JP7262270B2 - Power mode management system - Google Patents

Description

The present invention relates to a power mode management system that manages the power mode of an operating device that operates household equipment such as household fuel cell cogeneration equipment.

Patent Literature 1 discloses a remote control device for remotely controlling a water heater. The remote control device includes an operating state display section for displaying the operating state of the water heater, a time display section for displaying the time, and an operating state for switching between an operating state in which the hot water supply operation of the water heater is possible and a standby state in which the hot water supply operation is impossible. Equipped with a power-saving switch that turns on/off the switch and operation status display. By operating the operation switch, an operation state in which the hot water supply operation of the water heater is possible and a standby state in which the hot water supply operation of the water heater is impossible are switched. By operating the power saving switch, there is a "normal display mode" in which the operation status display is turned on, and a "power saving mode" in which the operation status display is turned off when the water heater is not in operation. display mode can be switched. This reduces unnecessary power consumption in the display.

Patent Document 2 discloses a hot water supply system including a water heater body and a hot water supply remote controller as a hot water supply operation device. The hot water supply remote controller has operation keys and a display device, and in a normal display mode such as immediately after the hot water supply operation is started, the hot water supply operation and the hot water supply temperature are displayed on the display device screen. Furthermore, even after the hot water supply operation has started, if the hot water supply associated with the opening operation or the operation of the operation keys is not performed for a certain period of time, the display mode is switched from the normal display mode to the power saving display mode, The screen is in a non-display state until the user performs some operation. This reduces power consumption and extends the life of the display device.

Patent Literature 3 discloses an air conditioning system that performs power-saving operation of a plurality of air conditioners installed in the same indoor space based on human detection information, suction temperature information, and operation information. In this system, the control unit of each air conditioner is connected by wire or wirelessly, and each of the control units of these multiple air conditioners detects information detected by the human sensor of all other air conditioners in the same space. , based on the set temperature information of the remote control and the suction temperature information of the suction temperature sensor, the air conditioning load of each air conditioned space and the air conditioning load of the entire air conditioned space based on these air conditioned spaces are calculated, and based on the calculation result Each compressor is driven and controlled.

Patent Literature 4 discloses a stay-at-home determination system that determines whether a person is at home or out of a house. This system consists of status information indicating whether a person is at home or out of the house, human detection information output from a human detection sensor that detects the presence of a person at the entrance, and whether or not the entrance door is closed. Opening/closing information output from an opening/closing sensor that detects opening/closing is acquired. Further, it is determined whether or not the order in which the human detection information and the opening/closing information are obtained is the order corresponding to the state information. This determination result is output to the information terminal via the network, and the status information is updated according to this determination result.

Japanese Patent Application Laid-Open No. 2006-162112 JP-A-2003-4296 JP 2013-137189 A JP 2017-220837 A

In the water heater disclosed in Patent Document 1, operating the power saving switch turns off the display section of the remote control device, thereby saving power. However, if the power saving switch is not operated, the remote control display is performed even when the remote control display is unnecessary, and power is wasted. In the hot water supply system according to Patent Document 2, when the hot water supply remote control is not operated for a certain period of time, the display mode is switched to a power saving display mode, the display unit is turned off, and power saving is achieved. However, even if a person approaches the hot water supply remote controller in order to see the power generation status, power usage status, or time, if the remote controller is not operated for a certain period of time, the display will not be displayed unless some operation is performed. is in the off state, it is impossible to check the usage status and time.

In the air conditioning system according to Patent Document 3, based on the detection information of the human sensor, the operation of the air conditioner in a part of the air-conditioned space with few people in the room is stopped, or the set temperature or the suction temperature is set to the power saving side. By shifting to , power saving is achieved. However, although this air conditioning system uses a human sensor to detect the distribution of people, it does not have a function of detecting a person approaching a specific operating device such as a remote controller and switching the power mode. The at-home determination system according to Patent Document 4 uses a human sensor at the entrance and an open/closed sensor at the entrance door to accurately determine the presence of a person in the house. However, even this system does not disclose a technique for detecting a person approaching a specific operating device such as a remote controller and switching the power mode.

SUMMARY OF THE INVENTION It is an object of the present invention to accurately estimate whether or not a person is present around an operating device for operating a household appliance, even without a human sensor, and to determine whether or not a person is present around the operating device. To provide a power mode management system capable of switching an operation device to a normal power mode when it is estimated that there is no person around the operation device and to a power saving mode when it is estimated that there is no person around the operation device.

A characteristic configuration of the power mode management system according to the present invention, which operates a home appliance and manages the power mode of an operation device that does not have a human sensor, is characterized in that the power mode of the home appliance caused by the operation of the operation device changes over time. A state information generating unit that generates state information about an operating state, an environment information generating unit that generates environmental information about the environmental state of the home appliance over time, and the presence of a person around the operating device is estimated. an estimating unit having an estimating operation function that outputs a human presence signal when the presence of a person is not estimated around the operation unit, and outputs a human absence signal when the presence of a person around the operation device is not estimated; a power mode setting unit that sets a power saving mode as a mode and sets a normal power mode as the power mode when the human presence signal is output;
The state information includes at least one of time-series usage pattern data for hot water supply, time-series usage pattern data for hot water heating, and time-series usage pattern data for electricity,
The environmental information includes at least one of time-series pattern data of water temperature, time-series pattern data of air temperature, date data, and time data,
Machine learning in which the estimating unit outputs the human presence signal or the human absence signal using the state information and the environment information generated by the operation device having no human sensor as input data. configured as a unit
Presence of a person around another operating device, wherein the machine learning unit generates supervised learning information for the machine learning unit by an operating device other than the operating device that does not have a human sensor . Alternatively, output machine-learned data using composite information indicating the relationship between human information , which is time-series data indicating the absence of a person , and the state information and the environment information,
The estimating unit is constructed using the machine-learned data .

According to this configuration, the estimating unit determines whether or not a person is present around the operating device based on the state information regarding the operation performed over time through the operating device and the environmental information regarding the environmental state during the operation. Estimate and output a human presence signal or a human absence signal. As supervised learning information, this estimation unit is composite information in which human information (teacher information) indicating the presence or absence of a person in the vicinity of the operation device, state information, and environment information are associated with each other over time. is configured as a machine learning unit learned to output a human presence signal or a human absence signal using state information and environmental information as input data. Therefore, switching between the normal power mode and the power saving mode based on the presence or absence of a person is possible even if the operation device or its periphery is not equipped with a human sensor.

In order to generate as easily as possible human information indicating the presence or absence of a person around the operation device, which is used as supervised information in supervised learning information, a human sensor such as an infrared sensor or a face detection sensor is used. is preferably used. Accordingly, in one preferred embodiment of the present invention, the human information is generated based on the detection result of a human sensor that detects the presence or absence of a human around the operating device.

When human information is generated using a human sensor, if the human sensor is incorporated in an operation device that generates state information and environment information, the operation device incorporating this human sensor can generate human information, state information, and environment information. can generate information and Combining these pieces of information together over time can be used as supervised learning information for an estimation unit (machine learning unit). For this reason, in one preferred embodiment of the present invention, the operating device incorporating the human sensor is provided as a human operating device, the human operating device generates the human information, and the composite Generate information.

In the case of a system in which household appliances and their operating devices are installed in many homes, it is possible to replace the operating device installed in a specific home with a human-sensing operating device that is an operating device incorporating a human-sensing sensor. , supervised learning information will be generated through daily operation in the specific household. For this reason, in one preferred embodiment of the present invention, the manipulators or the human-sensitive manipulators located in various places are connected to a management computer through a data communication line so as to be able to exchange data, The computer includes a machine learning section that performs learning of the machine learning unit, and the composite information is sent from the human-sensitive operation device to the machine learning section. Efficient machine learning is realized by network management of the operating devices and human-sensing operating devices installed in each home.

When the operation device and the human operation device are network-managed by the management computer, and the machine learning unit is provided in the management computer, the estimating unit as a result of machine learning may be provided in the management computer or at home It may be provided in the operating device to be installed.

In one of the embodiments in which the management computer is provided with an estimation unit as a learned machine learning unit constructed by a machine learning section, the power mode setting section is provided in the operation device, and the human presence signal or the human An absence signal is sent from the management computer to the power mode setting unit through the data communication line. In other words, the estimating unit installed in the management computer generates a human presence signal or a human absence signal based on the state information and environment information sent from each home's controller through a communication network (data communication line such as the Internet). Output. The output person presence signal or person absence signal is transmitted from the management computer to the corresponding home controller. The operator power mode setting unit sets the power mode to the normal power mode or the power saving mode according to the received human presence signal or human absence signal.

It is also possible to adopt an embodiment in which an estimating unit as a learned machine learning unit is provided in each home operator. In this configuration, each operator is provided with an estimating unit as a machine learning unit that has been learned by the machine learning unit. It outputs a human presence signal or a human absence signal by itself. The power mode setting unit of the operating device sets the power mode to the normal power mode or the power saving mode according to the output human presence signal or human absence signal.

It is difficult to achieve a 100% estimation probability of human presence around the operation device by the estimation unit constructed through the machine learning described above. In order to improve this estimated probability, it is necessary to set additional conditions. It is inconvenient in terms of cost to provide a special sensor or the like to introduce this additional condition. Therefore, in one of the preferred embodiments of the present invention, as a condition for setting the power saving mode, no operation on the operating device for a predetermined time or more is added. Predicting with a high probability that no one is in the vicinity of the operation unit by setting the condition that no operation has continued for a predetermined period of time after being operated once and that the estimation unit is outputting a no-person signal. It becomes possible to

INDUSTRIAL APPLICABILITY The power mode management system according to the present invention is applicable to controllers that operate many household appliances, and the power modes of the controllers are managed for power saving purposes. In particular, a remote control, which is an operation device attached to a domestic fuel cell cogeneration device, has a relatively large liquid crystal display, so application of this power mode management system is convenient. In addition, in home fuel cell cogeneration equipment, the state information and environment information contain data related to the presence of people around the remote control, which is advantageous for increasing the estimation probability of the estimation unit. Therefore, in one preferred embodiment of the present invention, the household appliance is a household fuel cell cogeneration appliance, and the state information includes time-series usage pattern data for hot water supply, time-series usage pattern data for hot water heating, and At least one of usage pattern data and time-series usage pattern data of electricity is included, and the environment information includes time-series pattern data of water temperature, time-series pattern data of air temperature, date data, and time data. contains at least one of

1 is a schematic diagram showing the overall configuration of a power mode management system; FIG. 1 is a schematic diagram showing the overall configuration of a domestic fuel cell cogeneration system; FIG. 1 is a functional block diagram of a power mode management system according to a first embodiment; FIG. FIG. 3 is a schematic diagram illustrating machine learning for an estimation unit configured as a convolutional neural network; FIG. 10 is a functional block diagram of a power mode management system according to a second embodiment; FIG. FIG. 11 is a functional block diagram of a power mode management system in a third embodiment; FIG. It is a functional block diagram of a remote control of another embodiment.

A specific embodiment of the power mode management system according to the present invention will be described with reference to the drawings. In this embodiment, the power mode management system is applied to a domestic fuel cell cogeneration system CS. As shown in FIG. 1, this domestic fuel cell cogeneration system CS consists of a management computer CC installed in a gas service center and a domestic fuel cell cogeneration device 1 installed in each home. and one or more remote controllers 2 as operating devices for operating the household fuel cell cogeneration equipment 1 . As shown in FIG. 2, the home-use fuel cell cogeneration equipment 1 includes a fuel cell unit 12, a hot water storage tank 11, and a heat source device 13 as components. to generate electricity. The heat generated by the fuel cell unit 12 is stored in the hot water storage tank 11 in the form of hot water. The heat source device 13 can heat hot water from the hot water storage tank 11 with combustion heat obtained by burning fuel. The heat source device 13 supplies hot water (for hot water supply, hot water heating, etc.) to the heat load section in the house. In the case of hot water supply, hot water is consumed in the target heat load portion and is not returned to the hot water storage tank 11 . In the case of hot water heating, only the heat of the hot water heated by the combustion heat of the heat source device 13 is consumed by the target heat load portion, and the hot water in the hot water storage tank 11 is not used. Hot water heating includes hot water floor heating, bathroom heater/dryer, and the like. Furthermore, bath reheating may be included in hot water heating. Note that the heat source device 13 can be replaced with a water heater. The remote controller 2 displays various information such as operation buttons for setting drive parameters (on/off of power generation, temperature of hot water in hot water supply, temperature of hot water in hot water heating, etc.), notification lights, drive setting information, drive status information, and date and time. It is equipped with a display for

The management computer CC and the remote controllers 2 in each home are connected through a data communication line DN such as the Internet or a public line. As a result, various types of information are transmitted and received between the management computer CC and the remote controller 2 . The information sent from the remote controller 2 to the management computer CC is status information about operations performed through the remote controller 2 over time and environmental information about the environmental conditions at the time of remote controller operation. The state information includes at least one of time-series usage pattern data for hot water supply, time-series usage pattern data for hot water heating, and time-series usage pattern data for electricity, and the environment information includes time-series pattern data for water temperature. , temperature time series pattern data, date data, and time data. The water temperature time-series pattern data includes the temperature of the water supply, the temperature of the hot water supply, or both. The air temperature time series pattern data includes the outdoor air temperature, the indoor temperature, or both.

The management computer CC has a function of estimating whether or not there is a person around the remote controller 2 based on the status information and environmental conditions sent from the remote controller 2 of each home, and sending back the estimation result to the remote controller 2. have When the remote controller 2 receives the estimation result that no person exists, the power mode of the household fuel cell cogeneration equipment 1 is set to the power saving mode, and the display and the notification lamp are turned off. When receiving the estimation result that a person is present, the power mode changes to the normal power mode, and the display, the notification lamp, etc. are turned on.

Next, the functions provided in the management computer CC and the remote controller 2 and particularly related to the present invention will be described. FIG. 3 shows a functional block diagram of the management computer CC and the remote controller 2. As shown in FIG.

In this embodiment, the remote controller 2 includes a state information generator 21 that generates state information, an environment information generator 22 that generates environment information, and a power mode setter 23 . The power mode setting unit 23 sets the power mode based on the human presence signal or the human absence signal sent from the management computer CC. The human presence signal is a signal indicating that the presence of a person is estimated around the remote controller 2, and the human absence signal is a signal indicating that the absence of a person around the remote controller 2 is estimated. be.

Furthermore, in this embodiment, the household fuel cell cogeneration system CS in a specific household incorporates a human sensor 30 in place of the remote control 2 to operate the household fuel cell cogeneration equipment 1. A remote controller 3 is provided as a human-sensitive operation device. The human-sensitive remote controller 3 detects the presence or absence of a person around the human-sensitive remote controller 3, and generates the detection result as human information. As with the remote controller 2, the motion remote controller 3 includes a state information generator 21, an environment information generator 22, and a power mode setting unit 23. Therefore, the human information is combined with the state information and the environment information. Composite information is created and sent to the management computer CC. In composite information, human information is associated with status information and environment information over time. Therefore, from this combined information, changes in the operating state of the household fuel cell cogeneration equipment 1 over time and changes in the environment over time can be read during the time period when people are present or absent around the human-sensing remote controller 3. It becomes possible.

