CN104571077A - Monitoring and diagnostic device of apparatus - Google Patents
Monitoring and diagnostic device of apparatus Download PDFInfo
- Publication number
- CN104571077A CN104571077A CN201410529204.4A CN201410529204A CN104571077A CN 104571077 A CN104571077 A CN 104571077A CN 201410529204 A CN201410529204 A CN 201410529204A CN 104571077 A CN104571077 A CN 104571077A
- Authority
- CN
- China
- Prior art keywords
- equipment
- grouping
- operating condition
- diagnostic device
- efficiency
- 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
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
- G05B23/0208—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterized by the configuration of the monitoring system
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention provides a monitoring and diagnostic device of an apparatus. Even if the monitoring and diagnostic device is unmatched with the apparatus at first in an operation stage, the monitoring and diagnostic device is capable of detecting defective operation conditions and highly precisely determining the apparatus which has deep relation with it. The monitoring and diagnostic device monitors the operation of machines which parallel form a region energy supply system of a workshop. The monitoring and diagnostic device is provided with a part which measures and stores process amounts during operation of devices and ON/OFF states of the devices in operation and performs grouping according to functions of water sending, cold and heat supplying, heat exchanging and so on. The monitoring and diagnostic device is also provided with a grouping/classification part which judges groups to which the functions of the targeted workshop belong and display process amounts and relations of the grouping in all function units according to processing amounts and ON/OFF states collected in a certain period.
Description
Technical field
The present invention relates to abnormal, deteriorated, particularly unit number control etc. and the monitoring diagnostic device not mating the equipment of the exceptions such as caused performance degradation, overload using the setting of relevant control system and actual load in the equipment waited between probe vehicles early.
Background technology
Have in the mansion of concentrated thermal source, region refrigerating and heating systems, by heat resource equipment, generating set, needing the various equipment such as end equipment to form the heat generated at this, the electrically transfer system such as heat exchanger, pipe arrangement, pump, fan coil units, air-conditioning package etc. delivered to as the occupant of consumer, lessee.Especially, in the equipment of process heat, fluid, according to faces such as its efficiency, in utilization, there is bound, when load condition changes larger than it, carry out multiple equipment different for size to configure in parallel, they are switched and the correspondences such as use.Such control is called unit number control, and it is according to the distribution of the bound of load, the generation of load, selects the equipment of combination be applicable to, and according to the specification of these equipment be equipped with in parallel, installs for each system.
In addition, about the combination of such equipment, usually except the situation of transformation, according to decisions such as imagination loads during system, designed capacity distributes, switch utilization does not compatibly correspond to actual load sometimes.
Such unit number control causes due to the load variations of every day, the load variations that accompanies with season, is subject to the impact of the various main cause such as week, meteorological condition.Such unit number control himself is not abnormal, but temporarily becomes the operating condition of transition after the handover, presents many from the situation of the characteristic different clearly that operates normally.But, if it considers due to reason as described above during the life cycle of equipment, very frequent and occur in irregular timing.
In addition, in the running of the equipment when running in the equipment when underload and high load capacity, characteristic is also different, but cannot an accepted argument one upright normal and a side is abnormal.In addition, the combination of the equipment of construction system has various, and for each combination, characteristic is also different.
, in patent documentation 1, disclose (1) in a packet herein, be conceived to the fluctuating of temporal data, track is divided into grouping by the time of reviewing; (2) for split packet group, as modelling in segment space, off-set value is calculated as abnormal candidate; (3) learning data is applied flexibly (compare with reference to etc.) for referencial use, grasp rheological parameters' change with time, environmental change, state transition caused by maintenance (part exchanging) duty; (4) in modelling, utilize adopt data N number of removal (N0,1,2) the partial area matching such as regression analysis, projection distance method (such as, in a case of n=1, be mixed into 1 abnormal data if be thought of as, then removed and modelling) or the method for detecting abnormal of Part portions space law and abnormality detection system.
But the every day of the timing of equipment switching etc. is different, so be conceived to the fluctuating of the temporal data shown in patent documentation 1, insignificant situation is also many.
In addition, number of units switch after transition state under, until by control eliminate during, the movement depending on the characteristic of equipment is also significantly changed sometimes, so the state of transition being removed in next modeled method, have the possibility of the information over sight (OS)s such as the deterioration of equipment.And then, be considered as correct by usually operating at first, detection from the method for the disengaging of the state of observing wherein, cannot by the utilization change of load side, initial time point need the detection such as not mate estimating the caused unit number control such as bad to be unaccommodated situation.
[patent documentation 1] Japanese Unexamined Patent Publication 2012-230703 publication
Summary of the invention
The present invention completes in view of above-mentioned problem, its object is to the monitoring diagnostic device that a kind of equipment is provided, even if from having unmatched situation at first in operational phase, also can detect the operational situation becoming problem, and the equipment having dark relation with it can be determined accurately.
In order to reach described object, the invention provides a kind of monitoring diagnostic device of equipment, in the workshop with the multiple equipment formed in parallel, monitor the monitoring diagnostic device of the running of these multiple equipment, it is characterized in that, possess: status data collection portion, for each function, collect the process variable from each equipment; Packetizing portion, measures and the ONOFF state of the process variable in memory device running and the equipment in running, according to functional unit packetizing; And packet classification portion, according to the ONOFF state of the process variable collected with certain cycle and equipment, judge which grouping is each function in object workshop belong to respectively, show the relation of process variable in each functional unit and grouping.
According to the monitoring diagnostic device of equipment of the present invention, even if from having unmatched situation at first in operational phase, the operational situation becoming problem also can be detected, and can determine the equipment having dark relation with it accurately.
Accompanying drawing explanation
Fig. 1 is the structural drawing of the embodiment 1 of the monitoring diagnostic device of equipment of the present invention.
Fig. 2 is the structural drawing of the embodiment 2 of the monitoring diagnostic device of equipment of the present invention.
Fig. 3 is the structural drawing of the embodiment 3 of the monitoring diagnostic device of equipment of the present invention.
Fig. 4 is the structural drawing of the embodiment 4 of the monitoring diagnostic device of equipment of the present invention.
Fig. 5 is the structural drawing of the embodiment 5 of the monitoring diagnostic device of equipment of the present invention.
Fig. 6 is the structural drawing of the embodiment 6 of the monitoring diagnostic device of equipment of the present invention.
Fig. 7 is the structural drawing of the embodiment 7 of the monitoring diagnostic device of equipment of the present invention.
Fig. 8 illustrates the figure switching the example that fault monitoring together shows with ONOFF.
Fig. 9 is the figure that the example utilizing the equipment monitor of diagnostic score preferentially to show is shown.
Figure 10 is the figure of the state indication example under the state illustrating that grouping distributes difficulty.
Figure 11 illustrates from different equipment ONOFF patterns and the figure of the comparison indication example of grouping corresponding to identical imagination operating condition.
Figure 12 is the figure of the maintenance picture that grouping is shown.
[symbol description]
100: monitoring diagnostic object-based device group; 110: status data collection portion; 112: process variable; 120: packetizing portion; 122: packet classification portion; 123: result data is extracted in grouping out; 124: packet classification result data; 130: grouping relational learning portion; 140: grouping efficiency study portion; 150: imagination operating condition storage part; 151: packet interpretation portion; 152: the imagination relational expanding unit of status packet; 160: transition state detection unit; 170: equipment switches transition and to study in groups portion; 180: incorrect equipment estimating unit; 190: diagnostic result display part; 1000,2000,3000,4000,5000,6000,7000: monitoring diagnostic device (the monitoring diagnostic device of equipment).
Embodiment
Below, with reference to accompanying drawing, the embodiment of the monitoring diagnostic device of equipment of the present invention is described.
(embodiment 1 of the monitoring diagnostic device of equipment)
Fig. 1 is the structural drawing of the embodiment 1 of the monitoring diagnostic device of equipment of the present invention.
The device cluster 100 becoming monitored object is by the such heat resource equipment of cooling tower 101, refrigerating machine 102, boiler 104, gas engine 105 and need family 103 to form.They all by the identical type of more than 1 equipment, need house to form.In addition, each needs family 103 to have the heat exchanger 106,107 and the load 108 that accept heat from thermal source.In addition, about heat exchanger 106,107, cannot an accepted argument be necessary equipment.In addition, although not shown herein, drive water, hot water etc. heat agent liquid-feeding pump, also object is become to its drive unit carried out, valve etc.
Monitoring diagnostic device 1000 comprises: status data collection portion 110, for each function that liquor charging, heat supply are such, collects the process variable from these object-based devices; Packetizing portion 120, extracts the grouping of the operating condition of representation feature out for each function from them; And packet classification portion 122, after extraction grouping, the process signal of each function of classifying is respectively corresponding to which grouping.
In the diagnostic device of monitoring diagnostic device, when the judgement of the appropriate property of operating condition, any one following method can be applied, according to the application of each method, there is different embodiments.
(1) method of the efficiency of Whole Equipment is used.
(2) method of the utilization condition determined at design time point etc. is in advance used
(3) method of above-mentioned (1), (2) these both sides is used.
(embodiment 2 of the monitoring diagnostic device of equipment)
Fig. 2 is the structural drawing of the embodiment 2 of the monitoring diagnostic device of equipment of the present invention, is to apply the monitoring diagnostic device 2000 that (1) uses the method for the efficiency of Whole Equipment when the judgement of the appropriate property of operating condition.
When the running efficiency of the equipment of use, according to the efficiency rating formula registered in efficiency rating formula storage part 142 in advance, according to the data collected by data collection unit 110, the efficiency of evaluation object Whole Equipment, in packet classification portion 122, use the result (packet classification result data 124) obtained process status now and belonged to the probability of each grouping, use the grouping efficiency study portion 140 each grouping be mapped with efficiency.In addition, about the efficiency rating formula recorded in efficiency rating formula storage part 142, in the same manner as aftermentioned status data obtaining section 110, with employ can uniquely identifying amount bookmark name and in the grand supervisor language of C, Excel the efficiency arithmetic expression of spendable function, operator etc. define.
According to the result in each portion, under efficiency reduces significant situation, equipment is listed according to the special synchronously movable possibility order from high to low of the generation reduced with such efficiency, in diagnostic result display part 190, its diagnostic result is shown to surveillance operator, thus when having carried out the not good running of the running different from initial imagination, efficiency, make the qualification of the equipment dark especially with its relation become easy, improve the efficiency of diagnosis.
(embodiment 3 of the monitoring diagnostic device of equipment)
Fig. 3 is the structural drawing of the embodiment 3 of the monitoring diagnostic device of equipment of the present invention, is to apply the monitoring diagnostic device 3000 that (2) use the method for the utilization condition determined at design time point etc. in advance when the judgement of the appropriate property of operating condition.
Utilization condition contemplated by design time point corresponds to the design philosophy that such as " liquor charging function is basic with pump Isobarically Control " is such.In order to process in monitoring diagnostic device 3000 for them, use: imagination operating condition storage part 150, defines such state with simple definition; Packet interpretation portion 151, according to the result of being extracted out by packetizing portion 120 (result data 123 is extracted in the grouping as the feature of each grouping out), evaluates the degree that this grouping corresponds to imagination operating condition; And the imagination relational expanding unit 152 of status packet, evaluate the corresponding relation of imagination operating condition and the grouping relevant with other functions defined about certain function.Thus, the generation of the state outside detection imagination.
In the same manner as monitoring diagnostic device 2000, monitoring diagnostic device 3000 has diagnostic result display part 190, when having carried out the not good running of the running different from initial imagination, efficiency, making the qualification of the equipment dark especially with its relation become easy, having improved the efficiency of diagnosis.
(embodiment 4 of the monitoring diagnostic device of equipment)
Fig. 4 is the structural drawing of the embodiment 4 of the monitoring diagnostic device of equipment of the present invention, be apply when the judgement of the appropriate property of operating condition that (1) uses the method for efficiency of Whole Equipment, (2) use the monitoring diagnostic device 4000 of these both sides of method of the utilization condition determined at design time point etc. in advance.
In the same manner as monitoring diagnostic device 2000,3000, in monitoring diagnostic device 4000, also in the same manner as monitoring diagnostic device 2000, monitoring diagnostic device 4000 has diagnostic result display part 190, when having carried out the not good running of the running different from initial imagination, efficiency, make the qualification of the equipment dark especially with its relation become easy, improve the efficiency of diagnosis.
(embodiment 5 of the monitoring diagnostic device of equipment)
Fig. 5 is the structural drawing of the embodiment 5 of the monitoring diagnostic device of equipment of the present invention, is to apply the monitoring diagnostic device 5000 that (2) use the method for the utilization condition determined at design time point etc. in advance when the judgement of the appropriate property of operating condition.
In monitoring diagnostic device 5000, using the transition relationship measuring the large grouping of degree particularly corresponding with being envisioned for stable state in imagination state and the grouping beyond it, extracting the transition state detection unit 160 all becoming the grouping of steady state (SS) out.Thus, even if become the state outside the imagination of transition, imagination state stable transiently, also suppress to be judged to be exception.
In the same manner as monitoring diagnostic device 2000,3000,4000, monitoring diagnostic device 5000 has diagnostic result display part 190, when having carried out the not good running of the running different from initial imagination, efficiency, making the qualification of the equipment dark especially with its relation become easy, having improved the efficiency of diagnosis.
(embodiment 6 of the monitoring diagnostic device of equipment)
Fig. 6 is the structural drawing of the embodiment 6 of the monitoring diagnostic device of equipment of the present invention.
Monitoring diagnostic device 6000 occurs while possessing the different grouping functionally of analysis, learns the grouping relational learning portion 130 that it is relevant.
In the same manner as monitoring diagnostic device 2000,3000,4000,5000, monitoring diagnostic device 6000 has diagnostic result display part 190, when having carried out the not good running of the running different from initial imagination, efficiency, make the qualification of the equipment dark especially with its relation become easy, improve the efficiency of diagnosis.
(embodiment 7 of the monitoring diagnostic device of equipment)
Fig. 7 is the structural drawing of the embodiment 7 of the monitoring diagnostic device of equipment of the present invention.
Monitoring diagnostic device 7000 possesses the portion for extracting the high equipment of the possibility that becomes its reason out.As with the first-described, the equipment becoming diagnosis object is the system of the equipment being combined with the such identical type of multiple pump, thermal source, for the equipment of identical type, compared to the equipment of identical specification, situation about being made up of multiple equipment that specification is different is many, and according to load, environmental baseline, carry out switching and control.
The quantity of state that the running of these equipment switches measuring involves large impact, so when diagnosing abnormal, deteriorated, become the main cause of wrong diagnosis.Therefore, use study to switch the be correlated with equipment of strong grouping of the timing occurred to equipment to switch transition and to study in groups portion 170.Wherein, evaluate the process status of the front and back of the change of the ONOFF state of major equipment and each corresponding degree of dividing into groups, and accordingly, it is relevant that study operates and switches and divide into groups.
In the same manner as monitoring diagnostic device 2000,3000,4000,5000,6000, monitoring diagnostic device 7000 has diagnostic result display part 190, when having carried out the not good running of the running different from initial imagination, efficiency, make the qualification of the equipment dark especially with its relation become easy, improve the efficiency of diagnosis.
(detailed description in each portion)
< status data obtaining section 110>
In status data obtaining section 110, in advance, for each function, define the operation rule of the relation defining the output of monitoring process amount, the sensor tag title becoming its source and these sensors and monitoring process amount.Such as, for liquor charging (water) function of cold water, monitoring process amount is lift and flow, about flow, if the sensor tag corresponding with the flowmeter configured after the junction of two streams of whole conveying pump is set to TAG1234, then define as TAG1234*1.For pump, individually there is flowmeter, when measuring TAG1201, TAG1202, TAG1203 these 3 respectively, defining as TAG1201*1+TAG1202*1+TAG1203*1.About lift, be also same.
Illustrated lift-flow Figure 111 illustrates the lift of situation and the relation of flow that illustrate equipment ONOFF state with 10010.In addition, in this example embodiment, the ONOFF state of equipment is binary number performance, and everybody 1 correspond to the ONOFF of major equipment.About the ONOFF state of equipment, be not only and lift and the related pump of flow, but also represent the ONOFF state of the major equipment forming object-based device entirety.
< packetizing portion 120>
In packetizing portion 120, according to these data, packetizing is carried out to data.About the method for packetizing, the various methods such as principal component analysis (PCA), multiple regression, the method obtaining the mass centres such as center of gravity, support vector machine also can be used.In illustrated example, the mode showing to make error become below setting is adjusted to the line segment of indefinite number to generate the example of grouping.In this example embodiment, grouping is represented with 4 line segments Chf001, Chf002, Chf003, Chf004.The state of the particularly Bypass Control in the cold and hot functions of physical supply returning at temperature-represent identical equipment ONOFF state toward temperature-refrigerating machine temperature in Figure 121 in figure.Owing to being three-dimensional, indigestion, extracts 2 groupings out, illustrates them with Ctrf001 and Ctrf002.
< divides into groups relational learning portion 130>
In grouping relational learning portion 130, when the probability that the process variable measured by the result 124 in packet classification portion 122 is in Chf001 in liquor charging function is high, as high in the probability being in Ctrf002 in cold and hot functions of physical supply, occur while analyzing different groupings functionally, learn it and be correlated with.131 " Ctrf002, Chf0010.95 " shown in bottom represents that their related coefficient is 0.95 such relation in the drawings.It can use and be speculated as under moment t probability P that measured process variable is Ctrf002 (Ctrf002, t) and to be speculated as be that the probability P of Chf001 (Chf001, t) to be calculated by following formula.
1/n·Σ[τ＝0，t，P(Ctrf002，τ)·P(Chf001，τ)]
Herein, Σ [τ=0, t, F (τ)] represents the summation of the F (τ) of τ=0 to t.
< grouping efficiency study portion 140>
In the grouping efficiency study portion 140 for judging the appropriate property operated, according to measured process variable, the efficiency of evaluation object entire system, saves as data such shown in pop-up window 141 by its result.
In the efficiency calculation of objective system, need calculating formula, but in the same manner as data acquisition 110, the pre-defined operation rule employing sensor tag, accordingly, according to measurement data, counting yield.Such as, according to each need the gateway enthalpy difference of the heat supply pipe arrangement of family 103 and mass rate long-pending, obtain each energy-consuming needing house, it be added for all needing house, thus the calculated load energy.Next, add the power consumption of cooling tower and refrigerating machine, thus obtain the required electric power energy.Especially, have when there being absorption refrigerating machine etc. to use the thermal source of steam, rate of fuel consumption amount is now multiplied by the steam use amount in these thermals source with from the ratio of whole vapor volumes of boiler, thus obtain rate of fuel consumption amount, heat equivalent for electric power and fuel is multiplied by the required electric power energy and fuel use amount, thus calculate the input energy, according to the load energy and the ratio dropping into the energy, obtain the methods such as efficiency.
Like this, according to measured process variable, the efficiency of calculating object equipment at any time, and according to the result as packet classification portion 122, the probability that belongs to each grouping about each function process variable now, obtain and efficiency is multiplied by it and average and variance on the time orientation of value that obtains, thus upgrade the evaluation of estimate of the efficiency of each grouping at any time.
< imagines operating condition storage part 150>
In imagination operating condition storage part 150, in advance, with the form that pop-up window 153 is such, store the restriction formula relevant with operating condition contemplated in the design phase.Such as, the such imagination state of S11 means, due to the restriction relevant with lift, lift equals normal number, and when this example, the control of pump means makes lift become certain." N " shown in the rightest means instruction, and it is the distinguished symbol of stable state contemplated in design.The such imagination state of S12 refers to, the result that the result obtained being multiplied by negative constant to flow and constant are added and obtain becomes lift, namely exists negative relevant in lift and flow.It represents the change of flow and the pressure accompanied with the opening and closing etc. of the valve of load side, but is not envisioned for stable state, so represent with mark as "-" the rightest.S21 is the definition relevant with cold and hot functions of physical supply, and what represent that refrigerating machine temperature in is equal to or less than cold water returns temperature, is positive constant toward temperature.It is also the state expected in stable utilization, so at the rightest expression " N ".In addition, the parameter that " lift ", " refrigerating machine temperature in " " toward temperature " and " lift " in data acquisition 110 described in this rule etc. are identical, in the mode that also can differentiate in diagnostic device of the present invention, in fact define with the mark that tag number etc. can identify uniquely.
< packet interpretation portion 151>
Packet interpretation portion 151 uses such definition information 153, analyzes the characteristic parameter of the grouping of the result 123 as packetizing portion 120, evaluates each grouping and each which kind of degree of imagination state consistency, and stores in inner form 154/upgrade.Such as, when Chf001, show by linear-apporximation, so the constant S11 of lift refers to, according to the coefficient for its flow, obtain the vector of inclination, calculate the absolute value with the cosine of the vector of lift constant (slope zero), thus can evaluate relevant.In addition, the value that obtains can also be used with set constant value and the constant value execution standardization of mean distance setting the straight line representing grouping.Relevant about to S12 is also same.About Chf002 ~ chf004, also can similarly calculate relevant.About the relation of the grouping of the specific function calculated like this and the imagination operating condition of specific function, in inner form 154, be expressed as the numerical value without ().
< imagines the relational expanding unit 152> of status packet
The relational expanding unit 152 of imagination status packet of the relation of the grouping obtaining the imagination operating condition that defines for certain function and obtain about the function different from it is also in the same manner as packet interpretation portion 151, store/upgrade relevant in inner form 154, but about here, write with () in Fig. 3 and Fig. 4 and represent.In the imagination relational expanding unit 152 of status packet, calculate according to relevant between the grouping of different functions obtained by grouping relational learning portion 130 and the grouping obtained by packet interpretation portion 151 and imagination the relevant of operating condition.
Such as, when the relation of the imagination state S21 relevant with cold and hot functions of physical supply and the grouping Ctrf002 relevant with cold and hot functions of physical supply is 0.8 as shown in inner form 154, use relevant 0.95 of grouping Ctrf002 and Chf001 shown in pop-up window 131, amassed by it, determine that the imagination state S21's relevant with cold and hot functions of physical supply and the grouping Chf001 relevant with liquor charging function is relevant.In the section that capable and S21 row intersect at the Chf001 of inner form, additional () and 0.76 (=0.8 × 0.95) obtained like this is shown.In addition, especially, about the relation between such function, also can only for the imagination state computation that the band " N " being envisioned for stable utilization in design indicates.
< transition state detection unit 160>
Transition state detection unit 160 is identified in the caused change of the deterioration, fault etc. of equipment in the grouping of extracting out about each function and switches etc. with running and the change that occurs of transition ground.
Use monitoring diagnostic device 5000 collection process amount in status data collection portion 110 of our department, in packet classification portion 122, about each function, obtain the probability that it belongs to the grouping of having extracted out, use the probability belonging to each grouping obtained according to and then measured before process variable and the probability obtained according to process variable now in advance, renewal migration probability form 161 at every turn.In transition state detection unit 160, for such obtained migration probability form 161, obtain A '=A+A
2+ A
3+ A
4++ A
m.
Herein, A represents migration probability matrix, is equivalent to the matrix of the value part of migration probability form 161.A
mrepresent that the m power of this matrix is taken advantage of.
In migration probability form 161, grouping before only then affiliated before migration is longitudinally illustrated, divide into groups laterally to illustrate by infer the migration that according to measured process variable after, the probability of the SNNP of the situation that the grouping before storing and then is in advance known, so be the grouping C (row of form 161) of more than certain threshold value to the probable value Pc (t-1) under time point above, add that the probability P l (t) belonging to each grouping l that the process variable after according to migration is inferred is averaged and both can.Therefore, in each matrix key element of form 161, only describe the migration probability value Ecl (t-1) from grouping c to grouping l, but such as, also there is the information of the sampling number N do not recorded in the drawings, and upgrade as Ecl (t)=(Ecl (t-1) * N+Pl (t))/(N+1).
When to grouping assigning process amount, be not limited to distribute to all groupings, so become | A|<1, if so m is fully large, then A^m moves closer to null matrix, so | certain threshold value is set in A^m|, stops calculating.N becomes the quantity for the detectable grouping of this merit.In Figure 5, addition of underscore to relevant high grouping design being envisioned for the state of stable utilization.In the drawings, the situation that first grouping is such grouping is imagined.
If for A '/m, be multiplied by component number and be n and second later arbitrary place k is only 1 at a position and residue is that the vector of 0 is (in (0 shown in the transition state detection unit 160 of Fig. 5,1,0,) t, wherein t represents transposed matrix herein), then can obtain representing that time which that become n kind from the state being in kth grouping within the m time divides into groups is the vector of the ratio of which kind of degree, so within it, be conceived to the value that the grouping corresponding with stable state is corresponding.If this value is fully large, then can be judged as that a kth state all becomes stable state, so can be transition state in imagination by such kth condition adjudgement.
The grouping that obtains like this and the evaluation (efficiency and be transition state in imagination operating condition or imagination) for it can be used, judge the state of diagnosis object system, but when in not imagining or efficiency significantly worsens, which is that the possibility of reason is high about, what represent pointer is incorrect equipment estimating unit 180, and what study was used for the information of this supposition is, and equipment switches transition studies in groups portion 170.
< equipment switches transition and to study in groups portion 170>
It is the state of the front and back of the change of the ONOFF state evaluating major equipment as mentioned above and each corresponding degree of divide into groups that equipment switches the transition portion 170 that studies in groups, and study running switching and the function of being correlated with of dividing into groups accordingly.
Use the ONOFF state of major equipment, the state before migration shown in being expert at, the state after migration shown in row, records the Frequency (not shown in the drawings) of the migration between different ONOFF states.Such as, obtain the quantity from certain state shown in being expert to whole migrations of the ensuing state shown in row within certain period, namely, obtain and add the generation number of all migrations and the result obtained about certain row, obtain the ratio at the frequency of the relatively each migration shown in its row and the result that obtains represents probability from from certain state shown in this row to the migration of the ensuing state shown in row, so use it, the migration of the corresponding relation of selection analysis and grouping.
The ONOFF state transition of Fig. 7 is analyzed form 171 and is represented that the migration exceeding certain value P to the probability of such migration addition of the appearance of mark.Running about equipment switches, and usually in system, is set to rule, so in all combinations can imagined in theory, the situation that the migration between a part of combination in fact only occurs is many, so implement the study of such equipment ONOFF migration.When the diversity of moving is few, also can become P=0, will all migrations of migration actual achievement be had as object.
Status packet-the transition relationship analyzing the equipment ONOFF state of migration front and back and the relation of grouping analyzes form 172 for each migration Tx shown in ONOFF state transition analysis form 171, according to front and back certain during process variable, record itself and each corresponding probability that divides into groups.Whenever there is migration, by the process variable according to front and back certain period, obtain itself and each corresponding probability that divides into groups, and equalization process of making even, update probability value.About the certain period before and after migration, have about before migration, certain period of a sampling period before comprising and then, after migration, until represent that such method appears in the grouping of the probability higher than certain value.In design, when contemplating with the time switching transition together, so certain period can also be used.
In certain period, when the process variable of multiple time point of sampling, adopt and obtain and each corresponding probability that divide into groups at every turn, the such method of use maximal value both can.
Also can implement in addition according to measured process variable, obtain the process that it belongs to the probability of each grouping, its result was kept during the regular hour, with reference to detecting the value of front and back of migration, from wherein using maximal value.
In addition, as the most simple structure, be not there is equipment switch transition and to study in groups the situation in portion 170.
In this case, transition state judges that hand 160 does not also need.
How the state that transition state detection unit 160 learns transition develops in the future.Equipment switch transition study in groups portion 170 learn ONOFF switch before and after state which type of divides into groups corresponding with.By using these, the state after just can having switched according to ONOFF, which type of state prediction is in the future stable at, but do not mate about using, the detection of deterioration, urgent situation is less, so judge after also can becoming stable state actually.
In this case, can by the efficiency value of grouping that obtained by grouping efficiency study portion 140 or, the degree of correspondence of imagination running that obtained by packet interpretation portion 151, evaluate the appropriate property of grouping.For each of the ONOFF pattern of equipment, generate grouping, thus the equipment operated during normality outside inefficient state, imagination can be considered as be use do not mate, the possibility of the reason of deterioration is high.Or, the equipment now stopped can be removed from reason candidate.
The incorrect equipment estimating unit 180> of <
Incorrect equipment estimating unit 180 is for the not good state of the operating condition outside the imagination detected by grouping efficiency study portion 140, packet interpretation portion 151, the imagination relational expanding unit of status packet 152, transition state detection unit 160, efficiency, the status packet-transition relationship being used as equipment to switch the result in transition class study portion 170 analyzes form 172, determines it is the equipment that the possibility of reason is high.
< is about considering the situation > switching transition together with the ONOFF of equipment
The situation considering to switch transition together with the ONOFF of equipment is described.
Such as, there are 7 at main equipment, when having showed its ONOFF with binary number, become 1100101.Herein, second equipment becomes OFF from ON, becomes 1000101.
Together in this, if analyze form 172 according to status packet-transition relationship, the grouping of known function y is from becoming the state transition of Cy1 with probability 0.8 to the state becoming Cy2 with probability 0.75, from the result of transition state detection unit 160, known Cy2 is transition state in the imagination in equipment ONOFF pattern 1000101, then alarm can be suppressed to occur as normal transition state, high at the migration probability to efficiency significantly reduced grouping Cy3, and process variable be afterwards calculated as the probability in fact becoming Cy3 high when, it is high that the stopping of known 2nd equipment and efficiency reduce related possibility.
< is about not corresponding to the specific grouping learnt situation > in the state after migration is during regulation
Known when not corresponding to the specific grouping learnt in the state after moving is during specifying, there is the possibility of certain problem high in the equipment becoming OFF in this migration or the equipment becoming ON, when such state move at same ONOFF, other ONOFF move in occur, when this relates to ON or OFF of same equipment, in the starting method, method of shutting down of this equipment, problematic possibility is high.
In such judgement, for becoming the equipment of ON, becoming each of equipment of OFF, form is set, ON or OFF recording each equipment becomes the number of times that the number of times of inefficient state or the probability corresponding to the grouping learnt maintain below certain value as former state and both can more than certain period, thus, even if different by the migration representing ONOFF pattern entirety as shown in Figure 8, in the migration that efficiency reduction etc. is caused, become ON or OFF and the equipment that relates to, can identify according to its number of times.
< is about priority > in diagnostic score
In the above description, about process variable correspond to grouping probability, grouping corresponds to imagination shape probability of state, the efficiency value of grouping has on average, variance, not corresponding to stable imagination state grouping reach stable imagination shape probability of state in the future, the ONOFF of equipment switches and the relevant probability etc. that divides into groups, except the calculating self of the ONOFF of equipment, efficiency value, all to process probabilityly.Therefore, about the estimation result of incorrect equipment, also for ON or OFF of each equipment, be recorded as the scoring based on described probability, according to scoring order from high to low, be shown in diagnostic result display part 190.
About diagnostic score, by using, to be multiplied by the efficiency value of grouping be the probability of this grouping and the value η that obtains, the η equalization that the state computation that to make for identical equipment be ON goes out and obtaining.In diagnostic result display part 190, according to form as shown in Figure 9, by diagnostic score (the equalization result of η) additional priority order, indicate the equipment of in-problem possibility, thus equipment that can be high for the possibility causing efficiency to reduce, preferentially take the correspondences such as investigation.
< is about imagining in design but the definition > of less desirable state
In the above description, in imagination operating condition storage part 150, be illustrated premised on the steady running state imagined, but also can define by-pass operation, the upper imaginations of design such as running of bleeding but less desirable state.In this case, being the definition of such state to identify, except with except " N " and "-" these 2 kinds of marks represented, as the value representing less desirable state, such as, adding " A " etc.Such state is not abnormal but takes place frequently, and ought to not imagine, so be called compromise operating condition in imagination.Similarly, the state after the emergent stopping of equipment is assigned to define by logo area such as " E ".
According to measured process variable, obtain the probability that it belongs to each grouping, and counting yield, exceed in the time-bands of certain value at it, according to probability order from high to low, point out the imagination state corresponding with produced grouping, if thus it is classified as " A ", " E ", then can be judged as that the reason of degradation in efficiency is that imagining interior compromise running, emergent stopping etc. there occurs more than imagination.
Also can obtain be move to be marked as " A ", " E " the high grouping of the probability of the large grouping of probability corresponding to imagination operating condition until the moving average of probability of time of moving short grouping, when it exceedes the value of regulation, display warning.Thus, the time point of the known state in transition, compromises before running, has this possibility in entering into extremely, imagining.
< about all only correspond to the such state of below certain low probability for any grouping of extracting out at this time point and continue certain period situation, in certain period, such state reaches the situation > of regular hour ratio
When all only corresponding to the situation during the such state of below certain low probability continues necessarily for any grouping of extracting out at this time point, such state reaching regular hour ratio in certain period, also can be as shown in Figure 10, in overlap on curve 111 and the such curve of curve 121, by the color can distinguished with other, shape (in figure ☆), display state during this period.There is the possibility of the state becoming non-experience before this thereby, it is possible to inspire, and easily can grasp process compared to the point that to become which type of state so in the past, be in which type of state compared to contemplated operating condition.
Other mode > of display packing in < diagnostic result display part 190
In addition, also can in diagnostic result display part 190, about for identical imagination operating condition with the probability of identical degree corresponding, grouping that equipment ONOFF pattern is different, the difference of the parameter of the efficiency between them, grouping is shown together with the difference of ONOFF pattern.Figure 11 illustrates an example.
Known adding at pump has been started after one, and the parameter (stationary value of pressure) of high grouping relevant to S11 is partial to upside than 550kPa.By carrying out such comparison display, passing through different equipment, when reaching operating condition in identical imagination, easily can grasp the difference caused by device structure, when the difference of efficient, parameter, can according to the difference etc. of constitution equipment, easily reduction becomes the candidate of the equipment of reason.
In the example of Figure 11, by the additional pump started, stable imagination operating condition becomes on high-tension side running, when the result of efficiency rating causes efficiency to reduce, can easily infer the setting of additional priming pump bad etc. be the possibility of reason.
About determining that according to process variable this packet of samples is contained in the process of which grouping, according to multiple process variable data, they is divided into the method for multiple grouping, there are generally known various grouping gimmicks, so use them both can.
Such as, about the process of extracting grouping from multiple process variable data out, the method generally known by kmeans method, SVM (support vector machine) etc. implements packet fragmentation, about the decision of model parameter, carry out principal component analysis (PCA) etc. about split data, thus straight line, face can be extracted out.As in advance by the method for straight line, face performance, also has the method using Hough transform.Also can when straight line, select arbitrary 2 samplings, when plane, select arbitrary 3 samplings, when n ties up lineoid, select arbitrary n sampling, investigate their slope, the distribution of normal vector, using frequency be certain above region as object, thus be limited to the space explored in Hough transform in advance.Also can be the parameter (mean value and variance, its probability of happening) of the Gaussian mixtures representing Frequency distribution as generation model setting, and the method inferring them to divide into groups.
Be contained in the process of which grouping as the sampled data judged according to process variable now, when using Gaussian mixtures in the model divided into groups, according to the distribution function of grouping, can determine that sampled data is contained in the probability of this grouping.The situation of straight line etc. has been obtained, when having extracted the characteristic quantities such as principal component analysis (PCA) out from grouping in Hough transform, after the distributional assumption of the error of the model will divided into groups from these (formula of straight line, plane) is Gaussian distribution etc., this distribution is determined according to learning data, according to the error of sampled process variable and model, according to its distribution, decisive probability.
In addition, when the conveyance function of such as Fig. 3, about the quantity of polynomial variable, also can as shown in curve 111, using lift and flow these 2 as process variable, with straight line implementation model.When cold and hot functions of physical supply, also using returning temperature, toward temperature, refrigerating machine temperature in these 3 or can also add that 4 of cooling water inlet temperature and flow or cooling water outlet temperature to 6 as process variable.
In the parameter of grouping, the parameter (average and variance) of the model that straight line, face etc. can be used to divide into groups, Multi-dimensional Gaussian distribution.By with certain cycle implementation model, the parameter between more corresponding model, when there being the change of more than certain threshold value compared to these variances, also can represent its result by diagnostic result display part 190.
Other mode > of display packing in < diagnostic result display part 190
About grouping and the correspondence imagining operating condition, user also can be made to confirm in the picture that Figure 12 is such via diagnostic result display part 190.There is the interface can selecting multiple grouping to indicate comprehensive or instruction to delete, when there being deletion, the data of association being deleted from various form (131,141,154,161,172 etc.).When having comprehensive, using belong to them data as a grouping, implement calculating again of parameter.Such as, for the sampling belonging to merged grouping, carry out principal component analysis (PCA), if take out the composition of necessary dimension, then can determine the parameter that the slope of straight line and intercept, the normal vector of plane and intercept etc. are necessary.In addition, the attribute (above-mentioned " N " " A " " E " "-" etc.) of grouping can also be set.And then, now, also the state of each function can be shown together with imagination operating condition.
In addition, in the point that the maintenance of grouping is such, when the efficiency of the grouping Cyi relevant with certain function y is meta and variance is large, this is grouped in efficiency face, cannot extract feature out fully, has the possibility of the set being multiple grouping.Under such a condition, according to the result 121 in grouping relational learning portion 130, when being Cyi, the possibility being simultaneously the grouping Czj of other function z is not high wittingly, but when being Czj, when being the high such result of the possibility of Cyi, by beyond the situation of more than the setting that the class Cyi of function y is separated into be the probability of Czj be and its these 2 kinds, the resolution of the grouping about function y can be improved.
Also can when the making of dividing into groups, the analysis of corresponding relation with grouping, according to summer with, winter use, the form such as middle Ji Yong, according to carrying out classification to learn to data season and diagnosing.And then, can be also condition and after having carried out data classifying, carry out learning and diagnosing at the trafficwise such as to stop doing business with week, feast day, summer etc.
Claims (8)
1. a monitoring diagnostic device for equipment, is in the workshop with the multiple equipment formed in parallel, monitors the monitoring diagnostic device of the running of these multiple equipment, it is characterized in that comprising:
Status data collection portion, for each function, collects the process variable from each equipment;
Packetizing portion, measures and the ONOFF state of the process variable in memory device running and the equipment in running, according to functional unit packetizing; And
Packet classification portion, according to the ONOFF state of the process variable collected with certain cycle and equipment, judges which grouping is each function in object workshop belong to respectively,
Show the relation of process variable in each functional unit and grouping.
2. the monitoring diagnostic device of equipment according to claim 1, characterized by further comprising:
Grouping efficiency study portion, measuring process amount in equipment operation, according to its measurement result, assess effectiveness, to grouping and the efficiency value additional relationships of each function being judged to belong to according to the process variable of this time point;
Incorrect equipment estimating unit, during certain, efficiency is lower than when exceeding certain ratio during set value, equipment in being operate with efficiency lower than the time point that the grouping of set value is corresponding and with efficiency exceed time point corresponding to the grouping of set value not operate in equipment in, according to the order of number of times from many to few exceeding described certain ratio, be set to the equipment that the possibility of problem is high;
Diagnostic result display part, according to the priority based on the estimation result obtained by described incorrect equipment estimating unit, has the equipment of the possibility of questions and prospect to operator prompting.
3. the monitoring diagnostic device of equipment according to claim 1, characterized by further comprising:
Imagination operating condition storage part, sets the operating condition imagined for each function by formula;
Packet interpretation portion, whether the relation of the grouping calculating each function of being extracted out by described packetizing portion and the formula representing operating condition, judges to divide into groups as contemplated state;
Incorrect equipment estimating unit, situation corresponding with contemplated operating condition during certain lower than set number of times during when exceeding certain ratio, in equipment in being operate in the time point corresponding with the grouping of not contemplated operating condition and the equipment in not operating in the time point corresponding with the grouping being contemplated operating condition, according to the order of number of times from many to few exceeding described certain ratio, be set to the equipment that the possibility of problem is high; And
Diagnostic result display part, according to the priority based on the estimation result obtained by described incorrect equipment estimating unit, has the equipment of the possibility of questions and prospect to operator prompting.
4. the monitoring diagnostic device of equipment according to claim 1, characterized by further comprising:
Grouping efficiency study portion, measuring process amount in equipment operation, according to its measurement result, assess effectiveness, to grouping and the efficiency value additional relationships of each function being judged to belong to according to the process variable of this time point;
Imagination operating condition storage part, sets the operating condition imagined for each function by formula;
Packet interpretation portion, whether the relation of the grouping calculating each function of being extracted out by described packetizing portion and the formula representing operating condition, judges to divide into groups as contemplated state;
Incorrect equipment estimating unit, efficiency during certain lower than set value during or when not exceeding certain ratio during imagination operating condition, in efficiency lower than set value or in the time point corresponding with the grouping not imagining operating condition operates and efficiency exceedes set value or in equipment in not operating when imagining operating condition, according to the order of number of times from many to few exceeding described certain ratio, be set to the equipment that the possibility of problem is high; And
Diagnostic result display part, according to the priority based on the estimation result obtained by described incorrect equipment estimating unit, has the equipment of the possibility of questions and prospect to operator prompting.
5. the monitoring diagnostic device of the equipment according to any one in Claims 1 to 4, characterized by further comprising:
Grouping relational learning portion, while evaluating between the grouping about different 2 functions extractions, probability of happening is to learn the relevant of them,
In described packetizing portion, the function of a side be the 1st grouping, the opposing party function be other the 2nd grouping probability high, the function of this opposing party is the 2nd grouping, the function of this side divide into groups the 3rd different possibilities of divide into groups from the 1st, the 2nd when exceeding certain value, uses the 1st result of determination of dividing into groups to split the 2nd and divide into groups.
6., according to the monitoring diagnostic device of the equipment of any one in claim 2 ~ 4, characterized by further comprising:
Transition state detection unit, store the migration between multiple groupings of extracting out about each function, migration probability is obtained according to its Frequency, the situation of the operating condition contemplated by not or efficiency are low states, the state of the transition of the operating condition contemplated by judgement or the high state of the efficiency that becomes or any one not in them
In described incorrect equipment estimating unit, when contemplated operating condition, become the state of transition of efficiency high state, not using when the possibility of the problem of determining apparatus is now the number of times of ON.
7. the monitoring diagnostic device of the equipment according to any one in Claims 1 to 4, it is characterized in that: in described packetizing portion, except the measured value of the ONOFF state of the equipment in the process variable in equipment operation and running, also store week and even season and even type of operation in the lump, packetizing is implemented according to functional unit, in described packet classification portion, use these information to judge which type of grouping is operating condition correspond to.
8., according to the monitoring diagnostic device of the equipment of any one in Claims 1 to 4, characterized by further comprising:
Equipment switches transition and to study in groups portion, according to the process status before and after the change of the ONOFF state of equipment, detects the migration change grouping of accompanying with the ONOFF of equipment,
In described incorrect equipment estimating unit, the efficiency of the grouping before and after switching according to the running of the equipment occurred in during certain or whether be contemplated operating condition, when it lower than the ratio of certain value or when being judged to be that the number of times of not contemplated operating condition exceedes certain ratio, about becoming the equipment of ON from OFF after before the handover or becoming the equipment of OFF from ON, according to the order of its number of times from many to few, be set to the equipment that the possibility of problem is high.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2013-213977 | 2013-10-11 | ||
| JP2013213977A JP2015076058A (en) | 2013-10-11 | 2013-10-11 | Facility monitoring diagnostic apparatus |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104571077A true CN104571077A (en) | 2015-04-29 |
| CN104571077B CN104571077B (en) | 2017-08-25 |
Family
ID=53000828
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410529204.4A Expired - Fee Related CN104571077B (en) | 2013-10-11 | 2014-10-10 | The monitoring diagnostic device of equipment |
Country Status (3)
| Country | Link |
|---|---|
| JP (1) | JP2015076058A (en) |
| CN (1) | CN104571077B (en) |
| IN (1) | IN2014DE02873A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107798161A (en) * | 2016-08-30 | 2018-03-13 | 阿自倍尔株式会社 | Design evaluatio accessory system |
| CN108604094A (en) * | 2016-03-31 | 2018-09-28 | 三菱日立电力系统株式会社 | The abnormality diagnostic method of equipment and the apparatus for diagnosis of abnormality of equipment |
| CN110226140A (en) * | 2017-01-25 | 2019-09-10 | Ntn株式会社 | State monitoring method and state monitoring apparatus |
| TWI728646B (en) * | 2020-01-09 | 2021-05-21 | 池御科技有限公司 | Temperature control method and temperature controller using the same |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6150934B1 (en) * | 2016-10-17 | 2017-06-21 | 三菱重工業株式会社 | Information processing method, information processing apparatus, program, and information processing system |
| JP6798968B2 (en) | 2017-11-22 | 2020-12-09 | ファナック株式会社 | Noise cause estimation device |
| JP6871877B2 (en) | 2018-01-04 | 2021-05-19 | 株式会社東芝 | Information processing equipment, information processing methods and computer programs |
| JPWO2021171387A1 (en) * | 2020-02-26 | 2021-09-02 | ||
| CN113533875B (en) * | 2020-04-22 | 2024-02-23 | 释普信息科技（上海）有限公司 | Method for intelligently judging running state of laboratory equipment and calculating equipment utilization rate |
| KR102472081B1 (en) * | 2020-11-26 | 2022-11-30 | (주)심플랫폼 | A System and Method for Monitoring Manufacturing Process |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH09248739A (en) * | 1996-03-14 | 1997-09-22 | Hitachi Ltd | Monitoring device for operation condition |
| CN201035376Y (en) * | 2006-12-22 | 2008-03-12 | 浙江大学 | Failure diagnosis device under small sample conditional in the process of manufacturing production |
| CN101158873A (en) * | 2007-09-26 | 2008-04-09 | 东北大学 | Non-linearity process failure diagnosis method |
| CN102339389A (en) * | 2011-09-14 | 2012-02-01 | 清华大学 | Fault detection method for one-class support vector machine based on density parameter optimization |
| JP2013164668A (en) * | 2012-02-09 | 2013-08-22 | Hitachi Systems Ltd | Fault monitoring system, incident tabulation method, and program |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5427107B2 (en) * | 2010-05-20 | 2014-02-26 | 株式会社日立製作所 | Monitoring and diagnosis apparatus and monitoring diagnosis method |
| JP2012137934A (en) * | 2010-12-27 | 2012-07-19 | Hitachi Ltd | Abnormality detection/diagnostic method, abnormality detection/diagnostic system, abnormality detection/diagnostic program and company asset management/facility asset management system |
| JP5498540B2 (en) * | 2012-07-19 | 2014-05-21 | 株式会社日立製作所 | Anomaly detection method and system |
-
2013
- 2013-10-11 JP JP2013213977A patent/JP2015076058A/en active Pending
-
2014
- 2014-10-08 IN IN2873DE2014 patent/IN2014DE02873A/en unknown
- 2014-10-10 CN CN201410529204.4A patent/CN104571077B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH09248739A (en) * | 1996-03-14 | 1997-09-22 | Hitachi Ltd | Monitoring device for operation condition |
| CN201035376Y (en) * | 2006-12-22 | 2008-03-12 | 浙江大学 | Failure diagnosis device under small sample conditional in the process of manufacturing production |
| CN101158873A (en) * | 2007-09-26 | 2008-04-09 | 东北大学 | Non-linearity process failure diagnosis method |
| CN102339389A (en) * | 2011-09-14 | 2012-02-01 | 清华大学 | Fault detection method for one-class support vector machine based on density parameter optimization |
| JP2013164668A (en) * | 2012-02-09 | 2013-08-22 | Hitachi Systems Ltd | Fault monitoring system, incident tabulation method, and program |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108604094A (en) * | 2016-03-31 | 2018-09-28 | 三菱日立电力系统株式会社 | The abnormality diagnostic method of equipment and the apparatus for diagnosis of abnormality of equipment |
| CN107798161A (en) * | 2016-08-30 | 2018-03-13 | 阿自倍尔株式会社 | Design evaluatio accessory system |
| CN110226140A (en) * | 2017-01-25 | 2019-09-10 | Ntn株式会社 | State monitoring method and state monitoring apparatus |
| TWI728646B (en) * | 2020-01-09 | 2021-05-21 | 池御科技有限公司 | Temperature control method and temperature controller using the same |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104571077B (en) | 2017-08-25 |
| JP2015076058A (en) | 2015-04-20 |
| IN2014DE02873A (en) | 2015-07-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104571077A (en) | Monitoring and diagnostic device of apparatus | |
| He et al. | Fault detection and diagnosis of chiller using Bayesian network classifier with probabilistic boundary | |
| Li et al. | Data partitioning and association mining for identifying VRF energy consumption patterns under various part loads and refrigerant charge conditions | |
| Seem | Using intelligent data analysis to detect abnormal energy consumption in buildings | |
| Li et al. | A review of virtual sensing technology and application in building systems | |
| Kocyigit | Fault and sensor error diagnostic strategies for a vapor compression refrigeration system by using fuzzy inference systems and artificial neural network | |
| US7444251B2 (en) | Detecting and diagnosing faults in HVAC equipment | |
| KR101316486B1 (en) | Error detection method and system | |
| JP7260292B2 (en) | Abnormality diagnosis device and abnormality diagnosis method | |
| CN102375452B (en) | Event-driven data mining method for improving fault code settings and isolating faults | |
| KR20110099245A (en) | Method for collecting device state information and device state information collection kit used for the method | |
| CN107796609B (en) | Water chilling unit fault diagnosis method based on DBN model | |
| CN105637432A (en) | Identifying anomalous behavior of a monitored entity | |
| JP2009289262A (en) | System and method for advanced condition monitoring of asset system | |
| Behfar et al. | Automated fault detection and diagnosis methods for supermarket equipment (RP-1615) | |
| CN102175282A (en) | Method for diagnosing fault of centrifugal air compressor based on information fusion | |
| EP3584656B1 (en) | Risk assessment device, risk assessment method, and risk assessment program | |
| Zucker et al. | Improving energy efficiency of buildings using data mining technologies | |
| WO2018181120A1 (en) | Information processing device, information processing method, and program | |
| CN105408828A (en) | Method of estimation on a curve of a relevant point for the detection of an anomaly of a motor and data processing system for the implementation thereof | |
| JP6915693B2 (en) | System analysis method, system analyzer, and program | |
| KR20130065844A (en) | System and method for managing energy equipments efficiency in intelligent building | |
| Mtibaa et al. | Refrigerant leak detection in industrial vapor compression refrigeration systems using machine learning | |
| Guo et al. | Intelligent outlier detection for HVAC system fault detection | |
| JP2007026134A (en) | Abnormality decision device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170825 Termination date: 20211010 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |