CN112909923A - Non-invasive household load behavior recognition device based on DTW algorithm - Google Patents

Non-invasive household load behavior recognition device based on DTW algorithm Download PDF

Info

Publication number
CN112909923A
CN112909923A CN202110081952.0A CN202110081952A CN112909923A CN 112909923 A CN112909923 A CN 112909923A CN 202110081952 A CN202110081952 A CN 202110081952A CN 112909923 A CN112909923 A CN 112909923A
Authority
CN
China
Prior art keywords
load
data
invasive
household
power
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
Application number
CN202110081952.0A
Other languages
Chinese (zh)
Inventor
李通
王博
王兆华
张伟
孙开宁
张斌
樊茂
徐森
张勇
王天军
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Institute of Technology BIT
State Grid Xinjiang Electric Power Co Ltd
Original Assignee
Beijing Institute of Technology BIT
State Grid Xinjiang Electric Power Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing Institute of Technology BIT, State Grid Xinjiang Electric Power Co Ltd filed Critical Beijing Institute of Technology BIT
Priority to CN202110081952.0A priority Critical patent/CN112909923A/en
Publication of CN112909923A publication Critical patent/CN112909923A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R21/00—Arrangements for measuring electric power or power factor
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R22/00—Arrangements for measuring time integral of electric power or current, e.g. electricity meters
    • G01R22/06—Arrangements for measuring time integral of electric power or current, e.g. electricity meters by electronic methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Electricity, gas or water supply
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
    • G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
    • H02J2203/10—Power transmission or distribution systems management focussing at grid-level, e.g. load flow analysis, node profile computation, meshed network optimisation, active network management or spinning reserve management
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
    • H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2310/00—The network for supplying or distributing electric power characterised by its spatial reach or by the load
    • H02J2310/10—The network having a local or delimited stationary reach
    • H02J2310/12—The local stationary network supplying a household or a building
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2310/00—The network for supplying or distributing electric power characterised by its spatial reach or by the load
    • H02J2310/10—The network having a local or delimited stationary reach
    • H02J2310/12—The local stationary network supplying a household or a building
    • H02J2310/14—The load or loads being home appliances
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2310/00—The network for supplying or distributing electric power characterised by its spatial reach or by the load
    • H02J2310/70—Load identification
    • 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
    • Y02B—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO BUILDINGS, e.g. HOUSING, HOUSE APPLIANCES OR RELATED END-USER APPLICATIONS
    • Y02B70/00—Technologies for an efficient end-user side electric power management and consumption
    • Y02B70/30—Systems integrating technologies related to power network operation and communication or information technologies for improving the carbon footprint of the management of residential or tertiary loads, i.e. smart grids as climate change mitigation technology in the buildings sector, including also the last stages of power distribution and the control, monitoring or operating management systems at local level
    • Y02B70/3225—Demand response systems, e.g. load shedding, peak shaving
    • 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
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S20/00—Management or operation of end-user stationary applications or the last stages of power distribution; Controlling, monitoring or operating thereof
    • Y04S20/20—End-user application control systems
    • Y04S20/222—Demand response systems, e.g. load shedding, peak shaving

Abstract

The invention discloses a non-invasive household load behavior recognition device based on a DTW algorithm, which comprises a household electric energy meter and a non-invasive load detection module which are uniformly installed at a user inlet, wherein the non-invasive load detection module comprises a threshold setting module, a trigger, a data acquisition module, a household electric appliance load database, a reference template database, a matching module and an analysis module which are sequentially connected; the output of electric energy meter is connected in data acquisition module's input, non-invasive load detection module's output is connected with domestic block terminal, domestic block terminal electric connection has domestic appliance. The invention can lead the user to know the household load consumption condition through non-invasive load identification, and generate the electricity bill of each household appliance load. Meanwhile, the data are uploaded to a power supply company, so that the power supply company can manage the users in a classified manner, and differentiated services of the customers are realized.

Description

Non-invasive household load behavior recognition device based on DTW algorithm
Technical Field
The invention relates to the technical field of household load behavior identification, in particular to a non-invasive household load behavior identification device based on a DTW algorithm.
Background
The development of the load monitoring technology is vigorous, and the identification of the electricity utilization behavior of the power customer also becomes an urgent problem to be solved. Most electric energy meters installed by the existing power supply companies cannot realize the classified metering of the electricity consumption of power customers, only count the total electricity quantity of users comprehensively, are not beneficial to the power supply companies to master the family load consumption condition, cannot further realize the classified management of the users, and cannot provide differentiated services.
How to select effective characteristic quantities and an identification algorithm to characterize and solve the power utilization behavior of the load becomes the greatest importance of research. Scholars at home and abroad propose a plurality of data processing methods, such as a clustering algorithm, a chicken flock algorithm, a differential evolution algorithm and the like. Establishing a household load standard database, detecting events through a variable point detection algorithm, collecting data and carrying out load identification through a clustering algorithm.
By representing any operating state of a household appliance load by 0 and 1, namely only relating and turning on, the household appliance load decomposition problem can be converted into solving an optimization combination problem, and then an improved chicken flock algorithm is applied to identify the operating state of each household appliance. And determining characteristic parameters of current and voltage steady-state data of the electric appliance by adopting a fuzzy clustering method, and then carrying out load decomposition by adopting a differential evolution algorithm. Although documents of many scholars at home and abroad realize the identification and the decomposition of the load, the problems of poor algorithm convergence, atypical feature quantity selection and the like exist.
The above algorithms mostly adopt load steady-state data as identification characteristic quantities, such as voltage, current, energy consumption change values, and the like, and it is difficult to implement capture on and off transient processes of household loads.
Therefore, the device for non-invasive household load behavior identification based on the DTW algorithm takes the transient waveform and the power change value of the household load during the on and off as the characteristic quantities, and can effectively identify various load electricity utilization data of the household electricity.
Disclosure of Invention
The invention aims to solve the defects in the prior art, and provides a non-invasive household load behavior identification device based on a DTW algorithm.
In order to achieve the purpose, the invention provides the following technical scheme:
a non-invasive household load behavior recognition device based on a DTW algorithm comprises a household electric energy meter and a non-invasive load detection module which are uniformly installed at a user inlet, wherein the non-invasive load detection module comprises a threshold setting module, a trigger, a data acquisition module, a household appliance load database, a reference template database, a matching module and an analysis module which are sequentially connected;
the output of electric energy meter is connected in data acquisition module's input, non-invasive load detection module's output is connected with domestic block terminal, domestic block terminal electric connection has domestic appliance.
The threshold setting module is used for setting a power threshold when the household appliance is used, and the trigger is used for starting the data acquisition module and acquiring data through the data acquisition module.
The data acquisition module takes the transient waveform and the power change value of the household appliances during switching on and off as characteristic quantities, acquires and identifies various load electricity consumption data of the household electricity, and identifies the consumed power of each household appliance along with time according to the total load consumption curve of the electric meter of the user house.
The load identification method of the household appliance specifically comprises the following steps:
step 1: the method comprises the steps of collecting data, uniformly installing a household electric energy meter and a non-invasive load detection module at a user inlet, monitoring active power, and when the data reach a set threshold value, actuating a trigger of the non-invasive load detection module to collect the data;
step 2: establishing a household appliance load database, extracting power waveforms of various types of household appliances to form a reference template database for subsequent matching;
and step 3: matching the reference template with the template to be identified, and calculating accumulated distortion amount to further solve the similarity between the reference template and the template to be identified;
and 4, step 4: comparing the similarity results of the groups, the test template belongs to the class with the similarity closest to 1.
The load characteristic data sample set of the home appliance may be represented as:
Figure BDA0002909692750000031
wherein Pm (n) is a power value corresponding to the nth sampling point of the electric appliance m; m is the type of the electric appliance; n is the number of sampling points;
according to the power data of each household appliance, k1, k2, …, Km is expressed as the opening time coefficient of the load, and then the calculation formula of the total power of the load is as follows:
Figure BDA0002909692750000032
let the actual sampled total load power data be expressed as:
y(n)＝[y(1)，y(2)，…，y(k)，…y(n)] (3)
where y (k) represents the total power monitored at the kth sample point: n represents the number of samples; d (x, y) is used for representing the correlation degree between the collected total power data y (n) and the fitted total power data x (n); defining d (x, y) as the amount of distortion between the two;
to reduce the programming effort, the similarity is characterized here using the chi-square test method:
Figure BDA0002909692750000033
x (n) and y (n) are two eigenvectors without negative numerical values, and the closer the distortion quantity is to zero, the higher the similarity is;
compared with the common Euclidean distance, the Caller method avoids root number operation, avoids calculation redundancy, and then performs normalization processing on the data to enable the data to be positioned between [0 and 1 ];
taking:
Figure BDA0002909692750000041
wherein a ═ max x (n); b ═ max y (n), then define the similarity:
Figure BDA0002909692750000042
for measuring the closeness between two data, when s (x, y) is larger, the two vectors are closer, and when s (x, y) is completely matched, s (x, y) is 1, so the solution target of the algorithm can be expressed as: an appropriate fitting signal x (n) is found such that s (x, y) is close to 1, i.e. a set of optimal k1, k2, …, km values is found such that the stronger the correlation between the fitting data and the total load data.
Compared with the prior art, the invention has the beneficial effects that: the transient waveform and the power change value of the household load during the on and off are used as characteristic quantities, various load electricity consumption data of the household electricity can be effectively identified, the user can know the household load consumption condition through non-invasive load identification, and the electricity consumption bill of each household appliance load is generated. Meanwhile, the data are uploaded to a power supply company, so that the power supply company can manage the users in a classified manner, and differentiated services of the customers are realized.
Drawings
FIG. 1 is a schematic diagram of the system of the present invention;
fig. 2 is a schematic flow chart of the load identification method of the household appliance according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1-2, the present invention provides a technical solution: a non-invasive household load behavior recognition device based on a DTW algorithm comprises a household electric energy meter and a non-invasive load detection module which are uniformly installed at a user inlet, wherein the non-invasive load detection module comprises a threshold setting module, a trigger, a data acquisition module, a household appliance load database, a reference template database, a matching module and an analysis module which are sequentially connected;
the output of electric energy meter is connected in data acquisition module's input, non-invasive load detection module's output is connected with domestic block terminal, domestic block terminal electric connection has domestic appliance.
The threshold setting module is used for setting a power threshold when the household appliance is used, and the trigger is used for starting the data acquisition module and acquiring data through the data acquisition module.
The data acquisition module takes the transient waveform and the power change value of the household appliances during switching on and off as characteristic quantities, acquires and identifies various load electricity consumption data of the household electricity, and identifies the consumed power of each household appliance along with time according to the total load consumption curve of the electric meter of the user house.
How to select effective characteristic quantities and an identification algorithm to characterize and solve the power utilization behavior of the load becomes the greatest importance of research. Scholars at home and abroad propose a plurality of data processing methods, such as a clustering algorithm, a chicken flock algorithm, a differential evolution algorithm and the like. Establishing a household load standard database, detecting events through a variable point detection algorithm, collecting data and carrying out load identification through a clustering algorithm.
By representing any operating state of a household appliance load by 0 and 1, namely only relating and turning on, the household appliance load decomposition problem can be converted into solving an optimization combination problem, and then an improved chicken flock algorithm is applied to identify the operating state of each household appliance. And determining characteristic parameters of current and voltage steady-state data of the electric appliance by adopting a fuzzy clustering method, and then carrying out load decomposition by adopting a differential evolution algorithm. Although documents of many scholars at home and abroad realize the identification and the decomposition of the load, the problems of poor algorithm convergence, atypical feature quantity selection and the like exist.
The above algorithms mostly adopt load steady-state data as identification characteristic quantities, such as voltage, current, energy consumption change values, and the like, and it is difficult to implement capture on and off transient processes of household loads.
Therefore, a non-invasive behavior identification method based on a Dynamic Time Warping (DTW) algorithm is provided, and each item of load electricity utilization data of the household electricity can be effectively identified by taking a transient waveform and a power change value of the household load during the opening and closing as characteristic quantities.
A behavior identification method is provided aiming at non-invasive household loads, and the consumed power of each household appliance along with time can be identified according to the total load consumption curve of a user house electric meter. Different devices have different actual power consumptions, the rated powers of all the electric appliances in one house are known, and the powers of all the electric appliances are combined together to form a database of the household.
And the change of the total load curve is calculated and compared with a database of the household, and the most possible load switch state is matched, so that the electricity utilization behavior of the household is identified. The behavior recognition generally comprises two steps of feature quantity selection and behavior recognition algorithm selection. First, selecting a feature amount is a key to recognition.
The load characteristic data sample set of the home appliance may be represented as:
Figure BDA0002909692750000061
wherein Pm (n) is a power value corresponding to the nth sampling point of the electric appliance m; m is the type of the electric appliance; n is the number of sampling points;
according to the power data of each household appliance, k1, k2, …, Km is expressed as the opening time coefficient of the load, and then the calculation formula of the total power of the load is as follows:
Figure BDA0002909692750000071
let the actual sampled total load power data be expressed as:
y(n)＝[y(1)，y(2)，…，y(k)，…y(n)] (3)
where y (k) represents the total power monitored at the kth sample point: n represents the number of samples; d (x, y) is used for representing the correlation degree between the collected total power data y (n) and the fitted total power data x (n); defining d (x, y) as the amount of distortion between the two;
to reduce the programming effort, the similarity is characterized here using the chi-square test method:
Figure BDA0002909692750000072
x (n) and y (n) are two eigenvectors without negative values, the closer the distortion amount is to zero, the higher the similarity,
compared with the common Euclidean distance, the method of the card method avoids root number operation, avoids calculation redundancy, and then performs normalization processing on the data to enable the data to be positioned between [0 and 1], so that the vector difference in different value ranges can be conveniently compared;
taking:
Figure BDA0002909692750000073
wherein a ═ max x (n); b ═ max y (n), then define the similarity:
Figure BDA0002909692750000081
for measuring the closeness between two data, when s (x, y) is larger, the two vectors are closer, and when s (x, y) is completely matched, s (x, y) is 1, so the solution target of the algorithm can be expressed as: an appropriate fitting signal x (n) is found such that s (x, y) is close to 1, i.e. a set of optimal k1, k2, …, km values is found such that the stronger the correlation between the fitting data and the total load data.
The algorithm flow is as follows:
and adopting a DTW matching algorithm to identify the household electricity consumption behavior. The DTW algorithm, i.e. the dynamic time warping algorithm, combines distance measure calculation with the time warping method. The method is an intelligent algorithm commonly used in the fields of voice recognition and behavior recognition. The method is extended to be used in the problem of power load identification, and the matching between templates with different lengths can be easily realized;
in general, methods that can be used to calculate the similarity difference between feature vectors can be used in the DTW algorithm, such as euclidean distance, mahalanobis distance, and the like. The present document uses a chi-square test to avoid root operations.
Establishing a reference template R (1), R (2), R (M) by a household appliance load information base to be identifiedAnother template, T ═ { T (1), T (2),.., T (n) }, where r (m) and T (n) are their internal feature vectors and are of the same dimension. Definition d [ T (n)i)，R(mi)]For calculating a feature vector T (n)i) And R (m)i) The resulting amount of distortion, DT (n)i)，R(mi) [ to calculate the accumulated amount of distortion for a path connecting pairs of feature vector nodes.
Figure BDA0002909692750000082
The goal of the DTW algorithm is to find an optimal path such that DT (N), R (M [ ]) calculated cumulatively over that path is the smallest of all paths. There is an iterative process as follows:
D[T(ni)，R(mi)]＝d[T(ni)，R(mi)]+D[T(ni-1)，R(mi-1)] (7)
wherein:
D[T(ni-1)，R(mi-1)]＝min(D[T(ni-1)，R(mi)]，
D[T(ni-1)，R(mi-1)]，D[T(ni＝1)，R(mi-2)]) (8)
and calculating the minimum accumulated distortion quantity of the reference template and the template to be identified by the iteration process. And matching the template to be recognized with the reference template one by one, wherein the template to be recognized belongs to the class with the similarity closest to 1 with the reference template. The following steps are summarized in the following regarding a method for identifying the load of a non-intrusive household appliance, and a flow chart is shown in fig. 2.
Step 1: the method comprises the steps of collecting data, uniformly installing a household electric energy meter and a non-invasive load detection module at a user inlet, monitoring active power, and when the data reach a set threshold value, actuating a trigger of the non-invasive load detection module to collect the data;
step 2: establishing a household appliance load database, extracting power waveforms of various types of household appliances to form a reference template database for subsequent matching;
and step 3: matching the reference template with the template to be identified, and calculating accumulated distortion amount to further solve the similarity between the reference template and the template to be identified;
and 4, step 4: comparing the similarity results of the groups, the test template belongs to the class with the similarity closest to 1.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.

