CN108700872A - Machine sort device - Google Patents
Machine sort device Download PDFInfo
- Publication number
- CN108700872A CN108700872A CN201680082718.7A CN201680082718A CN108700872A CN 108700872 A CN108700872 A CN 108700872A CN 201680082718 A CN201680082718 A CN 201680082718A CN 108700872 A CN108700872 A CN 108700872A
- Authority
- CN
- China
- Prior art keywords
- machine
- data
- sort
- equipment
- machine sort
- 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.)
- Granted
Links
- 238000011002 quantification Methods 0.000 claims abstract description 27
- 238000012544 monitoring process Methods 0.000 claims description 11
- 238000000034 method Methods 0.000 description 23
- 238000012545 processing Methods 0.000 description 19
- 238000013480 data collection Methods 0.000 description 14
- 230000015654 memory Effects 0.000 description 11
- 238000012423 maintenance Methods 0.000 description 6
- 241001269238 Data Species 0.000 description 5
- 238000004378 air conditioning Methods 0.000 description 5
- 230000002159 abnormal effect Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 4
- 238000006243 chemical reaction Methods 0.000 description 4
- 238000000491 multivariate analysis Methods 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000007417 hierarchical cluster analysis Methods 0.000 description 3
- 230000009467 reduction Effects 0.000 description 3
- 230000005856 abnormality Effects 0.000 description 2
- 230000015556 catabolic process Effects 0.000 description 2
- 238000006731 degradation reaction Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000000513 principal component analysis Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000010485 coping Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000007429 general method Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005191 phase separation Methods 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 230000003936 working memory Effects 0.000 description 1
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
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/04—Manufacturing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Manufacturing & Machinery (AREA)
- Business, Economics & Management (AREA)
- Automation & Control Theory (AREA)
- General Health & Medical Sciences (AREA)
- Economics (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Human Resources & Organizations (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Testing And Monitoring For Control Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Data acquisition (101) obtains machine sort achievement data from machine classification indicators database (201).The qualitative data that classification indicators quantification portion (102) is included in machine sort achievement data is converted into indicating the quantitative data of the similarity between qualitative data.Machine sort portion (103) uses quantitative data, is classified to equipment as unit of machine.
Description
Technical field
The present invention relates to the machine sort devices classified to equipment as unit of the machine of constitution equipment.
Background technology
About equipment such as elevator, air-conditionings, there are multiple equipment of identical type in a variety of environment, according to phase
It is useful that the equipment of same feature, which carries out classification,.Such as in the existing system recorded in patent document 1, with building
Equipment it is energy saving for the purpose of illumination and air-conditioning control in, elevator is divided according to the equipment of feature having the same
Class.In system recorded in patent document 1, using elevator operation information, according to what day, each period etc. is to public
The flow of the people in region, carries out medelling in room rate and plans control-time table at indoor occupancy.Here, building is carried out
Classification, to continue to use the solution of the similar building of feature having the same in the building that can not obtain elevator operation information
Analyse result.
Existing technical literature
Patent document
Patent document 1:Japanese Unexamined Patent Publication 2005-104635 bulletins
Invention content
The subject that the invention solves
In the case where carrying out the parsings such as failure and exception for the machine for constituting the equipment such as elevator, air-conditioning, by pressing
Multiple equipment and machine are classified and parsed according to the machine of identical type, same characteristic features, with the solution using only single equipment
Phase separation ratio, it is contemplated that the raising of failure and abnormal accuracy of detection.
However, in the existing method, classified with equipment unit as elevator, therefore there are following such
Problem:Even with different characteristic machine, as long as equipment unit is identical type, same characteristic features, can not to machine into
Row classification.For example, the door machine device in the door machine device and composition elevator B that constitute elevator A is homotype but the different machine of feature
In the case of, in the past, if elevator A and elevator B are as elevator, this equipment is judged as same characteristic features, both sides'
Door machine device can be classified into same characteristic features.
Also, index used in the previous method, classifying is elevator operation information, the purposes of elevator, rule
Mould, but parse failure or it is abnormal equal in the case ofs, the setting environment of fault history, equipment/machine before considering, machine are more
It changes the bulk informations such as information to classify, it is possible thereby to the raising of expected nicety of grading.These information are necessarily by numerical value structure
At quantitative data, sometimes include the qualitative data of character information.In addition, in the previous method, according to qualitative data
In the case of being classified, the similar evaluation of which kind of degree each other of different qualitative datas is not accounted for.As a result, can not fill
Divide and carry out abnormal reason parsing, existing leads to this problems such as abnormality detection precision reduction.
The present invention is to complete in order to solve this problem, it is intended that machine can precisely be carried out by providing
Failure and abnormal etc. parsing machine sort device.
Means for solving the problems
The machine sort device that the present invention is sent out has:Data acquisition is obtained as respectively by single or multiple machines
The machine sort achievement data of the intrinsic information of each machine in the multiple equipment of composition, the machine sort achievement data be from
It is obtained in the monitoring data of each machine;Classification indicators quantification portion is included in qualitative in machine sort achievement data
Data conversion at indicate qualitative data between similarity quantitative data;And machine sort portion, quantitative data is used, with
Machine is that unit classifies to equipment.
Invention effect
The qualitative data that the machine sort device of the present invention is included in machine sort achievement data is converted into indicating fixed
The quantitative data of similarity between property data classifies to equipment as unit of machine using the quantitative data.As a result,
The parsing of the failure and exception etc. of machine can precisely be carried out.
Description of the drawings
Fig. 1 is the structure chart of the machine sort device of embodiments of the present invention 1.
Fig. 2 is the explanation for showing to safeguard real data example used in the machine sort device of embodiments of the present invention 1
Figure.
Fig. 3 is the hardware structure diagram of the machine sort device of embodiments of the present invention 1.
Fig. 4 is the flow chart of the machine sort processing for the machine sort device for showing embodiments of the present invention 1.
Fig. 5 is the definition graph for showing quantitative data example used in the machine sort device of embodiments of the present invention 1.
Fig. 6 is the letter in advance of the qualitative data similarity for the machine sort device for showing to have embodiments of the present invention 1
The flow chart of machine sort processing in the case of breath.
Fig. 7 A, Fig. 7 B, Fig. 7 C are the explanations of the priori information example for the machine sort device for showing embodiments of the present invention 1
Figure.
Fig. 8 is the definition graph of the classification example of the quantitative data for the machine sort device for showing embodiments of the present invention 1.
Fig. 9 is the definition graph of the characteristic quantity of each machine for the machine sort device for showing embodiments of the present invention 1.
Figure 10 is the structure chart of the machine sort device of embodiments of the present invention 2.
Specific implementation mode
Hereinafter, in order to which the present invention will be described in more detail, based on the attached drawing of addition to side for carrying out the present invention
Formula illustrates.
Embodiment 1.
Fig. 1 is the structure chart of the monitoring system including the machine sort device 100 comprising present embodiment.
In the monitoring system of diagram, machine sort device 100 is connect with data collection managing device 200, data collection
Managing device 200 is connect via network 300 with supervision object 400.
Machine sort device 100 has data acquisition 101, classification indicators quantification portion 102 and machine sort portion
103.Data acquisition 101 is to obtain machine from the machine sort achievement data library 201 that data collection managing device 200 is managed
The processing unit of device classification indicators data.Classification indicators quantification portion 102 be included in it is qualitative in machine sort achievement data
Data conversion at quantitative data processing unit.Machine sort portion 103 is to use to be quantified by what classification indicators quantification portion 102 generated
Data, the processing unit classified to equipment according to machine units.
Data collection managing device 200 is to collect the monitoring data from supervision object 400 and as machine sort index
The device that database 201 is stored and managed.The monitoring data being stored in machine sort achievement data library 201 refer to basis
Data (such as safeguarding real data) that facility information and maintenance personnel generate the inspection of supervision object 400 etc., from monitoring
The data that object 400 directly or indirectly obtains.As the machine sort index being stored in machine sort achievement data library 201
The example for safeguarding real data by taking elevator as an example is shown in FIG. 2 in data.
It is shown in FIG. 2 according to the inspection of a machine of an equipment of facility information and maintenance personnel couple and obtains
Safeguard real data example.In safeguarding real data example, as the example of data item, device id, type ID, machine are described
ID, setting area, operator's name, upkeep operation content, have it is without exception etc..The value of these data item is an example.It can be in order to
It preserves the project for safeguarding real data being collected into from actual equipment, machine and changes data item.As long as also, can distinguish
Equipment, machine, can also by multiple equipment, machine data summarization at a table.As long as moreover, can be by equipment, machine pair
It should get up, the data of a machine of an equipment can also be divided into multiple tables.Also, maintenance usually can also be made
Industry with failure, exception has occurred in the case of the different upkeep operation of forms such as upkeep operation the segmentation of maintenance real data and
It is managed.That is, as the machine sort achievement data being stored in machine sort achievement data library 201, as long as machine institute
Intrinsic information can be then any information.
Supervision object 400 is, for example, elevator or the such equipment being made of single or multiple machine of air-conditioning.About prison
Depending on object 400, it is assumed that there are more than two equipment being made of identical machine.Can also be do not connect with network 300 and
The structure that supervision object 400 is directly connect with data collection managing device 200.No matter supervision object 400 and data collection management
How is the connection method of device 200, can be by 100 network connection of data collection managing device 200 and machine sort device
The structure got up.
Fig. 3 is the block diagram for the hardware configuration for showing the machine sort device for realizing present embodiment.It is shown in FIG. 3
The machine sort device 100 of Fig. 1 and data collection management device 200 be formed in the example on a hardware.Machine sort
Device 100 and data collection management device 200 have processor 11, memory (memory) 12, communication I/F (interface) device 13,
Memory (storage) 14 and output device 15.Processor 11 is for realizing machine sort device 100 and data collecting pipe
Manage the processor of the function of device 200.Memory 12 is as storage and machine sort device 100 and data collection management device
The program internal memory of the 200 corresponding various programs of function, when processor 11 carries out data processing the working memory that uses and
The storage parts such as the ROM and RAM used for memory etc. of signal data expansion.Communication I/F devices 13 are external with network 300 equal
Between communication interface.Memory 14 is the memory for storing various data and program.Output device 15 is for outside
Portion exports the device of handling result.
The processing that data acquisition 101, classification indicators quantification portion 102 and machine sort portion 103 in Fig. 1 are carried out
It is the program being stored in memory 12 to be read by processor 11 by is executed.The number that machine sort achievement data library 201 stores
According to being preserved in memory 14 via network 300 and by communicating I/F devices 13 from supervision object 400.Machine sort
The handling result in portion 103 preserves in memory 14 as needed, is exported to outside by output device 15.Alternatively, it is also possible to
Machine sort device 100 and data collection management device 200 are formed on different hardware.
Next, being illustrated to the action of the machine sort device 100 of present embodiment.
Data collection managing device 200 by the machine sort achievement data obtained from supervision object 400 continuously or interval
Ground inputs to machine sort achievement data library 201.Machine sort device 100 obtains machine from machine classification indicators database 201
Classification indicators data are simultaneously handled.Fig. 4 is the flow chart for the processing for showing machine sort device 100.
First, data acquisition 101 obtains machine sort achievement data (step from machine classification indicators database 201
ST1).In addition, in the case of including multiple data item in machine sort achievement data, figure is executed according to each data item
4 flow.For example, in the case where having input index of the machine ID as machine sort achievement data, outputs and classified
The list of machine ID.There is no limit as an example, give each classification distribution classification ID, by each machine ID and phase to the form of list
The sheet form that the classification ID answered is preserved in a row is exported.Also, as the example of other lists, there is also following sides
Method:A file is generated according to each classification, the machine ID for belonging to the classification is preserved in file.
In classification indicators quantification portion 102, it is included in determining from the machine sort achievement data that each machine obtains
Property data are according to can determine that the form of similitude is converted into the quantitative data being made of numerical value.It is defeated by institute in step ST2
Whether the machine sort achievement data entered is numerical value to determine whether being quantitative data, and branch is carried out to processing later.In step
In rapid ST2, (the step ST2 be quantitative data the case where:It is) under, classification indicators quantification portion 102 terminates the processing.That is, will
The machine sort achievement data for inputing to classification indicators quantification portion 102 is directly output to machine sort portion 103.On the other hand,
In step ST2, (the step ST2 be not quantitative data the case where:It is no) under, execute the processing of step ST3.In step ST3,
It calculates the distance between qualitative data and is used as the mutual similarity of qualitative data, each data will be distributed to apart from corresponding value,
Thus it is set as quantitative data.The distance between qualitative data is by the character string parsings method such as Hierarchical clustering analysis of n-gram
And calculate, quantitative data will be set as with apart from corresponding numerical value.Here, about qualitative data, the position of character with to away from
From influence between relationship (the character representation in front gets over the classification of major class, therefore the influence adjusted the distance is big, rear
Character, indicate the classification for getting over group, therefore the influence adjusted the distance is small, etc.) be known in the case of, can calculate apart from when
It carries out to processing such as the big character weightings of the influence adjusted the distance.For example, indicating master if it is the first half in the character string of machine ID
Will more new version number, later half expression secondary more new version number the case where etc., then the character string in front sometimes, the shadow adjusted the distance
Sound is bigger.
The quantitative data example converted to qualitative data is shown in FIG. 5.In Figure 5, it as simply example, shows
The title of machine ID is connected the main more new version number of machine by hyphen symbol "-" with secondary more new version number
Come and the quantitative data example in the case of table note.In the quantitative data example, machine ID AAA-01, AAA-02, AAA-03's
The main more new version number of machine is identical, and only secondary more new version number is different, therefore distributes similar value.Machine ID is BBB-
01, the main more new version number of the machine of BBB-02 is different from AAA-01, AAA-02, AAA-03, therefore distributes value farther out.
In the case that machine sort achievement data is quantitative data in step ST2 or carried out step ST3 processing it
Afterwards, machine sort portion 103 is by multivariate analysis method or machine learning method etc., according to value (the i.e. multivariable solution inputted
Analysis etc. in characteristic quantity) similar in machine classified (step ST4).The specific classification example of description below.
On the other hand, it in the case where learning the mutual similarity of qualitative data as priori information, can also use
The similarity of priori information.About priori information, a part for qualitative data can also be only specified.For example, only specified machine ID
In main update number similarity etc..Also, about priori information, each character position of qualitative data can also be specified
Weighted Rule.For example, the ratio etc. of the weight of main update number and secondary update number in specified machine ID.Also,
The qualitative data that do not converted to quantitative data can also be provided as priori information.
Classification indicators quantification flow in the case of being shown in FIG. 6 with priori information.Pair processing identical with Fig. 4
Mark identical number of steps.Classification indicators quantification portion 102 first step ST2 whether be quantitative data judgment step
The case where not to be quantitative data (step ST2:It is no) under, judge whether to have related with the machine sort achievement data inputted
Similarity priori information, to later processing carry out branch (step ST5).In step ST5, in the thing of not similarity
The case where preceding information (step ST5:It is no) under, the processing of implementation steps ST3 in the same manner as the flow of Fig. 4.With similarity
The case where priori information (step ST5:It is) under, in the processing of step ST6, the similarity of distribution and given priori information
Corresponding numerical value.Here, in the case where the similarity of priori information is not quantitative data but qualitative data, with step ST3
Similarly, it calculates the distance between qualitative data and is used as the mutual similarity of qualitative data, will be distributed to apart from corresponding value
Each data, are thus set as quantitative data.The distance between qualitative data is to calculate word by Hierarchical clustering analysis of n-gram etc.
The method of the distance between language and calculate, quantitative data will be set as with apart from corresponding numerical value.Between calculating qualitative data
The method of distance can also use the method different from step ST3.
The example of priori information is shown in FIG. 7.The example of the similarity of specified qualitative data is shown in fig. 7.Fig. 7 A
It is the example of the similarity of 3 characters in front of specified machine ID, shows a case that following:Machine ID is the machine of AAA and BBB
The similarity of device is higher, and the machine that machine ID is CCC similarity compared with machine ID is AAA and the machine of BBB is relatively low.Also,
The example of the Weighted Rule of each character position of specified qualitative data is shown in figure 7b.Fig. 7 B are each characters of machine ID
The example of the Weighted Rule of position is the example of following situation:The weight of the 1st~3 character of machine ID is increased, therefore will
The weight of 1st~3 character is set as 10, keeps the 1st~3 character of weight ratio of the 5th~6 character small, therefore by the 5th~6
The weight of character is set as 1.In addition, since the 4th character is hyphen, thus be excluded that except Weighted Rule.Also, in Fig. 7 C
In the example of the specified qualitative data that do not converted to quantitative data is shown.It shows in fig. 7 c and quantification is not carried out to device id
The case where.Fig. 7 A, Fig. 7 B, Fig. 7 C be the example for carrying out specified information, can also modification information give mode.
As another example, the free text recorded as the result of the upkeep operation of machine can also be used as machine
Classification indicators data.For example, being contained in the free text for the result for recording upkeep operation by morphemic analysis to extract
The words such as " having exception ", " disposition is completed ", " the reason is that event A " distribute similar numerical value etc. to the text more than similar morpheme.
In machine sort portion 103, classify to each machine according to the similar machine of feature, therefore input there are multiple machines
Device carries out transformed quantitative data by classification indicators quantification portion 102, is carried out according to the machine being worth similar in quantitative data
Classification.About quantitative data, it can be one data item of input, can also be that multiple data item are summarized and are exported.In Fig. 4 or
In the step ST4 of Fig. 6, non-hierarchical clusterings such as Hierarchical clustering analysis and k-means methods such as dendrogram etc. can also be used general
The common machine learning method such as logical multivariate analysis method and support vector machines.The example of classification is shown in FIG. 8.
