CN112082769A - Intelligent BIT design method of analog input module based on expert system and Bayesian decision maker - Google Patents
Intelligent BIT design method of analog input module based on expert system and Bayesian decision maker Download PDFInfo
- Publication number
- CN112082769A CN112082769A CN202010926904.2A CN202010926904A CN112082769A CN 112082769 A CN112082769 A CN 112082769A CN 202010926904 A CN202010926904 A CN 202010926904A CN 112082769 A CN112082769 A CN 112082769A
- Authority
- CN
- China
- Prior art keywords
- input module
- analog input
- fault
- probability
- bit
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M15/00—Testing of engines
- G01M15/14—Testing gas-turbine engines or jet-propulsion engines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N7/00—Computing arrangements based on specific mathematical models
- G06N7/01—Probabilistic graphical models, e.g. probabilistic networks
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Mathematical Analysis (AREA)
- Algebra (AREA)
- Computational Mathematics (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Chemical & Material Sciences (AREA)
- Combustion & Propulsion (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention provides an intelligent BIT design method of an analog input module based on an expert system and a Bayesian decision maker, and belongs to the field of testing and artificial intelligence. The method is used for improving the reliability of the heavy-duty gas turbine distributed control system and solving the problem of high BIT false alarm rate of the conventional analog input module. The method comprises the following steps: constructing an expert system knowledge base of an analog input module of the heavy gas turbine control system through a knowledge acquisition mechanism, and calculating the confidence coefficient of an expert system rule according to Bayesian probability; and recognizing the normal state, the transient state, the intermittent state and the fault state of the real-time BIT detection information by using a Bayesian decision maker, and loading the fault state information to an inference machine to realize the accurate positioning of the fault of the internal functional circuit of the analog input module. The efficiency and the accuracy of expert system reasoning are improved through the two technical means, and the false alarm rate of the existing analog input module BIT is reduced.
Description
Technical Field
The invention belongs to the field of testing and artificial intelligence, and particularly relates to an intelligent BIT design method of an analog input module based on an expert system and a Bayesian decision maker.
Background
The analog input module is a very important hardware module in the heavy-duty gas turbine distributed control system, and the reliability degree of the analog input module determines the overall reliability degree of the distributed control system. In recent years, distributed control systems are increasingly applied to a plurality of important control tasks, and in order to improve the reliability of the distributed control systems, reduce maintenance cost, improve Test maintenance efficiency and simplify equipment maintenance, a built-in Test (BIT for short) technology is often introduced.
Due to the theoretical limitations of conventional BIT, the BIT false alarm rate of electronic devices has remained high. High false alarm rates not only directly impact the effectiveness of the BIT system, but can also adversely affect the completion of system tasks and the availability, maintenance, and replacement of the system, even causing the loss of confidence in the user.
The method introduces the expert system technology into BIT, constructs a fact base and a rule base of an analog input module according to expert experience, and can reduce BIT false alarm rate to a great extent by depending on an efficient inference machine. However, as a core factor of reasoning, namely confidence coefficient of an expert system rule base, the accuracy of reasoning is determined to a great extent, a better algorithm of the confidence coefficient does not exist at present, and more importantly, information sources of a reasoning machine of the expert system are not screened, so that the BIT false alarm rate caused by system missing report and false report is still not reduced fundamentally, and therefore, the expert system is introduced into the BIT detection field and is far from sufficient.
Disclosure of Invention
In view of the technical problems, the invention provides an intelligent BIT design method of an analog input module based on an expert system and a Bayesian decision maker.
In order to solve the technical problems, the invention adopts the following technical scheme:
an intelligent BIT design method of an analog input module based on an expert system and a Bayesian decision maker comprises the following steps:
constructing an expert system knowledge base of an analog input module of the heavy gas turbine control system through a knowledge acquisition mechanism;
calculating the confidence coefficient of the expert system rule according to the Bayesian probability;
the Bayesian decision maker is used for identifying four states of normal, transient, intermittent and fault of the real-time BIT detection information;
and loading the information of the fault state to an inference machine to realize the inference of the fault reason.
Preferably, each piece of failure knowledge in the knowledge base contains all information of a complete failure, including rule ID, rule antecedent, rule back-piece, confidence, rule interpretation, failure principle.
The expert system rule base adopts a production rule to express the causal relationship between the fault and the reason, and the expression form is as follows:
IF E1 AND E2 AND … AND EN THEN H(CF)
wherein E1, E2, …, EN refer to evidence; h refers to the conclusion (or the assumption that holds under the corresponding evidence), and may be a single object or a plurality of objects; CF refers to confidence, determined by the trusted increased length BI and the untrusted increased length BD.
Preferably, the confidence coefficient of each rule in the expert system rule base is calculated by adopting a Bayesian probability theory;
determining the probability of occurrence of each conclusion H;
solving the posterior probability of H under the occurrence condition of the evidence E according to a Bayesian formula;
according to the obtained posterior probability P (H | E) and the prior probability P (H), obtaining the trust increased length BI and the distrust increased length BD;
and calculating the reliability CF according to the trust increased length BI and the distrust increased length BD:
preferably, the Bayesian decision maker identifies the state of the online BIT monitoring parameters and ensures the reliability of the information before inputting the online BIT monitoring parameters into the inference engine;
and performing probability calculation on the four BIT states, and establishing a Bayesian risk decision table.
Training a Bayesian model of BIT false alarm filtering;
making a decision for the four states by using a trained Bayesian decision maker;
and loading the information of the fault state to an inference machine to realize the inference of the fault reason.
The invention has the following beneficial effects: aiming at the defects that the expert system is low in efficiency in the reasoning process, the credibility factor algorithm is lack and the BIT information false alarm rate obtained by the expert system reasoning machine is high, the reliability factor is calculated by adopting the Bayesian probability, the reasoning accuracy is improved, the Bayesian decision maker is used for deciding which state the BIT information belongs to, the expert system reasoning is only carried out on the information of the fault state, and the reasoning efficiency and accuracy are improved. The combination of the analog input module and the distributed control system improves the reliability of the analog input module, and provides an effective method idea for improving the reliability of the distributed control system.
Drawings
FIG. 1 is a flow chart of an intelligent BIT design of an analog input module based on a professional system and a Bayesian decision maker according to an embodiment of the present invention;
FIG. 2 is a solution process for a Bayesian risk decision table;
FIG. 3 is a flow chart of the inference engine for implementing fault location of the analog input module.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, a flowchart of steps of an intelligent BIT design method for an analog input module based on an expert system and a bayesian decision maker according to an embodiment of the present invention is shown, which includes the following steps:
constructing an analog input module expert system knowledge base through a specific knowledge acquisition mechanism (acquired by adopting a manual acquisition mode);
calculating the rule confidence in the knowledge base according to the Bayesian probability;
the Bayesian decision maker is used for identifying four states of normal, transient, intermittent and fault of the real-time BIT detection information;
and loading the Bayesian decision as the information of the fault state, and realizing accurate fault positioning by using an expert system inference machine.
(1) Constructing an expert system knowledge base of an analog input module
Each fault knowledge in the knowledge base comprises all information of a complete fault, including rule ID, rule front piece, rule back piece, confidence coefficient, rule explanation and fault principle; and constructing a fault knowledge table (see table 1) for storing the fault knowledge of the analog input module.
Table 1 analog input module fault knowledge table
The rule base adopts production rules to express causal relationship between faults and reasons
IF E1 AND E2 AND … AND EN THEN H(CF)
Wherein E1, E2, …, EN refer to evidence; h refers to the conclusion (or the assumption that holds under the corresponding evidence), and may be a single object or a plurality of objects; CF refers to confidence, determined by the trusted increased length BI and the untrusted increased length BD.
(2) Calculating rule confidence according to a Bayesian probability formula;
in a first step, the probability of occurrence of each conclusion H is determined, where HNTo be able to deduce the total number of rules for which H can be concluded, S is the total number of rules, and P (H) is the prior probability of occurrence of H.
The second step is that: if an evidence E is found, the posterior probability of H under the occurrence condition of E is worked out according to a Bayesian formula, wherein NENumber of rules for existence of evidence E, HETo conclude the number of rules for the existence of evidence E and H, P (H | E) is the posterior probability that H occurs in the presence of E.
Thirdly, according to the obtained posterior probability P (H | E) and the prior probability P (H), obtaining the trust increment length BI and the distrust increment length BD:
wherein max (A, B) is the maximum value of A and B, and min (A, B) is the minimum value of A and B.
Fourthly, defining credibility CF according to the trust increased length BI and the distrust increased length BD:
CF＝BI(H,E)-BD(H,E)
if CF >0, the occurrence of E is shown to improve the probability of H occurrence; if CF <0, the occurrence of E is reduced by the probability of H; if CF is 0, the occurrence of E is not related to the probability of occurrence of H. CF is used to illustrate the degree of support of evidence E for conclusion H. Credibility CF is a core variable of an expert system, and uncertainty reasoning in the system is based on credibility.
(3) The Bayesian decision maker is used for identifying four states of normal, transient, intermittent and fault of the real-time BIT detection information;
firstly, probability calculation is carried out on four BIT states, and a Bayesian risk decision table is established.
Risk values are considered from the influence of misjudgment or missed judgment of four BIT states of internal components of the analog input module on the system, namely whether misjudgment or missed judgment of a decision can cause degraded operation of a decentralized control system due to reconstruction, whether meaningless or even wrong maintenance can be caused, and whether task reliability can be reduced or not are used for measuring risks. Through analysis, under the condition of false positive, the transient state is judged to be false negative caused by normal, and the risk is minimum (the risk value is set to be 0< k < 1); the risk of intermittent judgment as normal (the value of k < q ≦ 1); the failure is judged to be the failure caused by normal, and the risk is maximum (the value is set to be 1); under the condition of false alarm, normal false judgment is that the risk of the fault is the largest, transient false judgment is that the risk of the fault is the second highest, and intermittent false judgment is that the risk of the fault is the smallest, so a Bayesian decision table (see table 2) can be made (the variable in the table satisfies the relation that 0< k < q < 1). Different decision results can be obtained by taking different combination values of (k, q).
TABLE 2 Bayesian Risk decision Table
Secondly, approximating a BIT detection signal by using mixed Gaussian density;
firstly, calculating the power spectrum of the BIT state sample signal of each component of the analog input module by adopting a spectrum estimation method, wherein the spectrum estimation method adopts a Welch method.
Wherein x isi NAnd (n) is the sampled data of the ith segment, and a Hanning window is adopted for omega (n).
The probability distribution of the BIT detection signal state of the analog input module adopts a mixed Gaussian model to approximate the actual state distribution, theoretically speaking, if the Gaussian probability density functions are enough, the mixed Gaussian density can approximate any probability distribution function, each Gaussian probability density function has respective mean value and covariance matrix, the size of the probability distribution functions can be determined through samples, and the BIT detection signal state of the analog input module has the following form:
Pi(x) A mixed Gaussian model of the analog input module BIT detection signal state i, K is a mixed Gaussian number, the fitting degree of the probability density of the analog input module BIT detection signal is determined, and cijIs a mixing coefficient, Pij(x) A single gaussian probability density function for the jth component of state i, having the form:
and satisfies the conditions:
thirdly, training a Bayes model of BIT false alarm filtering;
the training of the whole model is divided into two steps: step 1, establishing an initial Bayesian model according to a training sample; and step 2, correcting the model according to the test sample, wherein the steps comprise:
initializing (k, q) ═ k, q, wherein k and q are equal and are numerical values close to zero;
the values of (k, q) can be obtained through a learning process of a decision model, a solving process of a decision risk table is shown in figure 2, and the steps comprise:
initializing (k, q) ═ k0, q0, k0, q0 are equal and are all values close to zero;
and selecting a Gaussian element number K equal to 3, setting a parameter updating step length to obtain an initialized risk table, and carrying out Bayesian decision on the test sample of the analog input module according to the risk table to obtain a group of decision results.
Fixing the value of k, and solving the value of q by using the updating step length alpha;
according to the obtained new risk table, carrying out Bayesian decision on the test sample of the analog input module again to obtain a new decision result;
repeating until q reaches 1;
updating the value of k by using an updating step length alpha, and enabling q to be q0；
The above process is repeated until the value of k reaches 1.
Therefore, a plurality of groups of decision results can be obtained, the decision error of the decision result under each (k, q) combination value is calculated, and the (k, q) value corresponding to the decision result with the minimum decision error is selected, so that the optimal decision model of the system can be established.
And thirdly, the Bayesian decision maker makes decisions on four states.
Training all 150 groups of training samples and 100 groups of testing samples, wherein the maximum decision accuracy is 89.73% after training, and the decision results are all correct when the training samples are input into the model. Table 3 shows the results of the decision making on 100 test samples.
| A | B | C | D | |
| A | 24 | 0 | 0 | 1 |
| B | 0 | 25 | 0 | 0 |
| C | 0 | 1 | 23 | 1 |
| D | 0 | 1 | 0 | 24 |
The experimental results show that the Bayesian decision model makes all decisions on the training samples correct, and some decision errors occur when the Bayesian decision model makes a popularization decision (the decision accuracy of the test samples for model correction is 89.73%, the false report rate is 6.55%, and the false report rate is 3.72%).
According to the establishing process of the Bayesian model, the magnitude of the error rate is related to the distribution of the error rate and the establishment of the decision risk. It should be noted that if the number of samples is insufficient, the prior probabilities have different values in different statistics, so the decision model obtained by the statistical method has a large uncertainty.
Fourthly, loading the fault information to an inference machine to realize reasoning of the fault reason;
the inference flow diagram is shown in fig. 3.
Loading Bayesian decision information as a fault state by adopting a forward reasoning method;
searching a knowledge base, and matching the characteristics of the fault information with the knowledge antecedents;
in the process of matching, the rule with higher weight is preferentially reasoned according to the rule confidence value obtained by Bayesian probability calculation;
and if the front part matching reaches the corresponding threshold value, the matching is successful, the corresponding fault reason is displayed, and the fault accurate positioning of the analog input module is realized.
Aiming at the defects that the efficiency of an expert system is low in the reasoning process, the credibility factor algorithm is lack and the false alarm rate of the BIT information obtained by an expert system reasoning machine is high, the Bayesian probability is adopted to calculate the credibility factor, so that the reasoning accuracy is improved, a Bayesian decision maker is used for deciding which state the BIT information belongs to, and the expert system reasoning is only carried out on the information of the fault state, so that the reasoning efficiency and the accuracy are improved. The combination of the analog input module and the distributed control system improves the reliability of the analog input module, and provides an effective method idea for improving the reliability of the distributed control system.
It is to be understood that the exemplary embodiments described herein are illustrative and not restrictive. Although one or more embodiments of the present invention have been described with reference to the accompanying drawings, it will be understood by those of ordinary skill in the art that various changes in form and details may be made therein without departing from the spirit and scope of the present invention as defined by the following claims.
Claims (5)
1. An analog input module BIT design method based on an expert system and a Bayesian decision maker is characterized by comprising the following steps:
constructing an expert system knowledge base of an analog input module of the heavy gas turbine control system through a knowledge acquisition mechanism;
calculating the confidence coefficient of the expert system rule according to the Bayesian probability;
the Bayesian decision maker is used for identifying four states of normal, transient, intermittent and fault of the real-time BIT detection information;
and loading the information of the fault state to an inference machine to realize the accurate positioning of the fault.
2. The BIT design method of the analog input module based on the expert system and the bayesian decision maker as claimed in claim 1, wherein each fault knowledge in the knowledge base includes all information of a complete fault, including rule ID, rule antecedent, rule consequent, confidence, rule interpretation, and fault principle, the expert system rule base adopts production formula rules to express the causal relationship between the fault and the cause in the form of:
IF E1 AND E2 AND … AND EN THEN H(CF)
wherein the content of the first and second substances,E1，E2，…，ENrefers to evidence;Ha conclusion (or an assumption that holds true under corresponding evidence), which may be a single object or a plurality of objects;CFfinger confidence, increased length by trustBIAnd degree of distrust growthBDAnd (6) determining.
3. The BIT design method of the analog input module based on the expert system and the bayesian decision maker as claimed in claim 2, wherein the confidence of each rule in the rule base of the expert system is calculated by using bayesian probability theory, comprising the following steps:
first, each conclusion is determinedHProbability of occurrence of whereinH N To be able to draw conclusions asHThe total number of rules of (a) is,Sthe number of the total rules is the number of the rules,P(H)is composed ofHPrior probability of occurrence:
the second step is that: if evidence is foundEIs obtained according to Bayes formulaHIn thatEThe posterior probability of the occurrence, where,N E for the existence of evidenceEThe number of rules of (a) is,H E for the existence of evidenceEAnd conclusionHThe number of rules of (a) is,P(H|E)is composed ofHIn thatEPosterior probability of occurrence in the occurrence:
thirdly, according to the determined posterior probabilityP(H|E)And prior probabilityP(H)Determining trust increment lengthBIAnd degree of distrust growthBD：
Wherein max (A, B) is the maximum value of A and B, and min (A, B) is the minimum value of A and B;
the fourth step, according to the trust increment degreeBIAnd degree of distrust growthBDDefinition confidenceCF：
If it is notCF>0, indicating that the occurrence of E is increasedHThe probability of occurrence; if it is notCF<0, descriptionEThe occurrence of the situation is reducedHThe probability of occurrence; if it is notCF=0, pair of cases where E occursHThe probability of occurrence is irrelevant;
CFfor explaining evidenceEFor the conclusionHThe degree of support of;
degree of confidenceCFIs a core variable of an expert system, and uncertainty reasoning in the system is carried out based on credibility.
4. The BIT design method for the analog input module based on the expert system and the bayesian decision maker as claimed in claims 1 to 3, wherein the bayesian decision maker performs state recognition on the online BIT monitoring parameters, improves the reliability of information from the acquisition source of the expert system information, and reduces the BIT false alarm rate caused by missing report and false report, the steps comprising:
firstly, carrying out probability calculation on four BIT states and establishing a Bayesian risk decision table;
the probability distribution of the BIT detection signal state of the analog input module adopts a mixed Gaussian model to approximate the actual state distribution, theoretically speaking, if the Gaussian probability density functions are enough, the mixed Gaussian density can approximate any probability distribution function, each Gaussian probability density function has respective mean value and covariance matrix, the size of the probability distribution functions can be determined through samples, and the BIT detection signal state of the analog input module has the following form:
P i (x)detecting signal states for analog input modules BITiThe model of the mixture of gaussian coefficients of (a),Kthe fitting degree of the probability density of the analog quantity input module BIT detection signal is determined for the mixed Gaussian number,c ij in order to obtain a mixing factor,P ij (x) Is in a stateiTo (1) ajA single gaussian probability density function of individual components, having the form:
and satisfies the conditions:
secondly, training a Bayes model of BIT false alarm filtering,
the training of the whole model is divided into two steps: step 1, establishing an initial Bayesian model according to a training sample; step 2, correcting the model according to the test sample;
thirdly, the Bayesian decision maker identifies the normal state, the transient state, the intermittent state and the fault state of the real-time BIT detection information;
fourthly, loading the fault information to an inference machine to realize inference of fault reasons, loading Bayesian decision information into a forward inference method to obtain fault state information, searching a knowledge base, and matching the characteristics of the fault information with the knowledge antecedents; in the process of matching, the rule with higher weight is preferentially reasoned according to the rule confidence value obtained by Bayesian probability calculation; and if the front part matching reaches the corresponding threshold value, the matching is successful, the corresponding fault reason is displayed, and the accurate fault positioning of the analog input module is realized.
5. The BIT design method of the analog input module based on the expert system and the bayesian decision maker as claimed in claims 1 to 4, characterized in that the information of the fault state through BIT false alarm filtering bayesian decision is loaded, the fault location of the analog input module is realized by using the inference engine, the inference engine can use the CLIPS development tool to perform high-efficiency inference, and the inference engine is based on the rule confidence value calculated by the bayesian probability value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010926904.2A CN112082769A (en) | 2020-09-07 | 2020-09-07 | Intelligent BIT design method of analog input module based on expert system and Bayesian decision maker |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010926904.2A CN112082769A (en) | 2020-09-07 | 2020-09-07 | Intelligent BIT design method of analog input module based on expert system and Bayesian decision maker |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112082769A true CN112082769A (en) | 2020-12-15 |
Family
ID=73732871
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010926904.2A Pending CN112082769A (en) | 2020-09-07 | 2020-09-07 | Intelligent BIT design method of analog input module based on expert system and Bayesian decision maker |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112082769A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117169717A (en) * | 2023-09-11 | 2023-12-05 | 江苏微之润智能技术有限公司 | Motor health assessment method and device based on single chip microcomputer and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120209575A1 (en) * | 2011-02-11 | 2012-08-16 | Ford Global Technologies, Llc | Method and System for Model Validation for Dynamic Systems Using Bayesian Principal Component Analysis |
| CN103198217A (en) * | 2013-03-26 | 2013-07-10 | X·Q·李 | Fault detection method and system |
| CN103426028A (en) * | 2013-08-23 | 2013-12-04 | 北京化工大学 | Fault diagnosis expert system based on inheritance pattern and fault diagnosis method |
| CN104267668A (en) * | 2014-09-02 | 2015-01-07 | 上海交通大学 | Bayes-method-based spaceflight valve part fault diagnosis method in machining process |
| CN104363106A (en) * | 2014-10-09 | 2015-02-18 | 国网辽宁省电力有限公司信息通信分公司 | Electric power information communication fault early warning analysis method based on big-data technique |
| CN110334383A (en) * | 2019-05-09 | 2019-10-15 | 上海电力学院 | Gas turbine fault diagnosis expert system method based on GA and L-M Combinatorial Optimization |
-
2020
- 2020-09-07 CN CN202010926904.2A patent/CN112082769A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120209575A1 (en) * | 2011-02-11 | 2012-08-16 | Ford Global Technologies, Llc | Method and System for Model Validation for Dynamic Systems Using Bayesian Principal Component Analysis |
| CN103198217A (en) * | 2013-03-26 | 2013-07-10 | X·Q·李 | Fault detection method and system |
| CN103426028A (en) * | 2013-08-23 | 2013-12-04 | 北京化工大学 | Fault diagnosis expert system based on inheritance pattern and fault diagnosis method |
| CN104267668A (en) * | 2014-09-02 | 2015-01-07 | 上海交通大学 | Bayes-method-based spaceflight valve part fault diagnosis method in machining process |
| CN104363106A (en) * | 2014-10-09 | 2015-02-18 | 国网辽宁省电力有限公司信息通信分公司 | Electric power information communication fault early warning analysis method based on big-data technique |
| CN110334383A (en) * | 2019-05-09 | 2019-10-15 | 上海电力学院 | Gas turbine fault diagnosis expert system method based on GA and L-M Combinatorial Optimization |
Non-Patent Citations (3)
| Title |
|---|
| 刘震等: "基于贝叶斯决策的机内测试虚警滤波技术", 《系统工程学报》 * |
| 张秦岭等: "大型无人机电源系统故障诊断专家系统", 《北京航空航天大学学报》 * |
| 杜鹏昊: "贝叶斯概率逻辑在人工智能专家系统中的应用", 《无线互联科技》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117169717A (en) * | 2023-09-11 | 2023-12-05 | 江苏微之润智能技术有限公司 | Motor health assessment method and device based on single chip microcomputer and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109086889B (en) | Terminal fault diagnosis method, device and system based on neural network | |
| CN109492193B (en) | Abnormal network data generation and prediction method based on deep machine learning model | |
| CN109141847B (en) | Aircraft system fault diagnosis method based on MSCNN deep learning | |
| CN112910859B (en) | Internet of things equipment monitoring and early warning method based on C5.0 decision tree and time sequence analysis | |
| CN110335168B (en) | Method and system for optimizing power utilization information acquisition terminal fault prediction model based on GRU | |
| CN110348615B (en) | Cable line fault probability prediction method based on ant colony optimization support vector machine | |
| CN112508243B (en) | Training method and device for multi-fault prediction network model of power information system | |
| CN108647707B (en) | Probabilistic neural network creation method, failure diagnosis method and apparatus, and storage medium | |
| CN112116010B (en) | Classification method for ANN-SNN conversion based on membrane potential pretreatment | |
| CN112163624A (en) | Data abnormity judgment method and system based on deep learning and extreme value theory | |
| CN112819107A (en) | Artificial intelligence-based fault prediction method for gas pressure regulating equipment | |
| CN112101765A (en) | Abnormal data processing method and system for operation index data of power distribution network | |
| US20220243347A1 (en) | Determination method and determination apparatus for conversion efficiency of hydrogen production by wind-solar hybrid electrolysis of water | |
| CN112507605A (en) | Power distribution network anomaly detection method based on AnoGAN | |
| CN112082769A (en) | Intelligent BIT design method of analog input module based on expert system and Bayesian decision maker | |
| CN114564513A (en) | Sea fog prediction method, device, equipment and storage medium based on neural network | |
| Liu et al. | Research on the strategy of locating abnormal data in IOT management platform based on improved modified particle swarm optimization convolutional neural network algorithm | |
| CN117216713A (en) | Fault delimiting method, device, electronic equipment and storage medium | |
| CN111934903A (en) | Docker container fault intelligent prediction method based on time sequence evolution genes | |
| CN114579962A (en) | AI safety attack and defense test method | |
| CN114330141A (en) | Fan main shaft bearing service life prediction method based on GRU (generalized regression Unit) super-parameter optimization | |
| CN113807462A (en) | AI-based network equipment fault reason positioning method and system | |
| CN111310907A (en) | Microwave assembly fault diagnosis method, device and equipment | |
| CN117149551B (en) | Test method of vehicle-mounted wireless communication chip | |
| CN113139645A (en) | Robustness evaluation method of neural network |
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 |