CN110416995B - Non-invasive load decomposition method and device - Google Patents
Non-invasive load decomposition method and device Download PDFInfo
- Publication number
- CN110416995B CN110416995B CN201910390156.8A CN201910390156A CN110416995B CN 110416995 B CN110416995 B CN 110416995B CN 201910390156 A CN201910390156 A CN 201910390156A CN 110416995 B CN110416995 B CN 110416995B
- Authority
- CN
- China
- Prior art keywords
- power
- value
- current
- electric equipment
- formula
- 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.)
- Active
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
Abstract
The invention discloses a non-invasive load decomposition method and a device, wherein the method comprises the following steps: acquiring operation records of electric equipment and power data on a bus, wherein the power data comprises: a current waveform; preprocessing the power data according to the operation record of the electric equipment to screen out first power data; determining four power parameter values corresponding to each moment according to the screened first power data; establishing a multi-target non-invasive load decomposition model with five optimization targets according to the power parameter values and the current waveforms; a type of powered device and a corresponding mode being operated on the bus are determined based on the multi-objective non-intrusive load decomposition model. The non-invasive load decomposition method provided by the invention has the advantages of high decomposition accuracy, stable result and stronger robustness.
Description
Technical Field
The invention relates to the technical field of pattern recognition, in particular to a non-invasive load decomposition method and device.
Background
In the last twenty years, the Chinese economy is developed and the living standard of people is greatly improved. The annual residential electricity consumption increase rate is about 8%. And an increase in power demand exacerbates the negative environmental impact. This contradicts the goal of the chinese government's efforts to reduce greenhouse gas emissions. The monitored household appliance energy consumption information can help decision makers and consumers to know the constitution, mode and characteristics of the residential energy requirements. The system has important effects on energy conservation and emission reduction, so that the design of a set of system capable of effectively monitoring the load is particularly important.
Existing load monitoring methods are largely divided into two types, invasive and non-invasive. Invasive load monitoring methods (ILMs) require that each powered device be equipped with an instrument with communication capabilities, which would increase deployment and maintenance of the measurement instrument costs; the non-invasive load monitoring method (NILM) requires that only one measuring instrument is installed at the user inlet of the power grid, and the collected total electricity consumption is analyzed through an algorithm, so that the electricity consumption condition of each electric equipment below the measuring instrument is monitored. For large scale deployments, non-invasive load monitoring systems can significantly reduce installation complexity and reduce maintenance costs.
Optimization and pattern recognition are 2 dominant approaches to solving the problem of load decomposition and recognition. The load recognition algorithm based on pattern recognition is to determine the structure and parameters of the recognition algorithm by learning the characteristics of each electric equipment in the database, and finally realize the recognition of the load. The non-invasive load monitoring system based on pattern recognition is firstly trained, but the trained classifier can only identify the existing electric equipment and the corresponding combination of the data set. Thus, the model needs to be trained each time a new device is added, which consumes a significant amount of time and computing resources. And when the electric equipment is more, the identification accuracy is lower. The load identification problem is converted into an optimization problem based on an optimized load identification algorithm. The load identification is achieved by minimizing the difference between the feature vector of the unknown electric equipment and the feature vector of the known electric equipment. The load characteristics of a plurality of electric equipment can be calculated by only acquiring the load characteristics of a single electric equipment based on the optimized load identification method, and the premise is that the load characteristics must meet the characteristic superposition standard.
Recent optimization-based studies use 1 to 3 objective functions, i.e. only 1 to 3 features. However, most existing studies weight multiple objective functions into one. This would lead to two problems: (1) the weighting parameters are sensitive to the dataset; (2) The weighting parameters corresponding to the different objective functions are also difficult to adjust. As errors in the load characteristics can seriously affect the accuracy of the optimization-based approach. If only one feature is used as an objective function, the obtained optimal solution is not the actual electric appliance running state; and if the consumers have similar or overlapping load characteristics, it is difficult to distinguish them using a single load characteristic, resulting in low accuracy of load resolution.
Therefore, how to provide a load decomposition method with high decomposition accuracy is a problem that needs to be solved by those skilled in the art.
Disclosure of Invention
In view of the above, the invention provides a non-invasive load decomposition method and device, which have high decomposition accuracy, stable results and stronger robustness.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
a non-invasive load splitting method comprising the steps of:
acquiring operation records of electric equipment and power data on a bus, wherein the power data comprises: a current waveform;
Preprocessing the power data according to the operation record of the electric equipment to screen out first power data;
determining four power parameter values corresponding to each moment according to the screened first power data;
establishing a multi-target non-invasive load decomposition model with five optimization targets according to the power parameter values and the current waveforms;
and determining the type of the electric equipment which is working on the bus and the corresponding working mode based on the multi-target non-invasive load decomposition model.
Preferably, preprocessing the power data according to the operation record of the electric equipment to screen out first power data specifically including;
dividing the power data according to the operation events of each electric equipment according to the operation records of the electric equipment to obtain a power data dividing result, so that the power data in the same interval are in the same mode of the same equipment;
carrying out abnormal value detection and elimination on the current and the voltage in the same interval to obtain an abnormal value elimination result;
and calculating average values of current and voltage in the interval without electric equipment according to the electric data dividing result and the abnormal value removing result, and subtracting the average values of the current and the voltage in the interval without electric equipment from the current values and the voltage values corresponding to the interval with electric equipment under the same circuit to obtain first electric data.
Preferably, the specific method for detecting and rejecting abnormal values of current and voltage in the same interval comprises the following steps:
mad=media (|a) according to the formula i Median (a) |), determining the median absolute deviation, wherein a represents the current or voltage value at the same interval, a i Representing a corresponding current value or voltage value at the ith moment, wherein MAD represents a median absolute deviation;
judging whether the current value or the voltage value in the same interval at any time deviates from the median value by more than N times of absolute deviation of the median value, if so, judging that the current value or the voltage value is an abnormal value at the moment and eliminating the abnormal value.
Preferably, determining four power parameter values corresponding to each moment according to the screened first power data specifically includes:
according to the formula: s=vi, determining the apparent power corresponding to each moment, where V represents a voltage value at a certain moment, I represents a current value at a certain moment, and S represents the apparent power;
according to the formula: p=vicos (Φ), determining the active power corresponding to each moment, wherein cos (Φ) represents a power factor and P represents the active power;
according to the formula: q=visin (Φ), determining reactive power corresponding to each moment, wherein Q represents reactive power;
according to the formula: 
The corresponding harmonics at each time instant are determined, wherein a (T) represents the current value at time instant T, provided that t=0, 1,2, …, T-1, T represents the number of sampling points in one current waveform period, and X (k) is the coefficient of the kth harmonic.
Preferably, a multi-objective non-invasive load decomposition model with five optimization objectives is established according to the power parameter values and the current waveforms, and specifically includes:
constructing five optimization targets according to the power parameter values and the current waveforms; wherein the power parameter values include active power, reactive power, apparent power and harmonics; the specific method comprises the following steps:
according to the formula:
constructing a first objective function based on the current waveform, wherein I ij (t) represents the current value when device i is operating independently in j mode at time t; t represents the number of sampling points in one current waveform period, t=0 represents that the voltage phase is at a specific time, the voltage waveform at this time is changing from the maximum value to the minimum value, N represents the total number of devices, M i Representing the total number of modes that device I contains, I representing the combined current waveform for an unknown device type; x is x ij Indicating that consumer i is in operating mode j, where i=1, 2, …, N, j=1, 2, …, M i ；
According to the formula:
constructing a second objective function based on reactive power, wherein Q ij The reactive power of the equipment i in independent operation in j mode is represented, and Q represents the reactive power of the unknown equipment type;
according to the formula:
constructing a third objective function based on the active power, wherein P ij Representing the active power of the device i when operating independently in j mode, P representing the active power of the unknown device type;
according to the formula:
constructing a fourth objective function based on apparent power, wherein S ij View representing independent operation of device i in j modeAt power, S represents the apparent power of the unknown device type;
according to the formula:
constructing a fifth objective function based on harmonics, wherein H ij (k) Representing the harmonics of the device i when operating independently in j mode, K being the order of the largest harmonics, H (K) representing the harmonics unknown to contain the device type;
according to the formula: minimum F (x) = (F) 1 (x)，f 2 (x)，f 3 (x)，f 4 (x)，f 5 (x) The five objective functions described above are constructed as a multi-objective non-invasive load decomposition model F (x), wherein,
Correspondingly, the type of the electric equipment working on the bus and the corresponding working mode are determined based on the multi-target non-invasive load decomposition model, and the method specifically comprises the following steps:
Solving the multi-target non-invasive load decomposition model F (x) by adopting a multi-target evolutionary algorithm to obtain the corresponding electric equipment type and the corresponding working mode when five optimization targets in the F (x) are simultaneously minimum;
in the multi-objective evolutionary algorithm, the constraint conditions are expressed by adopting an encoding and decoding mode: each bit of the code represents a type of powered device, where 0 represents the powered device being off and k represents the powered device being in the kth mode of operation.
A non-invasive load splitting apparatus comprising:
the system comprises an acquisition module, a control module and a control module, wherein the acquisition module is used for acquiring operation records of electric equipment and electric data on a bus, and the electric data comprises: a current waveform;
the preprocessing module is used for preprocessing the electric power data according to the operation record of the electric equipment and screening out first electric power data;
the parameter calculation module is used for determining four power parameter values corresponding to each moment according to the screened first power data;
the model building module is used for building a multi-target non-invasive load decomposition model with five optimization targets according to the power parameter values and the current waveforms;
and the determining module is used for determining the type of the electric equipment which is working on the bus and the corresponding mode based on the multi-target non-invasive load decomposition model.
Preferably, the preprocessing module specifically includes:
the dividing unit is used for dividing the power data according to the operation records of the electric equipment and each electric equipment operation event to obtain a power data dividing result, so that the power data in the same interval are in the same mode of the same equipment;
the abnormal value removing unit is used for detecting and removing abnormal values of the current and the voltage in the same interval to obtain abnormal value removing results;
and the calculation unit is used for calculating the average value of the current and the voltage in the interval without the electric equipment according to the electric data dividing result and the abnormal value removing result, and subtracting the average value of the current and the voltage in the interval without the electric equipment from the current value and the voltage value corresponding to the interval with the electric equipment under the same circuit to obtain the first electric data.
Preferably, the outlier rejection unit specifically includes:
a calculation subunit for calculating mad=media (|a) according to the formula i Median (a) |), determining the median absolute deviation, wherein a represents the current or voltage value at the same interval, a i Representing a corresponding current value or voltage value at the ith moment, wherein MAD represents a median absolute deviation;
And the eliminating subunit is used for judging whether the current value or the voltage value in the same interval at any moment deviates from the median value by more than N times of absolute deviation of the median value, and if so, the current value or the voltage value is an abnormal value at the moment and the abnormal value is eliminated.
Preferably, the parameter calculation module specifically includes:
an apparent power calculation unit for calculating apparent power according to the formula: s=vi, determining the apparent power corresponding to each moment, where V represents a voltage value at a certain moment, I represents a current value at a certain moment, and S represents the apparent power;
the active power calculation unit is used for calculating the active power according to the formula: p=vicos (Φ), determining the active power corresponding to each moment, wherein cos (Φ) represents a power factor and P represents the active power;
the reactive power calculation unit is used for calculating the reactive power according to the formula: q=visin (Φ), determining reactive power corresponding to each moment, wherein Q represents reactive power;
a harmonic calculation unit for calculating a harmonic according to the formula:
the corresponding harmonics at each time instant are determined, wherein a (T) represents the current value at time instant T, provided that t=0, 1,2, …, T-1, T represents the number of sampling points in one current waveform period, and X (k) is the coefficient of the kth harmonic.
Preferably, the model building module specifically includes:
A first objective function construction unit configured to, according to the formula:
constructing a first objective function based on the current waveform, wherein I ij (t) represents the current value when device i is operating independently in j mode at time t; t represents the number of sampling points in one current waveform period, t=0 represents that the voltage phase is at a specific time, the voltage waveform at this time is changing from the maximum value to the minimum value, N represents the total number of devices, M i Representing the total number of modes that device I contains, I representing the combined current waveform for an unknown device type; x is x ij Indicating that consumer i is in operating mode j, whereini＝1，2，…，N，j＝1，2，…，M i ；
A second objective function construction unit configured to, according to the formula:
constructing a second objective function based on reactive power, wherein Q ij The reactive power of the equipment i in independent operation in j mode is represented, and Q represents the reactive power of the unknown equipment type;
a third objective function construction unit configured to, according to the formula:
constructing a third objective function based on the active power, wherein P ij Representing the active power of the device i when operating independently in j mode, P representing the active power of the unknown device type;
a fourth objective function construction unit configured to, according to the formula:
constructing a fourth objective function based on apparent power, wherein S ij Representing the apparent power of the device i when operating independently in j mode, S representing the apparent power of the unknown device type;
a fifth objective function construction unit configured to, according to the formula:
constructing a fifth objective function based on harmonics, wherein H ij (k) Representing the harmonics of the device i when operating independently in j mode, K being the order of the largest harmonics, H (K) representing the harmonics unknown to contain the device type;
a load decomposition model building unit for building a load decomposition model according to the formula: minimum F (x) = (F) 1 (x)，f 2 (x)，f 3 (x)，f 4 (x)，f 5 (x) The five objective functions described above are constructed as a multi-objective non-invasive load decomposition model F (x), wherein,
representing the type of consumer required, the constraint x needs to be satisfied ij E {0,1} sum
Correspondingly, the determining module is specifically configured to solve the multi-objective non-invasive load decomposition model F (x) by adopting a multi-objective evolutionary algorithm, so as to obtain a corresponding electric equipment type and a corresponding working mode when five objective functions in the F (x) are simultaneously minimum;
in the multi-objective evolutionary algorithm, the constraint conditions are expressed by adopting an encoding and decoding mode: each bit of the code represents a type of powered device, where 0 represents the powered device being off and k represents the powered device being in the kth mode of operation.
As can be seen from the above technical solution, compared with the prior art, the present disclosure provides a non-invasive load decomposition method and apparatus, where the non-invasive load decomposition method specifically includes obtaining an operation record of an electrical device and power data on a bus, where the power data includes: a current waveform; preprocessing the power data according to the operation record of the electric equipment to screen out first power data; determining four power parameter values corresponding to each moment according to the screened first power data; establishing a multi-target non-invasive load decomposition model with five optimization targets according to the power parameter values and the current waveforms; the type of powered device and the corresponding mode being operated on the bus are determined based on the multi-objective non-intrusive load decomposition model. The non-invasive load decomposition method provided by the invention has the advantages of high decomposition accuracy, stable result and stronger robustness.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present invention, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a non-invasive load splitting method according to the present invention;
FIG. 2 is a flow chart of a method for preprocessing power data and screening out first power data according to the present invention;
FIG. 3 is a flow chart of a method for eliminating outliers provided by the present invention;
FIG. 4 is a schematic diagram of a non-invasive load splitting apparatus provided by the present invention;
FIG. 5 is a schematic diagram of a pretreatment module according to the present invention;
FIG. 6 is a schematic diagram of an outlier rejection unit according to the present invention;
FIG. 7 is a schematic diagram of current waveforms and voltage waveforms obtained by sampling 4 electric devices according to the present invention;
FIG. 8 is a schematic diagram of current-voltage waveforms before pretreatment according to the present invention;
fig. 9 is a schematic diagram of a current-voltage waveform after pretreatment according to the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, the embodiment of the invention discloses a non-invasive load decomposition method, which specifically comprises the following steps:
s1: acquiring operation records of electric equipment and power data on a bus, wherein the power data comprises: current, voltage, power factor, current waveform and voltage waveform;
s2: preprocessing the power data according to the operation record of the electric equipment to screen out first power data;
s3: determining four power parameter values corresponding to each moment according to the screened first power data;
s4: establishing a multi-target non-invasive load decomposition model with five optimization targets according to the power parameter values and the current waveforms;
s5: and determining the type of the electric equipment which is working on the bus and the corresponding working mode based on the multi-target non-invasive load decomposition model.
The invention provides a non-invasive load decomposition method, which has high decomposition accuracy, stable result and stronger robustness.
Referring to fig. 2, in order to further optimize the above technical solution, step S2: preprocessing and screening the first power data according to the operation record of the electric equipment, wherein the preprocessing and screening the first power data specifically comprises the following steps:
S21: dividing the power data according to the operation events of each electric equipment according to the operation records of the electric equipment to obtain a power data dividing result, so that the power data in the same interval are in the same mode of the same equipment;
s22: carrying out abnormal value detection and elimination on the current and the voltage in the same interval to obtain an abnormal value elimination result;
s23: and calculating average values of current and voltage in the interval without electric equipment according to the electric data dividing result and the abnormal value removing result, and subtracting the average values of the current and the voltage in the interval without electric equipment from the current values and the voltage values corresponding to the interval with electric equipment under the same circuit to obtain first electric data.
Referring to fig. 3, step S22 described above: the specific method for detecting and rejecting the abnormal values of the current and the voltage in the same interval further comprises the following steps:
s221: mad=media (|a) according to the formula i Median (a) |), determining the median absolute deviation, wherein a represents the same intervalDown current value or voltage value, A i Representing a corresponding current value or voltage value at the ith moment, wherein MAD represents a median absolute deviation;
s222: judging whether the current value or the voltage value in the same interval at any time deviates from the median value by more than N times of absolute deviation of the median value, if so, judging that the current value or the voltage value is an abnormal value at the moment and eliminating the abnormal value. Preferably, n=3.
In order to further optimize the above technical solution, step S3: determining four power parameter values corresponding to each moment according to the screened first power data specifically comprises the following steps:
according to the formula: s=vi, determining the apparent power corresponding to each moment, where V represents a voltage value at a certain moment, I represents a current value at a certain moment, and S represents the apparent power;
according to the formula: p=vicos (Φ), determining the active power corresponding to each moment, wherein cos (Φ) represents a power factor and P represents the active power;
according to the formula: q=visin (Φ), determining reactive power corresponding to each moment, wherein sin (Φ) represents sine Φ of a phase difference between voltage and current, no specific physical meaning exists, and Q represents reactive power;
according to the formula:
the corresponding harmonics at each time instant are determined, where a (T) represents the current value at time instant T, provided that t=0, 1,2, …, T-1, T represents the number of sampling points in one current waveform period (e.g., 1/50S), and X (k) is the coefficient of the kth harmonic.
Here, the order of calculation of the four power parameter values is not limited, and the four power parameter values may be calculated.
In order to further optimize the above technical solution, step S4: according to the power parameter value and the current waveform, a multi-objective non-invasive load decomposition model with five optimization objectives is established, and the method specifically comprises the following steps:
Constructing five optimization targets according to the power parameter values and the current waveforms; wherein the power parameter values include active power, reactive power, apparent power and harmonics; the specific method comprises the following steps:
according to the formula:
constructing a first objective function based on the current waveform, wherein I ij (t) represents the current value when device i is operating independently in j mode at time t; t represents the number of sampling points in one current waveform period (e.g., 1/50S), t=0 represents that the voltage phase is at a specific time, at which the voltage waveform is changing from maximum to minimum, N represents the total number of devices, M i Representing the total number of modes that device I contains, I representing the combined current waveform for an unknown device type; x is x ij Indicating that consumer i is in operating mode j, where i=1, 2, …, N, j=1, 2, …, M i ；
According to the formula:
constructing a second objective function based on reactive power, wherein Q ij The reactive power of the equipment i in independent operation in j mode is represented, and Q represents the reactive power of the unknown equipment type;
according to the formula:
constructing a third objective function based on the active power, wherein P ij Representing the active power of the device i when operating independently in j mode, P representing the active power of the unknown device type;
According to the formula:
constructing a fourth objective function based on apparent power, wherein S ij Representing the apparent power of the device i when operating independently in j mode, S representing the apparent power of the unknown device type;
according to the formula:
constructing a fifth objective function based on harmonics, wherein H ij (k) Representing the harmonics of the device i when operating independently in j mode, K is the order of the largest harmonics, H (K) represents the harmonics unknown to contain the device type.
Here, the above-mentioned five objective functions are not sequentially constructed, and the order of construction is not limited, so long as all of the five objective functions are constructed before the multi-objective invasive load decomposition model is built.
According to the formula: minimum F (x) = (F) 1 (x)，f 2 (x)，f 3 (x)，f 4 (x)，f 5 (x) The five objective functions described above are constructed as a multi-objective non-invasive load decomposition model F (x), wherein,
representing the type of consumer required, the constraint x needs to be satisfied ij E {0,1} and }>
Correspondingly, step S5: determining the type of the electric equipment which is working on the bus and the corresponding working mode based on the multi-target non-invasive load decomposition model, wherein the method specifically comprises the following steps of:
solving the multi-target non-invasive load decomposition model F (x) by adopting a multi-target evolutionary algorithm to obtain the corresponding electric equipment type and the corresponding working mode when five optimization targets in the F (x) are simultaneously minimum;
In the multi-objective evolutionary algorithm, the constraint conditions are expressed by adopting an encoding and decoding mode: each bit of the code represents a type of powered device, where 0 represents the powered device being off and k represents the powered device being in the kth mode of operation.
In addition, the embodiment of the invention also discloses a non-invasive load decomposition device, referring to fig. 4, the device comprises:
the system comprises an acquisition module 1, a control module and a control module, wherein the acquisition module is used for acquiring operation records of electric equipment and electric data on a bus, and the electric data comprises: a current waveform;
the preprocessing module 2 is used for preprocessing the electric power data according to the operation record of the electric equipment to screen out first electric power data;
the parameter calculation module 3 is used for determining four power parameter values corresponding to each moment according to the screened first power data;
the model building module 4 is used for building a multi-target non-invasive load decomposition model with five optimization targets according to the power parameter values and the current waveforms;
a determining module 5, configured to determine a type of electric equipment and a corresponding mode in operation on the bus based on the multi-objective non-invasive load decomposition model.
In order to further optimize the above technical solution, the preprocessing module 2 specifically includes:
The dividing unit 21 is configured to divide the power data according to the operation records of the electric devices and each electric device operation event to obtain a power data division result, so that the power data in the same interval are in the same mode of the same device;
an outlier rejection unit 22, configured to perform outlier detection and rejection on the current and the voltage in the same interval, so as to obtain an outlier rejection result;
and the calculating unit 23 is configured to calculate average values of current and voltage in the interval where no electric equipment works according to the electric power data dividing result and the abnormal value removing result, and subtract the average values of current and voltage in the interval where no electric equipment works from the current values and the voltage values corresponding to the interval where all electric equipment works under the same circuit to obtain the first electric power data.
Here, the same circuit is understood to be the same household.
In order to further optimize the above technical solution, the outlier rejection unit 22 specifically includes:
calculation subunit 221For determining according to the formula mad=media (|a) i Median (a) |), determining the median absolute deviation, wherein a represents the current or voltage value at the same interval, a i Representing a corresponding current value or voltage value at the ith moment, wherein MAD represents a median absolute deviation;
And a rejecting subunit 222, configured to determine whether the current value or the voltage value in the same interval at any time deviates from the median by more than N times of the median absolute deviation, and if so, then the current value or the voltage value is an abnormal value and reject the abnormal value. Preferably, n=3.
In order to further optimize the above technical solution, the parameter calculation module 3 specifically includes:
an apparent power calculation unit for calculating apparent power according to the formula: s=vi, determining the apparent power corresponding to each moment, where V represents a voltage value at a certain moment, I represents a current value at a certain moment, and S represents the apparent power;
the active power calculation unit is used for calculating the active power according to the formula: p=vicos (Φ), determining the active power corresponding to each moment, wherein cos (Φ) represents a power factor and P represents the active power;
the reactive power calculation unit is used for calculating the reactive power according to the formula: q=visin (Φ), determining reactive power corresponding to each moment, wherein Q represents reactive power;
a harmonic calculation unit for calculating a harmonic according to the formula:
the corresponding harmonics at each time instant are determined, wherein a (T) represents the current value at time instant T, provided that t=0, 1,2, …, T-1, T represents the number of sampling points in one current waveform period, and X (k) is the coefficient of the kth harmonic.
In order to further optimize the above technical solution, the model building module 4 specifically includes:
a first objective function construction unit configured to, according to the formula:
a first objective function is constructed based on the current waveform, wherein,I ij (t) represents the current value when device i is operating independently in j mode at time t; t represents the number of sampling points in one current waveform period, t=0 represents that the voltage phase is at a specific time, the voltage waveform at this time is changing from the maximum value to the minimum value, N represents the total number of devices, M i Representing the total number of modes that device I contains, I representing the combined current waveform for an unknown device type; x is x ij Indicating that consumer i is in operating mode j, where i=1, 2, …, N, j=1, 2, …, M i ；
A second objective function construction unit configured to, according to the formula:
constructing a second objective function based on reactive power, wherein Q ij The reactive power of the equipment i in independent operation in j mode is represented, and Q represents the reactive power of the unknown equipment type;
a third objective function construction unit configured to, according to the formula:
constructing a third objective function based on the active power, wherein P ij Representing the active power of the device i when operating independently in j mode, P representing the active power of the unknown device type;
A fourth objective function construction unit configured to, according to the formula:
constructing a fourth objective function based on apparent power, wherein S ij Representing the apparent power of the device i when operating independently in j mode, S representing the apparent power of the unknown device type;
a fifth objective function construction unit configured to, according to the formula:
constructing a fifth objective function based on harmonics, wherein H ij (k) Representing the harmonics of device i when operating independently in j mode, K being the most significantThe order of the large harmonics, H (k), represents the harmonics unknown to contain the device type.
A load decomposition model building unit for building a load decomposition model according to the formula: minimum F (x) = (F) 1 (x)，f 2 (x)，f 3 (x)，f 4 (x)，f 5 (x) The five objective functions described above are constructed as a multi-objective non-invasive load decomposition model F (x), wherein,
representing the type of consumer required, the constraint x needs to be satisfied ij E {0,1} sum
Correspondingly, the determining module 5 is specifically configured to solve the multi-objective non-invasive load decomposition model F (x) by using a multi-objective evolutionary algorithm, so as to obtain a corresponding type of electric equipment and a corresponding working mode when five objective functions in the F (x) are simultaneously minimized;
in the multi-objective evolutionary algorithm, the constraint conditions are expressed by adopting an encoding and decoding mode: each bit of the code represents a type of powered device, where 0 represents the powered device being off and k represents the powered device being in the kth mode of operation.
