CN108829084A - A kind of Fault diagnosis model distributed intelligence method and device thereof based on particle filter - Google Patents
A kind of Fault diagnosis model distributed intelligence method and device thereof based on particle filter Download PDFInfo
- Publication number
- CN108829084A CN108829084A CN201810640185.0A CN201810640185A CN108829084A CN 108829084 A CN108829084 A CN 108829084A CN 201810640185 A CN201810640185 A CN 201810640185A CN 108829084 A CN108829084 A CN 108829084A
- Authority
- CN
- China
- Prior art keywords
- data
- model
- particle filter
- fault diagnosis
- particle
- 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
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0259—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterized by the response to fault detection
- G05B23/0264—Control of logging system, e.g. decision on which data to store; time-stamping measurements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- 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
- G05B2219/00—Program-control systems
- G05B2219/20—Pc systems
- G05B2219/24—Pc safety
- G05B2219/24065—Real time diagnostics
Abstract
The Fault diagnosis model distributed intelligence method and device thereof based on particle filter that the invention discloses a kind of, technical solution are that normal model and various faults model are established according to system dynamics；Various types of data parameter is obtained by distributed data acquisition；Using the accurate estimating system state parameter of particle filter algorithm；Data parameters after filtering processing are compared with the normal model and various faults model, obtain the operational mode of current system by application mode recognizer.The invention has the advantages that：The acquisition and centralized processing of multivariate data are realized by sensing network, and the intelligent diagnostics of the accurate estimation and failure of monitoring object state are realized based on particle filter algorithm, system can be realized remote real-time monitoring and failure indication, effectively promote the application range and diagnostic level of existing fault diagnosis system.
Description
Technical field
The invention belongs to fault diagnosis fields, more particularly to a kind of Fault diagnosis model distributed intelligence based on particle filter
Method and device thereof.
Background technique
In modern industry, most of monitoring and controlling mechanism are on the basis of assuming that system mode is accurate measurable
Building.However, internal system state is often difficult to accurately obtain by sensor, actual measuring system is that have random error
, so that measurement vector is also contained random quantity, cannot directly find out state true value by preferably measuring equation.
As most important state estimation tool, filter had been subjected to from onrecurrent to recurrence, frequency domain to time domain, non-flat
Development course of the steady random process to state-space model.Nowadays, there are numerous filtering algorithms for state estimation, it is most typical
Have：Kalman filtering (Kalman Filter, KF), Extended Kalman filter (Extended Kalman Filter, EKF), nothing
Mark Kalman filtering (Unscented Kalman Filter, UKF) and particle filter (Particle Filter, PF).
Particle filter algorithm can be described as a kind of filtering method being most taken seriously in contemporary nonlinear filtering, it is in each neck
Great effect is suffered from domain, particle filter algorithm is integrated in state estimation by recent domestic scholars, constitutes base
In the state estimation of particle filter, economics, control science, aerospace, information science, electronic technology are pushed in this approach
Fast development with progress.For example, being applied to the prediction of economic data in Science of Economics；Its quilt in robot field
For to robot Global localization and tracking；It is used for radar, guided missile tracking airflight object in military field；Especially exist
It is used for fault diagnosis and the warning etc. of nonlinear system in industrial process and monitoring of tools.
Based on the extremely strong practicability of particle filter, numerous studies have been carried out to this, propose many by domestic and foreign scholars
Efficient algorithm for tracking.These methods can be mainly divided into two classes：(1) based drive method：It is strong according to certain
The point with Movement consistency in a period of time is classified as one kind by algorithm, such as optical flow method and method of characteristic point, but calculation amount compared with
Greatly.(2) based on the method for model：The main tracking that target is completed according to high-rise semantic expressiveness and knowledge description.Utilize mesh
The difference of message part in mark can be divided into based on object boundary, based on the method for target area.But due to the information of target itself
It is more, simplification is such as not added, a large amount of operations when will inevitably bring information matches.Therefore, very high for requirement of real-time
Moving target tracking technique for, how to choose clarification of objective information, and simplifying operation under the premise of reliable is mesh
Mark the key of tracking.
Failure usually means that the come to nothing situation result of output or output of system is unsatisfactory for routine request, and system is in
Abnormal operation.The fault diagnosis of system status is that determining system is in normal or abnormal patterns first, then carries out event
Barrier positioning and troubleshooting.Main task includes：Examined system status information acquisition, failure mode is screened, trend analysis research
With diagnosis decision.Fault diagnosis is generally possible to be summarised as following several forms：According to the method for model, according to the method for knowledge with
According to the method etc. of signal processing.According to the method that the method for model can be classified as expert system according to the difference of model, mould
Paste knows method for distinguishing, Neural Networks Learning Algorithm, the method for fuzzy reasoning；Observational measurement can be classified as according to the difference of symptom
Method, qualitative simulation method, knowledge measurement device method；System and the fault parameter estimation technique can be classified as according to parameter Estimation difference；Root
Viewer method and filter method can be classified as according to state estimation difference.
In terms of the algorithm for pattern recognition of fault diagnosis, there are the methods of neural network, support vector machine, Yang Yu et al. at present
It proposes based on EMD (Empirical Mode Decomposition) and VPMCD (Variable Predictive Model
Based Class Discriminate) application of the method for diagnosing faults in rolling bearing, avoid the knot of neural network
The select permeability of structure and type and support vector machines kernel function and its parameter, but complex and calculation amount is larger, to reality
It cannot be applied very well in the more demanding system of when property.
In addition, Chinese Patent Application No. CN201510031512.9 discloses a kind of rolling using particle filter and spectrum kurtosis
Dynamic bearing method for diagnosing faults equally exists diagnostic method complexity and computationally intensive problem.
Therefore, it is necessary to be improved to this.
Summary of the invention
The purpose of the invention is to overcome shortcoming and defect of the existing technology, and provide a kind of based on particle filter
Fault diagnosis model distributed intelligence method and apparatus, breach the limitation of wire transmission, improve the accurate of method for diagnosing faults
Property and real-time.
To achieve the above object, the technical scheme is that S01：According to system dynamics establish normal model and
Various faults model；
S02：Various types of data parameter is obtained by distributed data acquisition；
S03：Using the accurate estimating system state parameter of particle filter algorithm；
S04：Application mode recognizer, by the data parameters and the normal model and various faults mould after filtering processing
Type compares, and obtains the operational mode of current system, diagnose to the malfunction of system.
Further setting is that the step S01 includes：
The kinetic model of diagnosed object, including building normal model are established by system and structural analysis；
For fault model, first deduction system failure evolution process, typical fault Dynamic Evolution Characteristics, analysis system are refined
System mixes influent factor and propagation characteristic, according to system failure propagation law, establishes typical fault and propagates description relational expression, thus
The various faults model of object is established, and finally realizes the building of system state space.
Further setting is that the step S02 includes：
Sensor is distributed in each data collection point, timing acquiring data, data collected are with defined report
Literary format is packaged, then sends coordinator for the data of acquisition.
It is to be sent to coordinator by wireless receiving and dispatching radio frequency unit with wireless transmission method that further setting, which is the data,
The coordinator starts wireless communication networks by one channel of selection and a network ID.
Further setting is that the step S03 includes：
The algorithm detailed process is indicated with following pseudocode：
At the time of FOR k=1,2 ..., K are that each data collection point corresponds to；
(1) particle initializes：
At the k=0 moment, by known prior probability p (x0) generate populationAll particle weights are
(2) sequential importance sampling：
1. sampling：
The k moment samples the particle in significance distribution functionParticle assembly is at this time：
2. weight is estimated：
Acquire the observation z at k momentkAfterwards, the estimated value of available weights of importance：
WhereinFor the reference distribution of importance probability density；
3. weight normalizes：
(3) resampling：
On the basis of obtaining particle weights by Step2, the lesser particle of weight is weeded out, replicates the biggish grain of weight
Son finally obtains new particle collection
(4) it exports：
Calculate the state estimation at k moment：
This step can accurate estimating system state parameter.
Further setting is that the step S04 includes：
Data parameters after filtering processing are compared with the normal model and various faults model, specific steps are such as
Shown in lower：
Based on the obtained data of particle filter algorithm, by all kinds of state models of the data at preceding δ moment and corresponding moment
Calculating difference is simultaneously summed, and in obtained each and value, model corresponding to absolute value minimum is the current state in which of system；
Wherein x indicates state model classification, such as normal condition, malfunction 1,2 ..., fiAfter indicating the filtering of the i-th moment
Data value, mx,iFor the state value at the i-th moment of model x；
The method for distinguishing of sentencing of system running pattern is that time of day curve is corresponding with each failure or normal condition model
Curve compares, and is present mode of operation corresponding to most identical curve.
The present invention also provides a kind of Fault diagnosis model distributed intelligence device based on particle filter, it is characterised in that：The dress
It sets and fault diagnosis is carried out by the method, which includes：Distributed data acquisition system, coordinator, multiple routers
And multiple terminal devices；
Distributed data acquisition system includes multiple sensors, which is distributed in each data collection point, is used for
Acquire the data of each data collection point；
The coordinator is used to obtain the network address of terminal device, judges whether it networks, the receiving terminal if networking
The data that equipment is sent otherwise wait for its networking, which starts nothing by one channel of selection and a network ID
Line communication network, the router is for allowing equipment that network is added, realizing multihop routing and assisting the logical of the terminal device
News；Sensor timing acquiring data in the terminal device, then wirelessly it is sent to the coordinator.
Further setting be distributed data acquisition system further include have CC2530 wireless receiving and dispatching radio frequency unit, ZigBee without
Transmission network closes, USB turns serial communication and host computer, and the host computer that data are saved for real-time display is provided on the host computer
Software simultaneously carries out fault diagnosis algorithm calling analysis.
The present invention realizes the acquisition and centralized processing of multivariate data by sensing network, and real based on particle filter algorithm
The intelligent diagnostics of the accurate estimation and failure of existing monitoring object state, system can be realized remote real-time monitoring and failure indication,
Effectively promote the application range and diagnostic level of existing fault diagnosis system.
It is an advantage of the invention that improving the accuracy and real-time of method for diagnosing faults.Pass through data acquisition device timing
It acquires data and passes through ZigBee-network high speed transmission data, by the accurate estimating system state parameter of particle filter algorithm, finally
In host computer interface real-time display dynamic data, reach accurate and real-time purpose.
The present invention devises complex distributions formula data acquisition network and corresponding host computer system based on wireless sensor network
System, distributed, ambulant flexible feature are suitable for Large-scale Mobile sex object, extend the suitable of the intelligent Fault Diagnose Systems
With range, preferably resolve that the distribution of complex electromechanical systems signaling point is wide, monitoring parameters point distance is remote, the moment is kept in motion
Practical difficulty, relatively traditional cable data acquisition system is greatly improved.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention, for those of ordinary skill in the art, without any creative labor, according to
These attached drawings obtain other attached drawings and still fall within scope of the invention.
The flow chart of Fig. 1 Fault diagnosis model distributed intelligence device；
Fault diagnosis model distributed intelligence device of the Fig. 2 based on particle filter；
Fig. 3 data transmission stream journey figure；
The data transmission stream journey figure of Fig. 4 coordinator；
Fig. 5 software principle process；
Fig. 6 distributed data acquisition system general diagram.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, the present invention is made into one below in conjunction with attached drawing
Step ground detailed description.
It as shown in Figures 1 to 6, is a kind of Fault diagnosis model distributed intelligence based on particle filter in the embodiment of the present invention
The building process of device includes：Normal model and various faults model are established according to system dynamics；Pass through distributed number
Various types of data parameter is obtained according to acquisition；Using the accurate estimating system state parameter of particle filter algorithm；Application mode recognizer,
Data parameters after filtering processing are compared with the normal model and various faults model, obtain the operation of current system
Mode.
Step S01：Normal model and various faults model are established according to system dynamics
Diagnosed object kinetic model is established by system and structural analysis；For fault model, first deduction system
Failure evolution process refines typical fault Dynamic Evolution Characteristics, and analysis system mixes influent factor and propagation characteristic, according to system
Fault propagation rule establishes typical fault and propagates description relational expression, to establish the multiple faults evolutionary model of object, and final real
The building of existing system state space.This step is to realize failure in order to which the data parameters after being filtered have model to compare
Mode division.
Step S02：Various types of data parameter is obtained by distributed data acquisition
Sensor is distributed in each data collection point, and timing acquiring data, data collected are with defined message
Format is packaged, then is sent to coordinator by wireless receiving and dispatching radio frequency unit with wireless transmission method.Wherein, coordinator is responsible for
Start whole network, it is also first equipment of network, first selects a channel and a network ID (also referred to as by it
PAN ID, i.e. Personal Area Network ID), subsequent start-up whole network.
Entire distributed data acquisition network is to acquire Various types of data based on ZigBee-network.It is defined in ZigBee-network
Three kinds of logical device types：Coordinator (Coordinator), router (Router) and terminal device (End-Device).
The major function of router is to allow other equipment that network, multihop routing and the communication for assisting sub- terminal device is added.One
ZigBee-network is made of a coordinator and multiple routers and multiple terminal devices.Various kinds of sensors is fixed in terminal device
When acquire data, data collected are packaged with defined message format, then are wirelessly sent to coordination
Device.When data do not acquire, to reduce power consumption, then by chip suspend mode, waken up again when acquisition next time.As shown in Figure 3.
The core of control action is exactly coordinator node in whole system, it needs to obtain the network address of terminal node,
Judge whether it networks, the data that receiving terminal acquisition node is sent if networking otherwise wait for its networking；Again to data
It analyzed, handled, serial ports general's treated data information transfer to computer is then passed through based on RS232 agreement.Its process is such as
Shown in lower Fig. 4.The purpose that wherein serial communication reaches go here and there-and convert, to message transmission rate --- baud rate carries out
Selection and control etc..
Step S03：Using the accurate estimating system state parameter of particle filter algorithm
Algorithm detailed process can be used following pseudocode to indicate：
FOR k=1,2 ...
(1) particle initializes：
At the k=0 moment, by known prior probability p (x0) generate populationAll particle weights are
(2) sequential importance sampling：
1. sampling：
The k moment samples the particle in significance distribution functionParticle assembly is at this time：
2. weight is estimated：
Acquire the observation z at k momentkAfterwards, the estimated value of available weights of importance：
WhereinFor the reference distribution of importance probability density.
3. weight normalizes：
(3) resampling：
On the basis of obtaining particle weights by Step2, the lesser particle of weight is weeded out, replicates the biggish grain of weight
Son finally obtains new particle collection
(4) it exports：
Calculate the state estimation at k moment：
This step can accurate estimating system state parameter.
Step S04：Application mode recognizer obtains the operational mode of current system
Data parameters after filtering processing are compared with the normal model and various faults model, specific steps are such as
Shown in lower：
Based on the obtained data of particle filter algorithm, by all kinds of state models of the data at preceding δ moment and corresponding moment
Calculating difference is simultaneously summed.In obtained each and value, model corresponding to absolute value minimum is the current state in which of system.
Wherein x indicates state model classification, such as normal condition, malfunction 1,2 ..., fiAfter indicating the filtering of the i-th moment
Data value, mx,iFor the state value at the i-th moment of model x.
The basic thought of the differentiation of system running pattern is time of day curve and each failure or normal condition model pair
The curve answered compares, and is present mode of operation corresponding to most identical curve.
In order to clear and intuitive all kinds of required data of monitoring, host computer display interface of the invention is compiled using C# language
It writes.Wherein, the data for needing emphasis to monitor can save the data in designated position equipped with time history plot, and
Auxiliary monitoring only shows current value, and when being greater than given threshold, the green light by each data will become red light and warn.In interface
Serial ports receives baud rate need to be consistent with coordinator transmission rate.
Secondly, acquiring data by the sensor that serial ports receives, it is not real numerical value itself, but contains one
The data packet of the information such as serial frame head, packet length, control command.
It is logical according to above-mentioned rule by the data pack buffer received in array therefore, it is necessary to be parsed to data packet
Process ordered pair data packet carries out classification processing, finally obtains required data, is shown in the corresponding position at interface.Wherein, boundary
Face design cycle includes：Port1 port numbers and baud rate are set；Open Port1 serial ports and read port byte；By data conversion
At 16 system character string forms；The data of reading are analyzed from function is write；By character string display in corresponding position.Such as Fig. 5
It is shown.
Distributed data acquisition system altogether by sensor acquisition equipment physical parameter, CC2530 wireless receiving and dispatching radio frequency unit,
ZigBee wireless transmission gateway, USB turn serial communication, upper computer software real-time display saves data, fault diagnosis algorithm calls
Analysis composition.As shown in Figure 6.
Host computer monitoring software is transmitted to related data in the mobile device for being mounted with Android system by wireless network.
The communication of host computer and mobile phone terminal relies primarily on bluetooth module, realizes that communication function is based on Design of Serial Communication, mobile phone meeting
By send instructions between bluetooth and host computer, then data are sent to by Your Majesty position chance according to the instruction of transmission using serial ports
Bluetooth module then sends data to android mobile terminal by related protocol again.
The most important work of Android client is the data information for receiving pc client and sending, and display data, by data
The form for being processed into dynamic chart is shown, and is provided with a warning line to warn user.Major function and demand tool
Body is as follows：(1) Android interface is established；(2) it establishes Button and jumps interface；(3) analogue data is generated；(4) analogue data is turned
Turn to dynamic chart；(5) data are shown on curve；(6) historical data can check；(7) chart can draw high scaling；(8)
Set up warning line.
Those of ordinary skill in the art will appreciate that implement the method for the above embodiments be can be with
Relevant hardware is instructed to complete by program, the program can be stored in a computer readable storage medium,
The storage medium, such as ROM/RAM, disk, CD.
The above disclosure is only the preferred embodiments of the present invention, cannot limit the right model of the present invention with this certainly
It encloses, therefore equivalent changes made in accordance with the claims of the present invention, is still within the scope of the present invention.
Claims (8)
1. a kind of Fault diagnosis model distributed intelligence method based on particle filter, it is characterised in that including：
S01：Normal model and various faults model are established according to system dynamics；
S02：Various types of data parameter is obtained by distributed data acquisition；
S03：Using the accurate estimating system state parameter of particle filter algorithm；
S04：Application mode recognizer, by after filtering processing data parameters and the normal model and various faults model into
Row comparison, obtains the operational mode of current system, diagnose to the malfunction of system.
2. a kind of Fault diagnosis model distributed intelligence method based on particle filter according to claim 1, it is characterised in that
The step S01 includes：
The kinetic model of diagnosed object, including building normal model are established by system and structural analysis；
For fault model, first deduction system failure evolution process, typical fault Dynamic Evolution Characteristics are refined, analysis system is mixed
Miscellaneous influent factor and propagation characteristic establish typical fault and propagate description relational expression, to establish according to system failure propagation law
The various faults model of object, and finally realize the building of system state space.
3. a kind of Fault diagnosis model distributed intelligence method based on particle filter according to claim 1, it is characterised in that
The step S02 includes：
Sensor is distributed in each data collection point, timing acquiring data, data collected are with defined message lattice
Formula is packaged, then sends coordinator for the data of acquisition.
4. a kind of Fault diagnosis model distributed intelligence method based on particle filter according to claim 3, it is characterised in that：
The data are to be sent to coordinator by wireless receiving and dispatching radio frequency unit with wireless transmission method, and the coordinator passes through selection one
A channel and a network ID start wireless communication networks.
5. a kind of Fault diagnosis model distributed intelligence method based on particle filter according to claim 1,
It is characterized in that the step S03 includes：
The algorithm detailed process is indicated with following pseudocode：
At the time of FOR k=1,2 ..., K are that each data collection point corresponds to；
(1) particle initializes：
At the k=0 moment, by known prior probability p (x0) generate populationAll particle weights are
(2) sequential importance sampling：
1. sampling：
The k moment samples the particle in significance distribution functionParticle assembly is at this time：
2. weight is estimated：
Acquire the observation z at k momentkAfterwards, the estimated value of available weights of importance：
WhereinFor the reference distribution of importance probability density；
3. weight normalizes：
(3) resampling：
On the basis of obtaining particle weights by Step2, the lesser particle of weight is weeded out, replicates the biggish particle of weight, most
New particle collection is obtained eventually
(4) it exports：
Calculate the state estimation at k moment：
This step can accurate estimating system state parameter.
6. a kind of Fault diagnosis model distributed intelligence method based on particle filter according to claim 1, it is characterised in that
The step S04 includes：
Data parameters after filtering processing are compared with the normal model and various faults model, specific step is as follows institute
Show：
Based on the obtained data of particle filter algorithm, the data at preceding δ moment are calculated with all kinds of state models at corresponding moment
Difference is simultaneously summed, and in obtained each and value, model corresponding to absolute value minimum is the current state in which of system；
Wherein x indicates state model classification, such as normal condition, malfunction 1,2 ..., fiIndicate the i-th moment filtered data
Value, mx,iFor the state value at the i-th moment of model x；
The method for distinguishing of sentencing of system running pattern is time of day curve curve corresponding with each failure or normal condition model
It compares, is present mode of operation corresponding to most identical curve.
7. a kind of Fault diagnosis model distributed intelligence device based on particle filter, it is characterised in that：The device passes through claim
One of 1-6 the method carries out fault diagnosis, which includes：Distributed data acquisition system, coordinator, multiple routers
And multiple terminal devices；
Distributed data acquisition system includes multiple sensors, which is distributed in each data collection point, for acquiring
The data of each data collection point；
The coordinator is used to obtain the network address of terminal device, judges whether it networks, the receiving terminal equipment if networking
The data sent otherwise wait for its networking, which starts channel radio by one channel of selection and a network ID
Network is interrogated, the router is for allowing equipment that network is added, realizing multihop routing and assisting the communication of the terminal device；Institute
The sensor timing acquiring data in terminal device are stated, then are wirelessly sent to the coordinator.
8. a kind of Fault diagnosis model distributed intelligence device based on particle filter according to claim 7, it is characterised in that：
Distributed data acquisition system further includes having CC2530 wireless receiving and dispatching radio frequency unit, ZigBee wireless transmission gateway, USB to turn serial ports
Communication and host computer are provided on the host computer and save the upper computer software of data for real-time display and carry out fault diagnosis calculation
Method calls analysis.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810640185.0A CN108829084A (en) | 2018-06-20 | 2018-06-20 | A kind of Fault diagnosis model distributed intelligence method and device thereof based on particle filter |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810640185.0A CN108829084A (en) | 2018-06-20 | 2018-06-20 | A kind of Fault diagnosis model distributed intelligence method and device thereof based on particle filter |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108829084A true CN108829084A (en) | 2018-11-16 |
Family
ID=64142619
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810640185.0A Pending CN108829084A (en) | 2018-06-20 | 2018-06-20 | A kind of Fault diagnosis model distributed intelligence method and device thereof based on particle filter |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108829084A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114330599A (en) * | 2022-03-14 | 2022-04-12 | 中汽信息科技(天津)有限公司 | Engine fault identification method, equipment and system based on deep learning |
| CN114375270A (en) * | 2019-08-30 | 2022-04-19 | 捷豹路虎有限公司 | Distributed diagnostic architecture for vehicles |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104156612A (en) * | 2014-08-25 | 2014-11-19 | 福建师范大学 | Fault forecasting method based on particle filter forward and reverse direction prediction errors |
| CN105718733A (en) * | 2016-01-21 | 2016-06-29 | 福建师范大学 | Fault predicting method based on fuzzy nearness and particle filter |
-
2018
- 2018-06-20 CN CN201810640185.0A patent/CN108829084A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104156612A (en) * | 2014-08-25 | 2014-11-19 | 福建师范大学 | Fault forecasting method based on particle filter forward and reverse direction prediction errors |
| CN105718733A (en) * | 2016-01-21 | 2016-06-29 | 福建师范大学 | Fault predicting method based on fuzzy nearness and particle filter |
Non-Patent Citations (2)
| Title |
|---|
| 孟志强等: "基于粒子滤波的分布式智能故障诊断系统研究", 《湖南大学学报（自然科学版）》 * |
| 张磊等: "一种基于高斯混合模型粒子滤波的故障预测算法", 《航空学报》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114375270A (en) * | 2019-08-30 | 2022-04-19 | 捷豹路虎有限公司 | Distributed diagnostic architecture for vehicles |
| CN114375270B (en) * | 2019-08-30 | 2023-11-24 | 捷豹路虎有限公司 | Distributed diagnostic architecture for a vehicle |
| CN114330599A (en) * | 2022-03-14 | 2022-04-12 | 中汽信息科技(天津)有限公司 | Engine fault identification method, equipment and system based on deep learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107941537B (en) | A kind of mechanical equipment health state evaluation method | |
| CN108171259A (en) | A kind of wisdom tip water quality monitoring system and method based on NB-IoT | |
| CN108126987B (en) | Remote online monitoring and intelligent diagnosis method for high-speed wire rolling mill | |
| Zonzini et al. | Structural health monitoring and prognostic of industrial plants and civil structures: A sensor to cloud architecture | |
| Li et al. | Multi-sensor data-driven remaining useful life prediction of semi-observable systems | |
| CN109901537A (en) | Mechanical equipment method for monitoring operation states and system for edge calculations side | |
| CN107146004A (en) | A kind of slag milling system health status identifying system and method based on data mining | |
| CN112085261A (en) | Enterprise production status diagnosis method based on cloud fusion and digital twin technology | |
| CN109238455B (en) | A kind of characteristic of rotating machines vibration signal monitoring method and system based on graph theory | |
| KR102427205B1 (en) | Apparatus and method for generating training data of artificial intelligence model | |
| CN111130890A (en) | Network flow dynamic prediction system | |
| CN107473036A (en) | Elevator traction machine remote online detects and diagnostic system and its checkout and diagnosis method | |
| CN110597235A (en) | Universal intelligent fault diagnosis method | |
| CN108829084A (en) | A kind of Fault diagnosis model distributed intelligence method and device thereof based on particle filter | |
| CN114002597A (en) | Motor fault diagnosis method and system based on GRU network stator current analysis | |
| CN116519895A (en) | Water quality on-line monitoring analysis management system based on big data platform | |
| CN115238785A (en) | Rotary machine fault diagnosis method and system based on image fusion and integrated network | |
| CN113776794B (en) | Fault diagnosis method, device and system for embedded edge calculation | |
| Nguyen et al. | Ensemble based real-time adaptive classification system for intelligent sensing machine diagnostics | |
| CN117009841A (en) | Model training method, motor fault diagnosis method and microcontroller | |
| CN112417771A (en) | Method for establishing fault diagnosis model of machine tool spindle and method for diagnosing fault of machine tool spindle | |
| CN116704729A (en) | Industrial kiln early warning system and method based on big data analysis | |
| CN104168599B (en) | Wireless sensor network fault diagnosis method based on time weight K- nearest neighbour methods | |
| CN111382946B (en) | Autonomous evaluation method and system for health state of equipment and industrial internet equipment | |
| CN113449966A (en) | Gypsum board equipment inspection method and 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: 20181116 |
|
| RJ01 | Rejection of invention patent application after publication |