In fig. 8, the example as the classification of quantitative data inputs the quantitative data of multiple data item of three machines,
Feature quantity space when as multivariate analysis methods such as execution principal component analysis, schematically shows on two-dimentional scatter plot
Characteristic quantity 1 and characteristic quantity 2.The case where following is shown in FIG. 8:Characterizing magnitudes 801 and characterizing magnitudes 802 are on scatter plot
Distance it is closer, therefore be concluded for one classification 804.It shows a case that following:Characterizing magnitudes 803 and characterizing magnitudes 801
And distance of the characterizing magnitudes 802 on scatter plot is farther out, therefore be set as and 804 different classification 805 of classification.As in this way into
The method of row classification, the clustering methodology that nearest neighbor method and k-means methods etc. can be used common, the nearest neighbor method refer to calculating
Distance between characterizing magnitudes 801,802,803, classifies according to the threshold value of distance, which refers to advance
Determine the quantity of classification.
Can be the failure and analysis of anomaly of machine as the purposes of the present invention.For example, in order to formulate machine
Maintenance plan and in the case of predicting time for breaking down in the future, there is following method:According to the data obtained from machine,
According to statistically failure Frequency and degradation trend predict the probability (failure risk) that breaks down in the future, thus it is speculated that need
The time to be safeguarded.
Here, the probability and degradation trend to break down about the machine with similar feature also similar possibility
Height, therefore in order to predict failure risk and carry out classification to machine according to each similar features to be also useful.According to more like
Machine carry out classification and can improve the precision of prediction of failure risk.The data used to calculate failure risk can with
The data used in the machine sort device 100 of present embodiment are identical, can also use other data.In estimation failure risk
When the prediction of failure risk can be carried out as unit of machine, but can also synthetically judge to constitute multiple machines of identical equipment
The failure risk of multiple machines such as the correlativity of device, to predict the failure risk in equipment unit.
Next, being illustrated to the effect of embodiment 1.In fig.9, the example as the characteristic quantity of each machine,
For machine 1, machine 2 the two machines, they are shown from equipment a, equipment b, equipment c these three equipment and collects data simultaneously
On calculating characteristic quantity and the two-dimentional scatter plot generated according to the characteristic quantity of the two machines.By the equipment feature scale of machine 1
901 are shown as, the equipment characteristic quantity of machine 2 is expressed as 902.
In the conventional method, in the classification due to equipment unit and equipment a, equipment b, equipment c are classified into identical
Feature equipment in the case of, no matter the feature of machine 1, machine 2, be all set as identical classification.On the other hand, exist
In embodiment 1, for example, following wait of middle progress can be classified as unit of machine:Even in the equipment characteristic quantity of machine 1
In the case of being set as a classification by equipment a and equipment b in 901, equipment c is set as another classification, in the equipment feature of machine 2
In amount 902, equipment a and equipment c are set as a classification, equipment b is set as another classification.
By classifying according to the machine with same characteristic features, can expect the failure risk of machine precision of prediction,
The raising of the precision of failure and abnormality detection etc..Also, by classifying according to similar machine, in some machine discovery
In the case that failure and exception are equal, are safeguarded by machine of the extraction with same characteristic features, the event of other machines can be prevented
Barrier and exception can expect the maintenance efficient of the scheduling of the upkeep operation of each machine in possible trouble.For example, due to elevator
The torque of the door opening and closing motor of A reduce and cause it is trapped in the case of, check that the door of other elevators with same characteristic features is opened
Close whether motor has the sign of torque reduction and safeguarded, thus, it is possible to expect to reduce failure and accident.As another
Example, in the case that the torque of motor is opened and closed in the door for detecting elevator A to be reduced, although in other liters with same characteristic features
It is also possible to generate torque reduction in the door opening and closing motor of drop machine, but if It is not necessary to coping at once, then by suitably right
Upkeep operation is scheduled, and can expect the high efficiency of operation.
As described above, according to the machine sort device of embodiment 1, since the machine sort device has:Number
According to acquisition unit, the information intrinsic as each machine obtained from the monitoring data of each machine in multiple equipment is obtained
Machine sort achievement data, wherein multiple equipment is made of single or multiple machines respectively;Classification indicators quantification portion,
The qualitative data being included in machine sort achievement data is converted into indicating the quantitative data of the similarity between qualitative data;
And machine sort portion, classified to equipment as unit of machine using quantitative data, thus can precisely into
The parsing of the failure and exception etc. of row machine.
Embodiment 2.
In the embodiment 1, machine sort portion 103 is obtained according to quantification is carried out in classification indicators quantification portion 102
The machine sort achievement data arrived, has carried out the classification of each machine.In contrast, can also be that will determined by classification indicators
It is each in order to emphasize before machine sort achievement data obtained from the progress quantification of quantization unit 102 inputs to machine sort portion 103
The difference of the feature of machine and the machine sort achievement data is converted into characteristic quantity, machine sort portion 103 according to this feature amount,
Classify to each machine, is illustrated this as embodiment 2.
The purpose for being converted into characteristic quantity is in the case where being classified to machine according to multiple machine sort achievement datas
The difference of each machine is set to make clear.In the case where carrying out machine sort according only to machine sort achievement data, to
It gives the distribution of similar machine sort achievement data similar value when quantitative data is converted, therefore can only pass through machine sort index
The value of data is classified.But in the case where carrying out machine sort according to multiple machine sort achievement datas, even certain
A machine sort achievement data is the machine of similar value, and other machines classification indicators data also have separate value sometimes.
Under such circumstances, if keeping the original sample of the value of machine sort achievement data, the difference of each machine can not be expressly understood that
It is different, therefore can not accurately classify to machine.Therefore, each by seeking making according to multiple machine sort achievement datas
The difference of machine makes such characteristic quantity clear, can more accurately classify to machine.For example, having following method:
According to multiple machine sort achievement datas such as type ID, setting areas, characteristic quantity is set using the distance in MT methods.As spy
Sign amount, such as can be in each principal component, the regression coefficient in regression analysis and the error of principal component analysis, pattern match method
The general method such as the multivariate analysis method of similarity etc..
Figure 10 is the structure chart of the monitoring system for the machine sort device 100a for applying embodiment 2.Embodiment 2
Machine sort device 100a has data acquisition 101, classification indicators quantification portion 102, machine sort portion 103a and feature
Measure converter section 104.Here, data acquisition 101 and classification indicators quantification portion 102 are identical as embodiment 1.Characteristic quantity is converted
Portion 104 is that machine sort achievement data obtained from carrying out quantification by classification indicators quantification portion 102 is converted into characteristic quantity
Processing unit.Machine sort portion 103a be characteristic quantity obtained from use is converted by characteristic quantity converter section 104 to machine into
The processing unit of row classification.In addition, in Fig. 10, data collection managing device 200, network 300 and supervision object 400 and Fig. 1
Shown in embodiment 1 it is identical.
In the machine sort device 100a constituted in this way, in the quantitative number that will be generated by classification indicators quantification portion 102
According to before inputing to machine sort portion 103a, quantitative data is converted into characteristic quantity by characteristic quantity converter section 104.Machine sort portion
103a from characteristic quantity converter section 104 obtain characteristic quantity, using the similar machine of the value of characteristic quantity as the similar machine of feature and into
Row classification.In addition, characteristic quantity converter section 104 can also be following structure:It will not be generated by classification indicators quantification portion 102
Quantitative data is wholly converted into characteristic quantity, but only conversion is a part of.In the case of an only conversion part, machine sort portion
103a is classified using quantitative data and the characteristic quantity both sides being converted to.
As described above, according to the machine sort device of embodiment 2, since the machine sort device has:Number
According to acquisition unit, the information intrinsic as each machine obtained from the monitoring data of each machine in multiple equipment is obtained
Machine sort achievement data, wherein multiple equipment is made of single or multiple machines respectively;Classification indicators quantification portion,
The qualitative data being included in machine sort achievement data is converted into indicating the quantitative data of the similarity between qualitative data;
Quantitative data is converted into indicating the characteristic quantity of the difference of the feature of each machine by characteristic quantity converter section;And machine sort portion,
It classifies to equipment as unit of machine using machine similar in characteristic quantity as the similar machine of feature, therefore being capable of essence
Degree preferably carries out the parsing of the failure and exception etc. of machine.
In addition, the present application can freely combine each embodiment or to each embodiment party within the scope of the invention
The arbitrary structural element of formula is deformed, or omits arbitrary structural element in various embodiments.
Industrial availability
As above, machine sort device of the invention is for multiple equipment, according to every possessed by these equipment
A machine classifies to each equipment, is suitable for the equipment such as elevator or air-conditioning, there are identical types in different environment
Multiple equipment.
Label declaration
100,100a:Machine sort device;101:Data acquisition;102:Classification indicators quantification portion;103,103a:Machine
Device division;104:Characteristic quantity converter section;200:Data collection managing device;201:Machine sort achievement data library;300:Net
Network;400:Supervision object.
Claims (2)
1. a kind of machine sort device, which is characterized in that the machine sort device has:
It is intrinsic to obtain the conduct each machine institute obtained from the monitoring data of each machine in multiple equipment for data acquisition
Information machine sort achievement data, wherein multiple equipment is made of the single or multiple machines respectively;
Classification indicators quantification portion, the qualitative data being included in the machine sort achievement data are converted into indicating qualitative
The quantitative data of similarity between data;And
Machine sort portion is used the quantitative data, is classified to equipment as unit of machine.
2. a kind of machine sort device, which is characterized in that the machine sort device has:
It is intrinsic to obtain the conduct each machine institute obtained from the monitoring data of each machine in multiple equipment for data acquisition
Information machine sort achievement data, wherein multiple equipment is made of the single or multiple machines respectively;
Classification indicators quantification portion, the qualitative data being included in the machine sort achievement data are converted into indicating qualitative
The quantitative data of similarity between data;
The quantitative data is converted into indicating the characteristic quantity of the difference of the feature of each machine by characteristic quantity converter section;With
And
Machine sort portion pair is set using machine similar in the characteristic quantity as the similar machine of feature as unit of the machine
It is standby to classify.
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/JP2016/056050 WO2017149598A1 (en) | 2016-02-29 | 2016-02-29 | Apparatus classification device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108700872A true CN108700872A (en) | 2018-10-23 |
| CN108700872B CN108700872B (en) | 2021-08-06 |
Family
ID=59742555
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201680082718.7A Active CN108700872B (en) | 2016-02-29 | 2016-02-29 | Machine sorting device |
Country Status (4)
| Country | Link |
|---|---|
| JP (1) | JP6366852B2 (en) |
| CN (1) | CN108700872B (en) |
| TW (1) | TWI621951B (en) |
| WO (1) | WO2017149598A1 (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6909670B2 (en) * | 2017-08-03 | 2021-07-28 | 日立グローバルライフソリューションズ株式会社 | Anomaly detection method and anomaly detection system |
| CN111133396B (en) | 2017-10-16 | 2023-03-24 | 富士通株式会社 | Production facility monitoring device, production facility monitoring method, and recording medium |
| WO2020044533A1 (en) * | 2018-08-31 | 2020-03-05 | 東芝三菱電機産業システム株式会社 | Manufacturing process monitoring device |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101432864A (en) * | 2006-04-27 | 2009-05-13 | 夏普株式会社 | Method and system for classifying defect distribution, method and system for specifying causative equipment, computer program and recording medium |
| CN102404058A (en) * | 2010-07-26 | 2012-04-04 | 日本电气株式会社 | Electromagnetic wave discrimination device, electromagnetic wave discrimination method, and electromagnetic wave discrimination program |
| CN102750289A (en) * | 2011-04-19 | 2012-10-24 | 富士通株式会社 | Tag group classifying method and equipment as well as data mixing method and equipment |
| CN103679190A (en) * | 2012-09-20 | 2014-03-26 | 富士通株式会社 | Classification device, classification method and electronic equipment |
| CN103745229A (en) * | 2013-12-31 | 2014-04-23 | 北京泰乐德信息技术有限公司 | Method and system of fault diagnosis of rail transit based on SVM (Support Vector Machine) |
| CN104851054A (en) * | 2015-05-18 | 2015-08-19 | 国家电网公司 | Equipment maintenance method in 10kV voltage substation operation |
| CN104883278A (en) * | 2014-09-28 | 2015-09-02 | 北京匡恩网络科技有限责任公司 | Method for classifying network equipment by utilizing machine learning |
| US20150254890A1 (en) * | 2014-03-07 | 2015-09-10 | Fujitsu Limited | Classification method, classification device, and recording medium |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH049716A (en) * | 1990-04-27 | 1992-01-14 | Toshiba Corp | Forming apparatus for measuring reference signal of quantity of state of plant |
| JP3507607B2 (en) * | 1995-12-15 | 2004-03-15 | 株式会社東芝 | Plant data management device |
| TWI338129B (en) * | 2006-04-07 | 2011-03-01 | Rts Co Ltd | Apparatus and method for dual electronic part inspection |
| CN101315644B (en) * | 2008-05-09 | 2011-11-09 | 浙江工业大学 | Part classification method based on developable clustering |
| JP2011138251A (en) * | 2009-12-28 | 2011-07-14 | Fujitsu Telecom Networks Ltd | Monitoring control network system |
| TW201137653A (en) * | 2010-04-29 | 2011-11-01 | Hon Hai Prec Ind Co Ltd | System and method for providing a component derating process |
| JP5665635B2 (en) * | 2011-04-18 | 2015-02-04 | 三菱電機株式会社 | Parameter and equipment recommendation equipment for building facilities |
| JP5629237B2 (en) * | 2011-05-17 | 2014-11-19 | 日立オムロンターミナルソリューションズ株式会社 | Parts life management system |
-
2016
- 2016-02-29 JP JP2017541413A patent/JP6366852B2/en active Active
- 2016-02-29 WO PCT/JP2016/056050 patent/WO2017149598A1/en active Application Filing
- 2016-02-29 CN CN201680082718.7A patent/CN108700872B/en active Active
- 2016-07-29 TW TW105124046A patent/TWI621951B/en not_active IP Right Cessation
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101432864A (en) * | 2006-04-27 | 2009-05-13 | 夏普株式会社 | Method and system for classifying defect distribution, method and system for specifying causative equipment, computer program and recording medium |
| CN102404058A (en) * | 2010-07-26 | 2012-04-04 | 日本电气株式会社 | Electromagnetic wave discrimination device, electromagnetic wave discrimination method, and electromagnetic wave discrimination program |
| CN102750289A (en) * | 2011-04-19 | 2012-10-24 | 富士通株式会社 | Tag group classifying method and equipment as well as data mixing method and equipment |
| CN103679190A (en) * | 2012-09-20 | 2014-03-26 | 富士通株式会社 | Classification device, classification method and electronic equipment |
| CN103745229A (en) * | 2013-12-31 | 2014-04-23 | 北京泰乐德信息技术有限公司 | Method and system of fault diagnosis of rail transit based on SVM (Support Vector Machine) |
| US20150254890A1 (en) * | 2014-03-07 | 2015-09-10 | Fujitsu Limited | Classification method, classification device, and recording medium |
| CN104883278A (en) * | 2014-09-28 | 2015-09-02 | 北京匡恩网络科技有限责任公司 | Method for classifying network equipment by utilizing machine learning |
| CN104851054A (en) * | 2015-05-18 | 2015-08-19 | 国家电网公司 | Equipment maintenance method in 10kV voltage substation operation |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108700872B (en) | 2021-08-06 |
| WO2017149598A1 (en) | 2017-09-08 |
| TW201732638A (en) | 2017-09-16 |
| TWI621951B (en) | 2018-04-21 |
| JP6366852B2 (en) | 2018-08-01 |
| JPWO2017149598A1 (en) | 2018-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111475804B (en) | Alarm prediction method and system | |
| Sheu | Dynamic relief-demand management for emergency logistics operations under large-scale disasters | |
| JP5331774B2 (en) | Equipment state monitoring method and apparatus, and equipment state monitoring program | |
| CN106600115A (en) | Intelligent operation and maintenance analysis method for enterprise information system | |
| CN104636449A (en) | Distributed type big data system risk recognition method based on LSA-GCC | |
| CN112256761A (en) | Big data mining and fault prediction method based on elevator full life cycle | |
| KR20180072167A (en) | System for extracting similar patents and method thereof | |
| CN104050361A (en) | Intelligent analysis early warning method for dangerousness tendency of prison persons serving sentences | |
| CN113688169B (en) | Mine potential safety hazard identification and early warning system based on big data analysis | |
| RU2716029C1 (en) | System for monitoring quality and processes based on machine learning | |
| CN114327964A (en) | Method, device, equipment and storage medium for processing fault reasons of service system | |
| CN113190372B (en) | Multi-source data fault processing method and device, electronic equipment and storage medium | |
| ITMI20122255A1 (en) | METHOD AND SYSTEM FOR RISK ASSESSMENT FOR THE SAFETY OF AN INDUSTRIAL INSTALLATION | |
| CN108700872A (en) | Machine sort device | |
| Silva et al. | Damage detection in a benchmark structure using AR-ARX models and statistical pattern recognition | |
| CN114418175A (en) | Personnel management method and device, electronic equipment and storage medium | |
| Neves | Structural health monitoring of bridges: data-based damage detection method using machine learning | |
| CN114912678A (en) | Online automatic detection and early warning method and system for abnormal operation of power grid regulation and control | |
| CN112723075B (en) | Method for analyzing elevator vibration influence factors with unbalanced data | |
| KR101810860B1 (en) | Integrated monitoring system for personal information security product | |
| Zobel* et al. | An augmented neural network classification approach to detecting mean shifts in correlated manufacturing process parameters | |
| Gürbüz et al. | Classification rule discovery for the aviation incidents resulted in fatality | |
| Kaczmarek | Construction elements of bankruptcy prediction models in multi-dimensional Early Warning Systems | |
| Alawad et al. | Unsupervised machine learning for managing safety accidents in railway stations | |
| CN111221704B (en) | Method and system for determining running state of office management application 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |