CN112836381A - Multi-source information-based ship residual life prediction method and system - Google Patents

Multi-source information-based ship residual life prediction method and system Download PDF

Info

Publication number
CN112836381A
CN112836381A CN202110188897.5A CN202110188897A CN112836381A CN 112836381 A CN112836381 A CN 112836381A CN 202110188897 A CN202110188897 A CN 202110188897A CN 112836381 A CN112836381 A CN 112836381A
Authority
CN
China
Prior art keywords
sampling
state
time
value
moment
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110188897.5A
Other languages
Chinese (zh)
Other versions
CN112836381B (en
Inventor
魏慕恒
邱伯华
张羽
刘学良
李永杰
朱慧敏
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhendui Industrial Intelligent Technology Co ltd
Original Assignee
Zhendui Industrial Intelligent Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Zhendui Industrial Intelligent Technology Co ltd filed Critical Zhendui Industrial Intelligent Technology Co ltd
Priority to CN202110188897.5A priority Critical patent/CN112836381B/en
Publication of CN112836381A publication Critical patent/CN112836381A/en
Application granted granted Critical
Publication of CN112836381B publication Critical patent/CN112836381B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G06F30/15—Vehicle, aircraft or watercraft design
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/08—Thermal analysis or thermal optimisation

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Geometry (AREA)
  • General Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Automation & Control Theory (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Testing And Monitoring For Control Systems (AREA)

Abstract

The invention relates to a ship residual life prediction method and system based on multi-source information, and solves the problems of low reliability and poor prediction precision of the existing ship residual life prediction method. The method of the invention comprises the following steps: acquiring a system output signal vector of a ship sensor measurement sample, determining a system initial state and initial parameters, and calculating the system state according to the system initial state and the initial parameters; calculating the optimal state estimation value of the sampling time and each past sampling time according to the system output signal vector and the system state; estimating the state value of each future sampling moment according to the sampling moment and the state optimal estimation value of each past sampling moment; and comparing the state value of each future sampling moment with a threshold value to predict the residual life of the ship.

Description

Multi-source information-based ship residual life prediction method and system
Technical Field
The invention relates to the technical field of ship life prediction, in particular to a ship residual life prediction method based on multi-source information.
Background
In recent years, with the rapid development of technologies such as electronics and information, the prediction of the remaining service life is one of the important links of prediction and health management as the basis for the establishment of an equipment maintenance strategy. The method has the advantages that the residual service life can be accurately predicted, comprehensive, accurate and effective information can be provided for the formulation of the equipment maintenance strategy, the equipment failure is avoided, the loss caused by the failure is reduced, and the safe and reliable operation of the equipment is guaranteed.
The ship industry has a dynamic system with a plurality of independent and irrelevant sensors, and the system has an implicit performance degradation process, but the implicit performance degradation process is lacked in the residual life prediction process of the ship, so that the life prediction error is large, the reliability is low and the prediction precision is poor. Therefore, a method and a system for predicting the residual life of a ship based on multi-source information are lacked in the prior art.
Disclosure of Invention
In view of the foregoing analysis, embodiments of the present invention provide a method for predicting a remaining life of a ship based on multi-source information, so as to solve the problems of low reliability and poor prediction accuracy of the existing method for predicting a remaining life of a ship.
In one aspect, an embodiment of the present invention provides a method for predicting a remaining life of a ship based on multi-source information, including:
collecting a system output signal vector of a ship sensor measurement sample, and determining an initial state and initial parameters of the system;
calculating the sampling time and the state optimal estimation value of each past sampling time according to the system output signal vector, the system initial state and the initial parameters;
estimating the state value of each future sampling moment according to the sampling moment and the state optimal estimation value of each past sampling moment;
and comparing the state value of each future sampling moment with a threshold value to predict the residual life of the ship.
Further, the system state is xk，xkIs the system state value at the sampling instant k, x0In order to be in the initial state of the system,
Figure BDA0002944451650000021
as an initial parameter, the parameter is,
Figure BDA0002944451650000022
calculating parameters at the sampling time of k;
obtaining a system linear equation and an observation equation expression according to the system state and the system output signal vector:
xk+1＝xk+ηkτk+∈k (1)
yk＝ak+hkxk+ωk (2)
wherein, the formula (1) is a system linear equation with linear performance degradation, and the formula (2) is an observation equation; x is the number ofk+1For the predicted value of the system state at the sampling time k +1, ykA system output signal vector at the sampling moment of k is obtained, wherein k is more than or equal to 0, and t is the current moment;
Figure BDA0002944451650000023
the calculation parameters for the k sampling time comprise: etakFor the drift velocity of the linear degradation process, taukIs a sampling interval, ekIs noise, akFor sensor null shift, hkAs a system self-parameter, ωkTo measure noise.
Further, the calculating the best state estimation value of the sampling time and each past sampling time comprises:
calculating a system state estimation value at a sampling moment k according to a state optimal estimation value of a measurement sample at the sampling moment k-1 and a system output signal vector at the sampling moment k, wherein t is more than or equal to k and more than or equal to 0, and t is the current moment;
step two, according to the estimated value of the system state at the k sampling time and the parameter updated for the ith time at the k sampling time
Figure BDA0002944451650000031
The optimized state values at past sampling instants are updated,
Figure BDA0002944451650000032
for optimum parameters at the last moment
Figure BDA0002944451650000033
Step three, calculating a joint expectation according to the joint distribution of the system output signal vector and the optimized state value, maximizing the expectation, and obtaining the (l + 1) th updated parameter at the k sampling moment
Figure BDA0002944451650000034
Step four, if the updated parameters
Figure BDA0002944451650000035
If the requirements are met, calculating the state optimal estimation value at the k sampling moment; if the updated parameter does not meet the requirement, adding 1 to the l, and then repeatedly executing the steps from two to four until the state optimal estimation value of the k sampling time is calculated;
and circularly executing the steps from one to four until the state optimal estimation value of the current sampling time and each past sampling time is calculated.
Further, the estimated value of the system state at the k sampling time
Figure BDA0002944451650000036
Expressed as:
Figure BDA0002944451650000037
Pk|k-1＝Pk-1|k-1+Qk-1
Kk＝Pk|k-1CT(CPk|k-1CT+R)-1
Figure BDA0002944451650000038
Pk|k＝(I-KkC)Pk|k-1
wherein t is more than or equal to k and more than or equal to 0, t is the current moment,
Figure BDA0002944451650000039
to estimate the state at the k sample time based on the k-1 sample time,
Figure BDA00029444516500000310
is the best estimate of the system state at the time of k sampling, ηk-1For the drift velocity, tau, of the linear degradation process at the sampling instant k-1k-1For the sampling interval of k-1 sampling instants, Pk|k-1Predicting a variance matrix, P, for Kalman filtering from k-1 sample instantsk-1|k-1Variance matrix, Q, estimated for the k-1 th sampling instant of Kalman filteringk-1System noise at the sampling instant K-1, KkFor the kalman gain at the time of the k samples,
Figure BDA00029444516500000311
is the state estimate at k sampling instants, C is the measurement matrix, CTFor transposing the measurement matrix C, R is the measurement noise, ykAnd I is a system output signal vector at the sampling moment of k, and is an identity matrix.
Further, the optimized state value of the past sampling time is updated, and is expressed as:
Figure BDA0002944451650000041
Figure BDA0002944451650000042
Figure BDA0002944451650000043
wherein k is sampling time, j is past sampling time, t is current time, j is more than or equal to 0 and less than or equal to k and less than or equal to t; l isjIs an intermediate variable, Pj|jThe variance matrix estimated for the kalman filter j sample time,
Figure BDA0002944451650000044
the inverse of the variance matrix at sample time j +1 is predicted from sample time j for kalman filtering,
Figure BDA0002944451650000045
to optimize the value for the state at the j sample time based on the k sample time,
Figure BDA0002944451650000046
is the estimated value of the system state at the j sample time,
Figure BDA0002944451650000047
linear degradation process drift speed, tau, for the i-th update of the k sampling instantsjFor a sampling interval, Pj|kTo estimate the estimation error of the j-sample time state with k-sample times,
Figure BDA0002944451650000048
is an intermediate variable LjThe transposing of (1).
Further, the calculation is combined with expectation, expressed as:
Figure BDA0002944451650000049
wherein t is more than or equal to k and more than or equal to 0, t is the current moment,
Figure BDA00029444516500000416
a likelihood function for k sampling instants; gamma raykFor combined variables of state and observations at and all times before the sampling time k in the first update, i.e.
Figure BDA00029444516500000410
Theta is an independent variable, and theta which is expected to maximize k sampling time is
Figure BDA00029444516500000411
Figure BDA00029444516500000412
For the occurrence under theta condition in the first update
Figure BDA00029444516500000413
The probability of (a) of (b) being,
Figure BDA00029444516500000414
for the combination of the state variables at the sampling instant k and all preceding instants in the first update,
Figure BDA00029444516500000415
for the combination of the observed variables at the k sample time and all previous times in the first update,
Figure BDA0002944451650000051
is the sum of theta in the first update
Figure BDA0002944451650000052
Appear under the condition of
Figure BDA0002944451650000053
The probability of (a) of (b) being,
Figure BDA0002944451650000054
is the sum of theta in the first update
Figure BDA0002944451650000055
Under the condition of dischargingNow that
Figure BDA0002944451650000056
The probability of (a) of (b) being,
Figure BDA0002944451650000057
for the system state vector optimization value of the i-th update at the time of the t-sample,
Figure BDA0002944451650000058
