A motorcycle engine control system

Abstract

The present invention relates to a motorcycle engine control system. The control system collects motorcycle engine driving data in real time, including operating condition data and air-fuel ratio; inputs the driving data into a preset exhaust emission analysis model to predict the current concentration of harmful gases in the exhaust; constructs a generalized additive model to quantify the response relationship between the driving data and the concentration of harmful gases in the exhaust, and controls the air-fuel ratio of the current engine operating condition based on the response relationship. The present invention solves the problem in the prior art of difficulty in accurately controlling the air-fuel ratio according to different engine operating conditions, thereby increasing the concentration of harmful gases and polluting the environment. By constructing a random forest model as an exhaust emission analysis model, the concentration of harmful gases in exhaust emissions can be accurately predicted using engine operating conditions. By establishing a generalized additive model, the problem of the difficulty in quantifying the nonlinear relationship between driving data and harmful gas concentration is resolved, thereby improving the control accuracy of the air-fuel ratio under different engine operating conditions.

Description

Technical Field

[0001] The invention belongs to the technical field of motorcycles and relates to a motorcycle engine control system. Background Art

[0002] Motorcycle engines produce various pollutants during operation, such as carbon monoxide (CO), nitrogen oxides (NO x ), etc. These pollutants not only cause serious harm to the environment, but also affect people's health. In traditional motorcycle engine technology, the control of the engine is relatively extensive, especially in air-fuel ratio control and ignition control. There are many deficiencies.

[0003] Traditionally, motorcycle engines have relied on a relatively simple carburetor system to adjust the fuel mixture. This mechanical control method struggles to precisely control the air-fuel ratio based on varying engine operating conditions (such as speed and cylinder temperature). The carburetor primarily regulates fuel supply based on the vacuum created during the intake process, but its precision in regulating both air and fuel intake is limited. For example, during a cold engine start, the carburetor may provide an overly rich mixture to ensure a smooth engine start. However, as engine temperature rises, this overly rich mixture becomes unsuitable, leading to incomplete fuel combustion and increased pollutant emissions. Furthermore, under complex operating conditions, such as varying loads, the carburetor cannot flexibly and accurately adjust the air-fuel ratio, often preventing the engine from operating at the optimal air-fuel ratio. This wastes energy and exacerbates environmental pollution. Summary of the Invention

[0004] In order to solve the above-mentioned problems existing in the prior art, the present invention provides a motorcycle engine control system, which aims to configure the appropriate air-fuel ratio for different operating conditions of the motorcycle engine by establishing a response relationship model between the air-fuel ratio and the concentration of harmful gases in the exhaust gas under different engine operating conditions, thereby improving the control accuracy of the engine and reducing the emission of harmful gases.

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] The present application provides a motorcycle engine control system, comprising a driving data acquisition module, an exhaust emission analysis module, and an air-fuel ratio control module, wherein the driving data acquisition module, the exhaust emission analysis module, and the air-fuel ratio control module are communicatively connected, wherein:

[0007] The driving data acquisition module is used to collect the motorcycle engine driving data in real time, wherein the driving data includes the working condition data and the air-fuel ratio; the working condition data includes the engine speed and the engine cylinder temperature;

[0008] The exhaust emission analysis module is used to input the driving data into a preset exhaust emission analysis model to predict the concentration of harmful gases in the current exhaust;

[0009] The air-fuel ratio control module is used to construct a generalized additive model to quantify the response relationship between driving data and the concentration of harmful gases in exhaust gas, and control the air-fuel ratio of the current engine operating condition according to the response relationship.

[0010] Furthermore, in the driving data acquisition module, the engine speed is detected by an engine crankshaft speed sensor, the engine cylinder temperature is detected by a cylinder temperature sensor, and the air-fuel ratio is characterized by the oxygen concentration in the exhaust gas.

[0011] Furthermore, the exhaust emission analysis model includes the following construction steps:

[0012] S1. Collect historical motorcycle engine driving data and the corresponding harmful gas concentrations in exhaust gas;

[0013] S2. Divide the data samples into different exhaust emission levels according to the concentration of harmful gases;

[0014] S3. Train a random forest model using driving data as input features and exhaust emission levels as corresponding prediction labels.

[0015] S4. The random forest model is used as an exhaust emission analysis model to determine the exhaust emission levels under different driving data.

[0016] Furthermore, the generalized additive model includes the following construction steps:

[0017] T1. Data Preparation: Collect historical motorcycle engine driving data and the corresponding harmful gas concentrations in exhaust gas, and preprocess the collected data, including data cleaning and data standardization. The driving data is the response variable, and the harmful gas concentration in exhaust gas is the explanatory variable.

[0018] T2. Model selection and setting: Determine the distribution family and smoothing function of the model response variable, where the distribution family is configured as a normal distribution and the smoothing function is configured as a cubic spline function;

[0019] T3, Model Fitting: Substitute the preprocessed data into the selected generalized additive model form and use iterative weighted least squares method to fit the model;

[0020] T4. Model evaluation and diagnosis: Use evaluation indicators to judge the fitting effect of the model, including mean square error, mean absolute error and coefficient of determination;

[0021] T5. Model output: After the generalized additive model is established, a response relationship curve between the driving data and the concentration of harmful gases in the exhaust gas is output to determine the response relationship between the driving data and the concentration of harmful gases in the exhaust gas.

[0022] Furthermore, the generalized additive model is expressed as follows:

[0023] E(y)=μ+s(x1)+s(x2)+...+s(x i ),

[0024] Where, E(y) is the response variable, i.e., the concentration of harmful gases in the exhaust gas; μ is the intercept term; s(x i ) represents a smooth function; x1, x2, …, x i represents the explanatory variable, that is, the driving data.

[0025] Furthermore, controlling the air-fuel ratio of the current engine operating condition according to the response relationship includes the following steps:

[0026] E1. Set the concentration threshold of harmful gases according to relevant emission standards;

[0027] E2. Calculating the relative deviation between the current concentration of harmful gases in the exhaust gas and the harmful gas concentration threshold;

[0028] E3. When the relative deviation exceeds a set threshold, the air-fuel ratio control mode is activated, specifically including:

[0029] Identify current engine operating condition data;

[0030] Determine the appropriate air-fuel ratio data range based on the response relationship between driving data and the concentration of harmful gases in exhaust gas, combined with current operating conditions data;

[0031] According to the determined air-fuel ratio data range, the ratio of the intake air volume and the fuel volume is adjusted.

[0032] Furthermore, the relative deviation is calculated as follows:

[0033]

[0034] Where R t is the relative deviation; A t is the concentration of harmful gases in the current exhaust gas; B t is the harmful gas concentration threshold.

[0035] Furthermore, the air-fuel ratio control module also constructs an adaptive control model to coordinate the ignition advance angle and air-fuel ratio according to different operating conditions, specifically including:

[0036] Under acceleration conditions, the fuel injection amount is increased while the air-fuel ratio is increased, and the ignition advance angle is increased to provide power output;

[0037] Under idling conditions, the air-fuel ratio is controlled and reduced to ensure stable engine operation and compliance with emission standards.

[0038] Furthermore, the adaptive control model includes the following construction steps:

[0039] P1. Working condition data collection: Collect the ignition advance angle data and air-fuel ratio data of the motorcycle engine under different working conditions;

[0040] P2. Determine the model architecture: Using the operating condition data as input and the ignition advance angle data and air-fuel ratio data as output, determine the adaptive control model, which can be configured as a neural network model or a fuzzy logic model.

[0041] P3. Model parameter estimation: Initialize the model parameters according to the selected model type. Specifically, for neural network models, initialize the connection weights and thresholds between neurons; for fuzzy logic models, determine the membership function and rule base of the fuzzy set;

[0042] P4. Model training: The model is trained using a specific training algorithm, and the model parameters are continuously adjusted so that the error between the model's prediction results for the training data and the actual output meets the preset conditions. The training algorithm specifically uses the backpropagation algorithm for the neural network model and the fuzzy inference algorithm for the fuzzy logic model;

[0043] P5. Model evaluation and optimization: By calculating the error index between the predicted value and the actual value, the model's predictive ability and fitting effect on new data can be judged.

[0044] Furthermore, the air-fuel ratio control module also uses a real-time operating system to control the air-fuel ratio according to dynamic changes in motorcycle engine operating condition data.

[0045] Beneficial effects of the present invention:

[0046] (1) Real-time collection of motorcycle engine driving data, including operating condition data and air-fuel ratio; inputting the driving data into a preset exhaust emission analysis model to predict the current concentration of harmful gases in the exhaust; constructing a generalized additive model to quantify the response relationship between the driving data and the concentration of harmful gases in the exhaust, and controlling the air-fuel ratio of the current engine operating condition based on the response relationship. The present invention solves the problem that the existing technology is difficult to accurately control the air-fuel ratio according to different engine operating conditions, thereby increasing the concentration of harmful gases and polluting the environment.

[0047] (2) By constructing a random forest model as an exhaust emission analysis model, the concentration of harmful gases in exhaust emissions can be accurately predicted using engine operating conditions, providing a basis for subsequent control of the air-fuel ratio.

[0048] (3) By establishing a generalized additive model, the problem of the nonlinear relationship between driving data and harmful gas concentration being difficult to quantify is solved, thereby improving the control accuracy of the air-fuel ratio under different engine operating conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] To facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0050] Figure 1 This is a structural diagram of a motorcycle engine control system in the present invention.

[0051] Figure 2 The figure is a flow chart of constructing an exhaust emission analysis model in one embodiment of the present invention.

[0052] Figure 3 The flowchart of constructing a generalized additive model in one embodiment of the present invention is shown. DETAILED DESCRIPTION

[0053] In order to further illustrate the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0054] See also Figure 1-Figure 3 The present application provides a motorcycle engine control system, including a driving data acquisition module, an exhaust emission analysis module, and an air-fuel ratio control module, wherein the driving data acquisition module, the exhaust emission analysis module, and the air-fuel ratio control module are communicatively connected, wherein:

[0055] The driving data acquisition module is used to collect the motorcycle engine driving data in real time, wherein the driving data includes the working condition data and the air-fuel ratio; the working condition data includes the engine speed and the engine cylinder temperature;

[0056] In this embodiment, the driving data acquisition module plays a key role, responsible for collecting real-time driving data from the motorcycle engine. This data includes operating conditions (including engine speed and cylinder temperature) and the air-fuel ratio. Engine speed intuitively reflects different engine operating states, such as power output and fuel consumption during idling, acceleration, and high-speed driving. Engine cylinder temperature is related to engine performance, lifespan, and exhaust emissions, and different cylinder temperatures result in different requirements for mixture supply.

[0057] Furthermore, in the driving data acquisition module, the engine speed is detected by an engine crankshaft speed sensor, the engine cylinder temperature is detected by a cylinder temperature sensor, and the air-fuel ratio is characterized by the oxygen concentration in the exhaust gas.

[0058] In this embodiment, engine speed is detected using an engine crankshaft speed sensor. This sensor accurately senses the rotational speed of the engine crankshaft. Since the crankshaft speed directly reflects the overall engine speed, the sensor can accurately measure the engine speed at any given moment. This data is crucial for determining the engine's current operating state, such as idling, accelerating, or driving at a constant speed. It also provides an important basis for subsequent system control and analysis.

[0059] Engine cylinder temperature is monitored using a cylinder temperature sensor. During engine operation, fuel combustion generates heat within the cylinder, causing the cylinder temperature to fluctuate. Cylinder temperature sensors monitor cylinder temperature in real time. Because cylinder temperature not only affects engine performance and lifespan but also closely influences exhaust emissions and other factors, accurate cylinder temperature data is crucial for the entire engine control system to fully understand the engine's operating status.

[0060] Regarding air-fuel ratio data collection, an indirect method is used: exhaust gas oxygen concentration. The air-fuel ratio is the ratio of the mass of air to the mass of fuel in the mixture. After combustion, the oxygen concentration in the exhaust gas is closely related to the air-fuel ratio before combustion. If the air-fuel ratio of the mixture is appropriate and combustion is complete, the exhaust gas oxygen concentration will be within a specific and reasonable range. If the air-fuel ratio is too rich or too lean, the exhaust gas oxygen concentration will be correspondingly high or low. Therefore, by measuring the exhaust gas oxygen concentration, the current air-fuel ratio of the engine can be characterized to a certain extent, providing data support for subsequent air-fuel ratio-based control.

[0061] The exhaust emission analysis module is used to input the driving data into a preset exhaust emission analysis model to predict the concentration of harmful gases in the current exhaust;

[0062] In this embodiment, the preset exhaust emission analysis model is constructed based on extensive experimental data, theoretical research, and a deep understanding of motorcycle engine operating characteristics and exhaust emission patterns. It incorporates complex mathematical relationships and algorithmic logic, and can simulate the actual combustion process and exhaust gas production of the engine under current operating conditions based on different combinations of input driving data. This allows for accurate prediction of the concentration levels of harmful gases such as carbon monoxide, nitrogen oxides, and hydrocarbons in the exhaust gas under these operating conditions. Specifically, the exhaust emission analysis model includes the following construction steps:

[0063] Step S1: Data collection:

[0064] The first step was to conduct extensive data collection. This involved collecting historical motorcycle engine driving data and the corresponding data on harmful gas concentrations in exhaust. This driving data covered a wide range of information, including engine speed, cylinder temperature, and the air-fuel ratio, as represented by the oxygen concentration in the exhaust. By accumulating a large amount of real-world data under various operating conditions, we provided a rich and authentic foundation for subsequent model construction, enabling us to comprehensively and accurately reflect the relationship between motorcycle engine operating conditions and exhaust emissions.

[0065] Step S2: Data division:

[0066] After acquiring a large amount of data samples, they are classified and divided into different exhaust emission levels based on the concentration of harmful gases in the exhaust. This classification helps to more clearly sort out the characteristic differences in engine driving data under different emission levels, allowing subsequent model learning to better capture the inherent connection between different emission conditions and driving data, thereby improving the accuracy and pertinence of the model's exhaust emission predictions.

[0067] Step S3: Model training:

[0068] The random forest model is trained using organized driving data as input features and the classified exhaust emission levels as corresponding prediction labels. The random forest model is a decision tree-based ensemble learning model that improves prediction accuracy and stability by constructing multiple decision trees and combining their prediction results. During training, the model continuously learns the mapping between the input driving data features and exhaust emission levels, gradually adjusting its parameters to better fit the data, ultimately enabling it to accurately predict exhaust emission levels based on the given driving data.

[0069] Step S4: Model application:

[0070] After sufficient training, the random forest model was identified as an exhaust emissions analysis model, capable of determining exhaust emission levels under different driving data. When new motorcycle engine driving data is input into the model, it can quickly output the corresponding exhaust emission level based on previously learned patterns and relationships. This provides a strong basis for determining the current motorcycle exhaust emission status, allowing appropriate measures to be taken to control exhaust emissions, such as adjusting the air-fuel ratio, thereby achieving effective management of motorcycle exhaust emissions.

[0071] The air-fuel ratio control module is used to construct a generalized additive model to quantify the response relationship between driving data and the concentration of harmful gases in exhaust gas, and control the air-fuel ratio of the current engine operating condition according to the response relationship.

[0072] In this embodiment, the air-fuel ratio control module primarily achieves effective exhaust emissions control by constructing a generalized additive model. First, the generalized additive model takes various driving data as input, including engine speed, engine cylinder temperature, and the air-fuel ratio (represented by the oxygen concentration in the exhaust gas), and outputs the concentration of harmful gases in the exhaust gas. By analyzing and fitting a large amount of relevant data collected over a large period of time, the model accurately captures the complex, nonlinear response relationship between driving data and harmful gas concentrations. Second, during actual operation of the motorcycle, the driving data acquisition module collects the latest engine driving data in real time and feeds it into the generalized additive model. Based on this quantified response relationship, the model quickly calculates a predicted value for the harmful gas concentration under the current operating conditions. If the predicted value indicates that the harmful gas concentration exceeds a preset threshold, it indicates that the current air-fuel ratio may be problematic and requires adjustment. At this point, the air-fuel ratio control module uses relevant actuators, such as the fuel injection system, to take action. For example, if the carbon monoxide concentration in the exhaust gas is too high, it may mean that the mixture is too rich, that is, the air-fuel ratio is too small. The module will instruct the fuel injection system to reduce the fuel injection amount, increase the air-fuel ratio, and promote full combustion of the mixture, thereby reducing the concentration of harmful gases.

[0073] The Generalized Additive Model (GAM) is a statistical model used to analyze nonlinear relationships between variables. The following describes its principles and construction steps:

[0074] Principle: The generalized additive model is an extension of the traditional linear regression model. Traditional linear regression assumes a simple linear relationship between the dependent and independent variables, describing this relationship by fitting a straight line. However, in reality, many relationships between variables are not simple linear. The generalized additive model overcomes this limitation by allowing for nonlinear relationships between the dependent and independent variables. The basic idea is to express the expectation of the dependent variable as the sum of smooth functions of the independent variables, using these smooth functions to capture various possible nonlinear relationships between the independent and dependent variables.

[0075] Build steps:

[0076] T1. Data preparation:

[0077] Collect relevant data samples. For example, for a motorcycle engine control system, this involves collecting driving data such as engine speed, cylinder temperature, and air-fuel ratio as independent variables (explanatory variables), and the corresponding concentration of harmful gases in exhaust as the dependent variable (response variable). This data should be representative and sufficient to cover the engine's performance under various operating conditions.

[0078] The collected data are preprocessed, including data cleaning (removing outliers, processing missing values, etc.), data standardization (converting variables of different magnitudes into forms with similar scales for better model fitting), and other operations to ensure the quality and availability of the data.

[0079] T2. Model selection and setting:

[0080] Determine the specific form of the generalized additive model to use. Depending on the characteristics of the data and the purpose of the study, you can choose different distribution families to fit the dependent variable, such as normal distribution, Poisson distribution, binomial distribution, etc., depending on the nature of the dependent variable (such as continuous, count, binary, etc.).

[0081] Select a method for estimating smooth functions, common ones include spline functions (such as cubic splines) and local polynomial regression. These methods can generate smooth function curves by locally fitting the data to capture the nonlinear relationship between independent variables and dependent variables.

[0082] T3. Model fitting:

[0083] Substitute the preprocessed data into the selected generalized additive model form and use an appropriate algorithm to fit the model. Common fitting algorithms include iteratively reweighted least squares (IRLS). By continuously iteratively optimizing the model parameters, the model is able to best fit the data, that is, to minimize the difference between the predicted value and the actual value (such as the residual sum of squares).

[0084] T4. Model evaluation and diagnosis:

[0085] Some evaluation indicators are used to judge the fitting effect of the model, such as mean square error (MSE), mean absolute error (MAE), coefficient of determination (R 2 ) etc. These indicators can reflect the degree of fit of the model to the data from different perspectives. The smaller the value (for MSE and MAE) or the larger the value (for R 2 ) indicates that the model fits better.

[0086] Perform model diagnostics to check for overfitting or underfitting. Overfitting refers to a model that fits the training data too perfectly but performs poorly on new data. Underfitting refers to a model that fails to fully capture patterns in the data and performs poorly on both the training and new data. You can use methods such as cross-validation to determine whether the model exhibits these issues and make adjustments based on the diagnostic results, such as increasing or decreasing model complexity.

[0087] T5. Model output:

[0088] After the generalized additive model is established, the response relationship curve between different driving data such as engine speed, cylinder temperature and air-fuel ratio and harmful gas concentration is output to determine the response relationship between driving data and harmful gas concentration in exhaust gas.

[0089] Furthermore, the generalized additive model is expressed as follows:

[0090] E(y)=μ+s(x1)+s(x2)+...+s(x i ),

[0091] Where, E(y) is the response variable, i.e., the concentration of harmful gases in the exhaust gas; μ is the intercept term; s(x i ) represents a smooth function; x1, x2, …, x i represents the explanatory variable, that is, the driving data.

[0092] Furthermore, controlling the air-fuel ratio of the current engine operating condition according to the response relationship includes the following steps:

[0093] Step E1: Setting the Hazardous Gas Concentration Threshold

[0094] This step is a crucial foundation for the entire air-fuel ratio control process. Relevant departments will formulate emission standards based on environmental protection requirements and the actual operating characteristics of motorcycle engines. These standards clearly define the maximum concentrations of various harmful gases (such as carbon monoxide, nitrogen oxides, and hydrocarbons) permitted in exhaust.

[0095] In motorcycle engine control systems, the maximum allowable concentrations of various harmful gases are set as harmful gas concentration thresholds based on these authoritative emission standards. For example, the concentration threshold for carbon monoxide may be set at a specific value (e.g., X ppm), while that for nitrogen oxides may be set at another value (e.g., Y ppm). These thresholds serve as key references for determining whether engine exhaust emissions meet standards.

[0096] Step E2: Calculate relative deviation

[0097] During the actual operation of the motorcycle, the exhaust emission analysis module can obtain the actual concentration values of various harmful gases in the current exhaust in real time.

[0098] Then, for each harmful gas, its current actual concentration is compared with the corresponding harmful gas concentration threshold set in step E1, and the relative deviation between the two is calculated. The formula for calculating the relative deviation is generally: (Current Harmful Gas Concentration - Harmful Gas Concentration Threshold) / Harmful Gas Concentration Threshold × 100%. By calculating the relative deviation, we can intuitively understand the degree of deviation of the current harmful gas concentration in the exhaust gas from the standard threshold, thus providing a quantitative basis for determining whether to activate the air-fuel ratio control mode.

[0099] Step E3: Start the air-fuel ratio control mode and specific operations

[0100] Startup condition judgment: When the relative deviation calculated for one or more harmful gases exceeds the set threshold (this set threshold is also determined based on a comprehensive consideration of multiple factors such as engine performance and emission control), it indicates that the current engine exhaust emissions do not meet the requirements and the air-fuel ratio needs to be adjusted. At this time, the air-fuel ratio control mode is turned on.

[0101] Identifying Engine Operating Conditions: After activating the air-fuel ratio control mode, the first step is to accurately identify the current engine operating conditions. This data includes engine speed and cylinder temperature, collected in real time by appropriate sensors (such as the engine crankshaft speed sensor and cylinder temperature sensor). Accurately understanding engine operating conditions is crucial for determining the appropriate air-fuel ratio, as the engine's air-fuel ratio requirements vary under different operating conditions.

[0102] Determining the appropriate air-fuel ratio data range: Based on the previously established relationship between driving data (such as engine speed, engine cylinder temperature, and air-fuel ratio as measured by exhaust oxygen concentration) and exhaust gas concentrations, combined with the recently identified engine operating condition data, a comprehensive analysis is performed to determine the appropriate air-fuel ratio data range for the current operating conditions. For example, at a specific engine speed and cylinder temperature, if the exhaust carbon monoxide concentration is too high, the response relationship analysis indicates that the air-fuel ratio may need to be increased. Therefore, a corresponding appropriate air-fuel ratio data range can be determined.

[0103] Adjusting the ratio of intake air and fuel: Finally, the ratio of intake air and fuel is adjusted based on the determined air-fuel ratio data range. The air-fuel ratio is essentially the ratio of intake air to fuel. If it is determined that the air-fuel ratio needs to be increased, this means increasing the intake air or reducing the fuel. In practice, this is usually achieved by controlling the engine's intake system (such as the throttle) and fuel injection system. For example, the throttle opening can be adjusted to increase intake air, or the fuel injector can be controlled to reduce fuel injection, so that the ratio of intake air and fuel falls within the determined appropriate air-fuel ratio data range. This reduces the concentration of harmful gases in the exhaust and ensures that engine exhaust emissions meet relevant standards.

[0104] Furthermore, the relative deviation is calculated as follows:

[0105]

[0106] Where R t is the relative deviation; A t is the concentration of harmful gases in the current exhaust gas; B t is the concentration threshold of harmful gases.

[0107] Furthermore, the air-fuel ratio control module also constructs an adaptive control model to coordinate the ignition advance angle and air-fuel ratio according to different operating conditions, specifically including:

[0108] Under acceleration conditions, the fuel injection amount is increased and the ignition advance angle is increased to provide strong power output;

[0109] Under idling conditions, the air-fuel ratio is precisely controlled to ensure stable engine operation and compliance with emission standards.

[0110] In this embodiment, the adaptive control model aims to perform intelligent and coordinated control of two key parameters, the ignition advance angle and the air-fuel ratio, according to different operating conditions of the engine, so as to achieve the best performance of the engine under various operating conditions while meeting emission requirements.

[0111] The core principle of this model is based on a deep understanding and analysis of the engine's characteristics and requirements under different operating conditions. Through extensive testing, data collection, and a precise understanding of engine operating principles, the model clearly defines how changes in ignition advance angle and air-fuel ratio affect engine power output, fuel economy, and exhaust emissions under different operating conditions. Based on this, a control logic that automatically adapts to different operating conditions is constructed.

[0112] When a motorcycle is accelerating, the engine needs to quickly increase power output to meet the vehicle's acceleration requirements. At this time, the adaptive control model triggers the corresponding control strategy:

[0113] Increased fuel injection: During acceleration, engine speed rises rapidly, dramatically increasing energy demand. To ensure sufficient fuel for complete combustion and release more energy to propel the vehicle, the adaptive control model instructs the fuel injection system to increase the injection volume. This ensures sufficient fuel and air in the cylinder to form a combustible mixture, laying the material foundation for strong power output.

[0114] Adjusting the ignition advance angle: At the same time, simply increasing the amount of fuel injected is not enough; the ignition advance angle also needs to be adjusted appropriately. The ignition advance angle refers to the advance angle of the spark plug ignition moment relative to the piston reaching top dead center. Under acceleration conditions, appropriately advancing the ignition moment (i.e., increasing the ignition advance angle) allows the mixture to begin burning before the piston reaches top dead center, allowing the pressure generated by the combustion to more effectively push the piston downward, thereby improving the engine's power output efficiency. By precisely adjusting the ignition advance angle, the combustion process can be more coordinated with the engine's mechanical movement, further enhancing the engine's power performance during acceleration.

[0115] In idling conditions, the engine only needs to maintain its basic operation, and the control focus at this time is completely different from that in acceleration conditions:

[0116] Fine control of the air-fuel ratio: When idling, the engine speed is low and fuel consumption is relatively low, but the control requirements for the air-fuel ratio are more precise. Because if the air-fuel ratio is out of balance, it may cause unstable engine operation, jitter or even stalling, and also affect exhaust emissions. The adaptive control model will adjust the fuel injection amount based on the real-time collected engine operating data (such as engine cylinder temperature, intake volume, etc.) to accurately control the air-fuel ratio within an appropriate range that can ensure stable engine operation and meet exhaust emissions standards. For example, in the idle state, a relatively lean mixture is usually required to reduce the emission of pollutants such as carbon monoxide, while ensuring that the engine can run smoothly.

[0117] Through such targeted control strategies, the adaptive control model can effectively coordinate the ignition advance angle and air-fuel ratio under different operating conditions, so that the motorcycle engine can maintain good operating stability and environmental performance while meeting the power output requirements.

[0118] Furthermore, the adaptive control model includes the following construction steps:

[0119] Step 1: Working condition data collection and analysis:

[0120] Extensive data collection across various operating conditions: Detailed data is collected from motorcycle engines under various operating conditions, including idling, acceleration, deceleration, constant speed driving, and climbing. For each operating condition, data recorded includes engine speed, cylinder temperature, air intake, throttle opening, load, and corresponding ignition advance angle, air-fuel ratio settings, and exhaust emissions. By installing high-precision sensors on multiple motorcycles and conducting long-term data collection in various driving scenarios, we ensure that the collected data is comprehensive and representative, truly reflecting the actual operating status of the engine under different operating conditions.

[0121] Analyze the relationship between operating condition characteristics and parameters: Conduct in-depth analysis of the large amount of collected data from different operating conditions to study the engine's operating characteristics under each operating condition, as well as the intrinsic relationship between key parameters such as ignition advance angle and air-fuel ratio and other operating data. For example, under acceleration conditions, observe the changing trend of the intake volume, how the fuel injection volume is adjusted, and the impact of this adjustment on the ignition advance angle and air-fuel ratio when the engine speed rises rapidly. At the same time, analyze the relationship between these parameter changes and power output and exhaust emissions. Under idling conditions, explore the stability of cylinder temperature, the control method of intake volume, and the effect of fine-tuning the air-fuel ratio and ignition advance angle on engine stability and emissions when the engine maintains low speed operation. Through this meticulous analysis, the mutual influence between various parameters under different operating conditions is sorted out, providing a solid theoretical foundation for subsequent model construction.

[0122] Step 2: Determine the model architecture:

[0123] Select the appropriate model type: Based on the analysis of operating data and the requirements of engine control, select a suitable mathematical model as the basic architecture of the adaptive control model. Common choices include neural network models, fuzzy logic models, or rule-based expert system models. For example, neural network models have powerful nonlinear mapping capabilities and can well capture the complex nonlinear relationships between engine operating parameters; fuzzy logic models are good at handling situations with ambiguity and uncertainty, and are more suitable for fuzzy descriptions of engine operating conditions and fuzzy formulation of control decisions; rule-based expert system models can construct control logic based on existing professional knowledge and empirical rules, and are suitable for situations where there are clear rules to follow for engine control.

[0124] Define model inputs and outputs: Determine the input and output variables of the selected model. In this adaptive control model, input variables typically include real-time engine operating data such as engine speed, cylinder temperature, intake air volume, throttle opening, and load conditions. This data comprehensively describes the current engine operating state. Output variables are set to the ignition advance angle and air-fuel ratio, two key parameters that require control and adjustment. By clearly defining the model inputs and outputs, model construction has a clear goal and direction, facilitating subsequent parameter estimation and model training.

[0125] Step 3: Model parameter estimation and training:

[0126] Initialize model parameters: Depending on the selected model type, initialize the various model parameters. For example, for a neural network model, you need to initialize parameters such as the connection weights and thresholds between neurons; for a fuzzy logic model, you need to determine parameters such as the fuzzy set membership function and rule base. These initial parameters are typically based on empirical values or random assignments, but will be continuously optimized and adjusted during subsequent training.

[0127] Training with data: The large amount of collected data from different operating conditions is divided into training sets and test sets according to a certain ratio. First, the model is trained using the training set, with the engine operating data in the training set as input and the corresponding ignition advance angle and air-fuel ratio as output, allowing the model to learn the relationship between the input and output data. During the training process, specific training algorithms (such as the backpropagation algorithm for neural network models and the fuzzy inference algorithm for fuzzy logic models) are used to continuously adjust the model parameters to minimize the error between the model's predictions of the training data and the actual output. For example, for neural network models, the backpropagation algorithm continuously updates the connection weights and thresholds between neurons to minimize the sum of squared errors between the ignition advance angle and air-fuel ratio output by the model and the actual ignition advance angle and air-fuel ratio.

[0128] Step 4: Model evaluation and optimization:

[0129] Evaluating the model using a test set: After a certain number of rounds of model training, the trained model is evaluated using a reserved test set. The engine operating data from the test set is fed into the trained model to obtain the model-predicted ignition advance angle and air-fuel ratio, which are then compared with the actual ignition advance angle and air-fuel ratio from the test set. The model's predictive ability and fit for new data are assessed by calculating error metrics (such as mean squared error and mean absolute error) between the predicted and actual values. Excessively large error metrics indicate that the model may be overfitting or underfitting, and further optimization is required.

[0130] Optimize model performance: Based on the evaluation results, if the model is found to be overfitting or underfitting, take appropriate optimization measures. Overfitting can be addressed by increasing training data, adjusting model complexity (such as reducing the number of neural network layers or neurons), and employing regularization techniques. Underfitting can be addressed by increasing model complexity (such as increasing the number of neural network layers or neurons), improving the training algorithm, and extending the training time. Through continuous evaluation and optimization, ensure that the model has good predictive capabilities and fitting results, and can accurately predict the ignition advance angle and air-fuel ratio based on engine operating data to meet the requirements of the engine control system.

[0131] Furthermore, the air-fuel ratio control module also uses a real-time operating system to control the air-fuel ratio according to dynamic changes in motorcycle engine operating condition data.

[0132] In this embodiment, the real-time operating system (RTOS) plays a crucial role. The primary characteristic of an RTOS is its exceptional real-time performance. Motorcycle engine operating conditions are constantly changing, requiring immediate adjustments to the air-fuel ratio. The RTOS ensures that the entire process—from collecting operating data such as engine speed, cylinder temperature, and intake air volume, to analyzing and processing data to make air-fuel ratio adjustment decisions, and then transmitting control commands to actuators such as the fuel injection system—is completed in record time. For example, during engine acceleration, the RTOS responds swiftly, adapting the air-fuel ratio to the new operating conditions within mere milliseconds to avoid power and emissions issues. Its task scheduling and priority management are highly efficient. Air-fuel ratio control involves multiple tasks, including data collection, analysis, and command transmission. The RTOS prioritizes tasks based on their importance and urgency. For example, tasks such as adjusting the air-fuel ratio based on the latest operating conditions, which have a greater impact on engine performance and emissions, receive a higher priority. When multiple tasks request execution simultaneously, the higher-priority task is prioritized, ensuring that critical processes are not delayed and enabling precise control. The RTOS excels in resource allocation and management. Given the limited resources of motorcycle engine control systems, the RTOS effectively allocates processor computing power and memory space. Data acquisition requires temporary memory to ensure smooth execution, guaranteeing resource availability for other tasks, and maintaining a smooth control process. RTOS also offers high reliability and stability. Engine operation is susceptible to electromagnetic interference, which it addresses through fault-tolerance mechanisms and data protection measures. Data transmission errors can be verified and corrected, and sensor or actuator failures can be addressed using pre-set mechanisms, such as switching to backup components or adjusting strategies. This ensures safe and stable engine operation and effective air-fuel ratio control. In short, RTOS provides a solid foundation for air-fuel ratio control in motorcycle engines.

[0133] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as above in terms of a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art can, without departing from the scope of the technical solution of the present invention, make some changes or modifications to equivalent embodiments using the technical contents disclosed above. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A motorcycle engine control system, characterized in that: It includes a driving data acquisition module, an exhaust emission analysis module and an air-fuel ratio control module, which are communicatively connected to each other, wherein: The driving data acquisition module is used to collect the motorcycle engine driving data in real time, wherein the driving data includes the working condition data and the air-fuel ratio; the working condition data includes the engine speed and the engine cylinder temperature; The exhaust emission analysis module is used to input the driving data into a preset exhaust emission analysis model to predict the concentration of harmful gases in the current exhaust; The air-fuel ratio control module is used to construct a generalized additive model to quantify the response relationship between driving data and the concentration of harmful gases in exhaust gas, and control the air-fuel ratio of the current engine operating condition according to the response relationship; The generalized additive model includes the following construction steps: T1. Data Preparation: Collect historical motorcycle engine driving data and the corresponding harmful gas concentrations in exhaust gas, and preprocess the collected data, including data cleaning and data standardization. The driving data is the response variable, and the harmful gas concentration in exhaust gas is the explanatory variable. T2. Model selection and setting: Determine the distribution family and smoothing function of the model response variable, where the distribution family is configured as a normal distribution and the smoothing function is configured as a cubic spline function; T3, Model Fitting: Substitute the preprocessed data into the selected generalized additive model form and use iterative weighted least squares method to fit the model; T4. Model evaluation and diagnosis: Use evaluation indicators to judge the fitting effect of the model, including mean square error, mean absolute error and coefficient of determination; T5. Model output: After the generalized additive model is established, a response curve between driving data and the concentration of harmful gases in exhaust gas is output to determine the response relationship between driving data and the concentration of harmful gases in exhaust gas; The generalized additive model is expressed as follows: ， Where E(y) is the response variable, i.e., the concentration of harmful gases in exhaust gas; μ is the intercept term; s(xi) represents a smooth function; x1, x2, …, xi represent explanatory variables, i.e., driving data; The controlling of the air-fuel ratio of the current engine operating condition according to the response relationship comprises the following steps: E1. Set the concentration threshold of harmful gases according to relevant emission standards; E2. Calculating the relative deviation between the current concentration of harmful gases in the exhaust gas and the harmful gas concentration threshold; E3. When the relative deviation exceeds a set threshold, the air-fuel ratio control mode is activated, specifically including: Identify current engine operating condition data; Determine the appropriate air-fuel ratio data range based on the response relationship between driving data and harmful gas concentrations in exhaust gas, combined with current operating conditions data; According to the determined air-fuel ratio data range, the ratio of the intake air volume and the fuel volume is adjusted.

2. A motorcycle engine control system according to claim 1, characterized in that: In the driving data acquisition module, the engine speed is detected by an engine crankshaft speed sensor, the engine cylinder temperature is detected by a cylinder temperature sensor, and the air-fuel ratio is characterized by the oxygen concentration in the exhaust gas.

3. A motorcycle engine control system according to claim 1, characterized in that: The exhaust emission analysis model includes the following construction steps: S1. Collect historical motorcycle engine driving data and the corresponding harmful gas concentrations in exhaust gas; S2. Divide the data samples into different exhaust emission levels according to the concentration of harmful gases; S3. Train a random forest model using driving data as input features and exhaust emission levels as corresponding prediction labels. S4. The random forest model is used as an exhaust emission analysis model to determine the exhaust emission levels under different driving data.

4. A motorcycle engine control system according to claim 1, characterized in that: The relative deviation is calculated as follows: ， Where, R t is the relative deviation; A t is the concentration of harmful gases in the current exhaust gas; B t is the harmful gas concentration threshold.

5. A motorcycle engine control system according to claim 1, characterized in that: The air-fuel ratio control module also builds an adaptive control model to coordinate the control of the ignition advance angle and the air-fuel ratio according to different operating conditions, specifically including: Under acceleration conditions, the fuel injection amount is increased while the air-fuel ratio is increased, and the ignition advance angle is increased to provide power output; Under idling conditions, the air-fuel ratio is controlled and reduced to ensure stable engine operation and compliance with emission standards.

6. A motorcycle engine control system according to claim 5, characterized in that: The adaptive control model includes the following construction steps: P1. Working condition data collection: Collect the ignition advance angle data and air-fuel ratio data of the motorcycle engine under different working conditions; P2. Determine the model architecture: Using the operating condition data as input and the ignition advance angle data and air-fuel ratio data as output, determine the adaptive control model, which can be configured as a neural network model or a fuzzy logic model. P3. Model parameter estimation: Initialize the model parameters according to the selected model type. Specifically, for neural network models, initialize the connection weights and thresholds between neurons; for fuzzy logic models, determine the membership function and rule base of the fuzzy set; P4. Model training: The model is trained using a specific training algorithm, and the model parameters are continuously adjusted so that the error between the model's prediction results for the training data and the actual output meets the preset conditions. The training algorithm specifically uses the backpropagation algorithm for the neural network model and the fuzzy inference algorithm for the fuzzy logic model; P5. Model evaluation and optimization: By calculating the error index between the predicted value and the actual value, the model's predictive ability and fitting effect on new data can be judged.

7. A motorcycle engine control system according to claim 1, characterized in that: The air-fuel ratio control module also uses a real-time operating system to control the air-fuel ratio according to dynamic changes in motorcycle engine operating condition data.

Images