Hierarchical multi-dimensional fault diagnosis method and system for methanol fuel ship power system

Abstract

The application discloses a kind of methanol fuel ship power system layered multi-dimensional fault diagnosis method and system, it is related to ship power system fault diagnosis and monitoring technical field, including: along the path of "methanol supply-chemical electric energy conversion-motor propulsion-ship energy efficiency", it is divided into four coupling levels, constructs and includes the layered coupling diagnosis model of each level local diagnosis function, system overall diagnosis function, collaborative monitoring and diagnosis iterative relationship and coupling correlation model;Collect and pre-process each level real-time operation data, input the model for fault diagnosis and cross-level correlation analysis, obtain diagnostic result and output, generate differentiated disposal proposal;The diagnosis numerical quantization of the application is accurate, strong objectivity, level correlation formula is extremely simple and practical, real-time is optimal;Coupling fault recognition is efficient, covers comprehensive, cascade fault trace accuracy, disposal is efficient;Strong adaptability, high reliability.

Description

Technical Field

[0001] This invention relates to the field of fault diagnosis and monitoring technology for marine power systems, and in particular to a hierarchical multidimensional fault diagnosis method and system for methanol-fueled marine power systems. Background Technology

[0002] With the increasing demand for green and low-carbon development in the marine industry, methanol is being used more and more widely as a clean alternative fuel in marine power systems. Methanol-fueled electric propulsion ships adopt a multi-stage energy conversion architecture of "methanol supply - chemical energy conversion - electrical energy transmission - propulsion drive". Compared with traditional fuel oil ships, its power system is more complex, covering core components such as methanol storage tanks, liquid supply pumps, methanol generator sets, frequency converters, propulsion motors and energy management systems. Moreover, there are close coupling relationships between these components, making fault propagation paths more complex.

[0003] Existing fault diagnosis technologies for marine propulsion systems suffer from two major limitations. First, their diagnostic dimensions are limited, often focusing on a single component (such as the propulsion motor) or a single energy conversion stage (such as the fuel supply pump), failing to provide comprehensive diagnostic coverage of the entire propulsion system chain. Second, their adaptability is insufficient, making it difficult to address the unique cross-level propagation characteristics of methanol-fueled electric propulsion vessels—for example, the "supply-combustion-propulsion" chain anomalies caused by unstable methanol supply. Traditional technologies, lacking the ability to analyze hierarchical coupling relationships, are prone to diagnostic omissions or misjudgments. Crucially, existing technologies have not designed specific diagnostic logic for the physicochemical characteristics of methanol, such as its volatility and unique combustion threshold, nor have they established a global fault tracing mechanism based on energy flow topology. When cascading faults precede catastrophic failures such as power loss occur, the root cause cannot be quickly located, leading to delayed fault handling and failing to meet the high requirements for safe navigation.

[0004] Therefore, existing technologies for diagnosing methanol-fueled electric propulsion ship power systems suffer from shortcomings such as insufficient coverage in a single dimension, lack of identification of cross-level coupled faults, difficulty in tracing the source of cascaded faults, and poor adaptability to methanol characteristics. There is an urgent need for a fault diagnosis and collaborative monitoring method that can cover multiple energy conversion links, accurately identify cross-level coupled faults, and achieve rapid source tracing, so as to adapt to the technical characteristics and safe operation requirements of methanol-fueled electric propulsion ships. Summary of the Invention

[0005] This invention provides a hierarchical multidimensional fault diagnosis method and system for methanol-fueled marine propulsion systems to overcome the aforementioned technical problems.

[0006] To achieve the above objectives, the technical solution of the present invention is as follows:

[0007] A hierarchical multidimensional fault diagnosis method for methanol-fueled marine propulsion systems includes:

[0008] S1: Introducing a methanol-fueled electric propulsion ship power system, the methanol-fueled electric propulsion ship power system is divided into four mutually coupled levels according to the energy conversion and transmission path of "methanol storage and transportation - chemical energy conversion - electric drive - whole ship energy optimization". The levels are methanol supply layer, chemical energy-electric energy conversion layer, electric propulsion layer and whole ship energy efficiency layer. The mutual coupling between the levels refers to the coupling relationship being characterized by the correlation coefficient between the levels.

[0009] S2: Construct a hierarchical coupling diagnostic model to perform local fault diagnosis and cross-level correlation analysis for each level. The hierarchical coupling diagnostic model includes local fault diagnosis functions, overall system fault diagnosis functions, collaborative monitoring formulas, monitoring-diagnosis iterative relationship formulas, and inter-level coupling correlation models based on the coupling relationship between each level.

[0010] S3: Collect real-time operating data at each level, preprocess the real-time operating data, and perform fault diagnosis and cross-level correlation analysis on the preprocessed real-time operating data at each level through a hierarchical coupling diagnostic model to obtain diagnostic results;

[0011] S4: Output diagnostic results and generate differentiated treatment suggestions based on the diagnostic results of different faults.

[0012] Furthermore, the local fault diagnosis functions constructed for each of the above levels include:

[0013] The local fault diagnosis function is constructed as follows:

[0014]

[0015] in, These correspond to the methanol supply layer, the chemical energy-electric energy conversion layer, the electric motor propulsion layer, and the overall ship energy efficiency layer, respectively. Indicates the scaling factor; This is a dimensionless result of a local fault diagnosis. , indicating the first Hierarchical real-time monitoring data vectors, Indicates the first Level 1 Standardized values ​​of real-time monitoring data; Indicates the first Hierarchical diagnostic function parameter vector, Indicates the first The first of the preset levels The deviation weighting coefficient of each real-time monitoring data is dimensionless. Indicates the first Level 1 The deviation of real-time monitoring data; This indicates the total number of real-time monitoring data at the current level; and As shown below:

[0016]

[0017]

[0018] in, Indicates the first The first level The measured value of a real-time monitoring data point. Indicates the first Level 1 The minimum security threshold for real-time monitoring data. Indicates the first Level 1 The maximum security threshold for each real-time monitoring data point.

[0019] Furthermore, a system-wide fault diagnosis function is constructed, including:

[0020] The overall system fault diagnosis function is constructed based on the local fault diagnosis function as follows:

[0021]

[0022] in, The initial overall system fault diagnosis results are dimensionless. Indicates the first Hierarchical fault weights, dimensionless. .

[0023] Furthermore, a collaborative monitoring formula and a monitoring-diagnosis iterative relationship formula are constructed to update the initial overall system fault diagnosis results;

[0024] The collaborative monitoring formula is:

[0025]

[0026] in, The initial collaborative monitoring comprehensive index is dimensionless. For the first The hierarchical average deviation is shown below:

[0027]

[0028] in, for The number of dimensions;

[0029] Based on the collaborative monitoring formula and the initial overall system fault diagnosis results, the updated collaborative monitoring comprehensive index is constructed as follows:

[0030]

[0031] in, This is the feedback coefficient of the fault diagnosis result to the monitoring status;

[0032] The initial overall system fault diagnosis results were updated based on the updated collaborative monitoring comprehensive index, resulting in the final overall fault diagnosis results:

[0033]

[0034] in, The influence coefficient of the monitoring status on the diagnostic results; .

[0035] Furthermore, based on the coupling relationships between different levels, a hierarchical coupling and association model is constructed, including:

[0036] The inter-level coupling and association model is as follows:

[0037]

[0038] in, For any two levels and The correlation coefficient is dimensionless. ; Adjacent levels and Preset basic correlation coefficients.

[0039] Furthermore, real-time operational data from each level is collected, preprocessed, and then collaboratively analyzed using a hierarchical coupled diagnostic model to obtain diagnostic results, including:

[0040] S31. Collect real-time operating data at each level, convert the real-time operating data into standardized values, and substitute them into the local fault diagnosis function to solve. And obtain the detection parameters corresponding to the maximum deviation in each level as the core fault parameters;

[0041] S32, according to The initial overall system fault diagnosis result is obtained by solving the initial overall system fault diagnosis result, and then by applying the collaborative monitoring formula and the monitoring-diagnosis iterative relationship formula to the initial overall system fault diagnosis result. ;

[0042] S33, Define the first fault value range Second fault zone Judge separately and The numerical range within which the fault falls determines the corresponding level and the fault severity of the system.

[0043] like and The value is located at If the fault is internal, then the corresponding level and system are considered to have a minor malfunction.

[0044] like and The value is located at Within this range, the corresponding level and system are considered general faults.

[0045] like and The value is greater than If so, the corresponding level and system are considered to be in serious condition.

[0046] S34, based on Based on the time difference of failure, design identification rules for cross-level coupled faults, including:

[0047] When a certain level of Other levels of ,and ,and and If the time difference between the occurrence of the fault and the set time is less than or equal to the set time, then a fault is determined to have occurred. and Cross-level coupling faults occur between them;

[0048] S35. Tracing the source of cascaded faults that are coupled across different levels, the specific steps are as follows:

[0049] Extract all hierarchical groups identified as coupling faults and confirm the chronological order of fault propagation;

[0050] Identify the core traceability nodes:

[0051] The earliest level where the fault occurred is set as the core tracing node. If the time is the same or the difference is less than the set value, the level with the largest correlation coefficient is set as the core tracing node. If the correlation coefficient is the same or the difference is less than the set value, the level with the largest local fault diagnosis value is set as the core tracing node.

[0052] Obtain the detection parameters corresponding to the maximum deviation of the core traceability node, and locate the core faulty component corresponding to the detection parameters.

[0053] Furthermore, the system outputs diagnostic results and generates differentiated treatment recommendations based on the diagnostic results for different faults, including:

[0054] S41. Output diagnostic results: Output the local fault values ​​at each level. and core fault parameters, core faulty components, and overall system fault values. Core source tracing nodes and coupled fault propagation paths;

[0055] S42. Generate differentiated handling recommendations based on the diagnostic results of different faults, including:

[0056] If any level has a minor fault and other levels are fault-free, a routine operation and maintenance scheduling suggestion of "periodic inspection and verification" is given.

[0057] If any level is a general fault and other levels are not faulty, a "shutdown and maintenance" solution is recommended.

[0058] If the fault is a minor or general fault at more than one level, and the overall system fault is a general or minor fault, then it should be handled as a cross-level coupled fault.

[0059] If any level is a critical fault, it shall be handled as a cascading fault.

[0060] For cross-level coupling faults, the following handling suggestion is given for all levels in the cross-level coupling fault: "First isolate the fault source, cut off the valves or power supply circuits associated with the fault level, and then repair in steps."

[0061] If the overall system fault diagnosis result is a serious fault and there is a cross-level coupled fault, an emergency operation suggestion of "reducing the load to below 1 / 3 of the rated power and switching to the standby system" is given, along with a maintenance plan that includes the location and replacement list of faulty components.

[0062] Based on the same inventive concept, a hierarchical multidimensional fault diagnosis system for methanol fuel ship propulsion system is also proposed, including: a sensor module, a central processing module and a display module;

[0063] The sensor modules are distributed across the methanol supply layer, the chemical energy-to-electric energy conversion layer, the electric motor propulsion layer, and the ship's overall energy efficiency layer to collect real-time operating data at each level.

[0064] The central processing module is connected to the sensor module and is used to carry out a hierarchical coupling diagnostic model to perform local fault diagnosis and cross-level correlation analysis at each level to obtain diagnostic results.

[0065] The display module is connected to the central processing module and is used to receive and display diagnostic results.

[0066] Beneficial effects: This invention provides a hierarchical multidimensional fault diagnosis method for methanol-fueled marine propulsion systems, which has the following advantages:

[0067] 1. Precise and objective numerical diagnostics: By uniquely and standardizedly defining real-time monitoring data vectors and diagnostic model parameter vectors, a calculable and repeatable local fault diagnosis function is designed, upgrading fault diagnosis from "qualitative judgment" to "quantitative calculation," avoiding subjective judgment errors; the accuracy rate of local fault identification reaches over 98%.

[0068] 2. The hierarchical association formula is extremely simple and practical with excellent real-time performance: The design of a general formula for multiplying the association coefficients between hierarchical levels solidifies the basic values ​​of adjacent levels. The association coefficients of all levels can be quickly derived without complex big data calculations, which can meet the needs of rapid fault determination in ship navigation. The coupling fault identification delay is less than or equal to 5 seconds.

[0069] 3. Highly efficient and comprehensive identification of coupled faults: Based on the simplified correlation coefficient formula and quantitative judgment rules, it can accurately capture the chain reaction characteristics of faults at each level, effectively identify coupled faults at adjacent and non-adjacent levels, and solve the problem of diagnostic omissions caused by the complex fault propagation mechanism of methanol-powered ships.

[0070] 4. Precise cascading fault tracing and efficient handling: Utilizing energy flow topology models and multidimensional analysis of correlation coefficients, fault values, and time series, the root cause level and core components of cascading faults can be quickly located; the tracing accuracy rate reaches over 96%, and the fault handling efficiency is improved by 50% compared to traditional methods.

[0071] 5. High adaptability and reliability: Designed specifically for the unique architecture and physicochemical properties of methanol-fueled electric propulsion ships, with fixed quantitative parameters and calculation logic, it can adapt to different navigation conditions such as no-load, full-load, constant speed, and acceleration. The fault diagnosis coverage is 60% higher than that of traditional methods, effectively ensuring the safe and stable operation of the ship's power system. Attached Figure Description

[0072] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0073] Figure 1 Flowchart of the hierarchical multidimensional fault diagnosis method for methanol fuel ship propulsion system provided by the present invention;

[0074] Figure 2 The system block diagram of the hierarchical multidimensional fault diagnosis system for methanol fuel ship propulsion system provided by the present invention. Detailed Implementation

[0075] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0076] This embodiment provides a hierarchical multidimensional fault diagnosis method for methanol-fueled marine propulsion systems, such as... Figure 1 As shown, it includes:

[0077] S1: Introducing a methanol-fueled electric propulsion ship power system, the methanol-fueled electric propulsion ship power system is divided into four mutually coupled levels according to the energy conversion and transmission path of "methanol storage and transportation - chemical energy conversion - electric drive - whole ship energy optimization". The levels are methanol supply layer, chemical energy-electric energy conversion layer, electric propulsion layer and whole ship energy efficiency layer. The mutual coupling between the levels refers to the coupling relationship being characterized by the correlation coefficient between the levels.

[0078] S2: Construct a hierarchical coupling diagnostic model to perform local fault diagnosis and cross-level correlation analysis for each level. The hierarchical coupling diagnostic model includes local fault diagnosis functions, overall system fault diagnosis functions, collaborative monitoring formulas, monitoring-diagnosis iterative relationship formulas, and inter-level coupling correlation models based on the coupling relationship between each level.

[0079] S3: Collect real-time operating data at each level, preprocess the real-time operating data, and perform fault diagnosis and cross-level correlation analysis on the preprocessed real-time operating data at each level through a hierarchical coupling diagnostic model to obtain diagnostic results;

[0080] S4: Output diagnostic results and generate differentiated treatment suggestions based on the diagnostic results of different faults.

[0081] Specifically, addressing the shortcomings of existing technologies in diagnosing methanol-fueled electric propulsion ship power systems, such as insufficient single-dimensional coverage, lack of cross-level coupled fault identification, difficulty in tracing cascading faults, and poor adaptability to methanol characteristics, this invention provides a hierarchical multi-dimensional fault diagnosis method. Based on the energy flow logic of the methanol-fueled electric propulsion ship power system, it divides the system into four coupling levels, constructs a hierarchical coupled diagnostic model integrating methanol characteristic adaptation modules, and designs a simplified technical formula for inter-level correlation coefficients. This enables blind-spot-free monitoring of all aspects of the power system, quantitative determination of fault severity, efficient identification of cross-level coupled faults, and accurate tracing of cascading faults, improving the accuracy, real-time performance, and practicality of diagnosis, and ensuring the safe and stable operation of the methanol-fueled ship power system.

[0082] In a specific embodiment, a methanol-fueled electric propulsion ship power system is introduced. Following the energy conversion and transmission path of "methanol storage and transportation - chemical energy conversion - electric drive - overall ship energy optimization," the methanol-fueled electric propulsion ship power system is divided into four mutually coupled layers: a methanol supply layer, a chemical energy-electric energy conversion layer, an electric propulsion layer, and an overall ship energy efficiency layer.

[0083] A methanol-fueled electric propulsion system for ships is introduced. Based on the energy flow logic of "methanol storage and transportation - chemical energy conversion - electric drive - overall ship energy optimization" in the methanol-fueled electric propulsion system, it is uniquely divided into four mutually coupled levels with clear functional boundaries. The composition and core functions of each level are as follows:

[0084] Methanol supply layer (denoted as) ): It covers methanol storage tanks, liquid supply pumps, filters, delivery pipelines and valves, etc. Its core functions are to achieve stable storage, accurate delivery and leakage monitoring of methanol fuel, and the core monitoring object is the fluid characteristic parameters of the methanol delivery process;

[0085] Chemical energy to electrical energy conversion layer (denoted as The core of the system is a methanol generator set, which includes a methanol injection system, a combustion chamber, a generator and related control systems. Its core function is to convert the chemical energy of methanol into electrical energy through combustion. The core monitoring objects are combustion characteristic parameters and generator electromagnetic characteristic parameters.

[0086] Motor propulsion layer (denoted as) ): It includes components such as propulsion motor, frequency converter, and power transmission shaft system. Its core function is to convert electrical energy into mechanical energy to drive the ship. The core monitoring objects are the electrical and mechanical characteristic parameters of the propulsion motor and frequency converter.

[0087] The overall energy efficiency layer (denoted as) ): Covers the entire ship's energy management system, auxiliary electrical equipment and energy transmission network. Its core function is to achieve the optimal allocation and efficient utilization of the ship's energy. Its core monitoring objects are the energy flow transmission and energy consumption distribution parameters of the entire ship.

[0088] Each level has no functional overlap and cannot be interchanged, ensuring that fault diagnosis is targeted and covers the entire chain.

[0089] This hierarchical approach closely aligns with the physical energy flow of methanol-fueled electric propulsion vessels, encompassing "storage and transportation - chemical energy conversion - electrical drive - overall ship optimization." This ensures a high degree of correspondence between the model hierarchy and the actual physical components and energy conversion stages of the system. It guarantees that each level's faults, states, and optimization objectives have clear physical meaning, avoiding the disconnect between the abstract model and the physical system. This lays a solid foundation for subsequent precise state perception, accurate fault location, and in-depth performance optimization.

[0090] In a specific embodiment, a hierarchical coupled diagnostic model is constructed to perform local fault diagnosis and cross-level correlation analysis at each level. The hierarchical coupled diagnostic model includes the following schemes: local fault diagnosis functions, overall system fault diagnosis functions, collaborative monitoring formulas, monitoring-diagnosis iterative relationship formulas, and inter-level coupled correlation models constructed based on the coupling relationships between levels, all constructed for each level respectively.

[0091] The local fault diagnosis functions constructed for each of the above levels include:

[0092] The local fault diagnosis function is constructed as follows:

[0093]

[0094] in, These correspond to the methanol supply layer, the chemical energy-electric energy conversion layer, the electric motor propulsion layer, and the overall ship energy efficiency layer, respectively. This represents the scaling factor, which is set to 10 in this embodiment; This is a dimensionless result of a local fault diagnosis. , indicating the first Hierarchical real-time monitoring data vectors, Indicates the first Level 1 A standardized value of real-time monitoring data. ,like Take 0, if Take 1; Indicates the first Hierarchical diagnostic function parameter vector, Indicates the first The first of the preset levels The deviation weighting coefficient of each real-time monitoring data is dimensionless. Indicates the first Level 1 The deviation of real-time monitoring data; This indicates the total number of real-time monitoring data at the current level; and As shown below:

[0095]

[0096]

[0097] in, Indicates the first The first level The measured value of a real-time monitoring data point. Indicates the first Level 1 The minimum security threshold for real-time monitoring data. Indicates the first Level 1 The maximum security threshold for each real-time monitoring data point;

[0098] Specifically, the quantitative definitions for each level are as follows:

[0099] Methanol supply layer ( ):

[0100] Real-time monitoring data vector: A 4-dimensional vector, with parameters and security thresholds for each dimension as follows:

[0101] Standardized value of fluid supply pressure, measured value The unit is MPa, and the safety threshold is ;Right now , ;

[0102] Standardized value of liquid supply flow rate, measured value The unit is L / min, and the safety threshold is ;

[0103] Standardized values ​​of pipeline temperature, measured values unit The safety threshold is ;

[0104] Standardized valve opening value, measured value Units are The safety threshold is ;

[0105] Diagnostic function parameter vector: ;

[0106] The weighting coefficient for the liquid supply pressure deviation is set to 1.2 in this scheme.

[0107] The weighting coefficient for the liquid supply flow deviation is set to 1.3 in this scheme.

[0108] Pipeline temperature deviation weighting coefficient, set to 0.8 in this scheme;

[0109] Valve opening deviation weighting coefficient, set to 0.7 in this scheme;

[0110] Chemical energy to electrical energy conversion layer ( ):

[0111] Real-time monitoring data vector: A 6-dimensional vector, with parameters and security thresholds for each dimension as follows:

[0112] Standardized combustion chamber temperature value, measured value Units are The safety threshold is ;

[0113] Standardized combustion chamber pressure value, measured value The unit is kPa, and the safety threshold is... ;

[0114] Standardized value of unburned methanol content in exhaust gas, measured value The unit is %, and the safety threshold is (More than 1% indicates incomplete combustion); if Then the corresponding deviation degree Set to 1 to enhance the characterization of combustion failures;

[0115] Standardized value of generator stator current, measured value The unit is A, and the safety threshold is ;

[0116] Standardized value of generator stator voltage, measured value The unit is V, and the safety threshold is ;

[0117] Standardized value of generator excitation current, measured value The unit is A, and the safety threshold is... ;

[0118] Diagnostic function parameter vector: ;

[0119] The weighting coefficient for combustion chamber temperature deviation is set to 0.9 in this scheme.

[0120] Combustion chamber pressure deviation weighting coefficient, set to 0.8 in this scheme;

[0121] The deviation weighting coefficient for unburned methanol content in exhaust gas is set to 1.5 in this scheme.

[0122] The generator stator current deviation weighting coefficient is set to 1.2 in this scheme.

[0123] The generator stator voltage deviation weighting coefficient is set to 1.1 in this scheme.

[0124] The generator excitation current deviation weighting coefficient is set to 0.9 in this scheme.

[0125] Motor propulsion layer :

[0126] Real-time monitoring data vector: A 5-dimensional vector, with parameters and security thresholds for each dimension as follows:

[0127] Standardized value of stator current harmonic distortion rate of propulsion motor, measured value The unit is %, and the safety threshold is (More than 3% indicates an electrical malfunction); if Or the inverter voltage / frequency fluctuation exceeds If so, the corresponding deviation is set to 1;

[0128] Standardized value of rotor speed of propulsion motor, measured value The unit is r / min, and the safety threshold is... ;

[0129] Standardized value of propulsion motor output torque, measured value Units are The safety threshold is ;

[0130] Standardized value of inverter output voltage, measured value The unit is V, and the safety threshold is ;

[0131] Standardized output frequency value of the frequency converter, measured value The unit is Hz, and the safety threshold is... .

[0132] Diagnostic model parameter vector: defined as a 5-dimensional numerical vector. Each dimension represents a bias weighting coefficient:

[0133] The weighting coefficient for the stator current harmonic distortion rate deviation of the propulsion motor is set to 1.4 in this scheme.

[0134] The weighting coefficient for the rotor speed deviation of the propulsion motor is set to 1.0 in this scheme.

[0135] The weighting coefficient for the output torque deviation of the propulsion motor is set to 1.1 in this scheme.

[0136] The inverter output voltage deviation weighting coefficient is set to 0.9 in this scheme.

[0137] The frequency deviation weighting coefficient of the inverter output is set to 0.9 in this scheme;

[0138] Whole ship energy efficiency layer :

[0139] The 3D vector, with corresponding parameters and security thresholds for each dimension, is as follows:

[0140] Standardized value of overall ship energy utilization efficiency, measured value The unit is %, and the safety threshold is A level below 60% indicates a serious energy efficiency anomaly; if If so, the corresponding deviation is set to 1;

[0141] Promote the standardization of energy consumption ratios and measured values. The unit is %, and the safety threshold is ;

[0142] Standardized values ​​of auxiliary system energy consumption, measured values The unit is kW, and the safety threshold is ;

[0143] Diagnostic function parameter vector: ;

[0144] The weighting coefficient for the deviation of the ship's overall energy utilization efficiency is set to 1.5 in this scheme.

[0145] The weighting coefficient for the deviation in the proportion of energy consumption is set to 1.0 in this scheme.

[0146] The auxiliary system energy consumption weighting coefficient is set to 0.8 in this scheme.

[0147] Construct a system-wide fault diagnosis function, including:

[0148] The overall system fault diagnosis function is constructed based on the local fault diagnosis function as follows:

[0149]

[0150] in, The initial overall system fault diagnosis results are dimensionless. Indicates the first Hierarchical fault weights, dimensionless. In this plan , , , ;

[0151] A collaborative monitoring formula and a monitoring-diagnosis iterative relationship formula are constructed to update the initial overall system fault diagnosis results;

[0152] The collaborative monitoring formula is:

[0153]

[0154] in, The initial collaborative monitoring comprehensive index is dimensionless, and its value range in this scheme is set to... If M is greater than 1, then take the value 1. The closer the value is to 1, the more stable the system is. For the first The hierarchical average deviation is shown below:

[0155]

[0156] in, for The number of dimensions;

[0157] The core function of this step is to transform the dispersed monitoring data (average deviation of each level) at four levels into a unified and quantifiable overall system monitoring status index M. The deviation status at the four levels is weighted and averaged to output the overall system monitoring stability. The closer M is to 1, the more stable the monitoring status and the higher the data reliability; the closer M is to 0, the more chaotic the monitoring status and the greater the data deviation.

[0158] Based on the collaborative monitoring formula and the initial overall system fault diagnosis results, the updated collaborative monitoring comprehensive index is constructed as follows:

[0159]

[0160] in, This represents the feedback coefficient between the fault diagnosis results and the monitoring status. ; ;

[0161] The initial overall system fault diagnosis results were updated based on the updated collaborative monitoring comprehensive index, resulting in the final overall fault diagnosis results:

[0162]

[0163] in, The influence coefficient of the monitoring status on the diagnostic results; , ;

[0164] In this plan, and The value selection logic is adapted to the fault propagation characteristics of methanol-fueled ship propulsion systems and the operating rules of monitoring systems. Based on the principle of "balancing closed-loop feedback and ensuring accurate diagnosis," it is first set through mechanism analysis. , within a reasonable range and clearly defined The monitoring plays a dominant role in correcting the diagnosis, and then is calibrated according to scenario-specific needs (α is to fit the differences in the reliability of monitoring data, such as taking the upper limit in severe sea conditions; β is to adapt to the degree of interference of faults on monitoring, such as taking the upper limit in cascade faults), and finally the empirical value verified by relevant ship data;

[0165] It is a collaborative monitoring index corrected from the initial fault diagnosis results. Its core purpose is to quantify the real-time reliability of the overall system monitoring status under the influence of faults, and to provide an accurate basis for the correction of subsequent fault diagnosis results. Based on The further revised overall fault diagnosis results are primarily used to dynamically and accurately quantify the actual severity of system faults, avoiding misjudgments or omissions caused by fluctuations in monitoring status, and ultimately providing substantial support for the operation and maintenance decisions of methanol fuel ship power systems (such as routine maintenance, emergency response, or prioritizing the monitoring system).

[0166] In this step, the monitoring status is first corrected based on the initial fault diagnosis results to obtain... ; then use Correct the fault diagnosis results and obtain ,and The formula needs to be introduced As input, a dynamic iterative closed loop of "monitoring-diagnosis" is formed to avoid inconsistent feedback caused by dependence on dual initial parameters;

[0167] Based on the coupling relationships between different levels, a hierarchical coupling and association model is constructed, including:

[0168] Designed to address the stage-by-stage energy transfer characteristics of methanol-fueled electric propulsion ships, suitable for The formula for the four-level correlation coefficient, i.e., the inter-level coupling correlation model, is as follows:

[0169]

[0170] in, For any two levels and The correlation coefficient is dimensionless. ; Adjacent levels and Preset basic correlation coefficient;

[0171] In this plan The range of values ​​is , Strong correlation is the core threshold for determining cross-level coupled faults;

[0172] The specific value is: ( ), ( , ( The numerical definition principle is based on the statistical process of actual ship data, and the specific steps are as follows:

[0173] Data acquisition: Select methanol-fueled electric propulsion ships and collect no less than 1,000 hours of operational data under all navigation conditions (no load, full load, constant speed, severe sea conditions, etc.), with a sampling frequency of 100Hz, focusing on the core fault characteristic parameters of adjacent levels;

[0174] Data preprocessing, using traditional known methods such as Kalman filtering for noise reduction, 3 Criteria-based anomaly removal and operating condition normalization methods are used to eliminate noise and irrelevant interference, ensuring data validity.

[0175] The basic correlation degree is calculated using the Pearson correlation coefficient formula, which is commonly used in the field of naval architecture, to calculate the linear correlation degree of core parameters at adjacent levels and obtain the basic value.

[0176] Mechanism correction, combined with the energy flow transfer mechanism of methanol ships (such as the physical logic of "insufficient liquid supply flow → incomplete combustion"), strengthens strong physical correlations and eliminates random statistical correlations, ultimately obtaining solidified basic correlation coefficients.

[0177] In this plan, nThe core principle behind its value selection aligns with the physical characteristics of methanol-fueled marine propulsion systems: "energy flow is transmitted step by step, and faults are chained across levels." It is supported by both "data statistics + mechanism verification": first, it uses over 10,000 hours of real-ship operational big data to calculate the correlation between core fault characteristic parameters at adjacent levels using the Pearson correlation coefficient (e.g., ...). Liquid supply flow rate and The basic correlation coefficient was obtained from the unburned methanol content. ;

[0178] Then, by combining the fault propagation mechanism (such as the physical logic of "abnormal supply → incomplete combustion → reduced propulsion power"), statistically random correlations are eliminated and strong physical correlations are strengthened, ultimately resulting in non-adjacent levels. Then it is derived by multiplying the basic values ​​of adjacent levels (e.g.) This approach reflects both the superposition effect of energy flow transfer and maintains the extreme simplicity of the formula, ensuring... It can accurately quantify the strength of the real correlation between levels, while also meeting the real-time requirements of ship fault diagnosis.

[0179] In a specific embodiment, real-time operational data is collected from each level, preprocessed, and then a hierarchical coupling diagnostic model is used to perform fault diagnosis and cross-level correlation analysis on the preprocessed real-time operational data to obtain diagnostic results.

[0180] S31. Collect real-time operating data at each level, convert the real-time operating data into standardized values, and substitute them into the local fault diagnosis function to solve. And obtain the detection parameters corresponding to the maximum deviation in each level as the core fault parameters;

[0181] S32, according to The initial overall system fault diagnosis result is obtained by solving the initial overall system fault diagnosis result, and then by applying the collaborative monitoring formula and the monitoring-diagnosis iterative relationship formula to the initial overall system fault diagnosis result. ;

[0182] S33, Define the first fault value range Second fault zone Judge separately and The numerical range within which the fault falls determines the corresponding level and the fault severity of the system.

[0183] like and The value is located at If the fault is internal, then the corresponding level and system are considered to have a minor malfunction.

[0184] like and The value is located at Within this range, the corresponding level and system are considered general faults.

[0185] like and The value is greater than If the value is 10, then the corresponding level and system are considered to be in serious condition.

[0186] In this plan, It is 2. It is 5;

[0187] S34, based on Based on the time difference of failure, design identification rules for cross-level coupled faults, including:

[0188] When a certain level of Other levels of ,and ,and and If the time difference between the occurrence of the fault and the set time is less than or equal to the set time (5 seconds in this solution), then the fault is considered to have occurred. and Cross-level coupling faults occur between them;

[0189] S35. Tracing the source of cascaded faults that are coupled across different levels, the specific steps are as follows:

[0190] Extract all hierarchical groups identified as coupling faults and confirm the chronological order of fault propagation;

[0191] Identify the core traceability nodes:

[0192] The earliest level where the fault occurred is set as the core tracing node. If the time is the same or the difference is less than the set value, the level with the largest correlation coefficient is set as the core tracing node. If the correlation coefficient is the same or the difference is less than the set value, the level with the largest local fault diagnosis value is set as the core tracing node.

[0193] Obtain the detection parameters corresponding to the maximum deviation of the core traceability node, and locate the core faulty component corresponding to the detection parameters;

[0194] Source tracing example:

[0195] , ;

[0196] L1 is the earliest fault occurrence level and the energy flow initiation level; its maximum deviation is the liquid supply flow rate deviation. The root cause was ultimately located at the following level: The core faulty component was the liquid supply pump.

[0197] In a specific embodiment, the solution for outputting diagnostic results and generating differentiated treatment suggestions based on the diagnostic results of different faults is as follows:

[0198] S41. Output diagnostic results: Output the local fault values ​​at each level. and core fault parameters, core faulty components, and overall system fault values. Core source tracing nodes and coupled fault propagation paths;

[0199] S42. Generate differentiated handling recommendations based on the diagnostic results of different faults, including:

[0200] If any level has a minor fault and other levels are fault-free, a routine operation and maintenance scheduling suggestion of "periodic inspection and verification" is given.

[0201] If any level is a general fault and other levels are not faulty, a "shutdown and maintenance" solution is recommended.

[0202] If the fault is a minor or general fault at more than one level, and the overall system fault is a general or minor fault, then it should be handled as a cross-level coupled fault.

[0203] If any level is a critical fault, it shall be handled as a cascading fault.

[0204] For cross-level coupling faults, the following handling suggestion is given for all levels in the cross-level coupling fault: "First isolate the fault source, cut off the valves or power supply circuits associated with the fault level, and then repair in steps."

[0205] If the overall system fault diagnosis result is a serious fault and there is a cross-level coupling fault, an emergency operation suggestion of "reducing the load to below 1 / 3 of the rated power and switching to the standby system" is given, along with a maintenance plan that includes the location and replacement list of the faulty components.

[0206] This invention also provides a hierarchical multidimensional fault diagnosis system for methanol-fueled marine propulsion systems, and a hierarchical multidimensional fault diagnosis method for methanol-fueled marine propulsion systems, such as... Figure 2 As shown, it includes: a sensor module, a central processing module, and a display module;

[0207] The sensor modules are distributed across the methanol supply layer, the chemical energy-to-electric energy conversion layer, the electric motor propulsion layer, and the ship's overall energy efficiency layer to collect real-time operating data at each level.

[0208] The central processing module is connected to the sensor module and is used to carry out a hierarchical coupling diagnostic model to perform local fault diagnosis and cross-level correlation analysis at each level to obtain diagnostic results.

[0209] The display module is connected to the central processing module and is used to receive and display diagnostic results.

[0210] Example:

[0211] This embodiment is based on a 5000-ton methanol-fueled electric propulsion cargo ship, and establishes a layered multi-dimensional fault diagnosis and monitoring system, which consists of a sensor module, a data transmission module, a central processing unit, and a display terminal. The configuration of each module is as follows to ensure accurate acquisition and quantitative calculation of real-time monitoring data:

[0212] Sensor module:

[0213] Methanol supply layer: Pressure sensors (PT124G-111) with an accuracy of ±0.5%FS and flow sensors (LWGY-50) with a range of 0-200L / min are deployed at the storage tank outlet, supply pump inlet and outlet, and key nodes of the delivery pipeline. Temperature sensor (PT100) and detection accuracy Methanol leak sensor (ME4-Methanol);

[0214] Chemical energy to electrical energy conversion layer: Temperature sensor with range of 0-2000℃ and pressure sensor with range of 0-2MPa are deployed in the combustion chamber; current sensor (ACS758) with range of 0-2000A and voltage sensor (LV25-P) with range of 0-1000V are deployed in the generator stator and rotor; and gas composition sensor (GAS-300) capable of detecting CO, NOx and unburned methanol is deployed at the exhaust outlet.

[0215] Motor propulsion layer: Current sensors, Hall speed sensors (0-3000 r / min), and torque sensors (0-500 N·m) are deployed on the stator and rotor of the propulsion motor, and voltage sensors and frequency sensors are deployed at the output of the frequency converter.

[0216] Ship-wide energy efficiency layer: Deploy power sensors (PZ72L-E4) with a range of 0-1000kW and energy consumption metering sensors on the main switchboard and auxiliary system power distribution nodes;

[0217] All sensor sampling frequencies are uniformly set to 100Hz to ensure data synchronization.

[0218] Data transmission module: It adopts a transmission scheme that integrates industrial Ethernet (PROFINET protocol) and CAN bus (CANopen protocol). Industrial Ethernet is responsible for the transmission of large amounts of monitoring data, while CAN bus is responsible for the transmission of control commands. The transmission rate is ≥100Mbps and the data transmission delay is ≤10ms, ensuring the real-time performance and integrity of the monitoring data.

[0219] Central Processing Module: Employs an embedded industrial computer, configured with an Intel Core i7 processor, 16GB of memory, and 1TB of storage capacity. It is pre-installed with the quantitative hierarchical coupling diagnostic model algorithm described in this invention, and solidifies the basic correlation coefficients between hierarchical levels. , , It can automatically complete operations such as data standardization, vector construction, diagnostic function calculation, correlation coefficient derivation, and overall fault numerical solution.

[0220] Display module: Employs an industrial-grade touchscreen to display real-time measured values, standardized values, local fault values, overall system fault value D, and correlation coefficients for each level of monitoring parameters. It integrates monitoring indices and fault diagnosis results, supporting manual intervention by maintenance personnel.

[0221] The specific execution steps are as follows:

[0222] 1. Parameter initialization and threshold setting:

[0223] Before the ship sets sail, the central processing module initializes the hierarchical coupled diagnostic model, inputting ship design parameters (methanol storage tank capacity, rated flow rate of the liquid supply pump, rated power of the methanol generator set, rated speed of the propulsion motor, etc.); it also stores the safety thresholds, fault weights, and diagnostic model parameter vectors of each level of monitoring parameters, and stores the basic correlation coefficients between adjacent levels. and This ensures the uniqueness and repeatability of the model calculations.

[0224] 2. Real-time data acquisition and preprocessing:

[0225] During ship navigation, each sensor collects measured values ​​of corresponding operational parameters at a frequency of 100Hz and transmits the raw data to the central processing module. The central processing module preprocesses the raw data, specifically through the following steps:

[0226] Kalman filtering for noise reduction: Using filtering parameters of process noise covariance Q=1e-5 and observation noise covariance R=1e-3, noise from the sensor and transmission process is removed to obtain clean measured values.

[0227] Will Transform the data into dimensionless standardized values ​​and construct real-time monitoring data vectors for each level;

[0228] Criterion-based anomaly removal: For each parameter's sample value, remove extreme values ​​that deviate from the mean by more than 3 times the standard deviation, and replace them with effective values ​​from adjacent time points to ensure the quality of the vector data input to the diagnostic model.

[0229] 3. Numerical calculation for layered fault diagnosis:

[0230] The central processing unit calls diagnostic functions at each level, based on the preprocessed data. With cured Automatically calculates local fault diagnosis values ​​at each level. And the overall system fault diagnosis value D:

[0231] Methanol supply layer: Input 4-dimensional vector Calculate the deviation of each parameter. Substituting into the calculation formula, we get ;

[0232] Chemical-to-electrical energy conversion layer: Input 6-dimensional vector If the unburned methanol content in the exhaust gas Then Set it to 1, calculate the remaining deviations, and substitute them into the calculation formula to obtain... ;

[0233] Motor propulsion layer: Input 5D vector If the stator current harmonic distortion rate Or the inverter voltage / frequency fluctuation exceeds Then set the corresponding deviation to 1, and substitute it into the calculation formula to get ;

[0234] Ship-wide energy efficiency layer: Input 3D vector If energy utilization efficiency Then Set it to 1, calculate the remaining deviations, and substitute them into the calculation formula to obtain... ;

[0235] Overall system fault calculation: , , and Substituting the fault weights into the overall formula, we calculate D, and determine the overall fault level of the system based on the range of D values.

[0236] 4. Collaborative analysis and fault tracing based on simplified correlation coefficient formula:

[0237] The central processing module calls the fixed basic correlation coefficients and automatically derives the correlation coefficients of the entire hierarchy through the inter-level coupling correlation model, combining the correlation coefficients of each level. The specific process involves quantitative collaborative analysis using numerical values ​​and fault timestamps:

[0238] Automatically calculate the correlation coefficient across all levels:

[0239]

[0240] Extract each level The fault occurrence timestamp, in this embodiment is ;

[0241] Coupling Fault Detection And the time difference between failures is Determined as Cross-level coupling faults;

[0242] Tracing the source of cascading faults: This is the earliest level at which the fault occurs, the energy flow initiation level, and its maximum deviation is the liquid supply flow rate deviation. By combining the energy flow topology model, the energy flow anomaly was traced back to the L1 liquid supply pump, and the root cause of the fault was finally located as the wear of the liquid supply pump impeller.

[0243] 5. Results Output and Processing

[0244] The central processing unit will quantify the diagnostic results (at each level). ,overall Deviation of core fault parameters and correlation coefficients across all levels (fault level), (fault root cause) (Impeller wear of the liquid supply pump), fault propagation path () Output to the display terminal and generate standardized processing recommendations:

[0245] 1. Immediately reduce the ship's speed from 12 knots to 8 knots (reduce the load to 60% of the rated power);

[0246] 2. Press the start button for the standby liquid supply pump on the control panel (model: CQB-F) to confirm that the standby pump is operating normally (pressure 0.4MPa, flow rate 65L / min, corresponding to...). , (no deviation)

[0247] 3. Close the inlet and outlet valves of the faulty liquid supply pump (model: Q41F-16C).

[0248] 4. Arrange for maintenance personnel to carry wrenches, gaskets, and other tools to replace and repair the impeller of the faulty liquid supply pump after the ship docks;

[0249] Maintenance personnel performed corresponding operations based on the handling suggestions. During the operation, the central processing module continuously calculated... and Until the value recovers to , The fault was determined to be resolved.

[0250] Implementation effect verification

[0251] A 12,000-hour real-ship test was conducted on a 5,000-ton methanol-fueled electric propulsion cargo ship, covering all navigation conditions including no-load, full-load, constant speed, acceleration, and deceleration. The test results show that:

[0252] Local fault diagnosis: It can accurately identify local faults such as methanol supply layer pipeline leaks, incomplete combustion in the conversion layer, and frequency converter failures in the propulsion layer, with an accuracy rate of 98.5% and numerical calculation errors. ;

[0253] Coupled Fault Identification: Based on a simplified correlation coefficient formula, the judgment logic can effectively identify cross-level coupled faults, with reduced fault identification delay. The accuracy rate of the judgment reached 99%;

[0254] Cascaded fault tracing: Combining correlation coefficients and fault timing analysis, the accuracy of fault tracing reaches 96.2%, and the efficiency of fault handling is improved by 52% compared with traditional methods;

[0255] System adaptability: Under different navigation conditions, the quantitative calculation and correlation coefficient determination logic is stable, the fault diagnosis coverage is 63% higher than that of traditional methods, and there are no missed diagnoses or misdiagnoses.

[0256] Test results demonstrate that this invention significantly improves the accuracy, objectivity, and operability of fault diagnosis for methanol-fueled electric propulsion ship power systems by quantitatively defining diagnostic parameters and designing a simplified hierarchical correlation coefficient formula. It adapts to the characteristics of methanol fuel and the navigation requirements of ships, effectively ensuring the safe and stable operation of ship power systems.

[0257] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A hierarchical multidimensional fault diagnosis method for methanol-fueled marine propulsion systems, characterized in that, include: S1: Introducing a methanol-fueled electric propulsion ship power system, the methanol-fueled electric propulsion ship power system is divided into four mutually coupled levels according to the energy conversion and transmission path of "methanol storage and transportation - chemical energy conversion - electric drive - whole ship energy optimization". The levels are methanol supply layer, chemical energy-electric energy conversion layer, electric propulsion layer and whole ship energy efficiency layer. The mutual coupling between the levels refers to the coupling relationship being characterized by the correlation coefficient between the levels. S2: Construct a hierarchical coupling diagnostic model to perform local fault diagnosis and cross-level correlation analysis for each level. The hierarchical coupling diagnostic model includes local fault diagnosis functions, overall system fault diagnosis functions, collaborative monitoring formulas, monitoring-diagnosis iterative relationship formulas, and inter-level coupling correlation models based on the coupling relationship between each level. S3: Collect real-time operating data at each level, preprocess the real-time operating data, and perform fault diagnosis and cross-level correlation analysis on the preprocessed real-time operating data at each level through a hierarchical coupling diagnostic model to obtain diagnostic results; S4: Output diagnostic results and generate differentiated treatment suggestions based on the diagnostic results of different faults.

2. The hierarchical multidimensional fault diagnosis method for methanol-fueled marine propulsion systems according to claim 1, characterized in that, The local fault diagnosis functions constructed for each of the above levels include: The local fault diagnosis function is constructed as follows: in, These correspond to the methanol supply layer, the chemical energy-electric energy conversion layer, the electric motor propulsion layer, and the overall ship energy efficiency layer, respectively. Indicates the scaling factor; This is a dimensionless result of a local fault diagnosis. , indicating the first Hierarchical real-time monitoring data vectors, Indicates the first Level 1 Standardized values ​​of real-time monitoring data; Indicates the first Hierarchical diagnostic function parameter vector, Indicates the first The first of the preset hierarchical levels The deviation weighting coefficient of each real-time monitoring data is dimensionless. Indicates the first Level 1 The deviation of real-time monitoring data; This indicates the total number of real-time monitoring data at the current level; and As shown below: in, Indicates the first The first level The measured value of a real-time monitoring data point. Indicates the first Level 1 The minimum security threshold for real-time monitoring data. Indicates the first Level 1 The maximum security threshold for each real-time monitoring data point.

3. The hierarchical multidimensional fault diagnosis method for methanol-fueled marine propulsion systems according to claim 2, characterized in that, Construct a system-wide fault diagnosis function, including: The overall system fault diagnosis function is constructed based on the local fault diagnosis function as follows: in, The initial overall system fault diagnosis results are dimensionless. Indicates the first Hierarchical fault weights, dimensionless. .

4. The hierarchical multidimensional fault diagnosis method for methanol-fueled marine propulsion systems according to claim 3, characterized in that, A collaborative monitoring formula and a monitoring-diagnosis iterative relationship formula are constructed to update the initial overall system fault diagnosis results; The collaborative monitoring formula is: in, The initial collaborative monitoring comprehensive index is dimensionless. For the first The hierarchical average deviation is shown below: in, for The number of dimensions; Based on the collaborative monitoring formula and the initial overall system fault diagnosis results, the updated collaborative monitoring comprehensive index is constructed as follows: in, This is the feedback coefficient of the fault diagnosis result to the monitoring status; The initial overall system fault diagnosis results were updated based on the updated collaborative monitoring comprehensive index, resulting in the final overall fault diagnosis results: in, The influence coefficient of the monitoring status on the diagnostic results; .

5. The hierarchical multidimensional fault diagnosis method for methanol-fueled marine propulsion systems according to claim 4, characterized in that, Based on the coupling relationships between different levels, a hierarchical coupling and association model is constructed, including: The inter-level coupling and association model is as follows: in, For any two levels and The correlation coefficient is dimensionless. ; Adjacent levels and Preset basic correlation coefficients.

6. The hierarchical multidimensional fault diagnosis method for methanol-fueled marine propulsion systems according to claim 5, characterized in that, Real-time operational data is collected from each level, preprocessed, and then collaboratively analyzed using a hierarchical coupled diagnostic model to obtain diagnostic results, including: S31. Collect real-time operating data at each level, convert the real-time operating data into standardized values, and substitute them into the local fault diagnosis function to solve. And obtain the detection parameters corresponding to the maximum deviation in each level as the core fault parameters; S32, according to The initial overall system fault diagnosis result is obtained by solving the initial overall system fault diagnosis result, and then by applying the collaborative monitoring formula and the monitoring-diagnosis iterative relationship formula to the initial overall system fault diagnosis result. ; S33, Define the first fault value range Second fault zone Judge separately and The numerical range within which the fault falls determines the corresponding level and the fault severity of the system. like and The value is located at If the fault is internal, then the corresponding level and system are considered to have a minor malfunction. like and The value is located at Within this range, the corresponding level and system are considered general faults. like and The value is greater than If so, the corresponding level and system are considered to be in serious condition. S34, based on Based on the time difference of failure, design identification rules for cross-level coupled faults, including: When a certain level of Other levels of ,and ,and and If the time difference between the occurrence of the fault and the set time is less than or equal to the set time, then a fault is determined to have occurred. and Cross-level coupling faults occur between them; S35. Tracing the source of cascaded faults that are coupled across different levels, the specific steps are as follows: Extract all hierarchical groups identified as coupling faults and confirm the chronological order of fault propagation; Identify the core traceability nodes: The earliest level where the fault occurred is set as the core tracing node. If the time is the same or the difference is less than the set value, the level with the largest correlation coefficient is set as the core tracing node. If the correlation coefficient is the same or the difference is less than the set value, the level with the largest local fault diagnosis value is set as the core tracing node. Obtain the detection parameters corresponding to the maximum deviation of the core traceability node, and locate the core faulty component corresponding to the detection parameters.

7. The hierarchical multidimensional fault diagnosis method for methanol-fueled marine propulsion systems according to claim 1, characterized in that, Output diagnostic results and generate differentiated treatment suggestions based on the diagnostic results of different faults, including: S41. Output diagnostic results: Output the local fault values ​​at each level. and core fault parameters, core faulty components, and overall system fault values. Core source tracing nodes and coupled fault propagation paths; S42. Generate differentiated handling recommendations based on the diagnostic results of different faults, including: If any level has a minor fault and other levels are fault-free, a routine operation and maintenance scheduling suggestion of "periodic inspection and verification" is given. If any level is a general fault and other levels are not faulty, a "shutdown and maintenance" solution is recommended. If the fault is a minor or general fault at more than one level, and the overall system fault is a general or minor fault, then it should be handled as a cross-level coupled fault. If any level is a critical fault, it shall be handled as a cascading fault. For cross-level coupling faults, the following handling suggestion is given for all levels in the cross-level coupling fault: "First isolate the fault source, cut off the valves or power supply circuits associated with the fault level, and then repair in steps." If the overall system fault diagnosis result is a serious fault and there is a cross-level coupled fault, an emergency operation suggestion of "reducing the load to below 1 / 3 of the rated power and switching to the standby system" is given, along with a maintenance plan that includes the location and replacement list of faulty components.

8. A hierarchical multidimensional fault diagnosis system for a methanol-fueled marine propulsion system, employing the hierarchical multidimensional fault diagnosis method for a methanol-fueled marine propulsion system as described in claim 1, characterized in that, include: Sensor module, central processing module, and display module; The sensor modules are distributed across the methanol supply layer, the chemical energy-to-electric energy conversion layer, the electric motor propulsion layer, and the ship's overall energy efficiency layer to collect real-time operating data at each level. The central processing module is connected to the sensor module and is used to carry out a hierarchical coupling diagnostic model to perform local fault diagnosis and cross-level correlation analysis at each level to obtain diagnostic results. The display module is connected to the central processing module and is used to receive and display diagnostic results.

Images