for the system state vector optimization value of the i-th update at the time of the t-1 sample,
Figure BDA0002944451650000059
the system output signal value for the first update at the time of t samples.
Further, said maximizing said expectation, expressed as:
Figure BDA00029444516500000510
wherein t is more than or equal to k and more than or equal to 0, t is the current moment,
Figure BDA00029444516500000511
for the occurrence under theta condition in the first update
Figure BDA00029444516500000512
The probability of (a) of (b) being,
Figure BDA00029444516500000513
for the combination of the state variables at the sampling instant k and all preceding instants in the first update,
Figure BDA00029444516500000514
the combination of the observation variables at the sampling moment k and all previous moments in the first updating;
the obtained theta satisfying the condition is the updated parameter
Figure BDA00029444516500000515
The method comprises the following steps:
Figure BDA00029444516500000516
Figure BDA00029444516500000517
Figure BDA00029444516500000518
Figure BDA00029444516500000519
wherein k is sampling time, j is past sampling time, t is current time, j is more than or equal to 0 and less than or equal to k and less than or equal to t,
Figure BDA00029444516500000520
linear degradation process drift velocity, τ, for the i +1 th update of the k sampling instantsj-1The sampling interval at the sampling instant j-1,
Figure BDA00029444516500000521
the j-time system state for the i-th update of the k-sample time,
Figure BDA00029444516500000522
the system noise for the (l + 1) th update at the k sampling instant,
Figure BDA0002944451650000061
the estimate of the l +1 st update at the k time for the measurement coefficient corresponding to the ith sensor in the measurement matrix,
Figure BDA0002944451650000062
for the system output signal component corresponding to the moment of the ith sensor j,
Figure BDA0002944451650000063
for the null shift of the i-th sensor,
Figure BDA0002944451650000064
the updated estimate at time k, l +1, for the ith row in the measurement error matrix.
Further, the air conditioner is provided with a fan,
if the updated parameter
Figure BDA0002944451650000065
The parameters are converged and the parameters are,
Figure BDA0002944451650000066
calculating the optimal parameter of the k sampling time according to the optimal parameter to obtain the optimal estimated value of the system state of the k sampling time
Figure BDA0002944451650000067
If the updated parameter
Figure BDA0002944451650000068
Adding 1 to the l, and then repeating the steps two to four until the state optimal estimation value at the sampling moment k is calculated.
Further, the estimating the state value at each future sampling instant further comprises: and estimating the state value of each future sampling moment according to the optimal parameter and the optimal estimated value of the system state, wherein the expression is as follows:
xT＝xT-1+ηtτT-1
wherein T is more than T, T is future sampling time, T is current time, xTFor state estimates at future times, xT-1Initial value istBest estimate of the system state at the sampling instant, ηtFor optimum linear degradation process drift velocity at the moment of t-samplingT-1A sampling interval at time T-1;
setting a State threshold XthComparing the future sampling time state value with a state threshold value, if xT≤XthPredicting the T sampling time systemThe system fails.
On the other hand, the embodiment of the invention provides a ship residual life prediction system based on multi-source information, which comprises the following steps:
the data acquisition module is used for acquiring system output signal vectors of ship sensor measurement samples and determining the initial state and initial parameters of the system;
the system state optimization module is used for calculating the sampling time and the state optimal estimation value of each past sampling time according to the system output signal vector, the system initial state and the initial parameters;
the system state estimation module is used for estimating a state value of each future sampling moment according to the sampling moment and the state optimal estimation value of each past sampling moment;
and the life prediction module is used for comparing the state value at each future sampling moment with a threshold value so as to predict the residual life of the ship.
Compared with the prior art, the invention can at least realize the following beneficial effects:
the method comprises the steps that firstly, multi-group sensor measurement data are collected for predicting the residual life of a ship based on multi-source information, the multi-source property of the measurement data is guaranteed, an implicit performance degradation process of infinite separability is introduced into a calculation process, and a state estimation value of the next moment closest to a true value can be estimated by utilizing a state estimation value of the previous moment and a performance linear degradation process; optimizing each past state value according to the state estimation value closest to the true value to obtain the state optimal estimation value at each acquisition moment; estimating a state value at each future sampling moment according to the obtained state optimal estimation value, and then comparing the state value with a threshold value to predict the residual life of the ship; as the sampling point approaches the time of failure, the condition reliability prediction decays faster, and as the system time of failure approaches, the system condition reliability tends to zero.
In the invention, the technical schemes can be combined with each other to realize more preferable combination schemes. Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and drawings.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below. It should be apparent that the drawings in the following description are merely exemplary, and that other embodiments can be derived from the drawings provided by those of ordinary skill in the art without inventive effort.
Fig. 1 is a flowchart of a method for predicting the remaining life of a ship based on multi-source information according to an embodiment of the present application;
FIG. 2 is a graph illustrating a comparison of an actual degraded path and an estimated degraded path according to an embodiment of the present application;
FIG. 3 is a flowchart illustrating a method for calculating a best estimate of the state for each past sampling instant according to one embodiment of the present application;
FIG. 4 is a diagram illustrating a comparison of a true value of a parameter η and an estimated value of the parameter η according to an embodiment of the present application;
FIG. 5 shows a comparison of a true value of a parameter Q and an estimated value of the parameter Q according to an embodiment of the present application;
FIG. 6 shows a parameter h according to an embodiment of the present application1The true value of (1) and the parameter estimation value;
FIG. 7 shows a parameter h according to an embodiment of the present application2The true value of (1) and the parameter estimation value;
FIG. 8 shows a parameter h according to an embodiment of the present application3The true value of (1) and the parameter estimation value;
FIG. 9 shows a parameter R according to an embodiment of the present application1The true value of (1) and the parameter estimation value;
FIG. 10 shows a parameter R according to an embodiment of the present application2The true value of (1) and the parameter estimation value;
FIG. 11 shows a schematic view of the present applicationParameter R shown in the examples3The true value of (1) and the parameter estimation value;
FIG. 12 illustrates the results of a conditional reliability prediction that begins 180 steps in advance, according to one embodiment of the present application;
FIG. 13 shows the results of conditional reliability prediction starting 100 steps ahead, according to one embodiment of the present application;
FIG. 14 is a graph of a probability function of remaining useful life at each sampling time according to an embodiment of the present application;
FIG. 15 is a schematic structural diagram of a system for predicting the remaining life of a ship based on multi-source information according to another embodiment of the present application;
fig. 16 is a schematic hardware structure diagram of an electronic device for executing the method for predicting the remaining life of a ship based on multi-source information according to the embodiment of the present invention.
Detailed Description
The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate preferred embodiments of the invention and together with the description, serve to explain the principles of the invention and not to limit the scope of the invention.
As shown in fig. 1, a specific embodiment of the present invention discloses a method for predicting remaining life of a ship based on multi-source information, which includes:
s10, collecting a system output signal vector of a ship sensor measurement sample, and determining an initial state and initial parameters of the system;
specifically, as shown in fig. 2, a thin solid line is a ship actual degradation state path, a thick solid line is a ship estimated degradation state path, and the actual degradation state path of a ship device includes a degradation process, so in order to estimate the remaining life of the device, a device state degradation model needs to be established, a performance degradation process is implicitly included in a dynamic system of a plurality of independent and unrelated sensors existing in the ship industry, the degradation process often has a random characteristic, and the degradation process can only be obtained in an indirect manner through the plurality of sensors.
In this embodiment, the system state x of the ship sensor measurement sample is collectedk，xkIs a k miningDetermining the initial state x of the system according to the system state at the initial sampling time0Specifically, the system initial state value may be given based on the observed state value, or may be artificially given empirically, and as k increases, xkThe values may be closer to the actual vessel degradation state path,
Figure BDA0002944451650000101
for the calculated parameters at the time of the k samples,
Figure BDA0002944451650000102
the initial parameters can be given artificially; system output signal vector y for collecting ship sensor measurement samplek. Preferably, the different system sensors of the vessel are different and not limited to a particular type of sensor; for example, the engine system sensor may be selected from a vibration sensor, a temperature sensor, and specifically, such as: x is the number ofkRepresents the diesel generator temperature;
Figure BDA0002944451650000103
representing the diesel generator three-phase winding temperature (i is 3, representing the i-th sensor, and sequentially u-phase, v-phase, and w-phase), is a column vector of 3 x 1.
In particular, according to the system state xkAnd the system output signal vector ykObtaining a system linear equation and an observation equation expression:
xk+1＝xk+ηkτk+∈k (1)
yk＝ak+hkxk+ωk (2)
wherein, the formula (1) is a system linear equation with linear performance degradation, and the formula (2) is an observation equation; x is the number ofk+1The predicted value of the system state at the sampling moment of k +1 is obtained, t is more than or equal to k and more than or equal to 0, and t is the current moment;
Figure BDA0002944451650000104
calculating parameters for k sampling instantsThe method comprises the following steps: etakFor the drift velocity of the linear degradation process, taukIs a sampling interval, ekIs noise, akFor sensor null shift, hkAs a system self-parameter, ωkTo measure noise. Preferably, the sampling interval τkThe intervals may be unequal, for example: the sensor transmits data back at the beginning every second, and then transmits back once every five seconds;
Figure BDA0002944451650000105
initial parameters are given for the initial time, the initial parameters are generally given by experience, and the optimal parameters at a certain time are obtained through continuous optimization in subsequent calculation;
specifically, in a system linear equation with linear degradation of performance, a degradation state estimation and parameter identification algorithm is designed by using an implicit performance degradation process, so that the residual life contained in the degradation process is effectively predicted, and the life prediction result is closer to the real condition.
S20, calculating the best state estimation value of the sampling time and each past sampling time according to the system output signal vector, the system initial state and the initial parameters;
specifically, as shown in fig. 3, the best estimation value of the system state at the sampling time of step k can be obtained through the following sub-steps:
specifically, S201, calculating a system state estimated value at a sampling moment k according to a state optimal estimated value of a measurement sample at the sampling moment k-1 and a system output signal vector at the sampling moment k, wherein t is more than or equal to k and is more than or equal to 0, and t is the current moment;
the system state estimation value at the k sampling moment
Figure BDA0002944451650000111
Expressed as:
Figure BDA0002944451650000112
Pk|k-1＝Pk-1|k-1+Qk-1 (4)
Kk＝Pk|k-1CT(CPk|k-1CT+R)-1 (5)
Figure BDA0002944451650000113
Pk|k＝(I-KkC)Pk|k-1 (7)
wherein the content of the first and second substances,
Figure BDA0002944451650000114
to estimate the state at the k sample time based on the k-1 sample time,
Figure BDA0002944451650000115
is the best estimate of the system state at the time of k sampling, ηk-1For the drift velocity, tau, of the linear degradation process at the sampling instant k-1k-1For the sampling interval of k-1 sampling instants, Pk|k-1Predicting a variance matrix, P, for Kalman filtering from k-1 sample instantsk-1|k-1Variance matrix, Q, estimated for the k-1 th sampling instant of Kalman filteringk-1System noise at the sampling instant K-1, KkFor the kalman gain at the time of the k samples,
Figure BDA0002944451650000116
is the state estimate at k sampling instants, C is the measurement matrix, CTFor transposing the measurement matrix C, R is the measurement noise, ykThe system output signal vector at the sampling moment k is equal to or more than 0.
Specifically, formula (3)
Figure BDA0002944451650000117
Best estimation value of system state from k-1 sampling time
Figure BDA0002944451650000118
Calculated to obtain the system state estimation value of formula (6)
Figure BDA0002944451650000119
And the observed value y of the sensorkAnd the best estimation value of the system state based on the last moment
Figure BDA00029444516500001110
Correlation, I is an identity matrix;
when the method is implemented, the system state quantity interfered by noise is a random quantity, an accurate value cannot be measured, the system state quantity is predicted through a statistical viewpoint, and the estimated value of the system state is made to be as close to a true value as possible accurately according to a known observed value and the best estimated value of the previous system state, so that the method is beneficial to improving the accuracy of life prediction data of ship equipment.
S202, according to the estimated value of the system state at the k sampling moment and the parameter updated for the ith time at the k sampling moment
Figure BDA0002944451650000121
The optimized state values at past sampling instants are updated,
Figure BDA0002944451650000122
for optimum parameters at the last moment
Figure BDA0002944451650000123
Updating the optimized state value at the past sampling moment, and expressing as:
Figure BDA0002944451650000124
Figure BDA0002944451650000125
Figure BDA0002944451650000126
wherein k is sampling time, j is past sampling time, t is current time, j is more than or equal to 0 and less than or equal to k and less than or equal to t; l isjIs composed ofIntermediate variable, Pj|jThe variance matrix estimated for the kalman filter j sample time,
Figure BDA0002944451650000127
the inverse of the variance matrix at sample time j +1 is predicted from sample time j for kalman filtering,
Figure BDA0002944451650000128
to optimize the value for the state at the j sample time based on the k sample time,
Figure BDA0002944451650000129
is the estimated value of the system state at the j sample time,
Figure BDA00029444516500001210
linear degradation process drift speed, tau, for the i-th update of the k sampling instantsjA sampling interval of j sampling instants, Pj|kTo estimate the estimation error of the j-sample time state with k-sample times,
Figure BDA00029444516500001211
is an intermediate variable LjThe transposing of (1).
Specifically, the system state estimated value of k sampling time is obtained
Figure BDA00029444516500001212
The state of each past sampling moment is optimized and updated, the state value of each past sampling moment is optimized and updated in a reverse mode while the number of sampling data is increased, and the result accuracy can be further improved.
S203, calculating a joint expectation according to the joint distribution of the system output signal vector and the past time system state optimization value based on the k sampling time, maximizing the expectation value, and obtaining the (l + 1) th updated parameter of the k sampling time
Figure BDA00029444516500001213
The calculation is combined with expectation, expressed as:
Figure BDA0002944451650000131
wherein t is more than or equal to k and more than or equal to 0, t is the current moment,
Figure BDA00029444516500001323
a likelihood function for k sampling instants; gamma raykFor combined variables of state and observations at and all times before the sampling time k in the first update, i.e.
Figure BDA0002944451650000132
Theta is an independent variable, and theta which is expected to maximize k sampling time is
Figure BDA0002944451650000133
For the occurrence under theta condition in the first update
Figure BDA0002944451650000134
The probability of (a) of (b) being,
Figure BDA0002944451650000135
for the combination of the state variables at the sampling instant k and all preceding instants in the first update,
Figure BDA0002944451650000136
for the combination of the observed variables at the k sample time and all previous times in the first update,
Figure BDA0002944451650000137
is the sum of theta in the first update
Figure BDA0002944451650000138
Appear under the condition of
Figure BDA0002944451650000139
The probability of (a) of (b) being,
Figure BDA00029444516500001310
is the first timeIn new at θ and
Figure BDA00029444516500001311
appear under the condition of
Figure BDA00029444516500001312
The probability of (a) of (b) being,
Figure BDA00029444516500001313
for the system state vector optimization value of the i-th update at the time of the t-sample,
Figure BDA00029444516500001314
for the system state vector optimization value of the i-th update at the time of the t-1 sample,
Figure BDA00029444516500001315
the system output signal value for the first update at the time of t samples.
The maximum expectation, expressed as:
Figure BDA00029444516500001316
wherein t is more than or equal to k and more than or equal to 0, t is the current moment,
Figure BDA00029444516500001317
for the occurrence under theta condition in the first update
Figure BDA00029444516500001318
The probability of (a) of (b) being,
Figure BDA00029444516500001319
for the combination of the state variables at the sampling instant k and all preceding instants in the first update,
Figure BDA00029444516500001320
the combination of the observation variables at the sampling moment k and all previous moments in the first updating;
theta satisfying the condition is obtained, namelyUpdating parameters
Figure BDA00029444516500001321
The method comprises the following steps:
Figure BDA00029444516500001322
Figure BDA0002944451650000141
Figure BDA0002944451650000142
Figure BDA0002944451650000143
wherein j is more than or equal to 0 and less than or equal to k and less than or equal to t, k is sampling time, j is past sampling time, t is current time,
Figure BDA0002944451650000144
linear degradation process drift velocity, τ, for the i +1 th update of the k sampling instantsj-1The sampling interval at the sampling instant j-1,
Figure BDA0002944451650000145
the j-time system state for the i-th update of the k-sample time,
Figure BDA0002944451650000146
the system noise for the (l + 1) th update at the k sampling instant,
Figure BDA0002944451650000147
the estimate of the l +1 st update at the k time for the measurement coefficient corresponding to the ith sensor in the measurement matrix,
Figure BDA0002944451650000148
for the ith sensor at time jThe corresponding system output signal component(s) is (are),
Figure BDA0002944451650000149
for the null shift of the i-th sensor,
Figure BDA00029444516500001410
the updated estimate at time k, l +1, for the ith row in the measurement error matrix.
Specifically, as long as the amount of sensor information increases, the convergence speed of the above-described parameter update also increases.
S204, if the updated parameters
Figure BDA00029444516500001411
If the requirements are met, calculating the state optimal estimation value at the k sampling moment; if the updated parameter does not meet the requirement, adding 1 to the l, and then repeatedly executing the steps from two to four until the state optimal estimation value of the k sampling time is calculated;
in particular, if the parameters are updated
Figure BDA00029444516500001412
Satisfy the requirement of
Figure BDA00029444516500001413
The parameters are converged and the parameters are,
Figure BDA00029444516500001414
calculating the optimal state estimation value of the k sampling time according to the optimal parameter of the k sampling time;
if the updated parameter
Figure BDA00029444516500001415
The parameter is not the optimal parameter, 1 is continuously added to the l, and then S201 is repeated until the parameter converges.
Figure BDA00029444516500001416
Comprises a plurality of parameters, and when the convergence of the parameters is judged, all the parameters are required to be simultaneously usedConvergence is satisfied with a convergence value ζ > 0, ζ being dependent on the specific data, generally defined empirically.
In particular, in xkRepresents the diesel generator temperature;
Figure BDA0002944451650000151
representing the temperature of a three-phase winding of the diesel generator (i is 3 and represents the ith sensor, and the i phase, the v phase and the w phase are respectively and sequentially used) as the updated parameters of the specific embodiment
Figure BDA0002944451650000152
As shown in fig. 4 to 11, the comparison result of the parameter true value and the parameter estimated value is given, and after a non-stationary process, all unknown parameters: eta, Q, h1、h2、h3、R1、R2、R3The estimated values of (c) eventually converge to the true values.
In particular, when the parameter is
Figure BDA0002944451650000153
The convergence of the signals is carried out,
Figure BDA0002944451650000154
and for the optimal parameter of the k sampling time, after the optimal state estimation values of the k sampling time and the past sampling time are obtained through calculation according to the optimal parameter, circularly executing the steps from S201 to S204 until the optimal state estimation values of the t sampling time and the past sampling time are calculated.
S30, estimating the state value of each future sampling time according to the sampling time and the state optimal estimation value of each past sampling time;
the estimating the state value at each future sampling instant further comprises: and estimating the state value of each future sampling moment according to the optimal parameter and the optimal estimated value of the system state, wherein the expression is as follows:
xT＝xT-1+ηtτT-1
wherein T is more than T, T is future sampling time, T is current time, xTFor the state estimate at a future time instant,xT-1the initial value is the optimal estimated value of the system state, eta, obtained after the parameter convergence at the sampling time of ttFor optimum linear degradation process drift velocity at the moment of t-samplingT-1A sampling interval at time T-1;
specifically, in comparison with formula (1), the state value expression at each future sampling time is estimated to contain neither ekBecause the noise ekIs unpredictable, and does not take into account noise, η, when predicting future state valuestBased on the optimal linear degradation process drift velocity at the time of t sampling (the time of starting prediction), tauT-1In order that the sampling interval is a known quantity, it is preferably determined whether the sampling interval at different times changes according to actual requirements.
Specifically, as shown in fig. 12 and 13, a system initial state x is given0Initial parameters are given as 34.4 ℃ respectively: eta0＝0.08，Q＝0，a＝[3.78，-5.44，-0.13]，h＝[1,1，1]，R＝[0.5,0.5，0.5]The failure threshold is: phi-N (48,0.01), the failure time is at the 191 th sampling moment, the prediction step number delta is 180 from the 10 th sampling moment in the prediction of FIG. 12, the prediction step number delta is 100 from the 91 th sampling moment in the prediction of FIG. 13, and as can be seen by comparing FIG. 12 with FIG. 13, when the sampling point approaches the failure time, the prediction attenuation of the condition reliability is accelerated, the accuracy of the prediction curve is improved along with the increase of the sampling data at the sampling moment, and the condition reliability tends to zero along with the approach of the system failure event, and the prediction method can predict the possible failure event before the system failure occurs. As shown in fig. 14, at the same sampling time, the multi-source information fusion prediction result based on multiple sensors is more accurate than that of a single sensor.
And S40, comparing the state value at each future sampling moment with a threshold value to predict the remaining life of the ship.
Setting a State threshold XthComparing the future sampling time state value with a state threshold value, if xT≤XthAnd predicting that the system fails at the T sampling moment.
In particular, the state thresholdXthDepending on the specific data, it is generally defined empirically.
As shown in fig. 15, another embodiment of the present application provides a system for predicting remaining life of a ship based on multi-source information, including: the system comprises a data acquisition module 10, a system state optimization module 20, a system state estimation module 30 and a life prediction module 40.
The data acquisition module 10 is used for acquiring system output signal vectors of ship sensor measurement samples and determining the initial state and initial parameters of the system; specifically, the system state is xk，xkIs the system state value at the sampling instant k, x0In order to be in the initial state of the system,
Figure BDA0002944451650000161
calculating parameters at the sampling time of k; obtaining a system linear equation and an observation equation expression according to the system state and the system output signal vector:
xk+1＝xk+ηkτk+∈k (1)
yk＝ak+hkxk+ωk (2)
wherein, the formula (1) is a system linear equation with linear performance degradation, and the formula (2) is an observation equation; x is the number ofk+1For the predicted value of the system state at the sampling time k +1, ykA system output signal vector at the sampling moment of k is obtained, wherein k is more than or equal to 0, and t is the current moment;
Figure BDA0002944451650000171
the calculation parameters for the k sampling time comprise: etakFor the drift velocity of the linear degradation process, taukIs a sampling interval, ekIs noise, akFor sensor null shift, hkAs a system self-parameter, ωkTo measure noise.
A system state optimization module 20, configured to calculate a sampling time and a state optimal estimation value at each past sampling time according to the system output signal vector, a system initial state and an initial parameter; the method comprises the following steps:
the system state estimation unit is used for calculating the system state estimation value at the sampling moment k according to the state optimal estimation value of the measurement sample at the sampling moment k-1 and the system output signal vector at the sampling moment k, wherein t is more than or equal to k and is more than or equal to 0, and t is the current moment;
a parameter updating unit for updating the parameter according to the estimated value of the system state at the k sampling time and the updated parameter at the l-th time of the k sampling time
Figure BDA0002944451650000172
The optimized state values at past sampling instants are updated,
Figure BDA0002944451650000173
for optimum parameters at the last moment
Figure BDA0002944451650000174
According to the estimated value of the system state at the k sampling moment and the parameter updated for the I time at the k sampling moment
Figure BDA0002944451650000175
The optimized state values at past sampling instants are updated,
Figure BDA0002944451650000176
for optimum parameters at the last moment
Figure BDA0002944451650000177
A parameter confirmation unit for calculating a joint expectation according to the joint distribution of the system output signal vector and the optimized state value, maximizing the expectation, and obtaining the updated parameter of the (l + 1) th time at the sampling time k
Figure BDA0002944451650000181
An optimum state calculation unit for calculating optimum state of the object according to the parameters satisfying the requirement
Figure BDA0002944451650000182
Calculating k miningThe best estimate of the state at the sample time.
The system state estimation module 30 is configured to estimate a state value at each future sampling time according to the sampling time and the state optimal estimation value at each past sampling time; specifically, according to the optimal parameter and the state optimization value at the past sampling time, the expression is:
xT＝xT-1+ηtτT-1
wherein T is more than T, T is future sampling time, T is current time, xTFor state estimates at future times, xT-1The optimal estimation value of the system state is obtained after the parameter convergence at the sampling time with the initial value of t
Figure BDA0002944451650000183
ηtFor optimum linear degradation process drift velocity at the moment of t-samplingT-1Is the sampling interval at time T-1.
The life prediction module 40 is used for comparing the state value at each future sampling moment with a threshold value so as to predict the residual life of the ship; specifically, a state threshold value X is setthComparing the future sampling time state value with a state threshold value, if xT≤XthAnd predicting that the system fails at the T sampling moment.
Referring to fig. 16, another embodiment of the present invention further provides an electronic device for implementing the method for predicting the remaining life of a ship based on multi-source information in the foregoing embodiment. The electronic device includes:
one or more processors 710 and a memory 720, one processor 710 being illustrated in fig. 16.
The electronic device for executing the multi-source information-based ship remaining life prediction method may further include: an input device 730 and an output device 740.
The processor 710, the memory 720, the input device 730, and the output device 740 may be connected by a bus or other means, such as the bus connection in fig. 16.
The memory 720, which is a non-volatile computer-readable storage medium, may be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as program instructions/modules (units) corresponding to the multi-source information-based ship remaining life prediction method in the embodiment of the present invention. The processor 710 executes various functional applications of the server and data processing by running nonvolatile software programs, instructions and modules stored in the memory 720, that is, implements the icon display method of the above-described method embodiment.
The memory 720 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store information on the number of acquired reminders for the application program, and the like. Further, the memory 720 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some embodiments, memory 720 may optionally include memory located remotely from processor 710, which may be connected over a network to a processing device operating the list items. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The input device 730 may receive input numeric or character information and generate key signal inputs related to user settings and function control of the ship remaining life prediction device based on multi-source information. The output device 740 may include a display device such as a display screen.
The one or more modules are stored in the memory 720 and when executed by the one or more processors 710, perform a method for predicting remaining life of a ship based on multi-source information in any of the method embodiments described above.
The product can execute the method provided by the embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method. For technical details that are not described in detail in this embodiment, reference may be made to the method provided by the embodiment of the present invention.
The electronic device of embodiments of the present invention may exist in a variety of forms, including but not limited to:
(1) a mobile communication device: such devices are characterized by mobile communications capabilities and are primarily targeted at providing voice, data communications. Such terminals include: smart phones (e.g., iphones), multimedia phones, functional phones, and low-end phones, among others.
(2) Ultra mobile personal computer device: the equipment belongs to the category of personal computers, has calculation and processing functions and generally has the characteristic of mobile internet access. Such terminals include: PDA, MID, and UMPC devices, etc., such as ipads.
(3) A portable entertainment device: such devices can display and play multimedia content. Such devices include audio and video players (e.g., ipods), handheld game consoles, electronic books, as well as smart toys and portable car navigation devices.
(4) A server: the device for providing the computing service comprises a processor, a hard disk, a memory, a system bus and the like, and the server is similar to a general computer architecture, but has higher requirements on processing capacity, stability, reliability, safety, expandability, manageability and the like because of the need of providing high-reliability service.
(5) Other electronic devices with reminding item recording function.
The above-described embodiments of the apparatus are merely illustrative, and the units (modules) described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment.
The embodiment of the invention provides a non-transitory computer-readable storage medium, which stores computer-executable instructions, wherein when the computer-executable instructions are executed by an electronic device, the electronic device is caused to execute the multi-source information-based ship remaining life prediction method in any method embodiment.
Embodiments of the present invention provide a computer program product, where the computer program product includes a computer program stored on a non-transitory computer-readable storage medium, where the computer program includes program instructions, where the program instructions, when executed by an electronic device, cause the electronic device to execute a multi-source information-based ship remaining life prediction method in any of the above-mentioned method embodiments.
Through the above description of the embodiments, those skilled in the art will clearly understand that the embodiments may be implemented by software plus a necessary general hardware platform, and may also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (10)

1. A ship residual life prediction method based on multi-source information is characterized by comprising the following steps:
collecting a system output signal vector of a ship sensor measurement sample, and determining an initial state and initial parameters of the system;
calculating the sampling time and the state optimal estimation value of each past sampling time according to the system output signal vector, the system initial state and the initial parameters;
estimating the state value of each future sampling moment according to the sampling moment and the state optimal estimation value of each past sampling moment;
and comparing the state value of each future sampling moment with a threshold value to predict the residual life of the ship.
2. The multi-source information-based ship remaining life prediction method according to claim 1, wherein the system state is xk，xkIs the system state value at the sampling instant k, x0In order to be in the initial state of the system,
Figure FDA0002944451640000011
as an initial parameter, the parameter is,
Figure FDA0002944451640000012
calculating parameters at the sampling time of k;
obtaining a system linear equation and an observation equation expression according to the system state and the system output signal vector:
xk+1＝xk+ηkτk+∈k (1)
yk＝ak+hkxk+ωk (2)
wherein, the formula (1) is a system linear equation with linear performance degradation, and the formula (2) is an observation equation; x is the number ofk+1For the predicted value of the system state at the sampling time k +1, ykA system output signal vector at the sampling moment of k is obtained, wherein k is more than or equal to 0, and t is the current moment;
Figure FDA0002944451640000013
the calculation parameters for the k sampling time comprise: etakFor the drift velocity of the linear degradation process, taukIs a sampling interval, ekIs noise, akFor sensor null shift, hkAs a system self-parameter, ωkTo measure noise.
3. The method for predicting the remaining life of the ship based on the multi-source information according to claim 2, wherein the calculating of the best state estimation value at the sampling time and each past sampling time comprises:
calculating a system state estimation value at a sampling moment k according to a state optimal estimation value of a measurement sample at the sampling moment k-1 and a system output signal vector at the sampling moment k, wherein t is more than or equal to k and more than or equal to 0, and t is the current moment;
step two, according to the estimated value of the system state at the k sampling time and the parameter updated for the ith time at the k sampling time
Figure FDA0002944451640000021
The optimized state values at past sampling instants are updated,
Figure FDA0002944451640000022
for optimum parameters at the last moment
Figure FDA0002944451640000023
Step three, calculating a joint expectation according to the joint distribution of the system output signal vector and the optimized state value, maximizing the expectation, and obtaining the (l + 1) th updated parameter at the k sampling moment
Figure FDA0002944451640000024
Step four, if the updated parameters
Figure FDA0002944451640000025
If the requirements are met, calculating the state optimal estimation value at the k sampling moment; if the updated parameter does not meet the requirement, adding 1 to the l, and then repeatedly executing the steps from two to four until the state optimal estimation value of the k sampling time is calculated;
and circularly executing the steps from one to four until the state optimal estimation value of the current sampling time and each past sampling time is calculated.
4. The multi-source information-based ship remaining life prediction method according to claim 3, wherein the estimated value of the system state at the k sampling time is
Figure FDA0002944451640000026
Expressed as:
Figure FDA0002944451640000027
Pk|k-1＝Pk-1|k-1+Qk-1
Kk＝Pk|k-1CT(CPk|k-1CT+R)-1
Figure FDA0002944451640000028
Pk|k＝(I-KkC)Pk|k-1
wherein t is more than or equal to k and more than or equal to 0, t is the current moment,
Figure FDA0002944451640000029
to estimate the state at the k sample time based on the k-1 sample time,
Figure FDA00029444516400000210
is the best estimate of the system state at the time of k sampling, ηk-1For the drift velocity, tau, of the linear degradation process at the sampling instant k-1k-1For the sampling interval of k-1 sampling instants, Pk|k-1Predicting a variance matrix, P, for Kalman filtering from k-1 sample instantsk-1|k-1Variance matrix, Q, estimated for the k-1 th sampling instant of Kalman filteringk-1System noise at the sampling instant K-1, KkFor the kalman gain at the time of the k samples,
Figure FDA0002944451640000031
is the state estimate at k sampling instants, C is the measurement matrix, CTFor transposing the measurement matrix C, R is the measurement noise, ykAnd I is a system output signal vector at the sampling moment of k, and is an identity matrix.
5. The multi-source information-based ship remaining life prediction method according to claim 4, wherein the optimized state value at the past sampling time is updated and expressed as:
Figure FDA0002944451640000032
Figure FDA0002944451640000033
Figure FDA0002944451640000034
wherein k is sampling time, j is past sampling time, t is current time, j is more than or equal to 0 and less than or equal to k and less than or equal to t; l isjIs an intermediate variable, Pj|jThe variance matrix estimated for the kalman filter j sample time,
Figure FDA0002944451640000035
the inverse of the variance matrix at sample time j +1 is predicted from sample time j for kalman filtering,
Figure FDA0002944451640000036
to optimize the value for the state at the j sample time based on the k sample time,
Figure FDA0002944451640000037
is the estimated value of the system state at the j sample time,
Figure FDA0002944451640000038
linear degradation process drift speed, tau, for the i-th update of the k sampling instantsjFor a sampling interval, Pj|kTo estimate the estimation error of the j-sample time state with k-sample times,
Figure FDA0002944451640000039
is an intermediate variable LjThe transposing of (1).
6. The multi-source information-based ship remaining life prediction method according to claim 5, wherein the calculation joint expectation is expressed as:
Figure FDA00029444516400000310
wherein t is more than or equal to k and more than or equal to 0, t is the current moment,
Figure FDA0002944451640000041
a likelihood function for k sampling instants; gamma raykFor combined variables of state and observations at and all times before the sampling time k in the first update, i.e.
Figure FDA0002944451640000042
Theta is an independent variable, and theta which is expected to maximize k sampling time is
Figure FDA0002944451640000043
Figure FDA0002944451640000044
For the occurrence under theta condition in the first update
Figure FDA0002944451640000045
The probability of (a) of (b) being,
Figure FDA0002944451640000046
for the combination of the state variables at the sampling instant k and all preceding instants in the first update,
Figure FDA0002944451640000047
for the combination of the observed variables at the k sample time and all previous times in the first update,
Figure FDA0002944451640000048
is the sum of theta in the first update
Figure FDA0002944451640000049
Appear under the condition of
Figure FDA00029444516400000410
The probability of (a) of (b) being,
Figure FDA00029444516400000411
is the sum of theta in the first update
Figure FDA00029444516400000412
Appear under the condition of
Figure FDA00029444516400000413
The probability of (a) of (b) being,
Figure FDA00029444516400000414
for the system state vector optimization value of the i-th update at the time of the t-sample,
Figure FDA00029444516400000415
for the system state vector optimization value of the i-th update at the time of the t-1 sample,
Figure FDA00029444516400000416
the system output signal value for the first update at the time of t samples.
7. The multi-source information-based ship remaining life prediction method according to claim 6, wherein the maximization of the expectation is expressed as:
Figure FDA00029444516400000417
wherein t is more than or equal to k and more than or equal to 0, t is the current moment,
Figure FDA00029444516400000418
for the occurrence under theta condition in the first update
Figure FDA00029444516400000419
The probability of (a) of (b) being,
Figure FDA00029444516400000420
for the combination of the state variables at the sampling instant k and all preceding instants in the first update,
Figure FDA00029444516400000421
the combination of the observation variables at the sampling moment k and all previous moments in the first updating;
the obtained theta satisfying the condition is the updated parameter
Figure FDA00029444516400000422
The method comprises the following steps:
Figure FDA00029444516400000423
Figure FDA00029444516400000424
Figure FDA0002944451640000051
Figure FDA0002944451640000052
wherein k is sampling time, j is past sampling time, t is current time, j is more than or equal to 0 and less than or equal to k and less than or equal to t,
Figure FDA0002944451640000053
linear degradation process drift velocity, τ, for the i +1 th update of the k sampling instantsj-1The sampling interval at the sampling instant j-1,
Figure FDA0002944451640000054
the j-time system state for the i-th update of the k-sample time,
Figure FDA0002944451640000055
the system noise for the (l + 1) th update at the k sampling instant,
Figure FDA0002944451640000056
the estimate of the l +1 st update at the k time for the measurement coefficient corresponding to the ith sensor in the measurement matrix,
Figure FDA0002944451640000057
for the system output signal component corresponding to the moment of the ith sensor j,
Figure FDA0002944451640000058
for the null shift of the i-th sensor,
Figure FDA0002944451640000059
the updated estimate at time k, l +1, for the ith row in the measurement error matrix.
8. The multi-source information-based ship remaining life prediction method according to claim 7, characterized in that:
if the updated parameter
Figure FDA00029444516400000510
The parameters are converged and the parameters are,
Figure FDA00029444516400000511
calculating the optimal parameter of the k sampling time according to the optimal parameter to obtain the optimal estimated value of the system state of the k sampling time
Figure FDA00029444516400000512
If the updated parameter
Figure FDA00029444516400000513
Adding 1 to the l, and then repeating the steps two to four until the state optimal estimation value at the sampling moment k is calculated.
9. The method for predicting the remaining life of a ship based on multi-source information according to claim 1 or 8, wherein the estimating the state value at each future sampling moment further comprises: and estimating the state value of each future sampling moment according to the optimal parameter and the optimal estimated value of the system state, wherein the expression is as follows:
xT＝xT-1+ηtτT-1
wherein T is more than T, T is future sampling time, T is current time, xTFor state estimates at future times, xT-1The initial value is the optimal estimated value of the system state at the sampling time ttFor optimum linear degradation process drift velocity at the moment of t-samplingT-1A sampling interval at time T-1;
setting a State threshold XthComparing the future sampling time state value with a state threshold value, if xT≤XthAnd predicting that the system fails at the T sampling moment.
10. A ship residual life prediction system based on multi-source information is characterized by comprising:
the data acquisition module is used for acquiring system output signal vectors of ship sensor measurement samples and determining the initial state and initial parameters of the system;
the system state optimization module is used for calculating the sampling time and the state optimal estimation value of each past sampling time according to the system output signal vector, the system initial state and the initial parameters;
the system state estimation module is used for estimating a state value of each future sampling moment according to the sampling moment and the state optimal estimation value of each past sampling moment;
and the life prediction module is used for comparing the state value at each future sampling moment with a threshold value so as to predict the residual life of the ship.
CN202110188897.5A 2021-02-19 2021-02-19 Multi-source information-based ship residual life prediction method and system Active CN112836381B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110188897.5A CN112836381B (en) 2021-02-19 2021-02-19 Multi-source information-based ship residual life prediction method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110188897.5A CN112836381B (en) 2021-02-19 2021-02-19 Multi-source information-based ship residual life prediction method and system

Publications (2)

Publication Number Publication Date
CN112836381A true CN112836381A (en) 2021-05-25
CN112836381B CN112836381B (en) 2023-03-14

Family

ID=75933815

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110188897.5A Active CN112836381B (en) 2021-02-19 2021-02-19 Multi-source information-based ship residual life prediction method and system

Country Status (1)

Country Link
CN (1) CN112836381B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113221252A (en) * 2021-05-31 2021-08-06 震兑工业智能科技有限公司 Ship life prediction method and system

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7457785B1 (en) * 2000-08-25 2008-11-25 Battelle Memorial Institute Method and apparatus to predict the remaining service life of an operating system
CN102542155A (en) * 2011-12-05 2012-07-04 北京航空航天大学 Particle filter residual life forecasting method based on accelerated degradation data
US20130246006A1 (en) * 2012-03-13 2013-09-19 King Fahd University Of Petroleum And Minerals Method for kalman filter state estimation in bilinear systems
CN105300692A (en) * 2015-08-07 2016-02-03 浙江工业大学 Bearing fault diagnosis and prediction method based on extended Kalman filtering algorithm
CN109478045A (en) * 2016-04-28 2019-03-15 维利通阿尔法公司 Goal systems is controlled using prediction
CN110456199A (en) * 2019-08-14 2019-11-15 四川大学 A kind of method for predicting residual useful life of multisensor syste
KR102216306B1 (en) * 2019-11-29 2021-02-16 동명대학교산학협력단 System and method for predicting fatigue life of ship
US20210048807A1 (en) * 2018-03-14 2021-02-18 Shandong University Of Science And Technology Method of modeling multi-mode degradation process and predicting remaining useful life

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7457785B1 (en) * 2000-08-25 2008-11-25 Battelle Memorial Institute Method and apparatus to predict the remaining service life of an operating system
CN102542155A (en) * 2011-12-05 2012-07-04 北京航空航天大学 Particle filter residual life forecasting method based on accelerated degradation data
US20130246006A1 (en) * 2012-03-13 2013-09-19 King Fahd University Of Petroleum And Minerals Method for kalman filter state estimation in bilinear systems
CN105300692A (en) * 2015-08-07 2016-02-03 浙江工业大学 Bearing fault diagnosis and prediction method based on extended Kalman filtering algorithm
CN109478045A (en) * 2016-04-28 2019-03-15 维利通阿尔法公司 Goal systems is controlled using prediction
US20210048807A1 (en) * 2018-03-14 2021-02-18 Shandong University Of Science And Technology Method of modeling multi-mode degradation process and predicting remaining useful life
CN110456199A (en) * 2019-08-14 2019-11-15 四川大学 A kind of method for predicting residual useful life of multisensor syste
KR102216306B1 (en) * 2019-11-29 2021-02-16 동명대학교산학협력단 System and method for predicting fatigue life of ship

Non-Patent Citations (8)

* Cited by examiner, † Cited by third party
Title
QIAN C ET AL.: "Reliability analysis method of marine equipment multi-state system based on Markov model", 《SHIP ENGINEERING》 *
XIAO-SHENG SI ET AL.: "A wiener-process-based degradation model with a recursive filter algorithm for remaining useful life estimation", 《MECHANICAL SYSTEMS AND SIGNAL PROCESSING》 *
于震梁等: "基于支持向量机和卡尔曼滤波的机械零件剩余寿命预测模型研究", 《兵工学报》 *
李永杰 等: "船舶自主航行关键技术研究现状与展望", 《中国舰船研究》 *
李玥锌 等: "基于维纳过程的锂离子电池剩余寿命预测", 《大连理工大学学报》 *
邱伯华 等: "基于事件驱动的船载智能信息平台架构设计", 《科技导报》 *
马剑等: "船舶主推进系统故障预测与健康管理设计", 《南京航空航天大学学报》 *
魏慕恒 等: "基于多源监测信息的最优预测维护决策支持方法", 《控制与决策》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113221252A (en) * 2021-05-31 2021-08-06 震兑工业智能科技有限公司 Ship life prediction method and system
CN113221252B (en) * 2021-05-31 2023-01-20 震兑工业智能科技有限公司 Ship life prediction method and system

Also Published As

Publication number Publication date
CN112836381B (en) 2023-03-14

Similar Documents

Publication Publication Date Title
CN110326004B (en) Training a strategic neural network using path consistency learning
CN110574048B (en) Training action selection neural network
CN110378488B (en) Client-side change federal training method, device, training terminal and storage medium
CN109067773B (en) Vehicle-mounted CAN network intrusion detection method and system based on neural network
CN110546653B (en) Action selection for reinforcement learning using manager and worker neural networks
US10395646B2 (en) Two-stage training of a spoken dialogue system
CN118246513A (en) Training action selection neural network
US20150112891A1 (en) Information processor, information processing method, and program
US20230367934A1 (en) Method and apparatus for constructing vehicle dynamics model and method and apparatus for predicting vehicle state information
CN110413754B (en) Conversational (in) reward evaluation and conversational methods, media, apparatuses, and computing devices
CN114261400B (en) Automatic driving decision method, device, equipment and storage medium
CN112766497A (en) Deep reinforcement learning model training method, device, medium and equipment
CN112836381B (en) Multi-source information-based ship residual life prediction method and system
CN113052253A (en) Hyper-parameter determination method, device, deep reinforcement learning framework, medium and equipment
CN113627536A (en) Model training method, video classification method, device, equipment and storage medium
CN115098789A (en) Neural network-based multi-dimensional interest fusion recommendation method and device and related equipment
CN117400785A (en) Prediction method and device for pure electric endurance mileage, electronic equipment and vehicle
CN112600814B (en) Detection method and system for deception attack of ship CPS (cyber-physical system)
CN111957053A (en) Game player matching method and device, storage medium and electronic equipment
CN117151195A (en) Model optimization method, device, equipment and medium based on inversion normalization
CN116176531A (en) Method and device for determining performance index of opening degree adjustment and storage medium
CN113190960B (en) Parallel IMM maneuvering target tracking method based on non-equal-dimensional state hybrid estimation
CN111565118B (en) Virtualized network element fault analysis method and system based on multi-observation dimension HMM
CN112949850A (en) Hyper-parameter determination method, device, deep reinforcement learning framework, medium and equipment
CN112364292B (en) Ransac-based dense target tracking method, ransac-based dense target tracking device, ransac-based dense target tracking equipment and medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant