Fuel switching method and device for dual-fuel ship and storage medium

Abstract

The invention relates to the technical field of ship power, in particular to a dual-fuel ship fuel switching method and device and a storage medium, and the method comprises the steps that first engine performance data and engine state data corresponding to the first engine performance data are acquired; preprocessing the first engine performance data and the engine state data; the preprocessed first engine performance data are input into a target prediction control model for switching strategy prediction, and a fuel switching strategy corresponding to the engine is obtained; inputting the engine state data into the target fault prediction model for fault prediction to obtain a switching decision; a fuel switching strategy is sent to the control module according to the switching decision, so that the control module processes the fuel switching process according to the fuel switching strategy; and second engine performance data are received, and the fuel switching strategy is adjusted according to the second engine performance data. And seamless switching or cooperative use between liquefied natural gas and diesel oil of a ship can be met.

Description

Technical Field

[0001] The present application relates to the technical field of ship power, and particularly relates to a dual-fuel ship fuel switching method, device, and storage medium. Background Art

[0002] In the related art, the ship industry has long used a single-fuel propulsion system, such as diesel or liquefied natural gas. This single-fuel propulsion system has certain limitations and cannot fully utilize the advantages of different fuels, making it difficult to meet the requirements of energy conservation, emission reduction, and economy. Currently, the technology for seamless switching between liquefied natural gas and diesel on ships is not yet mature. The existing technology cannot meet the needs of seamless switching or combined use of different fuels on ships, and does not fully consider the differences in the physical and chemical properties of different fuels, as well as the safety and reliability issues during the switching process.

[0003] In summary, the technical problems existing in the related art need to be improved. Summary of the Invention

[0004] The main purpose of the embodiments of the present application is to propose a dual-fuel ship fuel switching method, device, and storage medium, which can meet the seamless switching or combined use of liquefied natural gas and diesel on ships.

[0005] To achieve the above object, on the one hand, an embodiment of the present application proposes a dual-fuel ship fuel switching method, the method including:

[0006] Obtaining first engine performance data and engine state data corresponding to the first engine performance data;

[0007] Preprocessing the first engine performance data and the engine state data;

[0008] Inputting the preprocessed first engine performance data into a target predictive control model for switching strategy prediction to obtain a fuel switching strategy corresponding to the engine;

[0009] Inputting the engine state data into a target fault prediction model for fault prediction to obtain a switching decision;

[0010] Sending the fuel switching strategy to a control module according to the switching decision, so that the control module processes the fuel switching process according to the fuel switching strategy;

[0011] Receiving second engine performance data and adjusting the fuel switching strategy according to the second engine performance data.

[0012] In some embodiments, the first engine performance data includes: engine speed stability data, fuel consumption rate, and pollutant emission data;

[0013] The engine state data includes: engine operation parameter verification data, fuel characteristic verification data, environmental parameter verification data, and system state verification data.

[0014] In some embodiments, inputting the preprocessed first engine performance data into a target predictive control model for switching strategy prediction to obtain a fuel switching strategy corresponding to the engine includes:

[0015] Obtaining real-time engine performance data corresponding to the dual-fuel ship at the current moment;

[0016] Updating the first historical performance data corresponding to the dual-fuel ship according to the real-time engine performance data to obtain second historical performance data;

[0017] Training the predictive control model at the current moment through the second historical performance data to obtain the target predictive control model;

[0018] Inputting the preprocessed first engine performance data into the target predictive control model for switching strategy prediction to obtain the fuel switching strategy corresponding to the engine.

[0019] In some embodiments, the construction and optimization process of the target predictive control model includes the following steps:

[0020] Establishing a discrete-time state-space model according to state variables, input variables, and output variables;

[0021] Constructing a predictive model based on the discrete-time state-space model;

[0022] Establishing an optimization objective function according to the predictive model and the performance index of fuel switching;

[0023] Converting the problem to be solved by the target predictive control model into a standard quadratic programming problem, and solving and optimizing it through an optimization algorithm to obtain the target predictive control model;

[0024] Implementing control input at each time step, comparing, analyzing, and correcting according to output feedback, and dynamically optimizing the target predictive control model through rolling optimization.

[0025] In some embodiments, the formula for the solution optimization under constraint conditions is:

[0026]

[0027] Wherein, U k represents the decision variable vector, H represents the Hessian matrix, and g represents the linear term coefficient vector.

[0028] In some embodiments, the constraint conditions include control input conditions and state variable conditions;

[0029] The control input conditions include: upper and lower limits of fuel flow rate and upper and lower limits of valve opening;

[0030] The state variable conditions include: upper and lower limits of engine speed, upper and lower limits of temperature, and upper and lower limits of pressure.

[0031] In some embodiments, the prediction of the engine state data by the target fault prediction model to generate a switching decision includes:

[0032] Obtaining the real-time state data corresponding to the dual-fuel ship at the current moment;

[0033] Updating the first historical state data corresponding to the dual-fuel ship according to the real-time state data to obtain second historical state data;

[0034] Training the fault prediction model at the current moment through the second historical state data to obtain a target fault prediction model;

[0035] Inputting the engine state data into the target fault prediction model for fault prediction to obtain the switching decision.

[0036] In some embodiments, the target fault prediction model includes: a fault classification model, a leakage risk prevention and control model, and an external factor risk prevention and control model.

[0037] To achieve the above object, another aspect of the embodiments of the present application provides a fuel switching device for a dual-fuel ship, and the device includes:

[0038] A first module, configured to obtain first engine performance data and engine state data corresponding to the first engine performance data;

[0039] A second module, configured to preprocess the first engine performance data and the engine state data;

[0040] A third module, configured to input the preprocessed first engine performance data into a target predictive control model for switching strategy prediction to obtain a fuel switching strategy corresponding to the engine;

[0041] A fourth module, configured to input the engine state data into a target fault prediction model for fault prediction to obtain a switching decision;

[0042] A fifth module, configured to send the fuel switching strategy to a control module according to the switching decision, so that the control module processes the fuel switching process according to the fuel switching strategy;

[0043] The sixth module is configured to receive the second engine performance data and adjust the fuel switching strategy according to the second engine performance data.

[0044] To achieve the above object, on the other hand, an embodiment of the present application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned dual-fuel ship fuel switching method.

