, the residual is calculated:

[0123] r(t)=y measured When the residual (t)-h(x(t),t) exceeds the threshold, it can be determined as a fault and the system enters the fault protection mode.

[0124] Fault Detection with Machine Learning and Deep Learning:

[0125] In order to improve the flexibility and accuracy of fault diagnosis, fault detection methods based on supervised learning and deep learning can be used. By collecting historical data (such as temperature, pressure, vibration, etc.), classifiers (such as support vector machines (SVM) or convolutional neural networks (CNN)) are trained to identify different fault modes.

[0126] For example, using a convolutional neural network to analyze the vibration signal of an engine can capture subtle changes when a fault occurs. The fault category can be predicted by the output of the network, such as knock, injector failure, etc.

[0127] The optimization goal of the fault-tolerant control strategy is to ensure that the engine can quickly recover to an acceptable working state when a fault occurs, minimizing the impact of the fault on performance. The following are several optimization strategies:

[0128] Control switching based on failure mode:

[0129] Depending on the fault detection results, different control strategies can be switched dynamically. For example, if a fuel injection system fault is detected, a more conservative fuel injection strategy can be switched; if the temperature is too high, resulting in potential knock, the compression ratio can be temporarily lowered or the ignition timing can be delayed.

[0130] The control switching rules can be defined by a discrete state machine (FSM), where each state represents an operating mode. For example:

[0131] Normal mode: The injection system, ignition timing, compression ratio and other parameters are controlled according to normal rules.

[0132] Fuel injection system failure mode: The fuel injection strategy is adjusted according to the remaining number of available injectors or the injection flow rate to ensure that the combustion process is as smooth as possible.

[0133] Overtemperature mode: adjust the intake system, improve cooling effect or adjust ignition timing to avoid detonation.

[0134] Fault reconstruction and state observer:

[0135] When certain sensors or actuators fail, a state observer (such as the Luenberger observer) can be used to infer the system state and perform control.

[0136] Assume the state equation of the system is:

[0137] Observer Design: Construct a state observer to calculate the state of the system given known inputs and outputs.

[0138] Where LL is the observer gain, is the observed state.

[0139] Through state reconstruction, even if a sensor fails, the system can still infer the normal operating state and continue to operate stably.

[0140] Fault-tolerant optimization and optimal control:

[0141] When a fault occurs, the control strategy can be optimized through optimal control methods. For example, a linear quadratic regulator (LQR) is used to calculate the optimal control input after the fault occurs, so that the system can still maintain the best performance within a certain range under the fault condition.

[0142] The goal of the LQR control problem is to minimize the following cost function:

[0143]

[0144] Where: Q and R are weighted matrices of state and control input, x(t) is the system state, and u(t) is the control input.

[0145] When a fault is detected, the state constraints of the system will change, and the LQR controller will recalculate the optimal control input u(t) so that the system can optimize its performance as much as possible under the fault condition.

[0146] Distributed fault-tolerant control:

[0147] In a multi-engine or multi-module system, a distributed fault-tolerant control method can be used to distribute tasks using a networked control architecture. Each module or engine unit has its own controller, and when a unit fails, other units can compensate by adjusting the load or changing the control strategy.

[0148] For example, when an aircraft uses hydrogen engines, if an engine unit fails, the remaining workload can be distributed to other healthy engine units through load sharing to ensure that the overall performance is not greatly affected.

[0149] Fault detection and fault-tolerant control system integration:

[0150] Integrating fault detection and fault-tolerant control strategies into a closed-loop system enables the entire system to sense, diagnose and respond in real time when a fault occurs. The workflow of the integrated framework is as follows:

[0151] Real-time monitoring and data acquisition: Collect engine status data such as temperature, pressure, vibration, etc.

[0152] Fault detection and diagnosis: Detect faults in real time using multi-sensor data fusion, model comparison, and machine learning models.

[0153] Fault response and control adjustment: Based on the detection results, switch to the appropriate control strategy and use optimized control methods to adjust operations.

[0154] Continuous feedback and learning: The system updates the control strategy in real time and continuously optimizes the control effect through the feedback mechanism to ensure the best system performance after a fault.

[0155] Finally, 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 aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A four-stroke engine for an agricultural tractor, characterized in that: include: The variable compression ratio engine automatically adjusts the compression ratio according to the demand for hydrogen combustion under different loads and speeds. It adopts a staged combustion method, first partially pre-combusting in the cylinder, and then gradually introducing more hydrogen for combustion; The cooling system uses gas cooling, liquid cooling, and turbine cooling to reduce the temperature generated during hydrogen combustion. A heat exchange device is designed near the injection system and combustion chamber to reduce the overall temperature of the engine. The fuel injection system combines dual injection technology of direct injection and port injection to optimize the fuel injection method under different working conditions. At low load, port injection technology is used to control the uniformity of the mixture; at high load, direct injection technology is used to accurately control the amount and timing of hydrogen injection to ensure the stability and efficiency of combustion. Hydrogen is injected in stages through multiple injection ports. A small amount of hydrogen is injected and ignited in the first stage, and hydrogen is injected in the subsequent stages to maintain a relatively stable combustion process. By controlling the time and amount of hydrogen injection, the combustion rate in the cylinder is avoided to be too fast and the temperature fluctuation is reduced. The variable injection time technology is used to adjust the timing of hydrogen injection under different working conditions to accurately match it with the intake and compression process, that is, under low load conditions, the injection time is delayed to reduce the combustion temperature, and under high load, the injection is carried out in advance to ensure sufficient hydrogen supply; The intelligent control system monitors the combustion status of the engine in real time and automatically adjusts the injection timing, injection amount, and combustion chamber pressure parameters.

2. A four-stroke engine for agricultural tractors according to claim 1, characterized in that: The intelligent control system also includes combustion control strategy, dynamic parameter adjustment, adaptive control and real-time fault detection.

3. A four-stroke engine for agricultural tractors according to claim 2, characterized in that: The combustion control strategy includes combustion rate and combustion temperature control. In a hydrogen engine, the combustion rate Q is closely related to the change of temperature T; The formula of burning rate Q and temperature T is: in: is the heat output power of the combustion process, η is the combustion efficiency, α is the calorific value constant of hydrogen, V f is the gas flow rate in the combustion chamber, p is the gas pressure, T is the instantaneous temperature, and T0 is the ambient temperature; The dynamic parameter adjustment includes a dynamic adjustment strategy: The temperature change in the cylinder is monitored by real-time temperature feedback temperature sensors and thermocouples. When the temperature rises to the threshold, that is, exceeds 1200°C, measures are taken to adjust the injection timing or injection amount to reduce the generation of NOx caused by excessive temperature. Use PID control to optimize injection quantity and timing: Where u(t) is the control signal, i.e. the adjustment of the injection amount or injection timing, e(t) = T set -T(t) is the error, T set is the set temperature, T(t) is the real-time temperature.

4. A four-stroke engine for agricultural tractors according to claim 3, characterized in that: The adaptive control and predictive control automatically adjust the control parameters to adapt to system changes by continuously estimating and updating the system model. In the hydrogen engine, the strategy of model reference adaptive control is adopted, and the target system is: y(t)=Cx(t)+v(t)y(t)=Cx(t)+v(t) Where: x(t) is the system state variable, i.e., temperature, pressure, and combustion rate; v(t) is the control input, i.e., injection amount and injection timing; w(t) and v(t) are external disturbances, i.e., load changes and fuel characteristic changes. By comparing the system output y(t) with the output y of the reference model ref (t), and make control adjustments based on the error: in: is the estimate of the control parameters, Γ is the adaptive gain matrix, and the update speed of the control system, e(t) = y(t) - y ref (t) is the output error.

5. A four-stroke engine for agricultural tractors according to claim 4, characterized in that: The predictive control adopts model predictive control, which is a method for predicting future states and optimizing control inputs based on a system dynamic model. Through future predictions, i.e., temperature and pressure in the next few time steps, the control strategy is adjusted in real time according to certain optimization criteria; The dynamic model of the system is: x(t)+Bu(t)x(t+1)=A x(t)+B u(t) The goal of MPC is to minimize the following cost function: Among them: (x(t+k) is the state prediction at the future time, x ref (t+k) is the reference state, i.e. the desired temperature and pressure, Q and R are weight matrices, controlling the cost of state deviation and control input, and u(t+k) is the control input at the future time, i.e. the injection amount and injection timing; By solving this optimization problem, we can obtain the optimal control input sequence u * (t),u * (t+1),…, and apply the optimal input at each moment.

6. A four-stroke engine for agricultural tractors according to claim 5, characterized in that: The real-time fault detection includes a fault detection model. When the fuel injection quantity control fails, the temperature fluctuation of the system will exceed the expected range. The temperature error threshold is set: ΔT=|T(t)-T ref (t)|>ΔT threshold When the temperature error exceeds the threshold, the system will determine it as a fault and trigger emergency control measures, that is, adjusting the injection amount or injection timing; Fault-tolerant control strategy: After a fault is detected, the system's operating mode is adjusted through fault-tolerant control technology. That is, when a problem occurs in the injection system, the stable operation of the engine is ensured by adjusting the injection timing or using a backup injection system.

7. A four-stroke engine for agricultural tractors according to claim 6, characterized in that: The fault detection model uses Kalman filtering or particle filtering technology, combined with multiple sensor data, namely pressure, temperature, vibration, flow, to dynamically estimate the system state, thereby improving the reliability and anti-interference ability of fault detection; The basic form of Kalman filtering is: xk∣k=xk∣k-1+Kk(zk-Hkxk∣k-1) Where: xk|k is the estimated system state, Kk is the Kalman gain, which determines the degree of dependence on the measurement value, zk is the sensor measurement value, and Hk is the measurement matrix; Model-based fault detection: Use the physical model of the engine, i.e. the thermodynamic model and the combustion model, to establish the expected normal working behavior and compare it with the measurement results of the actual system. When the difference between the output of the system and the expected value of the model exceeds the set threshold, it can be judged as a fault; The state of the engine is described by the following dynamic equations: y(t)=h(x(t),t) Where x(t) is the system state, u(t) is the control input, y(t) is the observed output, f(·) and h(·) are the system dynamic equation and output equation, respectively; By comparing with the actual measured value, the residual is calculated: r(t)=y measured When the residual (t)-h(x(t),t) exceeds the threshold, it can be determined as a fault and the system enters the fault protection mode.

8. The four-stroke engine for agricultural tractor according to claim 7, characterized in that: The fault-tolerant control strategy optimization includes control switching based on fault mode, dynamically switching different control strategies according to the fault detection result, and switching to a more conservative injection strategy when a fuel injection system fault is detected; When it is found that the temperature is too high, causing potential detonation, the compression ratio is temporarily lowered or the ignition timing is delayed. The control switching rules are defined by a discrete state machine, where each state represents an operating mode; Normal mode: the fuel injection system, ignition timing, and compression ratio parameters are controlled according to normal rules; Fuel injection system failure mode: adjust the injection strategy according to the remaining available number of injectors or injection flow to ensure that the combustion process is as smooth as possible; Overtemperature mode: adjust the intake system, improve cooling effect or adjust ignition timing to avoid knock; Fault reconstruction and state observer: When a sensor or actuator fails, the Luenberger observer is used to estimate the system state and perform control. Assume the state equation of the system is: Observer design: construct a state observer to calculate the state of the system given known inputs and outputs; Where: L is the observer gain, is the observed state; Through state reconstruction, when a sensor fails, the system infers the normal operating state and continues to operate stably; Fault-tolerant optimization and optimal control: When a fault occurs, the control strategy is optimized through the optimal control method, and the linear quadratic regulator is used to calculate the optimal control input after the fault occurs; The goal of the LQR control problem is to minimize the following cost function: Where: Q and R are weighted matrices of state and control input, x(t) is the system state, and u(t) is the control input; When a fault is detected, the state constraints of the system will change, and the LQR controller will recalculate the optimal control input u(t) so that the system can optimize its performance as much as possible under the fault condition.

Images