CN109460419A - A kind of equipment state change events monitoring method - Google Patents
A kind of equipment state change events monitoring method Download PDFInfo
- Publication number
- CN109460419A CN109460419A CN201811096084.8A CN201811096084A CN109460419A CN 109460419 A CN109460419 A CN 109460419A CN 201811096084 A CN201811096084 A CN 201811096084A CN 109460419 A CN109460419 A CN 109460419A
- Authority
- CN
- China
- Prior art keywords
- cluster
- power
- event
- group
- equipment
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- 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
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Remote Monitoring And Control Of Power-Distribution Networks (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of equipment state change events monitoring methods, including collect electricity consumption data；Threshold process is carried out to the changed power of total load；Figure signal is generated, all rising edges and failing edge of the node of figure signal to event are indexed；The power event extracted by initial threshold is classified, cluster is formed；Adaptive threshold modification and optimization cluster；Each positive event cluster and the immediate negative event cluster of amplitude are subjected to screening feature pairing；By calculating exemplary power, match the mean value of the changed power absolute value of each cluster group；The method of the present invention is by handling equipment electricity consumption data, realize the load decomposition of electricity consumption, to which real time monitoring is realized in the energy consumption to each equipment, grasp the electricity consumption situation of each equipment, facilitate user and understands each equipment electricity consumption situation, to carry out required energy consumption analysis according to the electricity consumption situation and formulate reasonable energy conservation program, and then unnecessary power consumption is reduced, economized on resources.
Description
Technical field
The present invention relates to field of power detection more particularly to a kind of equipment state change events monitoring methods.
Background technique
Electric energy is widely used in life, has become indispensable a part in our daily lifes, just because such as
This enters a unprecedented height to the research of electric energy with the development of the times, wherein the energy consumption ratio of household electricity
The ratio that example accounts for total energy consumption is quite high, it is desirable to achieve the effect that reduce the energy consumption waste in household electricity, to the consumption of household electricity
Energy situation carries out detection and is necessary step, and therefore, carrying out analysis and research to the electricity consumption energy consumption condition of the various electric appliances of family is to work as
A preceding problem urgently to be solved.
In the prior art, people detect household electricity energy consumption using two methods, and one is directly in circuit
Ammeter is installed directly to be measured on main road, this method can only obtain the overall energy consumption data of household electricity, and cannot obtain
Take the electricity consumption data of each equipment or electric appliance；Another method be each electrical equipment access load end install ammeter into
Row electricity consumption data metering, although this method can obtain the energy consumption data of each electrical equipment respectively, due to metering ammeter
It is connection relationship with electrical equipment, when electrical equipment is moved to elsewhere, metering ammeter also has to move to electrical equipment
Place, and be mounted on the mobile region of electrical equipment, the place where electrical equipment is moved to does not have installation metering
The environment of ammeter then cannot achieve energy consumption detection, and this method is complicated for operation, and inefficiency, fulfilling rate is low, and must be to each
Equipment matches a metering ammeter respectively, leads to measurement cost height.
Summary of the invention
The present invention provides a kind of equipment state change events monitoring methods, to solve the electricity consumption to each equipment or electric appliance
Complicated for operation when energy consumption data, inefficiency, fulfilling rate is low, problem at high cost, realizes and more saves, more intelligently obtains
The electricity consumption situation of each electrical equipment is taken to carry out electricity consumption data.
In order to solve the above-mentioned technical problem, the embodiment of the invention provides a kind of equipment state change events monitoring method,
Include:
Collect one or more electricity consumption datas of equipment and the power data of total load；
Threshold process is carried out to the changed power of total load, power event is extracted and forms candidate events group；
Figure signal is generated, all rising edges and failing edge of the node of figure signal to event are indexed；
The power event extracted by initial threshold is classified, cluster is formed；
Adaptive threshold modification and optimization cluster are carried out to sorted power event is carried out；
Each positive event cluster and the immediate negative event cluster of amplitude are subjected to screening feature pairing, and further screening obtains
Optimal matched group；
Exemplary power by calculating each equipment changes numerical value, is allowed to the changed power with each cluster group after characteristic matching
The mean value of absolute value is matched, and corresponding equipment Electrical change numerical value is obtained.
Preferably, described that the power event extracted by initial threshold is classified, specific as follows:
The candidate events group is divided into positive group and negative power group, and it is clustered respectively；
The similar all edges in power edge in the positive group and the negative power group are grouped into respectively same
In event cluster, and the edge being grouped into same event cluster is deleted from power packages；
It repeats the above steps, power edge is clustered, until all events realize cluster.
