JP6225079B2 - Air conditioner operation detection method and system - Google Patents

Description

  The present invention relates to a method for detecting an air conditioner operation by non-intrusive load monitoring (NILM) for estimating an electrical device connected to a destination from information on power and current measured by a distribution board, and a system thereof.

  Conventionally, there is a technique for identifying an operating device using time-series power or current data (see Non-Patent Document 1). When identifying specific electrical device operation from among a plurality of electrical devices using time-series power data (hereinafter referred to as power time sequence), the steep power inherent in each electrical device as shown in FIG. It features the rise and the increase in power with a certain slope after the steep rise. In this way, it was possible to detect the operation of an electrical device that showed a relatively simple change in power, that is, a constant slope after a steep rise.

Cole A.I., Albicki A, Data extraction for effective non-intrusive identification of residential power loads, Instrumentation and Measurement Technology Conference, 1998.IMTC / 98. Conference Proceedings.IEEE, pp.812-815 vol2

  On the other hand, in an air conditioner that has a thermometer inside and is controlled to keep the temperature constant, the power changes in a complex manner so that small fluctuations in power consumption with periodicity continue after a sharp rise in power. . Such power time series changes, especially small fluctuations in power consumption with periodicity, cannot be detected by the detection conditions in the conventional detection method, so the conventional detection method can accurately estimate the operation of the air conditioner. There is a problem that it is difficult.

  The present invention has been made in view of such problems, and an object of the present invention is to detect an air conditioner operation that can capture characteristic power fluctuations that occur at the start of the operation of an air conditioner, and a system thereof. Is to provide.

In order to solve the above-mentioned problem, the present invention is a method of detecting an air conditioner operation, the first time series power data is obtained from a power sensor capable of detecting only the power consumption of the air conditioner, From the power data of 1 time series, the rising time width, the rising power width, the maximum small waveform time width, the minimum small waveform time width, the maximum small waveform power width after the rising section from the start of the operation of the air conditioner , a step of computing a parameter for calculating the minimum small wave power width and, acquiring power data of the second time series from detectable power sensors the power supplied to the plurality of electric devices including the air conditioner Detecting a power fluctuation greater than or equal to the rising power width in the rising time width from the second time series power data; and After the power fluctuation of the power data of the series, at least, the minimum small waveform duration than said maximum small waveform waveform of one period in the range of time width is present, and, the power variation of the waveform, the minimum When the small waveform power width is equal to or greater than the maximum small waveform power width, the step of detecting the small waveform as the small waveform, and the small waveform is detected twice or more in succession, the power fluctuation And estimating an air conditioner operation that estimates that the air conditioner is operating from a start time.

  According to a second aspect of the present invention, in the air conditioner operation detection method according to the first aspect, the step of calculating the parameter includes a range specified as a small waveform range by a user of the first time-series power data. A step of calculating a time width at which the autocorrelation value is minimum multiplied by a predetermined constant as the minimum small waveform time width and the maximum small waveform time width, and the first time-series power data In the range designated as the small waveform range by the user, the maximum small waveform power width and the maximum small waveform power width are calculated by multiplying the maximum power fluctuation amount in one cycle of the small waveform by a predetermined constant. And a step of performing.

According to a third aspect of the present invention, in the method of detecting an air conditioner operation according to the first or second aspect, the step of detecting the small waveform further includes a step after the power fluctuation of the second time-series power data. The power data that can be considered to be the same in the power value and the sign of the slope with respect to the power data at the first time is a range from the first small time to the minimum small waveform time width to the maximum small waveform time width. The waveform is not detected as a small waveform until it is detected for the first time.

According to a fourth aspect of the present invention, there is provided a detection system for an air conditioner operation, wherein the first time-series power data and a plurality of electric devices including the air conditioner are supplied from a power sensor capable of detecting only power consumption of the air conditioner. An input interface for acquiring second time-series power data from a power sensor capable of detecting generated power, and a storage unit for storing the first time-series power data and the second time-series power data From the first time series power data, the rising time width of the rising section from the start of the operation of the air conditioner, the rising power width, the maximum small waveform time width after the rising section, the minimum small waveform time width, the maximum a first calculator for calculating the parameters for calculating the small wave power width, and the smallest small wave power width, from the power data of the second time series, contact the rise time width Wherein detecting the rising power width or more power fluctuations Te, said after the power fluctuation of the power data of the second time series, at least, one period in the range of less than the minimum small waveform duration than said maximum small waveform duration And the amount of power fluctuation of the waveform is not less than the minimum small waveform power width and not more than the maximum small waveform power width, the waveform is detected as a small waveform, and the small waveform is continuously detected. A second calculation unit that estimates an air conditioner operation that estimates that the air conditioner is operating from the start time of the power fluctuation when detected twice or more.

  According to a fifth aspect of the present invention, in the air conditioner operation detection system according to the fourth aspect, the first arithmetic unit is a range designated as a small waveform range by a user of the first time-series power data. The time width at which the autocorrelation value is minimized multiplied by a predetermined constant to obtain the minimum small waveform time width and the maximum small waveform time width, and the user of the first time series power data In the range designated as the small waveform range by the above, the minimum small waveform power width and the maximum small waveform power width are calculated by multiplying the maximum fluctuation amount of power in one cycle of the small waveform by a predetermined constant. It is characterized by.

According to a sixth aspect of the present invention, in the air conditioner operation detection system according to the fourth or fifth aspect, the second calculation unit further includes a second after the power fluctuation of the second time-series power data. The power data that can be considered to be the same in the power value and the sign of the slope of the power data at the time 1 is within the range from the minimum small waveform time width to the maximum small waveform time width after the first time. The waveform is not detected as a small waveform until it is detected for the first time.

  The present invention makes it possible to accurately detect an air conditioner operation by capturing characteristic power fluctuations that occur at the start of the air conditioner operation.

It is a figure which shows the example of a fluctuation | variation of the electric power at the time of an electric equipment operation | movement. It is a figure which shows the image of the steep rise and small waveform in the electric power time series when an air conditioner operate | moves. It is a figure which shows the function structure of the detection system of the air-conditioner operation | movement which concerns on one Embodiment of this invention. It is a figure which shows the procedure of the parameter calculation in the detection system of the air-conditioner operation | movement of this invention. It is a figure which shows the example of each area designated by the user. It is a figure which shows the procedure of the air-conditioner operation | movement estimation in the detection system of the air-conditioner operation | movement of the system of this invention. It is a figure explaining the 1st calculation method of the small waveform detection of this invention. It is a figure explaining the 2nd calculation method of the small waveform detection of this invention.

  Hereinafter, embodiments of the present invention will be described in detail.

  Since the present invention has a feature that small fluctuations in power consumption with periodicity continue after a steep rise in power at the time of start-up of the air conditioner, after detecting a steep rise in the power time series, Furthermore, it is estimated that the air conditioner is operating when small waves (hereinafter, small waveforms) are continuously detected. FIG. 2 shows an image of a steep rise and a small waveform in the power time series when the air conditioner is operated.

  In the present invention, in order to detect a small waveform generated when the air conditioner is operated, the maximum small waveform time width, the minimum small waveform time width, the maximum small waveform power width, and the minimum small waveform power width of the small waveform are used as the feature amount. These four features are defined, and these are extracted. By using these feature quantities, it was possible to improve the estimation accuracy of the operating state of the air conditioner. Furthermore, by detecting a small waveform with periodicity, which is the time series power fluctuation of the air conditioner, it is possible to improve the estimation accuracy of the operating state of the air conditioner.

  There are roughly two procedures for the operation of this system, and one is a parameter calculation procedure and an air conditioner operation estimation procedure. In the parameter calculation procedure, the feature quantity during the air conditioner operation is extracted using only the power time series of the air conditioner. On the other hand, in the air conditioner operation estimation procedure, the operation state of the air conditioner is estimated using the power time series of the total power of the entire house.

  Hereinafter, an embodiment of the configuration and operation procedure of the air conditioner operation detection system of the present invention will be described.

  FIG. 3 shows a functional configuration of an air conditioner operation detection system according to an embodiment of the present invention. The system 100 includes an apparatus having functions of an input interface 110, a storage unit 120, a calculation unit 130, and an output interface 140, or a program executed by a computer. The input interface 110 is connected to the power sensors 202-1 and 202-2 and the input device 150, and the output interface 140 is connected to the monitor 160.

<Input interface 110>
The power time series input unit 111 is an interface that inputs the power time series obtained from the power sensor 202 or the like as numerical information to the system. The power time-series sampling interval SI used in this system is preferably about 20 seconds, but is not limited to 20 seconds. The power time series input unit acquires the power time series supplied to the air conditioner 301 and the power time series of the total power of the entire house supplied from the power system to the distribution board 201 from the power sensors 202-1 and 202-2.

  The range designation input unit 112 is an interface for connecting the input device 150 used for designating the range of rising and small waveforms from the power time series.

<Storage unit 120>
The power time series storage unit 121 stores the power time series acquired by the power time series input unit 111 within an arbitrary period.

  The parameter storage unit 122 stores the parameters calculated by the parameter calculation unit 134. The parameters used in this system are shown below.

-Rise time width: Time from the start to the end of the rising section where the rate of change is greater than or equal to a predetermined threshold in the power time series-Rise power width: Minimum power of the rising section where the rate of change is greater than or equal to the predetermined threshold in the power time series -Maximum small waveform time width: Maximum time of one cycle of small waveform in power time series-Minimum small waveform time width: Minimum time of one cycle of small waveform of power time series-Maximum small waveform power width: Power Maximum fluctuation amount of power in one cycle of small waveform in time series-Minimum small waveform power width: Minimum fluctuation amount of power in one cycle of small waveform in time series

  The rising section storage unit 123 stores the rising section specified by the range specification input unit 112.

  The small waveform section storage unit 124 stores the small waveform section designated by the range designation input unit.

<Calculation unit 130>
As will be described later, the rise detection unit 131 detects a steep rise from the power time series stored in the power time series storage unit 121.

  The small waveform detection unit 132 detects a small waveform from the power time series stored in the power time series storage unit 121 as described later.

  The air conditioner operation estimation unit 133 estimates the operation state of the air conditioner based on the detection result of the small waveform continuous with the steep rise of the power time series.

  The parameter calculation unit 134 calculates parameters used for calculation for detecting rising and small waveforms.

<Output interface 140>
The power time series output unit 141 is an interface that displays the power time series stored in the power time series storage unit 121 in a graph and outputs the graph to a monitor or the like.

  The estimation result output unit 142 is an interface that outputs the estimation result of the operation state of the air conditioner calculated by the air conditioner operation estimation unit 133.

  FIG. 4 shows a parameter calculation procedure in the air conditioner operation detection system of the present invention.

  The power time series of the air conditioner 301 is acquired from the power time series input unit 111 and stored in the power time series storage unit 121 (S401). The power time series output unit 121 outputs a graph of the power time series of the air conditioner 301 to the monitor 160 via the power time series output unit 141 (S402).

  For the power time series graph displayed on the monitor 160, the data of the rising section and the small waveform section specified by the user using the input device 150 is acquired (S403). FIG. 5 shows an example of each section designated by the user. A time width and a power width are designated for the rising section and the small waveform section, respectively.

  The power time series of the rising section specified in S403 is stored in the rising section storage unit 123, and the power time series of the small waveform section specified in S403 is stored in the small waveform section storage unit 124 (S404).

  The parameter calculation unit 134 calculates the rising time width, the rising power width, the maximum small waveform time width, the minimum small waveform time width, the maximum small waveform power width, and the minimum small waveform power width (S405). The arithmetic expressions are shown below. In addition, the following arithmetic expressions are examples, and are not limited to the following arithmetic expressions.

  -Rising time width ST: When there are a plurality of data at the time of rising with respect to the data of the rising section stored in the rising section storage unit 123, the maximum time width is set as the value of the rising time width ST. The calculation formula is as follows.

Here, ST _i represents each value of the rise time width specified in S403, and MAX () represents a function for selecting the maximum value.

  -Rising power width SP: When there are a plurality of data at the time of rising for the data of the rising section stored in the rising section storage unit 123, the minimum power width is set as the value of the rising power width SP. The calculation formula is as follows.

Here, SP _i represents each value of the rising power width specified in S403, and MIN () represents a function for selecting the minimum value.

  -Maximum small waveform time width WT_MAX and minimum small waveform time width WT_MIN: The maximum small waveform time width and the minimum small waveform using the autocorrelation function formula for the data of the small waveform section stored in the small waveform section storage unit 124 Calculate the time width value. The calculation formula is as follows.

Here, R _k is an autocorrelation value, p _t is a power value at time t, p _{t−k · SI} is a power value at the time when k sample points have returned from time t, μ is an average of power values, E () Represents an expected value function, and Var () represents a dispersion function. Note that it is effective to apply a moving average of about 5 minutes to the power time series as preprocessing in order to suppress the influence of the periodicity of minute waves caused by noise or the like in the autocorrelation function.

A time width WT of a typical small waveform is calculated by the following formula. That is, a value obtained by multiplying the sampling interval SI by k at which the autocorrelation value _Rk is minimized in the small waveform section is set as the time width WT of the small waveform. The parentheses in the following formula indicate conditions.

In actual air conditioner power consumption fluctuations, there is a slight fluctuation in the time width of the small waveform, so the value of the maximum small waveform time width WT_MAX and the minimum small waveform time width WT_MIN is obtained by multiplying the time width WT of the small waveform by an appropriate constant Is calculated. The calculation formula is as follows.

  -Maximum small waveform power width WP_MAX and minimum small waveform power width WP_MIN: The maximum fluctuation amount of power in one cycle of the small waveform is acquired for the data of the small waveform section, and the fluctuation is taken into consideration and the maximum fluctuation amount is appropriate. The parameters of the maximum small waveform power width WP_MAX and the minimum small waveform power width WP_MIN are calculated by multiplying by a constant.

  Then, each parameter calculated in S405 is stored in the parameter storage unit 122 (S406).

  Based on the parameters thus obtained in advance, the air conditioner operation detection system of the system of the present invention estimates the air conditioner operation as follows.

  FIG. 6 shows an air conditioner operation estimation procedure in the air conditioner operation detection system of the system of the present invention.

  A power time series of the total power of the entire house is acquired from the power time series input unit 111 and stored in the power time series storage unit 121 (S601). In the power time series of the total power acquired in S601 and the rising time width ST stored in the parameter storage unit 122, when there is a power fluctuation greater than the rising power width SP, a sharp rise in power is detected (S602). ). The calculation formula is as follows.

Note that when Judge_S = 1, a sharp rise in power is detected, and when Judge_S = 0, no detection is performed.

  Under the condition where the rising edge is detected in S602, the power time series acquired in S601, the maximum small waveform time width WT_MAX, the minimum small waveform time width WT_MIN, the maximum small waveform power width WP_MAX, and the minimum small waveform stored in the parameter storage unit 122. A small waveform is detected using the power width WP_MIN (S604). The calculation method is as follows. A small waveform is detected when either <small waveform detection calculation 1> or <small waveform detection calculation 2> shown below is satisfied, and no small waveform is detected when neither of them is satisfied. .

  In the present invention, the small waveform detection means that a waveform of one cycle exists in the range of the minimum small waveform time width and the maximum small waveform time width, and the power fluctuation amount of the waveform is the minimum small waveform power width and the maximum small waveform power width. In this case, a small waveform is detected when it falls within the range. Specifically, the calculation is performed as follows.

<Calculation for small waveform detection 1>
FIG. 7 is a diagram for explaining the first calculation method of small waveform detection according to the present invention. It is positive at the slope of the power time series at time t, and there is a point where the difference between the power slope at plus and p _t is equal to or less than a predetermined value at time t + WT_MIN from time t + WT_MAX interval, and, from time t time t + inclination in a section WT_MIN is without that power is _{p t} plus also occur, even for once, and the difference between the maximum power and minimum power value in the interval from time t to time t + k · SI is more WP_MIN WP_MAX A small waveform is detected when the following is true, and a small waveform is not detected when this condition is not satisfied.

Incidentally, the point where the difference between the power p _t is equal to or less than the predetermined value, the presence of that point, in a power time series obtained as a discrete value, the power point where the p _t from the time t + WT_MIN of time t + WT_MAX It is a point that can be considered to exist in the section. the difference between p _t may be set to appropriate values in accordance with the sampling period.

<Calculation for small waveform detection 2>
FIG. 8 is a diagram for explaining a second calculation method of small waveform detection according to the present invention. And negative in the slope of the power time series at time t, and there is a point where the difference between the power slope at minus and p _t is equal to or less than a predetermined value at time t + WT_MIN time from t + WT_MAX interval, and, from time t time t + inclination in a section WT_MIN is without regard to power minus is _{p t} also occur, even for once, and the difference between the maximum power and minimum power value in the interval from time t to time t + k · SI is more WP_MIN WP_MAX A small waveform is detected when the following is true, and a small waveform is not detected when this condition is not satisfied. Note that the difference between p _t, like <calculation 1 small waveform detection> may be set to appropriate values in accordance with the sampling period.

  S604 is repeated, and it is determined whether the small waveform is detected twice or more continuously (S605). If the small waveform is detected twice or more continuously, the air conditioner operation is estimated from the start time of the steep rise to the current time (S606), and if the small waveform is not detected twice or more continuously, the air conditioner is not operated. Estimate (S607).

DESCRIPTION OF SYMBOLS 100 Air-conditioner operation detection system 110 Input interface 111 Power time series input unit 112 Range designation input unit 120 Storage unit 121 Power time series storage unit 122 Parameter storage unit 123 Rising section storage unit 124 Small waveform detection unit 130 Calculation unit 131 Rising detection unit 132 Small Waveform Detection Unit 133 Air Conditioner Operation Estimation Unit 134 Parameter Calculation Unit 140 Output Interface 141 Power Time Series Output Unit 142 Estimation Result Output Unit 150 Input Device 160 Monitor 201 Distribution Panel 202 Power Sensor 301 Air Conditioner

Claims (6)

Obtaining first time-series power data from a power sensor capable of detecting only power consumption of an air conditioner;
From the power data of the first time series, the rising time width, the rising power width, the maximum small waveform time width after the rising section, the minimum small waveform time width, and the maximum small waveform from the start of the operation of the air conditioner a step of computing a parameter for calculating the power width, and the minimum small wave power width,
Obtaining second time-series power data from a power sensor capable of detecting power supplied to a plurality of electric devices including the air conditioner;
Detecting power fluctuations greater than or equal to the rising power width in the rising time width from the power data of the second time series;
After the power fluctuation of the power data of the second time series, at least, the minimum small waveform duration than said maximum small waveform waveform of one period in the range of time width is present, and, the power fluctuation of the waveform Detecting a small waveform for detecting the waveform as a small waveform when the amount is not less than the minimum small waveform power width and not more than the maximum small waveform power width; and
Estimating the air conditioner operation that estimates that the air conditioner is operating from the start time of the power fluctuation when the small waveform is detected twice or more continuously;
A method for detecting an air conditioner operation characterized by comprising: The step of calculating the parameter includes:
A time width in which the autocorrelation value in the range designated as the small waveform range by the user of the first time series power data is multiplied by a predetermined constant is obtained by multiplying the minimum small waveform time width by the maximum small size. A step for calculating a waveform time width;
In the range specified as the small waveform range by the user of the first time-series power data, a value obtained by multiplying the maximum fluctuation amount of power in one cycle of the small waveform by a predetermined constant is the minimum small waveform power width and the A step of calculating the maximum small waveform power width,
The method of detecting an air conditioner operation according to claim 1, comprising: The step of detecting the small waveform further includes power that can be considered to be the same in the power value and the sign of the slope of the power data at the first time after the power fluctuation of the second time series power data. The waveform is not detected as a small waveform until data is detected for the first time in the range from the minimum small waveform time width to the maximum small waveform time width after the first time. The method for detecting an air conditioner operation according to 2. First time-series power data from a power sensor capable of detecting only the power consumption of the air conditioner, and second time-series power data from a power sensor capable of detecting power supplied to a plurality of electric devices including the air conditioner. An input interface to obtain and
A storage unit for storing the first time-series power data and the second time-series power data;
From the power data of the first time series, the rising time width, the rising power width, the maximum small waveform time width after the rising section, the minimum small waveform time width, and the maximum small waveform from the start of the operation of the air conditioner A first calculation unit for calculating a parameter for calculating a power width and a minimum small waveform power width;
From the second time series power data, a power fluctuation that is greater than or equal to the rising power width in the rise time width is detected, and at least after the power fluctuation of the second time series power data, at least the minimum small waveform When there is a waveform of one cycle in a range not less than the time width and not more than the maximum small waveform time width, and the power fluctuation amount of the waveform is not less than the minimum small waveform power width and not more than the maximum small waveform power width, A second arithmetic unit that detects an air conditioner operation that estimates that the air conditioner is operating from a start time of the power fluctuation when the waveform is detected as a small waveform and the small waveform is continuously detected twice or more; ,
An air conditioner operation detection system characterized by comprising: The first calculation unit includes:
A time width in which the autocorrelation value in the range designated as the small waveform range by the user of the first time series power data is multiplied by a predetermined constant is obtained by multiplying the minimum small waveform time width by the maximum small size. Calculate to be the waveform time width,
In the range specified as the small waveform range by the user of the first time-series power data, a value obtained by multiplying the maximum fluctuation amount of power in one cycle of the small waveform by a predetermined constant is the minimum small waveform power width and the 5. The air conditioner operation detection system according to claim 4, wherein the calculation is performed so as to obtain a maximum small waveform power width. The second calculation unit further includes power data that can be considered to be the same in the power value and the sign of the slope with respect to the power data at the first time after the power fluctuation of the second time-series power data. 6. The method according to claim 4, wherein the waveform is not detected as a small waveform until it is detected for the first time in the range from the minimum small waveform time width to the maximum small waveform time width after the first time. The air conditioner operation detection system described.

Images