CN109974837B - Ship structure damage identification method based on rule reasoning - Google Patents
Ship structure damage identification method based on rule reasoning Download PDFInfo
- Publication number
- CN109974837B CN109974837B CN201910314921.8A CN201910314921A CN109974837B CN 109974837 B CN109974837 B CN 109974837B CN 201910314921 A CN201910314921 A CN 201910314921A CN 109974837 B CN109974837 B CN 109974837B
- Authority
- CN
- China
- Prior art keywords
- damage
- ship
- srvr
- rule
- damage identification
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H9/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves by using radiation-sensitive means, e.g. optical means
- G01H9/004—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves by using radiation-sensitive means, e.g. optical means using fibre optic sensors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/04—Inference or reasoning models
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
- Testing Or Calibration Of Command Recording Devices (AREA)
Abstract
The invention relates to a ship structure damage identification method based on rule reasoning, and belongs to the field of ship structure state monitoring and fault diagnosis. According to the method, wavelet packet decomposition is carried out on monitoring signals of the fiber bragg grating sensor, damage identification indexes are obtained and input as a reliability reasoning model, structural damage types are set according to the damage structure position and the damage degree of a ship, the structural damage types are output as the reliability reasoning model, and a reliability rule base is constructed. And calculating the activation weight of all the rules according to the input value, fusing all the rules through an evidence reasoning algorithm, making a decision according to the confidence level of the damage category obtained by fusion, and judging the structural damage category to which the ship structure belongs. And constructing a target function training optimization parameter set to obtain an optimal parameter set of the inference model. And acquiring damage identification indexes on line, obtaining a fusion result based on the optimal reasoning model, making a decision, and judging the structure damage category to which the fusion result belongs. The invention can realize high-precision identification of structural damage of the ship.
Description
Technical Field
The invention relates to a ship structure damage identification method based on rule reasoning, and belongs to the field of ship structure state monitoring and fault diagnosis.
Background
The ship is a large-scale comprehensive system, has a complex structure, is in service in a severe marine environment for a long time, and is influenced by various loads, such as wind load, ocean current, wave load, ice load, deepwater pressure load and the like, and sometimes impacted by typhoon, hull collision, explosion and the like, and the structure itself is influenced by environmental corrosion and the like. Under the long-term action of the severe environmental loads, in addition to improper design or use, various types of damage are easily generated on the structure, so that the bearing capacity of the structure is reduced, disastrous accidents occur, and huge casualties and economic losses are caused. And as the ship structure becomes larger and the sailing speed becomes faster and faster, it has become very difficult to estimate the damage of the ship body caused by the load on the ship body through the experience of the crew. The online ship structure health monitoring system can provide objective and reliable information for ship operators in time, so that various risks of ships can be resisted in navigation, and the problem is urgently solved.
The T-shaped node structure is a typical component of a ship body in a stiffened plate structure of a surface ship or a ribbed straddle of a submersible, and generally affects the total longitudinal strength and the local strength of the ship body, such as the joint of a transverse bulkhead of the ship body and a deck layer, the joint of a side of the ship body and the deck layer, and the like.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a ship structure damage identification method based on rule reasoning. The method comprises the steps of monitoring vibration signals generated when loads such as waves and explosions impact a ship body by a fiber grating sensor, demodulating the vibration signals by using a fiber grating dynamic demodulator to obtain detection signals, and obtaining damage identification indexes after wavelet packet decomposition to be used as reliability inference model input. Setting structural damage categories as reliability reasoning model output according to ship damage structural positions and damage degrees, calculating activation weights of all rules according to input values based on given input reference values, fusing all rules through an evidence reasoning algorithm to obtain a fusion result for decision making, judging the affiliated structural damage categories, then constructing a target function to train evidence reliability reasoning model parameters, and finally making decision through the fusion result to judge the affiliated structural damage categories. According to the method, the data are monitored through the fiber bragg grating sensor, and the high-precision identification of the structural damage of the ship can be achieved.
The invention provides a ship structure damage identification method based on rule reasoning, which comprises the following steps:
(1) the set theta of the ship structure damage is set according to the position and the damage degree of the ship structure damage { F ═ F1,...Fi,...FN},FiThe representative damage of the ith category in the ship structure damage Θ is 1, 2.
(2) SRVR is a set theta capable of reflecting structural damage F of each type in ship structural damage setiThe damage identification index of (2), the damage identification index is specifically defined as follows:
when a ship runs, a fiber grating sensor monitors vibration signals generated by impact of loads such as waves, explosion and the like on a ship body, a fiber grating dynamic demodulator demodulates the vibration signals to obtain detection signals, the detection signals are set as f (t), J-layer wavelet packet decomposition is carried out on the detection signals, each node signal is reconstructed, and each node reconstruction signal is set asThen there are:
the energy of each node reconstructed signal can be defined as:
after removing nodes with smaller energy, retaining first m node signals and reconstructing signals from the signals
In order to ensure that the reconstructed signal retains the main frequency components of the original signal f (t) at this time, the main frequency components of the original signal f (t) can be obtained by calculationCorrelation coefficient with f (t), when the correlation coefficient of the two is more than 0.8, the front m node is consideredThe reconstructed signal is effective, the m nodes before wavelet packet decomposition can effectively reconstruct the original signal, and the proportion of the energy of the reconstructed signal of each node to the total energy of the signal is recorded as
Then the impairment recognition index may be defined as:
in the formula (5)The reference standard of the proportion of the energy of the reconstructed signal of each node in the total energy of the signal can be obtained by averaging the energy proportion of the reconstructed signal of each node by measuring the vibration signal for multiple times under the structural health state.
(3) When N types of structural damage in a ship structural damage set theta respectively occur, acquiring damage identification samples of each type of structural damage, wherein w damage identification indexes are provided for each type of structural damage, and a total of x ═ Nxw damage identification index sample set is used as a training sample and is recorded as F (x) ═ { SRVR1(x),SRVR2(x),...,SRVRn(x) N is the number of fiber grating sensors, SRVRnInput set of reference valuesJnIs the number of reference values.
(4) Constructing a rule base which consists of L rules, wherein the k rule in the established production rule base is described as follows:
in the formula: SRVRnRepresents the nth damage identification index;reference value representing the nth input variable in the kth rule, andL=J1×J2×…×Jn,
(5) sample set of damage identification indexes (SRVR)1(x),SRVR2(x),...,SRVRn(x) And (4) taking the model as an input, fusing and reasoning out the type of the ship structure damage to which the model belongs through a rule base, and specifically comprising the following steps:
(5-1) calculating the nth damage identification index SRVRn(x) Each reference value corresponding theretoA distance of (d) is as shown in the following formula (7)
(5-2) defining the nth Damage recognition index SRVRn(x) Each reference value corresponding theretoHas a degree of matching of
(5-3) calculating the nth Damage recognition index SRVR according to the formula (8) in the step (5-2)n(x) Weight of activation of each rule
WhereinIdentify indicator SRVR for ith impairmenti(x) Reference values corresponding to the respective under the kth ruleDegree of matching of (0 ≦ r)kThe weight of the k-th evidence is less than or equal to 1, and lambda is less than or equal to 0iThe reliability of each damage identification index is less than or equal to 1;
(5-4) obtaining the activation weight of each rule according to the step (5-3)Then, the confidence coefficient m of each rule is calculatedN,kThe fusion was carried out and the fusion result obtained was recorded asThe fusion formula is as follows:
(5-5) fusion results obtained according to the step (5-4)Making a decision to find out the maximum confidence coefficientIt can be judged that the damage identification index sample set belongs to the structural damage Fi。
(6) The method comprises the following steps of constructing a parameter optimization model based on Euclidean distance:
(6-1) determining an optimization parameter set P ═ { m ═ mi,k,rk,λj|i=1,2,...,N;k=1,2,...,L;j=1,2,...,n};
(6-2) minimizing the Euclidean distance as an optimization objective function,
when the damage identification index sample set actually belongs to FiWhen the class structure is damaged,
s.t. 0≤mi,k≤1 (12a)
0≤rk≤1 (12c)
0≤λi≤1 (12d)
equations (12b) - (12d) represent constraints that the optimization parameters need to satisfy;
and (6-3) obtaining an optimal parameter set P by using a GA genetic algorithm toolbox, monitoring vibration signals generated by impact of loads such as waves, explosion and the like on a ship body by using a fiber grating sensor when the ship runs, demodulating by using a fiber grating dynamic demodulator to obtain detection signals, calculating to obtain damage identification indexes, obtaining fusion results according to the step (5), making decisions, and judging the damage types of the ships.
The invention provides a ship structure damage identification method based on rule reasoning. The method comprises the steps of monitoring vibration signals generated when loads such as waves and explosions impact a ship body by a fiber bragg grating sensor, demodulating by using a fiber bragg grating dynamic demodulator to obtain detection signals, decomposing by using a wavelet packet, obtaining damage identification indexes, inputting as a reliability inference model, setting an input reference value according to an input value change range, setting structural damage types according to a ship damage structure position and a damage degree, outputting as the reliability inference model, and establishing the reliability inference model. And calculating the activation weight of all the rules according to the input value, fusing all the rules through an evidence reasoning algorithm to obtain a fusion result, making a decision according to the confidence level of the fusion result, and judging the type of the structure damage to which the decision belongs. Constructing a target function training optimization parameter set to obtain an optimal parameter set of a reasoning model, acquiring damage identification indexes on line, calculating activation weights of all rules, fusing all rules through an evidence reasoning algorithm to obtain a fusion result, making a decision, and judging the type of the structure damage to which the rule belongs. The program (Matlab) compiled by the method can run on a computer and is combined with hardware such as a sensor, a data collector and the like to form an online monitoring system, so that the damage of the ship structure can be monitored and diagnosed in real time.
Drawings
FIG. 1 is a block flow diagram of the process of the present invention;
FIG. 2 is a diagram of test set results in an embodiment of the method of the present invention.
Detailed Description
The invention provides a ship structure damage identification method based on rule reasoning, a flow chart of which is shown in figure 1, and the method comprises the following steps:
(1) the set theta of the ship structure damage is set according to the position and the damage degree of the ship structure damage { F ═ F1,...Fi,...FN},FiThe representative damage of the ith category in the ship structure damage Θ is 1, 2.
(2) SRVR is a set theta capable of reflecting structural damage F of each type in ship structural damage setiThe damage identification index of (2), the damage identification index is specifically defined as follows:
when a ship runs, a fiber grating sensor monitors vibration signals generated by impact of loads such as waves, explosion and the like on a ship body, a fiber grating dynamic demodulator demodulates the vibration signals to obtain detection signals, the detection signals are set as f (t), J-layer wavelet packet decomposition is carried out on the detection signals, each node signal is reconstructed, and each node reconstruction signal is set asThen there are:
the energy of each node reconstructed signal can be defined as:
after removing nodes with smaller energy, reserving signals of the first m nodes and reconstructing signals from signals
In order to ensure that the reconstructed signal retains the main frequency components of the original signal f (t) at this time, the main frequency components of the original signal f (t) can be obtained by calculationAnd f (t), when the correlation coefficient of the two nodes is more than 0.8, the reconstructed signal of the front m nodes is considered to be effective, the m nodes can effectively reconstruct the original signal before wavelet packet decomposition, and the ratio of the reconstructed signal energy of each node to the total signal energy is recorded as
Then the impairment recognition index may be defined as:
in the formula (5)The reference standard of the proportion of the energy of the reconstruction signal of each node in the total energy of the signal can be obtained by averaging the proportion of the energy of the reconstruction signal of each node by measuring the vibration signal for multiple times under the structural health state.
(3) When N types of structural damage in a ship structural damage set theta respectively occur, acquiring damage identification samples of each type of structural damage, wherein w damage identification indexes of each type are obtained, and a total of x ═ N × w damage identification index sample set is used as a training sample and is marked as F (x) ═ SRVR (SRVR)1(x),SRVR2(x),...,SRVRn(x) N is the number of fiber grating sensors, SRVRnInput set of reference valuesJnIs the number of reference values.
(4) Constructing a rule base which consists of L rules, wherein the k rule in the established production rule base is described as follows:
in the formula: SRVRnRepresents the nth damage identification index;reference value representing the nth input variable in the kth rule, andL=J1×J2×…×Jn,
(5) sample set of damage identification indexes (SRVR)1(x),SRVR2(x),...,SRVRn(x) As the input of the model, the method is deduced by fusing the rule baseThe method belongs to the category of ship structure damage, and comprises the following specific steps:
(5-1) calculating the nth damage identification index SRVRn(x) Each reference value corresponding theretoA distance of (d) is as shown in the following formula (7)
(5-2) defining the nth Damage recognition index SRVRn(x) Each reference value corresponding theretoHas a degree of matching of
(5-3) calculating the nth Damage recognition index SRVR according to the formula (8) in the step (5-2)n(x) Weight to activate each rule
WhereinIdentify indicator SRVR for ith impairmenti(x) Reference values corresponding to the respective under the kth ruleDegree of matching of (0 ≦ r)kWeight of k evidence is less than or equal to 1, λ is less than or equal to 0iThe reliability of each damage identification index is less than or equal to 1;
(5-4) obtaining according to the step (5-3)Activation weight to each ruleThen, each rule confidence coefficient m is determinedN,kThe fusion was carried out and the fusion result obtained was recorded asThe fusion formula is as follows:
(5-5) fusion results obtained according to the step (5-4)Making a decision to find out the maximum confidence coefficientIt can be judged that the damage identification index sample set belongs to the structural damage Fi。
To facilitate understanding, it is illustrated how all rules are inferentially fused using equations (7) - (10) in step (5), a two-input one-output model with input and output reference values set as shown in table 1 and a rule base as shown in table 2:
TABLE 1 semantic and reference values for inputs and outputs
TABLE 2 rule base
For example, if the input is {10.5,14.5} belonging to F1Obtained according to the formula (7) Is obtained from the formula (8) Let r bek=1,λ1=λ2Determining the activation weight of each rule according to equation (9) as 1 Fusing the confidence coefficient of each rule according to the formula (10) to obtain a fusion resultMaking a decision on the fusion result, wherein the confidence coefficient is the maximumIt can be determined that the input belongs to F1Damage to the class structure.
(6) The method comprises the following steps of constructing a parameter optimization model based on Euclidean distance:
(6-1) determining an optimization parameter set P ═ { m ═ mi,k,rk,λj|i=1,2,...,N;k=1,2,...,L;j=1,2,...,n};
(6-2) minimizing the Euclidean distance as an optimization objective function,
when the damage identification index sample set actually belongs to FiWhen the class structure is damaged,
s.t. 0≤mi,k≤1 (12a)
0≤rk≤1 (12c)
0≤λi≤1 (12d)
equations (12b) - (12d) represent constraints that the optimization parameters need to satisfy.
And (6-3) obtaining an optimal parameter set P by using a GA genetic algorithm toolbox, monitoring vibration signals generated by impact of loads such as waves, explosion and the like on a ship body by using a fiber grating sensor when the ship runs, demodulating by using a fiber grating dynamic demodulator to obtain detection signals, calculating to obtain damage identification indexes, obtaining fusion results according to the step (5), making decisions, and judging the damage types of the ships.
Embodiments of the method of the present invention are described in detail below with reference to the accompanying drawings:
the flow chart of the method of the invention is shown in figure 1, and the core part is as follows: wavelet packet decomposition is carried out on monitoring signals of the fiber bragg grating sensor, damage identification indexes are obtained and input as a reliability reasoning model, structural damage types are set according to the damage structure position and the damage degree of a ship, the structural damage types are output as the reliability reasoning model, and a reliability rule base is constructed. And calculating the activation weight of all the rules according to the input value, fusing all the rules through an evidence reasoning algorithm, making a decision according to the confidence level of the damage category obtained by fusion, and judging the structural damage category to which the ship structure belongs. And constructing a target function training optimization parameter set to obtain an optimal parameter set of the inference model. And acquiring damage identification indexes on line, obtaining a fusion result based on the optimal reasoning model, making a decision, and judging the structure damage category to which the fusion result belongs.
The steps of the method are described in detail below with reference to a T-shaped beam structure at the bilge of a ship, and the performance of structural damage identification of the method is verified through experimental data.
1. Collection and processing of experimental data
Collecting monitoring data of two fiber bragg grating sensors arranged on a T-shaped beam structure according to the steps (1-2), dividing the data into four types according to different damage degrees of the T-shaped beam structure, obtaining damage identification indexes according to the data, wherein the total number of the damage identification indexes is 500, and the damage identification indexes belong to a first type F1125 of (2) belonging to the second class F2125 of (2) belonging to the third class F3125 belong to the fourth class F4125. Any 100 samples in each class are taken as training samples, the rest 25 samples are taken as testing samples, the training samples account for 400, and the testing samples account for 100.
2. Selection of input reference values
According to the step (3), setting the damage identification index reference value of the first fiber grating sensor as A1J in total, 14,15,16,17,18,19,20,33,34,36,37}111 reference values; damage identification index reference value set A of second fiber grating sensor2J in total, 10,11,12,13,14,15,16,20,21,22,23211 reference values.
3. Building a rule base
According to step (4) of the present invention, damage recognition indexes SRVRs of two fiber grating sensors, each SRVR having 11 reference values, a total of 121 rules are constructed, as shown in table 3 below.
TABLE 3 rule base
3. Training optimization
According to step (6) of the present invention, a total of 607 optimized parameter sets are determined, where r is initiallyk=1,λ1=λ2=1,m1,kAnd m2,kFor the confidence of the rule base in table 3, 400 sets of data are used for training, the minimum Euclidean distance is used as an optimization objective function, and the GA genetic algorithm is used as an optimization algorithm to obtain optimized parameters.
3. Test experiments
And (4) obtaining a structural damage recognition decision result according to the step (5) by using the 100 groups of training data, wherein the calculation recognition accuracy is 89%, and the result is shown in figure 2, so that the expected design precision is achieved.
Claims (2)
1. A ship structure damage identification method based on rule reasoning is characterized by comprising the following steps:
(1) setting a ship structure damage set theta as { F ] according to the position and the damage degree of the damaged ship structure1,...Fi,...FN},FiRepresenting the i-th damage in the ship structure damage theta, wherein i is 1,2, N, and N is the number of the ship structure damage categories;
(2) defining an injury identification index SRVR, specifically:
when a ship runs, a fiber grating sensor monitors a vibration signal generated by the impact of a load on a ship body, a fiber grating dynamic demodulator demodulates the vibration signal to obtain a detection signal, the detection signal is set as f (t), the detection signal is subjected to j-layer wavelet packet decomposition, each node signal is reconstructed, and each node reconstruction signal is set asThen there are:
the energy of each node reconstruction signal is defined as:
after removing nodes with smaller energy, reserving signals of the first m nodes and reconstructing signals from signals
In order to ensure that the reconstructed signal retains the main frequency components of the detection signal f (t) at this time, the main frequency components are determinedCorrelation coefficients with f (t), and signals reconstructed by the first m nodes are considered to be valid when the correlation coefficients of the two are greater than 0.8; setting m nodes before wavelet packet decomposition to effectively reconstruct detection signals, and recording the proportion of reconstruction signal energy of each node to total signal energy as
Defining the damage identification index as:
in the formula (5)Reconstructing a reference standard of signal energy in proportion to the total energy of the signal for each node;
(3) when N types of structural damage in a ship structural damage set theta respectively occur, acquiring damage identification samples of each type of structural damage, wherein w damage identification indexes are provided for each type of structural damage, and a total of x ═ Nxw damage identification index sample set is used as a training sample and is recorded as F (x) ═ { SRVR1(x),SRVR2(x),...,SRVRn(x) N is the number of fiber grating sensors, SRVRn(x) Is An,JnIs the number of reference values;
(4) constructing a rule base which consists of L rules, wherein the k rule in the established production rule base is described as follows:
in the formula: SRVRnRepresents the nth damage identification index;reference value representing the nth input variable in the kth rule, andL=J1×J2×…×Jn,
(5) sample set of damage identification indexes (SRVR)1(x),SRVR2(x),...,SRVRn(x) The method comprises the following steps of inputting a model, fusing and reasoning the class of the structural damage of the ship to which the model belongs through a rule base, and carrying out the following stepsThe following:
(5-1) calculating the nth damage identification index SRVRn(x) Each reference value corresponding theretoA distance of (d) is as shown in the following formula (7)
(5-2) defining the nth Damage recognition index SRVRn(x) Each reference value corresponding theretoHas a degree of matching of
(5-3) calculating the nth Damage recognition index SRVR according to the formula (8) in the step (5-2)n(x) Weight to activate each rule
WhereinIdentify an indicator SRVR for the ith impairmenti(x) Reference values corresponding to the respective under the kth ruleDegree of matching of (0 ≦ r)kThe weight of the k-th evidence is less than or equal to 1, and lambda is less than or equal to 0iThe reliability of the ith damage identification index is not more than 1;
(5-4) obtaining the activation weight of each rule according to the step (5-3)Then, the confidence coefficient m of each rule is calculatedN,kThe fusion was carried out and the fusion result obtained was recorded asThe fusion formula is as follows:
2. The ship structure damage identification method based on rule reasoning according to claim 1, characterized in that: the method also comprises the step of constructing a parameter optimization model based on the Euclidean distance, and the specific steps are as follows:
(6-1) determining a parameter set P ═ { m ═ mi,k,rk,λs|i=1,2,…,N;k=1,2,...,L;s=1,2,...,n};
(6-2) minimizing the Euclidean distance as an optimization objective function,
when the damage identification index sample set actually belongs to FiWhen the class structure is damaged,
s.t.0≤mi,k≤1 (12a)
0≤rk≤1 (12c)
0≤λs≤1 (12d)
equations (12b) - (12d) represent constraints that the optimization parameters need to satisfy;
and (6-3) obtaining an optimal parameter set P by utilizing a GA genetic algorithm toolbox, monitoring a vibration signal generated by the impact of a load on a ship body by using a fiber grating sensor when the ship runs, demodulating by using a fiber grating dynamic demodulator to obtain a detection signal, calculating to obtain a damage identification index, obtaining a fusion result according to the step (5), making a decision, and judging the damage type of the ship to which the fusion result belongs.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910314921.8A CN109974837B (en) | 2019-04-18 | 2019-04-18 | Ship structure damage identification method based on rule reasoning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910314921.8A CN109974837B (en) | 2019-04-18 | 2019-04-18 | Ship structure damage identification method based on rule reasoning |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109974837A CN109974837A (en) | 2019-07-05 |
CN109974837B true CN109974837B (en) | 2022-05-20 |
Family
ID=67085250
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910314921.8A Active CN109974837B (en) | 2019-04-18 | 2019-04-18 | Ship structure damage identification method based on rule reasoning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109974837B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110850206A (en) * | 2019-11-13 | 2020-02-28 | 武汉理工大学 | Inverter capacitor aging fault diagnosis method based on confidence rule reasoning |
CN111912612A (en) * | 2020-07-16 | 2020-11-10 | 中国特种设备检测研究院 | Method and system for monitoring damage of gas cylinder of long-tube trailer |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105203130A (en) * | 2015-10-13 | 2015-12-30 | 哈尔滨工程大学 | Information fusion based fault diagnosis method for vessel integrated navigation system |
CN109115491A (en) * | 2018-10-16 | 2019-01-01 | 杭州电子科技大学 | A kind of evidence fusion method of Electrical Propulsion Ship shafting propulsion system mechanical fault diagnosis |
CN109615228A (en) * | 2018-12-12 | 2019-04-12 | 杭州电子科技大学 | Railway transportation risk probability calculation method based on mixing heuristic rule system |
-
2019
- 2019-04-18 CN CN201910314921.8A patent/CN109974837B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105203130A (en) * | 2015-10-13 | 2015-12-30 | 哈尔滨工程大学 | Information fusion based fault diagnosis method for vessel integrated navigation system |
CN109115491A (en) * | 2018-10-16 | 2019-01-01 | 杭州电子科技大学 | A kind of evidence fusion method of Electrical Propulsion Ship shafting propulsion system mechanical fault diagnosis |
CN109615228A (en) * | 2018-12-12 | 2019-04-12 | 杭州电子科技大学 | Railway transportation risk probability calculation method based on mixing heuristic rule system |
Non-Patent Citations (2)
Title |
---|
Intelligent wear mode identification system for marine diesel engines based on multi-level belief rule base methodology;Xinping Yan 等;《MEASUREMENT SCIENCE AND TECHNOLOGY》;20171215(第29期);1-13 * |
基于证据推理规则的船舶柴油机磨损类型辨识研究;徐晓健 等;《摩擦学学报》;20171130;第37卷(第6期);814-822 * |
Also Published As
Publication number | Publication date |
---|---|
CN109974837A (en) | 2019-07-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112232400B (en) | Stainless steel weld ultrasonic defect detection method based on depth feature fusion | |
Sohn et al. | Structural health monitoring using statistical pattern recognition techniques | |
Xu et al. | A survey on ship collision risk evaluation | |
Chandrashekhar et al. | Structural damage detection using modal curvature and fuzzy logic | |
CN109974837B (en) | Ship structure damage identification method based on rule reasoning | |
CN103076394A (en) | Safety evaluation method for ocean platform based on integration of vibration identification frequencies and vibration mode | |
CN103115789A (en) | Second generation small-wave support vector machine assessment method for damage and remaining life of metal structure | |
CN112597658B (en) | Multi-model fault diagnosis method for marine diesel engine based on working condition identification | |
CN107132760A (en) | The underwater robot state monitoring method described based on fuzzy support vector domain | |
CN110738753A (en) | ship power pipeline fatigue damage prediction system and prediction method | |
CN113538971A (en) | Risk ship detection method based on multi-source heterogeneous data fusion | |
CN109145972A (en) | A kind of watercraft electric propulsion system frequency converter alarm design method | |
CN117076935B (en) | Digital twin-assisted mechanical fault data lightweight generation method and system | |
CN117197415B (en) | Method, device and storage medium for detecting target in inspection area of natural gas long-distance pipeline | |
CN114330096A (en) | Deep learning-based rail transit motor bearing fault degree classification model construction method | |
Lucero et al. | Accelerometer placement comparison for crack detection in railway axles using vibration signals and machine learning | |
US20240149993A1 (en) | Integrated condition monitoring system and method for ice-going vessels | |
CN105976127B (en) | Risk assessment method and device for comprehensive electric propulsion system | |
CN112380782A (en) | Rotating equipment fault prediction method based on mixed indexes and neural network | |
CN111694997A (en) | Navigation data analysis control method and system for unmanned ship | |
CN116127846A (en) | Intelligent evaluation method and system for anti-rolling technology | |
Mikulić et al. | A review of artificial intelligence applications in ship structures | |
CN115410419A (en) | Ship mooring early warning method and system, electronic device and storage medium | |
Tang et al. | Autonomous recognition of a ship structure characteristic signal based on machine learning | |
Shi et al. | Structural damage estimation using short-time Fourier transform and improved convolution neural networks |
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 |