Claims (5)

1. The utility model provides a device of non-invasive house load action discernment based on DTW algorithm, includes the domestic ammeter and the non-invasive load detection module of unified installation of user's inlet port department, its characterized in that: the non-invasive load detection module comprises a threshold setting module, a trigger, a data acquisition module, a household appliance load database, a reference template database, a matching module and an analysis module which are connected in sequence;
the output of electric energy meter is connected in data acquisition module's input, non-invasive load detection module's output is connected with domestic block terminal, domestic block terminal electric connection has domestic appliance.
2. The device for non-invasive home load behavior recognition based on the DTW algorithm according to claim 1, wherein: the threshold setting module is used for setting a power threshold when the household appliance is used, and the trigger is used for starting the data acquisition module and acquiring data through the data acquisition module.
3. The device for non-invasive home load behavior recognition based on the DTW algorithm according to claim 1, wherein: the data acquisition module takes the transient waveform and the power change value of the household appliances during switching on and off as characteristic quantities, acquires and identifies various load electricity consumption data of the household electricity, and identifies the consumed power of each household appliance along with time according to the total load consumption curve of the electric meter of the user house.
4. The device for non-invasive home load behavior recognition based on the DTW algorithm according to claim 1, wherein: the load identification method of the household appliance specifically comprises the following steps:
step 1: the method comprises the steps of collecting data, uniformly installing a household electric energy meter and a non-invasive load detection module at a user inlet, monitoring active power, and when the data reach a set threshold value, actuating a trigger of the non-invasive load detection module to collect the data;
step 2: establishing a household appliance load database, extracting power waveforms of various types of household appliances to form a reference template database for subsequent matching;
and step 3: matching the reference template with the template to be identified, and calculating accumulated distortion amount to further solve the similarity between the reference template and the template to be identified;
and 4, step 4: comparing the similarity results of the groups, the test template belongs to the class with the similarity closest to 1.
5. The device for non-invasive home load behavior recognition based on the DTW algorithm according to claim 1, wherein: the load characteristic data sample set of the home appliance may be represented as:
Figure FDA0002909692740000021
wherein Pm (n) is a power value corresponding to the nth sampling point of the electric appliance m; m is the type of the electric appliance; n is the number of sampling points;
according to the power data of each household appliance, k1, k2, …, Km is expressed as the opening time coefficient of the load, and then the calculation formula of the total power of the load is as follows:
Figure FDA0002909692740000022
let the actual sampled total load power data be expressed as:
y(n)＝[y(1)，y(2)，…，y(k)，…y(n)] (3)
where y (k) represents the total power monitored at the kth sample point: n represents the number of samples; d (x, y) is used for representing the correlation degree between the collected total power data y (n) and the fitted total power data x (n); defining d (x, y) as the amount of distortion between the two;
to reduce the programming effort, the similarity is characterized here using the chi-square test method:
Figure FDA0002909692740000023
x (n) and y (n) are two eigenvectors without negative numerical values, and the closer the distortion quantity is to zero, the higher the similarity is;
compared with the common Euclidean distance, the Caller method avoids root number operation, avoids calculation redundancy, and then performs normalization processing on the data to enable the data to be positioned between [0 and 1 ];
taking:
Figure FDA0002909692740000031
wherein a ═ max x (n); b ═ max y (n), then define the similarity:
Figure FDA0002909692740000032
for measuring the closeness between two data, when s (x, y) is larger, the two vectors are closer, and when s (x, y) is completely matched, s (x, y) is 1, so the solution target can be expressed as: an appropriate fitting signal x (n) is found such that s (x, y) is close to 1, i.e. a set of optimal k1, k2, …, km values is found such that the stronger the correlation between the fitting data and the total load data.
CN202110081952.0A 2021-01-21 2021-01-21 Non-invasive household load behavior recognition device based on DTW algorithm Pending CN112909923A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110081952.0A CN112909923A (en) 2021-01-21 2021-01-21 Non-invasive household load behavior recognition device based on DTW algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110081952.0A CN112909923A (en) 2021-01-21 2021-01-21 Non-invasive household load behavior recognition device based on DTW algorithm

Publications (1)

Publication Number Publication Date
CN112909923A true CN112909923A (en) 2021-06-04

Family

ID=76118011

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110081952.0A Pending CN112909923A (en) 2021-01-21 2021-01-21 Non-invasive household load behavior recognition device based on DTW algorithm

Country Status (1)

Country Link
CN (1) CN112909923A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113567794A (en) * 2021-09-24 2021-10-29 国网江苏省电力有限公司营销服务中心 Electric bicycle indoor charging identification method and system based on dynamic time warping
CN114301061A (en) * 2021-12-23 2022-04-08 国网天津市电力公司营销服务中心 User intelligent load identification module analysis method
CN114598722A (en) * 2022-03-15 2022-06-07 北京汇智博艺科技有限公司 Non-invasive load identification and energy consumption monitoring system of Internet of things and implementation method thereof

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113567794A (en) * 2021-09-24 2021-10-29 国网江苏省电力有限公司营销服务中心 Electric bicycle indoor charging identification method and system based on dynamic time warping
CN114301061A (en) * 2021-12-23 2022-04-08 国网天津市电力公司营销服务中心 User intelligent load identification module analysis method
CN114301061B (en) * 2021-12-23 2023-10-27 国网天津市电力公司营销服务中心 Analysis method based on intelligent load identification module of user
CN114598722A (en) * 2022-03-15 2022-06-07 北京汇智博艺科技有限公司 Non-invasive load identification and energy consumption monitoring system of Internet of things and implementation method thereof

Similar Documents

Publication Publication Date Title
CN112909923A (en) Non-invasive household load behavior recognition device based on DTW algorithm
Wang et al. Non-intrusive load monitoring algorithm based on features of V–I trajectory
CN109490679B (en) Intelligent electricity stealing inspection system and method based on non-invasive load monitoring
CN106936129B (en) Power load identification method and system based on multi-feature fusion
Lin et al. Development of an improved time–frequency analysis-based nonintrusive load monitor for load demand identification
Liu et al. Admittance-based load signature construction for non-intrusive appliance load monitoring
CN111830347B (en) Two-stage non-invasive load monitoring method based on event
CN108021736A (en) A kind of load switching action monitoring method based on sliding window Remanent Model
CN106093565B (en) A kind of electricity subentry measurement method and device based on steady state characteristic Waveform Matching
WO2019128843A1 (en) Non-invasive ih electric cooker operating identification method based on hybrid criterion
CN106124850A (en) A kind of load identification system for domestic intelligent electricity meter based on similarity algorithm and load recognition methods thereof
Andrean et al. A hybrid method of cascade-filtering and committee decision mechanism for non-intrusive load monitoring
CN111027408A (en) Load identification method based on support vector machine and V-I curve characteristics
CN113361831B (en) Non-invasive load identification electric quantity decomposition method and system based on priority distribution
CN111242391A (en) Machine learning model training method and system for power load identification
CN111563827B (en) Load decomposition method based on physical characteristics of electrical appliance and electricity utilization behaviors of residents
CN114970633B (en) LSTM-based non-invasive electrical appliance identification method, system and equipment
Chen et al. Non-intrusive load monitoring based on feature extraction of change-point and xgboost classifier
Wu et al. Nonintrusive on-site load-monitoring method with self-adaption
CN113887912A (en) Non-invasive load identification method for deeply learning downward embedded equipment
Wu et al. A fast online load identification algorithm based on VI characteristics of high-frequency data under user operational constraints
CN112862163A (en) Short-term power load prediction system based on DTW algorithm
CN113206547A (en) Transformer substation power distribution system based on non-invasive power load monitoring
Cannas et al. NILM techniques applied to a real-time monitoring system of the electricity consumption
CN113363974B (en) Method and device for analyzing residential load composition based on accumulated electric quantity low-frequency sampling

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication