Method and system for estimating oil consumption of large marine diesel engine in high impurity oil environment

Abstract

The present application provides a method and system for estimating the oil consumption of a large ship diesel engine in a high impurity oil return environment, relating to the technical field of oil consumption analysis. The present application synchronously collects the vibration of the engine body, surface temperature and high-pressure oil pipe fuel sound velocity through non-invasive sensors, extracts the synthesized vibration velocity, vibration main frequency, temperature rise rate and fuel sound velocity deviation value; loads a multi-physical field digital twin model composed of thermodynamics, structural dynamics, fuel sound velocity and oil consumption simulation model; combines the temperature rise rate residual and the synthesized vibration velocity to determine the starting state of the engine, and calibrates the operating condition through vibration characteristic mode matching; then, the fuel sound velocity deviation value is used to invert the fuel impurity characteristic parameter, and the oil consumption simulation model is combined with the speed and load input to finally output the instantaneous theoretical oil consumption rate corrected by the fuel quality. Through the cooperation of multi-physical field digital twin and sound velocity inversion impurity characteristics, the present application realizes non-invasive, high-robustness real-time oil consumption soft measurement.

Description

Technical Field

[0001] This invention relates to the field of fuel consumption analysis technology, specifically to a method and system for predicting the fuel consumption of a large marine diesel engine in a high-impurity return oil environment. Background Technology

[0002] To ensure full utilization of fuel during long-distance ocean voyages, large marine diesel engines commonly employ fuel return systems, recycling unburned fuel from the main engine back into the fuel supply lines. However, under complex operating conditions of prolonged high load and high back pressure, the return process inevitably introduces air, moisture, and combustion byproducts, creating a high-impurity return environment. This dynamic degradation of fuel quality directly alters combustion characteristics, causing actual fuel consumption to consistently deviate from the theoretical value calibrated in bench tests. For large vessels undertaking transoceanic transport missions, accurate fuel consumption measurement is directly related to voyage cost accounting and compliance assessment of ship energy efficiency operation indices. Therefore, accurately predicting instantaneous fuel consumption of diesel engines under this specific high-impurity return environment is a critical technical challenge that urgently needs to be addressed in ship operation and management.

[0003] Currently, fuel consumption measurement for large marine diesel engines primarily relies on physical flow meters or open-loop estimation models calibrated through bench tests. While physical flow meters, such as Coriolis mass flow meters, offer high accuracy, they are extremely sensitive to air bubbles and impurities in the fuel. In high-impurity return fuel environments, they are prone to measurement signal drift and clogging failures, requiring frequent offline cleaning and calibration, making it difficult to meet the demands of long-term, stable online monitoring. Traditional open-loop estimation models typically use only engine speed and load as inputs, failing to detect and respond to real-time degradation in fuel quality. Some methods attempt to introduce small amounts of temperature or pressure parameters for correction, but their correction logic is based on the assumption of steady-state and clean fuel, essentially a passive, non-closed-loop, coarse compensation scheme that cannot accurately reflect the nonlinear impact of fuel impurity characteristics on the combustion process from a mechanistic perspective. Existing technologies cannot capture weak temperature rise, vibration, and acoustic signals from the engine block and high-pressure fuel lines without direct contact with the fuel. Through deep coupling and inversion of physical models, key characteristic parameters (such as equivalent air content) characterizing the degree of return fuel impurities can be calculated in real time, and the combustion model can be dynamically corrected accordingly. It is precisely this lack of perception across physical fields and insufficient model coupling that leads to a serious deficiency in the accuracy and robustness of existing fuel consumption assessment methods in the specific scenario of high-impurity oil return.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for predicting the fuel consumption of large marine diesel main engines in high-impurity return oil environments, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A method for predicting the fuel consumption of a large marine diesel engine in a high-impurity return oil environment, comprising the following steps: Step 1: During the simulation cycle, the operating parameters of the ship's diesel engine are acquired synchronously, and real-time feature extraction is performed to generate synthetic vibration velocity, vibration main frequency, temperature rise rate, and fuel sound velocity deviation values. Step 2: Load the pre-built multiphysics digital twin model of the marine diesel engine. The multiphysics digital twin model includes: a thermodynamic simulation model to describe the theoretical temperature rise curve under various operating conditions; a structural dynamics simulation model to describe the theoretical vibration response characteristics under various operating conditions; a fuel sound velocity simulation model to describe the theoretical mapping relationship between fuel impurity characteristic parameters and fuel sound velocity deviation values; and a fuel and fuel consumption simulation model to respond to boundary conditions and output the theoretical fuel consumption rate. Step 3: Determine whether the initial state of the main engine is a cold start based on the initial surface temperature of the engine body. If so, compare the temperature rise rate with the theoretical temperature rise curve from the cold start state using residual analysis. Based on the comparison results and the synthesized vibration velocity, determine whether the marine diesel engine meets the simulation start-up conditions. When the conditions are met, perform pattern matching between the synthesized vibration velocity, the dominant vibration frequency, and the theoretical vibration response characteristics of the structural dynamics simulation model under various operating conditions to calibrate the current operating conditions. Step 4: Using the fuel sound velocity deviation value as input, perform parameter inversion in the fuel sound velocity simulation model to solve for the fuel impurity characteristic parameters; use the speed, load and fuel impurity characteristic parameters corresponding to the current operating conditions as boundary conditions, input them into the fuel and fuel consumption simulation model, and obtain the instantaneous theoretical fuel consumption rate through simulation calculation.

[0007] Furthermore, the operating parameters are achieved through wireless composite sensors deployed on the marine diesel engine. The wireless composite sensors are magnetically attached to the surface of the marine diesel engine and clamped to the outer wall of the high-pressure oil pipe in a clamping manner, so as to synchronously collect the triaxial vibration velocity of the engine body, the surface temperature of the engine body, and the instantaneous sound velocity of fuel at the outer wall of the high-pressure oil pipe in a non-invasive manner. The synthesized vibration velocity is calculated from the vector sum of the three-axis vibration velocities of the engine body; the dominant vibration frequency is extracted by spectral analysis of the vibration velocity signal; the temperature rise rate is calculated from the temperature difference of the engine body surface in adjacent sampling periods; and the fuel sound velocity deviation value is calculated from the normalized difference between the instantaneous fuel sound velocity and the theoretical pure fuel sound velocity under the same temperature conditions. The theoretical pure fuel sound velocity refers to the sound velocity value of standard fuel without impurities under the same temperature conditions. The fuel sound velocity deviation value is calculated from the normalized difference between the instantaneous fuel sound velocity and the theoretical pure fuel sound velocity.

[0008] Furthermore, the thermodynamic simulation model is constructed as follows: using the surface temperature of the marine diesel engine under various operating conditions such as cold start, low idling load, economic cruise, heavy load variable operating conditions, and shutdown cooling, a nonlinear temperature rise model of the surface temperature of the engine changing with operating time is established based on the Arrhenius formula, and the corresponding temperature rise curve parameters are fitted for different operating conditions to form a theoretical temperature rise curve.

[0009] Furthermore, the structural dynamics simulation model is constructed as follows: historical data of the synthetic vibration velocity and dominant vibration frequency of the ship's diesel engine under various operating conditions are collected. Based on a pre-established simplified multi-degree-of-freedom parameter model, the correspondence between rotational speed and dominant vibration frequency, as well as the correspondence between rotational speed, load, and synthetic vibration velocity, are established to form theoretical vibration response characteristics under different operating conditions. Among them, the dominant vibration frequency is used to back-estimate the real-time rotational speed, and the synthetic vibration velocity is used to match and determine the real-time load.

[0010] Furthermore, the fuel sound velocity simulation model is constructed as follows: under laboratory conditions, a family of sound velocity curves showing the change of fuel sound velocity with temperature under different fuel impurity content ratios are measured, and a mapping model is established with engine surface temperature and fuel impurity characteristic parameters as inputs and fuel sound velocity deviation value as output; the fuel impurity characteristic parameters include equivalent air content or return fuel mixing ratio, which are used to characterize the degree of fuel quality degradation in a high-impurity return fuel environment.

[0011] Furthermore, the fuel and fuel consumption simulation model is constructed as follows: based on the modified Weber combustion model, the fuel and fuel consumption simulation model is established with speed, load, and fuel impurity characteristic parameters as input boundary conditions; the fuel and fuel consumption simulation model pre-stores the reference combustion efficiency curve and reference heat release rate curve obtained through bench test calibration, and corrects the reference combustion efficiency curve and reference heat release rate curve according to the input fuel impurity characteristic parameters, and simulates and outputs the instantaneous theoretical fuel consumption rate under the operating conditions.

[0012] Furthermore, the method for determining cold start is as follows: When the initial state of the main engine is determined to be a cold start based on the initial surface temperature of the engine body, the temperature rise rate is compared point by point with the theoretical temperature rise curve under the cold start condition in the thermodynamic simulation model on the same time series to calculate the temperature rise rate residual; at the same time, the synthesized vibration velocity is compared with the preset shutdown noise benchmark value; only when the temperature rise rate residual is within the preset range for multiple consecutive sampling periods and the synthesized vibration velocity exceeds the shutdown noise benchmark value is the marine diesel main engine determined to have entered the start-up state.

[0013] Furthermore, the specific method for calibrating the current operating condition is as follows: when it is determined that the ship's diesel engine has entered the start-up state, the synthetic vibration velocity and vibration dominant frequency are matched with the theoretical vibration response characteristics under each operating condition in the structural dynamics simulation model, and the operating condition with the highest similarity is taken as the current operating condition.

[0014] Furthermore, the fuel sound velocity deviation value calculated in real time is input into the fuel sound velocity simulation model. The engine surface temperature corresponding to the fuel sound velocity acquisition time is used as the input condition to inversely solve the current equivalent air content or return fuel mixing ratio as the characteristic parameter of fuel impurities. Determine the speed and load corresponding to the current operating conditions, and input the speed, load, and fuel impurity characteristic parameters obtained from the inversion solution as boundary conditions into the fuel and fuel consumption simulation model. First, the reference combustion efficiency curve is corrected by the fuel impurity characteristic parameters, and then the corrected reference combustion efficiency curve is coupled with the reference heat release rate curve to simulate and obtain the instantaneous theoretical fuel consumption rate of the current operating conditions.

[0015] The present invention also provides a system for predicting the fuel consumption of a large marine diesel engine in a high-impurity oil return environment. This system is used to implement the aforementioned method for predicting the fuel consumption of a large marine diesel engine in a high-impurity oil return environment, comprising: The multi-source signal real-time feature extraction module is used to synchronously acquire the operating parameters of the ship's diesel engine during the simulation cycle, extract its features in real time, and generate synthetic vibration velocity, vibration main frequency, temperature rise rate, and fuel sound velocity deviation value. A multiphysics digital twin model construction module is used to load a pre-built multiphysics digital twin model of a marine diesel engine. The multiphysics digital twin model includes: a thermodynamic simulation model to describe the theoretical temperature rise curve under various operating conditions; a structural dynamics simulation model to describe the theoretical vibration response characteristics under various operating conditions; a fuel sound velocity simulation model to describe the theoretical mapping relationship between fuel impurity characteristic parameters and fuel sound velocity deviation values; and a fuel and fuel consumption simulation model to respond to boundary conditions and output the theoretical fuel consumption rate. The operating condition start-up determination and calibration module is used to determine whether the initial state of the main engine is a cold start based on the initial surface temperature of the engine body. If so, the temperature rise rate is compared with the theoretical temperature rise curve from the cold start state. Based on the comparison result and the synthetic vibration velocity, it is determined whether the marine diesel engine meets the simulation start-up conditions. When the conditions are met, the synthetic vibration velocity, vibration frequency and the theoretical vibration response characteristics of the structural dynamics simulation model under each operating condition are pattern matched to calibrate the current operating condition. The inversion and theoretical fuel consumption rate simulation module is used to perform parameter inversion in the fuel sound velocity simulation model with the fuel sound velocity deviation value as input, and solve for the characteristic parameters of fuel impurities. The speed, load and fuel impurity characteristic parameters corresponding to the current operating conditions are used as boundary conditions and input into the fuel and fuel consumption simulation model to obtain the instantaneous theoretical fuel consumption rate through simulation calculation.

[0016] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention directly addresses the persistent technical problems of physical flow meters being prone to clogging and drifting in high-impurity return oil, and the inability of traditional open-loop models to detect dynamic deterioration of fuel quality, achieving a breakthrough effect. This invention employs a magnetic and clamp-type non-invasive wireless composite sensor, simultaneously acquiring engine temperature rise, vibration, and fuel sound velocity signals without disassembling the oil circuit or contacting the fuel, completely avoiding sensor interference with the fuel system and achieving maintenance-free long-term online monitoring in harsh engine room environments. Secondly, by comparing the temperature rise rate with the theoretical temperature rise curve of cold start, and the vibration characteristics with the theoretical vibration response characteristics using dual time-series residuals, it can accurately eliminate false alarms related to shutdown noise and reliably determine the true start-up status of the main engine. This invention uses the deviation between engine surface temperature and fuel sound velocity to invert fuel impurity characteristic parameters within a multi-physics digital twin framework, and uses this to dynamically correct the combustion model, achieving closed-loop adaptive adjustment of fuel consumption prediction to fuel quality deterioration. Ultimately, it transforms real-time fuel consumption, which cannot be directly measured, into a high-confidence instantaneous theoretical fuel consumption rate, providing solid data support for precise energy efficiency management and intelligent navigation decision-making in large ships. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall method flow of the present invention; Figure 2 This is a schematic diagram showing the changes of the synthesized vibration velocity and the measured temperature rise rate over time during the cold start-up process of this invention; Figure 3 This is a schematic diagram of the system structure of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0019] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0020] Example: Please see Figures 1 to 2 The present invention provides a technical solution: A method for predicting the fuel consumption of a large marine diesel engine in a high-impurity return oil environment, comprising the following steps: Step 1: During the simulation cycle, the operating parameters of the ship's diesel engine are acquired synchronously, and real-time feature extraction is performed to generate synthetic vibration velocity, vibration main frequency, temperature rise rate, and fuel sound velocity deviation values.

[0021] In this embodiment, the operating parameters are achieved through a wireless composite sensor deployed on the marine diesel engine. The wireless composite sensor is magnetically attached to the surface of the marine diesel engine and clamped to the outer wall of the high-pressure fuel line using a clamping method. It synchronously collects the three-axis vibration velocity of the engine body, the surface temperature of the engine body, and the instantaneous sound velocity of the fuel at the outer wall of the high-pressure fuel line in a non-invasive manner. The wireless composite sensor integrates a three-axis MEMS vibration sensing unit, a contact thermocouple temperature sensing unit, and a pair of ultrasonic transducers for sound velocity measurement. The acquisition actions of each sensing unit are triggered by a unified microcontroller via a synchronous clock, ensuring that each set of vibration velocity, surface temperature, and instantaneous fuel sound velocity data is strictly aligned in timestamps, forming a complete data frame for a simulation cycle. The simulation cycle, i.e., the data frame acquisition interval, is set to a fixed value between 0.1 seconds and 1 second, preferably 0.5 seconds.

[0022] The synthesized vibration velocity is calculated from the vector sum of the three-axis vibration velocities of the host machine. Specifically, if the three-axis vibration velocities of the host machine collected within a simulation cycle are the radial velocity, axial velocity, and vertical velocity, then the synthesized vibration velocity is the square root of the sum of the squares of these three velocities. This calculation is performed directly in the microcontroller to reduce the load on wireless data transmission. This synthesized value reflects the overall intensity of the host machine's vibration and is used for subsequent load matching with the structural dynamics simulation model.

[0023] The dominant vibration frequency is extracted through spectral analysis of the vibration velocity signal. To balance real-time computation and frequency resolution, the microcontroller's built-in Fast Fourier Transform module performs spectral analysis on a sliding time window of the synthesized vibration velocity signal, with a length that is an integer multiple of one simulation period, such as 2 to 5 seconds. The spectral resolution is set between 0.5 Hz and 2 Hz. Within a preset frequency band of interest, such as 5 Hz to 200 Hz for low-speed marine diesel engines and 10 Hz to 500 Hz for medium-speed engines, the frequency point with the largest amplitude is searched and taken as the dominant vibration frequency at that moment. This dominant vibration frequency has a deterministic correspondence with the engine speed. For example, for a four-stroke, four-cylinder diesel engine, its ignition frequency is the engine speed multiplied by the number of cylinders and then divided by 120. The dominant vibration frequency usually corresponds to this ignition frequency or its harmonics.

[0024] The temperature rise rate is calculated from the difference in body surface temperature between adjacent sampling periods. In a simulation period, the temperature rise rate at the current moment is equal to the difference between the body surface temperature collected in the current simulation period and the body surface temperature collected in the previous simulation period, divided by the duration of the simulation period. To avoid amplifying the difference results due to measurement noise, a recursive average filter can be applied to the original body surface temperature sequence, with a filter window length of 3 to 5 simulation periods.

[0025] The fuel sound velocity deviation value is calculated from the normalized difference between the instantaneous fuel sound velocity and the theoretical pure fuel sound velocity under the same temperature conditions; the theoretical pure fuel sound velocity refers to the sound velocity value of standard fuel without impurities under the same temperature conditions. This standard fuel refers to new fuel that conforms to a specific grade in the international standard ISO 8217 and is free from backfill impurities.

[0026] The formula for calculating the normalized difference is as follows: in, This represents the deviation value of the fuel sound velocity, which is a dimensionless quantity. This represents the instantaneous sound velocity of fuel collected during the current simulation cycle, in meters per second. The theoretical pure sound velocity of the standard fuel is given at a temperature T equal to the surface temperature of the aircraft at that moment, expressed in meters per second. As a marker to distinguish between theoretical and measured values, it is used to indicate the instantaneous state of the current simulation cycle.

[0027] in, The method for determining the value is as follows: Under laboratory conditions, the sound velocity value of a selected grade of standard pure oil sample is measured using a sound velocity measuring instrument, and the data of sound velocity changing with temperature are then analyzed using a quadratic polynomial. Perform least squares fitting to obtain the model coefficients. And store it inside the microcontroller.

[0028] For example: for a certain batch of standard pure fuel oil, the measured coefficient Values , The value is -4.1. The value is 1450. When the collected temperature T is 80℃, the calculated value is... m / s; if the instantaneous speed of sound of fuel is measured at the same moment If the speed is 1050 m / s, then the deviation value of the fuel sound speed is... This positive value directly quantifies the degree to which the sound velocity is reduced due to impurities in the returned fuel; the larger the deviation value, the more severe the deterioration of fuel quality.

[0029] The fuel sound velocity deviation value is calculated from the normalized difference between the instantaneous fuel sound velocity and the theoretical pure fuel sound velocity.

[0030] Step 2: Load the pre-built multiphysics digital twin model of the marine diesel engine. The multiphysics digital twin model includes: a thermodynamic simulation model to describe the theoretical temperature rise curve under various operating conditions; a structural dynamics simulation model to describe the theoretical vibration response characteristics under various operating conditions; a fuel sound velocity simulation model to describe the theoretical mapping relationship between fuel impurity characteristic parameters and fuel sound velocity deviation values; and a fuel and fuel consumption simulation model to respond to boundary conditions and output the theoretical fuel consumption rate.

[0031] The multiphysics digital twin model is deployed on a shipboard host computer or a shore-based cloud server, with each sub-model existing as a callable function module or data table. During model loading, the corresponding calibrated model instance is indexed and matched from the model library based on the identification parameters of the current ship's diesel engine, such as model number, number of cylinders, number of strokes, rated power, and rated speed.

[0032] In this embodiment, the thermodynamic simulation model is constructed as follows: the surface temperature of the marine diesel engine is used under various operating conditions such as cold start, low idling load, economic cruise, heavy load variable conditions, and shutdown cooling. A nonlinear temperature rise model of the surface temperature of the engine changing with operating time is established based on the Arrhenius formula, and the corresponding temperature rise curve parameters are fitted for different operating conditions to form a theoretical temperature rise curve.

[0033] The specific expression for this nonlinear temperature rise model is as follows: in, Indicates the start of the operating condition The theoretical surface temperature of the organism at any given time, in degrees Celsius; This represents the thermal equilibrium temperature under this operating condition, i.e., theoretically... The steady-state surface temperature reached when approaching infinity, expressed in degrees Celsius. This indicates the initial surface temperature of the machine body when entering this operating condition, in degrees Celsius. This indicates the running time from the start of this operating condition, in seconds; This represents the rate of temperature rise coefficient, expressed as the reciprocal of the rate per second. This coefficient characterizes the speed at which an organism approaches thermal equilibrium, as determined by the Arrhenius equation. It is confirmed that, among them, Pre-exponential factor, Equivalent activation energy; is the gas constant.

[0034] For each operating condition and The two parameters are determined by conducting multiple actual machine or bench tests under this operating condition, recording the surface temperature of the machine body from the initial temperature. The complete time series data, rising to near steady state, was fitted using the nonlinear least squares method with the aforementioned nonlinear temperature rise model to obtain the optimal temperature rise under this operating condition. and Value selection. For example, during an economic cruise test of a certain type of low-speed marine diesel engine, a temperature sequence data of 1800 seconds was obtained, which was then fitted to obtain... , The reciprocal of each second. This forms the theoretical temperature rise curve for the economic cruise condition, used for residual comparison in step 3.

[0035] The equivalent activation energy is determined as follows: At least two temperature rise tests are conducted under the same operating condition with different initial temperatures to obtain at least two sets of time-series data on the change of the machine surface temperature over time; each set of data is fitted using the aforementioned nonlinear temperature rise model to obtain the corresponding temperature rise rate coefficient. and thermal equilibrium temperature Taking the natural logarithm of both sides of the Arrhenius formula yields a linear relationship. ; each group The slope of the straight line obtained by performing linear regression on the data is... Multiply the slope by the gas constant. And take the absolute value, that is, calibrate to obtain the equivalent activation energy. The value of . Among them, the gas constant. This is a universal physical constant, with a value of 8.314 J / (mol·K). The equivalent activation energy, calibrated using the above method... The thermal inertia of complex thermal processes such as in-cylinder combustion heat release, cooling water heat exchange, and engine body heat conduction is equivalent to a macroscopic parameter, enabling the nonlinear temperature rise model to accurately extrapolate the temperature rise behavior at different initial temperatures under limited experimental data conditions.

[0036] In this embodiment, the structural dynamics simulation model is constructed as follows: historical data of the synthetic vibration velocity and vibration dominant frequency of the ship's diesel engine under various operating conditions are collected. Based on a pre-established simplified multi-degree-of-freedom parameter model, the correspondence between rotational speed and vibration dominant frequency, as well as the correspondence between rotational speed, load, and synthetic vibration velocity, are established to form theoretical vibration response characteristics under different operating conditions. Among them, the vibration dominant frequency is used to back-estimate the real-time rotational speed, and the synthetic vibration velocity is used to match and determine the real-time load.

[0037] The simplified multi-degree-of-freedom parametric model is a lumped parametric model that equates the main body, vibration isolators, and hull base to lumped mass and spring-damped elements, respectively. This model is used to simulate the vibration response of the main body under known excitation force frequency and amplitude conditions. However, its reverse application, i.e., the deduction of operating parameters from measured vibration signals, is achieved by directly establishing data mapping relationships.

[0038] The relationship between rotational speed and vibration frequency is established by extracting data pairs of vibration frequency and measured rotational speed from historical data under various stable operating conditions and establishing a linear regression equation. For a four-stroke N-cylinder diesel engine, the following equation exists: in, The dominant frequency of vibration is expressed in Hertz (Hz). This refers to the main unit's rotational speed, measured in revolutions per minute. This refers to the number of cylinders; This is the order coefficient, typically set to 1 to represent the ignition fundamental frequency and 2 to represent the multiplier. In practice, the order with the strongest correlation and most stable amplitude from historical data is selected as the basis for back-calculating the engine speed. For example, for a six-cylinder four-stroke engine, if the ignition fundamental frequency (m=1) is determined to be the optimal order, then the formula for back-calculating the engine speed is: When the measured dominant vibration frequency is 45 Hz, the real-time rotational speed is calculated to be 900 revolutions per minute.

[0039] The correspondence between rotational speed, load, and synthetic vibration velocity is stored in the form of a three-dimensional data table or a surface fitting function. The horizontal axis represents rotational speed, the vertical axis represents the load factor (i.e., the current load as a percentage of rated power), and the table values ​​are the corresponding synthetic vibration velocities. This table is constructed by interpolating and smoothing the rotational speed, load factor, and synthetic vibration velocity data collected under various stable operating conditions. During load matching, the real-time load factor can be determined by performing a two-dimensional lookup or inverse distance weighted interpolation in the data table based on the inferred real-time rotational speed and the measured synthetic vibration velocity.

[0040] In this embodiment, the fuel sound velocity simulation model is constructed as follows: under laboratory conditions, a family of sound velocity curves showing the change of fuel sound velocity with temperature under different fuel impurity content ratios are measured, and a mapping model is established with engine surface temperature and fuel impurity characteristic parameters as inputs and fuel sound velocity deviation value as output; the fuel impurity characteristic parameters include equivalent air content or return fuel mixing ratio, which are used to characterize the degree of fuel quality degradation in a high impurity return fuel environment.

[0041] The equivalent air content is defined as the equivalent volume fraction of microbubbles and dissolved air in the fuel after return fuel mixing, expressed as a percentage; the return fuel mixing ratio is defined as the percentage of return fuel mass to the total fuel mass in the high-pressure fuel line. The two parameters can be converted through calibration. The return fuel mixing ratio is preferred as a characteristic parameter of fuel impurities because it is more readily available on board.

[0042] The mapping model is stored in the form of a two-dimensional interpolation table. The two input dimensions of the table are the engine surface temperature (resolution 5℃ or 10℃) and the return fuel mixing ratio (resolution 5% or 10%). The output value of the table is the fuel sound velocity deviation value. The specific operation for constructing this table is as follows: Under constant temperature conditions in the laboratory, a series of fuel samples with return fuel mixing ratios of 0%, 10%, 20%, 30%, 40%, and 50% are configured. For each sample, its sound velocity value is measured in 10℃ increments within a temperature range of 30℃ to 120℃. Then, according to the normalized difference formula given in step 1, the fuel sound velocity deviation value under different temperatures and different return fuel mixing ratios is calculated and filled into the interpolation table. For example, under the conditions of 80℃ temperature and 30% return fuel mixing ratio, the measured and calculated corresponding fuel sound velocity deviation value is 0.052, which is then stored in the corresponding grid cell. In step 4, the table takes the engine surface temperature and fuel sound velocity deviation value as inputs and inverts the fuel impurity characteristic parameters through reverse lookup or optimization solution.

[0043] In this embodiment, the fuel and fuel consumption simulation model is constructed as follows: based on the modified Weber combustion model, the fuel and fuel consumption simulation model is established with speed, load and fuel impurity characteristic parameters as input boundary conditions; the fuel and fuel consumption simulation model pre-stores the reference combustion efficiency curve and reference heat release rate curve obtained through bench test calibration, and corrects the reference combustion efficiency curve and reference heat release rate curve according to the input fuel impurity characteristic parameters, and simulates and outputs the instantaneous theoretical fuel consumption rate under the operating conditions.

[0044] The baseline combustion efficiency curve and baseline heat release rate curve were obtained through bench tests at various stable speeds and load points using standard pure fuel (i.e., a 0% return fuel mixture ratio). The baseline combustion efficiency curve records the effective thermal efficiency at different speeds and loads; the baseline heat release rate curve records the instantaneous heat release rate distribution corresponding to the crankshaft angle under the corresponding operating conditions. These two baseline curves are stored in a two-dimensional lookup table format, indexed by speed and load rates.

[0045] The correction method involves introducing a combustion efficiency correction factor, which is defined as a function of fuel impurity characteristic parameters. A specific form is... In the formula, This is a combustion efficiency correction factor, dimensionless, with a value range of (0,1]. The impurity influence coefficient is obtained through regression testing data under different oil return mixing ratios, with typical values ​​ranging from 0.1 to 0.5. The corrected combustion efficiency curve is equal to each value in the baseline combustion efficiency curve multiplied by this combustion efficiency correction factor. The corrected heat release rate curve is obtained by multiplying the baseline heat release rate curve by the same correction factor and performing phase shifting based on the measured combustion duration variation.

[0046] The calculation process for the instantaneous theoretical fuel consumption rate output by simulation is as follows: Based on the current real-time engine speed and load rate, the baseline effective thermal efficiency for using standard pure fuel is retrieved from the pre-stored baseline combustion curve. Then, based on the fuel impurity characteristic parameters obtained through inversion, the current effective thermal efficiency is determined from the corrected combustion efficiency curve, i.e., the current effective thermal efficiency = combustion efficiency correction factor. Baseline effective thermal efficiency.

[0047] According to the calibrated load factor and the rated power of the diesel engine (Unit: kilowatts) Calculate the current required effective engine power. : Combining the corrected effective thermal efficiency, effective power, and the physicochemical property constants of the fuel (i.e., the lower heating value of the fuel, denoted as...) (Unit: kilojoules per kilogram) Calculate the instantaneous theoretical fuel consumption rate. The formula is: in, The unit is kilograms per second, and the instantaneous theoretical fuel consumption rate in kilograms per hour can be obtained by unit conversion (×3600). This indicates the corrected effective thermal efficiency.

[0048] Step 3: Determine whether the initial state of the main engine is a cold start based on the initial surface temperature of the engine body. If so, compare the temperature rise rate with the theoretical temperature rise curve when starting from a cold state. Based on the comparison results and the synthesized vibration velocity, determine whether the marine diesel engine meets the simulation start-up conditions. When it does, perform pattern matching between the synthesized vibration velocity, the dominant vibration frequency and the theoretical vibration response characteristics of the structural dynamics simulation model under various operating conditions to calibrate the current operating conditions.

[0049] The prerequisite for cold start determination is that the initial engine surface temperature collected in the current simulation cycle is lower than a preset cold start temperature threshold. This cold start temperature threshold is determined by recording the ambient temperature statistics of a specific type of marine diesel engine in its shutdown state, and then adding a preset temperature rise increment to this statistical value. The significance of introducing this preset temperature rise increment is to construct a fuzzy upper bound for judgment, thereby solving the problem of misjudgment caused by ambient temperature fluctuations when relying solely on absolute temperature thresholds. On the one hand, it compensates for the thermal stratification effect of the actual ship's engine room environment and the measurement deviation of sensors in the near-room temperature range, forming a protective zone to prevent frequent false triggering of the cold start identification process due to ambient temperature fluctuations in high-temperature sea areas during summer. On the other hand, considering the large thermal inertia of the main engine, only when the surface temperature drops below this threshold, which is slightly higher than the average ambient temperature, does it indicate that the internal core hot spots have been sufficiently cooled, thus providing an effective and repeatable initial benchmark for subsequent theoretical temperature rise curve comparisons, ensuring the timeliness of the start determination.

[0050] A typical example of setting this value is as follows: taking the average ambient temperature of the cabin (35℃) and adding a temperature increase increment of 5℃, the cold-state temperature threshold is 40℃. When the initial surface temperature of the fuselage is below 40℃, the host is determined to be in a cold state and enters the cold-state startup identification process; otherwise, it is determined to be in a hot state and directly proceeds to the operating condition calibration stage.

[0051] In this embodiment, the method for determining cold start is as follows: Once the initial state of the main engine is determined to be a cold start based on the initial surface temperature of the engine body, the temperature rise rate is compared point by point with the theoretical temperature rise curve under the cold start condition in the thermodynamic simulation model on the same time series to calculate the temperature rise rate residual. At the same time, the synthesized vibration velocity is compared with the preset shutdown noise benchmark value. Only when the temperature rise rate residual is within the preset range for multiple consecutive sampling periods and the synthesized vibration velocity exceeds the shutdown noise benchmark value is the ship's diesel main engine determined to have entered the start-up state, in order to avoid misjudgment caused by instantaneous vibration spikes caused by wave impact or hull flutter during ship navigation.

[0052] In this embodiment, a cold start test was conducted on a marine four-stroke six-cylinder diesel engine to verify the anti-false alarm capability in the cold start determination process. The test environment, engine room temperature, was 35°C, and the initial surface temperature of the main engine stabilized at 35.5°C, which was lower than the preset cold start temperature threshold of 40°C, thus the initial state was determined to be cold. The simulation period was set to 0.5 seconds, and the shutdown noise baseline value was pre-calibrated to 1.4 mm / s. The theoretical temperature rise curve parameters for the cold start condition were obtained through bench testing: thermal equilibrium temperature was 80°C, initial temperature was 35°C, and temperature rise rate coefficient was 0.008. The theoretical temperature rise rate is calculated in real time using the first derivative of the nonlinear temperature rise model with respect to time. The preset allowable range for the residual temperature rise rate is an upper deviation of +0.5℃ / min and a lower deviation of -0.3℃ / min. The number of sampling periods required for continuous judgment is set to 5, meaning that both conditions must be continuously met within a 2.5-second time window to confirm entry into the start-up state. Specifically, the complete data sequence from shutdown waiting to successful start-up judgment and then transition to idle low-load conditions during this cold start process was recorded, as shown in the table below: Table 1. Example of joint criterion triggering logic data for cold start process: Combining Table 1 above and Figure 2As shown, during the time interval from 0s to 2s, the main engine is in a stopped waiting state. The synthesized vibration velocity fluctuates below the stop noise baseline of 1.4mm / s, and the measured temperature rise rate fluctuates randomly around zero. Although the residual temperature rise rate is within the preset range, the system maintains the stop judgment because the vibration conditions are not met. At 2.5s, the start-up motor runs, and the synthesized vibration velocity jumps instantly to 3.5mm / s, significantly exceeding the 1.4mm / s baseline. However, at this time, in-cylinder combustion has not yet stabilized, and there is a significant deviation between the measured temperature rise rate and the theoretical temperature rise rate curve. The residuals reach 0.26℃ / min, 0.88℃ / min, and 0.7℃ / min respectively between 3s and 4s, all exceeding the preset upper deviation limit. The data in this stage shows that if only a single vibration threshold criterion is relied upon, the system will mistakenly judge that the main engine has entered the start-up state at 2.5s, when in fact combustion heat generation has not yet entered a normal pattern. This misjudgment will lead to subsequent operating condition calibration and fuel consumption calculations based on an incorrect time starting point. As the combustion process stabilizes, starting from 4.5s, the measured temperature rise rate falls back into the allowable deviation zone of the theoretical temperature rise rate curve, while the synthesized vibration velocity remains consistently above the 1.4 mm / s reference value. By 6.5s, these two conditions have been met for five consecutive sampling cycles, at which point the system reliably determines that it has entered the start-up state. Thereafter, the vibration velocity stabilizes in the range of 5.5–8.8 mm / s, the temperature rise rate decays exponentially with the cylinder block's thermal inertia, and the deviation from the theoretical curve gradually converges. Based on this, the system further completes the calibration of its operating conditions.

[0053] Therefore, this embodiment introduces a dynamic comparison between the residual temperature rise rate and the theoretical temperature rise curve as a second-dimensional criterion. This effectively eliminates false triggering of a single vibration threshold caused by vibration shock or other external interference during startup, ensuring that a successful startup is only confirmed under the dual physical conditions of a genuinely established combustion heat generation law within the main engine cylinder and a vibration level synchronously reaching the operating state. This joint determination mechanism significantly improves the accuracy and robustness of cold-state startup identification, providing a reliable time benchmark for subsequent fuel consumption prediction based on operating conditions.

[0054] The above-mentioned discrimination logic employs a combined approach of thermodynamic and vibration characteristics. Its rationale lies in the fact that a single vibration threshold judgment cannot distinguish between the continuous vibration generated by the actual ignition and start-up of the main unit and the instantaneous vibration spikes caused by external random impacts. By introducing a comparison of the residual between the temperature rise rate and the theoretical temperature rise curve, a physical constraint is added on whether the combustion heat generation process inside the main unit actually occurs. Start-up is only confirmed when the temperature change trend of the main unit conforms to the heat generation law of cold start and the vibration level exceeds that of the shutdown state, thus significantly reducing the probability of false positives.

[0055] The residual temperature rise rate is calculated as follows: the temperature rise rate calculated in the current simulation cycle is compared with the theoretical temperature rise rate at the same moment in the theoretical temperature rise curve of the cold start condition. The theoretical temperature rise rate is the expression of the nonlinear temperature rise model. Regarding time The first derivative of is in the form of: The derivative is solved in real time in the microcontroller or edge computing gateway.

[0056] The preset range for the residual temperature rise rate, i.e., the allowable upper and lower deviations, is determined as follows: In the actual ship cold start test, the deviation distribution between the temperature rise rate and the theoretical temperature rise rate during multiple normal start-up processes is recorded, and the 95% confidence interval of the deviation distribution is used as the preset range. A typical example is: upper deviation of +0.5℃ / min and lower deviation of -0.3℃ / min. The number of consecutive sampling periods required for continuous determination is 3 to 5 periods to further filter out short-term interference.

[0057] The method for determining the baseline value of shutdown noise is as follows: Under the condition that the main engine is in a confirmed shutdown state and only the background vibration of the ship's navigation exists, continuously collect synthetic vibration velocity data, and take the statistical mean plus three times the standard deviation as the baseline value of shutdown noise. A typical example is: if the average value of the synthetic vibration velocity collected in the shutdown state is 0.8 mm / s and the standard deviation is 0.2 mm / s, then the baseline value of shutdown noise is 0.8 + 3 × 0.2 = 1.4 mm / s. When the measured synthetic vibration velocity exceeds this value, it is determined that the vibration level is significantly higher than the background noise of shutdown.

[0058] In this embodiment, the specific method for calibrating the current operating condition is as follows: after determining that the ship's diesel engine has entered the start-up state, the synthetic vibration velocity and vibration dominant frequency are matched with the theoretical vibration response characteristics under each operating condition in the structural dynamics simulation model, and the operating condition with the highest similarity is taken as the current operating condition.

[0059] The similarity matching is achieved using a weighted Euclidean distance minimization criterion. Specifically, for each candidate operating condition pre-stored in the structural dynamics simulation model, the theoretical dominant vibration frequency and theoretical composite vibration velocity from its theoretical vibration response characteristics are taken, and these are combined with the currently measured dominant vibration frequency and composite vibration velocity to form two vectors. The weighted Euclidean distance between these two vectors is then calculated. The formula for calculating the weighted Euclidean distance is: in, For weighted Euclidean distance; and These are the currently measured dominant vibration frequency and the synthesized vibration velocity, respectively. and These are the theoretical dominant vibration frequency and the theoretical synthesized vibration velocity under the candidate operating conditions, respectively. and These are the weighting coefficients for the dominant vibration frequency and the synthesized vibration velocity, respectively.

[0060] Weighting coefficient and The weighting is based on the contribution of the dominant vibration frequency and the synthesized vibration velocity to the distinguishing effect of different operating conditions. The dominant vibration frequency is directly related to the rotational speed, has significant distinguishing effect across different operating conditions, and is less affected by disturbances, thus it is assigned a higher weight. The synthesized vibration velocity is affected by the hull transmission characteristics, has relatively low distinguishing effect, and is assigned a lower weight. A typical example of this value is as follows: Take 0.7, If we take 0.3, the sum of the two is 1.

[0061] After traversing all candidate operating conditions, select the one that makes the weighted Euclidean distance equal. The candidate operating condition with the minimum value is selected as the current operating condition obtained through matching, thereby synchronously determining the current speed and load. If the minimum weighted Euclidean distance still exceeds the preset matching tolerance threshold, it is determined that the current vibration characteristics cannot be reliably matched to any known operating condition. The system issues a calibration failure prompt and maintains the previous valid calibration result, waiting for the next simulation cycle to re-match. The matching tolerance threshold is determined based on the statistical distribution of historical matching distances under normal operating conditions, taking the 99th percentile of this distribution.

[0062] When the simulation start-up conditions are not met—that is, the residual temperature rise rate exceeds the preset range for multiple consecutive sampling cycles, the synthesized vibration velocity does not exceed the shutdown noise reference value, or neither of these conditions is met—the system maintains the current state of the main unit as shut down or not started, and does not trigger subsequent operating condition calibration and fuel consumption calculation processes. Simultaneously, the feature data collected in the current simulation cycle is marked as pending confirmation. The system continuously checks these two conditions in subsequent consecutive sampling cycles until the conditions are met and start-up is confirmed, or the accumulated waiting time exceeds the preset start-up timeout threshold.

[0063] The method for determining the start-up timeout threshold is as follows: Statistically analyze the typical time required from the issuance of the start-up command to the first simultaneous fulfillment of both the residual temperature rise rate and vibration velocity conditions during historical cold starts of this type of marine diesel engine. Take 1.5 to 2 times this time as the timeout threshold. A typical example is: if the average time required to meet both conditions for a normal cold start of this engine is 15 seconds, then the start-up timeout threshold is set to 30 seconds. When the cumulative waiting time exceeds this threshold and the two conditions are still not met, the system issues an alarm indicating a start-up judgment timeout and marks the data collected during this start-up process as an abnormal event for subsequent manual verification to confirm whether it is due to start-up failure, sensor malfunction, or model mismatch. Afterward, the system resets the cold start judgment state and restarts a new round of initial state judgment.

[0064] Step 4: Using the fuel sound velocity deviation value as input, perform parameter inversion in the fuel sound velocity simulation model to solve for the fuel impurity characteristic parameters; use the speed, load and fuel impurity characteristic parameters corresponding to the current operating conditions as boundary conditions, input them into the fuel and fuel consumption simulation model, and obtain the instantaneous theoretical fuel consumption rate through simulation calculation.

[0065] In this embodiment, the specific method for calibrating the current operating condition is as follows: after determining that the ship's diesel engine has entered the start-up state, the synthetic vibration velocity and vibration dominant frequency are matched with the theoretical vibration response characteristics under each operating condition in the structural dynamics simulation model, and the operating condition with the highest similarity is taken as the current operating condition.

[0066] The operating condition calibration has been completed in step 3. Step 4 directly calls the speed and load corresponding to the calibrated operating condition. If the operating condition calibration in step 3 fails, the most recent valid calibration result will be maintained and a prompt will be issued; if the calibration fails for multiple consecutive simulation cycles, fuel consumption prediction will be paused and an alarm will be triggered.

[0067] In this embodiment, the fuel sound velocity deviation value calculated in real time is input into the fuel sound velocity simulation model. The surface temperature of the engine corresponding to the fuel sound velocity acquisition time is used as the input condition to inversely solve the current equivalent air content or return fuel mixing ratio as the characteristic parameter of fuel impurities.

[0068] The specific method for inversion is as follows: using the measured surface temperature of the engine block and the fuel sound velocity deviation value calculated in step 1 as known quantities, a reverse lookup is performed in the two-dimensional interpolation table of the fuel sound velocity simulation model constructed in step 2. Since the interpolation table is a monotonically nonlinear mapping, the reverse lookup is implemented using a two-dimensional search or successive approximation method. One specific implementation method is as follows: first, fix the engine block surface temperature dimension and locate the row containing that temperature in the interpolation table; then, search for the two adjacent nodes in that row that are closest to the current fuel sound velocity deviation value, and inversely calculate the corresponding return fuel mixing ratio through linear interpolation. For example, under the condition that the surface temperature of the engine is 80℃, the measured deviation value of the fuel sound velocity is 0.047. Looking up the table, we find that in the 80℃ row, the deviation value is 0.035 when the return fuel mixing ratio is 20% and 0.052 when the return fuel mixing ratio is 30%. Then, by linear interpolation, we can calculate that the current return fuel mixing ratio is approximately 20% + (30% - 20%) × (0.047 - 0.035) / (0.052 - 0.035) ≈ 27.1%.

[0069] To further improve the robustness of the inversion solution, a moving average can be applied to the fuel sound velocity deviation values ​​for multiple consecutive simulation cycles, and then the average value can be used for the inversion solution; the window length of the moving average can be 3 to 5 simulation cycles.

[0070] Determine the speed and load corresponding to the current operating conditions, and input the speed, load, and fuel impurity characteristic parameters obtained from the inversion solution as boundary conditions into the fuel and fuel consumption simulation model. First, the reference combustion efficiency curve is corrected by the fuel impurity characteristic parameters, and then the corrected reference combustion efficiency curve is coupled with the reference heat release rate curve to simulate and obtain the instantaneous theoretical fuel consumption rate of the current operating conditions.

[0071] The simulation calculation process is as follows: First, using the current engine speed and load rate obtained from step 3, the baseline effective thermal efficiency when using standard pure oil is obtained by looking up a table in the baseline combustion efficiency curve, and recorded as the baseline thermal efficiency. Then, the combustion efficiency correction factor described in step 2 is applied. The baseline thermal efficiency is corrected, and the corrected effective thermal efficiency is equal to the baseline thermal efficiency multiplied by 1 / 2. The impurity influence coefficient is given here. Specific value example: Run the main engine on a test bench under rated operating conditions, and conduct tests using fuel with a return oil mixing ratio of 0% and a known non-zero return oil mixing ratio. Record the percentage decrease in actual effective thermal efficiency relative to the effective thermal efficiency of standard pure oil under each return oil mixing ratio. Plot the return oil mixing ratio on the x-axis and the relative decrease in effective thermal efficiency on the y-axis, and perform a linear regression through the origin. The slope obtained from the regression is the value of the return oil mixing ratio. .

[0072] The final formula for calculating the instantaneous theoretical fuel consumption rate is: Instantaneous theoretical fuel consumption rate = Effective power / (Corrected effective thermal efficiency × Lower heating value of fuel oil). Wherein, effective power is determined by the currently calibrated engine speed and load rate, as well as the rated power of the main engine, i.e., Effective power = Load rate × Rated power; the lower heating value of fuel oil is a physicochemical property constant of the fuel oil, measured in kilojoules per kilogram, typically around 40-200 kilojoules per kilogram for marine heavy fuel oil. For example: A certain type of main engine has a rated power of 10,000 kW, and the current rated load rate is 75%, so the effective power is 7,500 kW; the corrected effective thermal efficiency is 0.9512 × the reference thermal efficiency. Assuming the reference thermal efficiency under this condition is 0.42, the corrected effective thermal efficiency is 0.3995; the lower heating value of the fuel is taken as 40,200 kJ / kg; then the instantaneous theoretical fuel consumption rate = 7500 / (0.3995 × 40,200) ≈ 0.467 kg / s, which is equivalent to 1681 kg / hour. This output value is the estimated fuel consumption of the main engine after fuel quality correction under the current simulation cycle.

[0073] Please see Figure 3 The present invention also provides a system for predicting the fuel consumption of a large marine diesel engine in a high-impurity oil return environment. This system is used to implement the aforementioned method for predicting the fuel consumption of a large marine diesel engine in a high-impurity oil return environment, comprising: The multi-source signal real-time feature extraction module is used to synchronously acquire the operating parameters of the ship's diesel engine during the simulation cycle, extract its features in real time, and generate synthetic vibration velocity, vibration main frequency, temperature rise rate, and fuel sound velocity deviation value. A multiphysics digital twin model construction module is used to load a pre-built multiphysics digital twin model of a marine diesel engine. The multiphysics digital twin model includes: a thermodynamic simulation model to describe the theoretical temperature rise curve under various operating conditions; a structural dynamics simulation model to describe the theoretical vibration response characteristics under various operating conditions; a fuel sound velocity simulation model to describe the theoretical mapping relationship between fuel impurity characteristic parameters and fuel sound velocity deviation values; and a fuel and fuel consumption simulation model to respond to boundary conditions and output the theoretical fuel consumption rate. The operating condition start-up determination and calibration module is used to determine whether the initial state of the main engine is a cold start based on the initial surface temperature of the engine body. If so, the temperature rise rate is compared with the theoretical temperature rise curve from the cold start state. Based on the comparison result and the synthetic vibration velocity, it is determined whether the marine diesel engine meets the simulation start-up conditions. When the conditions are met, the synthetic vibration velocity, vibration frequency and the theoretical vibration response characteristics of the structural dynamics simulation model under each operating condition are pattern matched to calibrate the current operating condition. The inversion and theoretical fuel consumption rate simulation module is used to perform parameter inversion in the fuel sound velocity simulation model with the fuel sound velocity deviation value as input, and solve for the characteristic parameters of fuel impurities. The speed, load and fuel impurity characteristic parameters corresponding to the current operating conditions are used as boundary conditions and input into the fuel and fuel consumption simulation model to obtain the instantaneous theoretical fuel consumption rate through simulation calculation.

[0074] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0075] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0076] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0077] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for predicting the fuel consumption of a large marine diesel engine in a high-impurity return oil environment, characterized in that, The specific steps include: Step 1: During the simulation cycle, the operating parameters of the ship's diesel engine are acquired synchronously, and real-time feature extraction is performed to generate the synthetic vibration velocity, vibration main frequency, temperature rise rate, and fuel sound velocity deviation value. Step 2: Load the pre-built multiphysics digital twin model of the marine diesel engine. The multiphysics digital twin model includes: a thermodynamic simulation model to describe the theoretical temperature rise curve under various operating conditions; a structural dynamics simulation model to describe the theoretical vibration response characteristics under various operating conditions; a fuel sound velocity simulation model to describe the theoretical mapping relationship between fuel impurity characteristic parameters and fuel sound velocity deviation values; and a fuel and fuel consumption simulation model to respond to boundary conditions and output the theoretical fuel consumption rate. Step 3: Determine whether the initial state of the main engine is a cold start based on the initial surface temperature of the engine body. If so, compare the temperature rise rate with the theoretical temperature rise curve from the cold start state using residual analysis. Based on the comparison results and the synthesized vibration velocity, determine whether the marine diesel engine meets the simulation start-up conditions. When the conditions are met, perform pattern matching between the synthesized vibration velocity, the dominant vibration frequency, and the theoretical vibration response characteristics of the structural dynamics simulation model under various operating conditions to calibrate the current operating conditions. Step 4: Using the fuel sound velocity deviation value as input, perform parameter inversion in the fuel sound velocity simulation model to solve for the fuel impurity characteristic parameters; use the speed, load and fuel impurity characteristic parameters corresponding to the current operating conditions as boundary conditions, input them into the fuel and fuel consumption simulation model, and obtain the instantaneous theoretical fuel consumption rate through simulation calculation. The fuel sound velocity simulation model is constructed as follows: Under laboratory conditions, a family of sound velocity curves showing the change of fuel sound velocity with temperature under different fuel impurity content ratios are measured, and a mapping model is established with engine surface temperature and fuel impurity characteristic parameters as inputs and fuel sound velocity deviation value as output; the fuel impurity characteristic parameters include equivalent air content or return fuel mixing ratio, which are used to characterize the degree of fuel quality degradation in a high impurity return fuel environment. The fuel and fuel consumption simulation model is constructed as follows: based on the modified Weber combustion model, the fuel and fuel consumption simulation model is established with speed, load and fuel impurity characteristic parameters as input boundary conditions; the fuel and fuel consumption simulation model pre-stores the reference combustion efficiency curve and reference heat release rate curve obtained through bench test calibration, and corrects the reference combustion efficiency curve and reference heat release rate curve according to the input fuel impurity characteristic parameters, and simulates and outputs the instantaneous theoretical fuel consumption rate under the operating conditions; The method for determining cold start is as follows: When the initial state of the main engine is determined to be a cold start based on the initial surface temperature of the engine body, the temperature rise rate is compared point by point with the theoretical temperature rise curve under the cold start condition in the thermodynamic simulation model on the same time series to calculate the temperature rise rate residual; at the same time, the synthesized vibration velocity is compared with the preset shutdown noise benchmark value; only when the temperature rise rate residual is within the preset range for multiple consecutive sampling periods and the synthesized vibration velocity exceeds the shutdown noise benchmark value is the marine diesel main engine determined to have entered the start-up state.

2. The method for predicting the fuel consumption of a large marine diesel engine in a high-impurity return oil environment according to claim 1, characterized in that, The operating parameters are achieved by wireless composite sensors deployed on the marine diesel engine. The wireless composite sensors are magnetically attached to the surface of the marine diesel engine and clamped to the outer wall of the high-pressure oil pipe. They synchronously collect the triaxial vibration velocity of the engine body, the surface temperature of the engine body, and the instantaneous sound velocity of the fuel at the outer wall of the high-pressure oil pipe in a non-invasive manner. The synthesized vibration velocity is calculated from the vector sum of the three-axis vibration velocities of the engine body; the dominant vibration frequency is extracted by spectral analysis of the vibration velocity signal; the temperature rise rate is calculated from the temperature difference of the engine body surface in adjacent sampling periods; and the fuel sound velocity deviation value is calculated from the normalized difference between the instantaneous fuel sound velocity and the theoretical pure fuel sound velocity under the same temperature conditions. The theoretical pure fuel sound velocity refers to the sound velocity value of standard fuel without impurities under the same temperature conditions. The fuel sound velocity deviation value is calculated from the normalized difference between the instantaneous fuel sound velocity and the theoretical pure fuel sound velocity.

3. The method for predicting the fuel consumption of a large marine diesel engine in a high-impurity return oil environment according to claim 1, characterized in that, The thermodynamic simulation model is constructed as follows: the surface temperature of the marine diesel engine is used under various operating conditions, including cold start, low idling load, economic cruise, heavy load variable conditions, and shutdown cooling. A nonlinear temperature rise model of the surface temperature of the engine as a function of operating time is established based on the Arrhenius formula. The corresponding temperature rise curve parameters are fitted for different operating conditions to form a theoretical temperature rise curve.

4. The method for predicting the fuel consumption of a large marine diesel engine in a high-impurity return oil environment according to claim 1, characterized in that, The structural dynamics simulation model is constructed as follows: historical data of the synthetic vibration velocity and dominant vibration frequency of the ship's diesel engine under various operating conditions are collected. Based on a pre-established simplified multi-degree-of-freedom parameter model, the correspondence between rotational speed and dominant vibration frequency, as well as the correspondence between rotational speed, load and synthetic vibration velocity, are established to form theoretical vibration response characteristics under different operating conditions. Among them, the dominant vibration frequency is used to back-estimate the real-time rotational speed, and the synthetic vibration velocity is used to match and determine the real-time load.

5. The method for predicting the fuel consumption of a large marine diesel engine in a high-impurity return oil environment according to claim 1, characterized in that, The specific method for calibrating the current operating condition is as follows: when it is determined that the ship's diesel engine has entered the start-up state, the synthetic vibration velocity and vibration main frequency are matched with the theoretical vibration response characteristics under each operating condition in the structural dynamics simulation model, and the operating condition with the highest similarity is taken as the current operating condition.

6. The method for predicting the fuel consumption of a large marine diesel engine in a high-impurity return oil environment according to claim 1, characterized in that, The fuel sound velocity deviation value calculated in real time is input into the fuel sound velocity simulation model. The engine surface temperature corresponding to the fuel sound velocity acquisition time is used as the input condition to inversely solve the current equivalent air content or return fuel mixing ratio as the characteristic parameter of fuel impurities. Determine the speed and load corresponding to the current operating conditions, and input the speed, load, and fuel impurity characteristic parameters obtained from the inversion solution as boundary conditions into the fuel and fuel consumption simulation model. First, the reference combustion efficiency curve is corrected by the fuel impurity characteristic parameters, and then the corrected reference combustion efficiency curve is coupled with the reference heat release rate curve to simulate and obtain the instantaneous theoretical fuel consumption rate of the current operating conditions.

7. A system for predicting the fuel consumption of a large marine diesel engine in a high-impurity return oil environment, characterized in that, The system for predicting the fuel consumption of a large marine diesel engine in a high-impurity oil return environment is used to implement the method for predicting the fuel consumption of a large marine diesel engine in a high-impurity oil return environment as described in any one of claims 1-6, comprising: The multi-source signal real-time feature extraction module is used to synchronously acquire the operating parameters of the ship's diesel engine during the simulation cycle, extract its features in real time, and generate synthetic vibration velocity, vibration main frequency, temperature rise rate, and fuel sound velocity deviation value. A multiphysics digital twin model construction module is used to load a pre-built multiphysics digital twin model of a marine diesel engine. The multiphysics digital twin model includes: a thermodynamic simulation model to describe the theoretical temperature rise curve under various operating conditions; a structural dynamics simulation model to describe the theoretical vibration response characteristics under various operating conditions; a fuel sound velocity simulation model to describe the theoretical mapping relationship between fuel impurity characteristic parameters and fuel sound velocity deviation values; and a fuel and fuel consumption simulation model to respond to boundary conditions and output the theoretical fuel consumption rate. The operating condition start-up determination and calibration module is used to determine whether the initial state of the main engine is a cold start based on the initial surface temperature of the engine body. If so, the temperature rise rate is compared with the theoretical temperature rise curve from the cold start state. Based on the comparison result and the synthetic vibration velocity, it is determined whether the marine diesel engine meets the simulation start-up conditions. When the conditions are met, the synthetic vibration velocity, vibration frequency and the theoretical vibration response characteristics of the structural dynamics simulation model under each operating condition are pattern matched to calibrate the current operating condition. The inversion and theoretical fuel consumption rate simulation module is used to perform parameter inversion in the fuel sound velocity simulation model with the fuel sound velocity deviation value as input, and solve for the characteristic parameters of fuel impurities. The speed, load and fuel impurity characteristic parameters corresponding to the current operating conditions are used as boundary conditions and input into the fuel and fuel consumption simulation model to obtain the instantaneous theoretical fuel consumption rate through simulation calculation.

Images