Preferably, sorted power event progress adaptive threshold modification and optimization cluster are carried out for described pair,
It is specific as follows:
Using that the smallest event cluster of the absolute value of the mean value of power cluster group as new adaptive threshold, carries out threshold value and repair
Change；
Improve the cluster degree of cluster group, optimization cluster.
Preferably, described using that the smallest event cluster of the absolute value of the mean value of power cluster group as newly adaptive
Threshold value is answered, threshold modifying is carried out, specific as follows:
Using that the smallest event cluster of the absolute value of the mean value of positive cluster group as new adaptive positive threshold value；
Using that the smallest event cluster of the absolute value of the mean value of negative power cluster group as new adaptive negative threshold value；
The lesser changed power event of changed power in positive cluster group and negative power cluster group is removed respectively.
Preferably, the cluster degree for improving cluster group, optimization cluster are specific as follows:
The cluster degree of positive cluster group and the event cluster in negative power cluster group is respectively increased；
The similar all edges in power edge in the positive cluster group and the negative power cluster group are grouped into respectively
In same event cluster, cluster is advanced optimized.
Preferably, it is described advanced optimize cluster after, further includes: the cluster group of less element is merged into adjacent member
Phylogenetic group more than element is to keep just clustering the balance between the quantity of negative cluster.
Preferably, described each positive event cluster and the immediate negative event cluster of amplitude are subjected to screening feature to match
It is right, and further screening obtains optimal matched group, it is specific as follows:
Positive event cluster and negative power event cluster are rearranged from big to small by mean amplitude, so that positive event
Cluster and negative power event cluster are corresponding, to be more convenient for matching；
Time of origin cluster corresponding with the power event of positive event cluster and negative event cluster is formed, so that characteristic matching is carried out,
Filter out candidate matched group；
In conjunction with the matched group that figure signal is optimal to the matched group further screening filtered out；
It repeats the above steps, optimal matched group is filtered out to each power event cluster, until to all power things
Part cluster completes screening.
Preferably, after described pair of all power event cluster completes screening, further includes: to the power for failing pairing
It is deleted from event cluster at edge.
Preferably, the exemplary power by calculating each equipment changes numerical value, be allowed to after characteristic matching
The mean value of the changed power absolute value of each cluster group is matched, and corresponding equipment Electrical change numerical value is obtained, specific as follows:
The typical change power that each equipment is calculated by the power data being collected into establishes device label library and by each equipment
Typical change power data be stored in the device label library as label；
By the exemplary power of equipment in the mean value and tag library of the changed power absolute value of each cluster group after characteristic matching
Variation matches, if matching meets, which belongs to corresponding equipment, if marked without corresponding label
Label add new label in library；
According to the result of pairing by the time difference of each power-up ramp and failing edge multiplied by changed power, and cumulative cluster
All changed power events, obtain the total electricity using the equipment in group.
A kind of equipment state change events monitoring method, comprising:
Collect one or more electricity consumption datas of equipment and the power data of total load；
Threshold process is carried out to the changed power of total load, power event is extracted and forms candidate events group；
Figure signal is generated, all rising edges and failing edge of the node of figure signal to event are indexed；
The power event extracted by initial threshold is classified, cluster is formed；
Adaptive threshold modification and optimization cluster are carried out to sorted power event is carried out；
Each positive event cluster and the immediate negative event cluster of amplitude are subjected to screening feature pairing, and further screening obtains
Optimal matched group；
Exemplary power by calculating each equipment changes numerical value, is allowed to the changed power with each cluster group after characteristic matching
The mean value of absolute value is matched, and corresponding equipment Electrical change numerical value is obtained.
It is described that the power event extracted by initial threshold is classified, specific as follows:
The candidate events group is divided into positive group and negative power group, and it is clustered respectively；
The similar all edges in power edge in the positive group and the negative power group are grouped into respectively same
In event cluster, and the edge being grouped into same event cluster is deleted from power packages；
It repeats the above steps, power edge is clustered, until all events realize cluster.
Described pair carries out sorted power event and carries out adaptive threshold modification and optimization cluster, specific as follows:
Using that the smallest event cluster of the absolute value of the mean value of power cluster group as new adaptive threshold, carries out threshold value and repair
Change；
Improve the cluster degree of cluster group, optimization cluster.
It is described using that the smallest event cluster of the absolute value of the mean value of power cluster group as new adaptive threshold, carry out threshold
Value modification, specific as follows:
Using that the smallest event cluster of the absolute value of the mean value of positive cluster group as new adaptive positive threshold value；
Using that the smallest event cluster of the absolute value of the mean value of negative power cluster group as new adaptive negative threshold value；
The lesser changed power event of changed power in positive cluster group and negative power cluster group is removed respectively.
The cluster degree for improving cluster group, optimization cluster are specific as follows:
The cluster degree of positive cluster group and the event cluster in negative power cluster group is respectively increased；
The similar all edges in power edge in the positive cluster group and the negative power cluster group are grouped into respectively
In same event cluster, cluster is advanced optimized.
It is described advanced optimize cluster after, further includes: the cluster group of less element is merged into the phylogenetic group more than adjacent element
To keep the balance between positive cluster and the quantity of negative cluster.
It is described that each positive event cluster and the immediate negative event cluster of amplitude are subjected to screening feature pairing, and further screen
Optimal matched group is obtained, specific as follows:
Positive event cluster and negative power event cluster are rearranged from big to small by mean amplitude, so that positive event
Cluster and negative power event cluster are corresponding, to be more convenient for matching；
Time of origin cluster corresponding with the power event of positive event cluster and negative event cluster is formed, so that characteristic matching is carried out,
Filter out candidate matched group；
In conjunction with the matched group that figure signal is optimal to the matched group further screening filtered out；
It repeats the above steps, optimal matched group is filtered out to each power event cluster, until to all power things
Part cluster completes screening.
After described pair of all power event cluster completes screening, further includes: to failing the power edge of pairing from event cluster
Middle deletion.
The exemplary power by calculating each equipment changes numerical value, is allowed to the power with each cluster group after characteristic matching
The mean value of change absolute value is matched, and corresponding equipment Electrical change numerical value is obtained, specific as follows:
The typical change power that each equipment is calculated by the power data being collected into establishes device label library and by each equipment
Typical change power data be stored in the device label library as label；
By the exemplary power of equipment in the mean value and tag library of the changed power absolute value of each cluster group after characteristic matching
Variation matches, if matching meets, which belongs to corresponding equipment, if marked without corresponding label
Label add new label in library；
According to the result of pairing by the time difference of each power-up ramp and failing edge multiplied by changed power, and cumulative cluster
All changed power events, obtain the total electricity using the equipment in group.
Compared with the prior art, the embodiment of the present invention has the following beneficial effects:
1, it by handling electrical equipment electricity consumption general power, realizes the load decomposition for carrying out electricity consumption, reaches essence
The energy consumption data for really identifying each equipment grasp the electricity consumption situation of each equipment, realize the high-efficient and high effect of fulfilling rate
Fruit；
2, it does not need that metering ammeter is installed respectively at each electrical equipment, save the cost is easy to operate.
Detailed description of the invention
Fig. 1: for the specific steps flow chart of monitoring method embodiment of the present invention；
Fig. 2: for the specific flow chart of step S4 in monitoring method embodiment of the present invention；
Fig. 3: for the specific flow chart of step S5 in monitoring method embodiment of the present invention；
Fig. 4: for the specific flow chart of step S51 in monitoring method embodiment of the present invention；
Fig. 5: for the specific flow chart of step S52 in monitoring method embodiment of the present invention；
Fig. 6: for the specific flow chart of step S6 in monitoring method embodiment of the present invention；
Fig. 7: for the specific flow chart of step S7 in monitoring method embodiment of the present invention；
Fig. 8: for the schematic diagram for forming candidate events group in monitoring method embodiment of the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
Fig. 1 is please referred to, the preferred embodiment of the present invention provides a kind of equipment state change events monitoring method, comprising:
S1 collects one or more electricity consumption datas of equipment and the power data of total load；
S2 carries out threshold process to the changed power of total load, extracts power event and form candidate events group；
S3 generates figure signal, all rising edges and failing edge of the node of figure signal to event is indexed；
S4 classifies the power event extracted by initial threshold, forms cluster；
S5, to carrying out, sorted power event carries out adaptive threshold modification and optimization clusters；
Each positive event cluster and the immediate negative event cluster of amplitude are carried out screening feature pairing, and further screened by S6
Obtain optimal matched group；
S7, the exemplary power by calculating each equipment change numerical value, are allowed to the power with each cluster group after characteristic matching
The mean value of change absolute value is matched, and corresponding equipment Electrical change numerical value is obtained.
Referring to Fig. 2, in the present embodiment, step S4 is specifically included:
The candidate events group is divided into positive group and negative power group, and clustered respectively to it by S41；
S42 is respectively grouped into the similar all edges in power edge in the positive group and the negative power group
In same event cluster, and the edge being grouped into same event cluster is deleted from power packages；
S43 repeats the above steps, and clusters to power edge, until all events realize cluster.
Referring to Fig. 3, in the present embodiment, step S5 is specifically included:
S51 carries out threshold using that the smallest event cluster of the absolute value of the mean value of power cluster group as new adaptive threshold
Value modification；
S52 improves the cluster degree of cluster group, optimization cluster.
Referring to Fig. 4, in the present embodiment, step S51 is specifically included:
S511, using that the smallest event cluster of the absolute value of the mean value of positive cluster group as new adaptive positive threshold value；
S512, using that the smallest event cluster of the absolute value of the mean value of negative power cluster group as new adaptive negative threshold value；
S513 respectively goes the lesser changed power event of changed power in positive cluster group and negative power cluster group
It removes.
Referring to Fig. 5, in the present embodiment, step S52 is specifically included:
The cluster degree of positive cluster group and the event cluster in negative power cluster group is respectively increased in S521；
S522, respectively to the similar all edges point in power edge in the positive cluster group and the negative power cluster group
Group advanced optimizes cluster into same event cluster.
In the present embodiment, after executing step S522, further includes: it is more that the cluster group of less element is merged into adjacent element
Phylogenetic group to keep just clustering the balance between the quantity of negative cluster.
Referring to Fig. 6, in the present embodiment, step S6 is specifically included:
S61 from big to small rearranges positive event cluster and negative power event cluster by mean amplitude, so that positive
Event cluster and negative power event cluster are corresponding, to be more convenient for matching；
S62 forms time of origin cluster corresponding with the power event of positive event cluster and negative event cluster, to carry out feature
Match, filters out candidate matched group；
S63, in conjunction with the matched group that figure signal is optimal to the matched group further screening filtered out；
S64 repeats the above steps, and filters out optimal matched group to each power event cluster, until to all function
Rate event cluster completes screening.
In the present embodiment, after executing step S64, further includes: deleted from event cluster to the power edge for failing pairing
It removes.
Referring to Fig. 7, in the present embodiment, step S7 is specifically included:
S71 calculates the typical change power of each equipment by the power data being collected into, and establishes device label library and will be each
The typical change power data of equipment is stored in the device label library as label；
S72, by the typical case of equipment in the mean value and tag library of the changed power absolute value of each cluster group after characteristic matching
Changed power matches, if matching meets, which belongs to corresponding equipment, if without corresponding label,
New label is added in tag library；
S73 according to the result of pairing by the time difference of each power-up ramp and failing edge multiplied by changed power, and adds up
All changed power events, obtain the total electricity using the equipment in phylogenetic group.
The specific implementation process of this method is as follows:
Referring to Fig. 8, in embodiment, we obtain three days electricity consumption datas of a certain household electricity, in order to minimize
The power of open/close state conversion of influence of the household appliances fluctuation of load to load decomposition without losing all electrical equipments becomes
Change, we are first by using T0=10W, which changes total load active power, carries out threshold process, threshold process formula:
ΔPi∈(-∞,T0)∪(T0,+∞)
Wherein, T0Initial, the small threshold value for one predetermined, as unit of watt；|ΔPi|>T0Reference event.
Event is extracted to the energy data for carrying out threshold process, generates split time, forms rising edge and failing edge.
Figure signal is generated based on Euclidean distance algorithm, by the node of figure signal to all rising edges of event and decline
Edge is indexed, Euclidean distance formula:
Wherein: A is matrix, Ai,jIndicate its entry in ith row and jth column；xi,xjIndicate its i-th row, jth column
Element, ρ are zoom factors.
Δ P in rising edge and failing edgeiCorrespond respectively to a node vi of figure signal, and using Euclid away from
From formula, x is utilizedi=Δ PiAnd xj=Δ PjIt calculates its Euclidean distance and carrys out distribution node viAnd vjBetween side.
The power event screened by initial threshold is classified, for this purpose, we are first divided into power event group
Positive group and negative power group cluster it respectively, and in first time cluster, we will be similar with the first power edge
All edges are grouped into same event cluster.For this purpose, finding out matrix first with Euclidean distance, figure signal s is then formed,
Wherein each element of s corresponds to a node in figure, right to ensure that figure signal keeps smooth by diagonal matrix algorithm
Angle matrix algorithm:
Di,i=∑jAi,j
Wherein D is the diagonal matrix with nonzero term.
Finally, threshold process is carried out to A by using high threshold q=0.98, if smoothness is greater than threshold value, then it is assumed that should
Edge is similar to first edge, and assigns to first event cluster, and the edge is deleted from power event group.We are to non-sub-clustering
Event repeats identical cluster process, until all affair clusterings, finally, in embodiment, form eight and initially just clustering, eight
A original negative cluster.
For the event cluster of each cluster, the resultant error of cluster is larger, needs to optimize and clusters again.We, which are arranged, comments
The superset clustering class estimated event cluster aggregate quality value and set 20 for original zoom factor ρ to avoid most of household electrical appliance,
And reduce the cluster power edge of mistake.
In embodiment, still there are many lesser changed power events for initial clustering result, therefore pass through modification threshold
With filter events again.For this purpose, then by that the smallest event cluster of the absolute value of positive event cluster group and the mean value of negative event cluster group
It is selected out, and using the event cluster mean value as new adaptive positive threshold and negative threshold, after filtering some changed powers
Lesser changed power event.
In embodiment, the event cluster in the event cluster in positive event cluster and negative event cluster needs to reduce zoom factor ρ use
In improving cluster degree, we repeat the process for being similar to the initial clustering stage, the two event clusters are formed to thinner grouping.It is excellent
After changing cluster, positive event cluster is divided into 14 groups, and negative event cluster then has 15 groups.Note that some clusters only one in positive and negative event cluster
A or two changed power events, so the grouping with less element will be combined into adjacent element in end of clustering
Balance between quantity of more phylogenetic groups to keep positive and negative cluster.In this way after sub- Cluster merging, we are obtaining 7 just
Phylogenetic group and 7 negative phylogenetic groups.
Then, we match each positive event cluster with the immediate negative event cluster of amplitude；We are first positive event
Cluster and negative power event cluster are rearranged from big to small by mean amplitude, one-to-one correspondence when in order to match.Then formed with just
The corresponding time of origin cluster of the power event of event cluster and negative event cluster, screening candidate pairing when this is characterized matching.
For each positive and negative cluster pair, each rising edge and best failing edge are matched using characteristic matching, to find out pair
It should be in the failing edge of each rising edge.Rising edge specific for one, we first filter out the failing edge event of candidate pairing
Group (from physical significance, failing edge be must be present in after rising edge, i.e., electric appliance, which only opens, to close), then base
Figure signal is formed in the distance of their power edge amplitudes and temporal distance.We construct two figure signals: (1) one
The distance of difference of the figure for indicating the absolute power level of rising edge and the absolute value power of candidate failing edge；(2) second charts
Show the time difference between rising edge and candidate failing edge.We by the two combined factors get up consider, finishing screen is selected optimal
Pairing.
We repeat above-mentioned steps to each positive event cluster, until circulation terminates.The number sampled due to us
Be in a period of time according to collection, thus there are electric switch state not within the period that we acquire, consider further that calculation
The reason of method error, the power edge for failing pairing are then deleted from event cluster
Finally, first calculate the Δ P of each electric appliance to the appliance power come out by data set, form the tag library of electric appliance,
Then by the mean value of each group of changed power absolute value after characteristic matching with the typical changed power phase of electric appliance in tag library
It compares, if matching result meeting with electric appliance tag library, is considered electric appliance corresponding to the label for the group, if do not had
Corresponding label then adds new label in tag library.In addition, according to the result of pairing by each power-up ramp and decline
The time difference on edge is multiplied by changed power, and all changed power events in cumulative phylogenetic group, obtains user and uses the electricity in this three days
The total electricity of device.
The method of the present invention realizes the load decomposition of electricity consumption, thus to each by handling equipment electricity consumption data
Real time monitoring is realized in the energy consumption of equipment, grasps the electricity consumption situation of each equipment, is facilitated user and is understood each equipment electricity consumption situation,
To carry out required energy consumption analysis according to the electricity consumption situation and formulate reasonable energy conservation program, and then reduce unnecessary electric energy
Consumption, economizes on resources.
Particular embodiments described above has carried out further the purpose of the present invention, technical scheme and beneficial effects
It is described in detail, it should be understood that the above is only a specific embodiment of the present invention, the protection being not intended to limit the present invention
Range.It particularly points out, to those skilled in the art, all within the spirits and principles of the present invention, that is done any repairs
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of equipment state change events monitoring method characterized by comprising
Collect one or more electricity consumption datas of equipment and the power data of total load；
Threshold process is carried out to the changed power of total load, power event is extracted and forms candidate events group；
Figure signal is generated, all rising edges and failing edge of the node of figure signal to event are indexed；
The power event extracted by initial threshold is classified, cluster is formed；
Adaptive threshold modification and optimization cluster are carried out to sorted power event is carried out；
Each positive event cluster and the immediate negative event cluster of amplitude are subjected to screening feature pairing, and further screening acquisition is optimal
Matched group；
Exemplary power by calculating each equipment changes numerical value, is allowed to absolute with the changed power of each cluster group after characteristic matching
The mean value of value is matched, and corresponding equipment Electrical change numerical value is obtained.
2. equipment state change events monitoring method as described in claim 1, which is characterized in that described to pass through initial threshold
The power event extracted is classified, specific as follows:
The candidate events group is divided into positive group and negative power group, and it is clustered respectively；
Same event is grouped into the similar all edges in power edge in the positive group and the negative power group respectively
In cluster, and the edge being grouped into same event cluster is deleted from power packages；
It repeats the above steps, power edge is clustered, until all events realize cluster.
3. equipment state change events monitoring method as described in claim 1, which is characterized in that described pair of progress is sorted
Power event carries out adaptive threshold modification and optimization cluster, specific as follows:
Using that the smallest event cluster of the absolute value of the mean value of power cluster group as new adaptive threshold, threshold modifying is carried out；
Improve the cluster degree of cluster group, optimization cluster.
4. equipment state change events monitoring method as claimed in claim 3, which is characterized in that described by the equal of power cluster group
That the smallest event cluster of the absolute value of value carries out threshold modifying as new adaptive threshold, specific as follows:
Using that the smallest event cluster of the absolute value of the mean value of positive cluster group as new adaptive positive threshold value；
Using that the smallest event cluster of the absolute value of the mean value of negative power cluster group as new adaptive negative threshold value；
The lesser changed power event of changed power in positive cluster group and negative power cluster group is removed respectively.
5. equipment state change events monitoring method as claimed in claim 3, which is characterized in that the cluster for improving cluster group
Degree, optimization cluster are specific as follows:
The cluster degree of positive cluster group and the event cluster in negative power cluster group is respectively increased；
The similar all edges in power edge in the positive cluster group and the negative power cluster group are grouped into respectively same
In event cluster, cluster is advanced optimized.
6. equipment state change events monitoring method as claimed in claim 5, which is characterized in that described to advanced optimize cluster
Afterwards, further includes: the cluster group of less element is merged into the phylogenetic group more than adjacent element to keep just clustering and the quantity of negative cluster
Between balance.
7. equipment state change events monitoring method as described in claim 1, which is characterized in that described by each positive event cluster
Screening feature pairing is carried out with the immediate negative event cluster of amplitude, and further screening obtains optimal matched group, specific as follows:
Positive event cluster and negative power event cluster are rearranged from big to small by mean amplitude so that positive event cluster and
Negative power event cluster is corresponding, to be more convenient for matching；
Time of origin cluster corresponding with the power event of positive event cluster and negative event cluster is formed, to carry out characteristic matching, is screened
Candidate matched group out；
In conjunction with the matched group that figure signal is optimal to the matched group further screening filtered out；
It repeats the above steps, optimal matched group is filtered out to each power event cluster, until to all power event clusters
Complete screening.
8. equipment state change events monitoring method as claimed in claim 7, which is characterized in that described pair of all power thing
After part cluster completes screening, further includes: deleted from event cluster the power edge for failing pairing.
9. equipment state change events monitoring method as described in claim 1, which is characterized in that described by calculating each equipment
Exemplary power change numerical value, be allowed to be matched with the mean value of the changed power absolute value of each cluster group after characteristic matching,
Corresponding equipment Electrical change numerical value is obtained, specific as follows:
The typical change power that each equipment is calculated by the power data being collected into establishes device label library and by the allusion quotation of each equipment
Type variation power data is stored in the device label library as label；
By the exemplary power variation of equipment in the mean value and tag library of the changed power absolute value of each cluster group after characteristic matching
Match, if matching meets, which belongs to corresponding equipment, if without corresponding label, in tag library
The new label of middle addition；
According to the result of pairing by the time difference of each power-up ramp and failing edge multiplied by changed power, and in cumulative phylogenetic group
All changed power events, obtain the total electricity using the equipment.
10. a kind of equipment state change events monitoring method characterized by comprising
Collect one or more electricity consumption datas of equipment and the power data of total load；
Threshold process is carried out to the changed power of total load, power event is extracted and forms candidate events group；
Figure signal is generated, all rising edges and failing edge of the node of figure signal to event are indexed；
The power event extracted by initial threshold is classified, cluster is formed；
Adaptive threshold modification and optimization cluster are carried out to sorted power event is carried out；
Each positive event cluster and the immediate negative event cluster of amplitude are subjected to screening feature pairing, and further screening acquisition is optimal
Matched group；
Exemplary power by calculating each equipment changes numerical value, is allowed to absolute with the changed power of each cluster group after characteristic matching
The mean value of value is matched, and corresponding equipment Electrical change numerical value is obtained；
It is described that the power event extracted by initial threshold is classified, specific as follows:
The candidate events group is divided into positive group and negative power group, and it is clustered respectively；
Same event is grouped into the similar all edges in power edge in the positive group and the negative power group respectively
In cluster, and the edge being grouped into same event cluster is deleted from power packages；
It repeats the above steps, power edge is clustered, until all events realize cluster；
Described pair carries out sorted power event and carries out adaptive threshold modification and optimization cluster, specific as follows:
Using that the smallest event cluster of the absolute value of the mean value of power cluster group as new adaptive threshold, threshold modifying is carried out；
Improve the cluster degree of cluster group, optimization cluster；
It is described using that the smallest event cluster of the absolute value of the mean value of power cluster group as new adaptive threshold, carry out threshold value and repair
Change, specific as follows:
Using that the smallest event cluster of the absolute value of the mean value of positive cluster group as new adaptive positive threshold value；
Using that the smallest event cluster of the absolute value of the mean value of negative power cluster group as new adaptive negative threshold value；
The lesser changed power event of changed power in positive cluster group and negative power cluster group is removed respectively；
The cluster degree for improving cluster group, optimization cluster are specific as follows:
The cluster degree of positive cluster group and the event cluster in negative power cluster group is respectively increased；
The similar all edges in power edge in the positive cluster group and the negative power cluster group are grouped into respectively same
In event cluster, cluster is advanced optimized；
It is described advanced optimize cluster after, further includes: the cluster group of less element is merged into the phylogenetic group more than adjacent element to protect
Hold the balance between positive cluster and the quantity of negative cluster；
It is described that each positive event cluster and the immediate negative event cluster of amplitude are subjected to screening feature pairing, and further screening obtains
Optimal matched group, specific as follows:
Positive event cluster and negative power event cluster are rearranged from big to small by mean amplitude so that positive event cluster and
Negative power event cluster is corresponding, to be more convenient for matching；
Time of origin cluster corresponding with the power event of positive event cluster and negative event cluster is formed, to carry out characteristic matching, is screened
Candidate matched group out；
In conjunction with the matched group that figure signal is optimal to the matched group further screening filtered out；
It repeats the above steps, optimal matched group is filtered out to each power event cluster, until to all power event clusters
Complete screening；
After described pair of all power event cluster completes screening, further includes: deleted from event cluster to the power edge for failing pairing
It removes；
The exemplary power by calculating each equipment changes numerical value, is allowed to the changed power with each cluster group after characteristic matching
The mean value of absolute value is matched, and corresponding equipment Electrical change numerical value is obtained, specific as follows:
The typical change power that each equipment is calculated by the power data being collected into establishes device label library and by the allusion quotation of each equipment
Type variation power data is stored in the device label library as label；
By the exemplary power variation of equipment in the mean value and tag library of the changed power absolute value of each cluster group after characteristic matching
Match, if matching meets, which belongs to corresponding equipment, if without corresponding label, in tag library
The new label of middle addition；
According to the result of pairing by the time difference of each power-up ramp and failing edge multiplied by changed power, and in cumulative phylogenetic group
All changed power events, obtain the total electricity using the equipment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811096084.8A CN109460419A (en) | 2018-09-19 | 2018-09-19 | A kind of equipment state change events monitoring method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811096084.8A CN109460419A (en) | 2018-09-19 | 2018-09-19 | A kind of equipment state change events monitoring method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109460419A true CN109460419A (en) | 2019-03-12 |
Family
ID=65606813
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811096084.8A Pending CN109460419A (en) | 2018-09-19 | 2018-09-19 | A kind of equipment state change events monitoring method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109460419A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110702408A (en) * | 2019-09-30 | 2020-01-17 | 佛山科学技术学院 | Bearing state change event monitoring method and device |
| CN111693765A (en) * | 2020-05-29 | 2020-09-22 | 杭州电子科技大学 | Non-invasive load monitoring method based on improved spectral clustering |
| CN113449894A (en) * | 2020-03-25 | 2021-09-28 | 中关村海华信息技术前沿研究院 | Method, system, device and storage medium for analyzing on-off state of energy-using device |
| CN113792767A (en) * | 2021-08-27 | 2021-12-14 | 国网福建省电力有限公司 | Load electricity utilization characteristic monitoring and analyzing method based on graph signal processing |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103105556A (en) * | 2013-01-30 | 2013-05-15 | 西安交通大学 | Intelligent power grid load testing and recognition method based on steady state and transient state characteristic joint matching |
-
2018
- 2018-09-19 CN CN201811096084.8A patent/CN109460419A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103105556A (en) * | 2013-01-30 | 2013-05-15 | 西安交通大学 | Intelligent power grid load testing and recognition method based on steady state and transient state characteristic joint matching |
Non-Patent Citations (3)
| Title |
|---|
| 刘然: "结合改进最近邻法与支持向量机的住宅用电负荷识别研究", 《中国优秀硕士学位论文全文数据库（电子期刊）》 * |
| 李雨轩: "非侵入式负荷分解算法的综合研究", 《中国优秀硕士学位论文全文数据库（电子期刊）》 * |
| 李静: "一种用于家用智能电表的负荷识别方法", 《现代电力》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110702408A (en) * | 2019-09-30 | 2020-01-17 | 佛山科学技术学院 | Bearing state change event monitoring method and device |
| CN113449894A (en) * | 2020-03-25 | 2021-09-28 | 中关村海华信息技术前沿研究院 | Method, system, device and storage medium for analyzing on-off state of energy-using device |
| CN111693765A (en) * | 2020-05-29 | 2020-09-22 | 杭州电子科技大学 | Non-invasive load monitoring method based on improved spectral clustering |
| CN113792767A (en) * | 2021-08-27 | 2021-12-14 | 国网福建省电力有限公司 | Load electricity utilization characteristic monitoring and analyzing method based on graph signal processing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109460419A (en) | A kind of equipment state change events monitoring method | |
| CN110514889A (en) | A kind of method and system of non-intrusion type household electricity remained capacity | |
| CN109633368B (en) | VMD and DFA-based power quality disturbance detection method for power distribution network with distributed power supplies | |
| CN111428355B (en) | Modeling method for intelligent comprehensive statistics of power load numbers | |
| CN105447082A (en) | Distributed clustering method for mass load curves | |
| CN106022976A (en) | Demand side oriented power consumer classification method and system | |
| Chui et al. | Appliance signature identification solution using K-means clustering | |
| CN109946543B (en) | V-I pattern sample data set screening method for non-invasive load monitoring | |
| CN111651448B (en) | Low-voltage topology identification method based on noise reduction differential evolution | |
| CN110084719A (en) | A kind of distribution network load type device for identifying | |
| CN113627289A (en) | Deep learning-based power equipment identification method and device | |
| CN111191883B (en) | Intelligent distribution transformer terminal micro-application selection method and system | |
| CN108614154A (en) | A kind of power consumer phase recognition methods based on spectrum analysis | |
| CN115829168B (en) | New energy equipment remote coordination operation and maintenance management system based on data analysis | |
| Zhibin et al. | Research on household appliances recognition method based on data screening of deep learning | |
| Mei et al. | Random forest based adaptive non-intrusive load identification | |
| CN108090647B (en) | Method for obtaining load operation probability of power consumers and method for obtaining load operation probability of power consumer group | |
| CN112241812B (en) | Topology identification method for low-voltage distribution network based on single-side optimization and genetic algorithm cooperation | |
| CN113131487B (en) | Transformer area identification method and device based on voltage regulation pattern, storage medium and electronic equipment | |
| CN114046873A (en) | Reactor vibration monitoring system based on LOF-FCM fuzzy clustering algorithm | |
| Li et al. | A fusion framework using integrated neural network model for non-intrusive load monitoring | |
| Qaisar et al. | Adaptive rate sampling and machine learning based power quality disturbances interpretation | |
| Lin et al. | Computer aided analysis and control of power system based on data mining technology | |
| CN111948446A (en) | Self-adaptive load identification method and intelligent electric energy meter | |
| Liu et al. | Home appliances classification based on multi-feature using ELM |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190312 |
|
| RJ01 | Rejection of invention patent application after publication |