Operation protection method and system for ship electric propulsion system

Abstract

The invention discloses a ship electric propulsion system operation protection method and system, and the method comprises the steps: generating a time sequence signal set through a real-time operation data set of a ship electric propulsion system, constructing a plurality of fault monitoring models through an HSMM algorithm, processing different time sequence signals, and enhancing the monitoring capability of the system for multi-dimensional faults; the fuzzy processing is carried out on the system operation state through the membership degree function, effective fuzzy evaluation on the operation state of the ship electric propulsion system can be realized, reasonable judgment and decision can still be made in the face of partial inaccurate or incomplete data, and compared with traditional simple threshold value judgment, the method has the advantages that the judgment accuracy is improved, and the judgment efficiency is improved. The fuzzy rule can make more reasonable judgment in the state transition stage, misjudgment caused by boundary fuzziness is reduced, the fault monitoring accuracy is improved, the most appropriate fault protection scheme can be further matched according to the fault evaluation result, and it is ensured that the ship electric propulsion system is pertinently and effectively protected.

Description

Technical Field

[0001] The present invention relates to the technical field of state monitoring, and in particular to an operation protection method and system for a ship electric propulsion system. Background Art

[0002] A ship electric propulsion system is a ship propulsion system that uses an electric motor to drive the ship's propeller or other propulsion device instead of a traditional internal combustion engine or steam turbine. In a ship electric propulsion system, the main engine or generator generates electrical energy, which is then transmitted to the motor through the power transmission system, and the motor then drives the propeller or other propulsion device. With its high efficiency, environmental protection, and energy saving, the ship electric propulsion system has become the mainstream of ship power devices. However, due to its structural complexity and harsh working environment (such as high humidity, high temperature, vibration, etc.), the ship electric propulsion system frequently fails, seriously affecting the normal operation of the ship and the life safety of the crew. Therefore, timely and accurate fault monitoring and diagnosis of the ship electric propulsion system is crucial to ensure the safe operation of the ship and improve the work efficiency of the crew.

[0003] The existing ship electric propulsion system status monitoring detects the faults of the electric propulsion system through real-time parameters. However, due to the structural complexity of the ship electric propulsion system and the large amount of operating data, the fault detection of the electric propulsion system only by setting thresholds is prone to singleness and low accuracy, and it is impossible to comprehensively evaluate the status of the ship electric propulsion system. Summary of the invention The present invention provides a method and system for protecting the operation of a ship electric propulsion system, so as to improve the comprehensiveness of fault prediction of the ship electric propulsion system and improve the operation stability of the ship electric propulsion system.

[0004] In order to solve the above technical problems, an embodiment of the present invention provides a method for protecting the operation of a ship electric propulsion system, comprising: Acquire a real-time operation data set of a ship electric propulsion system, preprocess the real-time operation data set, and acquire a time series signal set; Constructing a plurality of first fault monitoring models based on the time series signal set and the HSMM algorithm, and optimizing each of the first fault monitoring models based on the ant colony algorithm to obtain a plurality of second fault monitoring models; Acquire a hidden state sequence of the second fault monitoring model, generate a corresponding initial membership function based on each hidden state sequence, and perform fuzzification processing on the initial membership function to acquire a target membership function; Determining the operating state of the ship electric propulsion system based on the target membership function and the preset fuzzy rules; When the operating state is abnormal, performing defuzzification processing on the operating state to obtain an operating evaluation result of the electric propulsion system; A fault protection scheme is matched based on the operation evaluation result, and an operation parameter of the ship electric propulsion system is adjusted based on the fault protection scheme to perform operation protection on the ship electric propulsion system.

[0005] The present invention generates a time series signal set through the real-time operation data set of the ship's electric propulsion system, and then constructs multiple fault monitoring models through the HSMM (Hidden Semi-Markov Model) algorithm, which can process different time series signals respectively and enhance the system's monitoring ability for multi-dimensional faults. At the same time, the optimization of the ant colony algorithm can automatically adjust the parameters of the model to ensure that the model can maintain high accuracy in a dynamic environment and avoid fault monitoring failure caused by overfitting or underfitting of the model; in addition, the system operation state is fuzzy processed by the membership function, which can realize effective fuzzy evaluation of the operation state of the ship's electric propulsion system, so that when facing partially inaccurate or incomplete data, reasonable judgments and decisions can still be made. Compared with traditional simple threshold judgments, fuzzy rules can make more reasonable judgments in the state transition stage, reduce misjudgments caused by boundary fuzziness, improve fault monitoring accuracy, and further match the most suitable fault protection scheme according to the fault evaluation results to ensure that the ship's electric propulsion system is targeted and effectively protected.

[0006] Furthermore, the step of obtaining a real-time operation data set of the ship electric propulsion system and preprocessing the real-time operation data set includes: Acquire a real-time operating data set of a ship electric propulsion system, and perform filtering and normalization processing on the real-time operating data set; the real-time operating data set includes current data, voltage data, speed data, vibration data and temperature data; A window length is set based on the operation mode of the ship electric propulsion system, and data in the real-time operation data set is divided based on the window length to obtain a time series signal set.

[0007] After filtering, the present invention removes high-frequency noise or interference signals, ensures the smoothness and accuracy of the data, and normalizes the data to a unified range, making different data sources comparable, avoiding the impact of dimensional differences in different data on model processing, and setting window lengths according to different operating modes to optimize data division and make fault monitoring more accurate. Furthermore, through time window division, real-time data can be converted into independent time series signals, so that each signal can be analyzed independently, facilitating subsequent model processing.

[0008] Furthermore, the constructing of a plurality of first fault monitoring models based on the time series signal set and the HSMM algorithm includes: Based on the operating state of the ship electric propulsion system, respectively set a hidden state and a corresponding hidden state duration distribution for the time series signals in the time series signal set; A first fault detection model is constructed based on the HSMM algorithm and the time series signal, wherein the first fault detection model includes a hidden state duration distribution, a hidden state transfer matrix, and an observation probability distribution. The present invention can obtain a refined fault monitoring model by combining the HSMM algorithm with the modeling of the time series signal set. By combining the hidden semi-Markov model (HSMM) with time series data, hidden state duration, transfer matrix and observation probability distribution, it can comprehensively describe the operation and failure mode of the ship's electric propulsion system, thereby improving the reliability of fault detection, improving the adaptability and generalization ability of the model, and enabling it to perform effective fault diagnosis and prediction under different operating environments and conditions.

[0009] Furthermore, the optimization of each first fault monitoring model based on the ant colony algorithm to obtain a plurality of second fault monitoring models includes: Initializing the ant colony size and search space based on the ant colony algorithm, and initializing the parameters of the first fault monitoring model; Iteratively performing ant colony search based on the ant colony size and search space to update parameters of the first fault monitoring model; In each iteration, the fitness of the updated first fault detection model is evaluated based on the fitness function until the fitness or the number of iterations reaches a preset condition, the iteration is stopped, and the second fault detection model is output.

[0010] The present invention can adjust the search strategy according to the fitness of each ant through the ant colony algorithm to achieve dynamic adaptive optimization. At the same time, the ant colony algorithm has good global search capabilities and can stably and effectively optimize the parameters of the fault monitoring model when facing multi-modal or complex objective functions.

[0011] Further, the step of obtaining the hidden state sequence of the second fault monitoring model, generating a corresponding initial membership function based on each hidden state sequence, and performing fuzzification processing on the initial membership function to obtain a target membership function includes: Based on each second fault monitoring model, respectively obtain the hidden state sequence at each time point, and record the probability distribution corresponding to each hidden state; Calculating the probability value of the hidden state sequence based on the probability distribution corresponding to each hidden state, and generating the initial membership function corresponding to each second fault monitoring model based on the probability value; The initial membership function is mapped to a fuzzy set based on a fuzzification function to obtain a target membership function.

[0012] By converting the probability distribution of the hidden state into a membership function, the system can capture the slight changes in the operating state more carefully, thereby giving early warning of potential faults. Moreover, through fuzzy processing and the generation of the target membership function, the system can still perform effective reasoning and decision-making in the face of incomplete information, thereby improving the fault monitoring accuracy of the ship's electric propulsion system.

[0013] Furthermore, when the operating state is abnormal, the operating state is defuzzified to obtain an operating evaluation result of the ship electric propulsion system, including: When the operating state is abnormal, a weighted average value of the target membership function is calculated based on a centroid method to obtain an operating evaluation result, wherein the operating evaluation result includes a system state and a fault type.

[0014] The present invention can extract a clear system state and fault type from the fuzzy target membership function through defuzzification processing, and performs defuzzification processing through the centroid method, thereby improving the accuracy of fault detection and enhancing the system's response capability to abnormal situations.

[0015] Further, matching a fault protection scheme based on the operation evaluation result, and adjusting the operation parameters of the ship electric propulsion system based on the fault protection scheme to perform operation protection on the ship electric propulsion system, includes: Matching a fault protection scheme in a preset strategy library based on the fault type and the system state; A control strategy is acquired based on the fault protection scheme, and an operating parameter of the ship electric propulsion system is adjusted based on the control strategy to perform operation protection on the ship electric propulsion system.

[0016] The present invention combines the system evaluation results to automatically match and execute the fault protection plan. The ship's electric propulsion system can respond quickly when a fault occurs and take appropriate control measures. This not only ensures the safety of the ship, but also improves the intelligence and adaptability of the system, providing a strong guarantee for the long-term stable operation of the ship's electric propulsion system.

[0017] In a second aspect, the present invention provides a ship electric propulsion system operation protection system, including: a data acquisition module, a monitoring model building module, a fuzzy module, an evaluation module, a defuzzification module and a maintenance module; The data acquisition module is used to obtain a real-time operation data set of the ship electric propulsion system, pre-process the real-time operation data set, and obtain a time series signal set; The monitoring model building module is used to build a plurality of first fault monitoring models based on the time series signal set and the HSMM algorithm, and optimize each first fault monitoring model based on the ant colony algorithm to obtain a plurality of second fault monitoring models; The fuzzy module is used to obtain the hidden state sequence of the second fault monitoring model, generate a corresponding initial membership function based on each hidden state sequence, and perform fuzzification on the initial membership function to obtain a target membership function; The evaluation module is used to determine the operating state of the ship electric propulsion system based on the target membership function and the preset fuzzy rules; The defuzzification module is used to perform defuzzification processing on the operating state when there is an abnormality in the operating state, so as to obtain an operating evaluation result of the electric propulsion system; The maintenance module is used to match the fault protection scheme based on the operation evaluation result, and adjust the operation parameters of the ship electric propulsion system based on the fault protection scheme to perform operation protection on the ship electric propulsion system.

[0018] Furthermore, the monitoring model construction module is used to: Based on the operating state of the ship electric propulsion system, respectively set a hidden state and a corresponding hidden state duration distribution for the time series signals in the time series signal set; A first fault detection model is constructed based on the HSMM algorithm and the time series signal, wherein the first fault detection model includes a hidden state duration distribution, a hidden state transfer matrix, and an observation probability distribution. Furthermore, the monitoring model building module is also used to: Initializing the ant colony size and search space based on the ant colony algorithm, and initializing the parameters of the first fault monitoring model; Iteratively performing ant colony search based on the ant colony size and search space to update parameters of the first fault monitoring model; In each iteration, the fitness of the updated first fault detection model is evaluated based on the fitness function until the fitness or the number of iterations reaches a preset condition, the iteration is stopped, and the second fault detection model is output. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A schematic diagram of a flow chart of a method for protecting an operation of a ship electric propulsion system provided in an embodiment of the present invention; Figure 2 A structural schematic diagram of a ship electric propulsion system operation protection system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0021] The terms "first" and "second" and the like in the specification, claims and drawings of this application are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device comprising a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products or devices.

[0022] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0023] Example 1 See also Figure 1 , Figure 1 The flowchart of a ship electric propulsion system operation protection method provided in an embodiment of the present invention is as follows. The embodiment of the present invention provides a ship electric propulsion system operation protection method, including steps 101 to 106, which are as follows: Step 101: obtaining a real-time operation data set of a ship electric propulsion system, preprocessing the real-time operation data set, and obtaining a time series signal set; In this embodiment, the step of obtaining a real-time operating data set of the ship electric propulsion system and preprocessing the real-time operating data set includes: Acquire a real-time operating data set of a ship electric propulsion system, and perform filtering and normalization processing on the real-time operating data set; the real-time operating data set includes current data, voltage data, speed data, vibration data and temperature data; A window length is set based on the operation mode of the ship electric propulsion system, and data in the real-time operation data set is divided based on the window length to obtain a time series signal set.

[0024] In this embodiment, the real-time operation data set includes sensor data from various monitoring points of the ship's electric propulsion system, such as current (A), voltage (V), speed (rpm), vibration (mm / s), temperature (°C), etc. These data are usually collected in the form of time series, and the data may be biased due to noise or sensor errors.

[0025] In this embodiment, filtering and normalization are performed on various types of data in the real-time running data set.

[0026] In this embodiment, a low-pass filter or a median filter is used to filter out high-frequency noise to ensure that the signal is more consistent with the actual physical characteristics. After filtering, high-frequency noise or interference signals are removed to ensure the smoothness and accuracy of the data.

[0027] In this embodiment, the maximum and minimum values ​​of each sensor data are normalized to between 0 and 1 (for example, using maximum value normalization) to ensure that all signals are within the same scale range. The normalized data eliminates the dimensional differences between different types of data, making subsequent analysis and modeling more consistent and stable.

[0028] In this embodiment, the operation modes of the ship electric propulsion system include: normal operation, low-load operation, high-load operation, etc. Different operation modes have different effects on various performance indicators of the system (such as current, voltage, etc.). Therefore, according to the actual operation mode and application requirements of the ship electric propulsion system, a suitable window length (for example, 10 seconds, 1 minute, etc.) is set to ensure that each time window contains sufficient historical data to identify potential fault modes.

[0029] In another optional embodiment, the system response speed can be determined according to the operation mode. If the system response is relatively stable and the signal changes slowly, a longer window length (such as 10 seconds, 30 seconds or even 1 minute) can be selected. If the system dynamic response is fast (such as starting, stopping, etc.), a shorter time window should be selected to capture the rapidly changing signal characteristics (such as 1 second, 3 seconds).

[0030] In another optional embodiment, the sampling frequency of the signal can be obtained according to the operation mode, and the length of the window can be determined according to the sampling frequency of the signal. For example, if the sampling frequency of the signal is high (collecting data multiple times per second), a shorter time window can be selected to capture detailed information. If the sampling frequency is low, the window length can be appropriately extended.

[0031] In this embodiment, the pre-processed real-time data is divided into a plurality of small windows in chronological order, and the data in the windows are input as time series signals for subsequent fault monitoring and analysis.

[0032] In the embodiment of this city, high-frequency noise or interference signals are removed through filtering processing, ensuring the smoothness and accuracy of the data. At the same time, the data is normalized to a unified range to make different data sources comparable, avoiding the impact of dimensional differences in different data on model processing. By setting the window length according to different operating modes, data division can be optimized to make fault monitoring more accurate. Furthermore, through time window division, real-time data can be converted into independent time series signals, so that each signal can be analyzed independently, which is convenient for subsequent model processing.

[0033] Step 102: constructing a plurality of first fault monitoring models based on the time series signal set and the HSMM algorithm, and optimizing each of the first fault monitoring models based on the ant colony algorithm to obtain a plurality of second fault monitoring models; In this embodiment, the construction of several first fault monitoring models based on the time series signal set and the HSMM algorithm includes: Based on the operating state of the ship electric propulsion system, respectively set a hidden state and a corresponding hidden state duration distribution for the time series signals in the time series signal set; A first fault detection model is constructed based on the HSMM algorithm and the time series signal, wherein the first fault detection model includes a hidden state duration distribution, a hidden state transfer matrix, and an observation probability distribution. In this embodiment, since the time series signal set includes time series signals of various types of data, a fault monitoring model needs to be constructed for each different time series signal.

[0034] In this embodiment, for each different time series signal, a hidden state is defined according to the operation mode of the ship electric propulsion system.

[0035] In this embodiment, the hidden state includes a normal state, a fault warning state and a fault state, wherein the normal state indicates that the system runs smoothly and there is no fault; the fault warning state indicates that the system may fail and there are certain abnormalities; the fault state indicates that the system fails and requires intervention.

[0036] In this embodiment, the characteristics of the signal may be different in each hidden state, so it needs to be set based on the actual operating state.

[0037] In this embodiment, one of the core features of HSMM is the ability to model the duration of hidden states. Each hidden state has not only a transition probability (the probability of going from one hidden state to another), but also a duration distribution, that is, the length of time the system stays in a hidden state. For each hidden state, its duration distribution can be estimated based on historical data (such as signal fluctuations before and after a fault occurs). Statistical methods such as Gamma distribution and exponential distribution are usually used to fit the duration of hidden states.

[0038] As a specific example of an embodiment of the present invention, assuming that in the "fault warning" state, the ship electric propulsion system may be in this state for a period of time (for example, several minutes or hours), while in the "normal" state, the duration may be longer. Based on historical data (such as fluctuations in sensor data and changes in operating modes), the duration distribution under different hidden states can be calculated.

[0039] In this embodiment, the first fault monitoring model structure includes a hidden state duration distribution, a hidden state transfer matrix, and an observation probability distribution.

[0040] In this embodiment, the hidden state transfer matrix represents the probability of transition between different hidden states. For example, there may be a lower transition probability from "normal state" to "fault warning state", while there may be a higher transition probability from "fault warning state" to "fault state".

[0041] In this embodiment, the transfer matrix is ​​a matrix of size N*N, where N is the number of hidden states, and each element P(i,j) represents the probability of transitioning from hidden state 𝑖 to hidden state 𝑗.

[0042] In this embodiment, the observation probability distribution represents the probability distribution of the observation data (i.e., time series signal) generated by the system in each hidden state. The generation process of the observation data can be modeled using Gaussian distribution, multinomial distribution, etc. For example, in the "fault state", the observed value of the current signal may be high, while in the "normal state", the fluctuation of the current signal is small.

[0043] In this embodiment, according to the actual situation of the ship electric propulsion system, an initial probability distribution is set for each hidden state (such as "normal", "fault warning", "fault"), so as to construct a first fault monitoring model. The first fault monitoring model is trained by historical time series data (including current, voltage, temperature and other signals) to estimate the hidden state transfer matrix, hidden state duration distribution and observation probability distribution using the expectation maximization (EM) algorithm.

[0044] In this embodiment, the use of the hidden semi-Markov model (HSMM) can take into account the duration characteristics of the hidden state, and can more accurately describe the performance of the ship's electric propulsion system in different states, especially when there are hidden faults or transition states, and can identify potential faults in advance to avoid system operation failures.

[0045] The present invention can obtain a refined fault monitoring model by combining the HSMM algorithm with the modeling of the time series signal set. By combining the hidden semi-Markov model (HSMM) with time series data, hidden state duration, transfer matrix and observation probability distribution, it can comprehensively describe the operation and failure mode of the ship's electric propulsion system, thereby improving the reliability of fault detection, improving the adaptability and generalization ability of the model, and enabling it to perform effective fault diagnosis and prediction under different operating environments and conditions.

[0046] In this embodiment, the optimization of each first fault monitoring model based on the ant colony algorithm to obtain a plurality of second fault monitoring models includes: Initializing the ant colony size and search space based on the ant colony algorithm, and initializing the parameters of the first fault monitoring model; Iteratively performing ant colony search based on the ant colony size and search space to update parameters of the first fault monitoring model; In each iteration, the fitness of the updated first fault detection model is evaluated based on the fitness function until the fitness or the number of iterations reaches a preset condition, the iteration is stopped, and the second fault detection model is output.

[0047] In this embodiment, the parameters of the ant colony algorithm, namely, the ant colony size and the search space, are initialized. Since the basic idea of ​​the ant colony algorithm is to simulate the foraging process of ants in nature, the number of ants and the exploration process will affect the optimization effect. In this process, the size of the ant colony, namely, the number of ants participating in the optimization, needs to be set first. The number of ants usually selected ranges from dozens to hundreds.

[0048] In this embodiment, the search space is a parameter space that the ant colony algorithm needs to search.

[0049] In this embodiment, the search space mainly includes the following parameters: hidden state transfer matrix, hidden state duration distribution and observation probability distribution. The setting of the search space determines the parameter range that the ant colony needs to explore in each iteration.

[0050] In this embodiment, for each ant, its solution needs to be randomly initialized, that is, an initial value of a hidden state transfer matrix, duration distribution, and observation probability distribution is given. These initial values ​​should be reasonable to ensure that the ant colony can effectively explore the optimization space. In this way, each ant will search according to the quality of the current solution (i.e., fitness) and select a possible path (i.e., update parameters).

[0051] In this embodiment, when updating parameters, the ants will choose the path based on the fitness information of the previous iteration, which is probabilistic. For the solution with high fitness, the ants will be more inclined to choose this path, thereby strengthening the search for this path. In the process of selecting a path, the ants will release pheromones on the path. The concentration of pheromones is proportional to the fitness of the path. A path with high fitness will release more pheromones, attracting more ants to choose this path.

[0052] In this embodiment, after each iteration, the ants update the pheromone concentration of the path according to the fitness function. The pheromone update formula is: (1) in, It is a pheromone volatile factor. is the concentration of pheromone on the path, is the pheromone left by each ant on the path.

[0053] In this embodiment, ants explore in local space and share pheromones globally to promote global search.

[0054] In this embodiment, in each iteration, the ants evaluate the quality of the current solution according to the fitness function.

[0055] In this embodiment, a fitness function is constructed based on model accuracy, fault detection rate, and false alarm rate, wherein model accuracy is the error between the predicted fault state and the actual fault state, such as mean square error (MSE) or other regression error metrics. Fault detection rate is whether the system fault can be accurately detected. False alarm rate is used to measure whether the model will mistakenly report a non-fault state as a fault.

[0056] As a specific example of an embodiment of the present invention, the fitness function is: (2) Among them, F is the fitness, and are weight coefficients, For detection accuracy, is the false alarm rate.

[0057] In this embodiment, the iteration process will continue until any of the following conditions are met: the preset maximum number of iterations is reached. In several iterations, the value of the fitness function does not change significantly, indicating that the optimization has reached a local optimal solution.

[0058] In this embodiment, after multiple iterations, the optimal second fault monitoring model is output, which includes an optimized hidden state transfer matrix, hidden state duration distribution, and observation probability distribution. Compared with the initial first fault monitoring model, this model has significantly improved the accuracy and stability of fault detection and can better adapt to different operating states of ship electric propulsion systems.

[0059] In this embodiment, the ant colony algorithm can adjust the search strategy according to the fitness of each ant to achieve dynamic adaptive optimization. At the same time, the ant colony algorithm has good global search capabilities and can stably and effectively optimize the parameters of the fault monitoring model when facing multimodal or complex objective functions.

[0060] Step 103: Obtain the hidden state sequence of the second fault monitoring model, generate a corresponding initial membership function based on each hidden state sequence, and perform fuzzification processing on the initial membership function to obtain a target membership function; In this embodiment, the step of obtaining the hidden state sequence of the second fault monitoring model, generating a corresponding initial membership function based on each hidden state sequence, and performing fuzzification processing on the initial membership function to obtain a target membership function includes: Based on each second fault monitoring model, respectively obtain the hidden state sequence at each time point, and record the probability distribution corresponding to each hidden state; Calculating the probability value of the hidden state sequence based on the probability distribution corresponding to each hidden state, and generating the initial membership function corresponding to each second fault monitoring model based on the probability value; The initial membership function is mapped to a fuzzy set based on a fuzzification function to obtain a target membership function.

[0061] In this embodiment, the second fault monitoring model is a model optimized by an ant colony algorithm, including information such as a hidden state transfer matrix, hidden state duration distribution, and observation probability distribution. Through this model, the hidden state sequence of the ship's electric propulsion system can be obtained.

[0062] In this embodiment, the system status data corresponding to each time point is acquired through the time series signal set.

[0063] In this embodiment, the optimized second fault monitoring model is used to infer the real-time operation data through the HSMM algorithm to obtain the hidden state at each moment. The hidden state is obtained by decoding the observed time series signal and represents the possible internal state of the system at a certain moment.

[0064] For each time point , calculate the probability that the system is in a certain hidden state, and the result is a hidden state sequence: (3) in, For at the moment The hidden state of the system.

[0065] In this embodiment, for each time point , hidden state There is a corresponding probability , which indicates the probability that the system is in the hidden state at that moment. Based on the probability distribution corresponding to the hidden state, the probability value of the hidden state sequence can be calculated. That is, at each moment, the probability of the hidden state appearing is used to describe the current state uncertainty of the system. For each hidden state , its probability value is used to generate the membership function.

[0066] In this embodiment, the degree to which an event or state belongs to a fuzzy set can be represented by the membership function in fuzzy logic. The membership function generally takes a value range of [0,1], representing the degree of membership of a state.

[0067] In this embodiment, the membership value can be calculated based on the hidden state and its probability distribution, specifically: (4) in, Represents hidden state The normalized membership value of .

[0068] In this embodiment, the hidden state sequence and the corresponding membership value are combined to generate an initial membership function. The initial membership function represents the membership values ​​of different hidden state sequences.

[0069] In this embodiment, the initial membership function is mapped to a fuzzy set, the target membership function is obtained, the initial membership function is fuzzified, and it is mapped to the fuzzy set. The purpose of fuzzification is to convert the original numerical data into a fuzzy set.

[0070] In this embodiment, various fuzzification functions, such as triangle, trapezoid, normal distribution and other functions, can be used to map the membership function of the hidden state to the fuzzy set. For example, the triangle membership function is used to define the membership relationship of different hidden states in the fuzzy set. The membership of each hidden state According to the selected fuzzification function, it will be mapped to a fuzzy set. These fuzzy sets can be classified according to the operating status of the ship's electric propulsion system, such as "normal", "minor fault" and "serious fault".

[0071] As a specific example of an embodiment of the present invention, a triangular membership function or a trapezoidal membership function is used as a fuzzification function, and the initial membership values ​​are allocated to different fuzzy sets according to fuzzy rules: Normal operation state: The closer the initial membership value is to 1, the more stable the system is.

[0072] Minor fault state: The initial membership value is in the middle range (for example, 0.3~0.7), indicating that the system has some deviation but has not reached the fault standard.

[0073] Severe fault state: The initial membership value is close to 0 and the system is in an abnormal state.

[0074] The fuzzy rules are preset. If the initial membership is higher than 0.8, it belongs to the "normal operation state"; if the initial membership is between 0.3 and 0.8, it belongs to the "slight abnormal state"; if the initial membership is lower than 0.3, it belongs to the "serious fault state".

[0075] In this embodiment, the hidden state sequence of each signal , calculate the membership value of the corresponding state , where j represents the state category (normal, abnormal, fault), so that the fuzzy result is obtained in matrix form: Among them, each column corresponds to the fuzzy membership of "normal", "slightly abnormal" and "serious fault".

[0076] In this embodiment, the fuzzy membership matrix of all signals is Perform weighted fusion to obtain the target membership function, specifically: (5) in, For signal The weights reflect the importance of different signals in evaluating the system status. The weights can be set based on historical data or expert experience. It represents the fused target membership function, which is finally used to evaluate the operating status of the system.

[0077] Step 104: determining the operating state of the ship electric propulsion system based on the target membership function and a preset fuzzy rule; In this embodiment, by converting the probability distribution of the hidden state into a membership function, the system can capture subtle changes in the operating state more carefully, thereby providing early warning of potential faults. Moreover, through fuzzy processing and the generation of target membership functions, the system can still perform effective reasoning and decision-making in the face of incomplete information, thereby improving the fault monitoring accuracy of the ship's electric propulsion system.

[0078] Step 105: When the operating state is abnormal, defuzzification is performed on the operating state to obtain an operating evaluation result of the electric propulsion system; In this embodiment, further, when the operating state is abnormal, the operating state is defuzzified to obtain an operating evaluation result of the ship electric propulsion system, including: When the operating state is abnormal, a weighted average value of the target membership function is calculated based on a centroid method to obtain an operating evaluation result, wherein the operating evaluation result includes a system state and a fault type.

[0079] In this embodiment, the centroid method calculates the weighted average value according to the membership of each fuzzy set in the target membership function and the system state value represented by each fuzzy set. Specifically: (6) in, is the state in the fuzzy set The membership function value of (i.e. the membership of the state), For each state value (such as different fault types, system operation modes, etc.), is the centroid value, which represents the system state value obtained after defuzzification.

[0080] In this embodiment, matching a fault protection scheme based on the operation evaluation result, and adjusting the operation parameters of the ship electric propulsion system based on the fault protection scheme to perform operation protection on the ship electric propulsion system, includes: Matching a fault protection scheme in a preset strategy library based on the fault type and the system state; A control strategy is acquired based on the fault protection scheme, and an operating parameter of the ship electric propulsion system is adjusted based on the control strategy to perform operation protection on the ship electric propulsion system.

[0081] In this embodiment, the system status reflects the overall operation of the current ship electric propulsion system, such as "normal", "minor fault", "serious fault", etc. The fault type describes the type of fault that may occur in the system, such as "voltage abnormality", "current overload", "overtemperature", etc.

[0082] The present invention combines the system evaluation results to automatically match and execute the fault protection plan. The ship's electric propulsion system can respond quickly when a fault occurs and take appropriate control measures. This not only ensures the safety of the ship, but also improves the intelligence and adaptability of the system, providing a strong guarantee for the long-term stable operation of the ship's electric propulsion system.

[0083] Step 106: Matching a fault protection scheme based on the operation evaluation result, and adjusting the operation parameters of the ship electric propulsion system based on the fault protection scheme to perform operation protection on the ship electric propulsion system.

[0084] In this embodiment, matching a fault protection scheme based on the operation evaluation result, and adjusting the operation parameters of the ship electric propulsion system based on the fault protection scheme to perform operation protection on the ship electric propulsion system, includes: Matching a fault protection scheme in a preset strategy library based on the fault type and the system state; A control strategy is acquired based on the fault protection scheme, and an operating parameter of the ship electric propulsion system is adjusted based on the control strategy to perform operation protection on the ship electric propulsion system.

[0085] In this embodiment, the preset policy library contains protection schemes for different fault types and system states. The policy library is designed through experience, historical data, and expert knowledge to ensure effective protection and adjustment when a system fault occurs. According to the fault type and system state obtained from the operation evaluation results, the system will search for the corresponding fault protection scheme in the preset policy library. The policy library can provide different levels of protection schemes according to different fault types, system states, and their severity.

[0086] In this embodiment, if the fault type is "low voltage" and the system status is "minor fault", it may match the "current limiting" or "automatic boosting" protection strategy. If the fault type is "current overload" and the system status is "serious fault", it may match the "power off" or "speed limiting" protection strategy.

[0087] In this embodiment, after determining the fault protection scheme, the control strategy related to the scheme is extracted. The control strategy includes how to adjust the operating parameters of the ship's electric propulsion system to minimize the impact of the fault and restore normal operation. For the "current overload" fault, the control strategy may be to reduce power output, limit current load, etc. For the "overtemperature" fault, the control strategy may be to turn on the cooling system, adjust the operating frequency of the equipment, etc. According to the selected control strategy, the system will automatically or manually adjust the operating parameters of the electric propulsion system to ensure that the system can continue to operate in the event of a fault, or quickly return to normal.

[0088] The present invention combines the system evaluation results to automatically match and execute the fault protection plan. The ship's electric propulsion system can respond quickly when a fault occurs and take appropriate control measures. This not only ensures the safety of the ship, but also improves the intelligence and adaptability of the system, providing a strong guarantee for the long-term stable operation of the ship's electric propulsion system.

[0089] In this embodiment, a time series signal set is generated through the real-time operation data set of the ship's electric propulsion system, and then multiple fault monitoring models are constructed through the HSMM algorithm. Different time series signals can be processed separately to enhance the system's monitoring capability for multi-dimensional faults. At the same time, the optimization of the ant colony algorithm can automatically adjust the parameters of the model to ensure that the model can maintain high accuracy in a dynamic environment and avoid failure of fault monitoring due to overfitting or underfitting of the model. In addition, the system operation state is fuzzy processed by the membership function, which can realize effective fuzzy evaluation of the operation state of the ship's electric propulsion system, so that reasonable judgments and decisions can still be made when facing partially inaccurate or incomplete data. Compared with traditional simple threshold judgments, fuzzy rules can make more reasonable judgments in the state transition stage, reduce misjudgments caused by boundary fuzziness, improve fault monitoring accuracy, and further match the most suitable fault protection scheme according to the fault evaluation results to ensure that the ship's electric propulsion system is targeted and effectively protected.

[0090] Please refer to Figure 2 , Figure 2 A structural diagram of a ship electric propulsion system operation protection system provided by an embodiment of the present invention includes: a data acquisition module 201, a monitoring model construction module 202, a fuzzy module 203, an evaluation module 204, a defuzzification module 205 and a maintenance module 206; The data acquisition module 201 is used to obtain a real-time operation data set of the ship electric propulsion system, pre-process the real-time operation data set, and obtain a time series signal set; The monitoring model building module 202 is used to build a plurality of first fault monitoring models based on the time series signal set and the HSMM algorithm, and optimize each first fault monitoring model based on the ant colony algorithm to obtain a plurality of second fault monitoring models; The fuzzy module 203 is used to obtain the hidden state sequence of the second fault monitoring model, generate a corresponding initial membership function based on each hidden state sequence, and perform fuzzification processing on the initial membership function to obtain a target membership function; The evaluation module 204 is used to determine the operating state of the ship electric propulsion system based on the target membership function and the preset fuzzy rules; The defuzzification module 205 is used to perform defuzzification processing on the operating state when there is an abnormality in the operating state, so as to obtain an operating evaluation result of the electric propulsion system; The maintenance module 206 is used to match the fault protection scheme based on the operation evaluation result, and adjust the operation parameters of the ship electric propulsion system based on the fault protection scheme to perform operation protection on the ship electric propulsion system.

[0091] In this embodiment, the data acquisition module is used to: Acquire a real-time operating data set of a ship electric propulsion system, and perform filtering and normalization processing on the real-time operating data set; the real-time operating data set includes current data, voltage data, speed data, vibration data and temperature data; A window length is set based on the operation mode of the ship electric propulsion system, and data in the real-time operation data set is divided based on the window length to obtain a time series signal set.

[0092] In this embodiment, the monitoring model building module is used to: Based on the operating state of the ship electric propulsion system, respectively set a hidden state and a corresponding hidden state duration distribution for the time series signals in the time series signal set; A first fault detection model is constructed based on the HSMM algorithm and the time series signal, wherein the first fault detection model includes a hidden state duration distribution, a hidden state transfer matrix, and an observation probability distribution. In this embodiment, the monitoring model building module is also used to: Initializing the ant colony size and search space based on the ant colony algorithm, and initializing the parameters of the first fault monitoring model; Iteratively performing ant colony search based on the ant colony size and search space to update parameters of the first fault monitoring model; In each iteration, the fitness of the updated first fault detection model is evaluated based on the fitness function until the fitness or the number of iterations reaches a preset condition, the iteration is stopped, and the second fault detection model is output.

[0093] In this embodiment, the fuzzy module is used to: Based on each second fault monitoring model, respectively obtain the hidden state sequence at each time point, and record the probability distribution corresponding to each hidden state; Calculating the probability value of the hidden state sequence based on the probability distribution corresponding to each hidden state, and generating the initial membership function corresponding to each second fault monitoring model based on the probability value; The initial membership function is mapped to a fuzzy set based on a fuzzification function to obtain a target membership function.

[0094] In this embodiment, the defuzzification module is used to: When the operating state is abnormal, a weighted average value of the target membership function is calculated based on a centroid method to obtain an operating evaluation result, wherein the operating evaluation result includes a system state and a fault type.

[0095] In this embodiment, the maintenance module includes: Matching a fault protection scheme in a preset strategy library based on the fault type and the system state; A control strategy is acquired based on the fault protection scheme, and an operating parameter of the ship electric propulsion system is adjusted based on the control strategy to perform operation protection on the ship electric propulsion system.

[0096] In an embodiment of the present invention, a terminal device is also provided, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, the above-mentioned ship electric propulsion system operation protection method is implemented.

[0097] In an embodiment of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the above-mentioned ship electric propulsion system operation protection method.

[0098] Exemplarily, the computer program may be divided into one or more modules, one or more modules are stored in a memory and executed by a processor to implement the present invention. One or more modules may be a series of computer program instruction segments capable of implementing specific functions, and the instruction segments are used to describe the execution process of the computer program in a terminal device.

[0099] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, and a cloud server. The terminal device may include, but is not limited to, a processor, a memory, and a display. Those skilled in the art will appreciate that the above components are merely examples of the terminal device and do not constitute a limitation on the terminal device. The terminal device may include more or fewer components than the components, or may combine certain components, or different components. For example, the terminal device may also include input and output devices, network access devices, buses, etc.

[0100] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the terminal device, and uses various interfaces and lines to connect various parts of the entire terminal device.

[0101] The memory can be used to store computer programs and / or modules. The processor implements various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function (such as a sound playback function, a text conversion function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, text message data, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0102] Among them, if the module based on the operation protection of the ship electric propulsion system is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. Computer-readable media may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0103] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for protecting the operation of a ship electric propulsion system, characterized in that: include: Acquire a real-time operation data set of a ship electric propulsion system, preprocess the real-time operation data set, and acquire a time series signal set; Constructing a plurality of first fault monitoring models based on the time series signal set and the HSMM algorithm, and optimizing each of the first fault monitoring models based on the ant colony algorithm to obtain a plurality of second fault monitoring models; Obtaining a hidden state sequence of the second fault monitoring model, generating a corresponding initial membership function based on each hidden state sequence, and performing fuzzification processing on the initial membership function to obtain a target membership function; Determining the operating state of the ship electric propulsion system based on the target membership function and the preset fuzzy rules; When the operating state is abnormal, performing defuzzification processing on the operating state to obtain an operating evaluation result of the electric propulsion system; A fault protection scheme is matched based on the operation evaluation result, and an operation parameter of the ship electric propulsion system is adjusted based on the fault protection scheme to perform operation protection on the ship electric propulsion system.

2. A method for protecting the operation of a ship electric propulsion system according to claim 1, characterized in that: The step of obtaining a real-time operation data set of the ship electric propulsion system and preprocessing the real-time operation data set includes: Acquire a real-time operating data set of a ship electric propulsion system, and perform filtering and normalization processing on the real-time operating data set; the real-time operating data set includes current data, voltage data, speed data, vibration data and temperature data; A window length is set based on the operation mode of the ship electric propulsion system, and data in the real-time operation data set is divided based on the window length to obtain a time series signal set.

3. A method for protecting the operation of a ship electric propulsion system according to claim 2, characterized in that: The constructing of a plurality of first fault monitoring models based on the time series signal set and the HSMM algorithm comprises: Based on the operating state of the ship electric propulsion system, respectively set a hidden state and a corresponding hidden state duration distribution for the time series signals in the time series signal set; A first fault detection model is constructed based on the HSMM algorithm and the time series signal, wherein the first fault detection model includes a hidden state duration distribution, a hidden state transfer matrix, and an observation probability distribution.

4. A method for protecting the operation of a ship electric propulsion system according to claim 3, characterized in that: The step of optimizing each first fault monitoring model based on the ant colony algorithm to obtain a plurality of second fault monitoring models includes: Initializing the ant colony size and search space based on the ant colony algorithm, and initializing the parameters of the first fault monitoring model; Iteratively performing ant colony search based on the ant colony size and search space to update parameters of the first fault monitoring model; In each iteration, the fitness of the updated first fault detection model is evaluated based on the fitness function until the fitness or the number of iterations reaches a preset condition, the iteration is stopped, and the second fault detection model is output.

5. A method for protecting the operation of a ship electric propulsion system according to claim 4, characterized in that: The step of obtaining the hidden state sequence of the second fault monitoring model, generating a corresponding initial membership function based on each hidden state sequence, and performing fuzzification processing on the initial membership function to obtain a target membership function includes: Based on each second fault monitoring model, respectively obtain the hidden state sequence at each time point, and record the probability distribution corresponding to each hidden state; Calculating the probability value of the hidden state sequence based on the probability distribution corresponding to each hidden state, and generating the initial membership function corresponding to each second fault monitoring model based on the probability value; The initial membership function is mapped to a fuzzy set based on a fuzzification function to obtain a target membership function.

6. A method for protecting the operation of a ship electric propulsion system according to claim 5, characterized in that: When the operating state is abnormal, performing defuzzification processing on the operating state to obtain an operating evaluation result of the ship electric propulsion system includes: When the operating state is abnormal, a weighted average value of the target membership function is calculated based on a centroid method to obtain an operating evaluation result, wherein the operating evaluation result includes a system state and a fault type.

7. A method for protecting the operation of a ship electric propulsion system according to claim 6, characterized in that: The matching of the fault protection scheme based on the operation evaluation result, and adjusting the operation parameters of the ship electric propulsion system based on the fault protection scheme to perform operation protection on the ship electric propulsion system, includes: Matching a fault protection scheme in a preset strategy library based on the fault type and the system state; A control strategy is acquired based on the fault protection scheme, and an operating parameter of the ship electric propulsion system is adjusted based on the control strategy to perform operation protection on the ship electric propulsion system.

8. A ship electric propulsion system operation protection system, characterized in that: include: Data acquisition module, monitoring model building module, fuzzy module, evaluation module, defuzzification module and maintenance module; The data acquisition module is used to obtain a real-time operation data set of the ship electric propulsion system, pre-process the real-time operation data set, and obtain a time series signal set; The monitoring model building module is used to build a plurality of first fault monitoring models based on the time series signal set and the HSMM algorithm, and optimize each first fault monitoring model based on the ant colony algorithm to obtain a plurality of second fault monitoring models; The fuzzy module is used to obtain the hidden state sequence of the second fault monitoring model, generate a corresponding initial membership function based on each hidden state sequence, and perform fuzzification on the initial membership function to obtain a target membership function; The evaluation module is used to determine the operating state of the ship electric propulsion system based on the target membership function and the preset fuzzy rules; The defuzzification module is used to perform defuzzification processing on the operating state when there is an abnormality in the operating state, so as to obtain an operating evaluation result of the electric propulsion system; The maintenance module is used to match the fault protection scheme based on the operation evaluation result, and adjust the operation parameters of the ship electric propulsion system based on the fault protection scheme to perform operation protection on the ship electric propulsion system.

9. A ship electric propulsion system operation protection system as claimed in claim 8, characterized in that: The monitoring model building module is used to: Based on the operating state of the ship electric propulsion system, respectively set a hidden state and a corresponding hidden state duration distribution for the time series signals in the time series signal set; A first fault detection model is constructed based on the HSMM algorithm and the time series signal, wherein the first fault detection model includes a hidden state duration distribution, a hidden state transfer matrix, and an observation probability distribution.

10. A ship electric propulsion system operation protection system according to claim 9, characterized in that: The monitoring model building module is also used to: Initializing the ant colony size and search space based on the ant colony algorithm, and initializing the parameters of the first fault monitoring model; Iteratively performing ant colony search based on the ant colony size and search space to update parameters of the first fault monitoring model; In each iteration, the fitness of the updated first fault detection model is evaluated based on the fitness function until the fitness or the number of iterations reaches a preset condition, the iteration is stopped, and the second fault detection model is output.

Images