CN107015486A - A kind of air-conditioner water system regulating valve intelligent fault diagnosis method - Google Patents
A kind of air-conditioner water system regulating valve intelligent fault diagnosis method Download PDFInfo
- Publication number
- CN107015486A CN107015486A CN201710260512.5A CN201710260512A CN107015486A CN 107015486 A CN107015486 A CN 107015486A CN 201710260512 A CN201710260512 A CN 201710260512A CN 107015486 A CN107015486 A CN 107015486A
- Authority
- CN
- China
- Prior art keywords
- model
- regulating valve
- fault
- parameter
- 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
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B17/00—Systems involving the use of models or simulators of said systems
- G05B17/02—Systems involving the use of models or simulators of said systems electric
Abstract
The present invention discloses a kind of air-conditioner water system regulating valve intelligent fault diagnosis method, based on the air-conditioner water system regulating valve for carrying out fault detection and diagnosis that SVMs is theoretical, so as to carry out the detection and diagnosis of failure when air-conditioner water system regulating valve incipient failure occurs.So as to solve the limitation in terms of the Machine Learning Problems under high-dimensional problem, nonlinear problem, Small Sample Size;And solve air-conditioner water system regulating valve fault accommodation sensitiveness, accuracy problem at this stage.
Description
Technical field
Type of the present invention is related to regulating valve fault diagnosis and detection technique field, and system is controlled in particular to building automatic
System.
Background technology
Lasting highest attention with China to building energy utilization rate, finds effective method and improves building energy
Utilization rate, reduction building energy consumption it is extremely urgent.In the Practical Project operation of building, even if experienced more perfect building
Operation debugging, can also be gradually deviated from optimal operational effect in actual motion.
Air-conditioner water system is responsible for building the regulation of heating ventilation air-conditioning system cooling and heating load, and the energy consumption to heating ventilation air-conditioning system is played
Key effect.And in air-conditioner water system, valve is then very important control unit, it is used for changing the path of pipe-line system
Section and media flow direction, with blocking, non-return, regulation, peace congruous function.
When air-conditioner water system regulating valve breaks down, the execution of system control logic is influenced whether, end water system is caused
The serious water conservancy of system is unbalance, while can not ensure the defencive function to Cooling and Heat Source unit, the normal fortune of unit can be influenceed when serious
When turning and service life, unit operation security and indoor comfort are not only influenceed, and increase cost of equipment and whole system
The energy consumption of system, causes energy waste, so as to cause unnecessary economic loss.Therefore, air-conditioner water system valve is run in real time
State is monitored, and prevents the generation of failure using reliable fault detection and diagnosis strategy, high all the time to air-conditioner water system
Effect operation has huge realistic meaning.
The traditional approach of fault detection and diagnosis (FDD) relies on the professional standing and experience of attendant, take time and effort and
Higher to attendant's competency profiling, reliability and sustainability can not meet the requirement of modern society.
At present, the fault diagnosis technology based on intelligent theory is the main flow of FDD system development.It is extensive at this stage
The FDD methods of receiving have three classes, the i.e. method for diagnosing faults based on analytic modell analytical model, the method for diagnosing faults based on signal transacting with
And Knowledge based engineering method for diagnosing faults.
The probability that gradual soft fault occurs for air-conditioner water system regulating valve is larger, can influence system running state, day accumulates the moon
It is tired even to damage whole valve body.But, the fault detection and diagnosis of valve is carried out on the basis of raw water system pipeline is not destroyed
One of problem that always air-conditioner water system is debugged.
Recent study personnel propose a variety of failsafe valve diagnosis and detection method, and a kind of trend is in pipe-line system
The special sensor of increase and detection device carry out the detection of failure, and this detection method is mostly for detecting valve body in itself
Failure, such as reveal, block etc..
Another detection trend is developed to Intelligent fault diagnosis and detection method, and this detection method is to fault type
Differentiation more fully, be mainly used in the related Valve controlling breakdown judge of executing agency.
With continuous maximization, complication and the non-linearization of modern comfort, it is difficult often or can not sets up system essence
True mathematical modeling, so as to greatly limit the promotion and application of the method for diagnosing faults based on analytic modell analytical model.Solving higher-dimension
Machine Learning Problems under degree problem, nonlinear problem, Small Sample Size have significant limitation, it is difficult to avoid neutral net
Structure choice and local minimum point's problem.And diagnostic method at this stage is higher to the dependence of data prediction, sensitivity with
And the degree of accuracy is relatively low.
The content of the invention
For above-mentioned the deficiencies in the prior art, type patent technical problem to be solved of the present invention is:1. higher-dimension is being solved
The limitation in terms of Machine Learning Problems under degree problem, nonlinear problem, Small Sample Size;2. air-conditioning water system at this stage is solved
Regulating valve fault accommodation of uniting sensitiveness, accuracy problem.
To achieve the above object, the present invention is based on the theoretical air conditioner water for carrying out fault detection and diagnosis of SVMs
System fading margin valve, so as to carry out the detection and diagnosis of failure when air-conditioner water system regulating valve incipient failure occurs, that is, discloses one kind
Air-conditioner water system regulating valve intelligent fault diagnosis method, it is characterised in that comprise the following steps:
1) air-conditioner water system model is set up
Simulation model is set up according to the pipe network system of the water system of actual air-conditioning system in building;
2) actual condition parameter collection, characterisitic parameter are extracted
According to step 1) described in simulation model, test setting nominal situation and several regulating valve fault condition are (delayed
Property, controlling dead error, band out of control, viscosity, leakage, blocking) under can characterization failure characterisitic parameter:Controller output signal, tune
Save flow after the valve position feedback signal and regulating valve of valve;
3) the Fault Model training based on SVM
3-1) selecting step 2) the valve position feedback signal that obtains regulating valve under several regulating valve fault condition merges
It is used as a fault condition data set;
Selecting step 2) the valve position feedback signal of regulating valve under several regulating valve nominal situation is obtained as normal
Floor data collection;
The input that fault condition data set and nominal situation data set are trained as SVM fault diagnosis model;
Data are carried out equidistantly to be split as two parts, a part of training set as svm classifier model, so as to be adjusted
Valve Fault Model is saved, another part is then as model measurement collection.
3-2) data in training set and test set are carried out after visualization processing, the normalization letter carried using MATLAB
Several normalizeds that data are carried out with deviation standardization;
3-3) after the normalized of data is completed, the parameter for proceeding by SVMs failure modes model is excellent
Select process.The purpose of preferred process is the nuclear parameter g and penalty factor c for obtaining optimal SVM;
3-4) according to step 3-3) optimized parameter that obtains carries out the training of model, and training process passes through LIBSVM tool boxes
Complete, finally give Fault Model;
4) the fault diagnosis model training based on SVM
4-1) by step 2) obtained carry out the data of every kind of operating mode be equidistantly split as two parts, a part of conduct
The training set of svm classifier model, so as to obtain regulating valve Fault Model, another part is then as model measurement collection.
4-2) data in training set and test set are carried out after visualization processing, the normalization letter carried using MATLAB
Several normalizeds that data are carried out with deviation standardization;
4-3) after the normalized of data is completed, the parameter for proceeding by SVMs failure modes model is excellent
Select process.The purpose of preferred process is the nuclear parameter g and penalty factor c for obtaining optimal SVM, to obtain failure the most accurate
Detect disaggregated model;
4-4) according to step 4-3) optimized parameter that obtains carries out the training of model, and training process passes through LIBSVM tool boxes
Complete, finally give fault diagnosis model;
5) the fault diagnosis model training based on SVM, obtains diagnostic result
Gather the duty parameter and characterisitic parameter of the water system of air-conditioning system;According to step 3) obtained fault detect mould
Type, judges whether the water system of central air conditioner system occurs regulating valve failure:If so, by step 4) obtain fault detect mould
Type detects fault type.
What deserves to be explained is, SVMs (SVM) is built upon the structural risk minimization of machine Learning Theory
On, its main thought is to be directed to binary classification problems, and a hyperplane is found in high-dimensional space as the segmentation of two classes,
To ensure minimum classification error rate, and one important advantage of SVMs is exactly that can handle the feelings of linearly inseparable
Condition.
6 kinds of most common failures of regulating valve in air-conditioner water system have:Hysteresis quality, controlling dead error, band out of control, viscosity, let out
Dew, blocking.Wherein, the first four kinds nonlinear fault features for valve actuator, latter two is valve body and valve seat failure.
This patent considers that when regulating valve has failure corresponding change can also occur for each characterisitic parameter, in order to improve
The operability and applicability of fault detect flow, the 3 typical valves most easily measured using air-conditioner water system regulating valve
Energy parameter is used as the characterisitic parameter for indicating failsafe valve:Flow after controller output signal, valve position feedback signal, regulating valve (or
Pressure difference before and after person's regulating valve).FDD based on SVM is set up according to the analogue data under nominal situation and under all kinds of fault conditions
Disaggregated model, the disaggregated model that training data is obtained is classified to test data, and event is carried out to valve member according to classification results
Barrier detection, judges whether valve breaks down.
When classification and Detection Model Diagnosis valve characteristic parameter drift-out nominal situation, after judging that valve breaks down, further according to
Fault characteristic supplemental characteristic under all kinds of fault conditions sets up the fault diagnosis regression model based on SVM, and model predication value
As benchmark, the measured value of estimated performance parameter and the residual error of a reference value are gone forward side by side so as to be distributed failure judgement type by residual error t
Row diagnosis.So far, a kind of air-conditioner water system intelligent diagnostics regulating valve can just be arrived.
Brief description of the drawings
Fig. 1 is Failure Diagnostic Code figure.
Fig. 2 is building air conditioning water system model.
In Fig. 2, dotted line frame inner assembly designates position of the regulating valve in air-conditioner water system described in type patent of the present invention.Member
Part 8 is water pump;Element 9-21 is ducting module and end resistance devices module;Element 26-29 is ball valve (regulating valve), element 53
Diaphragm valve is bypassed for pressure difference, element 50,51 is butterfly valve;Element 30,32 is signal measuring apparatus;Element 33,35,36,49 is valve
Controller, element 31,48 is signal controller;Element 25 is level pressure pressure source.
Fig. 3 is fault diagnosis technology route map.
Embodiment
With reference to embodiment, the invention will be further described, but should not be construed above-mentioned subject area of the invention only
It is limited to following embodiments.Without departing from the idea case in the present invention described above, according to ordinary skill knowledge and used
With means, various replacements and change are made, all should be included within the scope of the present invention.
A kind of air-conditioner water system regulating valve intelligent fault diagnosis method, it is characterised in that comprise the following steps:
1) air-conditioner water system model is set up
Simulation model, the foundation of model are set up according to the pipe network system of actual constant-primary-flow air-conditioner water system in building
Need to ignore some parts little on result influence and parameter.(Fig. 2 is the building air conditioning water system model of the present embodiment)
2) actual condition parameter collection, characterisitic parameter are extracted
According to step 1) described in simulation model, test setting nominal situation, delayed fault condition, viscous fault condition,
Tape jam operating mode out of control, controlling dead error fault condition, leakage fault condition, under blocking fault condition can characterization failure characteristic ginseng
Number:Flow (or pressure difference before and after regulating valve) after controller output signal, the valve position feedback signal of regulating valve and regulating valve.And root
Faulty tag is added to initial data according to fault condition type, such as 0 nominal situation, 1 representing fault operating mode is represented.
Show that data are handled by Flowmaster simulations, it is met under MATLAB LIBSVM tool boxes to training
The file format requirements of data and inspection data.The data that every kind of operating mode is included, are added according to fault type to initial data
Faulty tag, e.g., 0 represents nominal situation, 1 representing fault operating mode.
3) the Fault Model training based on SVM
3-1) selecting step 2) the valve position feedback signal that obtains regulating valve under several regulating valve fault condition merges
It is used as a fault condition data set;
Selecting step 2) the valve position feedback signal of regulating valve under several regulating valve nominal situation is obtained as normal
Floor data collection;
Data are carried out equidistantly to be split as two parts, a part of training set as svm classifier model, according to training set
The training of svm classifier model is carried out with the faulty tag of division, so as to obtain regulating valve Fault Model, another part is then made
For model measurement collection.
The input that fault condition data set and nominal situation data set are trained as SVM Fault Model;
3-2) data in training set and test set are carried out after visualization processing, the normalization letter carried using MATLAB
Several normalizeds that data are carried out with deviation standardization, the method is the linear transformation to initial data, maps end value
To between [0-1].
3-3) after the normalized of data is completed, the parameter for proceeding by SVMs failure modes model is excellent
Select process.The purpose of preferred process is the nuclear parameter g and penalty factor c for obtaining optimal SVM.What deserves to be explained is, for core
Parameter g and penalty factor c selection, this patent carry out optimization of parameters using cross-validation method (K-CV).For given interval interior
Nuclear parameter g and penalty factor c, training sample is divided into three groups, i.e. K=3 first for each group of collocation, then will be each
Subset data makees one-time authentication collection respectively, while remaining two subset data can so obtain 3 models as training set,
The average of so final classification accuracy rate of their checking collection is exactly accuracy rate when grader takes the model parameter.Then
The collocation for choosing next group of nuclear parameter g and penalty factor c proceeds cross validation, until complete in given interval all takes
With type.
3-4) according to step 3-3) optimized parameter that obtains carries out the training of model, and training process passes through LIBSVM tool boxes
Complete.Afterwards, failure detection result is obtained.
4) the fault diagnosis model training based on SVM
4-1) by step 2) obtained carry out the data of every kind of operating mode be equidistantly split as two parts, a part of conduct
The training set of svm classifier model, according to training set and valve position feedback signal, so that regulating valve fault diagnosis model is obtained, it is another
Part is then as model measurement collection.
4-2) data in training set and test set are carried out after visualization processing, the normalization letter carried using MATLAB
Several normalizeds that data are carried out with deviation standardization, the method is the linear transformation to initial data, maps end value
To between [0-1].
4-3) after the normalized of data is completed, the parameter for proceeding by SVMs failure modes model is excellent
Select process.The purpose of preferred process is the nuclear parameter g and penalty factor c for obtaining optimal SVM, to obtain failure the most accurate
Detect disaggregated model.As previously mentioned, for nuclear parameter g and penalty factor c selection, entered herein using cross-validation method (K-CV)
Row optimization of parameters.It is equal first by training sample for each group of collocation for the nuclear parameter g and penalty factor c in given interval
It is divided into three groups, i.e. K=3, each subset data is then made into one-time authentication collection respectively, while remaining two subset data conduct
Training set, can so obtain 3 models, then the average of the final classification accuracy rate of their checking collection is exactly that grader takes
The accuracy rate during model parameter.Then the collocation for choosing next group of nuclear parameter g and penalty factor c proceeds cross validation,
Until completing all collocation types in given interval.
4-4) according to step 4-3) optimized parameter that obtains carries out the training of model, and training process passes through LIBSVM tool boxes
Complete.Finally give fault diagnosis model;
5) the fault diagnosis model training based on SVM, obtains diagnostic result
Gather the duty parameter and the characterisitic parameter (data gathered with step 2 of the water system of water-cooled central air-conditioning
Data structure it is identical);According to step 3) obtained Fault Model, judge whether the water system of central air conditioner system occurs
Regulating valve failure:If so, by step 4) fault diagnosis model that obtains detects fault type.
Below by taking the delayed failure of regulating valve as an example, the air-conditioner water system regulating valve fault diagnosis model based on SVM is introduced:Press
According to step 2) the performance parameter data of model training collection and test set under fault condition are taken, will be according to step 4) training completes
The predicted value of diagnostic model draws test set as benchmark, the test set data and the residual error of a reference value of estimated performance parameter
Root-mean-square error and coefficient correlation between data and a reference value, are used as the basis of fault diagnosis.
Bring the characteristic parameter data under nominal situation and other 5 kinds of failures into the fault diagnosis as new test set again
Model, completes the diagnostic process of this failure.
, can be according to delayed fault diagnosis model according to the root-mean-square error and the numerical value of coefficient correlation obtained under each operating mode
Whether the deviation of predicted value and actual measured value currently occurs delayed failure to diagnose.Now, fault diagnosis result is just obtained.
The diagnosis of other fault conditions is carried out with above-mentioned steps.Diagnostic techniques route map is shown in Fig. 3.
Claims (1)
1. a kind of air-conditioner water system regulating valve intelligent fault diagnosis method, it is characterised in that comprise the following steps:
1) the air-conditioner water system model is set up
Simulation model is set up according to the pipe network system of the water system of actual air-conditioning system in building;
2) the actual condition parameter collection, characterisitic parameter are extracted
According to step 1) described in simulation model, test setting nominal situation and several regulating valve fault condition are repeatedly extracted
Under various operating modes, the characterisitic parameter of characterization failure;The characterisitic parameter of described characterization failure includes controller output signal, regulation
Flow after the valve position feedback signal and regulating valve of valve;
3) the Fault Model training based on SVM
3-1) selecting step 2) regulating valve under whole regulating valve fault conditions for being obtained valve position feedback signal, as
One fault condition data set;
Selecting step 2) regulating valve under the regulating valve nominal situation that is obtained valve position feedback signal, as nominal situation
Data set;
The input that fault condition data set and nominal situation data set are trained as SVM fault diagnosis model;
Data are carried out equidistantly to be split as two parts, a part of training set as svm classifier model, so as to obtain regulating valve
Fault Model, another part is then as model measurement collection.
3-2) data in training set and test set are carried out after visualization processing, the normalized function pair carried using MATLAB
Data carry out the normalized of deviation standardization;
3-3) after the normalized of data is completed, the preferred mistake of parameter of SVMs failure modes model is proceeded by
Journey.The purpose of preferred process is the nuclear parameter g and penalty factor c for obtaining optimal SVM;
3-4) according to step 3-3) optimized parameter that obtains carries out the training of model, and training process is complete by LIBSVM tool boxes
Into finally giving Fault Model;
4) the fault diagnosis model training based on SVM
4-1) by step 2) obtained carry out the data of every kind of operating mode be equidistantly split as two parts, and a part is used as SVM
The training set of disaggregated model, so as to obtain regulating valve Fault Model, another part is then as model measurement collection.
4-2) data in training set and test set are carried out after visualization processing, the normalized function pair carried using MATLAB
Data carry out the normalized of deviation standardization;
4-3) after the normalized of data is completed, the preferred mistake of parameter of SVMs failure modes model is proceeded by
Journey.The purpose of preferred process is the nuclear parameter g and penalty factor c for obtaining optimal SVM, to obtain fault detect the most accurate
Disaggregated model;
4-4) according to step 4-3) optimized parameter that obtains carries out the training of model, and training process is complete by LIBSVM tool boxes
Into finally giving fault diagnosis model;
5) the fault diagnosis model training based on SVM, obtains diagnostic result
Gather the duty parameter and characterisitic parameter of the water system of water-cooled central air-conditioning;According to step 3) inspection of obtained failure
Model is surveyed, judges whether the water system of air-conditioning system occurs regulating valve failure:If so, by step 4) obtain fault diagnosis mould
Type detects fault type.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710260512.5A CN107015486A (en) | 2017-04-20 | 2017-04-20 | A kind of air-conditioner water system regulating valve intelligent fault diagnosis method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710260512.5A CN107015486A (en) | 2017-04-20 | 2017-04-20 | A kind of air-conditioner water system regulating valve intelligent fault diagnosis method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107015486A true CN107015486A (en) | 2017-08-04 |
Family
ID=59447292
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710260512.5A Pending CN107015486A (en) | 2017-04-20 | 2017-04-20 | A kind of air-conditioner water system regulating valve intelligent fault diagnosis method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107015486A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108388227A (en) * | 2018-01-30 | 2018-08-10 | 杭州深渡科技有限公司 | Air-conditioning remote fault diagnosis method and system |
CN108895195A (en) * | 2018-07-23 | 2018-11-27 | 中国矿业大学 | A kind of control method of pneumatic control valve intelligent Fault Diagnose Systems |
CN109163415A (en) * | 2018-07-23 | 2019-01-08 | 成都慧云阵信息技术有限公司 | Central air conditioning water system adjustment method, method for diagnosing faults, running optimizatin method |
CN112989522A (en) * | 2021-05-10 | 2021-06-18 | 创新奇智(成都)科技有限公司 | Model training method, fault prediction method and device and electronic equipment |
CN115095953A (en) * | 2022-06-16 | 2022-09-23 | 青岛海信日立空调系统有限公司 | Training method and device of fault diagnosis model |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110016070A1 (en) * | 2009-07-14 | 2011-01-20 | Nikovski Daniel N | Method for Predicting Future Environmental Conditions |
CN104573740A (en) * | 2014-12-22 | 2015-04-29 | 山东鲁能软件技术有限公司 | SVM classification model-based equipment fault diagnosing method |
CN105627504A (en) * | 2015-05-28 | 2016-06-01 | 重庆大学 | Energy consumption estimation method for water chilling unit of variable-air-volume central air conditioning system based on support vector machine |
-
2017
- 2017-04-20 CN CN201710260512.5A patent/CN107015486A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110016070A1 (en) * | 2009-07-14 | 2011-01-20 | Nikovski Daniel N | Method for Predicting Future Environmental Conditions |
CN104573740A (en) * | 2014-12-22 | 2015-04-29 | 山东鲁能软件技术有限公司 | SVM classification model-based equipment fault diagnosing method |
CN105627504A (en) * | 2015-05-28 | 2016-06-01 | 重庆大学 | Energy consumption estimation method for water chilling unit of variable-air-volume central air conditioning system based on support vector machine |
Non-Patent Citations (2)
Title |
---|
廖袖锋等: "重庆某综合楼空调水系统控制策略优化", 《建设科技》 * |
罗天: "基于支持向量机的空调水系统调节阀FDD方法研究", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108388227A (en) * | 2018-01-30 | 2018-08-10 | 杭州深渡科技有限公司 | Air-conditioning remote fault diagnosis method and system |
CN108388227B (en) * | 2018-01-30 | 2020-09-04 | 杭州深渡科技有限公司 | Remote fault diagnosis method and system for air conditioner |
CN108895195A (en) * | 2018-07-23 | 2018-11-27 | 中国矿业大学 | A kind of control method of pneumatic control valve intelligent Fault Diagnose Systems |
CN109163415A (en) * | 2018-07-23 | 2019-01-08 | 成都慧云阵信息技术有限公司 | Central air conditioning water system adjustment method, method for diagnosing faults, running optimizatin method |
CN108895195B (en) * | 2018-07-23 | 2019-11-26 | 中国矿业大学 | A kind of control method of pneumatic control valve intelligent Fault Diagnose Systems |
CN112989522A (en) * | 2021-05-10 | 2021-06-18 | 创新奇智(成都)科技有限公司 | Model training method, fault prediction method and device and electronic equipment |
CN112989522B (en) * | 2021-05-10 | 2021-07-30 | 创新奇智(成都)科技有限公司 | Model training method, fault prediction method and device and electronic equipment |
CN115095953A (en) * | 2022-06-16 | 2022-09-23 | 青岛海信日立空调系统有限公司 | Training method and device of fault diagnosis model |
CN115095953B (en) * | 2022-06-16 | 2024-04-09 | 青岛海信日立空调系统有限公司 | Training method and device for fault diagnosis model |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107015486A (en) | A kind of air-conditioner water system regulating valve intelligent fault diagnosis method | |
Zhao et al. | Diagnostic Bayesian networks for diagnosing air handling units faults–Part II: Faults in coils and sensors | |
US7444251B2 (en) | Detecting and diagnosing faults in HVAC equipment | |
Li et al. | A sensor fault detection and diagnosis strategy for screw chiller system using support vector data description-based D-statistic and DV-contribution plots | |
Wang et al. | A robust fault detection and diagnosis strategy for multiple faults of VAV air handling units | |
CN107111286B (en) | Automated functional testing for diagnostics and control | |
Wang et al. | Fault-tolerant control for outdoor ventilation air flow rate in buildings based on neural network | |
CN104390657B (en) | A kind of Generator Unit Operating Parameters measurement sensor fault diagnosis method and system | |
CN107063663A (en) | A kind of air-conditioner water system intelligent diagnostics regulating valve | |
WO2005111806A3 (en) | Sensor fault diagnostics and prognostics using component model and time scale orthogonal expansions | |
CN102799171A (en) | Detecting anomalies in fault code settings and enhancing service documents using analytical symptoms | |
CN109213127A (en) | A kind of HVAC system gradual failure diagnostic method based on deep learning | |
KR20170031985A (en) | Fault detection and diagnostics method of air-conditioning system | |
Zhao et al. | A proactive fault detection and diagnosis method for variable-air-volume terminals in building air conditioning systems | |
KR102279351B1 (en) | A fault diagnosis method of based ai on optimization of air conditioner | |
CN102110187A (en) | Method and system for diagnosing mixed failure based on PCA and artificial immune system | |
CN110673515A (en) | Monitoring of industrial facilities | |
CN112836396B (en) | Building real-time energy consumption abnormity diagnosis system | |
Wang et al. | Monte Carlo analysis of the effect of uncertainties on model-based HVAC fault detection and diagnostics | |
Zhang et al. | Evaluate the impact of sensor accuracy on model performance in data-driven building fault detection and diagnostics using Monte Carlo simulation | |
EP3627261B1 (en) | Diagnosis system and method using parallel analysis paths | |
CN109084971B (en) | A kind of pneumatic control valve method for diagnosing faults based on particle group optimizing | |
CN104216397B (en) | Failure recognition and detection method for intelligent drive axle system | |
CN105090000B (en) | A kind of method for diagnosing faults of the thick pumped vacuum systems of space simulator | |
CN110705114B (en) | Ventilation fault diagnosis method without training sample |
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: 20170804 |
|
RJ01 | Rejection of invention patent application after publication |