Method and device for monitoring the state of a marine diesel engine

Abstract

The present application provides a marine diesel engine state monitoring method and device, relates to the diesel engine state monitoring technical field, and specifically comprises the following steps: collecting three-axis vibration velocity components, three-axis direct current gravity components and scavenging pressure, obtaining three-axis combined vibration velocity and ship roll angle, calculating the cumulative starting working hours of the diesel engine in combination with the basic start-stop threshold value, decoupling the normalized response ratio of vibration and pressure combined with the reference parameter and the ship roll angle, weighting the degradation divergence index through the life cycle weighting coefficient, dividing the sub-interval through the time domain sliding window and counting the number, taking the arithmetic mean of the dominant response interval as the degradation index, and comparing the preset degradation index threshold value to determine the health state. The present application filters out the false vibration amplification caused by the sea state swing and load through multi-dimensional decoupling to overcome false positives, relies on frequency statistics to isolate data pollution, combines working hours to give adaptive fatigue sensitivity, realizes the nonlinear amplification of weak degradation, and improves the diagnosis accuracy.

Description

Technical Field

[0001] This invention relates to the field of diesel engine condition monitoring technology, specifically to a method and apparatus for monitoring the condition of marine diesel engines. Background Technology

[0002] As the core of a ship's power, the stability of its operating state directly affects the overall safety of the navigation system. With the development of the Industrial Internet of Things and edge computing technologies, marine engineering is gradually evolving from traditional periodic preventive maintenance to intelligent predictive maintenance. Currently, using sensors to collect low-level physical parameters of diesel engines, such as vibration and thermodynamics, in real time, and combining this with software algorithm models for condition monitoring and diagnosis, has become the mainstream technical approach in this field. However, the operating environment of a ship's engine room is extremely complex, with multiple sources of physical fields, including dynamics, thermodynamics, and the marine environment, coupled and intertwined. The deep integration of existing data processing algorithms with the special physical conditions of ships is still in its early stages, making it difficult to meet the high-precision monitoring requirements under extreme sea conditions.

[0003] In existing technologies, vibration speed signals are typically collected by installing vibration sensors on the machine surface. The core technical solution generally involves: collecting vibration signals in a fixed direction using sensors, extracting the absolute effective value of the vibration, and directly comparing it with a preset global fixed alarm threshold to determine the equipment's health status. In some improved existing technologies, the algorithm simultaneously collects the engine's input load parameters as an independent auxiliary reference dimension, or combines this with the equipment's cumulative operating time to trigger fixed maintenance work orders.

[0004] However, the aforementioned existing technologies have technical defects in actual monitoring. These defects primarily manifest in the fact that the algorithm fails to decouple the dynamic response of the diesel engine from the ship's heel attitude and the dynamics of the work load execution. This results in monitoring results being strictly dependent on the precise installation and alignment of sensors. Furthermore, under full-load acceleration or severe sea conditions, the attenuation of the normal constraint force of the engine mount easily leads to amplified false vibrations and continuous false alarms. Simultaneously, the existing technologies lack a nonlinear gain amplification mechanism that matches the evolution of mechanical fatigue life, making them lack adaptive sensitivity to early, subtle structural degradation characteristics. In addition, the global averaging or static fixed interval segmentation used in the underlying data filtering is problematic. The former is easily amplified by transient extreme pulses such as large waves, triggering false alarms, while the latter, due to its inherent boundary truncation effect, not only fails to adapt to the natural changes in background divergence caused by long-term equipment aging but also fragments the truly dense steady-state dominant characteristics, ultimately leading to a significant decrease in early warning accuracy throughout the entire lifespan.

[0005] 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

[0006] The purpose of this invention is to provide a method and apparatus for monitoring the condition of marine diesel engines, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: The condition monitoring method for marine diesel engines includes the following specific steps: Step 1: Real-time acquisition of triaxial vibration velocity components, triaxial DC gravity components, and scavenging pressure at different times during diesel engine operation; calculate the triaxial composite vibration velocity based on the triaxial vibration velocity components. Step 2: Compare the triaxial composite vibration velocity with the preset basic start-stop threshold, obtain the diesel engine start-stop status based on the comparison results, and count the cumulative start-up time of the diesel engine; Step 3: Calculate the ship's heel angle based on the three-axis DC gravity components, normalize the three-axis composite vibration velocity and scavenging pressure with the corresponding preset reference parameters, and decouple the calculations based on the ship's heel angle to obtain the normalized response ratio. Step 4: Based on the cumulative start-up time of the diesel engine, combined with the preset material fatigue experience constant and the diesel engine design life time, the life cycle weighting coefficient is obtained, and the normalized response ratio is used for weighted calculation. The degradation divergence index is generated in real time at each moment, forming a continuously changing degradation divergence index time series. Step 5: Set a time-domain sliding window with a fixed number of sampling points, continuously input the real-time output degradation divergence index time series data, construct a scanning cursor based on the interquartile range and the number of sampling points in the time-domain sliding window, and use the scanning cursor to slide and traverse from the minimum to the maximum value in the time-domain sliding window data to define the dominant response interval of the degradation divergence index, and output the arithmetic mean of the data in this interval as the degradation index. Step 6: Compare the degradation index with the preset degradation index threshold, and judge the health status of the diesel engine based on the comparison results.

[0008] Furthermore, with the geometric center point of the diesel engine as the origin, the X-axis and Y-axis are set in the horizontal direction, and the Z-axis is set in the vertical direction to establish a spatial geometric coordinate system; The triaxial vibration velocity components include the effective values ​​of the vibration velocities of the diesel engine's X-axis, Y-axis, and Z-axis. The three-dimensional spatial vector magnitude calculation is used to calculate the triaxial composite vibration velocity from the triaxial vibration velocity components.

[0009] Furthermore, the triaxial composite vibration velocity is compared with the preset basic start-stop threshold. When the triaxial composite vibration velocity is greater than or equal to the basic start-stop threshold, the diesel engine is determined to be in operation. The running time of this start-up is continuously accumulated according to the sampling interval, and the accumulated start-up time is the value after the last shutdown settlement. When the triaxial composite vibration velocity is less than the basic start-stop threshold, the diesel engine is determined to be in a stopped state. The accumulated running time of this start is added to the total machine start time, and the running time of this start is reset to zero, waiting for the next start to reset the timing.

[0010] Furthermore, the ship's heel angle is calculated based on the three-axis DC gravity components, with the following specific logic: The three-axis DC gravity components include the DC gravity components of the X-axis, Y-axis, and Z-axis, and the preset three-axis reference DC gravity components include the reference DC gravity components corresponding to the X-axis, Y-axis, and Z-axis. The collected X, Y, and Z axis DC gravity components are combined to form a measured triaxial DC gravity vector, and the collected three reference DC gravity components are combined to form a reference triaxial DC gravity vector. Calculate the dot product of the two sets of spatial vectors, then solve for the magnitudes of the two vectors separately. Divide the dot product of the vectors by the product of the magnitudes of the two vectors to obtain the cosine of the included angle. Convert the angle of heel of the ship by the inverse cosine operation.

[0011] Furthermore, the normalized response ratio is obtained, with the following specific logic: the preset reference parameters include the reference triaxial composite vibration velocity and the reference scavenging pressure; The ratio of the triaxial composite vibration velocity to the preset reference triaxial composite vibration velocity is calculated to obtain the normalized triaxial composite vibration velocity. Calculate the ratio of the scavenging pressure to the preset reference scavenging pressure to obtain the normalized scavenging pressure; Calculate the square of the ratio of the normalized triaxial composite vibration velocity to the normalized scavenging pressure, and multiply it by the square of the cosine of the ship's heel angle to obtain the normalized response ratio.

[0012] Furthermore, the life cycle weighting coefficient is obtained based on the following formula: in, Represents the life-cycle weighting coefficient. This represents a preset empirical constant for material fatigue. This indicates the cumulative starting hours of the diesel engine. This indicates the preset design life hours for the diesel engine; The degradation divergence index is obtained by weighting the life cycle weighting coefficient and the normalized response ratio, based on the following formula: in, Indicates the degradation divergence index. This represents the normalized response ratio.

[0013] Furthermore, a scanning cursor is constructed based on the interquartile range and the number of sampling points within the time-domain sliding window. The scanning cursor slides across the data from the minimum to the maximum value within the time-domain sliding window to define the dominant response interval of the degradation divergence index. The specific logic is as follows: Input degradation divergence index time series data into a time-domain sliding window with a fixed number of sampling points, extract the interquartile range of the degradation divergence index within the time-domain sliding window, and calculate the vernier width based on the fixed number of sampling points in the time-domain sliding window. The formula used is as follows: in, Indicates the cursor width. The interquartile range represents the degradation divergence index within the time-domain sliding window. This indicates the preset fixed number of sampling points; Generate a range-scanning cursor based on the cursor width; Using the minimum value of the degradation divergence index within the time-domain sliding window as the starting boundary, the value range scanning cursor is controlled to continuously overlap and slide along the direction of numerical increase with the cursor width as the step size, thus dividing the area into multiple overlapping sub-intervals. Count the number of degradation divergence indices contained in each sub-interval; Compare the number of degradation divergence indices in each sub-interval, and select the sub-interval with the largest number of degradation divergence indices as the dominant response interval. If there are two or more sub-intervals with the same maximum value for the number of degradation divergence indices, then the sub-interval with the largest upper limit of the value range is selected as the dominant response interval. The upper limit of the value range is the critical value at the right end of the value range of a single sub-interval.

[0014] Furthermore, the degradation index is compared with a preset degradation index threshold: If the degradation index is greater than or equal to the degradation index threshold, the diesel engine is determined to be in a faulty state. If the degradation index is less than the degradation index threshold, the diesel engine is considered to be in a healthy state.

[0015] To achieve the above objectives, the present invention also provides the following technical solution: A condition monitoring device for a marine diesel engine, the device being used to perform the condition monitoring method for a marine diesel engine as described in any of the preceding claims, comprising: Data acquisition module: Real-time acquisition of triaxial vibration velocity components, triaxial DC gravity components and scavenging pressure at different times during diesel engine operation, and calculation of triaxial composite vibration velocity based on triaxial vibration velocity components; Start-stop status judgment module: compares the triaxial composite vibration velocity with the preset basic start-stop threshold, obtains the diesel engine start-stop status based on the comparison result, and counts the cumulative start-up time of the diesel engine; Data decoupling module: Calculates the ship's heel angle based on the three-axis DC gravity components, normalizes the three-axis composite vibration velocity and scavenging pressure with the corresponding preset reference parameters, and decouples the calculation based on the ship's heel angle to obtain the normalized response ratio; Data processing module: Based on the cumulative start-up time of the diesel engine, combined with the preset material fatigue experience constant and the diesel engine design life time, the life cycle weighting coefficient is obtained, and the normalized response ratio is used for weighted calculation. The degradation divergence index is generated in real time at each moment, forming a continuously changing degradation divergence index time series. Interval Extraction Module: A time-domain sliding window with a fixed number of sampling points is preset, continuously inputting the real-time output degradation divergence index time-series data, constructing a scanning cursor based on the interquartile range and the number of sampling points in the time-domain sliding window, and traversing the data in the time-domain sliding window from the minimum value to the maximum value to define the dominant response interval of the degradation divergence index, and outputting the arithmetic mean of the data in this interval as the degradation index. Health status assessment module: compares the degradation index with the preset degradation index threshold, and judges the health status of the diesel engine based on the comparison result.

[0016] Compared with the prior art, the beneficial effects of the present invention are: This invention calculates the triaxial composite vibration velocity and ship heel angle based on the triaxial vibration velocity components, triaxial DC gravity components, and scavenging air pressure collected during diesel engine operation. The composite vibration velocity and scavenging air pressure are then normalized using corresponding preset reference parameters, and the calculation is completely decoupled from the ship heel angle to obtain the normalized response ratio. This logic successfully and accurately separates load fluctuations and spatial attitude changes from a single vibration signal, eliminating the false amplification of vibration intensity caused by external wind and waves and internal air supply load conversion. It represents a leap from traditional single fixed absolute value judgment to multi-dimensional relative characteristic monitoring, significantly reducing the false alarm rate of condition monitoring under complex operating conditions. This invention also compares the triaxial synthetic vibration velocity with a preset basic start-stop threshold to obtain the diesel engine's start-stop status and accurately calculate the cumulative start-up time. Then, it combines this with preset material fatigue empirical constants and the diesel engine's design lifespan to construct a life-cycle weighted coefficient. This invention utilizes the life-cycle weighted coefficient combined with a normalized response ratio to generate a degradation divergence index, perfectly aligning with the objective laws of mechanical fatigue evolution and giving the monitoring features the physical property of automatically increasing sensitivity with increasing operating time. Subsequently, the real-time output degradation divergence index time-series data is continuously input. A scanning cursor is constructed based on the interquartile range and the number of sampling points within the time-domain sliding window. The scanning cursor slides and overlaps within the time-domain sliding window data from minimum to maximum value, defining the dominant response interval of the degradation divergence index. The arithmetic mean of the data within this interval is output as the degradation index, and compared with the degradation index threshold to determine the health status. This vernier construction and overlapping scanning mechanism based on interquartile range replaces the practice of blindly averaging and statically fixed equal-width cutting. It not only immunizes against the distortion caused by extreme physical extremes on the statistical span, but also ensures that the densest physical real response is not truncated and crushed by the boundary, accurately extracting the core data representing the stable operation of the equipment, and greatly improving the fault diagnosis accuracy throughout the entire life cycle. 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 of the overall device 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 Figure 1 The present invention provides a technical solution: The condition monitoring method for marine diesel engines includes the following specific steps: Step 1: Real-time acquisition of triaxial vibration velocity components, triaxial DC gravity components, and scavenging pressure at different times during diesel engine operation; calculate the triaxial composite vibration velocity based on the triaxial vibration velocity components. Among them, the triaxial vibration velocity component is obtained by collecting the triaxial vibration acceleration signal through a triaxial accelerometer and performing integration processing; the triaxial DC gravity component is obtained by collecting the triaxial raw acceleration signal through a triaxial accelerometer and performing low-pass filtering processing; and the scavenging pressure is directly collected by a pressure sensor installed at the diesel engine scavenging box or intake manifold. In this embodiment, a spatial geometric coordinate system is established with the geometric center point of the diesel engine as the origin, the X-axis and Y-axis are set in the horizontal direction, and the Z-axis is set in the vertical direction. The triaxial vibration velocity components include the effective values ​​of the vibration velocities of the diesel engine's X-axis, Y-axis, and Z-axis. The three-dimensional spatial vector magnitude calculation is used to calculate the triaxial composite vibration velocity from the triaxial vibration velocity components; The formula used to calculate the triaxial composite vibration velocity is as follows: in, Indicates the triaxial composite vibration velocity. This represents the effective value of the X-axis vibration velocity of the diesel engine. This represents the effective value of the Y-axis vibration velocity of the diesel engine. This represents the effective value of the Z-axis vibration velocity of the diesel engine.

[0021] Based on the above, it should be noted that, since ships are inevitably affected by the heeling of the hull caused by severe winds and waves when actually sailing on the sea, and the combustion load inside the diesel engine is also in dynamic fluctuation with the operation of changing conditions, the measured mechanical vibration response is easily modulated by the cross-modulation of external spatial attitude changes and internal thermodynamic state transformation. This requires that the three-axis vibration velocity component, the three-axis DC gravity component, and the scavenging pressure be collected at the same time to completely lock the bottom state of the multidimensional physical field on the same time section.

[0022] In this set of synchronously acquired real-time parameters, the three-axis vibration velocity component fully carries the mechanical vibration energy of the diesel engine in space, the three-axis DC gravity component accurately maps the distribution ratio of gravitational acceleration caused by the hull attitude drift, and the scavenging pressure, as a sensitive indicator directly related to thermal load, can characterize the intensity of the actual excitation force inside the cylinder due to combustion work hitting the engine base. The simultaneous coordination of these three heterogeneous parameters directly determines whether external attitude interference and internal load illusion can be separated from the mixed signal.

[0023] After successfully acquiring the three-axis vibration velocity components, they are transformed into a unified three-axis composite vibration velocity through three-dimensional spatial vector magnitude calculation. The scientific nature of this is that the energy transfer of mechanical vibration exhibits continuity and scattering in three-dimensional space. Vibration indicators of any single or partial axis are easily subject to the strict limitations of the physical alignment accuracy of the sensor. Furthermore, with the slight deformation of the aircraft structure during long-term navigation, the energy distribution between the axes in space often undergoes relative migration, which in turn leads to inherent defects such as omission or distortion of single-axis data.

[0024] Therefore, the triaxial composite vibration velocity obtained by performing vector superposition on three mutually perpendicular spatial components achieves lossless normalization extraction of the full energy amplitude at the mathematical level. This makes the mechanical response indicators used in subsequent diagnosis spatially isotropic, which not only eliminates the strict dependence on the horizontality of sensor installation in traditional monitoring processes, but also prevents false alarms and misjudgments caused by misalignment of measuring points or uniaxial energy drift. This greatly ensures the authenticity and objectivity of the underlying physical characteristic data, laying an extremely solid digital foundation for achieving highly sensitive degradation state assessment throughout the entire life cycle.

[0025] Step 2: Compare the triaxial composite vibration velocity with the preset basic start-stop threshold, obtain the diesel engine start-stop status based on the comparison results, and count the cumulative start-up time of the diesel engine; In this embodiment, the triaxial composite vibration velocity is compared with the preset basic start-stop threshold. When the triaxial composite vibration velocity is greater than or equal to the basic start-stop threshold, the diesel engine is determined to be in operation. The running time of this start-up is continuously accumulated according to the sampling interval. The accumulated start-up time is the value after the last shutdown settlement. When the triaxial composite vibration velocity is less than the basic start-stop threshold, the diesel engine is determined to be in a stopped state. The accumulated running time of this start is added to the total machine start time, and the running time of this start is cleared to zero, waiting for the next start to reset the timing. The diesel engine is divided into a main unit and an auxiliary unit. Historical vibration data of the main unit and auxiliary unit under different conditions are collected, and the triaxial composite vibration velocity per minute is calculated. The triaxial composite vibration velocity in the stopped state is observed to be approximately... Fluctuations between these ranges; the triaxial composite vibration velocity during main unit operation is typically within... The above-mentioned triaxial composite vibration velocity during auxiliary machine operation is typically within the range of... The above. After observing the synthesis speed trend in conjunction with the actual on-site operation, the basic start-up and shutdown thresholds for the main unit are manually determined. Auxiliary machine basic start / stop threshold Verification has shown that this basic start / stop threshold can clearly distinguish between shutdown and operation states.

[0026] Based on the above, it should be noted that during the long-term operation of marine diesel engines, determining the start-stop status and calculating the cumulative start-up time directly determines the accuracy of the subsequent physical fatigue evolution assessment parameters. The microscopic damage and macroscopic stiffness degradation of mechanical materials physically correspond strictly to the cumulative operating period of the equipment under real alternating loads. However, during daily navigation, berthing, or anchoring, diesel engines experience frequent start-stop switching and long periods of offline idleness. If the operating and shutdown conditions are not strictly defined and the natural time span is directly accumulated, the vibration response caused by hull swaying, transmission from side-operating machinery, and environmental clutter in the engine room during the idle period will be incorrectly included in the diesel engine's own operating losses, leading to a numerical drift in the time base for structural life assessment.

[0027] To address this technical bias, the real-time acquired triaxial composite vibration velocity is numerically compared with the preset basic start-stop threshold. The rationale is that the mechanical vibration energy amplitude generated by the combustion of the cylinder and the reciprocating motion of the piston in the running state of the diesel engine is significantly higher than the ambient noise in the stopped state. Through numerical comparison, the physical operation definition between the running excitation and the stationary state can be directly realized at the data extraction level.

[0028] Based on this judgment mechanism, when the triaxial composite vibration velocity is greater than or equal to the preset basic start-stop threshold, the diesel engine is determined to be in operation, and the running time of this start-up is continuously accumulated according to the sampling interval. When the triaxial composite vibration velocity is less than the preset basic start-stop threshold, the diesel engine is determined to be in shutdown, the accumulated running time of this start-up is added to the total machine startup time, and the startup time is reset to zero.

[0029] This time-segmented calculation logic eliminates the data accumulation during non-operation periods of the diesel engine, ensuring that the cumulative start-up time parameters of the diesel engine only reflect the actual mechanical operation and wear time. Its beneficial effect is that it effectively isolates the equipment operation time quantification process from the external non-operation state, providing an objective time axis basis for the subsequent construction of life cycle weighted coefficients, excluding the interference of downtime state changes. This enables the algorithm model to strictly correspond to the actual operation and wear trajectory of the mechanical structure and perform feature adjustments, avoiding the deviation in the calculation of degradation features caused by inflated working hours, and improving the accuracy of fault diagnosis and judgment under multiple operating conditions.

[0030] Step 3: Calculate the ship's heel angle based on the three-axis DC gravity components, normalize the three-axis composite vibration velocity and scavenging pressure with the corresponding preset reference parameters, and decouple the calculations based on the ship's heel angle to obtain the normalized response ratio. In this embodiment, the ship's roll angle is calculated based on the three-axis DC gravity components. The specific logic is as follows: The three-axis DC gravity components include the DC gravity components of the X-axis, Y-axis, and Z-axis, and the preset three-axis reference DC gravity components include the reference DC gravity components corresponding to the X-axis, Y-axis, and Z-axis. The collected X, Y, and Z axis DC gravity components are combined to form a measured triaxial DC gravity vector, and the collected three reference DC gravity components are combined to form a reference triaxial DC gravity vector. Calculate the dot product of the two sets of spatial vectors, then solve for the magnitudes of the two vectors separately. Divide the dot product by the product of the magnitudes of the two vectors to obtain the cosine of the included angle. Convert this cosine value to the ship's heel angle using the inverse cosine operation. The formula used is: in, Indicates the ship's heel angle. This represents the DC component of gravity along the X-axis. Represents the DC component of gravity along the Y-axis. Represents the DC gravity component along the Z-axis. Represents the DC gravity component of the X-axis reference. Represents the DC gravity component of the Y-axis reference. Represents the DC gravity component of the Z-axis reference; The normalized response ratio is obtained, and the specific logic is as follows: The preset reference parameters include the reference triaxial composite vibration velocity and the reference scavenging pressure; The ratio of the triaxial composite vibration velocity to the preset reference triaxial composite vibration velocity is calculated to obtain the normalized triaxial composite vibration velocity. Calculate the ratio of the scavenging pressure to the preset reference scavenging pressure to obtain the normalized scavenging pressure; The normalized response ratio is obtained by calculating the square of the ratio of the normalized triaxial composite vibration velocity to the normalized scavenging pressure, and then multiplying this product by the square of the cosine of the ship's heel angle. The formula used is as follows: in, The normalized response ratio represents the relative vibration response of a diesel engine per unit work excitation after eliminating disturbances from sea state inclination and load fluctuations. Represents the normalized triaxial composite vibration velocity. Indicates the normalized scavenging pressure. This represents the preset baseline triaxial composite vibration velocity, obtained by collecting multiple triaxial composite vibration velocities of a diesel engine under healthy conditions and calculating their arithmetic mean. Indicates scavenging pressure, The reference scavenging pressure is obtained by collecting multiple scavenging pressures of the diesel engine under healthy conditions and calculating their arithmetic average.

[0031] Building upon the above, it should be noted that, in order to achieve high-precision state perception in the complex marine environment characterized by varying sea states and intricate operating conditions, the ship's heel angle is precisely calculated based on the reference triaxial DC gravity vector formed by the combination of reference triaxial DC gravity components and the measured triaxial DC gravity vector formed by the combination of measured components. This is necessary to explicitly reconstruct the spatial tilt attitude of the hull under wave impact, thereby providing a quantitative attitude reference to eliminate vibration measurement errors caused by engine base constraint force attenuation. Furthermore, the normalized response ratio is obtained by further decoupling and calculating the normalized triaxial synthetic vibration velocity and normalized scavenging pressure. Its core significance lies in establishing a dimensionless mechanical state characterization index that eliminates interference from external wind and waves and internal air supply loads, revealing the inherent response characteristics of the diesel engine under unit work excitation from a mechanistic perspective.

[0032] In the formula calculation of this design, the ship's heel angle, as the calculation result, specifically reflects the geometric deflection of the sensor coordinate axes relative to the absolute horizontal plane at the current sampling moment, achieving the technical effect of accurately quantifying the intensity of attitude deviation. Meanwhile, the normalized response ratio profoundly reflects the degree of steady-state decay or degradation of the internal mechanical structure of the diesel engine. Its technical effect lies in converting the traditional absolute vibration amplitude into a relative variability index that better reflects the machine's internal health. Furthermore, the changes in each component of the measured three-axis DC gravity vector are directly related to the spatial redistribution of gravitational acceleration caused by the ship's tilt. The proportional relationship between the three-dimensional vector dot product and the modulus product directly maps the geometric span of the attitude deviation. The normalized scavenging pressure, as a core physical quantity directly reflecting the cylinder's power excitation force, is coupled with the normalized three-axis synthesized vibration velocity, a variable representing the mechanical terminal spatial response, jointly determining the dynamic mapping relationship of the mechanical structure's transfer function.

[0033] From the positive and negative correlation in the formula, since the ship's heel angle is obtained by dividing the dot product of the measured triaxial DC gravity vector and the reference triaxial DC gravity vector by the product of the magnitudes of the two vectors and then converting it by inverse cosine, the ship's heel angle and the cosine value of the ratio of the dot product of the vectors show an inverse correspondence. That is, when the ship tilts and the ratio of the dot product decreases, the calculated ship's heel angle increases accordingly.

[0034] In the formula for calculating the normalized response ratio, the normalized response ratio and the normalized triaxial composite vibration velocity exhibit a positive square correlation. This means that when the work load remains constant, the increase in composite vibration energy caused by mechanical structure wear or increased clearance will directly lead to a significant quadratic increase in the normalized response ratio. Conversely, the normalized response ratio and the normalized scavenging pressure exhibit a negative square correlation. By using the normalized scavenging pressure as the denominator, it is possible to ensure that when normal variable operating conditions such as full-load acceleration cause a sharp increase in scavenging pressure and a natural amplification of excitation force, the mathematical suppression effect of increasing the denominator can be used to offset the synchronous increase in vibration amplitude, thereby avoiding continuous false alarms under variable operating conditions.

[0035] At the same time, the normalized response ratio is positively correlated with the square of the cosine of the ship's heel angle, while it is negatively correlated with the ship's heel angle itself. This design cleverly utilizes the physical characteristic that the cosine value decreases when the hull is tilted, and performs reverse mathematical suppression and gain compensation on the spurious vibration amplification caused by the weakening of the normal constraint force of the engine base under severe sea conditions, ensuring that the final output normalized response ratio can be purely focused on the early degradation of the diesel engine mechanical structure itself.

[0036] Step 4: Based on the cumulative start-up time of the diesel engine, combined with the preset material fatigue experience constant and the diesel engine design life time, the life cycle weighting coefficient is obtained, and the normalized response ratio is used for weighted calculation. The degradation divergence index is generated in real time at each moment, forming a continuously changing degradation divergence index time series. In this embodiment, the life cycle weighting coefficient is obtained based on the following formula: in, Represents the life-cycle weighting coefficient. This represents a preset empirical constant for material fatigue, obtained by referring to the fatigue life guidelines for marine rotating machinery issued by the International Organization for Standardization and major classification societies, based on the engine type listed on the diesel engine's nameplate. This indicates the cumulative starting hours of the diesel engine. The preset design life hours for the diesel engine are provided by the diesel engine manufacturer. The degradation divergence index is obtained by weighting the life cycle weighting coefficient and the normalized response ratio, based on the following formula: in, The degradation divergence index represents the degree to which the mechanical structure of a diesel engine deviates from the absolute health benchmark at the current moment.

[0037] Based on the above, it should be noted that the mechanical structure of marine diesel engines will inevitably suffer progressive fatigue damage under the action of alternating stress over the years. Therefore, after eliminating external dynamic interference, it is natural to introduce the evolution law of time dimension to dynamically correct the state assessment standard. This constitutes the inherent physical logic of calculating the life cycle weighting coefficient.

[0038] The same vibration response deviation represents drastically different levels of physical damage in the break-in phase of a new machine versus the end of its lifespan. To quantify this difference, a life-cycle weighted coefficient is constructed by combining the ratio of the diesel engine's cumulative start-up hours to its design lifespan hours with a preset material fatigue empirical constant. This parameter directly reflects the current physical aging level of the equipment, and its technical effect is to endow purely static vibration monitoring with adaptive sensitivity that evolves dynamically over time.

[0039] A deeper exploration of the underlying variable relationships in this calculation process reveals that the cumulative starting hours of the diesel engine truly record the effective total span of the actual working load borne by the equipment. It exhibits a clear positive correlation with the life cycle weighting coefficient, meaning that as the actual operating hours continue to accumulate and extend, the value of the life cycle weighting coefficient continues to rise, which aligns with the objective physical laws of gradually increasing mechanical clearance and gradually decreasing material stiffness.

[0040] Based on this, the life cycle weighting coefficient is further multiplied and weighted with the normalized response ratio after taking the natural logarithm, thereby deriving the degradation divergence index. This index profoundly characterizes the true degree of degradation of the internal mechanical structure of the diesel engine at the current moment, which deviates from the absolute health benchmark. Its core technical effect lies in the integration of the dual dimensions of transient physical excitation response and long-cycle historical fatigue wear.

[0041] In this calculation formula, the normalized response ratio reflects the pure relative vibration intensity after removing external disturbances at the current moment. In order to capture the weak local micro-damage in the early stage of the machine, the normalized response ratio is subjected to natural logarithmic operation. The nonlinear mapping characteristics of the logarithmic function are used to significantly amplify the small abnormal increments in the early stage. The degradation divergence index calculated in this way, the life cycle weighting coefficient, and the logarithmized normalized response ratio all show a strict positive correlation.

[0042] This deep-level weighted calculation method, through the dual positive joint gain of time parameters and response parameters, not only achieves nonlinear and sensitive capture of early weak structural degradation characteristics, but also endows the monitoring model with the natural attribute of automatically improving the early warning and defense level as the service life of the diesel engine increases. This avoids the risk of missed reports caused by the overly sluggish judgment threshold of aging equipment, and ensures high-precision early warning capabilities throughout the entire life cycle of the equipment.

[0043] Step 5: Set a time-domain sliding window with a fixed number of sampling points, continuously input the real-time output degradation divergence index time series data, construct a scanning cursor based on the interquartile range and the number of sampling points in the time-domain sliding window, and use the scanning cursor to slide and traverse from the minimum to the maximum value in the time-domain sliding window data to define the dominant response interval of the degradation divergence index, and output the arithmetic mean of this interval as the degradation index. The preset fixed sampling point number is set by combining the diesel engine's operating speed and the actual generation frequency of the degradation divergence index. Assuming that the monitoring device is required to output a confirmation result within one minute after the anomaly occurs, and the continuous generation frequency of the index is once per second, the fixed sampling point number is strictly set to 60. In specific edge monitoring implementation scenarios, the value range of the preset fixed sampling point number is preferably limited to between 30 and 120. In this embodiment, based on the computing power limitation of the ship's engine room microprocessor and the frequency of regular data transmission, the preset fixed sampling point number is set to 50. In this embodiment, a scanning cursor is constructed based on the interquartile range and the number of sampling points within the time-domain sliding window. The scanning cursor slides and traverses the data from the minimum to the maximum value within the time-domain sliding window to define the dominant response interval of the degradation divergence index. The specific logic is as follows: Input degradation divergence index time series data into a time-domain sliding window with a fixed number of sampling points, extract the interquartile range of the degradation divergence index within the time-domain sliding window, and calculate the vernier width based on the fixed number of sampling points in the time-domain sliding window. The formula used is as follows: in, Indicates the cursor width. The interquartile range represents the degradation divergence index within a time-domain sliding window. The interquartile range is the difference between the upper and lower quartile values ​​of the degradation divergence index time-series data within the time-domain sliding window. This indicates the preset fixed number of sampling points; Generate a range-scanning cursor based on the cursor width; Using the minimum value of the degradation divergence index within the time-domain sliding window as the starting boundary, the value range scanning cursor is controlled to continuously overlap and slide along the direction of numerical increase with the cursor width as the step size, thus dividing the area into multiple overlapping sub-intervals. The number of degradation divergence indices contained in each sub-interval is counted one by one; the number of degradation divergence indices in each sub-interval is compared, and the sub-interval with the largest number of degradation divergence indices is selected as the dominant response interval. If there are two or more sub-intervals with the same maximum value for the number of degradation divergence indices, then the sub-interval with the largest upper limit of the value range is selected as the dominant response interval. The upper limit of the value range is the critical value at the right end of the value range of a single sub-interval.

[0044] Based on the above, it should be noted that when processing continuously generated deterioration divergence index time series, due to the extreme extreme values ​​often caused by wave impact or transient mechanical pulses under complex ship operating conditions, if the global arithmetic mean is directly calculated for all data within the preset time domain sliding window, the calculation result is easily affected by outlier noise and the overall baseline shift occurs. Therefore, establishing the highest density distribution area of ​​the data in the value domain space by statistical frequency to extract the dominant response interval becomes an objective requirement for filtering transient interference and characterizing the real physical mechanical steady state.

[0045] To achieve this high-precision state extraction, extracting the interquartile range of data within the time-domain sliding window and combining it with a fixed number of sampling points to calculate the specific width of the value range scanning vernier is crucial. In the calculation logic of this formula, the vernier width and the interquartile range, which represents the macroscopic divergence of the data, show a positive correlation. To ensure that the vernier width accurately maps the physical properties of the diesel engine's background vibration, which approximates a Gaussian distribution, the formula sets a constant 2 as the base ratio coefficient. This specific value is directly controlled by the rigid transformation ratio between the interquartile range and the standard deviation in the normal distribution.

[0046] Specifically, the interquartile range, in its absolute definition, only encompasses the middle 50% of the data distribution span. In the standard Gaussian distribution model, this 50% core data span is mathematically equal to approximately 1.35 times the standard deviation. To enable the vernier width to span the local core data and accurately define the complete main peak of the entire mechanical steady state, the interquartile range must be expanded by a specific ratio. When the scaling factor is strictly set to a constant of 2, the calculated vernier width is twice the interquartile range, which is mathematically equivalent to approximately 2.7 times the standard deviation span. In data statistics, 2.7 times the standard deviation precisely constitutes the optimal geometric critical boundary for distinguishing between high-frequency, dense true steady-state signals and long-tailed random outlier noise. If this basic multiplier coefficient is discarded, it will directly undermine the unbiased premise of state extraction. If the value is too small, it will cause the vertex to be too narrow, which will easily misjudge high-frequency transient clutter as the dominant response. If the value is too large, it will cause serious over-smoothing, which will forcibly incorporate the weak early degradation divergence exponential characteristics into the normal steady-state region and thus mask them.

[0047] In the above calculation logic, the cursor width and the fixed number of sampling points in the time-domain sliding window exhibit a negative exponential correlation. If the cursor width depends only on the base multiplier and the interquartile range, its span will encompass too much broad data, causing the algorithm to lose its ability to extract and locate the true steady-state peak from the data set. Therefore, the formula introduces the negative third power of the fixed number of sampling points as a nonlinear convergence factor. The setting of this specific exponent directly follows the objective geometric dimensionality reduction law when the sample information capacity is mapped to a one-dimensional numerical space.

[0048] Specifically, the fixed number of sampling points within the time-domain sliding window constitutes a three-dimensional information benchmark for evaluating the mechanical state in terms of statistical characteristics. When it is necessary to accurately locate the densest physical peaks in a single-dimensional value space, it is essential to ensure that the value-domain scanning cursor still has sufficient local data density to support it during the narrowing process. According to the mathematical relationship of this dimensionality reduction mapping, the total number of samples must exhibit a cubic growth, and only then can the linear feature density accumulated in the one-dimensional space be sufficient to support the cursor width to perform a precise proportional reduction of half while maintaining the same statistical confidence. Therefore, the adaptive convergence ratio of the cursor width is strictly locked to the reciprocal of the cube root of the fixed number of sampling points, i.e., the power of -3 / 1. Using any other attenuation rate will lead to physical mismatch. If the absolute value of the attenuation exponent is too large, the cursor will collapse rapidly with the increase of the number of samples, shattering the continuous steady-state features into isolated noise points with no statistical significance. If the absolute value of the attenuation exponent is too small, it is impossible to effectively utilize the sample bonus for convergence, resulting in the local early degradation peaks always being masked by the broad background. This dynamic mechanism, which deeply integrates macroscopic divergence characteristics with negative third power-law dimensionality reduction convergence, enables the cursor to automatically sharpen as the sample size increases.

[0049] Based on the processed cursor width, a range scanning cursor is generated and a continuous overlapping sliding scan is performed from the minimum value to the maximum value. This not only overcomes the boundary truncation effect caused by conventional static equal-width grid cutting, but also ensures from the mathematical level that no matter how the deterioration divergence index drifts and evolves in the range, the densest physical dominant features that have not been cut can be completely captured.

[0050] By statistically analyzing the number of degradation divergence indices falling within each sub-interval and selecting the sub-interval with the highest number as the dominant response interval, the highest frequency consensus in statistics is used to confirm the main state, thus eliminating the statistical weight of sparsely distributed transient anomalies that interfere with subsequent calculations. Furthermore, when two or more sub-intervals have the same maximum value for their degradation divergence indices, selecting the sub-interval with the largest upper limit of the value range as the dominant response interval is a prudent judgment strategy that tends to reveal a more severe degradation trend when features of equal density overlap.

[0051] By extracting only the degradation divergence index within the dominant response interval, performing arithmetic averaging, and outputting it as the degradation index, the path of random discrete noise participating in feature calculation is cut off from the mathematical level. This accurately extracts the core data representing the stable operating trajectory of the diesel engine's mechanical structure, significantly improving the fault diagnosis accuracy and baseline output stability under the interference of harsh marine environments.

[0052] Step 6: Compare the degradation index with the preset degradation index threshold, and determine the health status of the diesel engine based on the comparison results; In this embodiment, the degradation index is compared with a preset degradation index threshold: If the degradation index is greater than or equal to the degradation index threshold, the diesel engine is determined to be in a faulty state. If the degradation index is less than the degradation index threshold, the diesel engine is considered to be in a healthy state. The degradation index threshold is obtained by adding 20% ​​to the baseline value of the degradation index obtained from historical data.

[0053] Please see Figure 2 The present invention also provides a condition monitoring device for a marine diesel engine, the device being used to perform the condition monitoring method for a marine diesel engine as described in any of the preceding claims, comprising: Data acquisition module: Real-time acquisition of triaxial vibration velocity components, triaxial DC gravity components and scavenging pressure at different times during diesel engine operation, and calculation of triaxial composite vibration velocity based on triaxial vibration velocity components; Start-stop status judgment module: compares the triaxial composite vibration velocity with the preset basic start-stop threshold, obtains the diesel engine start-stop status based on the comparison result, and counts the cumulative start-up time of the diesel engine; Data decoupling module: Calculates the ship's heel angle based on the three-axis DC gravity components, normalizes the three-axis composite vibration velocity and scavenging pressure with the corresponding preset reference parameters, and decouples the calculation based on the ship's heel angle to obtain the normalized response ratio; Data processing module: Based on the cumulative start-up time of the diesel engine, combined with the preset material fatigue experience constant and the diesel engine design life time, the life cycle weighting coefficient is obtained, and the normalized response ratio is used for weighted calculation. The degradation divergence index is generated in real time at each moment, forming a continuously changing degradation divergence index time series. Interval Extraction Module: A time-domain sliding window with a fixed number of sampling points is preset, continuously inputting the real-time output degradation divergence index time-series data, constructing a scanning cursor based on the interquartile range and the number of sampling points in the time-domain sliding window, and traversing the data in the time-domain sliding window from the minimum value to the maximum value to define the dominant response interval of the degradation divergence index, and outputting the arithmetic mean of the data in this interval as the degradation index. Health status assessment module: compares the degradation index with the preset degradation index threshold, and judges the health status of the diesel engine based on the comparison result.

[0054] 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.

[0055] 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.

[0056] 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.

[0057] 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 condition monitoring of marine diesel engines, characterized in that, The specific steps include: Step 1: Real-time acquisition of triaxial vibration velocity components, triaxial DC gravity components, and scavenging pressure at different times during diesel engine operation; calculate the triaxial composite vibration velocity based on the triaxial vibration velocity components. Step 2: Compare the triaxial composite vibration velocity with the preset basic start-stop threshold, obtain the diesel engine start-stop status based on the comparison results, and count the cumulative start-up time of the diesel engine; Step 3: Calculate the ship's heel angle based on the three-axis DC gravity components, normalize the three-axis composite vibration velocity and scavenging pressure with the corresponding preset reference parameters, and decouple the calculations based on the ship's heel angle to obtain the normalized response ratio. Step 4: Based on the cumulative start-up time of the diesel engine, combined with the preset material fatigue experience constant and the diesel engine design life time, the life cycle weighting coefficient is obtained, and the normalized response ratio is used for weighted calculation. The degradation divergence index is generated in real time at each moment, forming a continuously changing degradation divergence index time series. Step 5: Set a time-domain sliding window with a fixed number of sampling points, continuously input the real-time output degradation divergence index time series data, construct a scanning cursor based on the interquartile range and the number of sampling points in the time-domain sliding window, and use the scanning cursor to slide and traverse from the minimum to the maximum value in the time-domain sliding window data to define the dominant response interval of the degradation divergence index, and output the arithmetic mean of the data in this interval as the degradation index. Step 6: Compare the degradation index with the preset degradation index threshold, and judge the health status of the diesel engine based on the comparison results.

2. The condition monitoring method for marine diesel engines according to claim 1, characterized in that: A spatial geometric coordinate system is established with the geometric center point of the diesel engine as the origin, the X and Y axes are set in the horizontal direction, and the Z axis is set in the vertical direction. The triaxial vibration velocity components include the effective values ​​of the vibration velocities of the diesel engine's X-axis, Y-axis, and Z-axis. The three-dimensional spatial vector magnitude calculation is used to calculate the triaxial composite vibration velocity from the triaxial vibration velocity components.

3. The condition monitoring method for marine diesel engines according to claim 2, characterized in that: The triaxial composite vibration velocity is compared with the preset basic start-stop threshold. When the triaxial composite vibration velocity is greater than or equal to the basic start-stop threshold, the diesel engine is determined to be in operation. The running time of this start-up is continuously accumulated according to the sampling interval. The accumulated start-up time is the value after the last shutdown settlement. When the triaxial composite vibration velocity is less than the basic start-stop threshold, the diesel engine is determined to be in a stopped state. The accumulated running time of this start is added to the total machine start time, and the running time of this start is reset to zero, waiting for the next start to reset the timing.

4. The condition monitoring method for marine diesel engines according to claim 1, characterized in that: The ship's roll angle is calculated based on the three-axis DC gravity components, and the specific logic is as follows: The three-axis DC gravity components include the DC gravity components of the X-axis, Y-axis, and Z-axis, and the preset three-axis reference DC gravity components include the reference DC gravity components corresponding to the X-axis, Y-axis, and Z-axis. The collected X, Y, and Z axis DC gravity components are combined to form a measured triaxial DC gravity vector, and the collected three reference DC gravity components are combined to form a reference triaxial DC gravity vector. Calculate the dot product of the two sets of spatial vectors, then solve for the magnitudes of the two vectors separately. Divide the dot product of the vectors by the product of the magnitudes of the two vectors to obtain the cosine of the included angle. Convert the angle of heel of the ship by the inverse cosine operation.

5. The condition monitoring method for marine diesel engines according to claim 1, characterized in that: The normalized response ratio is obtained, and the specific logic is as follows: the preset reference parameters include the reference triaxial composite vibration velocity and the reference scavenging pressure; The ratio of the triaxial composite vibration velocity to the preset reference triaxial composite vibration velocity is calculated to obtain the normalized triaxial composite vibration velocity. Calculate the ratio of the scavenging pressure to the preset reference scavenging pressure to obtain the normalized scavenging pressure; Calculate the square of the ratio of the normalized triaxial composite vibration velocity to the normalized scavenging pressure, and multiply it by the square of the cosine of the ship's heel angle to obtain the normalized response ratio.

6. The condition monitoring method for marine diesel engines according to claim 5, characterized in that: The life cycle weighting coefficient is obtained based on the following formula: in, Represents the life-cycle weighting coefficient. This represents a preset empirical constant for material fatigue. This indicates the cumulative starting hours of the diesel engine. This indicates the preset design life hours for the diesel engine; The degradation divergence index is obtained by weighting the life cycle weighting coefficient and the normalized response ratio, based on the following formula: in, Indicates the degradation divergence index. This represents the normalized response ratio.

7. The condition monitoring method for marine diesel engines according to claim 1, characterized in that: A scanning cursor is constructed based on the interquartile range and the number of sampling points within the time-domain sliding window. The scanning cursor slides across the data within the time-domain sliding window from the minimum to the maximum value to define the dominant response interval of the degradation divergence index. The specific logic is as follows: Input degradation divergence index time series data into a time-domain sliding window with a fixed number of sampling points, extract the interquartile range of the degradation divergence index within the time-domain sliding window, and calculate the vernier width based on the fixed number of sampling points in the time-domain sliding window. The formula used is as follows: in, Indicates the cursor width. The interquartile range represents the degradation divergence index within the time-domain sliding window. This indicates the preset fixed number of sampling points; Generate a range-scanning cursor based on the cursor width; Using the minimum value of the degradation divergence index within the time-domain sliding window as the starting boundary, the value range scanning cursor is controlled to continuously overlap and slide along the direction of numerical increase with the cursor width as the step size, thus dividing the area into multiple overlapping sub-intervals. Count the number of degradation divergence indices contained in each sub-interval; Compare the number of degradation divergence indices in each sub-interval, and select the sub-interval with the largest number of degradation divergence indices as the dominant response interval. If there are two or more sub-intervals with the same maximum value for the number of degradation divergence indices, then the sub-interval with the largest upper limit of the value range is selected as the dominant response interval. The upper limit of the value range is the critical value at the right end of the value range of a single sub-interval.

8. The condition monitoring method for marine diesel engines according to claim 1, characterized in that: The degradation index is compared with the preset degradation index threshold: If the degradation index is greater than or equal to the degradation index threshold, the diesel engine is determined to be in a faulty state. If the degradation index is less than the degradation index threshold, the diesel engine is considered to be in a healthy state.

9. A condition monitoring device for marine diesel engines, characterized in that: The device is used to perform the condition monitoring method for marine diesel engines according to any one of claims 1-8, including: Data acquisition module: Real-time acquisition of triaxial vibration velocity components, triaxial DC gravity components and scavenging pressure at different times during diesel engine operation, and calculation of triaxial composite vibration velocity based on triaxial vibration velocity components; Start-stop status judgment module: compares the triaxial composite vibration velocity with the preset basic start-stop threshold, obtains the diesel engine start-stop status based on the comparison result, and counts the cumulative start-up time of the diesel engine; Data decoupling module: Calculates the ship's heel angle based on the three-axis DC gravity components, normalizes the three-axis composite vibration velocity and scavenging pressure with the corresponding preset reference parameters, and decouples the calculation based on the ship's heel angle to obtain the normalized response ratio; Data processing module: Based on the cumulative start-up time of the diesel engine, combined with the preset material fatigue experience constant and the diesel engine design life time, the life cycle weighting coefficient is obtained, and the normalized response ratio is used for weighted calculation. The degradation divergence index is generated in real time at each moment, forming a continuously changing degradation divergence index time series. Interval Extraction Module: A time-domain sliding window with a fixed number of sampling points is preset, continuously inputting the real-time output degradation divergence index time-series data, constructing a scanning cursor based on the interquartile range and the number of sampling points in the time-domain sliding window, and traversing the data in the time-domain sliding window from the minimum value to the maximum value to define the dominant response interval of the degradation divergence index, and outputting the arithmetic mean of the data in this interval as the degradation index. Health status assessment module: compares the degradation index with the preset degradation index threshold, and judges the health status of the diesel engine based on the comparison result.

Images