It should be noted that there are many known multi-objective algorithms, for example: reference vector directed evolution algorithms (RVEA), but solving the multi-objective optimization problem using multi-objective evolution algorithms has been rarely reported.
In order to further discuss the non-invasive load decomposition method and device provided by the invention in detail, the technical scheme of the invention is described in detail by taking 241 bus power data effectively recorded in a certain household as an example.
The embodiment of the invention obtains four electric power parameter values based on current, voltage, power factor, current waveform and voltage waveform, utilizes the obtained electric power parameter values and the obtained electric current waveform to construct five optimization targets, proposes a multi-target non-invasive load decomposition model, and solves the model by combining a multi-target evolutionary algorithm, thereby determining the type of electric equipment working on a bus and the corresponding working mode.
Firstly, acquiring operation records of electric equipment and electric data on a bus, wherein the electric data comprise current, voltage, power factor, current waveform and voltage waveform. And preprocessing the electric power data according to the operation record of the electric equipment to screen reasonable data, and finally determining four electric power parameter values corresponding to each moment according to the screened electric power data.
The collected three macroscopic power parameters comprise current, voltage and power factor, and the sampling rate of the parameters is 1Hz; the microloading characteristics obtained are current and voltage waveforms, the sampling rate of the above parameters in a 50Hz system of 6400Hz, and for the purpose of reducing the storage space, only the first 1/50 seconds of the waveform per second is used, as shown in fig. 7. A total of 16 operating modes were tested for 8 appliances, 8 appliances were fans, kettles, televisions, incandescent lamps, printers, computers, water dispensers and blowers as shown in table 1. The data set A only collects the situation that only one electric appliance is running; the data set B collected contains a case where a plurality of appliances (less than 4) are simultaneously operated. Using data set a, a 5-dimensional feature vector of 16 modalities can be obtained as a known load database; the data set B is used to verify the load decomposition method we propose.
Table 18 appliances 16 modes correspond to current, voltage and power factor
Using only one load feature does not distinguish between all consumers. Five load signatures are used for decomposing the load including macroscopic signatures (i.e., active power, reactive power, and apparent power) and microscopic signatures (i.e., current waveforms and harmonics). These features are defined as follows:
Feature superposition criteria: the load characteristics used by the proposed load decomposition model must meet the characteristic superposition criteria. The detailed definition of the feature overlay criteria is as follows:
wherein ψ is l (t) generating a superposition of the characteristic l for the simultaneous operation of the K consumers at time t,
representing a value corresponding to the generated characteristic l in the mode j of the electric equipment a; if the consumer a is in j-mode instantaneous operation at the time t+Δt and equation 1 is satisfied, it is indicated that the feature l satisfies the feature superposition criterion. If feature/meets the feature overlay criterion, this feature can be used to estimate the overlay load feature, i.e., the load feature value for simultaneous operation of more than one appliance.
The calculation formulas of the active power, the reactive power and the apparent power are as follows:
P＝VIcos(φ)， (2)
Q＝VIsin(φ)， (3)
S＝VI， (4)
where V and I are the voltage and current values, respectively, and cos phi is the power factor. The above features all meet the feature overlay criteria.
Current waveform: the current waveform is 128 data points sampled in one cycle (1/50 s). Because of the high sampling rate, the current waveform contains detailed electrical characteristics, but the voltage waveform is not significantly different as shown in fig. 7 (b). The current waveform is selected and it meets the characteristic overlap-add criterion.
Harmonic wave: and performing fast Fourier transform on the current waveform to obtain current harmonic waves. The rectangular form of the harmonic (a+jb) satisfies the feature superposition criterion. But the polar coordinate form A & lt theta with physical meaning does not meet the characteristic superposition criterion. The calculation formula of the rectangular coordinate form of the harmonic is as follows:
Wherein A (t) represents a current value at time t; the constraint is t=0, 1,2, …, T-1; t represents the number of sampling points in one current waveform period (1/50 s); x (k) is a coefficient of the kth harmonic. Table 3 is the power parameter values determined from the power data and this is a macroscopic feature prior to preprocessing. The harmonics are calculated after the preprocessing.
Table 2 pretreatment of the previous 8 appliances 16 modes corresponding to active, apparent and reactive power
The preprocessing of the data includes: (1) Separating the switching event of the electric equipment by using the operation record of the electric equipment; (2) Simultaneously, the current waveform and the voltage waveform are sampled to ensure that the initial phase (changing from the crest to the trough) of the voltage waveform is the same at each moment, and meanwhile, the current waveform and the voltage waveform are required to have a correct phase relation; (3) removing outliers to ensure accuracy of post-processing; (4) The load characteristic of the operation of the electrical appliance is subtracted from the load characteristic of the operation of the electrical appliance so as to reduce the influence of noise. The processed partial data are shown in Table 3.
TABLE 3 active, apparent and reactive Power for 16 modes of 8 appliances after pretreatment
Step two: and according to the five electric power parameter values, establishing a multi-target load decomposition model with five optimization targets, and finally solving the proposed multi-target load decomposition model by using a multi-target evolutionary algorithm to determine the type and corresponding mode of the electric equipment running on the bus at the current moment.
The load decomposition problem is converted into an optimization problem, and each feature can form an optimization problem. A conventional optimization-based load decomposition approach is to weight multiple objective functions into one. However, because of the high degree of positive or negative correlation between certain features, the different optimization problems cannot be simply weighted; moreover, the weighting parameters are sensitive to different data sets and are difficult to adjust, and the weighting parameters have a significant effect on the load decomposition accuracy. To address these shortcomings, the transformation of the load decomposition problem into a multi-objective optimization problem can be represented by the following formula:
minimize F(x)＝(f 1 (x)，f 2 (x)，f 3 (x)，f 4 (x)，f 5 (x))，
the constraint conditions are as follows:
x ij ∈{0，1}，
wherein the F (x) table is a five-dimensional objective function that must be optimized simultaneously; x is x ij Representing the running state of the electric equipment i in the mode j (0 represents off, 1 represents on), N is the maximum electric equipment number in the database, M i Representing the total pattern number of the electric equipment i. These five optimization problems are constructed using five different load characteristics. The calculation formula of the five objective functions is as follows:
I ij (t) represents the current value of the electric equipment i when independently operating in the mode j at the moment t; t represents the number of sampling points in one current waveform period (1/50 s); t=0 indicates that the voltage phase is at a specific time, at which the voltage waveform is changing from the maximum value to the minimum value; n represents the total number of electric equipment; m is M i Representing the total number of modes contained by the electric equipment i; i represents a current waveform of an unknown consumer type; q (Q) ij Representing reactive power when the device i is independently operated in j mode; q represents reactive power of unknown consumer type; p (P) ij Representing the active power of device i when operating independently in j mode; p represents the active power of the unknown electric equipment type; s is S ij Representing the apparent power of device i when operating independently in j mode; s represents apparent power of unknown electric equipment type; h ij (k) Representing harmonics of device i when operating independently in j mode; k is the order of the largest harmonics and H (K) represents the harmonics of the unknown consumer type.
And solving the multi-target load decomposition model by using a multi-target evolutionary algorithm to obtain the working state of each electric equipment. Firstly, a new coding mode is proposed to solve the constraint condition x ij E {0,1} sum
Each bit of the new coding scheme represents oneAnd the type of electric equipment, wherein 0 represents the electric equipment, and k represents that the electric equipment is in a kth working mode. Genetic operators (bit mutation operators and single point crossover operators) are then used to generate offspring. Thereafter, the target rank R is calculated using the assigned target ranks, and the adaptation value is evaluated using a multi-target evolutionary algorithm for obtaining adaptation values, while the next generation is screened from the same target ranks using these adaptation values. And screening the filial generation according to the target grade, then further screening from the same target grade according to the adaptive value obtained by the multi-target evolutionary algorithm, and finally, outputting the final population while meeting the stop condition.
Assigning a target grade: and a method for distributing the target grade is provided to solve the constraint of the upper limit of the electric equipment. Firstly, calculating the number of electric equipment of an individual, and then distributing the individual containing 1 to U electric equipment numbers to R 1 . Assigning an individual containing N powered devices to R N-U+2 。
The following will further describe the above technical solutions in conjunction with specific examples, please refer to tables 4 and 5:
TABLE 4 active power, apparent power, reactive power, harmonic and current waveform values for 4 modes total for 3 appliances after pretreatment
Table 5 analysis on bus to obtain active, apparent, reactive, harmonic and current waveform values
| Active power (W) | Apparent power (W) | Reactive power (W) | Harmonic wave (A) | Current waveform (A) |
| 1710 | 1720 | 87 | 30 | 3 |
Table 4 is the corresponding active power, apparent power, reactive power, harmonic and current waveform values for 4 modes for 3 appliances. Table 5 is the values of the active power, apparent power, reactive power, harmonics and current waveforms measured at a certain point in time on the bus and obtained after the first processing step. Solving through a multi-objective evolutionary algorithm to obtain the type of electric equipment and a corresponding mode which enable 5 objective functions to be optimal; it is determined that the hot water kettle mode 1 and the printer mode 2 are in operation at this point.
According to the non-invasive load monitoring method based on the multi-objective evolutionary algorithm, the electric power data including current, voltage, power factor, current waveform and voltage waveform are preprocessed according to the operation record of electric equipment, and reasonable data are screened out; according to the screened power data, determining that four power parameter values corresponding to each moment comprise useful power, reactive power, apparent power and harmonic waves, establishing a multi-target non-invasive load decomposition model with five optimization targets by utilizing the power parameter values and current waveforms, solving the multi-target load decomposition model by utilizing a multi-target evolutionary algorithm, and determining the type and corresponding mode of electric equipment working on a bus. Therefore, the load monitoring method provided by the invention has the advantages of high decomposition accuracy, stable result and stronger robustness.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (6)
1. A method of non-intrusive load shedding comprising the steps of:
acquiring operation records of electric equipment and power data on a bus, wherein the power data comprises: a current waveform;
preprocessing the power data according to the operation record of the electric equipment to screen out first power data;
determining four power parameter values corresponding to each moment according to the screened first power data;
establishing a multi-target non-invasive load decomposition model with five optimization targets according to the power parameter values and the current waveforms;
determining the type of the electric equipment which is working on the bus and the corresponding working mode based on the multi-target non-invasive load decomposition model;
The determining four power parameter values corresponding to each moment according to the screened first power data specifically includes:
according to the formula: s=vi, determining the apparent power corresponding to each moment, where V represents a voltage value at a certain moment, I represents a current value at a certain moment, and S represents the apparent power;
according to the formula: p=vicos (Φ), determining the active power corresponding to each moment, wherein cos (Φ) represents a power factor and P represents the active power;
according to the formula: q=visin (Φ), determining reactive power corresponding to each moment, wherein Q represents reactive power;
according to the formula:
determining a harmonic corresponding to each moment, wherein A (T) represents a current value at the moment T, and the constraint condition is that t=0, 1,2, …, T-1, T represents the number of sampling points in one current waveform period, and X (k) is a coefficient of a kth harmonic;
the method for establishing the multi-objective non-invasive load decomposition model with five optimization objectives according to the electric power parameter values and the current waveforms specifically comprises the following steps:
constructing five optimization targets according to the power parameter values and the current waveforms; wherein the power parameter values include active power, reactive power, apparent power and harmonics; the specific method comprises the following steps:
According to the formula:
constructing a first objective function based on the current waveform, wherein I ij (T) represents the current value when the device i is independently operated in j mode at time T, T represents the number of sampling points in one current waveform period, t=0 represents the voltage phase at a specific time, the voltage waveform at this time is changing from the maximum value to the minimum value, N represents the total number of devices, M i Representing the total number of modes contained by the consumer I, I representing the combined current waveform of an unknown device type, x ij Indicating that consumer i is in operating mode j, where i=1, 2, …, N, j=1, 2, …, M i ；
According to the formula:
constructing a second objective function based on reactive power, wherein Q ij The reactive power of the equipment i in independent operation in j mode is represented, and Q represents the reactive power of the unknown equipment type;
according to the formula:
constructing a third objective function based on the active power, wherein P ij Representing the active power of the device i when operating independently in j mode, P representing the active power of the unknown device type;
according to the formula:
constructing a fourth objective function based on apparent power, wherein S ij Representing the apparent power of the device i when operating independently in j mode, S representing the apparent power of the unknown device type;
According to the formula:
constructing a fifth objective function based on harmonics, wherein H ij (k) Representing the harmonics of the device i when operating independently in j mode, K being the order of the largest harmonics, H (K) representing the harmonics unknown to contain the device type;
according to the formula: minimum F (x) = (F) 1 (x)，f 2 (x)，f 3 (x)，f 4 (x)，f 5 (x) The five objective functions described above are constructed as a multi-objective non-invasive load decomposition model F (x), wherein,
representing the type of consumer required, the constraint x needs to be satisfied ij E {0,1} and }>
Correspondingly, the type of the electric equipment working on the bus and the corresponding working mode are determined based on the multi-target non-invasive load decomposition model, and the method specifically comprises the following steps:
solving the multi-target non-invasive load decomposition model F (x) by adopting a multi-target evolutionary algorithm to obtain the corresponding electric equipment type and the corresponding working mode when five optimization targets in the F (x) are simultaneously minimum;
in the multi-objective evolutionary algorithm, the constraint conditions are expressed by adopting an encoding and decoding mode: each bit of the code represents a type of powered device, where 0 represents the powered device being off and k represents the powered device being in the kth mode of operation.
2. The non-invasive load splitting method according to claim 1, wherein preprocessing the power data to screen the first power data according to the operation record of the electric equipment specifically comprises:
Dividing the power data according to the operation events of each electric equipment according to the operation records of the electric equipment to obtain a power data dividing result, so that the power data in the same interval are in the same mode of the same equipment;
carrying out abnormal value detection and elimination on the current and the voltage in the same interval to obtain an abnormal value elimination result;
and calculating average values of current and voltage in the interval without electric equipment according to the electric data dividing result and the abnormal value removing result, and subtracting the average values of the current and the voltage in the interval without electric equipment from the current values and the voltage values corresponding to the interval with electric equipment under the same circuit to obtain first electric data.
3. A non-invasive load decomposition method according to claim 2, wherein the specific method for detecting and rejecting abnormal values of current and voltage in the same interval comprises:
mad=media (|a) according to the formula i Media (a) |), determinationA median absolute deviation, wherein A represents a current value or a voltage value in the same interval, A i Representing a corresponding current value or voltage value at the ith moment, wherein MAD represents a median absolute deviation;
judging whether the current value or the voltage value in the same interval at any time deviates from the median value by more than N times of absolute deviation of the median value, if so, judging that the current value or the voltage value is an abnormal value at the moment and eliminating the abnormal value.
4. A non-invasive load splitting apparatus, comprising:
the system comprises an acquisition module, a control module and a control module, wherein the acquisition module is used for acquiring operation records of electric equipment and electric data on a bus, and the electric data comprises: a current waveform;
the preprocessing module is used for preprocessing the electric power data according to the operation record of the electric equipment and screening out first electric power data;
the parameter calculation module is used for determining four power parameter values corresponding to each moment according to the screened first power data;
the model building module is used for building a multi-target non-invasive load decomposition model with five optimization targets according to the power parameter values and the current waveforms;
the determining module is used for determining the type of the electric equipment which is working on the bus and the corresponding mode based on the multi-target non-invasive load decomposition model;
the parameter calculation module specifically comprises:
an apparent power calculation unit for calculating apparent power according to the formula: s=vi, determining the apparent power corresponding to each moment, where V represents a voltage value at a certain moment, I represents a current value at a certain moment, and S represents the apparent power;
the active power calculation unit is used for calculating the active power according to the formula: p=vicos (Φ), determining the active power corresponding to each moment, wherein cos (Φ) represents a power factor and P represents the active power;
The reactive power calculation unit is used for calculating the reactive power according to the formula: q=visin (Φ), determining reactive power corresponding to each moment, wherein Q represents reactive power;
a harmonic calculation unit for calculating a harmonic according to the formula:
determining a harmonic corresponding to each moment, wherein A (T) represents a current value at the moment T, and the constraint condition is that t=0, 1,2, …, T-1, T represents the number of sampling points in one current waveform period, and X (k) is a coefficient of a kth harmonic;
the model building module specifically comprises:
a first objective function construction unit configured to, according to the formula:
constructing a first objective function based on the current waveform, wherein I ij (t) represents the current value when device i is operating independently in j mode at time t; t represents the number of sampling points in one current waveform period, t=0 represents that the voltage phase is at a specific time, the voltage waveform at this time is changing from the maximum value to the minimum value, N represents the total number of devices, M i Representing the total number of modes contained by the electric equipment I, wherein I represents the combined current waveform of the unknown equipment type; x is x ij Indicating that consumer i is in operating mode j, where i=1, 2, …, N, j=1, 2, …, M i ；
A second objective function construction unit configured to, according to the formula: 
Constructing a second objective function based on reactive power, wherein Q ij The reactive power of the equipment i in independent operation in j mode is represented, and Q represents the reactive power of the unknown equipment type;
a third objective function construction unit configured to, according to the formula:
constructing a third objective function based on the active power, wherein P ij Indicating that device i is in j modeActive power when running independently, P represents active power of unknown device type;
a fourth objective function construction unit configured to, according to the formula:
constructing a fourth objective function based on apparent power, wherein S ij Representing the apparent power of the device i when operating independently in j mode, S representing the apparent power of the unknown device type;
a fifth objective function construction unit configured to, according to the formula:
constructing a fifth objective function based on harmonics, wherein H ij (k) Representing the harmonics of the device i when operating independently in j mode, K being the order of the largest harmonics, H (K) representing the harmonics unknown to contain the device type;
a load decomposition model building unit for building a load decomposition model according to the formula: minimum F (x) = (F) 1 (x)，f 2 (x)，f 3 (x)，f 4 (x)，f 5 (x) The five objective functions described above are constructed as a multi-objective non-invasive load decomposition model F (x), wherein, 
Representing the type of consumer required, the constraint x needs to be satisfied ij E {0,1} sum
Correspondingly, the determining module is specifically configured to solve the multi-objective non-invasive load decomposition model F (x) by adopting a multi-objective evolutionary algorithm, so as to obtain a corresponding electric equipment type and a corresponding working mode when five objective functions in the F (x) are simultaneously minimum;
in the multi-objective evolutionary algorithm, the constraint conditions are expressed by adopting an encoding and decoding mode: each bit of the code represents a type of powered device, where 0 represents the powered device being off and k represents the powered device being in the kth mode of operation.
5. The non-invasive load splitting apparatus according to claim 4, wherein the preprocessing module specifically comprises:
the dividing unit is used for dividing the power data according to the operation records of the electric equipment and each electric equipment operation event to obtain a power data dividing result, so that the power data in the same interval are in the same mode of the same equipment;
the abnormal value removing unit is used for detecting and removing abnormal values of the current and the voltage in the same interval to obtain abnormal value removing results;
and the calculation unit is used for calculating the average value of the current and the voltage in the interval without the electric equipment according to the electric data dividing result and the abnormal value removing result, and subtracting the average value of the current and the voltage in the interval without the electric equipment from the current value and the voltage value corresponding to the interval with the electric equipment under the same circuit to obtain the first electric data.
6. The non-invasive load decomposition apparatus according to claim 5, wherein the outlier rejection unit specifically comprises:
a calculation subunit for calculating mad=media (|a) according to the formula i Median (a) |), determining the median absolute deviation, wherein a represents the current or voltage value at the same interval, a i Representing a corresponding current value or voltage value at the ith moment, wherein MAD represents a median absolute deviation;
and the eliminating subunit is used for judging whether the current value or the voltage value in the same interval at any moment deviates from the median value by more than N times of absolute deviation of the median value, and if so, the current value or the voltage value is an abnormal value at the moment and the abnormal value is eliminated.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910390156.8A CN110416995B (en) | 2019-05-10 | 2019-05-10 | Non-invasive load decomposition method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910390156.8A CN110416995B (en) | 2019-05-10 | 2019-05-10 | Non-invasive load decomposition method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110416995A CN110416995A (en) | 2019-11-05 |
| CN110416995B true CN110416995B (en) | 2023-05-05 |
Family
ID=68358167
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910390156.8A Active CN110416995B (en) | 2019-05-10 | 2019-05-10 | Non-invasive load decomposition method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110416995B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110979079B (en) * | 2019-12-17 | 2021-06-29 | 深圳华建电力物联技术有限公司 | Method and device for positioning electric vehicle to tail end |
| CN111092434B (en) * | 2019-12-25 | 2023-07-04 | 天津大学 | Residential community power load control method and device based on non-invasive electricity consumption data |
| CN111091911A (en) * | 2019-12-30 | 2020-05-01 | 重庆同仁至诚智慧医疗科技股份有限公司 | System and method for screening stroke risk |
| CN111190067B (en) * | 2020-01-21 | 2021-03-19 | 山东建筑大学 | Catering disinfection equipment use monitoring method, server and system based on Internet of things |
| CN111934318B (en) * | 2020-08-13 | 2024-02-27 | 彭浩明 | Non-invasive power load decomposition method, apparatus, device and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103001230A (en) * | 2012-11-16 | 2013-03-27 | 天津大学 | Non-invasive power load monitoring and decomposing current mode matching method |
| CN105186693A (en) * | 2015-09-28 | 2015-12-23 | 南方电网科学研究院有限责任公司 | Non-intrusive mode electrical load identification system and method |
| CN105514984A (en) * | 2015-12-07 | 2016-04-20 | 河南许继仪表有限公司 | Plug-and-play non-intrusive load decomposition device |
| CN107525964A (en) * | 2017-10-23 | 2017-12-29 | 云南电网有限责任公司电力科学研究院 | A kind of recognition methods of non-intrusion type load and device based on fusion decision-making |
| CN109492667A (en) * | 2018-10-08 | 2019-03-19 | 国网天津市电力公司电力科学研究院 | A kind of feature selecting discrimination method for non-intrusive electrical load monitoring |
| CN109596912A (en) * | 2018-11-21 | 2019-04-09 | 河海大学 | A kind of decomposition method of non-intrusion type power load |
-
2019
- 2019-05-10 CN CN201910390156.8A patent/CN110416995B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103001230A (en) * | 2012-11-16 | 2013-03-27 | 天津大学 | Non-invasive power load monitoring and decomposing current mode matching method |
| CN105186693A (en) * | 2015-09-28 | 2015-12-23 | 南方电网科学研究院有限责任公司 | Non-intrusive mode electrical load identification system and method |
| CN105514984A (en) * | 2015-12-07 | 2016-04-20 | 河南许继仪表有限公司 | Plug-and-play non-intrusive load decomposition device |
| CN107525964A (en) * | 2017-10-23 | 2017-12-29 | 云南电网有限责任公司电力科学研究院 | A kind of recognition methods of non-intrusion type load and device based on fusion decision-making |
| CN109492667A (en) * | 2018-10-08 | 2019-03-19 | 国网天津市电力公司电力科学研究院 | A kind of feature selecting discrimination method for non-intrusive electrical load monitoring |
| CN109596912A (en) * | 2018-11-21 | 2019-04-09 | 河海大学 | A kind of decomposition method of non-intrusion type power load |
Non-Patent Citations (2)
| Title |
|---|
| Load_Signature_StudyPart_II_Disaggregation_Framework_Simulation_and_Applications;Jian Liang等;《Load signature study—part II: disaggregation framework, simulation, and applications》;20101231;第561-562页 * |
| 非侵入式电力负荷多目标分解框架;杨立余;《电力系统保护与控制》;20200316;第48卷(第6期);第100-107页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110416995A (en) | 2019-11-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110416995B (en) | Non-invasive load decomposition method and device | |
| CN106600074B (en) | DFHSMM-based non-invasive power load monitoring method and system | |
| CN108021736B (en) | Load switching action monitoring method based on sliding window residual error model | |
| US8892376B2 (en) | Data processing device, data processing method, and program | |
| US9104189B2 (en) | Methods and apparatuses for monitoring energy consumption and related operations | |
| Panapakidis et al. | Pattern recognition algorithms for electricity load curve analysis of buildings | |
| US20170351288A1 (en) | Non-invasive online real-time electric load identification method and identification system | |
| CN103135009B (en) | Electric appliance detection method and system based on user feedback information | |
| CN109856299A (en) | A kind of transformer online monitoring differentiation threshold value dynamic setting method, system | |
| CN110223196A (en) | Analysis method of opposing electricity-stealing based on typical industry feature database and sample database of opposing electricity-stealing | |
| CN109359665B (en) | Household appliance load identification method and device based on support vector machine | |
| Lee | Electric load information system based on non-intrusive power monitoring | |
| Andrean et al. | A hybrid method of cascade-filtering and committee decision mechanism for non-intrusive load monitoring | |
| Li et al. | Energy data generation with wasserstein deep convolutional generative adversarial networks | |
| Eskander et al. | A complementary unsupervised load disaggregation method for residential loads at very low sampling rate data | |
| Hošovský et al. | Comparative study of week-ahead forecasting of daily gas consumption in buildings using regression ARMA/SARMA and genetic-algorithm-optimized regression wavelet neural network models | |
| CN111932051A (en) | Malicious behavior detection method based on non-invasive power terminal time sequence monitoring | |
| 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 | |
| CN114942344A (en) | Non-invasive electrical appliance identification method, system, medium, equipment and terminal | |
| CN113193654A (en) | Event-driven non-intrusive power load monitoring method based on transient and steady state combination characteristics | |
| CN113762355A (en) | User abnormal electricity consumption behavior detection method based on non-invasive load decomposition | |
| CN116861316B (en) | Electrical appliance monitoring method and device | |
| CN113595071A (en) | Transformer area user identification and voltage influence evaluation method | |
| Mariño et al. | NILM: Multivariate DNN performance analysis with high frequency features |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |