Method and system for detecting occasional faults in ship main engine control system

By collecting and analyzing sensor data from the ship host control system online, using square prediction error and Hotling statistics combined with principal component analysis model, the problem of sensor accidental fault detection is solved, and the accurate fault detection of the ship host control system is achieved. Improves security and reduces costs.

CN112199781BActive Publication Date: 2024-11-26ZHENDUI IND ARTIFICIAL INTELLIGENCE CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202011176457.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2020-10-28
Publication Date
2024-11-26
Estimated Expiration
2040-10-28

AI Technical Summary

Technical Problem

There is a lack of accurate detection methods for the occasional failure of the ship host control system sensor in the prior art, resulting in fewer research on fault detection, affecting the safety and cost of ship operation.

Method used

By collecting sensor measurement samples online, calculating square prediction error SPE and Hotling T2 statistics, and comparing them with preset thresholds, combining the principal component analysis model, dynamic process data modeling is constructed to realize real-time monitoring and fault detection of sensor status .

Benefits of technology

It realizes accurate detection of the occasional failure of the ship's main engine control system sensor, detects problems early, improves ship operation safety and reduces operating costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN112199781B_ABST
    Figure CN112199781B_ABST
Patent Text Reader

Abstract

The present invention relates to a method and system for detecting sporadic faults in a ship main engine control system, which solves the technical problem that sporadic faults of sensors in existing ship main engine control systems are difficult to detect accurately. The method of the present invention comprises: online acquisition of measurement samples of sensors of a ship main engine control system; calculation of squared prediction error SPE and Hotelling T according to the measurement samples; 2 Statistics; Squared Prediction Error SPE and Hotelling T 2 The statistics are compared with the preset threshold value to realize the occasional fault detection function of the ship's main engine control system.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of ship fault detection, and in particular to a method and system for detecting sporadic faults of a ship main engine control system. Background Art

[0002] In recent years, with the rapid development of electronic and information technologies, a special type of fault, namely, sporadic fault, which is different from the traditional continuous fault form, has gradually attracted people's attention. Sporadic fault refers to a type of fault with a limited duration that can disappear on its own without external compensation measures, allowing the system to restore acceptable performance.

[0003] At present, fault detection of ship main engine control systems has become a hot research topic. However, for many years, people have mostly focused on the detection of persistent faults, while there has been little research on the detection of sporadic faults. The ship main engine control system is composed of complex electronic circuits. Welding and aging can cause the control circuit to loosen and cause sporadic controller failures. In addition, the operating environment of the ship main engine control system is complex, and the sensors are susceptible to vibration, electromagnetic interference, etc., which leads to sporadic failure of the sensors.

[0004] Therefore, the prior art lacks a method and system that can accurately detect occasional faults of sensors in a ship main engine control system. Summary of the invention

[0005] The present invention aims to provide a method and system for detecting sporadic faults in a ship main engine control system, so as to solve the problem of detecting sporadic faults in the existing ship main engine control system.

[0006] In order to achieve the above object, the embodiments of the present invention provide the following technical solutions:

[0007] In a first aspect of an embodiment of the present invention, a method for detecting sporadic faults of a ship main engine control system is provided, comprising:

[0008] Online collection of measurement samples of sensors of ship host control systems;

[0009] Calculate the squared prediction error SPE and Hotelling T based on the measurement samples 2 Statistics;

[0010] The squared prediction error SPE and Hotelling's T 2 The statistics are compared with the preset threshold to determine whether there are occasional faults in the ship's main engine control system.

[0011] Furthermore, the method for detecting sporadic faults in a ship main engine control system further includes a modeling step:

[0012] Collect the measurement samples x of the sensors of the ship's main engine control system under normal working conditions. The measurement samples at time k are recorded as x(k), where x(k) is an m-dimensional column vector, and m is the number of sensors. In the modeling process, a total of N groups of measurement samples at N times are collected;

[0013] The measurement samples under the normal working condition are subjected to time delay processing to construct an augmented measurement matrix:

[0014]

[0015] Where d is the delay sample length, x T (k) represents the transpose of the online collected measurement samples at time k, k = 1-d, 2-d, 3-d...N;

[0016] Each column of the amplification measurement matrix X is standardized to have zero mean and unit variance to obtain a measurement matrix.

[0017] Furthermore, the modeling step further includes: performing eigenvalue decomposition on the covariance matrix S of the measurement matrix, expressed as:

[0018]

[0019] in is a diagonal matrix containing the m×(d+1) eigenvalues ​​of the covariance matrix S, arranged in descending order. is the corresponding eigenvector matrix.

[0020] Furthermore, the modeling step further includes: according to the diagonal matrix and the eigenvector matrix The number of principal components l is determined according to the cumulative variance contribution rate criterion, where the diagonal matrix And the eigenvector matrix P is further expressed as:

[0021]

[0022] Where P is a [m×(d+1)]×l-order matrix, which is the load matrix, given by The first l columns of is the non-principal component matrix, λ j yes The j-th diagonal element j=1,2,3...m×(d+1), Λ=diag{λ1,…,λ l} means that the diagonal elements are λ1,…,λ l The diagonal matrix of Indicates that the diagonal elements are λ l+1 ,…,λ m(d+1) The diagonal matrix of ;

[0023] The cumulative variance contribution rate criterion is to limit the sum of the variance contributions of the first l principal components to be greater than a preset value so that the selected principal components effectively represent the original data.

[0024] Furthermore, the modeling step further comprises: according to the diagonal matrix The diagonal elements of and a given confidence level α determine the squared prediction error SPE and Hotelling T 2 Threshold value of the statistic:

[0025]

[0026]

[0027] Among them, J th,SPE is the threshold of the squared prediction error SPE statistic, Hotelling T 2 The threshold value of the statistic; α is the confidence level; c α is the threshold value when the confidence level α is set for the standard normal distribution, F α Represents the threshold value at Fisher distribution confidence level α; It is the i-th power of the j-th diagonal element, j=1,2,3...m×(d+1), i=0,1.

[0028] Furthermore, in the method for detecting sporadic faults in a ship main engine control system, the squared prediction error SPE and Hotelling T 2 The statistical expression is:

[0029]

[0030] T 2 = z(k) T PΛ -1 P T z(k)

[0031] Among them, z(k) is the online measurement sample at time k, z(k) T is the transpose of the online measurement sample at time k, P is a [m×(d+1)]×l-order matrix, is the load matrix, is the part of P except the first l columns, Λ -1 is the inverse matrix of Λ.

[0032] Furthermore, the squared prediction error SPE and Hotelling T 2 The steps of comparing the statistic with the preset threshold value to determine whether there is an occasional fault in the ship main engine control system include:

[0033] The squared prediction error SPE and Hotelling T calculated online 2The statistic is compared with the preset threshold. If SPE≤J th,SPE and The ship's main engine control system has no accidental failures, J th,SPE is the threshold of the squared prediction error SPE statistic, Hotelling T 2 The threshold value of the statistic.

[0034] In another aspect of the embodiments of the present invention, a system for detecting sporadic faults of a ship main engine control system is provided, comprising:

[0035] Online data acquisition module, used to collect online measurement samples of sensors of the ship's main engine control system;

[0036] A data calculation module is used to calculate the square prediction error SPE and Hotelling T according to the measurement samples. 2 Statistics;

[0037] Fault detection module, used to convert the squared prediction error SPE and Hotelling T 2 The statistics are compared with the preset threshold to determine whether there are occasional faults in the ship's main engine control system.

[0038] Furthermore, the ship main engine control system occasional fault detection system also includes a modeling module, including:

[0039] A data acquisition unit, used to collect measurement samples x of sensors of the ship's main engine control system under normal working conditions;

[0040] Principal Component Analysis unit for:

[0041] The measurement samples collected by the data acquisition module under normal working conditions are subjected to time delay processing according to a certain time delay d to construct an augmented measurement matrix X;

[0042] Standardizing each column of the amplified measurement matrix X to zero mean and unit variance to obtain a measurement matrix;

[0043] Perform eigenvalue decomposition on the covariance matrix S of the measurement matrix to obtain a diagonal matrix and the eigenvector matrix

[0044] According to the cumulative variance contribution rate criterion, the number of principal components l is determined, and the diagonal matrix is ​​obtained Loading matrix P and non-principal component matrix

[0045] A threshold determination unit is used to determine the threshold value according to the diagonal matrix The diagonal elements of and a given confidence level α determine the squared prediction error SPE and Hotelling T 2The threshold value of the statistic J th,SPE and

[0046] Furthermore, the data calculation module is also used to calculate the squared prediction error SPE and Hotelling T according to the following formula: 2 Statistics:

[0047]

[0048] T 2 = z(k) T PΛ -1 P T z(k)

[0049] Among them, z(k) is the online measurement sample at time k, z(k) T is the transpose of the online measurement sample, P is a [m×(d+1)]×l-order matrix, is the load matrix, for Except for the first l columns, Λ -1 is the inverse matrix of Λ;

[0050] The fault detection module is specifically used to: calculate the square prediction error SPE and Hotelling T 2 The statistic is compared with the preset threshold. If SPE≤J th,SPE and The ship main engine control system has no accidental failure; among them, J th,SPE is the threshold of the squared prediction error SPE statistic, Hotelling T 2 The threshold value of the statistic.

[0051] Beneficial effects of the present invention:

[0052] Using the observation samples under normal working conditions as the basis, by extending the principal component analysis model, a part of the time-delay samples are included in the data matrix, and a dynamic principal component analysis model is constructed to model the dynamic process data related to the time series. The obtained model is then used to "evaluate" the measurement samples collected online, check whether there are abnormalities in the measurement samples (abnormality means that there are obvious differences from the observation samples under normal working conditions), and monitor the sensor status online in real time without interrupting the normal operation of the sensor. It can effectively realize the accurate detection of occasional faults of sensors in the ship's main engine control system, discover problems early, and improve the safety of ship operation and reduce operating costs.

[0053] In the present invention, the above-mentioned technical solutions can also be combined with each other to achieve more preferred combination solutions. Other features and advantages of the present invention will be described in the subsequent description, and some advantages can become obvious from the description, or can be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained through the contents particularly pointed out in the description and the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0055] Figure 1 This is a flow chart of a method for detecting sporadic faults in a ship host control system according to an embodiment of the present application;

[0056] Figure 2 This is a flow chart of a principal component analysis method in a method for detecting sporadic faults in a ship main engine control system according to an embodiment of the present application;

[0057] Figure 3 This is a schematic structural diagram of an occasional fault detection device for a ship main engine control system according to another embodiment of the present application;

[0058] Figure 4 A schematic diagram of a modeling structure of a device for detecting sporadic faults of a ship main engine control system shown in another embodiment of the present application;

[0059] Figure 5 Hotelling T is another embodiment of the present application showing the occasional fault detection function of the ship main engine control system 2 Statistics simulation data graph;

[0060] Figure 6 This is a simulation data diagram of the square prediction error SPE statistic of the sporadic fault detection function of the ship main engine control system shown in another embodiment of the present application;

[0061] Figure 7 A schematic diagram of the hardware structure of an electronic device for executing the method for detecting sporadic faults in a ship main engine control system provided in an embodiment of the invention of this application. DETAILED DESCRIPTION

[0062] The preferred embodiments of the present invention are described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not used to limit the scope of the present invention.

[0063] like Figure 1 As shown, a specific embodiment of the present invention discloses a method for detecting sporadic faults of a ship main engine control system, comprising:

[0064] S100, online collection of measurement samples of sensors of the ship host control system;

[0065] Specifically, the ship main engine control system is the core of ship control as a control system that directly operates the main engine. Whether the main engine can operate normally and provide power to the ship is the determining factor for whether the ship can complete the navigation mission normally. The purpose of the ship main engine control system is to ensure that the main engine works safely, reliably and effectively, and can promptly eliminate the fault and send out fault alarm signals and display the fault content when a fault occurs. The main engine is operated, controlled and monitored through various instruments, alarms and control devices. As a device belonging to the ship main engine control system, the sensor provides the ship main engine control system with various important information for controlling the ship, such as various power parameter information, operating parameter information, environmental parameter information, etc.

[0066] In this embodiment, measurement samples under normal working conditions are collected as training data to generate the parameters required by the algorithm and form a model. The obtained model is then used to "evaluate" the measurement samples collected online to check whether there are abnormalities in the measurement samples. The abnormality in this embodiment refers to obvious differences from the observation samples under normal working conditions. The measurement samples under normal working conditions and the measurement samples collected online refer to the readings of sensors arranged in the ship's main engine control system, including but not limited to temperature, pressure, vibration and other types of data.

[0067] Specifically, the measurement samples x of the ship main engine control system sensors under normal working conditions are first collected. The measurement samples at time k are recorded as x(k), where x(k) is an m-dimensional column vector, and m is the number of sensors. A total of N groups of measurement samples at N times are collected during the modeling process.

[0068] Furthermore, the measurement samples z(k) of the sensors of the ship main engine control system are collected online to determine whether an occasional fault occurs in the sensor at the current moment.

[0069] S200, calculating the squared prediction error SPE and Hotelling T according to the measurement samples 2 Statistics;

[0070] Specifically, Figure 2 As shown, this step can obtain the number of principal components through the following sub-steps, and further calculate the squared prediction error SPE and Hotelling T 2 Statistics and their thresholds:

[0071] Specifically, the measurement samples under normal working conditions are subjected to time delay processing to construct an augmented measurement matrix:

[0072]

[0073] Where d is the delay sample length, x T (k) represents the transpose of the online collected measurement samples at time k, k = 1-d, 2-d, 3-d...N;

[0074] Preferably, the time delay sample length d is determined according to the following criteria: by continuously increasing d, the correlation between the sample with a time delay of d and the original sample is calculated. If the correlation no longer changes significantly with the increase of d, that is, it tends to be stable, then d at this time is selected as the time delay sample length. Optionally, it is defined that as d increases by 1, the correlation increases by no more than 0.1%, which is considered to be stable. Optionally, the data x in the measurement matrix is ​​expanded. T The time k in (k) can be a negative number, that is, the observation sample is a measurement sample at a past time.

[0075] Further, each column of the amplification measurement matrix X is standardized to zero mean unit variance to obtain a measurement matrix. Specifically, the zero mean unit variance refers to obtaining data with zero mean unit variance by calculating the mean and variance of all samples, then subtracting the mean from the sample, and dividing by the variance. Through the above-mentioned standardization process of this embodiment, the data is standardized to facilitate subsequent calculations.

[0076] Furthermore, the covariance matrix S of the measurement matrix is ​​subjected to eigenvalue decomposition, which is expressed as:

[0077]

[0078] in is a diagonal matrix containing the m×(d+1) eigenvalues ​​of the covariance matrix S, arranged in descending order. is the corresponding eigenvector matrix.

[0079] Further, according to the diagonal matrix and the eigenvector matrix Determine the number of principal components l, where the diagonal matrix and the eigenvector matrix Further expressed as:

[0080]

[0081] Where P is a [m×(d+1)]×l row matrix, which is the load matrix, given by The first l columns of is the non-principal component matrix, λ j yes The j-th diagonal element j=1,2,3...m×(d+1), Λ=diag{λ1,…,λ l} means that the diagonal elements are λ1,…,λ l The diagonal matrix of Indicates that the diagonal elements are λ l+1 ,…,λ m(d+1) The diagonal matrix of ;

[0082] Preferably, by limiting the sum of the variance contributions of the first l principal components to be greater than a preset value, the selected principal component effectively represents the original data. Specifically, the eigenvalue of the matrix corresponding to the principal component represents the variance contribution of the principal component. By limiting the sum of the variance contributions of the first l principal components to be greater than a certain threshold, it can be ensured that the selected principal component can effectively represent the original data. Preferably, the threshold is selected to be 95%.

[0083] Specifically, the squared prediction error SPE and Hotelling T 2 The statistical expression is:

[0084]

[0085] T 2 = z(k) T PΛ -1 P T z(k)

[0086] Among them, z(k) is k Measure samples online at all times, z(k) T for k The transpose of the online measurement samples at all times, P is a [m×(d+1)]×l-order matrix, is the load matrix, for Except for the first l columns, Λ -1 is the inverse matrix of Λ.

[0087] S300, squared prediction error SPE and Hotelling T 2 The statistics are compared with the preset threshold value to determine whether there is an occasional fault in the ship's main engine control system;

[0088] Specifically, according to the diagonal matrix The diagonal elements of and a given confidence level α determine the squared prediction error SPE and Hotelling T 2 Threshold value of the statistic:

[0089]

[0090]

[0091] Among them, J th,SPEis the threshold of the squared prediction error SPE statistic, Hotelling T 2 The threshold value of the statistic; α is the confidence level. The higher the error requirement, the higher α is. Optionally, α is 95%; c α is the threshold value when the confidence level α is set for the standard normal distribution, F α Represents the threshold value at Fisher distribution confidence level α; It is the i-th power of the j-th diagonal element, j=1,2,3...m×(d+1), i=0,1.

[0092] Furthermore, the squared prediction error SPE and Hotelling T calculated online are 2 The statistic is compared with the preset threshold. If SPE≤J th,SPE and The ship's main engine control system has no accidental failures, J th,SPE is the threshold of the squared prediction error SPE statistic, Hotelling T 2 The threshold value of the statistic.

[0093] More preferably, the maximum interval between the online measurement samples of the sensors of the ship's main engine control system and the measurement samples under normal operating conditions within a maintenance cycle is 30 days. If the interval between the two exceeds 30 days or after maintenance, the measurement samples under normal operating conditions must be re-collected. Because if the time interval is too long or after maintenance, the external environment may have changed significantly, and it is impossible to determine whether the difference between normal data and online measurement data is due to a fault in the sensor itself or the external environment.

[0094] The method for detecting sporadic faults of a ship main engine control system provided by an embodiment of the present invention realizes accurate detection of sporadic faults of sensors of a ship main engine control system, and discovers problems early to improve the safety of ship operation and reduce operation costs.

[0095] See also Figure 3 Another embodiment of the present application provides a system for detecting occasional faults of a ship main engine control system, including an online data acquisition module 101, a data calculation module 201, and a fault detection module 301.

[0096] In this embodiment, the online data acquisition module 101 is used to collect online measurement samples z(k) of sensors of the ship host control system, which are used as input data for calculation and judgment of whether an occasional fault occurs in the sensor at the current moment.

[0097] like Figure 4 As shown, this embodiment also includes a modeling module. In order to explain the present invention more clearly, the following is combined with Figure 4Each part in this embodiment is described in detail.

[0098] In this embodiment, the data calculation module 201 is used to calculate the square prediction error SPE and Hotelling T based on the data obtained by the principal component analysis unit and the data collected by the online data collection module. 2 Statistics;

[0099] The data acquisition unit 400 is used to collect measurement samples x of sensors of the ship host control system under normal working conditions;

[0100] The principal component analysis unit 500 performs time delay processing on the measurement samples under normal working conditions collected by the data collection module according to a certain time delay d to construct an augmented measurement matrix X;

[0101] Further, each column of the amplified measurement matrix X is standardized to zero mean and unit variance to obtain a measurement matrix;

[0102] Furthermore, the covariance matrix S of the measurement matrix is ​​subjected to eigenvalue decomposition to obtain a diagonal matrix and the eigenvector matrix

[0103] Furthermore, according to the cumulative variance contribution rate criterion, the number of principal components l is determined, and the diagonal matrix is ​​obtained Loading matrix P and non-principal component matrix

[0104] Furthermore, the data calculation module 201 is also used to calculate the squared prediction error SPE and Hotelling T according to the following formula: 2 Statistics:

[0105]

[0106] T 2 = z(k) T PΛ -1 P T z(k)

[0107] Among them, z(k) is the online measurement sample at time k, z(k) T is the transpose of the online measurement sample, P is a [m×(d+1)]×l-order matrix, is the load matrix, for Except for the first l columns, Λ -1 is the inverse matrix of Λ;

[0108] In this embodiment, the fault detection module 301 is used to convert the square prediction error SPE and Hotelling T 2 The statistics are compared with the preset threshold to determine whether there are occasional faults in the ship's main engine control system.

[0109] Specifically, the threshold determination unit 600 is used to determine the threshold value according to the diagonal matrix The diagonal elements of and a given confidence level α determine the squared prediction error SPE and Hotelling T 2 The threshold value of the statistic J th,SPE and

[0110] Furthermore, the squared prediction error SPE and Hotelling T calculated online are 2 The statistic is compared with the preset threshold. If SPE≤J th,SPE and The ship main engine control system has no accidental failure; among them, J th,SPE is the threshold of the squared prediction error SPE statistic, Hotelling T 2 The threshold value of the statistic.

[0111] like Figure 5 As shown, the horizontal axis data represents the sampling time k, and the vertical axis solid line data represents Hotelling T 2 Statistical value, the dotted line represents Hotelling T 2 The threshold of the statistic When Hotelling T 2 If the statistic exceeds the threshold value, the data at that moment is considered abnormal.

[0112] like Figure 6 As shown, the horizontal axis data represents the sampling time k, the vertical axis solid line data represents the square prediction error SPE statistic, and the dotted line represents the threshold J of the square prediction error SPE statistic. th,SPE , when the squared prediction error SPE statistic exceeds the threshold data, the data at that moment is considered abnormal.

[0113] Compared with the prior art, the method and system for detecting sporadic faults of the ship main engine control system proposed in the present invention firstly statistically models the data features with dynamic changes under normal working conditions through training data, analyzes the sporadic fault forms of the ship main engine control system, and obtains relevant fault parameters. Then, by comparing the dynamic features of the new measurement samples with the normal data, the detection statistics are calculated, the number of dynamic principal components is selected, and the threshold is set. Finally, fault detection is performed in the online stage to detect problems early and ensure that the main engine works safely, reliably and effectively, so as to improve the safety of ship operation and reduce the operation cost.

[0114] See also Figure 7 Another embodiment of the present invention further provides an electronic device for executing the method for detecting occasional faults in the above embodiment. The electronic device includes:

[0115] One or more processors 710 and memory 720, Figure 7A processor 710 is taken as an example.

[0116] The electronic device for executing the method for detecting sporadic faults of a ship main engine control system may further include: an input device 730 and an output device 740 .

[0117] The processor 710, the memory 720, the input device 730 and the output device 740 may be connected via a bus or other means. Figure 7 The example of connecting through bus is taken in the following.

[0118] The memory 720, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer executable programs and modules, such as program instructions / modules (units) corresponding to the method for detecting sporadic faults of a ship host control system in an embodiment of the present invention. The processor 710 executes various functional applications and data processing of the server by running the non-volatile software programs, instructions and modules stored in the memory 720, that is, realizing the icon display method of the above method embodiment.

[0119] The memory 720 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and applications required for at least one function; the data storage area may store information on the number of reminder items of the acquired application, etc. In addition, the memory 720 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 720 may optionally include a memory remotely disposed relative to the processor 710, and these remote memories may be connected to a processing device for list item operations via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0120] The input device 730 can receive input digital or character information, and generate key signal input related to user settings and function control of the sporadic fault detection device of the ship main engine control system. The output device 740 can include display devices such as a display screen.

[0121] The one or more modules are stored in the memory 720, and when executed by the one or more processors 710, the ship main engine control system occasional fault detection method in any of the above method embodiments is executed.

[0122] The above-mentioned product can execute the method provided by the embodiment of the present invention, and has the functional modules and beneficial effects corresponding to the execution method. For technical details not fully described in this embodiment, please refer to the method provided by the embodiment of the present invention.

[0123] The electronic device of the embodiment of the present invention may exist in various forms, including but not limited to:

[0124] (1) Mobile communication devices: These devices are characterized by their mobile communication functions and their main purpose is to provide voice and data communications. These terminals include: smart phones (such as iPhone), multimedia phones, functional phones, and low-end phones.

[0125] (2) Ultra-mobile personal computer devices: These devices fall into the category of personal computers, have computing and processing capabilities, and generally also have mobile Internet access features. These terminals include: PDA, MID and UMPC devices, such as iPad.

[0126] (3) Portable entertainment devices: These devices can display and play multimedia content. They include audio and video players (such as iPods), handheld game consoles, e-books, as well as smart toys and portable car navigation devices.

[0127] (4) Server: A device that provides computing services. The server consists of a processor, hard disk, memory, system bus, etc. The server is similar to a general computer architecture, but because it needs to provide highly reliable services, it has higher requirements in terms of processing power, stability, reliability, security, scalability, and manageability.

[0128] (5) Other electronic devices with reminder recording function.

[0129] The device embodiments described above are merely illustrative, wherein the units (modules) described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, i.e., they may be located in one place or distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment.

[0130] An embodiment of the present invention provides a non-transitory computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are executed by an electronic device, the ship main engine control system sporadic fault detection method in any of the above-mentioned method embodiments is executed on the electronic device.

[0131] An embodiment of the present invention provides a computer program product, wherein the computer program product includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions, wherein when the program instructions are executed by an electronic device, the electronic device executes the method for detecting sporadic faults of a ship main engine control system in any of the above-mentioned method embodiments.

[0132] Through the description of the above implementation methods, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiment.

[0133] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the principles or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for detecting sporadic faults in a ship main engine control system, characterized in that: include: Online collection of measurement samples of sensors of ship host control systems; Calculate the squared prediction error SPE and Hotelling T based on the measurement samples 2 Statistics, including modeling steps: Collect measurement samples x of the sensors of the ship main engine control system under normal working conditions, and the measurement samples at time k are recorded as x(k), where x(k) is an m-dimensional column vector, and m is the number of sensors. The modeling process collects N groups of measurement samples at N times, wherein the maximum interval between the online measurement samples of the sensors of the ship main engine control system and the measurement samples x under normal working conditions within a maintenance cycle is 30 days; The measurement samples under the normal working condition are subjected to time delay processing to construct an augmented measurement matrix: Where d is the delay sample length, x T (k) represents the transpose of the online collected measurement sample at time k, k = 1-d, 2-d, 3-d, ... N, wherein the time delay sample length d is determined according to the following criteria: By continuously increasing d, the correlation between the sample with a time delay of d and the original sample is calculated. If the correlation no longer changes significantly with the increase of d, that is, it tends to be stable, then d at this time is selected as the time delay sample length, where, The stabilization is defined as the correlation increasing by no more than 0.1% with each increase of d; Each column of the amplification measurement matrix X is standardized to zero mean and unit variance to obtain a measurement matrix, where: The zero mean and unit variance refers to obtaining data with zero mean and unit variance by calculating the mean and variance of all samples, subtracting the mean from the sample, and dividing by the variance; The covariance matrix S of the measurement matrix is ​​subjected to eigenvalue decomposition, which is expressed as: in is a diagonal matrix containing the m×(d+1) eigenvalues ​​of the covariance matrix S, arranged in descending order. is the corresponding eigenvector matrix; According to the diagonal matrix and the eigenvector matrix The number of principal components l is determined according to the cumulative variance contribution rate criterion, where the diagonal matrix and the eigenvector matrix Further expressed as: Where P is a [m×(d+1)]×l-order matrix, which is the load matrix, given by The first l columns of is the non-principal component matrix, λ j yes The j-th diagonal element j=1,2,3...m×(d+1), Λ=diag{λ1,...,λ l } means that the diagonal elements are λ1,…,λ l The diagonal matrix of Indicates that the diagonal elements are λ l+1 , …,λ m(d+1) The diagonal matrix of ; The cumulative variance contribution rate criterion is to limit the sum of the variance contributions of the first l principal components to be greater than a preset value so that the selected principal component effectively represents the original data; The squared prediction error SPE and Hotelling's T 2 The statistical expression is: T 2 =z(k) T PΛ -1 P T z(k) Among them, z(k) is the online measurement sample at time k, z(k) T is the transpose of the online measurement sample at time k, P is a [m×(d+1)]×l-order matrix, is the load matrix, for Except for the first l columns, Λ -1 is the inverse matrix of Λ; According to the diagonal matrix The diagonal elements of and a given confidence level α determine the squared prediction error SPE and Hotelling T 2 Threshold value of the statistic: Among them, J th,SPE is the threshold of the squared prediction error SPE statistic, Hotelling T 2 The threshold value of the statistic; α is the confidence level; c α is the threshold value when the confidence level α is set for the standard normal distribution, F α Represents the threshold value at Fisher distribution confidence level α; is the i-th power of the j-th diagonal element, j=1,2,3...m×(d+1), i=0,1; The squared prediction error SPE and Hotelling's T 2 The statistics are compared with the preset threshold to determine whether there are occasional faults in the ship's main engine control system, including: The squared prediction error SPE and Hotelling T calculated online 2 The statistic is compared with the preset threshold. If SPE≤J th,SPE and The ship's main engine control system has no accidental failures, J th,SPE is the threshold of the squared prediction error SPE statistic, Hotelling T 2 The threshold value of the statistic.

2. A system for detecting occasional faults in a ship main engine control system, characterized in that: include: Online data acquisition module, used to collect online measurement samples of sensors of the ship's main engine control system; A data calculation module is used to calculate the square prediction error SPE and Hotelling T according to the measurement samples and the following formula: 2 Statistics, the following formula is specifically expressed as: T 2 =z(k) T PΛ -1 P T z(k) Among them, z(k) is the online measurement sample at time k, z(k) T is the transpose of the online measurement sample, P is a [m×(d+1)]×l-order matrix, is the load matrix, for Except for the first l columns, Λ -1 is the inverse matrix of Λ; Fault detection module, used to convert the squared prediction error SPE and Hotelling T 2 The statistics are compared with the preset threshold value to determine whether there is an occasional fault in the ship's main engine control system; The fault detection module is specifically used to: calculate the square prediction error SPE and Hotelling T 2 The statistic is compared with the preset threshold. If SPE≤J th,SPE and The ship main engine control system has no accidental failure; among them, J th,SPE is the threshold of the squared prediction error SPE statistic, Hotelling T 2 The threshold of the statistic is expressed as the following expression In the formula, α is the confidence level; c α is the threshold value when the confidence level α is set for the standard normal distribution, F α Represents the threshold value at Fisher distribution confidence level α; is the i-th power of the j-th diagonal element, j=1,2,3...m×(d+1), i=0,1; The fault detection system also includes a modeling module, including: A data acquisition unit is used to collect measurement samples x of sensors of the ship main engine control system under normal working conditions, the measurement samples at time k are recorded as x(k), x(k) is an m-dimensional column vector, m is the number of sensors, and a total of N groups of measurement samples at N times are collected during the modeling process, wherein the maximum interval between the online measurement samples of the sensors of the ship main engine control system and the measurement samples x under normal working conditions within a maintenance cycle is 30 days; Principal Component Analysis unit for: The measurement samples collected by the data acquisition module under normal working conditions are processed with a certain delay d to construct an augmented measurement matrix X, which is expressed as: Where d is the delay sample length, x T (k) represents the transpose of the online collected measurement samples at time k, k = 1-d, 2-d, 3-d, ... N, The time delay d is determined according to the following criteria: By continuously increasing d, the correlation between the sample with a time delay of d and the original sample is calculated. If the correlation no longer changes significantly with the increase of d, that is, it tends to be stable, then d at this time is selected as the time delay sample length, where, The stabilization is defined as the correlation increasing by no more than 0.1% with each increase of d; The columns of the amplified measurement matrix X are standardized to have a zero mean and unit variance to obtain a measurement matrix, wherein the zero mean and unit variance refer to obtaining data with a zero mean and unit variance by calculating the mean and variance of all samples, then subtracting the mean from the sample and dividing by the variance; Perform eigenvalue decomposition on the covariance matrix S of the measurement matrix to obtain a diagonal matrix and the eigenvector matrix in, The covariance matrix S of the measurement matrix is ​​subjected to eigenvalue decomposition, which is expressed as: The diagonal matrix and the eigenvector matrix Further expressed as: Where P is a [m×(d+1)]×l-order matrix, which is the load matrix, given by The first l columns of is the non-principal component matrix, λ j yes The j-th diagonal element of j=1,2,3...m×(d+1), Λ=diag{λ1,…,λ l } means the diagonal elements are λ1,…λ l The diagonal matrix of Indicates that the diagonal elements are λ l+1 ,…λ m(d+1) The diagonal matrix of ; According to the cumulative variance contribution rate criterion, the number of principal components l is determined, and the diagonal matrix is ​​obtained Loading matrix P and non-principal component matrix A threshold determination unit is used to determine the threshold value according to the diagonal matrix The diagonal elements of and a given confidence level α determine the squared prediction error SPE and Hotelling T 2 The threshold value of the statistic J th,SPE and J th,T2 .