CN108376264A - A kind of handpiece Water Chilling Units method for diagnosing faults based on support vector machines incremental learning - Google Patents
A kind of handpiece Water Chilling Units method for diagnosing faults based on support vector machines incremental learning Download PDFInfo
- Publication number
- CN108376264A CN108376264A CN201810159234.9A CN201810159234A CN108376264A CN 108376264 A CN108376264 A CN 108376264A CN 201810159234 A CN201810159234 A CN 201810159234A CN 108376264 A CN108376264 A CN 108376264A
- Authority
- CN
- China
- Prior art keywords
- data
- water chilling
- chilling units
- handpiece water
- collection data
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
According to the handpiece Water Chilling Units method for diagnosing faults according to the present invention based on support vector machines incremental learning, which is characterized in that include the following steps：S1 collects handpiece Water Chilling Units malfunction history data, including initial training collection data；S2 pre-processes the historical data in S1, the initial training collection data that obtain that treated；S3 selects Radial basis kernel function as the kernel function of SVM (support vector machines), builds the diagnostor based on SVM models；S4 obtains initial diagnosis device after initial training collection data are trained in diagnostor using treated in S2；S5, collection in worksite new data, and new data is pre-processed to obtain new samples collection；S6 obtains last diagnostic device using Increment Learning Algorithm adaptive updates initial diagnosis device；S7 acquires the operation data of handpiece Water Chilling Units and is pre-processed in real time；Pretreated operation data in S7 is input in the last diagnostic device in S6, obtains diagnostic result by S8.
Description
Technical field
The invention belongs to refrigerating fields, and in particular to a kind of handpiece Water Chilling Units method for diagnosing faults.
Background technology
Heating ventilation air-conditioning system is multivariable, multimode, strongly coupled system, is chronically at variable working condition, operation at part load shape
State.It is always industrial equipment, the big power consumer in household electrical appliance.And handpiece Water Chilling Units are most important in heating ventilation air-conditioning system set
It is standby, and the maximum equipment of energy consumption.Due to many reasons, equipment is made often to break down.Also, with the automation of equipment and
Highgrade integration, the cost and maintenance cost of equipment are also sharply increasing.Therefore, its failure symptom is found in time and is incited somebody to action in failure
It is excluded when occurring, to reducing equipment energy consumption, the comfort level for improving people has very important significance.In the past few decades
In, the development of handpiece Water Chilling Units fault detection and diagnosis (fault detection and diagnosis, FDD) is always to study
One of hot spot.
Currently, it is mainly three categories to apply the FDD methods in heating ventilation air-conditioning system：Quantitative model-based method, base
Method, the method for Kernel-based methods historical data in qualitative model.Because refrigeration system is complex, knowledge acquisition is difficult, before being
The research of the two brings inconvenience；And Kernel-based methods historical data method then fully relies on data itself and establishes model, is not required to
Want priori, you can useful information is extracted from abundant data, therefore the method for Kernel-based methods historical data is being made
Good popularization has been obtained in the FDD of cooling system.
Support vector machines (Support Vector Machine, abbreviation SVM) is that the research group of Vapnik leaders proposes
A kind of novel universal the machine learning method for having supervision.It is established on structure minimization principle basis, is had very strong
Learning ability and Generalization Capability, the problems such as capable of preferably solving small sample, high dimension, non-linear, local minimum, Ke Yiyou
Effect classify, return, density estimation etc..Chiller system in Heating, Ventilation and Air Conditioning (HVAC) systems is difficult the whole fault samples of disposable acquisition, is needed
Machine by constantly learning abundant diagnosis detecting system, adapts to handpiece Water Chilling Units characteristic complicated and changeable, carries in the process of running
The accuracy of height diagnosis detection, can make training data sample huge in this way, and what SVM was substantially solved is a convex double optimization
Problem, time complexity and space complexity are higher, and especially when handling large-scale data, learning efficiency is very low, needs to seek
Look for an efficient solution.
Invention content
The present invention is to carry out to solve the above-mentioned problems, and being designed to provide for one aspect of the present invention is a kind of based on branch
Hold the handpiece Water Chilling Units method for diagnosing faults of vector machine incremental learning.
The present invention provides a kind of handpiece Water Chilling Units method for diagnosing faults based on support vector machines incremental learning, feature exists
In including the following steps：
S1 collects handpiece Water Chilling Units malfunction history data, including initial training collection data；
S2 pre-processes the historical data in S1, the initial training collection data that obtain that treated；
S3 selects Radial basis kernel function as the kernel function of SVM (support vector machines), builds the diagnosis based on SVM models
Device；
S4 obtains initial diagnosis device after initial training collection data are trained in diagnostor using treated in S2；
S5, collection in worksite new data are demarcated fault category to the new data of collection in worksite and are pre-processed to new data；
S6 obtains last diagnostic device using Increment Learning Algorithm adaptive updates initial diagnosis device；
S7 acquires handpiece Water Chilling Units operation data and is pre-processed in real time；
Pretreated operation data in S7 is input in the last diagnostic device in S6, obtains diagnostic result by S8.
In the handpiece Water Chilling Units method for diagnosing faults provided by the invention based on support vector machines incremental learning, can also have
There is such feature：Wherein, pretreatment uses normalized.
In addition, in the handpiece Water Chilling Units method for diagnosing faults provided by the invention based on support vector machines incremental learning, also
It can have the feature that：
Wherein, the expression formula of Radial basis kernel function is：
σ is the width of radial basis function, and g is radial base nuclear parameter, and x is to differentiate sample；xiFor training sample.
In addition, in the handpiece Water Chilling Units method for diagnosing faults provided by the invention based on support vector machines incremental learning, also
It can have the feature that：Wherein, the expression formula of diagnostor model is：
Sgn is sign function, yiTo judge the classification of sample；α is Lagrange multipliers；B is threshold value, and i indicates sample sequence
Row number, N indicate sample total number.
In addition, in the handpiece Water Chilling Units method for diagnosing faults provided by the invention based on support vector machines incremental learning, also
It can have the feature that：Wherein, using the method for cross validation optimize penalty coefficient C and nuclear parameter g, to training sample into
When row training, between log2c settings -10 to 10, kernel parameter log2g settings -10 to 10 execute cross validation parameter optimization journey
Sequence chooses optimized parameter best c as punishment parameter, chooses optimum core parameter best g as nuclear parameter.
In addition, in the handpiece Water Chilling Units method for diagnosing faults provided by the invention based on support vector machines incremental learning, also
It can have the feature that：Wherein, update initial diagnosis device includes the following steps：
S6-1 obtains initial diagnosis device using initial training collection data training SVM, and initial training collection data include first
Hold vector set data；
S6-2, for new samples collection data application after initial diagnosis device, obtained data include the data of diagnostic error, i.e., and
One test errors collection data；If the first test errors collection data are zero, iteration terminates, and does not update initial diagnosis device, otherwise
Using the first test errors collection data and the first supporting vector collection data as the first training sample set data to initial diagnosis device into
Row training obtains new diagnostor, while obtaining the second supporting vector collection data；
S6-3 verifies new diagnostor with new samples collection data and new data, and obtained data include the data of diagnostic error,
That is the second test errors collection data；If the second test errors collection data are equal to zero, new diagnostor is last diagnostic device；Otherwise will
First supporting vector collection data, the second supporting vector collection data, the first test errors collection data and the second test errors collection number
New diagnostor is trained according to as the second training sample set data, obtains last diagnostic device.
In addition, in the handpiece Water Chilling Units method for diagnosing faults provided by the invention based on support vector machines incremental learning, also
It can have the feature that：Wherein, as a result include breaking down or fault-free, if a failure occurs, carry out fault alarm, such as
Fruit fault-free carries out artificial judgment.
In addition, in the handpiece Water Chilling Units method for diagnosing faults provided by the invention based on support vector machines incremental learning,
It is characterized in that：Further include S9, judges whether to stop refrigeration system operation, be judged as NO, return to S5, be judged as YES, refrigeration system
Operation stops, and diagnostic work terminates.
The effect of invention
According to a kind of handpiece Water Chilling Units adaptive failure diagnosis based on support vector machines incremental learning according to the present invention
Method.This method uses the parameter of cross validation algorithm optimization support vector machines first, builds handpiece Water Chilling Units fault diagnosis device；When
Occur to new handpiece Water Chilling Units operation data, diagnostor needs to adapt to new operating condition, is first classified with former diagnostor, to classification
The data of mistake merge the new diagnostor of training with the supporting vector of diagnostor, in next step with new diagnostor to all data point
Class obtains finally after the data of double classification mistake are merged the new diagnostor of training with the supporting vector collection data of two diagnostors
Diagnostor.
Compared to it is original it is simple all data are used for Training diagnosis device, method of the invention is ensureing measuring accuracy
While can effectively cast out the data useless to final classification result, to reduce the training time.
Method proposed by the present invention can be efficiently applied to handpiece Water Chilling Units fault diagnosis, have adaptable, diagnosis correctly
The features such as rate is high, the training time is few, calculation amount is small.
Description of the drawings
Fig. 1 is support vector machines incremental learning diagnostic flow chart in the embodiment of the present invention.
Specific implementation mode
It is real below in order to make the technical means, the creative features, the aims and the efficiencies achieved by the present invention be easy to understand
Example combination attached drawing is applied to be specifically addressed the handpiece Water Chilling Units method for diagnosing faults based on support vector machines incremental learning of the present invention.
Embodiment
Handpiece Water Chilling Units method for diagnosing faults based on support vector machines incremental learning includes the following steps：
Establish handpiece Water Chilling Units diagnostor module
S1 collects handpiece Water Chilling Units malfunction history data, including initial training collection data.
Collect the handpiece Water Chilling Units malfunction history data of several experiments or scene storage.Experimental data in the present embodiment comes from
The centrifugal refrigerating machines of one 90 standard ton (about 316kW).It is simulated in laboratory conditions by the testing stand of special designing cold
7 kinds of typical gradual failures of water dispenser group, each failure are all tested under 27 kinds of operating modes, and a large amount of experiment number is had collected
According to.These data include the data under normal operation and failure operation.The present embodiment is not easy to detect gradually for 7 kinds of handpiece Water Chilling Units
Become failure and carry out diagnostic analysis, local fault has gas containing incoagulability in condenser structure, condenser side water flow deficiency refrigerant
Body, vaporizer side water flow are insufficient, and the system failure has insufficient refrigerant leakage/charging amount, lubricating oil excess, refrigerant charging mistake
Amount.Data set shares 8 kinds of classifications, seven kinds of failures and a kind of normal category, acquires and be calculated totally 64 spies from refrigeration system
Sign, they can be used for building diagnostic model, the fault category that diagnosis refrigeration system occurs.
1600 (200 × 8) a historical datas are randomly selected from experimental data as initial training collection, totally 8 type, often
200 data of type.
S2 pre-processes the historical data in S1, the initial training collection data that obtain that treated.
Historical data in S1 is pre-processed, pretreatment uses normalized.
The value of feature is zoomed to section [0,1] by Max-Min normalization；
Wherein, xiIt is a specific sample, xminAnd xmaxIt is the minimum value and maximum value of certain characteristic series respectively.
S3 selects Radial basis kernel function as the kernel function of SVM (support vector machines), builds the diagnosis based on SVM models
Device.
It uses Radial basis kernel function as the kernel function of SVM, penalty coefficient C, nuclear parameter is optimized using the method for cross validation
G builds the fault diagnosis device based on SVM models, in refrigeration systems.
SVM is built upon the Novel learning machine on the basis of Statistical Learning Theory and structural risk minimization, its root
Seek optimal compromise according between the complexity and learning ability of finite sample information in a model, to obtain strongest popularization energy
Power.Breakdown of refrigeration system diagnosis is nonlinear problem of classifying more, and more classification can be reduced to two classification problems, non-linear to ask
Topic can be converted into the linear problem in some higher dimensional space by nonlinear change, be found in transformation space optimal super flat
Face.
Here two classification problem of support vector machines is mainly introduced.Given training sample setBy nonlinear function Ф by data sample
From luv spaceIt is mapped to a high-dimensional feature space Middle construction optimal classification surface so that training data is by hyperplane
It separates, the kernel function of Mercer conditions is met by introducingSolve quadratic programming problem：
In formula：MaxQ (α) is quadratic programming function；xi, xjFor the sample of training set；yiTo judge the classification of sample；α is
Lagrange multipliers；C is penalty coefficient, it can be achieved that compromise between class interval and error rate.Utilize KKT (Karush-
Kuhn-Tucker) optimal condition acquires threshold value b, to obtain optimal classification decision function, as fault diagnosis device mould
Type.
The expression formula of fault diagnosis device model is：
In formula：X is to differentiate sample；xiFor training sample；Sgn is sign function, yiTo judge the classification of sample；α is
Lagrange multipliers；B is threshold value, and i indicates that sample sequence number, N indicate sample total number.
By the positive and negative classification that can be differentiated belonging to sample x of f (x).
The present invention chooses Radial basis kernel function：
In formula：G is nuclear parameter
Determine parameter：Breakdown of refrigeration system diagnosis belongs to Nonlinear Classification, which need to set punishment parameter C, nuclear parameter
G, when being trained to training sample, between log2c settings -10 to 10, then kernel parameter log2g settings -10 to 10 are held
Row cross validation parameter optimization program chooses optimized parameter bestc as between punishment parameter, chooses optimum core parameter best g
As nuclear parameter.
S4, using treated in S2, initial training collection data are trained to obtain initial diagnosis device in diagnostor.
S5 demarcates fault category to the new data of collection in worksite and is pre-processed to new data, obtains new samples collection.
When there are the situations such as new handpiece Water Chilling Units operating mode, diagnosis performance decline, the new data of collection in worksite is demarcated
Fault category, 6 groups of new datas of collection in worksite in embodiment, every group of 800 data, and new data is pre-processed, it obtains new
Sample set, new samples collection are divided into 6 groups of new datas (increment 1, increment 2, increment 3, increment, increment 5, increment 6), every group of 800 numbers
According to, successively carry out incremental learning simplation verification.
S6, using Increment Learning Algorithm adaptive updates initial diagnosis device.
Using the method adaptive updates initial diagnosis device of incremental learning, last diagnostic device is obtained, for chilled water system event
Barrier diagnosis, update initial diagnosis device include the following steps：
S6-1 obtains initial diagnosis device using initial training collection data training SVM, and initial training collection data include first
Hold vector set data；
S6-2, support vector machines are supervised learning, i.e., know the characteristic and fault category of failure in advance, when will instruct
Experienced diagnostor be used for fault diagnosis when, i.e., bring characteristic into diagnostor, fault category can be obtained, this classification with it is known
Classification it is different, then be the data of diagnostic error, be otherwise the correct data of diagnosis.New samples collection data application is initially being examined
After disconnected device, obtained data include the data of diagnostic error, i.e. the first test errors collection data；If the first test errors collection number
According to being zero, then iteration terminates, and does not update initial diagnosis device, otherwise by the first test errors collection data and the first supporting vector collection number
Initial diagnosis device is trained according to as the first training sample set data, obtains new diagnostor, at the same obtain second support to
Quantity set data；
S6-3 verifies new diagnostor with new samples collection data and new data, and obtained data include the data of diagnostic error,
That is the second test errors collection data；If the second test errors collection data are equal to zero, new diagnostor is last diagnostic device；Otherwise will
First supporting vector collection data, the second supporting vector collection data, the first test errors collection data and the second test errors collection number
New diagnostor is trained according to as the second training sample set data, obtains last diagnostic device.
Diagnostor after each incremental learning is detected diagnosis to test data, verifies its diagnosis performance.
Diagnostic module
S7 acquires handpiece Water Chilling Units operation data and is pre-processed in real time.
Acquisition handpiece Water Chilling Units operation data in real time, randomly selects 1600 groups of operation datas and is pre-processed.
S8, operation data is input in the last diagnostic device in S6 after being pre-processed in S7, is broken down or without reason
The diagnostic result of barrier carries out fault alarm, otherwise carries out artificial judgment if it is breaking down.
S9, whether artificial checkout and diagnosis result is correct, judges whether to stop refrigeration system operation.It is judged as NO, returns to S5,
It is judged as YES, refrigeration system operation stops, and diagnostic work terminates.
As shown in Figure 1, support vector machines incremental learning diagnostic process
A1 begins setting up handpiece Water Chilling Units diagnostor module；
A2 collects handpiece Water Chilling Units malfunction history data；
A3 pre-processes the historical data in A2；
A4 builds the diagnostor based on SVM models；
A5, collection in worksite new data update the initial diagnosis device in A4, if to update, are returned using Increment Learning Algorithm
To A4；If do not updated, into next step；
Diagnostor module is established in A6, end.
B1 begins setting up diagnostic module；
B2 acquires the data of handpiece Water Chilling Units process of refrigerastion in real time；
B3 pre-processes the data in B2；
Pretreated data in B2 are input in the diagnostor of A4 foundation, obtain diagnostic result by B4, in case of event
Barrier, into next step, if fault-free, into lower B5；
B41, after carrying out fault alarm, into next step；
Whether B5, artificial judgment stop refrigeration system operation, are judged as NO, return to B2, be judged as YES, refrigeration system operation
Stop, into next step；
B6, diagnostic work terminate.
Interpretation of result
Table 1 is two kinds of diagnostor diagnostic result analyses of handpiece Water Chilling Units, wherein classical SVM study refers to using total data
Training obtains diagnostic model, and increment number represents the number of acquisition new data, and accuracy indicates that test sample is correctly diagnosed
The ratio for testing total sample is accounted for, number of training indicates that the number of samples for participating in establishing diagnostic model, time expression establish model
It takes.The rate of correct diagnosis of diagnostor rises to 97.87%, new data with the addition of new data from initial accuracy for 83.88%
Addition improve rate of correct diagnosis, improve the stability of diagnostic model, so in refrigeration system actual motion, need root
New data is acquired according to operating condition, to adapt to refrigeration system and improve the stability of diagnostic model.Classical SVM learning outcomes and increasing
Amount SVM rate of correct diagnosis gaps are smaller, and have diminishing trend, maximum difference 1.00%, two in last learning outcome
The accuracy difference 0.15% of person.Increment SVM is obviously upper less than classics SVM learning methods, the instruction of increment SVM in number of training
Practice sample number and be less than classics SVM, but the former computational complexity is higher than the latter, so starting both study stages training sample
When number difference unobvious, i.e., preceding 4 incremental learnings, increment SVM is not better than classics SVM on the training time；It is continuous to go deep into
Study, increment SVM show jump, can be seen that in learning from increment 5 and increment 6.Above analysis shows with study
Deepen continuously, method of the invention has high accuracy, training time few, and the advantage that calculation amount is small.
1 handpiece Water Chilling Units of table, two kinds of diagnostor diagnostic result analyses
The effect of embodiment
The handpiece Water Chilling Units adaptive failure diagnostic method based on support vector machines incremental learning of the present embodiment, SVM conducts
Diagnostor needs to add new training data, makes to examine when new operating mode occurs in refrigeration system or diagnosis performance declines
Disconnected device adapts to current handpiece Water Chilling Units operational mode.Method proposed by the present invention can be efficiently applied to handpiece Water Chilling Units fault diagnosis,
Have the characteristics that adaptable, rate of correct diagnosis is high, the training time is few, calculation amount is small.
In addition, the new data of acquisition is further learnt using Increment Learning Algorithm, this method only selects may be to diagnostor
There are the data of significant contribution to learn simultaneously Training diagnosis device, rejecting does not have contributive data, so as to maximumlly save the time,
Ensure the raising of rate of correct diagnosis simultaneously.
The above embodiment is the preferred case of the present invention, is not intended to limit protection scope of the present invention.
Claims (8)
1. a kind of handpiece Water Chilling Units method for diagnosing faults based on support vector machines incremental learning, which is characterized in that including following step
Suddenly：
S1 collects handpiece Water Chilling Units malfunction history data, including initial training collection data；
S2 pre-processes the historical data in S1, the initial training collection data that obtain that treated；
S3 selects Radial basis kernel function as the kernel function of SVM (support vector machines), builds the diagnostor based on SVM models；
S4 obtains initial diagnosis after the initial training collection data are trained in the diagnostor using treated in S2
Device；
S5, collection in worksite new data, and the new data is pre-processed to obtain new samples collection data；
S6 obtains last diagnostic device using initial diagnosis device described in Increment Learning Algorithm adaptive updates；
S7 acquires the operation data of handpiece Water Chilling Units and is pre-processed in real time；
The pretreated operation data in S7 is input in the last diagnostic device in S6, obtains diagnostic result by S8.
2. the handpiece Water Chilling Units method for diagnosing faults of support vector machines incremental learning according to claim 1, it is characterised in that：
Wherein, the pretreatment uses normalized.
3. the handpiece Water Chilling Units method for diagnosing faults according to claim 1 based on support vector machines incremental learning, feature
It is：
Wherein, the expression formula of the Radial basis kernel function is：
σ is the width of the radial basis function, and g is radial base nuclear parameter, and x is to differentiate sample；xiFor training sample.
4. the handpiece Water Chilling Units method for diagnosing faults according to claim 3 based on support vector machines incremental learning, feature
It is：
Wherein, the expression formula of the diagnostor model is：
Sgn is sign function, yiTo judge the classification of sample；α is Lagrange multipliers；B is threshold value, and i indicates sample sequence number, N
Indicate sample total number.
5. the handpiece Water Chilling Units method for diagnosing faults according to claim 3 based on support vector machines incremental learning, feature
It is：
Wherein, the penalty coefficient C and nuclear parameter g is optimized using the method for cross validation, when being trained to training sample,
Between log2c settings -10 to 10, kernel parameter log2g settings -10 to 10 execute cross validation parameter optimization program, choose most
Excellent parameter best c choose optimum core parameter best g as nuclear parameter as punishment parameter.
6. the handpiece Water Chilling Units method for diagnosing faults according to claim 1 based on support vector machines incremental learning, feature
It is：
Wherein, the initial diagnosis device is updated to include the following steps：
S6-1 obtains initial diagnosis device using initial training collection data training SVM, and the initial training collection data include the
One supporting vector collection data；
S6-2, for the new samples collection data application after the initial diagnosis device, obtained data include the first test errors collection
Data；If the first test errors collection data are zero, iteration terminates, and does not update the initial diagnosis device, otherwise by institute
The first test errors collection data and the first supporting vector collection data are stated as the first training sample set data to described initial
Diagnostor is trained, and obtains new diagnostor, while obtaining the second supporting vector collection data；
S6-3 verifies the new diagnostor with the new samples collection data and the new data, and obtained data include the second survey
Try Error Set data；If the second test errors collection data are equal to zero, the new diagnostor is last diagnostic device；Otherwise will
The first supporting vector collection data, the second supporting vector collection data, the first test errors collection data and described
Second test errors collection data are trained the new diagnostor as the second training sample set data, obtain described finally examining
Disconnected device.
7. the handpiece Water Chilling Units method for diagnosing faults according to claim 1 based on support vector machines incremental learning, feature
It is：
The result includes breaking down or fault-free,
If a failure occurs, artificial judgment is carried out after carrying out fault alarm, if fault-free, carries out artificial judgment.
8. the handpiece Water Chilling Units method for diagnosing faults according to claim 1 based on support vector machines incremental learning, feature
It is：
Further include S9, whether artificial judgment stops refrigeration system operation, be judged as NO, return to S5, be judged as YES, refrigeration system fortune
Row stops, and diagnostic work terminates.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810159234.9A CN108376264A (en) | 2018-02-26 | 2018-02-26 | A kind of handpiece Water Chilling Units method for diagnosing faults based on support vector machines incremental learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810159234.9A CN108376264A (en) | 2018-02-26 | 2018-02-26 | A kind of handpiece Water Chilling Units method for diagnosing faults based on support vector machines incremental learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108376264A true CN108376264A (en) | 2018-08-07 |
Family
ID=63018054
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810159234.9A Pending CN108376264A (en) | 2018-02-26 | 2018-02-26 | A kind of handpiece Water Chilling Units method for diagnosing faults based on support vector machines incremental learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108376264A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109190719A (en) * | 2018-11-30 | 2019-01-11 | 长沙理工大学 | Support vector machines learning method, device, equipment and computer readable storage medium |
| CN109540562A (en) * | 2018-12-12 | 2019-03-29 | 上海理工大学 | A kind of water cooler method for diagnosing faults |
| CN110263846A (en) * | 2019-06-18 | 2019-09-20 | 华北电力大学 | The method for diagnosing faults for being excavated and being learnt based on fault data depth |
| CN110705631A (en) * | 2019-09-27 | 2020-01-17 | 中国船舶工业系统工程研究院 | SVM-based bulk cargo ship equipment state detection method |
| CN110702438A (en) * | 2019-09-23 | 2020-01-17 | 天津大学 | Centrifugal water chilling unit and fresh air system combined fault diagnosis method |
| CN111337263A (en) * | 2020-02-12 | 2020-06-26 | 中国民航大学 | Fault diagnosis method for engine turbine disk |
| CN112964469A (en) * | 2021-02-28 | 2021-06-15 | 哈尔滨理工大学 | Online fault diagnosis method for rolling bearing under variable load of transfer learning |
| CN114120592A (en) * | 2021-10-15 | 2022-03-01 | 青岛海尔空调电子有限公司 | Method and device for fault alarm, electronic equipment and storage medium |
| CN114429153A (en) * | 2021-12-31 | 2022-05-03 | 苏州大学 | Lifetime learning-based gearbox increment fault diagnosis method and system |
| CN115127192A (en) * | 2022-05-20 | 2022-09-30 | 中南大学 | Semi-supervised water chilling unit fault diagnosis method and system based on graph neural network |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1563857A (en) * | 2004-03-24 | 2005-01-12 | 上海理工大学 | Double-temp-raising produced overcooling absorption and single effect, afterheat absorbing refrgerating system |
| US20130268469A1 (en) * | 2012-04-06 | 2013-10-10 | Applied Materials, Inc. | Increasing signal to noise ratio for creation of generalized and robust prediction models |
| CN104462846A (en) * | 2014-12-22 | 2015-03-25 | 山东鲁能软件技术有限公司 | Intelligent device failure diagnosis method based on support vector machine |
| CN107144430A (en) * | 2017-06-27 | 2017-09-08 | 电子科技大学 | A kind of Method for Bearing Fault Diagnosis based on incremental learning |
| CN107316046A (en) * | 2017-03-09 | 2017-11-03 | 河北工业大学 | A kind of method for diagnosing faults that Dynamic adaptiveenhancement is compensated based on increment |
-
2018
- 2018-02-26 CN CN201810159234.9A patent/CN108376264A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1563857A (en) * | 2004-03-24 | 2005-01-12 | 上海理工大学 | Double-temp-raising produced overcooling absorption and single effect, afterheat absorbing refrgerating system |
| US20130268469A1 (en) * | 2012-04-06 | 2013-10-10 | Applied Materials, Inc. | Increasing signal to noise ratio for creation of generalized and robust prediction models |
| CN104462846A (en) * | 2014-12-22 | 2015-03-25 | 山东鲁能软件技术有限公司 | Intelligent device failure diagnosis method based on support vector machine |
| CN107316046A (en) * | 2017-03-09 | 2017-11-03 | 河北工业大学 | A kind of method for diagnosing faults that Dynamic adaptiveenhancement is compensated based on increment |
| CN107144430A (en) * | 2017-06-27 | 2017-09-08 | 电子科技大学 | A kind of Method for Bearing Fault Diagnosis based on incremental learning |
Non-Patent Citations (2)
| Title |
|---|
| 孔筱筱: "基于增量学习支持向量机集成算法的轨道电路故障诊断方法", 《哈尔滨铁道科技》 * |
| 滕月阳等: "一种新的支持向量机增量学习算法", 《计算机工程与应用》 * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109190719A (en) * | 2018-11-30 | 2019-01-11 | 长沙理工大学 | Support vector machines learning method, device, equipment and computer readable storage medium |
| CN109540562A (en) * | 2018-12-12 | 2019-03-29 | 上海理工大学 | A kind of water cooler method for diagnosing faults |
| CN110263846A (en) * | 2019-06-18 | 2019-09-20 | 华北电力大学 | The method for diagnosing faults for being excavated and being learnt based on fault data depth |
| CN110702438B (en) * | 2019-09-23 | 2020-09-18 | 天津大学 | Centrifugal water chilling unit and fresh air system combined fault diagnosis method |
| CN110702438A (en) * | 2019-09-23 | 2020-01-17 | 天津大学 | Centrifugal water chilling unit and fresh air system combined fault diagnosis method |
| CN110705631A (en) * | 2019-09-27 | 2020-01-17 | 中国船舶工业系统工程研究院 | SVM-based bulk cargo ship equipment state detection method |
| CN110705631B (en) * | 2019-09-27 | 2022-12-09 | 中国船舶工业系统工程研究院 | SVM-based bulk cargo ship equipment state detection method |
| CN111337263A (en) * | 2020-02-12 | 2020-06-26 | 中国民航大学 | Fault diagnosis method for engine turbine disk |
| CN112964469A (en) * | 2021-02-28 | 2021-06-15 | 哈尔滨理工大学 | Online fault diagnosis method for rolling bearing under variable load of transfer learning |
| CN114120592A (en) * | 2021-10-15 | 2022-03-01 | 青岛海尔空调电子有限公司 | Method and device for fault alarm, electronic equipment and storage medium |
| CN114120592B (en) * | 2021-10-15 | 2023-10-20 | 青岛海尔空调电子有限公司 | Method and device for fault alarm, electronic equipment and storage medium |
| CN114429153A (en) * | 2021-12-31 | 2022-05-03 | 苏州大学 | Lifetime learning-based gearbox increment fault diagnosis method and system |
| CN114429153B (en) * | 2021-12-31 | 2023-04-28 | 苏州大学 | Gear box increment fault diagnosis method and system based on life learning |
| CN115127192A (en) * | 2022-05-20 | 2022-09-30 | 中南大学 | Semi-supervised water chilling unit fault diagnosis method and system based on graph neural network |
| CN115127192B (en) * | 2022-05-20 | 2024-01-23 | 中南大学 | Semi-supervised water chilling unit fault diagnosis method and system based on graph neural network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108376264A (en) | A kind of handpiece Water Chilling Units method for diagnosing faults based on support vector machines incremental learning | |
| Li et al. | An improved fault detection method for incipient centrifugal chiller faults using the PCA-R-SVDD algorithm | |
| Li et al. | A semi-supervised approach to fault detection and diagnosis for building HVAC systems based on the modified generative adversarial network | |
| Li et al. | A data-driven strategy for detection and diagnosis of building chiller faults using linear discriminant analysis | |
| Zhao et al. | Artificial intelligence-based fault detection and diagnosis methods for building energy systems: Advantages, challenges and the future | |
| Guo et al. | Deep learning-based fault diagnosis of variable refrigerant flow air-conditioning system for building energy saving | |
| CN108491580A (en) | A kind of equipment fault diagnosis apparatus and system | |
| CN107844799B (en) | Water chilling unit fault diagnosis method of integrated SVM (support vector machine) mechanism | |
| Liu et al. | Transfer learning-based strategies for fault diagnosis in building energy systems | |
| Li et al. | An improved decision tree-based fault diagnosis method for practical variable refrigerant flow system using virtual sensor-based fault indicators | |
| Zhou et al. | A comparison study of basic data-driven fault diagnosis methods for variable refrigerant flow system | |
| CN113792762B (en) | Water chilling unit fault diagnosis method, system and medium based on Bayesian optimization LightGBM | |
| Han et al. | Important sensors for chiller fault detection and diagnosis (FDD) from the perspective of feature selection and machine learning | |
| Li et al. | Interpretation of convolutional neural network-based building HVAC fault diagnosis model using improved layer-wise relevance propagation | |
| Zhou et al. | Data-driven fault diagnosis for residential variable refrigerant flow system on imbalanced data environments | |
| CN109447236A (en) | A kind of method for diagnosing faults of hybrid vehicle heat management system | |
| CN114484731A (en) | Method and device for diagnosing faults of central air conditioner based on stacking fusion algorithm | |
| CN112097365A (en) | Air conditioner fault detection and identification method and device based on prediction and classification model | |
| Sun et al. | Optimization of support vector regression model based on outlier detection methods for predicting electricity consumption of a public building WSHP system | |
| Li et al. | Diagnosis for multiple faults of chiller using ELM-KNN model enhanced by multi-label learning and specific feature combinations | |
| CN112101519B (en) | Converter valve cooling margin prediction operation and maintenance method and system based on LSTM | |
| CN112990258A (en) | Fault diagnosis method and system for water chilling unit | |
| CN114169254A (en) | Abnormal energy consumption diagnosis method and system based on short-term building energy consumption prediction model | |
| CN117032120A (en) | Integrated intelligent cloud control system and control method for air compression station | |
| Behravan et al. | Deep learning application in mechatronics systems’ fault diagnosis, a case study of the demand-controlled ventilation and heating system |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180807 |
|
| RJ01 | Rejection of invention patent application after publication |