A fuel control method for a methanol-diesel dual fuel engine

Abstract

The present application relates to dual-fuel engine technical field, especially to a kind of methanol-diesel dual-fuel engine's fuel control method, comprising: obtaining the real-time data of engine speed value, load value and temperature value;Real-time data is input into the engine operating condition model established in advance, and the fuel demand characteristic parameter of current operating condition is output;Fuel demand characteristic parameter is input into the injection ratio model established in advance, and the injection ratio value of diesel and methanol is output;Based on injection ratio value, generate the fuel injection scheme of diesel and methanol;According to fuel injection scheme, time sequence control algorithm is used to generate diesel injection command and methanol injection command with time phase difference, and the mixing effect of diesel and methanol is monitored in real time;Not only can the injection ratio of diesel and methanol be accurately adjusted, but also the sufficient mixing of two kinds of fuels can be ensured through real-time monitoring, the fuel utilization efficiency is improved, the emission is reduced, and the stable operation and performance output of engine are ensured.

Description

Technical Field

[0001] This invention relates to the field of dual-fuel engine technology, and more particularly to a fuel control method for a methanol-diesel dual-fuel engine. Background Technology

[0002] In dual-fuel engine systems, the injection control of diesel and methanol is a critical technical aspect. Diesel, as the primary fuel, ignites the air-fuel mixture, while methanol, as an auxiliary fuel, optimizes the combustion process and reduces emissions. However, in current systems, diesel and methanol injection are managed by separate control units (ECUs), leading to insufficient coordination between the two fuels. Specifically, the injection timing and quantity of diesel cannot be dynamically matched with those of methanol. For instance, when engine operating conditions change, the diesel injection control unit adjusts its injection strategy based on load and speed, while the methanol injection control unit may respond with lag or lead, resulting in an incoordination in the timing and proportion of the two fuels entering the cylinder. This incoordination directly affects the quality of the air-fuel mixture and consequently, combustion stability.

[0003] Therefore, how to achieve dynamic coordination of the two fuel injections is a technical challenge that urgently needs to be solved in current dual-fuel engine systems. Summary of the Invention

[0004] The purpose of this invention is to overcome the above-mentioned shortcomings and provide a fuel control method for a methanol-diesel dual-fuel engine. This method can not only accurately adjust the injection ratio of diesel and methanol, but also ensure the full mixing of the two fuels through real-time monitoring, thereby improving fuel utilization efficiency, reducing emissions, and ensuring stable engine operation and performance output.

[0005] To achieve the above objectives, the specific solution of the present invention is as follows: a fuel control method for a methanol-diesel dual-fuel engine, specifically comprising: acquiring real-time data of engine speed, load, and temperature; inputting the real-time data into a pre-established engine operating condition model; and outputting fuel demand characteristic parameters of the current operating condition.

[0006] Input the fuel demand characteristic parameters into the pre-established injection ratio model and output the injection ratio values ​​of diesel and methanol.

[0007] Based on the injection ratio, fuel injection schemes for diesel and methanol are generated;

[0008] Based on the fuel injection scheme, a timing control algorithm is used to generate diesel injection commands and methanol injection commands with a time phase difference, and the mixing effect of diesel and methanol is monitored in real time.

[0009] Optionally, the step of inputting real-time data into a pre-established engine operating condition model and outputting fuel demand characteristic parameters of the current operating condition includes: inputting speed, load and temperature values ​​into an engine operating condition model based on polynomial regression to obtain fuel demand characteristic parameters of the current operating condition.

[0010] If the fuel demand characteristic parameter is greater than the preset threshold, the relationship between the fuel demand characteristic parameter and the speed, load and temperature values ​​is fitted by the least squares method to generate the corrected fuel demand characteristic parameter.

[0011] Optionally, the step of inputting the speed, load, and temperature values ​​into the engine operating condition model based on polynomial regression to obtain the fuel demand characteristic parameters of the current operating condition includes:

[0012] ;

[0013] In the formula, The parameters represent fuel demand characteristics, where n represents engine speed, L represents load, T represents temperature, β0 represents constant coefficient, and β1 to β6 represent regression coefficients for each term.

[0014] Optionally, the step of fitting the relationship between the fuel demand characteristic parameters and the speed, load, and temperature values ​​using the least squares method to generate the corrected fuel demand characteristic parameters includes:

[0015] ;

[0016] In the formula, Let n represent the fuel demand characteristic parameter, L represent the engine speed value, T represent the engine load value, and k1, k2, k3 and k4 represent the coefficients to be fitted.

[0017] Optionally, the step of inputting fuel demand characteristic parameters into a pre-established injection ratio model and outputting the injection ratio value of diesel and methanol includes: inputting fuel demand characteristic parameters into a pre-established injection ratio model, which adopts a linear regression algorithm and is trained using fuel demand characteristic parameters from historical engine operating conditions; the training data includes the injection ratio of diesel and methanol and their corresponding fuel demand characteristic parameters.

[0018] The trained injection ratio model is used to predict the diesel to methanol injection ratio under current operating conditions; whereby...

[0019] ;

[0020] In the formula, R represents the injection ratio of diesel to methanol, α1, α2 and α3 are regression coefficients, X1, X2 and X3 are fuel demand characteristic parameters, and θ is the bias constant.

[0021] Optionally, a trained injection ratio model can be used to predict the diesel to methanol injection ratio under the current operating conditions, including:

[0022] Based on the prediction results, the model prediction error is calculated, where...

[0023] ;

[0024] In the formula, E represents the model prediction error, n is the sample size, and y i This is the actual spray ratio value. This represents the injection ratio predicted by the model.

[0025] Optionally, generating a fuel injection scheme for diesel and methanol based on the injection ratio includes: determining power load parameters and emission index parameters by combining the performance curve of the diesel engine;

[0026] Based on the power load parameters and emission index parameters, calculate the combustion efficiency coefficient of diesel and the critical mixture ratio threshold of methanol;

[0027] The diesel combustion efficiency coefficient is calculated using the thermal efficiency formula by calculating the calorific value of diesel fuel and measuring the actual output power.

[0028] Based on the injection ratio and the physicochemical properties of methanol, the critical mixing ratio threshold of methanol was determined using a flash point tester.

[0029] The target blending range of methanol and diesel is generated based on the injection ratio value, and the injection timing distribution is determined in combination with the combustion efficiency coefficient.

[0030] Determine whether the injection timing distribution exceeds the constraint of the critical mixing ratio threshold. If it does, adjust the boundary conditions of the target mixing ratio range.

[0031] Based on the adjusted boundary conditions, a fuel dual-pulse control map is generated, which includes the injection quantity, phase difference, and duration.

[0032] Optionally, the calculation of the diesel combustion efficiency coefficient using the thermal efficiency formula includes:

[0033] The formula for thermal efficiency is: Where η is the combustion efficiency coefficient, P is the actual output power, and m d For diesel quality, Q d This refers to the calorific value of diesel fuel.

[0034] Optionally, the step of generating the fuel dual-pulse control spectrum includes:

[0035] Based on the target ratio range and injection timing distribution, the injection quantity and phase difference at each time point are calculated, and the duration of each injection pulse is determined.

[0036] Optionally, the step of generating diesel injection commands and methanol injection commands with a time phase difference using a timing control algorithm includes: using a PID control algorithm to calculate the injection quantity difference between diesel and methanol based on their injection ratio values, and generating an initial injection command.

[0037] Based on the initial injection command and fuel injection scheme, a timer control tool is used to generate injection commands with time differences.

[0038] This invention offers the following advantages: By acquiring engine operating condition data in real time and inputting it into a pre-established operating condition model, the fuel demand characteristic parameters under the current operating conditions are obtained. Subsequently, these parameters are input into an injection ratio model to calculate the optimal injection ratio of diesel and methanol. Based on this ratio, a detailed fuel injection scheme is generated, and a timing control algorithm is used to produce injection commands with a time phase difference. This not only allows for precise adjustment of the diesel and methanol injection ratio but also ensures thorough mixing of the two fuels through real-time monitoring, improving fuel utilization efficiency, reducing emissions, and contributing to stable engine operation and performance output. Attached Figure Description

[0039] Figure 1 This is a schematic flowchart of the fuel control method for the methanol-diesel dual-fuel engine of the present invention. Detailed Implementation

[0040] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, but this is not to limit the scope of the invention to this.

[0041] like Figure 1 As shown in the figure, the fuel control method for a methanol-diesel dual-fuel engine provided in this embodiment may specifically include the following steps:

[0042] Step S100: Obtain real-time data of engine speed, load, and temperature; input the real-time data into a pre-established engine operating condition model and output the fuel demand characteristic parameters of the current operating condition.

[0043] Specifically, real-time engine data, including engine speed, load, and temperature, is acquired. These values ​​are then input into a pre-established engine operating condition model based on multinomial regression. The training process for this model involves using the diesel-to-methanol injection ratio and its corresponding fuel demand characteristic parameters as training data. These parameters include engine speed, load, and temperature. A linear regression algorithm is used to train the injection ratio model. The model then calculates the fuel demand characteristic parameters for the current operating condition.

[0044] The formula for calculating fuel demand characteristic parameters can be set as follows:

[0045] ;

[0046] In the formula, The parameters represent fuel demand characteristics, where n represents engine speed, L represents load, T represents temperature, β0 represents constant coefficient, and β1 to β6 represent regression coefficients for each term.

[0047] Furthermore, if the fuel demand characteristic parameter exceeds a preset threshold, the relationship between the fuel demand characteristic parameter and the engine speed, load, and temperature values ​​is fitted using the least squares method to generate a corrected fuel demand characteristic parameter. By monitoring and correcting the fuel demand characteristic parameter in real time, it can be ensured that the engine maintains optimal fuel economy under different operating conditions.

[0048] The fitting formula for the fuel demand characteristic parameters can be set as follows:

[0049] ;

[0050] In the formula, Let n represent the fuel demand characteristic parameter, L represent the engine speed value, T represent the engine load value, and k1, k2, k3 and k4 represent the coefficients to be fitted.

[0051] For example, sensors collect real-time data on engine speed (e.g., 2000 rpm), load (e.g., 80%), and temperature (e.g., 90°C). This data is then input into a pre-built engine operating condition model based on multinomial regression. The model is in the following form:

[0052] Where β0=1.2, β1=0.003, β2=0.005, β3=0.002, β4=-0.0001, β5=-0.0002, β6=-0.0001; through model calculation, the fuel demand characteristic parameter F=15.6 for the current operating condition is obtained. If F is greater than the preset threshold of 15, the relationship between F and n, L, and T is fitted using the least squares method to generate the corrected fuel demand characteristic parameter F'=14.8, and the fitting formula is:

[0053] , where k1=0.002, k2=0.004, k3=0.001, k4=1.1.

[0054] In practical applications, the fuel demand characteristics of an engine during operation depend on three key indicators: engine speed, load, and temperature. A polynomial regression model can accurately describe the relationship between these parameters, thus enabling precise prediction of fuel demand. Engine speed may vary between 2,000 and 6,000 revolutions per minute (rpm), load values ​​range from 10% to 90%, and temperature fluctuates between 60 and 120 degrees Celsius. For example, when the engine speed is 3,000 rpm, the load is 60%, and the temperature is 80 degrees Celsius, the fuel demand characteristic parameter calculated by the polynomial regression model is 0.75. This value reflects the engine's actual fuel demand under this operating condition. If the fuel demand characteristic parameter exceeds a preset threshold, such as 0.85, it needs to be corrected. The correction process uses the least squares method, establishing a system of linear equations to fit the relationship between the parameters. For instance, when a heavy truck is transporting goods over a long distance, due to the heavy load and long duration, the fuel demand characteristic parameter may exceed the threshold. In this case, the correction coefficient obtained by fitting the data using the least squares method can more accurately reflect the engine's fuel demand characteristics under high load conditions.

[0055] In a polynomial regression model, the coefficients of each term reflect the degree of influence of different parameters on fuel demand. A larger coefficient for the engine speed term indicates that changes in engine speed have a significant impact on fuel demand, while a relatively smaller coefficient for the temperature term indicates that changes in temperature have a relatively smaller impact. For example, during the startup phase, when the engine speed rises rapidly, the fuel demand characteristic parameters will increase significantly, while changes in ambient temperature will have a relatively small impact on fuel demand. The calculation and correction process of fuel demand characteristic parameters is, in fact, an important means of optimizing engine performance under different operating conditions.

[0056] This embodiment effectively reduces fuel consumption and extends engine life through this dynamic adjustment mechanism. By precisely controlling fuel demand characteristics, the engine can achieve optimal operating conditions under different operating circumstances. Dynamically adjusted fuel demand characteristics ensure that the engine maintains good power output and fuel economy throughout all stages of operation.

[0057] Step S200: Input the fuel demand characteristic parameters into the pre-established injection ratio model and output the injection ratio values ​​of diesel and methanol.

[0058] Specifically, fuel demand characteristic parameters are input into a pre-established injection ratio model, which employs a linear regression algorithm and is trained using fuel demand characteristic parameters from historical engine operating conditions. The training data includes the injection ratios of diesel and methanol and their corresponding fuel demand characteristic parameters. The trained injection ratio model is then used to predict the diesel-methanol injection ratio under the current operating conditions. In practical applications, the fuel demand characteristic parameters under the current operating conditions are input into the trained injection ratio model to predict the diesel-methanol injection ratio.

[0059] The diesel to methanol injection ratio can be obtained by the following formula:

[0060] ;

[0061] In the formula, R represents the injection ratio of diesel to methanol, α1, α2 and α3 are regression coefficients, X1, X2 and X3 are fuel demand characteristic parameters, and θ is the bias constant.

[0062] Furthermore, based on the prediction results, the model prediction error is calculated:

[0063] ;

[0064] In the formula, E represents the model prediction error, n is the sample size, and y i This is the actual spray ratio value. This represents the injection ratio predicted by the model.

[0065] In practical applications, the key to the injection ratio model lies in predicting the optimal diesel-methanol blend ratio using fuel demand characteristic parameters. For example, when the fuel demand characteristic parameter is 0.75, the diesel-methanol injection ratio calculated by the linear regression model is approximately 7:3. This ratio ensures that the engine achieves optimal combustion efficiency and emission performance under this operating condition. The linear regression algorithm establishes the relationship between the injection ratio and the fuel demand characteristic parameter by analyzing historical data. For instance, if the fuel demand characteristic parameter ranges from 0.6 to 0.85, the corresponding diesel-methanol injection ratio varies between 8:2 and 6:4. This dynamic adjustment mechanism can automatically optimize the fuel blend ratio according to changes in operating conditions.

[0066] The quality of model training data directly affects prediction accuracy. When collecting training data, it's necessary to cover the engine's operating state under different load conditions. Recording fuel demand characteristic parameters and corresponding optimal injection ratios from no-load to full-load conditions forms a complete dataset, providing a reliable foundation for model training. Calculating and controlling prediction errors is crucial for ensuring the model's practicality. For example, the average error between the model's predicted injection ratio and the actual optimal ratio should be controlled within 3%. This level of accuracy meets actual production needs while allowing for appropriate adjustment margins. Regression coefficients reflect the degree of influence of different characteristic parameters on the injection ratio. For instance, the coefficient for the characteristic parameter related to engine speed is the largest, indicating that changes in engine speed have the most significant impact on the fuel ratio. Conversely, the coefficient for the characteristic parameter related to temperature is smaller, indicating that changes in temperature have a relatively smaller impact on the injection ratio.

[0067] The dynamic predictive capabilities of injection ratio models play a crucial role in engine performance optimization. For example, when frequent start-stop cycles cause significant changes in operating conditions, adjusting the diesel-methanol injection ratio in real time can maintain optimal combustion. From idle to acceleration, the injection ratio automatically adjusts according to changes in fuel demand characteristics, ensuring both power output and emission control. In practical applications, the model's predictions need to consider the engine's physical constraints. For instance, when operating in low-temperature environments, to ensure reliability, the diesel injection ratio needs to be appropriately increased. By correcting the injection ratio model's output, the vehicle's adaptability to low-temperature conditions is improved.

[0068] Step S201: This injection ratio model uses a linear regression algorithm and is trained using fuel demand characteristic parameters from historical engine operating conditions. Specifically, it includes:

[0069] Obtain historical fuel demand characteristic parameters of the engine, including engine speed, load, and temperature. Collect diesel to methanol injection ratio data and match them with the fuel demand characteristic parameters. Construct a training dataset using the fuel demand characteristic parameters as independent variables and the diesel to methanol injection ratio as the dependent variable. Initialize the regression model using the linear regression algorithm from the Scikit-learn library. Train the linear regression model using the training dataset to minimize the prediction error. During training, determine the values ​​of the regression coefficients α1, α2, α3, and the bias constant θ.

[0070] For example, obtain the fuel demand characteristic parameters of the engine under historical operating conditions, including engine speed of 2000 rpm, load of 75%, and temperature of 85℃. Collect data on the diesel-to-methanol injection ratio, for example, diesel injection of 10 mL and methanol injection of 5 mL, and match them with the fuel demand characteristic parameters. Use the fuel demand characteristic parameters as independent variables and the diesel-to-methanol injection ratio as dependent variables to construct a training dataset, specifically including 1000 sets of historical operating condition data. Use the linear regression algorithm from the Scikit-learn library to initialize the regression model, setting the random seed to 42. Train the linear regression model using the training dataset, minimizing the prediction error, using the mean squared error as the loss function, and iterating 1000 times. During training, determine the regression coefficients α1 (0.8), α2 (0.6), α3 (0.4), and the bias constant θ (0.1). Input the fuel demand characteristic parameters under the current operating conditions into the trained linear regression model, for example, input parameters of engine speed of 2500 rpm, load of 80%, and temperature of 90℃. The model was used to calculate and predict the injection ratio of diesel and methanol under the current operating conditions, resulting in a diesel injection volume of 12 mL and a methanol injection volume of 6 mL.

[0071] For example, 1000 sets of data are extracted from a historical database, each set containing the injection ratio of diesel to methanol and the corresponding engine speed, load, and temperature. Based on the training dataset, fuel demand characteristic parameters are defined as engine speed, load, and temperature, corresponding to variables X1, X2, and X3, respectively. For example, the engine speed range is 800-4000 rpm, the load range is 20%-100%, and the temperature range is 60-120℃. Using the Scikit-learn library in Python, a linear regression algorithm model is initialized, setting the model parameters as regression coefficients α1, α2, α3, and the bias constant θ. For example, α1=0.5, α2=0.3, α3=0.2, and θ=0.1 are initialized. By minimizing the prediction error, the linear regression model is trained, and the specific values ​​of the regression coefficients α1, α2, α3, and the bias constant θ are determined. For example, after training, α1=0.6, α2=0.4, α3=0.1, and θ=0.05 are obtained. Save the trained linear regression model as an injection ratio model for subsequent prediction of the diesel-methanol injection ratio. For example, save the model as the file "Injection Ratio Model.pkl". Obtain the fuel demand characteristics parameters under the current operating conditions, including engine speed, load, and temperature. For example, the current engine speed is 1500 rpm, the load is 80%, and the temperature is 90℃. Input the fuel demand characteristics parameters under the current operating conditions into the trained injection ratio model to obtain the predicted diesel-methanol injection ratio R. Based on the predicted value R and the actual injection ratio value y... i Calculate the model prediction error E, for example, if E = 0.02. If the prediction error E exceeds the preset threshold, re-extract historical operating data, adjust the model parameters, and optimize the injection ratio model. For example, when E > 0.05, retrain the model and update the regression coefficients and bias constants.

[0072] Step S300: Generate a fuel injection scheme for diesel and methanol based on the injection ratio value.

[0073] Specifically, power load parameters and emission index parameters are determined by combining the diesel engine performance curve. Based on the power load parameters and emission index parameters, the combustion efficiency coefficient of diesel and the critical mixture ratio threshold of methanol are calculated. The diesel combustion efficiency coefficient is calculated using a thermal efficiency formula based on the diesel combustion calorific value calculation and actual output power measurement. The thermal efficiency formula is as follows: Where η is the combustion efficiency coefficient, P is the actual output power, and m d For diesel quality, Q d This refers to the calorific value of diesel fuel.

[0074] Based on the injection ratio and the physicochemical properties of methanol, the critical mixing ratio threshold of methanol was determined using a flash point tester.

[0075] The target blending ratio range of methanol and diesel is generated based on the injection ratio value (the target blending ratio range of methanol and diesel is generated using linear interpolation, and the linear interpolation formula is: M...). ratio =M min +(M max -M min )×(PP min ) / (P max -P min In the formula, M ratio M represents the mixing ratio of methanol and diesel. min and M max The preset minimum and maximum mixing ratios are given, and P represents the current power load parameter. min and P min These are the preset minimum and maximum power load parameters.

[0076] The injection timing distribution is determined by combining the combustion efficiency coefficient; it is determined whether the injection timing distribution exceeds the constraint of the critical mixture ratio threshold. If it does, the boundary conditions of the target ratio range are adjusted; a fuel dual-pulse control spectrum containing injection quantity, phase difference and duration is generated based on the adjusted boundary conditions to achieve precise mixing and combustion control of methanol and diesel.

[0077] Furthermore, the method for generating the fuel dual-pulse control spectrum is as follows: based on the target ratio range and injection timing distribution, calculate the injection quantity and phase difference at each time point, and determine the duration of each injection pulse.

[0078] For example, diesel engine performance curves are crucial for analyzing power load parameters and emission index parameters. By collecting engine performance data using speed and torque sensors, it can be observed that at different vehicle speeds, the power load parameter varies between 0.4 and 0.9, while the emission index parameter shows an upward trend with increasing load. When traveling at a constant speed of 80 km / h, the diesel combustion efficiency coefficient reaches 42%, at which point the critical methanol mixture ratio threshold is 35%. Under these conditions, the diesel combustion calorific value is 42 megajoules per kilogram, and the actual output power is 200 kilowatts. The combustion efficiency coefficient calculated using the thermal efficiency formula indicates good combustion. For example, when agricultural tractors are tilling the land, the power load parameters fluctuate significantly, requiring real-time adjustment of the target methanol-diesel mixture ratio. When the power load parameter increases from 0.5 to 0.8, the methanol mixture ratio calculated using linear interpolation increases from 20% to 30%. This dynamic adjustment ensures stable engine operation under varying load conditions. Similarly, when hydraulic excavators are performing digging operations, the injection timing distribution needs precise control based on the combustion efficiency coefficient. When the combustion efficiency coefficient is 38%, the system advances diesel injection to 15 degrees before top dead center and delays methanol injection to 5 degrees after top dead center. This staggered injection strategy effectively improves fuel mixing. For example, in fishing vessels operating in the open sea, the generation of a dual-pulse fuel control pattern is particularly crucial. At an engine speed of 1600 rpm, the diesel injection quantity is 60 mg per cycle, and the methanol injection quantity is 25 mg per cycle, with a phase difference maintained at 20 degrees, and injection durations of 1.2 milliseconds and 0.8 milliseconds, respectively. This precise injection control allows the engine to maintain optimal combustion conditions during long-term operation. For example, in urban bus operations, frequent starts and stops cause load changes, requiring real-time monitoring of the injection timing distribution. When the injection timing distribution is detected to be close to the critical mixture ratio threshold, the control system automatically adjusts the upper limit of the target mixture range from 35% to 30%, ensuring the safety and stability of the combustion process. This adaptive adjustment mechanism significantly enhances the adaptability of dual-fuel engines under complex operating conditions.

[0079] Step S400: Based on the fuel injection scheme, a timing control algorithm is used to generate diesel injection commands and methanol injection commands with a time phase difference, and the mixing effect of diesel and methanol is monitored in real time.

[0080] Specifically, a PID control algorithm is employed to calculate the injection quantity difference between diesel and methanol based on their injection ratio, generating an initial injection command. Based on the initial injection command and the fuel injection scheme, a timer control tool generates injection commands with a time difference. The mixing effect of diesel and methanol is monitored in real time to determine if the mixing ratio meets a preset threshold. If the mixing ratio does not meet the threshold, the time phase difference of the injection command is adjusted. Based on the adjusted time phase difference, the injection quantities of diesel and methanol are recalculated. A new fuel blending scheme is generated using the adjusted injection quantities. Precise injection control of diesel and methanol is then performed using this new fuel blending scheme.

[0081] For example, in a diesel and methanol dual-fuel injection system, precise mixing of the two fuels can be achieved through proportional-integral-derivative (PID) control. For instance, a heavy-duty truck engine using a transverse pump nozzle injector has an initial diesel injection pressure of 160 MPa and a methanol injection pressure of 60 MPa. Under high-speed cruising conditions, the injection quantity difference is approximately 0.04 ml per cycle. The proportional control stage calculates the injection command, the integral stage eliminates steady-state errors, and the derivative stage improves dynamic response performance.

[0082] The timer control tool uses a high-precision crystal oscillator with a clock frequency of 20 MHz, enabling microsecond-level injection phase control. For example, in a certain type of engineering machinery, the diesel injection advance angle is set to 12 degrees crankshaft angle, and the methanol injection is delayed by 1.5 milliseconds, forming a stepped injection sequence. This timing arrangement ensures that the diesel first forms a high-temperature ignition source, followed by methanol injection and full atomization.

[0083] Real-time monitoring employs optical sensors to detect the concentration distribution of the mixed fuel. For example, in a bus engine, if the mixture ratio threshold is set at 25%, the control system automatically adjusts the injection phase difference when the mixture ratio exceeds the threshold. By increasing the phase difference to two milliseconds, the mixture ratio can be reduced to a safe range. Simultaneously, combustion data fed back from the cylinder pressure sensor shows that the adjusted phase difference maintains a stable combustion process.

[0084] When recalculating the injection quantity is required, the control system comprehensively considers factors such as engine speed and load. For example, during harvesting operations, frequent load fluctuations in agricultural machinery lead to unstable air-fuel mixture ratios. The system stabilizes combustion by adjusting the baseline diesel injection quantity. When the load suddenly increases by 30%, the diesel injection quantity increases by 20%, while the methanol injection quantity only increases by 10% to ensure reliable ignition. During the execution of the fuel blending scheme, the control system can use adaptive algorithms to dynamically optimize injection parameters. For instance, under low-speed, heavy-load conditions, the main engine of a fishing vessel improves combustion by adjusting the injection duration in real time. When the engine speed drops to 1000 rpm, the diesel injection pulse width is appropriately extended to 0.8 milliseconds, while the methanol injection quantity remains unchanged to avoid producing an overly rich mixture. This precise control method not only improves fuel economy but also significantly reduces emissions.

[0085] The above description is only a preferred embodiment of the present invention. Therefore, any equivalent changes or modifications made to the structure, features and principles described in the claims of this patent application are included within the protection scope of this patent application.

Claims

1. A fuel control method for a methanol-diesel dual-fuel engine, characterized in that, include: Acquire real-time data on engine speed, load, and temperature; Real-time data is input into a pre-established engine operating condition model based on multinomial regression, and the model outputs fuel demand characteristic parameters for the current operating condition, including: ； In the formula, This represents the fuel demand characteristic parameters, where n represents the engine speed, L represents the load, T represents the temperature, β0 represents the constant term coefficient, and β1 to β6 represent the regression coefficients for each term. If the fuel demand characteristic parameters exceed a preset threshold, the relationship between the fuel demand characteristic parameters and the engine speed, load, and temperature is fitted using the least squares method to generate corrected fuel demand characteristic parameters, including: ； In the formula, The parameters represent fuel demand characteristics, where n represents engine speed, L represents engine load, T represents engine temperature, and k1, k2, k3, and k4 represent coefficients to be fitted. Input the fuel demand characteristic parameters into a pre-established injection ratio model, and output the injection ratio values ​​of diesel and methanol, including: The fuel demand characteristic parameters are input into a pre-established injection ratio model, which is trained using a linear regression algorithm and fuel demand characteristic parameters from historical engine operating conditions. The training data includes the injection ratios of diesel and methanol and their corresponding fuel demand characteristic parameters. The trained injection ratio model is used to predict the diesel to methanol injection ratio under current operating conditions; whereby... ； In the formula, R represents the injection ratio of diesel to methanol, α1, α2 and α3 are regression coefficients, X1, X2 and X3 are fuel demand characteristic parameters, and θ is the bias constant. Based on the injection ratio, fuel injection schemes for diesel and methanol are generated; Based on the fuel injection scheme, a timing control algorithm is used to generate diesel injection commands and methanol injection commands with a time phase difference, and the mixing effect of diesel and methanol is monitored in real time.

2. The fuel control method for a methanol-diesel dual-fuel engine according to claim 1, characterized in that, Predict the diesel to methanol injection ratio under current operating conditions using a trained injection ratio model, including: Based on the prediction results, the model prediction error is calculated, where... ； In the formula, E represents the model prediction error, n is the sample size, and y i This is the actual spray ratio value. This represents the injection ratio predicted by the model.

3. The fuel control method for a methanol-diesel dual-fuel engine according to claim 1, characterized in that, The process of generating a fuel injection scheme for diesel and methanol based on the injection ratio includes: Determine the power load parameters and emission index parameters by combining the performance curves of the diesel engine; Based on the power load parameters and emission index parameters, calculate the combustion efficiency coefficient of diesel and the critical mixture ratio threshold of methanol; The diesel combustion efficiency coefficient is calculated using the thermal efficiency formula by calculating the calorific value of diesel fuel and measuring the actual output power. Based on the injection ratio and the physicochemical properties of methanol, the critical mixing ratio threshold of methanol was determined using a flash point tester. The target blending range of methanol and diesel is generated based on the injection ratio value, and the injection timing distribution is determined in combination with the combustion efficiency coefficient. Determine whether the injection timing distribution exceeds the constraint of the critical mixing ratio threshold. If it does, adjust the boundary conditions of the target mixing ratio range. Based on the adjusted boundary conditions, a fuel dual-pulse control map is generated, which includes the injection quantity, phase difference, and duration.

4. The fuel control method for a methanol-diesel dual-fuel engine according to claim 3, characterized in that, The calculation of the diesel combustion efficiency coefficient using the thermal efficiency formula includes: The formula for thermal efficiency is: Where η is the combustion efficiency coefficient, P is the actual output power, and m d For diesel quality, Q d This refers to the calorific value of diesel fuel.

5. The fuel control method for a methanol-diesel dual-fuel engine according to claim 3, characterized in that, The steps for generating the fuel dual-pulse control pattern include: Based on the target ratio range and injection timing distribution, the injection quantity and phase difference at each time point are calculated, and the duration of each injection pulse is determined.

6. The fuel control method for a methanol-diesel dual-fuel engine according to claim 1, characterized in that, The generation of diesel injection commands and methanol injection commands with a time phase difference using a timing control algorithm includes: The PID control algorithm is used to calculate the difference in injection quantity between diesel and methanol based on their injection ratio, and to generate the initial injection command. Based on the initial injection command and fuel injection scheme, a timer control tool is used to generate injection commands with time differences.

Images