Intelligent detection method and system for ship performance state

Abstract

The invention provides a ship performance state intelligent detection method and system. The method provided by the invention comprises the steps of determining a to-be-monitored variable in a ship operation process; determining the measuring range of a monitoring sensor corresponding to each variable type according to the alarm threshold value of the to-be-monitored variable; determining a monitoring position point of each to-be-monitored variable according to the space structure information of the ship; determining a plurality of monitoring sensors according to the ship space structure in the first neighborhood of the monitoring position point and the distance from the monitoring position point to the acquisition module, and installing the monitoring sensors at the monitoring position point; according to the monitoring position point and the variable type corresponding to each monitoring sensor, matching the corresponding relationship between the monitoring sensor and the acquisition module; a plurality of acquisition modules in the data acquisition and transmission unit acquire the real-time monitoring state of each monitoring sensor, and send the identified real-time monitoring information in the abnormal state to the central processing unit; and the central processing unit diagnoses the real-time monitoring information and determines the fault position and the fault type of the ship.

Description

Technical Field

[0001] The present application relates to the technical field of ship fault diagnosis, and in particular to a method and system for intelligently detecting ship performance status. Background Art

[0002] With the rapid development of the shipping industry, the demand for safe, efficient and intelligent management of ship operations is growing. Ship internal signal acquisition is essential to ensure the safe, stable and efficient operation of ships. By collecting the operating data of various ship systems in real time, such as engine performance, navigation status, environmental parameters, etc., it can provide strong support for ship maintenance, fault diagnosis and navigation decision-making.

[0003] The current ship signal collection methods mostly use distributed sensor nodes and centralized data collection mode. Although this method can cover a variety of signal types, it usually has the following disadvantages: 1. Accuracy limitation: Some sensors cannot provide high-precision data in complex environments due to their weak anti-interference ability.

[0004] 2. Lack of real-time performance: The data transmission and processing speeds are limited and cannot meet the needs of rapid response, especially in emergency situations or highly dynamic navigation conditions.

[0005] 3. Compatibility issues: The acquisition interface standards between different systems are different, which makes signal integration and collaborative analysis more complicated.

[0006] 4. Complexity of data processing: The collection and storage of massive data requires high bandwidth and computing power. Existing equipment is inefficient in processing large-scale multi-dimensional data, resulting in the possibility of delays or omissions of key information.

[0007] Therefore, there is an urgent need for an intelligent detection method for ship performance status to realize the automatic collection, processing and transmission of ship fault signals, improve the safety and reliability of ship operation, reduce operating costs, and enhance the intelligence level of shipping management. Summary of the invention

[0008] The present application provides a method and system for intelligent detection of ship performance status, which are used to realize automatic collection, processing and transmission of ship fault signals, improve the safety and reliability of ship operation, reduce operating costs, and enhance the intelligence level of shipping management.

[0009] Specifically, the present application is implemented through the following technical solutions: The first aspect of the present application provides a method for intelligently detecting a ship's performance status, the method comprising: Determining variables to be monitored during the operation of the ship, wherein the variables to be monitored include a plurality of different variable types; Determine the range of the monitoring sensor corresponding to each variable type according to the alarm threshold of the variable to be monitored; Determining the monitoring location points of each variable to be monitored according to the spatial structure information of the ship; Determine a plurality of monitoring sensors according to the spatial structure of the ship in the first neighborhood of the monitoring location point and the distance between the monitoring location point and the data acquisition and transmission unit, and install the monitoring sensors at the monitoring location point; According to the monitoring location point and the variable type corresponding to each of the monitoring sensors, the corresponding relationship between the monitoring sensor and the acquisition module is matched, the acquisition module is multiple, each acquisition module corresponds to multiple monitoring sensors, and the monitoring sensors corresponding to each acquisition module are completely different; The multiple acquisition modules in the data acquisition and transmission unit acquire the real-time monitoring status of each monitoring sensor, each acquisition module cyclically acquires the corresponding multiple monitoring sensors in sequence, and the sampling frequency of each acquisition module is different; The data acquisition and transmission unit identifies an abnormal state in the real-time monitoring state, and sends the real-time monitoring information in the abnormal state to the central processing unit, wherein the real-time monitoring signal in the abnormal state exceeds the range of the corresponding monitoring sensor; The central processing unit diagnoses the real-time monitoring information to determine the fault location and fault type of the ship.

[0010] The second aspect of the present application provides a ship performance status intelligent detection system, the system comprising a data acquisition and transmission unit and a central processing unit connected to each other; wherein, The data acquisition and transmission unit is used to determine the variables to be monitored during the operation of the ship, and the variables to be monitored include multiple different variable types; The data acquisition and transmission unit is also used to determine the range of the monitoring sensor corresponding to each variable type according to the alarm threshold of the variable to be monitored; The data acquisition and transmission unit is also used to determine the monitoring position points of each variable to be monitored according to the spatial structure information of the ship; The data acquisition and transmission unit is further used to determine a plurality of monitoring sensors according to the ship spatial structure in the first neighborhood of the monitoring location point and the distance from the monitoring location point to the data acquisition and transmission unit, and install the monitoring sensors at the monitoring location point; The data acquisition and transmission unit is also used to match the corresponding relationship between the monitoring sensor and the acquisition module according to the monitoring location point and the variable type corresponding to each of the monitoring sensors, the acquisition modules are multiple, each acquisition module corresponds to multiple monitoring sensors, and the monitoring sensors corresponding to each acquisition module are completely different; The data acquisition and transmission unit is also used to acquire the real-time monitoring status of each monitoring sensor based on multiple acquisition modules, each acquisition module cyclically acquires the corresponding multiple monitoring sensors in sequence, and the sampling frequency of each acquisition module is different; The data acquisition and transmission unit is also used to identify an abnormal state in the real-time monitoring state, and send the real-time monitoring information in the abnormal state to the central processing unit, and the real-time monitoring signal in the abnormal state exceeds the range of the corresponding monitoring sensor; The central processing unit is used to diagnose the real-time monitoring information and determine the fault location and fault type of the ship.

[0011] The intelligent detection system and method for ship performance status provided by the present application, in the first aspect, the range of the monitoring sensor is set according to the alarm threshold of the variable to be monitored, ensuring that the most suitable monitoring sensor can be used for each variable type. Specifically, by selecting the appropriate range and monitoring sensor for different types of variables, it is possible to avoid excessive or insufficient data collection, ensure that the collected data information is accurate and effective, and reduce the transmission and storage of irrelevant data. In addition, the acquisition module collects data from different monitoring sensors in a patrol manner, but only uploads data when an abnormal signal occurs, which effectively reduces the amount of data transmission and avoids excessive waste of resources. Through multiple acquisition modules of the data acquisition and transmission unit, multiple acquisition modules collect real-time monitoring status in turn at different sampling frequencies, and upload relevant information to the central processing unit according to the abnormal status. Each acquisition module performs data acquisition according to different types of monitoring variables, and can obtain key information in real time without interfering with the normal operation of ship equipment. The central processing unit uses these data to quickly identify the fault location and category, thereby helping offshore workers to more quickly discover equipment failures and repair them in a timely manner. Through this intelligent and automated approach, the intensity of work is reduced and the response time is greatly shortened. On the second aspect, this application builds a full-process, closed-loop intelligent detection system for ship performance status from real-time monitoring to data analysis to fault diagnosis. This application is real-time, and multiple acquisition modules in the data acquisition and transmission unit collect the real-time monitoring status of each monitoring sensor in real time, and send the real-time monitoring information under abnormal conditions to the central processing unit. This application is accurate, and the central processing unit diagnoses and analyzes the real-time monitoring information under abnormal conditions to achieve rapid fault location and classification. Through accurate grasp of the ship's operating status, this system provides a scientific basis for equipment maintenance, fault handling and operation optimization, and the closed-loop system effectively solves the defects of traditional monitoring methods with strong lag and diagnosis relying on manual experience, improves the safety and efficiency of ship operation, and enables offshore workers to more quickly discover faulty equipment and repair it in a timely manner, while enhancing the intelligence level of my country's ship system. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 A flow chart of a method for intelligently detecting ship performance status provided in Example 1 of the present application; Figure 2 A schematic diagram of the structure of the intelligent detection system for ship performance status provided in Example 2 of the present application; Figure 3 This is a schematic diagram of the structure of the intelligent detection system for ship performance status provided in Example 3 of the present application. DETAILED DESCRIPTION

[0013] Here, exemplary embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application.

[0014] The terms used in this application are only for the purpose of describing specific embodiments and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in this article refers to and includes any or all possible combinations of one or more associated listed items.

[0015] It should be understood that although the terms first, second, third, etc. may be used in the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0016] Specific embodiments are given below to introduce the technical solution of the present application in detail.

[0017] Figure 1 This is a flow chart of the intelligent detection method for ship performance status provided in Example 1 of this application. Figure 1 , the method provided in this embodiment may include: S101. Determine variables to be monitored during ship operation, where the variables to be monitored include multiple different variable types.

[0018] Specifically, the variables to be monitored refer to the data that need to be monitored during the operation of the ship. Multiple different types of variables include temperature, pressure, speed, electricity, and gas concentration (SO2, CO).

[0019] In specific implementation, the operation status information of the ship is obtained during operation, and the variable types closely related to the ship's performance are identified. According to the ship's operating requirements, the monitoring priority of different variable types is determined, and the variable type with a higher monitoring priority is selected from the identified variable types based on the monitoring priority as the variable to be monitored during the ship's operation.

[0020] S102: Determine the range of the monitoring sensor corresponding to each variable type according to the alarm threshold of the variable to be monitored.

[0021] Specifically, the alarm threshold represents a critical value of the monitored variable outside the normal operating range. When the monitored variable exceeds the alarm threshold, it indicates that the monitoring sensor may be abnormal. The alarm threshold is usually set based on the safe operation standards, empirical data or historical monitoring information of the equipment. It represents the maximum limit to which the monitored variable can operate safely during operation. Exceeding the alarm threshold may cause equipment failure or safety hazards. It should be noted that different variables to be monitored have different corresponding alarm thresholds.

[0022] Furthermore, the range refers to the range that the monitoring sensor can accurately measure, usually defined by its minimum measurable value to its maximum measurable value. Between these upper and lower limits, the monitoring sensor can effectively sense changes in the variables to be monitored and output corresponding data. The setting of the range should not only consider the alarm threshold, but also need to be adjusted according to the normal fluctuation range of the variables to be monitored and external disturbance factors to ensure that the monitoring sensor can work efficiently and accurately throughout the normal operating range, and can effectively capture data changes when abnormalities occur.

[0023] In specific implementation, the range of the monitoring sensor corresponding to each variable type is determined according to the alarm threshold of the variable to be monitored, including: analyzing the fluctuation range of each variable to be monitored under normal operating conditions based on historical monitoring data, and determining the normal measurement value interval of each variable to be monitored; determining the floating interval of each variable to be monitored based on external disturbance factors; determining the lower limit of the range of each variable to be monitored based on the sum of the normal measurement value interval and the floating interval, and determining the upper limit of the range of each variable to be monitored based on the normal measurement value interval, the floating interval and the alarm threshold of the variable to be monitored; wherein the alarm threshold is used to determine the critical value of the abnormal state outside the normal measurement value interval.

[0024] Specifically, collect historical monitoring data, analyze the maximum and minimum measured values ​​of each variable to be monitored in the historical monitoring data during normal operation, and determine the normal measurement value interval of each variable to be monitored with the maximum measured value as the upper limit of the interval and the minimum measured value as the lower limit of the interval. Identify and evaluate the impact of external disturbance factors (such as temperature changes, load changes, etc.) on the monitored variables, and determine the floating interval of each variable to be monitored based on the impact of external disturbance factors on the monitored variables. Further, add the normal measurement value interval and the minimum value of the floating interval (i.e., the lower limit of the interval) to obtain the lower limit of the range of each variable to be monitored. Add the maximum values ​​of the normal measurement value interval and the floating interval (i.e., the upper limit of the interval) to obtain the upper limit of the range of each variable to be monitored. Further, adjust the upper limit of the range by comparing the upper limit of the range with the alarm threshold of the variable to be monitored. If the upper limit of the range is less than the alarm threshold, increase the upper limit of the range according to the difference between the two, and the increased upper limit of the range is greater than the alarm threshold. If the upper limit of the range is greater than the alarm threshold, there is no need to adjust the upper limit of the range. According to the determined upper and lower limits of the measuring range, the measuring range of the monitoring sensor corresponding to each variable type is determined.

[0025] The method provided in this embodiment provides a comprehensive and accurate method for determining the range of the monitoring sensor. By analyzing the historical monitoring data, considering the external disturbance factors and setting the alarm threshold, the rationality and accuracy of the monitoring sensor range are ensured. First, based on the historical monitoring data, the fluctuation range of each variable to be monitored under normal operating conditions is analyzed, and a reasonable normal measurement value interval can be set for each variable to be monitored, so that the sensor will not exceed its monitoring range when the ship equipment is operating normally, and the measurement error or data loss caused by improper range setting of the sensor is avoided. By determining the normal measurement value interval of each variable to be monitored, it is ensured that the system monitors the normal operating state stably and reliably. Secondly, considering the external disturbance factors further expands the adaptability of the range. During the operation of the ship, it is affected by a variety of environmental factors, such as climate, ocean conditions and dynamic changes of equipment, which may cause certain fluctuations or fluctuations in the monitoring variables. By determining the floating interval based on the external disturbance factors, the influence of these external changes on the monitoring variables can be effectively captured, and false alarms or missed reports caused by environmental changes can be avoided. Through the reasonable setting of the floating interval, the monitoring system can adapt to the changing operating environment and improve the robustness of data acquisition. On this basis, by taking the sum of the normal measurement value interval and the floating interval as the lower limit of the range, and combining the alarm threshold to determine the upper limit of the range, a dynamic monitoring range is set for the monitoring sensor. This setting can flexibly respond to changes in the normal operation of the equipment, while ensuring that once the monitoring value exceeds the set upper limit of the range, the alarm mechanism can be triggered in time. The setting of the alarm threshold can help the system quickly identify potential abnormal conditions during the operation of the ship, provide real-time warnings to the crew, avoid the spread of faults and improve maintenance efficiency.

[0026] S103: Determine monitoring location points of various variables to be monitored according to the spatial structure information of the ship.

[0027] Specifically, the spatial structure information of a ship represents the overall layout and shape of the interior and exterior of the ship, as well as the relative positional relationship of each component. It describes the spatial arrangement of various areas, cabins, equipment, pipelines, sensor locations, etc. of the ship. The monitoring location point refers to a location in the spatial structure of the ship that is specially selected for installing monitoring sensors to collect data related to the variables to be monitored.

[0028] In specific implementation, the detailed spatial layout of the ship is obtained through ship design drawings, three-dimensional modeling or other methods, including information such as each cabin, equipment, pipeline, and sensor installation location. According to the variable type of the variable to be monitored, determine the location where each variable to be monitored needs to be monitored. According to the characteristics of the ship's spatial structure and the variables to be monitored, consider factors such as easy installation of sensors, effective reflection of variable changes, and ensuring data collection accuracy to select the most suitable monitoring location point. According to the spatial structure information of the ship, record the coordinate information of the selected monitoring location point and obtain it through coordinate system calibration or three-dimensional modeling.

[0029] For example, in one embodiment, the monitoring point for oil temperature is at the inlet and outlet of the main and auxiliary engine oil ports, the monitoring point for oil pressure is inside the fuel tank, and the monitoring point for gas concentration is in the cargo hold.

[0030] S104, determining a plurality of monitoring sensors according to the spatial structure of the ship in the first neighborhood of the monitoring location point and the distance between the monitoring location point and a data acquisition and transmission unit, and installing the monitoring sensors at the monitoring location point.

[0031] Specifically, the first neighborhood is set according to actual needs, and is not limited in this embodiment. Different monitoring sensors correspond to different monitoring locations, and each monitoring location needs to be installed with multiple types of monitoring sensors.

[0032] In specific implementation, the spatial structure of the ship in the first neighborhood of the monitoring location is obtained according to the spatial structure information of the ship, including the equipment, structures and other factors around the monitoring location that may affect the installation of the monitoring sensor. Calculate the distance from the monitoring location to the data acquisition and transmission unit. According to the characteristics of the monitoring location (such as environmental conditions, variable types of variables to be monitored, etc.), determine the type and number of monitoring sensors suitable for installation. When determining the monitoring sensor, comprehensively consider the measurement range of the monitoring sensor, which should cover the distance from the monitoring location to the data acquisition and transmission unit. Install multiple monitoring sensors at the monitoring location. After the installation is completed, connect the monitoring sensor to the data acquisition and transmission unit to ensure that the data acquisition and transmission unit can receive real-time data from the monitoring sensor.

[0033] Specifically, the first neighborhood environment of the monitoring location is obtained in combination with the spatial structure information of the ship. The first neighborhood includes the physical environment around the monitoring location (such as cabins, equipment, walls, etc.) and factors that may affect the installation of sensors (such as vibration, temperature changes, etc.). According to the results of the neighborhood environment analysis, determine whether the location is suitable for installing the required type of monitoring sensor. For example, if the temperature is monitored, it may be necessary to avoid extreme areas of too high or too low temperature; if the pressure or vibration is monitored, it is necessary to select a location that will not be disturbed by mechanical equipment or vibration sources. According to the spatial structure of the ship and the type of variables to be monitored, determine the layout of multiple monitoring sensors. These monitoring sensors should cover different dimensions of the monitoring location to ensure the accuracy of the monitoring data. Each monitoring location may require multiple sensors to monitor different variables (such as temperature, pressure, humidity, vibration, etc.). According to the layout of the ship, calculate the distance from the monitoring location to the data acquisition and transmission unit to ensure that the signal transmission path of the selected monitoring location is the shortest and is not interfered by the outside world, so as to ensure the real-time and accuracy of the data. Too far distance may cause signal attenuation or transmission delay. After determining the monitoring location and the layout of the monitoring sensor, select an appropriate installation method to fix the monitoring sensor on the monitoring location. When installing, you need to consider the safety, maintainability, and long-term stability of the equipment. For example, a temperature sensor can be directly fixed on the surface of the equipment to be monitored; a vibration sensor may need to be installed on a structural surface with a certain rigidity.

[0034] S105. Match the correspondence between the monitoring sensors and the acquisition modules according to the monitoring location points and the variable types corresponding to the respective monitoring sensors. There are multiple acquisition modules, each of which corresponds to multiple monitoring sensors, and the monitoring sensors corresponding to each acquisition module are completely different.

[0035] Specifically, the data acquisition and transmission unit includes a plurality of acquisition modules, each of which is used to acquire variables to be monitored from a plurality of monitoring sensors of the same type, and different acquisition modules correspond to different types of monitoring sensors.

[0036] In a specific implementation, the correspondence between the monitoring sensors and the acquisition modules is matched according to the monitoring location points and the variable types corresponding to the respective monitoring sensors, including: classifying the monitoring sensors according to the variable types corresponding to the respective monitoring sensors to obtain a plurality of first categories, wherein the variable types of the respective monitoring sensors in each of the first categories are the same, and the variable types corresponding to different first categories are different; determining the monitoring period of each first category according to the variable type; clustering the distances between the respective monitoring sensors in each first category and the data acquisition and transmission unit according to the monitoring period and the number of monitoring sensors in each first category, dividing each first category based on the clustering results to obtain a plurality of second categories, and establishing a correspondence between each second category and each acquisition module.

[0037] Specifically, different monitoring sensors correspond to different variable types. For example, a gas concentration sensor corresponds to gas concentration, a speed sensor corresponds to speed, and a temperature sensor corresponds to temperature. The variable types corresponding to each monitoring sensor in each first category are the same, and the variable types corresponding to each monitoring sensor in different first categories are different. For example, one first category is a temperature sensor, and the other first category is a pressure sensor, and the variable types corresponding to the two first categories are different.

[0038] It should be noted that the first category is classified according to the variable type of the monitoring sensor, that is, all monitoring sensors of the same type will be classified into the same category. For example, if multiple monitoring sensors are monitoring temperature, they belong to the same first category; if other sensors monitor pressure, they will belong to different first categories. The second category is to further subdivide the monitoring sensors in the first category according to the distance and monitoring period to the data acquisition and transmission unit. Each monitoring sensor in the first category will be divided into multiple second categories according to the distance and monitoring period between it and the data acquisition and transmission unit, and monitoring sensors with similar distances and similar monitoring periods will be classified into the same second category. For example, in one embodiment, there are 3 temperature sensors in the first category, temperature sensor 1: 10 meters away from the data acquisition and transmission unit, monitoring period is 10 seconds; temperature sensor 2: 20 meters away from the data acquisition and transmission unit, monitoring period is 15 seconds; temperature sensor 3: 30 meters away from the data acquisition and transmission unit, monitoring period is 12 seconds. The temperature sensors in the first category are further subdivided into the second category according to their distance from the data acquisition and transmission unit and the monitoring period, category 21: temperature sensor 1 (closer to the data acquisition and transmission unit, with a monitoring period of 10 seconds), category 2: temperature sensor 2, temperature sensor 3 (these temperature sensors are far from the data acquisition and transmission unit, with monitoring periods of 15 seconds and 12 seconds respectively).

[0039] In the specific implementation, all monitoring sensors are classified according to the corresponding types of variables to be monitored, and monitoring sensors with the same variable type are classified into one category to obtain multiple first categories. According to the characteristics of the variable type corresponding to each first category and the monitoring requirements, the monitoring period (i.e., sampling frequency) of each monitoring sensor in each first category is determined. The distance between each monitoring sensor in each first category and the data acquisition and transmission unit is calculated. For each monitoring sensor in the first category, it is clustered according to the distance between it and the data acquisition and transmission unit. At the same time, considering the number and monitoring period of the monitoring sensors, the monitoring sensors with small differences in monitoring period and number are classified into the same category to obtain multiple second categories. In each second category, the distance between the monitoring sensor and the data acquisition and transmission unit is relatively close. Furthermore, for each monitoring sensor in the second category, a corresponding acquisition module is established. Each acquisition module will be responsible for collecting data from one or more monitoring sensors in the second category and transmitting data according to the sampling frequency.

[0040] Optionally, there are N variable types of the variables to be monitored, and there are M monitoring components corresponding to the variables to be monitored. N and M are positive integers greater than or equal to 2, N>M, and the variable types collected by the monitoring sensors corresponding to each acquisition module are the same.

[0041] Specifically, the number of variable types of the variables to be monitored is greater than the number of monitoring components, that is, at least one monitoring component has more than one different type of variables to be monitored. The monitoring sensors corresponding to each acquisition module acquire the same variable type.

[0042] For example, in one embodiment, a CO concentration sensor, a SO2 concentration sensor, and an O2 concentration sensor are arranged in the cargo hold, and the gas concentration is monitored in a roving manner through the acquisition module 1; speed sensors are arranged in the main and auxiliary engines, and the speeds of the main and auxiliary engines are monitored in a roving manner through the acquisition module 2; temperature sensors are arranged in the engine oil tank and the main and auxiliary engine oil tanks, and the oil temperature is monitored in a roving manner through the acquisition module 3; pressure sensors are arranged in the engine oil tank and the main and auxiliary engine mailboxes, and the oil pressure is monitored in a roving manner through the acquisition module 4; power sensors are installed at each group of generators on the ship, and the power of each generator is monitored in a roving manner through the acquisition module 5.

[0043] S106. The multiple acquisition modules in the data acquisition and transmission unit acquire the real-time monitoring status of each monitoring sensor. Each acquisition module cyclically acquires the corresponding multiple monitoring sensors in sequence. The sampling frequencies of the acquisition modules are different.

[0044] Specifically, in combination with the above description, the monitoring sensors corresponding to each acquisition module collect different types of variables, and the monitoring sensors corresponding to different acquisition modules collect different types of variables. In this step, each acquisition module sequentially and cyclically collects the real-time monitoring status of the corresponding multiple monitoring sensors. The cycle period is set according to actual needs. In this embodiment, it is not limited to this. For example, in one embodiment, the acquisition module collects data once every 4 seconds.

[0045] It should be noted that different acquisition modules have different sampling frequencies. For example, for a temperature sensor corresponding to an acquisition module, the sampling frequency may be once per minute, while for a vibration sensor corresponding to an acquisition module, the sampling frequency may be once per second.

[0046] S107, the data acquisition and transmission unit identifies an abnormal state in the real-time monitoring state, and sends the real-time monitoring information in the abnormal state to the central processing unit, and the real-time monitoring signal in the abnormal state exceeds the range of the corresponding monitoring sensor.

[0047] Specifically, when the signal value of real-time monitoring under abnormal conditions exceeds the range of the corresponding monitoring sensor, that is, when the monitoring information detected by the monitoring sensor is not within its corresponding range, the monitoring information is determined to be real-time monitoring information under abnormal conditions. The data acquisition and transmission unit is used to perform local mean decomposition on the real-time monitoring state detected by the monitoring sensor, perform abnormality detection on the decomposed detail curve, identify the real-time monitoring information under abnormal conditions, and transmit the real-time monitoring information under abnormal conditions to the central processing unit.

[0048] In a specific implementation, the data acquisition and transmission unit identifies an abnormal state in the real-time monitoring state, including: (1) The data acquisition and transmission unit performs local mean decomposition on the real-time monitoring state to obtain a local mean curve.

[0049] Specifically, the local mean curve represents the low-frequency component of the real-time monitoring state, which reflects the trend or overall change trend of the real-time monitoring state. The local mean curve is the basic waveform or envelope of the real-time monitoring state, usually contains the slowly changing information of the real-time monitoring state, and is the stable part of the real-time monitoring state.

[0050] In a specific implementation, local extreme point detection is performed on the real-time monitoring state to determine all local maximum values ​​and local minimum values ​​in the real-time monitoring state; the local maximum values ​​and local minimum values ​​are screened to eliminate pseudo extreme points or abnormal points; all the screened local maxima and all the local minimum values ​​are connected by linear interpolation to obtain a first envelope curve and a second envelope curve; the average value of the first envelope curve and the second envelope curve is determined as the local mean curve.

[0051] Specifically, the signal curve of the real-time monitoring state changing over time is analyzed, each sampling point on the signal curve is traversed, the amplitude of the current point is compared with the amplitude of the adjacent points before and after it, and the maximum and minimum values ​​are determined at the turning point of the signal curve. The local maximum value is the peak value with a smaller value of the adjacent point, and the local minimum value is the valley value with a larger value of the adjacent point. Based on the amplitude difference between extreme points (if the amplitude difference between extreme points is small and close to the noise level, it is regarded as a pseudo-extreme point), spacing (if the time interval between two extreme points is too short, it may be caused by noise, and the abnormal points should be removed), smoothness, etc., the pseudo-extreme points that may be caused by noise or interference are screened out. Further, linear interpolation is used for the local maximum values ​​after screening, and the first envelope curve is obtained by connection. Linear interpolation is used for the local minimum values ​​after screening, and the second envelope curve is obtained by connection. By calculating the average value of the amplitude of the first envelope curve and the second envelope curve at each time point, the local mean point at each time point is obtained, and all local mean points are connected to obtain the local mean curve.

[0052] (2) Removing the local mean curve from the real-time monitoring state to obtain a detail curve.

[0053] Specifically, the detail curve represents the high-frequency component of the real-time monitoring state, which reflects the fluctuation or detail change of the real-time monitoring state. The detail curve reflects the local characteristics of the real-time monitoring state, such as short-term changes, noise, pulse components, etc.

[0054] In specific implementation, combined with the above description, the real-time monitoring state includes a local mean curve and a detail curve. In this step, the local mean curve in the real-time monitoring state is removed, and the remaining part of the real-time monitoring state after removal is the detail curve.

[0055] (3) For each time point in the detail curve, an abnormal state in the real-time monitoring state is determined based on the difference between the real-time monitoring state corresponding to the time point and a preset threshold.

[0056] Specifically, the preset threshold is set according to actual needs, and is not limited in this embodiment.

[0057] In specific implementation, each time point in the detail curve is traversed, and the magnitude relationship between the amplitude at each time point and the preset threshold is used to determine whether it is an abnormal state. When the amplitude is greater than the preset threshold, the real-time monitoring state corresponding to the time point is determined to be an abnormal state. After traversing all time points, all abnormal states are obtained.

[0058] The method provided in this embodiment introduces a local mean decomposition method in the data acquisition and transmission unit, decomposes the real-time monitoring state into a local mean and a detail curve for analysis, and obtains an abnormal state. In the subsequent process, only the screened abnormal state is subjected to fault analysis to avoid processing a large number of normal signals, thereby greatly reducing the complexity of data processing and resource consumption. By concentrating computing resources on key data, the speed and efficiency of fault diagnosis are improved, and the system can respond to potential faults more quickly. In addition, normal signals do not participate in fault analysis, which can reduce the interference of background noise and normal fluctuations on the diagnosis results, thereby reducing the possibility of false alarms and missed alarms.

[0059] S108. The central processing unit diagnoses the real-time monitoring information to determine the fault location and fault type of the ship.

[0060] In a specific implementation, the central processing unit diagnoses the real-time monitoring information to determine the fault location and fault category of the ship, including: constructing a training data set of a multi-classification support vector machine based on historical monitoring data and the real-time monitoring information; the training data set includes measurement values ​​of multiple variables to be monitored and corresponding fault locations and fault categories; training a multi-classification support vector machine model based on the training data set, and optimizing the penalty parameters and kernel parameters in the model through the sparrow algorithm; wherein, a set of penalty parameters and kernel parameters are randomly selected for each sparrow, and the training error of the initial model under each set of penalty parameters and kernel parameters is calculated, and the penalty parameters and kernel parameters are adjusted based on the training error until the training error is less than a preset error value or reaches a preset number of iterations; retraining the model based on the optimized penalty parameters and kernel parameters, inputting the real-time monitoring information into the trained model, and outputting the corresponding ship's fault location and fault category; wherein, the model determines different fault states as different categories, establishes multiple decision boundaries, and each decision boundary determines the attribution of one category to all other categories.

[0061] Specifically, historical monitoring data during the operation of the ship is collected, including the measured values ​​of each variable to be monitored (such as temperature, pressure, vibration, etc.), the corresponding fault location and fault category. The historical monitoring data is combined with the real-time monitoring information to form a training data set for a multi-classification support vector machine (SVM). Each data point in the training data set contains the measured values ​​of multiple variables to be monitored and their corresponding fault locations and fault categories. The multi-classification support vector machine model is trained using the training data set. During the training process, the penalty parameters and kernel parameters in the multi-classification support vector machine model are optimized using the sparrow algorithm. In each iterative training, a set of penalty parameter and kernel parameter combinations are randomly selected, and the multi-classification support vector machine model is trained using the training data set to calculate the training error under the combination parameters. Based on the difference between the training error and the preset error value, iterative training is continuously performed through the feedback mechanism of the sparrow algorithm, and the penalty parameters and kernel parameters are repeatedly optimized until the training error is less than the preset error value or the number of iterations reaches the preset number of iterations, and the optimal penalty parameters and kernel parameters are determined. Further, based on the optimal penalty parameters and kernel parameters, the multi-classification support vector machine model is retrained using the training data set. During the training process, the multi-classification support vector machine model builds multiple decision boundaries according to different fault states to distinguish different fault locations and fault categories. After the multi-classification support vector machine model is trained, the real-time monitoring data is input into it. The multi-classification support vector machine model determines the corresponding fault location and fault category based on the relationship between the measured value of the input data and the decision boundary.

[0062] It should be noted that the multi-classification support vector machine model regards different fault states (fault locations and fault categories) as different categories and establishes multiple decision boundaries, each of which determines the attribution of one fault state to other fault states.

[0063] Optionally, before the central processing unit diagnoses the real-time monitoring information, the method also includes: the central processing unit performs data cleaning on the real-time monitoring information; filters and amplifies the real-time monitoring information in the time domain, so that the processed real-time monitoring information is free of noise and has a higher signal-to-noise ratio than before processing; performs statistical feature calculation and correlation analysis on the filtered and amplified real-time monitoring information, and extracts the similarity and correlation of the real-time monitoring information at different times; based on the similarity and correlation, obtains characteristic parameters during the operation of the ship, and determines the time domain feature graph of the real-time monitoring information based on the characteristic parameters.

[0064] In the specific implementation, the central processing unit performs data cleaning on the received real-time monitoring information, including removing invalid data, filling missing values, eliminating duplicate data, etc., to ensure the quality of the data. Further, the real-time monitoring information after cleaning is filtered in the time domain, and a suitable filter (such as a low-pass, high-pass or band-pass filter) is selected to remove the high-frequency noise in the real-time monitoring information. The filtered real-time monitoring information is amplified, and the strength of the real-time monitoring information is increased by appropriate gain to ensure that the signal-to-noise ratio of the real-time monitoring information is higher than before processing. The real-time monitoring information after filtering and amplification is calculated for statistical features, including mean, variance, peak, skewness, kurtosis, etc. The calculated statistical features are subjected to correlation analysis to determine the correlation between different monitoring variables. For example, the correlation between vibration and temperature of certain parts of the ship, or the mutual influence of different monitoring variables under different fault conditions, is analyzed. Based on the results of the correlation analysis, the similarity and correlation of the real-time monitoring information at different times are extracted by methods such as correlation coefficient, mutual information, and Pearson correlation. Based on the similarity and correlation of the real-time monitoring information at different times, the characteristic parameters of the ship operation process are further extracted. These characteristic parameters represent the key indicators of ship performance and can effectively reflect the ship's operating status and fault conditions. Based on the extracted characteristic parameters, a time domain characteristic graph is generated. The time domain characteristic graph shows the changes in the ship's status at different time points, which can reflect the ship's operating process and abnormal status, and facilitate subsequent fault diagnosis.

[0065] The method provided in this embodiment brings significant benefits to the central processing unit for performing data cleaning, filtering, amplification, feature calculation and correlation analysis on the real-time monitoring information before performing real-time monitoring information diagnosis. First, the data cleaning process is to remove the noise and abnormal data generated by the monitoring sensor or signal transmission process to ensure the data quality input into the subsequent processing flow. This step can significantly improve the accuracy of fault diagnosis, avoid misdiagnosis caused by noise or inaccurate data, and reduce the probability of false alarms and missed reports. Filtering and amplification processing improves the signal-to-noise ratio of the signal, making the monitoring information clearer and more reliable, especially in a noisy environment. This step effectively enhances the effective components of the signal, thereby improving the perception of the real fault state. Secondly, statistical feature calculation and correlation analysis extract the similarity and correlation of real-time monitoring information at different time points, which provides rich feature information for subsequent fault diagnosis. Through the analysis of these features, the system can identify the regular changes of the signal, capture potential fault precursors, and thus improve the sensitivity and predictive ability of fault diagnosis. Similarity and correlation analysis further reveals the inherent connection between different monitoring variables, helping to build a more accurate model that can perform a multi-dimensional and comprehensive assessment of changes in the ship's operating status. In addition, based on these similarities and correlations, the extracted characteristic parameters can better characterize the key state changes during the ship's operation. The generation of time domain feature maps visualizes these extracted characteristic parameters, facilitates a comprehensive and intuitive analysis of real-time monitoring information, and helps the system identify fault states and their development trends. This visual analysis not only helps the machine automatically identify faults, but also provides crew members with concise and clear fault information, improving decision-making efficiency and response speed.

[0066] Optionally, the central processing unit diagnoses the real-time monitoring information to determine the fault location and fault category of the ship, including: identifying and recording the fault location and fault category of different equipment on the ship; selecting a preset proportion of first real-time monitoring information from the real-time monitoring information, extracting time domain features and frequency domain features of the first real-time monitoring information, and obtaining a first signal feature vector; constructing a training set and a test set with the first signal feature vector as input and the fault location and fault category as output, and training a random forest model; performing feature extraction on the second real-time monitoring information that was not selected, and inputting the extracted second signal feature vector into the trained random forest model to obtain the fault location and fault category of the ship.

[0067] Specifically, the real-time monitoring information under abnormal conditions is mapped to the monitoring sensor positions corresponding to different equipment on the ship, and the equipment to which the real-time monitoring information under abnormal conditions belongs is identified. According to historical data or expert experience, different types of fault categories (such as motor overload, sensor failure, etc.) are preset, and the identified real-time monitoring information under abnormal conditions is classified into specific fault categories. Further, a preset proportion of first real-time monitoring information is selected from the real-time monitoring information, and time domain features and frequency domain features are extracted from the first real-time monitoring information, and the time domain features and frequency domain features are combined into a first signal feature vector. With the first signal feature vector as input, the fault location (such as equipment number) and fault category (such as motor overload, sensor failure, etc.) as output, a training set and a test set are constructed through the annotated historical data, the training set is used to train the random forest model, the test set is used to verify the performance of the random forest model, and the parameters are continuously adjusted to optimize the performance of the random forest model. Finally, the trained random forest model is used to perform fault diagnosis on the second real-time monitoring information that has not been selected, and the time domain features and frequency domain features are extracted for the input second real-time monitoring information to generate the corresponding second signal feature vector. The second signal feature vector is input into the trained random forest model, and the random forest model outputs the fault location and fault category.

[0068] In the intelligent detection method for ship performance status provided by this embodiment, in the first aspect, the range of the monitoring sensor is set according to the alarm threshold of the variable to be monitored, ensuring that the most suitable monitoring sensor can be used for each variable type. Specifically, by selecting appropriate ranges and monitoring sensors for different types of variables, it is possible to avoid excessive or insufficient data collection, ensure that the collected data information is accurate and effective, and reduce the transmission and storage of irrelevant data. In addition, the acquisition module collects data from different monitoring sensors in a patrol manner, but only uploads data when an abnormal signal occurs, which effectively reduces the amount of data transmission and avoids excessive waste of resources. Through multiple acquisition modules of the data acquisition and transmission unit, multiple acquisition modules collect real-time monitoring status in turn at different sampling frequencies, and upload relevant information to the central processing unit according to the abnormal status. Each acquisition module performs data acquisition according to different types of monitoring variables, and can obtain key information in real time without interfering with the normal operation of ship equipment. The central processing unit uses these data to quickly identify the fault location and category, thereby helping offshore workers to more quickly discover equipment failures and repair them in a timely manner. Through this intelligent and automated approach, the intensity of work is reduced and the response time is greatly shortened. On the second hand, the present application constructs a full-process, closed-loop intelligent detection system for ship performance status from real-time monitoring to data analysis to fault diagnosis. The present application is real-time, and multiple acquisition modules in the data acquisition and transmission unit collect the real-time monitoring status of each monitoring sensor in real time, and send the real-time monitoring information under abnormal conditions to the central processing unit. The present application is accurate, and the central processing unit performs diagnostic analysis on the real-time monitoring information under abnormal conditions to achieve rapid fault location and classification. Through the precise grasp of the ship's operating status, the present system provides a scientific basis for equipment maintenance, fault handling and operation optimization, and the closed-loop system effectively solves the defects of traditional monitoring methods with strong lag and diagnosis relying on manual experience, improves the safety and efficiency of ship operation, enables marine workers to find faulty equipment more quickly and repair it in time, and at the same time enhances the intelligence of my country's ship system. On the third hand, the training data set constructed by the present application based on historical monitoring data and real-time monitoring information can provide comprehensive data support for the fault diagnosis model. This enables the fault diagnosis model to not only diagnose based on current monitoring data, but also combine historical data to improve the recognition ability and accuracy of ship faults. In this way, the application can accumulate more fault samples, optimize its fault diagnosis capabilities, and make more accurate judgments when facing new fault conditions. Secondly, the application of the sparrow algorithm has significant advantages in model optimization. The sparrow algorithm can optimize the penalty parameters and kernel parameters in the multi-classification support vector machine model by simulating the group search mechanism, avoiding the limitations of traditional parameter adjustment methods.This adaptive optimization method improves the generalization ability of the multi-classification support vector machine model, so that the model can more accurately identify potential problems and perform classification diagnosis when facing complex and uncertain fault data. When optimizing parameters, the sparrow algorithm randomly selects parameter combinations and calculates training errors, and continuously adjusts until the preset conditions are met. This process avoids the tediousness and subjective factors of manual adjustment, reduces the risk of overfitting, and ensures the effectiveness of the model in different types of faults. In addition, retraining the model based on the optimized parameters can not only better adapt to different types of fault data, but also ensure the efficient processing of real-time monitoring information and output the fault location and category in real time. The advantage of this process is that it can respond to ship faults quickly and accurately locate different types of faults, reducing maintenance delay time. In addition, the multi-classification support vector machine model divides different fault states into different categories by establishing multiple decision boundaries in the classification task. In this way, ship faults can not only be accurately identified, but also effective maintenance and repair decisions can be made according to different categories of fault states. This provides crews with clear fault classification and location, helping them locate problems and take measures more quickly, avoiding maintenance costs and time waste due to misjudgment.

[0069] Corresponding to the aforementioned embodiment of a method for intelligently detecting the performance status of a ship, the present application also provides an embodiment of a system for intelligently detecting the performance status of a ship.

[0070] Figure 2 This is a schematic diagram of the structure of the intelligent detection system for ship performance status provided in Example 2 of this application. Figure 2 The system provided in this embodiment includes a data acquisition and transmission unit and a central processing unit connected to each other; wherein, The data acquisition and transmission unit is used to determine the variables to be monitored during the operation of the ship, and the variables to be monitored include multiple different variable types; The data acquisition and transmission unit is also used to determine the range of the monitoring sensor corresponding to each variable type according to the alarm threshold of the variable to be monitored; The data acquisition and transmission unit is also used to determine the monitoring position points of each variable to be monitored according to the spatial structure information of the ship; The data acquisition and transmission unit is further used to determine a plurality of monitoring sensors according to the ship spatial structure in the first neighborhood of the monitoring location point and the distance from the monitoring location point to the data acquisition and transmission unit, and install the monitoring sensors at the monitoring location point; The data acquisition and transmission unit is also used to match the corresponding relationship between the monitoring sensor and the acquisition module according to the monitoring location point and the variable type corresponding to each of the monitoring sensors, the acquisition modules are multiple, each acquisition module corresponds to multiple monitoring sensors, and the monitoring sensors corresponding to each acquisition module are completely different; The data acquisition and transmission unit is also used to acquire the real-time monitoring status of each monitoring sensor based on multiple acquisition modules, each acquisition module cyclically acquires the corresponding multiple monitoring sensors in sequence, and the sampling frequency of each acquisition module is different; The data acquisition and transmission unit is also used to identify an abnormal state in the real-time monitoring state, and send the real-time monitoring information in the abnormal state to the central processing unit, and the real-time monitoring signal in the abnormal state exceeds the range of the corresponding monitoring sensor; The central processing unit is used to diagnose the real-time monitoring information and determine the fault location and fault type of the ship.

[0071] Optionally, the data acquisition and transmission unit includes an analog multiplexing switch, a programmable gain amplifier and a digital-to-analog converter connected in sequence; wherein, The analog multiplexing switch is used to filter the normal state in the real-time monitoring state. When the normal state passes through the analog multiplexing switch, the analog multiplexing switch is closed and will not be transmitted to the central processing unit; when the abnormal state in the real-time monitoring state passes through the analog multiplexing switch, the analog multiplexing switch is opened and transmitted to the programmable gain amplifier; The programmable gain amplifier is used to adjust the amplitude based on the amplitude characteristics of the real-time monitoring information under the abnormal state, and the amplitude of the real-time monitoring information under the abnormal state after adjustment is adapted to the input range of the digital-to-analog converter; The digital-to-analog converter is used to convert the real-time monitoring information under abnormal conditions processed by the programmable gain amplifier into a digital signal.

[0072] Specifically, the data acquisition and transmission unit includes an analog multiplexing switch, a programmable gain amplifier, and a digital-to-analog converter, wherein the analog multiplexing switch is connected to the programmable gain amplifier, and the programmable gain amplifier is connected to the digital-to-analog converter.

[0073] Furthermore, the analog multiplexing switch selectively transmits the signal based on the signal state (normal state or abnormal state). When the input signal is judged to be a normal signal, the analog multiplexing switch remains in a closed state, and the normal signal will not continue to be transmitted backward. When the input signal is judged to be an abnormal signal, the analog multiplexing switch is opened, and the abnormal signal is transmitted to the subsequent modules (programmable gain amplifier, digital-to-analog converter) through the analog multiplexing switch for processing. The programmable gain amplifier is used to dynamically adjust the amplitude of the abnormal signal according to the amplitude characteristics of the abnormal signal so that it adapts to the input range of the digital-to-analog converter. By monitoring the original amplitude of the abnormal signal, if the original amplitude is too large (exceeding the input range of the digital-to-analog converter), the gain is reduced for compression. If the original amplitude is too small (lower than the resolution capability of the digital-to-analog converter), the gain is increased for amplification. The amplitude of the abnormal signal after adjustment remains within the optimal working range of the digital-to-analog converter. The digital-to-analog converter converts the abnormal signal (usually an analog signal) that has been gain-adjusted into a digital signal for subsequent analysis and processing by the central processing unit. The digital-to-analog converter receives the abnormal signal from the output of the programmable gain amplifier, processes the input signal to ensure that its amplitude is within its resolution range, samples the input signal at a fixed sampling frequency, and quantizes and maps the sampled analog signal amplitude into a digital code to obtain a digital signal.

[0074] In specific implementation, after abnormal detection of the real-time monitoring state, the real-time monitoring information under the normal state detected will not pass through the analog multiplexing switch, while the real-time monitoring information under the abnormal state detected will be transmitted to the programmable gain amplifier through the analog multiplexing switch. The programmable gain amplifier adjusts the amplitude of the real-time monitoring information under the abnormal state and transmits it to the digital-to-analog converter. The digital-to-analog converter converts the real-time monitoring information under the abnormal state into a digital signal and transmits it to the central processing unit through a communication interface (such as a serial port, CAN bus, Ethernet, etc.).

[0075] Furthermore, Figure 3 This is a schematic diagram of the structure of the intelligent detection system for ship performance status provided in Example 3 of this application. Figure 3 The system also includes an alarm unit, which is connected to the central processing unit and includes an audible and visual alarm and a radio communication system; wherein the audible and visual alarm is used to send out an alarm signal when an abnormal state is determined to exist, prompting offshore workers to deal with the fault in time; the radio communication system is used to transmit the fault location and fault type to a remote terminal device or a central control room, and the offshore workers promptly repair the faulty equipment based on the fault location and fault type.

[0076] For details, please refer to Figure 3The ship performance status intelligent detection system also includes an alarm unit, which is connected to the central processing unit and includes an audible and visual alarm and a radio communication system.

[0077] Furthermore, when the central processing unit detects an abnormal state and confirms the existence of a fault, the sound and light alarm is immediately triggered. The sound and light alarm emits a high-decibel alarm sound through a buzzer or speaker, and emits flashing or continuous light signals through high-brightness LED lights, strobe lights, etc. to attract the attention of offshore workers. The radio communication system packages the fault location and fault category into a digital format of a radio signal and sends it to the remote terminal device or central control room using a communication module (such as VHF, UHF, 4G / 5G module). The remote terminal device or central control room obtains information through a receiving device, and the fault information is displayed on a screen or graphical interface for staff to view and make decisions and perform corresponding maintenance.

[0078] The system of this embodiment can be used to perform Figure 1 The steps, specific implementation principles and implementation processes of the method embodiment shown are similar and will not be repeated here.

[0079] The implementation process of the functions and effects of each unit in the above system is specifically described in the implementation process of the corresponding steps in the above method, which will not be repeated here.

[0080] For the system embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The system embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present application. A person of ordinary skill in the art can understand and implement it without creative work.

[0081] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for intelligent detection of ship performance status, characterized in that: The method comprises: Determining variables to be monitored during the operation of the ship, wherein the variables to be monitored include a plurality of different variable types; Determine the range of the monitoring sensor corresponding to each variable type according to the alarm threshold of the variable to be monitored; Determining the monitoring location points of each variable to be monitored according to the spatial structure information of the ship; Determine a plurality of monitoring sensors according to the spatial structure of the ship in the first neighborhood of the monitoring location point and the distance between the monitoring location point and the data acquisition and transmission unit, and install the monitoring sensors at the monitoring location point; According to the monitoring location point and the variable type corresponding to each of the monitoring sensors, the corresponding relationship between the monitoring sensor and the acquisition module is matched, the acquisition module is multiple, each acquisition module corresponds to multiple monitoring sensors, and the monitoring sensors corresponding to each acquisition module are completely different; The multiple acquisition modules in the data acquisition and transmission unit acquire the real-time monitoring status of each monitoring sensor, each acquisition module cyclically acquires the corresponding multiple monitoring sensors in sequence, and the sampling frequency of each acquisition module is different; The data acquisition and transmission unit identifies an abnormal state in the real-time monitoring state, and sends the real-time monitoring information in the abnormal state to the central processing unit, wherein the real-time monitoring signal in the abnormal state exceeds the range of the corresponding monitoring sensor; The central processing unit diagnoses the real-time monitoring information to determine the fault location and fault type of the ship.

2. The method according to claim 1, characterized in that: There are N variable types of the variables to be monitored, and there are M monitoring components corresponding to the variables to be monitored. N and M are positive integers greater than or equal to 2, N>M, and the variable types collected by the monitoring sensors corresponding to each collection module are the same.

3. The method according to claim 1, characterized in that The step of determining the range of the monitoring sensor corresponding to each variable type according to the alarm threshold of the variable to be monitored includes: Analyze the fluctuation range of each variable to be monitored under normal operating conditions based on historical monitoring data, and determine the normal measurement value range of each variable to be monitored; Determine the floating range of each variable to be monitored based on external disturbance factors; The lower limit of the range of each variable to be monitored is determined based on the sum of the normal measurement value interval and the floating interval, and the upper limit of the range of each variable to be monitored is determined based on the normal measurement value interval, the floating interval and the alarm threshold of the variable to be monitored; wherein the alarm threshold is used to determine the critical value of the abnormal state outside the normal measurement value interval.

4. The method according to claim 1, characterized in that: Before the central processing unit diagnoses the real-time monitoring information, the method further includes: The central processing unit performs data cleaning on the real-time monitoring information; The real-time monitoring information is filtered and amplified in the time domain, and the processed real-time monitoring information has no noise and a higher signal-to-noise ratio than before the processing; Performing statistical feature calculation and correlation analysis on the real-time monitoring information after filtering and amplification processing, and extracting the similarity and correlation of the real-time monitoring information at different times; Based on the similarity and correlation, characteristic parameters in the ship operation process are obtained, and based on the characteristic parameters, a time domain characteristic graph of the real-time monitoring information is determined.

5. The method according to claim 1, characterized in that The central processing unit diagnoses the real-time monitoring information to determine the fault location and fault type of the ship, including: Constructing a training data set for a multi-classification support vector machine based on historical monitoring data and the real-time monitoring information; the training data set includes measurement values ​​of multiple variables to be monitored and corresponding fault locations and fault categories; Training a multi-classification support vector machine model based on the training data set, and optimizing the penalty parameters and kernel parameters in the model by using the sparrow algorithm; Wherein, a set of penalty parameters and kernel parameters are randomly selected for each sparrow, the training error of the initial model under each set of penalty parameters and kernel parameters is calculated, and the penalty parameters and kernel parameters are adjusted based on the training error until the training error is less than a preset error value or reaches a preset number of iterations; The model is retrained based on the optimized penalty parameters and kernel parameters, the real-time monitoring information is input into the trained model, and the corresponding fault location and fault category of the ship are output; wherein the model determines different fault states as different categories, establishes multiple decision boundaries, and each decision boundary determines the attribution of one category to all other categories.

6. The method according to claim 1, characterized in that The matching of the corresponding relationship between the monitoring sensor and the acquisition module according to the monitoring location point and the variable type corresponding to each of the monitoring sensors includes: Classifying the monitoring sensors according to the variable types corresponding to the monitoring sensors to obtain a plurality of first categories, wherein the variable types of the monitoring sensors in each of the first categories are the same, and different first categories correspond to different variable types; Determine the monitoring period for each first category according to the variable type; According to the monitoring period and the number of monitoring sensors in each first category, the distances between each monitoring sensor in each first category and the data acquisition and transmission unit are clustered, and each first category is divided based on the clustering results to obtain multiple second categories, and a corresponding relationship between each second category and each acquisition module is established.

7. The method according to claim 1, characterized in that The data acquisition and transmission unit identifies an abnormal state in the real-time monitoring state, including: The data acquisition and transmission unit performs local mean decomposition on the real-time monitoring state to obtain a local mean curve; Removing the local mean curve from the real-time monitoring state to obtain a detail curve; For each time point in the detail curve, an abnormal state in the real-time monitoring state is determined based on the difference between the real-time monitoring state corresponding to the time point and a preset threshold.

8. The method according to claim 1, characterized in that The central processing unit diagnoses the real-time monitoring information to determine the fault location and fault type of the ship, including: Identify and record the fault location and fault type of different equipment on board the ship; Selecting a preset proportion of first real-time monitoring information from the real-time monitoring information, extracting time domain features and frequency domain features of the first real-time monitoring information, and obtaining a first signal feature vector; Taking the first signal feature vector as input and the fault location and fault category as output, a training set and a test set are constructed to train a random forest model; Feature extraction is performed on the unselected second real-time monitoring information, and the extracted second signal feature vector is input into the trained random forest model to obtain the fault location and fault category of the ship.

9. An intelligent detection system for ship performance status, characterized in that: The system includes a data acquisition and transmission unit and a central processing unit connected to each other; wherein, The data acquisition and transmission unit is used to determine the variables to be monitored during the operation of the ship, and the variables to be monitored include multiple different variable types; The data acquisition and transmission unit is also used to determine the range of the monitoring sensor corresponding to each variable type according to the alarm threshold of the variable to be monitored; The data acquisition and transmission unit is also used to determine the monitoring position points of each variable to be monitored according to the spatial structure information of the ship; The data acquisition and transmission unit is further used to determine a plurality of monitoring sensors according to the ship spatial structure in the first neighborhood of the monitoring location point and the distance from the monitoring location point to the data acquisition and transmission unit, and install the monitoring sensors at the monitoring location point; The data acquisition and transmission unit is also used to match the corresponding relationship between the monitoring sensor and the acquisition module according to the monitoring location point and the variable type corresponding to each of the monitoring sensors, the acquisition modules are multiple, each acquisition module corresponds to multiple monitoring sensors, and the monitoring sensors corresponding to each acquisition module are completely different; The data acquisition and transmission unit is also used to acquire the real-time monitoring status of each monitoring sensor based on multiple acquisition modules, each acquisition module cyclically acquires the corresponding multiple monitoring sensors in sequence, and the sampling frequency of each acquisition module is different; The data acquisition and transmission unit is also used to identify an abnormal state in the real-time monitoring state, and send the real-time monitoring information in the abnormal state to the central processing unit, and the real-time monitoring signal in the abnormal state exceeds the range of the corresponding monitoring sensor; The central processing unit is used to diagnose the real-time monitoring information and determine the fault location and fault type of the ship.

10. The system according to claim 9, characterized in that The data acquisition and transmission unit includes an analog multiplexing switch, a programmable gain amplifier and a digital-to-analog converter connected in sequence; wherein, The analog multiplexing switch is used to filter the normal state in the real-time monitoring state. When the normal state passes through the analog multiplexing switch, the analog multiplexing switch is closed and will not be transmitted to the central processing unit; when the abnormal state in the real-time monitoring state passes through the analog multiplexing switch, the analog multiplexing switch is opened and transmitted to the programmable gain amplifier; The programmable gain amplifier is used to adjust the amplitude based on the amplitude characteristics of the real-time monitoring information under the abnormal state, and the amplitude of the real-time monitoring information under the abnormal state after adjustment is adapted to the input range of the digital-to-analog converter; The digital-to-analog converter is used to convert the real-time monitoring information under abnormal conditions processed by the programmable gain amplifier into a digital signal.

Images