[0045] The embodiments of the present application have at least the following beneficial effects: The present application provides a dual-fuel ship fuel switching method, device and storage medium. This solution obtains the first engine performance data and engine state data, obtains the fuel switching strategy in cooperation with the target predictive control model for the first engine performance data, obtains the switching decision in cooperation with the fault prediction model for the engine state data, and sends the fuel switching strategy to the control module for fuel switching according to the switching decision. The switching decision ensures the safety of fuel switching, determines whether to perform fuel switching based on the safety of fuel switching, and the fuel switching strategy ensures the energy conservation, emission reduction and economic requirements of fuel switching. The combination of the two jointly ensures the reliability of fuel switching. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 is a flowchart of the dual-fuel ship fuel switching method provided by the embodiment of the present application;

[0047] Figure 2 is Figure 1 a flowchart of step S300 in

[0048] Figure 3 is Figure 1 a flowchart of step S400 in

[0049] Figure 4 is a schematic structural diagram of a dual-fuel ship fuel switching device provided by the embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of the present application. They are only examples of devices and methods consistent with some aspects of the embodiments of the present application.

[0051] It will be appreciated that the terms "first", "second", etc. used in the present application may be used herein to describe various concepts, but unless otherwise specified, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of the embodiments of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the words "if", "when" as used herein may be interpreted as "when...", "while...", or "in response to determining".

[0052] The terms "at least one", "a plurality of", "each", "any one", etc. used in the present application, "at least one" includes one, two or more than two, "a plurality of" includes two or more than two, "each" refers to each of the corresponding plurality, and "any one" refers to any one of the plurality.

[0053] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present application belongs. The terms used herein are only for the purpose of describing the embodiments of the present application and are not intended to limit the present application.

[0054] Figure 1 is an alternative flowchart of the dual-fuel ship fuel switching method provided by the embodiments of the present application. Figure 1 The method in may include but is not limited to steps S100 to S600:

[0055] Step S100: Obtain first engine performance data and engine state data corresponding to the first engine performance data;

[0056] Step S200: Preprocess the first engine performance data and the engine state data;

[0057] Step S300: Input the preprocessed first engine performance data into the target predictive control model for switching strategy prediction to obtain the fuel switching strategy corresponding to the engine;

[0058] Step S400: Input the engine state data into the target fault prediction model for fault prediction to obtain a switching decision;

[0059] Step S500: Send the fuel switching strategy to the control module according to the switching decision so that the control module processes the fuel switching process according to the fuel switching strategy;

[0060] Step S600: Receive the second engine performance data and adjust the fuel switching strategy according to the second engine performance data.

[0061] In some embodiments, during the operation of the engine, the first engine performance data and the engine state data are obtained through sensors or other instrument devices. The first engine performance data is processed by a target prediction control model to obtain a fuel switching strategy corresponding to the engine. The engine state data is processed by a target fault prediction model to obtain a switching decision. According to the switching decision, the fuel switching strategy is sent to a control module, and the control module processes the fuel switching process according to the fuel switching strategy. The engine operation data processed according to the fuel switching strategy is continuously obtained to obtain the second engine performance data, and then the fuel switching strategy is adjusted according to the second engine performance data.

[0062] In some embodiments, in step S100, the first engine performance data and the engine state data of the first engine performance data are obtained. The first engine performance data includes: engine speed stability data, fuel consumption rate, and pollutant emission data. Among them:

[0063] The engine speed stability data is obtained by a speed sensor installed on the engine crankshaft or transmission shaft. These sensors usually convert the rotational motion of the engine into an electrical signal based on principles such as electromagnetic induction and Hall effect. The electronic control unit can read these signals in real time and calculate the engine speed, and then analyze the stability of the speed.

[0064] The fuel consumption rate is measured by installing a flow sensor on the fuel supply pipeline to measure the fuel flow rate per unit time, and combining information such as the engine operation time, the fuel consumption rate is calculated. For some large engines or industrial equipment, the fuel consumption rate may also be estimated by indirectly analyzing the pressure change of the fuel system.

[0065] The pollutant emission data is obtained using a dedicated emission detection device, such as an exhaust gas analyzer. These devices can measure the concentration of various pollutants in the exhaust gas. The pollutants include substances such as carbon monoxide, hydrocarbons, and nitrogen oxides, so as to obtain the pollutant emission data, and some emission-related data can be monitored and reported in real time.

[0066] The engine state data includes: engine operation parameter verification data, fuel characteristic verification data, environmental parameter verification data, and system state verification data. Among them:

[0067] The engine operating parameter verification data includes temperature verification data, pressure verification data, and rotational speed verification data. The temperature verification data includes verification values such as the engine coolant temperature, lubricating oil temperature, exhaust temperature, etc. Exemplarily, for instance, the normal range verification value of the coolant temperature can ensure that the engine operates in a suitable thermal state. If the actual temperature exceeds or is lower than the verification range, it may indicate that there is a fault in the engine, that it is not suitable for fuel switching, or that special measures need to be taken when switching fuels. The pressure verification data are the standard values or thresholds such as fuel pressure, oil pressure, etc. The verification data of the fuel pressure can help determine whether the fuel supply system is normal. If the pressure is abnormal, it may lead to fuel switching failure or affect the engine performance. The rotational speed verification data are the reference values such as the rated rotational speed and idle speed of the engine. By comparing the actual rotational speed with the verification rotational speed data, the operating state of the engine can be understood, ensuring that the engine rotational speed is stable during fuel switching and avoiding problems such as excessive rotational speed fluctuations.

[0068] The fuel property verification data includes density verification data, viscosity verification data, and flash point and ignition point verification data. For the density verification data, different fuels have different densities, such as heavy oil, diesel, liquefied natural gas, etc. Accurate density verification data can help determine the mass - volume relationship of the fuel, be used to calculate the fuel supply amount, ensure that the engine obtains the correct fuel amount after fuel switching, and maintain normal combustion. For the viscosity verification data, the viscosity of the fuel affects its fluidity and atomization effect. For example, heavy oil has a higher viscosity. When switching to heavy oil, parameters such as the heating temperature and fuel supply pressure need to be adjusted according to the viscosity verification data to ensure that the fuel can be sprayed and burned well. The flash point and ignition point verification data are important indicators for measuring the safety and combustion characteristics of the fuel. During the fuel switching process, it is necessary to ensure that the stored and used fuel is within the safe temperature range to avoid dangers such as fire or explosion. Therefore, the verification data of the flash point and ignition point are crucial.

[0069] The environmental parameter verification data includes temperature and humidity verification data and air pressure verification data. The temperature and humidity verification data affect the physical properties of the fuel and the performance of the engine. For example, in a cold and humid environment, moisture in the fuel may condense, affecting the combustion effect. Through the verification data of the environmental temperature and humidity, the fuel switching strategy can be adjusted accordingly, such as adding antifreeze or adjusting the combustion parameters. The air pressure verification data affects the intake air volume and combustion pressure of the engine. At different altitudes or weather conditions, the air pressure is different. According to the air pressure verification data, the fuel injection amount and intake control can be optimized to ensure that the engine can maintain good performance after fuel switching.

[0070] The system status verification data includes sensor status verification data and valve and pipeline status verification data. The sensor status verification data is used to determine whether the sensors are working properly to ensure the accuracy and reliability of the obtained engine performance data. If a certain sensor fails, the data transmitted by it may mislead the fuel switching decision, so the verification of the sensor status is crucial. For the valve and pipeline status verification data, the fuel switching process involves the opening and closing of various valves and the conduction of pipelines. The verification data of the valve opening, sealing performance, and whether the pipeline is blocked can ensure that the fuel can be switched according to the predetermined path and flow rate, preventing problems such as leakage or supply interruption.

[0071] In some embodiments, in step S200, the first engine performance data and engine status data are preprocessed. The preprocessing methods include data cleaning, normalization, smoothing processing, and feature extraction and algorithms. Among them, data cleaning is to remove noise, outliers, and incorrect data in the data. For example, data points that deviate significantly from the normal range due to sensor failures or interference need to be identified, removed, or corrected; normalization is to convert data with different ranges and units into a unified standard range for easy model processing and comparison; smoothing processing is to reduce data fluctuations and spikes so that the data can better reflect the true trends and characteristics. Common smoothing methods include the moving average method or the exponential smoothing method, etc.; feature extraction and selection is to extract features that are important for the predictive controller from the original data and select the most representative and informative features as inputs. For example, by calculating the statistical features and spectral features of the data, the statistical features include: mean, variance, or standard deviation, etc., to extract higher-level feature information, reduce the data dimension, and improve the efficiency and accuracy of the model.

[0072] Figure 2 is Figure 1 The flowchart of step S300 in Figure 1 Step S300 of includes but is not limited to steps S310 to S340:

[0073] Step S310: Obtain the real-time engine performance data corresponding to the dual-fuel ship at the current moment;

[0074] Step S320: Update the first historical performance data corresponding to the dual-fuel ship according to the real-time engine performance data to obtain the second historical performance data;

[0075] Step S330: Train the predictive control model at the current moment through the second historical performance data to obtain the target predictive control model;

[0076] Step S340: Input the preprocessed first engine performance data into the target predictive control model for switching strategy prediction to obtain the fuel switching strategy corresponding to the engine.

[0077] In some embodiments, in step S310, at the current moment when the dual-fuel ship is operating, various sensors and instruments installed on the ship engine and related systems are used to collect real-time engine operation parameter verification data, fuel characteristic verification data, environmental parameter verification data, and system status verification data. These data cover various aspects of information such as engine speed stability, fuel consumption rate, pollutant emissions, fuel flow rate, valve opening degree, etc. The collected real-time data is preprocessed and integrated to form real-time engine performance data corresponding to the current moment.

[0078] In some embodiments, in step S320, the obtained real-time engine performance data is incorporated into the existing first historical performance data. The first historical performance data is various data records about the performance of the dual-fuel ship engine collected and stored over a past period of time. By merging and updating the real-time data with the first historical performance data, more comprehensive second historical performance data that can better reflect the current ship operation state is obtained. This process involves operations such as data storage, database update, and data format unification to ensure that the new data can be accurately added to the historical dataset.

[0079] In some embodiments, in steps S330 to S340, the time series data in the second historical performance data is adapted according to the requirements of the discrete-time state space model. This involves discretizing the continuous sensor data at a specific time step to ensure that the data format is consistent with the model input requirements. For example, the engine speed data collected once per second is reorganized according to a discrete time step of 0.1 seconds. At the same time, the data is normalized to unify variables with different dimensions (such as fuel flow rate in liters per minute and valve opening in percentage) into a similar numerical range to improve the stability and efficiency of model training. Using the adapted data, the parameters of the discrete-time state space model are estimated through system identification methods. Exemplarily, taking the state transition matrix as an example, algorithms such as the least squares method are used to estimate the values of each element in the matrix based on the changes in state variables in the historical data. For another example, according to the change relationship between state variables such as fuel flow rate and engine speed at different times, the corresponding elements in the state transition matrix are adjusted to accurately reflect the natural evolution law of the system state. During the training process, these parameters are continuously iteratively optimized so that the model can better fit the system dynamic behavior in the historical data. Using part of the second historical performance data as a validation set, the model prediction results are compared and analyzed with the actual data. Metrics such as mean square error and mean absolute error are calculated to evaluate the accuracy of the model. If the model error is large, the parameter estimation method is further adjusted or the data volume is increased to optimize the model to ensure that the discrete-time state space model can accurately describe the dynamic characteristics of the dual-fuel ship engine system in the discrete-time domain.

[0080] Extract features related to engine performance prediction from the second historical performance data. In addition to the direct performance index data, new features can also be generated through data transformation, combination, etc. For example, calculate the change rate of fuel consumption rate over a period of time, or weight-combine the flow rates of different types of fuels as a new feature. Through methods such as correlation analysis or principal component analysis, select the features that have a greater impact on the prediction target to reduce the complexity of model training and improve the prediction accuracy. Initialize the selected model, including determining the network structure, setting model parameters, etc. Divide the data processed by feature engineering into a training set and a validation set. During the training process, the model adjusts its own parameters by continuously learning the relationship between the features in the training set and the prediction target. During the training process, use the validation set to regularly evaluate the model, adopt techniques such as cross-validation, adjust the hyperparameters of the model, prevent the model from overfitting, and improve the generalization ability and prediction accuracy of the model.

[0081] According to the operation objectives of the dual-fuel ship, the operation objectives include: reducing fuel costs, reducing pollutant emissions, or maintaining stable engine operation, etc., and an optimization objective function is constructed. During the training process, according to the actual operation conditions and feedback data, the weight coefficients of the objective function are continuously refined to more accurately reflect the relative importance of each objective. For example, if the environmental protection requirements are increased, the weight of the pollutant emission cost in the objective function is appropriately increased. During the training process, according to factors such as the maintenance of the ship equipment and the change of the operation environment, these constraint conditions are updated in real time. For example, when the ship engine is maintained, the safety upper limit of its rotational speed may increase, and the rotational speed constraint conditions are updated accordingly. During the training process, the algorithm continuously searches for the optimal values of the decision variables according to the current objective function and constraint conditions. The decision variables include: fuel flow control sequence or valve opening control sequence. In each iteration process, according to the change of the objective function and the satisfaction degree of the constraint conditions, the search direction and step size are adjusted to improve the convergence speed and solution accuracy of the optimization algorithm. At the same time, combined with the idea of real-time rolling optimization, within each time step, according to the newly obtained second historical performance data, the optimization problem is reconstructed and solved to ensure that the control strategy can timely adapt to the change of the ship operation state.

[0082] In some embodiments, the construction and optimization process of the target predictive control model includes the following steps:

[0083] Establish a discrete-time state space model based on state variables, input variables, and output variables;

[0084] Construct a prediction model based on the discrete-time state space model;

[0085] Establish an optimization objective function according to the prediction model and the performance index of fuel switching;

[0086] Transform the problem to be solved by the target predictive control model into a standard quadratic programming problem, and solve and optimize it through an optimization algorithm to obtain the target predictive control model;

[0087] Implement control input at each time step, compare, analyze, and correct according to the output feedback, and dynamically optimize the target predictive control model through rolling optimization.

[0088] Specifically, after the data is input, a continuous-time state-space model is constructed using the input state variables, input variables, and output variables to describe the dynamic characteristics of the ship's fuel system, establishing the mathematical relationship between the internal state, input, and output of the system. The continuous-time state-space model is transformed into a discrete-time state-space model to adapt to the digital processing of the computer. The discrete-time step is determined to convert the continuous system behavior into a discrete-time series representation. The current system state data output from the discrete-time state-space model is combined with historical data as the input to the prediction model. Based on the input data, the algorithm and mechanism of the prediction model are used to predict the future state of the ship's engine, including changes in future engine speed stability, fuel consumption rate, pollutant emissions, and other indicators.

[0089] An optimization objective function is constructed around the key performance indicators of ship dual-fuel switching, such as minimizing engine speed fluctuations, minimizing fuel consumption, and minimizing pollutant emissions. The reference trajectory, state error weight matrix, and control input weight matrix are defined. The setting of these parameters determines the degree of emphasis on different performance indicators. If more attention is paid to fuel economy, the weight of fuel consumption in the objective function can be appropriately increased to guide the optimization direction of the control input. An appropriate optimization algorithm is determined according to the characteristics of the optimization objective function and the complexity of the problem. The model predictive control problem is transformed into a standard quadratic programming problem, and the Hessian matrix and linear term are determined. With the help of mature quadratic programming solution algorithms, such as the interior point method, sequential quadratic programming method, etc., the optimization problem is solved to obtain the optimal control input sequence. During the solution process, the control input constraints and state variable constraints are fully considered to ensure that the optimization results meet the actual operation limits of the ship. The optimal control strategy obtained by solving the optimization problem is output to guide the actual operation of the ship's fuel switching system, achieving precise control of the ship's engine.

[0090] During the optimization process of the model, at each time step, the first control input in the obtained optimal control input sequence is applied to the control module to control the operating state of the engine by adjusting the valve openings of diesel and gas, etc. According to the output feedback of the actual system, the actual state data at the current moment is obtained and compared with the model prediction results. If there is a deviation, the parameters of the discrete-time state-space model are corrected in real time to improve the prediction accuracy of the model. At the same time, based on the new state data, the prediction model and the optimization objective function are updated, and the optimization problem is solved again to obtain the optimal control input for the next time step. For rolling optimization, at each time step, the first control input in the optimal control input sequence obtained by solving the optimization problem is applied to the ship dual-fuel system, and then the prediction window is moved forward by one time step. The current system state data is re-obtained, and the parameters of the discrete-time state-space model are updated. Based on the updated model, the above steps of constructing the prediction model, setting the optimization objective function, and solving the optimization problem are repeated, and the control input is continuously adjusted according to the real-time state of the system to achieve real-time and dynamic optimization switching of the ship dual-fuel and ensure the stable and efficient operation of the engine under different working conditions. The performance indicators of the output system in the current and future states, such as the predicted engine speed stability, fuel consumption rate, pollutant emissions, etc., are output to provide reference for the operating personnel on the system operating state for further analysis and decision-making.

[0091] In some embodiments, for the discrete-time state-space model, the discrete-time state-space equation is:

[0092] State equation: x(k + 1) = A d x(k) + B d u(k);

[0093] Output equation: y(k) = Cx(k) + Du(k);

[0094] where x(k) represents the state vector, u(k) represents the control input vector, y(k) represents the output vector, A d represents the state transition matrix, B d represents the input matrix, C represents the output matrix, and D represents the feedforward matrix.

[0095] It should be noted that the state variable, input variable, and output variable are the basic descriptive quantities of the system. The state vector is composed of multiple state variables, the control input vector is composed of input variables, and the output vector is composed of output variables. The vector is the set form of the corresponding variables and is used for system mathematical modeling and analysis.

[0096] The state vector x(k) includes the engine operating state and the fuel system state, which are used to describe the internal state of the system at time k. The engine operating state includes the rotational speed, temperature, and pressure. The rotational speed of the engine is an important indicator reflecting its operating state. During the fuel switching process, changes in the rotational speed will affect the power output and stability of the engine. For example, when switching from diesel to fuel, if the rotational speed fluctuates too much, it will cause instability in the ship's power. The temperature includes the coolant temperature and exhaust temperature of the engine. Fuel switching may affect the combustion process, thereby causing changes in the temperatures of various parts of the engine. Excessive or too low temperatures may affect the performance and lifespan of the engine. The pressure, such as the intake pressure, fuel pressure, and fuel pressure, etc. Different fuels have different combustion characteristics. When switching fuels, these pressure parameters need to be adjusted accordingly to ensure the normal operation of the engine. The fuel system state includes the fuel inventory and the fuel mixing ratio. The fuel inventory is the remaining amount of diesel and natural gas, which is very important for determining whether fuel switching can be carried out and the timing of switching. For the fuel mixing ratio, in a dual-fuel engine, there is a situation where diesel and liquefied natural gas are combusted in combination. The state information of the fuel mixing ratio is crucial for controlling fuel switching and optimizing the combustion process.

[0097] The control input vector u(k) includes fuel flow control and valve opening, which are variables used to control the system behavior. Among them, fuel flow control includes diesel flow and gas flow. The diesel flow controls the supply of diesel by adjusting the diesel flow, thereby realizing the switching process from diesel to gas or from gas to diesel. For example, when switching to gas, the diesel flow is gradually reduced. The gas flow adjusts the gas flow to control the gas supply. During the fuel switching process, it is necessary to accurately control the increase or decrease of the gas flow according to the engine operating state and performance requirements. The valve opening includes the diesel valve opening and the gas valve opening. The diesel valve opening controls the size of the passage for diesel to enter the engine and is closely related to the diesel flow control. The gas valve opening controls the size of the passage for gas to enter the engine and is one of the key control parameters for realizing fuel switching.

[0098] The output vector y(k) includes performance indicators and pollutant emission concentrations, representing the output results of the system at time k. The performance indicators include power output and fuel consumption rate. The power output directly affects the sailing speed and power performance of the ship. During the fuel switching process, it is necessary to ensure the smoothness of the power output and avoid excessive fluctuations. The fuel consumption rate reflects the fuel utilization efficiency of the engine under different fuel supply conditions. Optimizing the fuel consumption rate is one of the important goals of ship dual-fuel switching. The pollutant emission concentration is the emission concentration of pollutants such as nitrogen oxides, carbon monoxide, and hydrocarbons. Reducing pollutant emissions is an important requirement for the environmental protection operation of ships. During the fuel switching process, it is necessary to monitor and control the emission indicators in real time.

[0099] A d Describes the transition relationship of the system state from time k to time k+1, reflecting the internal dynamic characteristics of the system, such as the inertia of the engine itself, the dynamic response of the fuel combustion process, etc. In the scenario of dual-fuel switching of ships, it includes the dynamic characteristic information of the engine under different fuel states, such as the influence relationship of different fuels on state variables such as engine speed, temperature, and pressure.

[0100] B d Represents the influence degree of the control input u(k) on the system state x(k). For example, how the changes in diesel fuel flow and gas fuel flow affect state variables such as engine speed and temperature. The elements in B d reflect the coupling relationship between fuel flow control and engine state.

[0101] C describes how the system state x(k) is mapped to the output vector y(k). For example, how state variables such as engine speed and temperature affect output indicators such as power output, fuel consumption rate, and pollutant emissions. The matrix reflects the relationship between the internal state of the system and the externally observable output.

[0102] D represents the direct influence of the control input u(k) on the output vector y(k). In the scenario of dual-fuel ship switching, the direct change in fuel flow will immediately have a certain impact on output indicators such as power output and fuel consumption rate. The matrix is used to describe this direct influence relationship.

[0103] At time k, predict the state x p of the future N k time steps and the control input U k , and the specific expression is:

[0104]

[0105] where, x k represents the state of the time step, N p represents the prediction step length, U k represents the control input, and N c represents the control step length.

[0106] According to the discrete-time state-space equation, the formula can be obtained:

[0107]

[0108] where, i represents the index of the prediction step length, and j represents the internal index used for iteration in the summation.

[0109] Integrating the above formula into matrix form, the formula can be obtained:

[0110] X k= Φx k + θU k ;

[0111] Where Φ represents the state transition matrix and θ represents the control input matrix.

[0112] For the optimization objective function, it is to minimize the weighted sum of the system error and the control input, so the formula is:

[0113]

[0114] Where R k represents the reference trajectory, Q represents the state error weight matrix, and W represents the control input weight matrix.

[0115] After expanding the above formula, we can get:

[0116]

[0117] Where E represents the term independent of the control input, and E = Φx k - R k .

[0118] The solution optimization of the target predictive control model is transformed into a standard quadratic programming problem, and the formula can be obtained:

[0119]

[0120] Where H is the Hessian matrix, and H = 2(θ T Qθ + W), g is the linear term, and g = 2E T QΘ.

[0121] In some embodiments, during the solution optimization process, constraint conditions need to be considered, including control input constraints and state variable constraints. Among them, the control input constraints conform to the formula:

[0122] U lb ≤ U k ≤ U ub ;

[0123] Specifically, the control input constraint refers to the variable that can be directly adjusted by the control system. In the fuel switching scenario, it mainly involves fuel flow and valve opening. The front and rear limits of the control input specify the allowable change range of these control variables during the fuel switching process.

[0124] U lbIndicates the lower limit of fuel flow and the lower limit of valve opening. For the flow rates of diesel and gas, the lower limit represents the amount of fuel required to maintain the minimum stable operation of the engine. If the fuel flow is below this lower limit, the engine may experience unstable phenomena such as stalling and shaking, and cannot operate properly. For example, during the process of switching from gas to diesel, the diesel flow gradually increases. Before increasing to a certain value, if it is below the lower limit, the engine may not be able to sustain stable combustion. In addition, the lower limit of fuel flow is also related to the operating characteristics of equipment such as the fuel injection system. Some fuel injection systems may not be able to operate properly when the flow rate is too low, resulting in uneven fuel injection and affecting the engine performance. The lower limit of the opening of the diesel valve and the gas valve indicates the minimum opening degree of the valve. This limit is to prevent the valve from closing too much, resulting in fuel supply interruption. At the same time, the lower limit of valve opening is also related to the mechanical structure and control accuracy of the valve. If the valve opening is less than the lower limit, it may cause the valve to jam or be unable to accurately control the flow rate.

[0125] U ub Indicates the upper limit of fuel flow and the upper limit of valve opening. The upper limit of fuel flow is mainly determined by the design and safety requirements of the engine. Exceeding this upper limit, the engine may experience problems such as knocking and overheating due to excessive fuel supply, damaging engine components. For example, during the fuel switching process when the ship is accelerating, if the gas flow increases too quickly and exceeds the upper limit, it may cause incomplete combustion of the engine, generating a large amount of black smoke, and at the same time increasing the heat load of the engine, affecting its service life. In addition, the conveying capacity of the fuel supply system is also one of the factors limiting the upper limit of fuel flow. If the fuel flow exceeds the maximum conveying capacity of the supply system, it will lead to unstable fuel supply and affect the normal operation of the engine. The upper limit of valve opening indicates the maximum opening degree of the valve. Exceeding this upper limit may cause valve damage or inability to accurately control the fuel flow. Moreover, too large a valve opening may make the fuel supply too sudden and intense, having an adverse impact on the stability of the engine.

[0126] The state variable constraints conform to the formula:

[0127] X lb ≤X k ≤X ub ;

[0128] Specifically, the state variable reflects the internal state of the system. In the fuel switching scenario, it includes the engine speed, temperature, and pressure. The front and rear boundaries of the state variable specify the allowable range of change of these states during the fuel switching process.

[0129] X lbIndicates the lower limits of engine speed, temperature, and pressure. The lower limit of engine speed is the minimum speed required to ensure the normal operation of the engine and the basic power demand of the ship. If the speed is lower than this lower limit, the engine may lose power and be unable to drive the ship forward. At the same time, it may also cause the lubrication system, cooling system, etc. of the engine to malfunction, affecting the reliability of the engine. During the fuel switching process, it is necessary to ensure that the engine speed is not lower than the lower limit to maintain the stable navigation of the ship. For the coolant temperature, exhaust temperature, etc. of the engine, the lower limit represents the minimum temperature required for the normal operation of the engine. Too low a temperature may cause incomplete fuel combustion, increase pollutant emissions, and may also affect the lubrication performance of the engine, exacerbating component wear. For example, when switching fuels in cold weather, it is necessary to ensure that the engine reaches a certain temperature before switching to ensure normal fuel combustion. Intake pressure, fuel pressure, gas pressure, etc. all have their respective lower limits. The lower limit of pressure is a necessary condition to ensure the normal supply and combustion of fuel. If the intake pressure is too low, it will cause insufficient air intake by the engine, affecting the combustion effect of fuel. If the fuel or gas pressure is too low, it may cause poor fuel injection and inability to enter the engine combustion chamber normally.

[0130] X ub Indicates the upper limits of engine speed, temperature, and pressure. The upper limit of engine speed is determined by the mechanical structure and design performance of the engine. Exceeding this upper limit, the engine components will bear excessive stress and are prone to damage. Components such as pistons and connecting rods may break due to excessive speed. During the fuel switching process, it is necessary to strictly control the engine speed not to exceed the upper limit to ensure the safe operation of the engine. There are upper limit requirements for various temperature indicators of the engine. Too high a coolant temperature may cause the engine to overheat and damage components such as cylinder gaskets and cylinder liners. Too high an exhaust temperature may damage components of the exhaust system, such as turbochargers. During the fuel switching process, it is necessary to closely monitor the temperature changes and avoid exceeding the upper limit. The upper limits of intake pressure, fuel pressure, gas pressure, etc. are to prevent damage to the engine and related systems caused by excessive pressure. Excessive pressure may cause problems such as pipeline rupture and valve damage, affecting the safety and reliability of the system. During the fuel switching process, it is necessary to monitor and control the pressure in real time to ensure that it does not exceed the upper limit.

[0131] Simultaneously perform real-time rolling optimization. At each time step K, solve the above optimization problem to obtain the optimal control input U k and only apply the first control input U k . Subsequently, move the prediction window forward by one time step and repeat the above process. Through the above steps, a smooth switch between diesel and gas can be achieved.

[0132] In some embodiments, after the optimized output data is solved, fuel switching of the ship needs to be controlled through data transmission and reception, control instruction generation and distribution, actuator motion control, feedback regulation and real-time monitoring, and safety interlock and protection mechanisms.

[0133] Data transmission and reception: The output data is first transmitted from the calculation control unit to the ship's automation control system. Specific data interfaces and communication protocols are required to achieve reliable data transmission. After receiving the data, the controller in the ship automation control system parses it and converts the digital control instructions into signals that the system can understand and execute.

[0134] Control instruction generation and distribution: According to the parsed data, the controller generates specific control instructions. For example, for fuel flow control, an electrical signal corresponding to the output flow set value is generated, and the magnitude of this signal is proportional to the desired flow; for valve opening control, corresponding pulse signals or analog signals are generated to control the opening degree of the valve. The generated control instructions are distributed to each actuator, such as diesel pumps, gas pumps, diesel valve actuators, and gas valve actuators. Each actuator has its corresponding control channel to ensure that the control instructions can be accurately delivered to the target device.

[0135] Actuator motion control: The controller adjusts the speed or stroke of the diesel pump according to the output diesel flow control sequence, thereby changing the supply of diesel. For example, if the control instruction requires an increase in diesel flow, the controller sends a signal to the drive motor of the diesel pump to increase the motor speed, so that the diesel pump delivers more diesel to the engine. Similarly, for the gas pump, the controller adjusts its operating parameters according to the gas flow control sequence to ensure that the gas supply meets the control requirements. After receiving the valve opening control instruction from the controller, the valve actuator drives the valve to perform corresponding actions. The control method of the gas valve is similar to that of the diesel valve, and the actuator accurately adjusts the opening degree of the gas valve according to the control instruction to achieve gas flow control.

[0136] Feedback regulation and real-time monitoring: During the fuel switching process, various sensors installed in the engine and fuel system collect the operation status data of the system in real time and feed this data back to the controller. The controller compares the feedback signal with the output control target value and calculates the error. If there is a deviation between the actual flow and the set flow, the controller adjusts the control instruction according to the control algorithm to make the actual operation status gradually approach the target value, realizing precise fuel switching control. The ship's monitoring system will display various parameters and status information during the fuel switching process in real time, and the operator can understand the operation of the system at any time through the monitoring interface. If an abnormal situation occurs, the monitoring system will send an alarm signal to remind the operator to take measures in time.

[0137] Safety interlock and protection mechanism: To ensure the safety of the fuel switching process, the system sets a series of interlock control logics. For example, before switching to gas, it will check whether parameters such as the pressure and temperature of the gas system are normal. Only when all conditions are met, the gas valve is allowed to be opened. At the same time, during the process of closing the diesel valve, it will ensure the stability of parameters such as the engine speed and power to avoid power interruption. When the system encounters serious faults or dangerous situations, the safety protection mechanism will be immediately activated. For example, when it detects that the engine temperature is too high or fuel leakage occurs, the system will automatically cut off the fuel supply, stop the fuel switching process, and take corresponding emergency measures to ensure the safety of the ship and personnel.

[0138] Figure 3 Yes Figure 1 The flowchart of step S400 in Figure 1 Step S400 of includes but is not limited to steps S410 to S440:

[0139] Step S410: Obtain the real-time status data corresponding to the dual-fuel ship at the current moment;

[0140] Step S420: Update the first historical status data corresponding to the dual-fuel ship according to the real-time status data to obtain the second historical status data;

[0141] Step S430: Train the current moment fault prediction model with the second historical status data to obtain the target fault prediction model;

[0142] Step S440: Input the engine status data into the target fault prediction model for fault prediction to obtain the switching decision.

[0143] In some embodiments, the current moment fault prediction model is trained through steps S410 to S430 to obtain the target fault prediction model, and the switching decision is obtained through step S440.

[0144] In some embodiments, the target fault prediction model includes: a fault classification model, a leakage risk prevention and control model, and an external factor risk prevention and control model. Specifically, the fault classification model extracts time-domain features and frequency-domain features from the data collected by sensors based on all components of the liquefied natural gas system listed in detail, including storage tanks, pipelines, valves, pumps, and connecting components. For each component, it analyzes the possible fault modes and evaluates the impact on the ship operation. The leakage risk prevention and control model combines the sensor data monitored in real time, the engine room detection alarms, and the external data. According to the risk assessment index system in the model, it regularly compares the newly collected data with the model data and makes an assessment. According to the warning thresholds set for different risk levels, it adopts various warning methods to ensure that the operation and maintenance personnel can receive the warning information in time and make a response. The external factor risk prevention and control model obtains environmental data such as meteorological conditions, sea conditions, and traffic information in the ship's navigation area through the data interfaces of external institutions such as satellite communications, meteorological data providers, and maritime management departments accessed. By integrating relevant data, it issues a warning to the natural gas system before bad weather arrives, enabling the operation and maintenance personnel to judge whether to take necessary protective measures such as inspections and reinforcements.

[0145] More specifically, during the operation of a dual-fuel ship, sensors installed at key parts of the ship are used to collect real-time status data, which includes: engine operation parameter verification data, fuel characteristic verification data, environmental parameter verification data, and system status verification data. The data is transmitted to the ship's data processing system to form a set of real-time status data of the ship at the current moment. The obtained real-time status data is integrated into the existing first historical status data. If the first historical status data is stored in a database, the new real-time data is automatically classified and inserted according to the time sequence and data category. The newly collected data such as fuel pressure and temperature is inserted into the time series of the corresponding fuel status data; the engine speed and vibration data are supplemented into the engine operation status data record. In this way, the updated first historical status data is transformed into second historical status data, which completely records the change of the ship operation status over time.

[0146] For the training of the fault classification model, time-domain features and frequency-domain features are extracted from the second historical status data. For the fuel pressure data, time-domain statistics such as its mean and variance are calculated, and it is transformed into the frequency domain through Fourier transform to obtain the frequency distribution characteristics. For each component of the ship's liquefied natural gas system, such as storage tanks, pipelines, and valves, the characteristic changes in different operating states are analyzed respectively. Through supervised learning algorithms, such as support vector machine or decision tree algorithms, the data marked as normal and faulty are used as training samples. In the decision tree algorithm training, the data is divided according to the importance of the features to construct a decision tree model, enabling the model to accurately judge whether the operating state of the component is normal or faulty based on the input features and identify the specific fault type.

[0147] For the training of the leakage risk prevention and control model, a risk assessment index system is established by combining the real-time monitoring sensor data, engine room detection alarm data, and external environment data in the second historical state data. This system covers all aspects that may be involved in fuel leakage, such as fuel concentration, pressure change rate, equipment operation duration, etc. Using these data, according to the set risk assessment rules, the newly collected data is regularly compared and analyzed with the data in the model. If the fuel concentration is close to the warning threshold and the pressure change is abnormal, the model determines that the current state is at a high leakage risk. During the training process, the thresholds and rules for risk assessment are continuously optimized to improve the prediction accuracy of the model for leakage risks.

[0148] For the training of the external factor risk prevention and control model, environmental data such as meteorological conditions, sea conditions, and traffic information in the ship's navigation area are obtained from external data interfaces such as satellite communication, Automatic Identification System (AIS), and meteorological data providers. These external data are combined with the ship's own operating state data to analyze the risks that the natural gas system may face under different external environments. In strong wind weather, the ship's swaying may cause the pipe joints to loosen, increasing the leakage risk. Using supervised learning algorithms, the model is trained to learn from these data containing external factors and corresponding risk situations, and accurately issue a risk warning for the natural gas system according to the real-time obtained external data before the arrival of bad weather.

[0149] Integrate the trained fault classification model, leakage risk prevention and control model, and external factor risk prevention and control model to form a target fault prediction model. This comprehensive model can evaluate and predict the operating state of the ship's dual-fuel system from multiple perspectives and comprehensively identify potential faults. When new data input is received, the model will analyze it through each sub-model in turn to quickly determine whether there is a fault, the type of fault, and the degree of fault risk. Input the engine state data into the target fault prediction model. The model analyzes and processes the engine state data according to the knowledge and rules obtained from training. If the fault classification model detects that a certain component has a fault, the leakage risk prevention and control model evaluates that the leakage risk is high, and the external factor risk prevention and control model indicates that the external environment may exacerbate the fault risk, the model synthesizes this information to generate a switching decision, suggesting that the ship switch the fuel supply mode or stop adopting other corresponding measures such as the fuel switching strategy to ensure the safety of ship operation.

[0150] In some embodiments, a fuel switching strategy is sent to a control module according to a switching decision, so that the control module processes the fuel switching process according to the fuel switching strategy. When the switching decision determines that the ship is suitable for fuel switching, the fuel switching strategy is allowed to be transmitted to the control module, and the control module mobilizes the fuel pump and valves. When the switching decision determines that fuel switching is not suitable, the transmission of the fuel switching strategy to the control module is stopped, and a stop switching instruction or a switch to diesel instruction is sent to the control module until the switching decision determines that switching is allowed, and the latest fuel switching strategy is called and sent to the control module, and the control module mobilizes the fuel pump and valves to complete the fuel switching.

[0151] In some embodiments, new engine performance data is received at all times to update the second engine performance data, and at the same time, new real-time status data is received and the second historical status data is updated. According to the second engine performance data, a fuel switching strategy is obtained and updated through a target predictive control model; according to the second historical status data, a switching decision is obtained and updated through a target fault prediction model. The updated fuel switching strategy is sent to the control module according to the updated switching decision, and the control module processes the fuel switching process according to the updated fuel switching strategy.

[0152] As Figure 4 shown, a dual-fuel ship fuel switching device, the device includes:

[0153] A first module for obtaining first engine performance data and engine status data corresponding to the first engine performance data;

[0154] A second module for preprocessing the first engine performance data and the engine status data;

[0155] A third module for inputting the preprocessed first engine performance data into a target predictive control model for switching strategy prediction to obtain a fuel switching strategy corresponding to the engine;

[0156] A fourth module for inputting the engine status data into a target fault prediction model for fault prediction to obtain a switching decision;

[0157] A fifth module for sending a fuel switching strategy to a control module according to the switching decision, so that the control module processes the fuel switching process according to the fuel switching strategy;

[0158] A sixth module for receiving second engine performance data and adjusting the fuel switching strategy according to the second engine performance data.

[0159] It can be understood that the content in the above method embodiments is applicable to the device embodiments of the present application. The functions specifically implemented in the device embodiments of the present application are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those in the above method embodiments.

[0160] The embodiments of the present application also provide a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the above dual-fuel ship fuel switching method is implemented.

[0161] It can be understood that the content in the above method embodiments is applicable to the storage medium embodiments of the present application. The functions specifically implemented in the storage medium embodiments of the present application are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those in the above method embodiments.

[0162] As a non-transitory computer-readable storage medium, the memory can be used to store non-transitory software programs and non-transitory computer-executable programs. In addition, the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor through a network. Examples of the above networks include, but are not limited to, the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0163] The dual-fuel ship fuel switching method, device, and storage medium provided by the embodiments of the present application obtain a fuel switching strategy by obtaining first engine performance data and engine state data in cooperation with a target predictive control model, obtain a switching decision in cooperation with a fault prediction model based on the engine state data, and send the fuel switching strategy to a control module for fuel switching according to the switching decision. The switching decision ensures the safety of fuel switching, and the fuel switching strategy ensures the energy conservation, emission reduction, and economic requirements of fuel switching. The combination of the two jointly ensures the reliability of fuel switching.

[0164] The embodiments described in the embodiments of the present application are for more clearly illustrating the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application. Those skilled in the art will know that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of the present application are equally applicable to similar technical problems.

[0165] Those skilled in the art can understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than those shown in the figures, or combine certain steps, or different steps.

[0166] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0167] Those of ordinary skill in the art can understand that all or some of the steps in the methods disclosed above, and the functional modules / units in systems and devices, can be implemented as software, firmware, hardware, and appropriate combinations thereof.

[0168] It should be understood that in this application, "at least one (item)" means one or more, and "a plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships can exist. For example, "A and / or B" can mean: only A exists, only B exists, and both A and B exist at the same time. Among them, A and B can be singular or plural. The character " / " generally means that the associated objects before and after are in an "or" relationship. "At least one (one) of the following" or similar expressions refer to any combination of these items, including any combination of single item (one) or plural items (ones). For example, at least one (one) of a, b, or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0169] In addition, in each embodiment of this application, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0170] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in each embodiment of this application. The foregoing storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROM for short), random access memories (RAM for short), magnetic disks, or optical discs, and other various media that can store programs.

[0171] The preferred embodiments of the embodiments of the present application have been described above with reference to the accompanying drawings. This does not limit the scope of the rights of the embodiments of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and essence of the embodiments of the present application shall fall within the scope of the rights of the embodiments of the present application.

Claims

1. A dual-fuel ship fuel switching method, characterized in that, The method includes: Obtaining first engine performance data and engine state data corresponding to the first engine performance data; Preprocessing the first engine performance data and the engine state data; Inputting the preprocessed first engine performance data into a target predictive control model for switching strategy prediction to obtain a fuel switching strategy corresponding to the engine; Inputting the engine state data into a target fault prediction model for fault prediction to obtain a switching decision; Sending the fuel switching strategy to a control module according to the switching decision so that the control module processes the fuel switching process according to the fuel switching strategy; Receiving second engine performance data and adjusting the fuel switching strategy according to the second engine performance data.

2. The method according to claim 1, characterized in that, The first engine performance data includes: engine speed stability data, fuel consumption rate, and pollutant emission data; The engine state data includes: engine operation parameter verification data, fuel characteristic verification data, environmental parameter verification data, and system state verification data.

3. The method according to claim 1, wherein The step of inputting the preprocessed first engine performance data into a target predictive control model for switching strategy prediction to obtain a fuel switching strategy corresponding to the engine includes: Obtaining real-time engine performance data corresponding to the dual-fuel ship at the current moment; Updating first historical performance data corresponding to the dual-fuel ship according to the real-time engine performance data to obtain second historical performance data; Training a predictive control model at the current moment through the second historical performance data to obtain the target predictive control model; Inputting the preprocessed first engine performance data into the target predictive control model for switching strategy prediction to obtain the fuel switching strategy corresponding to the engine.

4. The method according to claim 3, wherein The construction and optimization process of the target predictive control model includes the following steps: Establishing a discrete-time state space model according to state variables, input variables, and output variables; Constructing a predictive model based on the discrete-time state space model; Establishing an optimization objective function according to the predictive model and performance indicators of fuel switching; Converting the problem to be solved by the target predictive control model into a standard quadratic programming problem and solving and optimizing it through an optimization algorithm to obtain the target predictive control model; Implementing control input at each time step, comparing, analyzing, and correcting according to output feedback, and dynamically optimizing the target predictive control model through rolling optimization.

5. The method according to claim 4, characterized in that, The calculation formula for the solution optimization under the constraint conditions is: Among them, U k represents the decision variable vector, H represents the Hessian matrix, and g represents the linear term coefficient vector.

6. The method according to claim 5, wherein The constraint conditions include control input conditions and state variable conditions; The control input conditions include: upper and lower limits of fuel flow rate and upper and lower limits of valve opening; The state variable conditions include: upper and lower limits of engine speed, upper and lower limits of temperature, and upper and lower limits of pressure.

7. The method according to claim 1, characterized in that, The step of inputting the engine state data into a target fault prediction model for fault prediction to obtain a switching decision includes: Obtaining real-time state data corresponding to the dual-fuel ship at the current moment; Updating first historical state data corresponding to the dual-fuel ship according to the real-time state data to obtain second historical state data; Training the current moment fault prediction model with the second historical state data to obtain a target fault prediction model; Inputting the engine state data into the target fault prediction model for fault prediction to obtain the switching decision.

8. The method according to claim 7, wherein The target fault prediction model includes: a fault classification model, a leakage risk prevention and control model, and an external factor risk prevention and control model.

9. A dual-fuel ship fuel switching device, characterized in that The device includes: A first module, configured to obtain first engine performance data and engine state data corresponding to the first engine performance data; A second module, configured to preprocess the first engine performance data and the engine state data; A third module, configured to input the preprocessed first engine performance data into a target predictive control model for switching strategy prediction to obtain a fuel switching strategy corresponding to the engine; A fourth module, configured to input the engine state data into a target fault prediction model for fault prediction to obtain a switching decision; A fifth module, configured to send the fuel switching strategy to a control module according to the switching decision, so that the control module processes the fuel switching process according to the fuel switching strategy; A sixth module, configured to receive second engine performance data and adjust the fuel switching strategy according to the second engine performance data.

10. A computer-readable storage medium storing a computer program, characterized in that, The computer program, when executed by a processor, implements the method according to any one of claims 1 to 8.

Images