The management computer CC includes an information storage section 51 , a data preprocessing section 52 , a machine learning section 53 and an estimation unit 4 . The information storage unit 51 stores the state information and the environmental state sent from the remote controller 2 of each home. The data preprocessing unit 52 reads information stored in the information storage unit 51 , performs necessary data processing, and supplies the data to the machine learning unit 53 and the estimation unit 4 .

The estimating unit 4 uses state information and environment information from the remote controller 2 as input data, outputs a human presence signal when the presence of a person around the remote controller 2 is estimated, and detects the presence of a person around the operation device. It has an estimation calculation function that outputs an absence signal when the existence of the person is not estimated. The output human presence signal or human absence signal is sent to the remote controller 2 and used for switching the power mode (normal power mode or power saving mode) by the power mode setting unit 23 . Since the estimation unit 4 is configured as a machine learning unit, a machine learning process is required to build this estimation computation function. In the machine learning process executed by the machine learning unit 53, the composite information sent from the human-sensitive remote controller 3 is used, and the human information included in the composite information sent from the human-sensitive remote controller 3 is used as teacher information. becomes.

The estimation unit model learned by the machine learning unit 53 has substantially the same configuration as the estimation unit (machine learning unit) 4. Here, a neural network model such as a recursive neural network model, a recurrent neural network, An optimal model selected from a convolutional neural network model or the like can be constructed, but a recurrent neural network is advantageous when dealing with time-series data. FIG. 4 shows the configuration of the neural network model, an input layer containing seven neurons w1, w2, w3, . . . , w7 and a first hidden layer containing three neurons x1, x2, x3 (first hidden layer), a second intermediate layer (second hidden layer) containing three neurons y1, y2 and y3, and an output layer containing one neuron z1. In FIG. 4, only the first and second intermediate layers are shown, but in reality, an input layer corresponding to the amount of data to be input is required, and it is preferable that the number of hidden layers also be multi-layered. is.

This neural network receives chronological state information and environmental information as input, and human information (time-series data of the presence or absence of people), which is the detection result of the human sensor 30 among the supervised learning information, is used as supervised information. , the relationship between state information over time, environment information, and human information. Specifically, as state information, a time-series usage pattern for hot water supply is input to neuron w1, a time-series usage pattern for hot water heating is input to neuron w2, and a time-series usage pattern for electricity is input to neuron w3. As environmental information, date data is input to neuron w4, time data is input to neuron w5, time-series pattern data of water temperature is input to neuron w6, and time-series pattern data of air temperature is input to neuron w7.

Each of the time-series patterns described above is conveniently created through statistical calculation from the time-series data read from the state information and the environment information. can be done with

The hidden layers have a first hidden layer (hidden layer) consisting of one or more convolutional layers and a pooling layer, and a second hidden layer consisting of a fully connected layer. The first hidden layer acquires a feature map and abstracts patterns to recognize and classify input time-series patterns.

When the learning process is completed, the machine-learned data containing the final weighting factors in the estimation unit model is generated, and the estimation unit 4 is constructed using this machine-learned data. From the state information and environment information from the remote controller 2 without the human sensor 30, the time-series usage pattern of hot water supply, the time-series usage pattern of hot water heating, the time-series usage pattern of electricity, date data, time data, and water temperature. time-series pattern data of temperature and time-series pattern data of air temperature are created and input to the estimation unit 4, it is estimated whether or not a person is present around the remote controller 2. FIG. A human presence signal is output when it is estimated that a person exists, and a human absence signal is output when it is estimated that a person does not exist. The output human presence signal is sent to the power mode setting section 23 of the remote controller 2 corresponding to the human absence signal, and is used to set the power mode (normal power mode or power saving mode). When the remote controller 2 is set to the power saving mode, the display and notification lamp of the remote controller 2 are turned off.

[Another embodiment]
(1) FIG. 5 shows a functional block diagram of another embodiment of the power mode management system. This embodiment differs from the previous embodiment (see FIG. 3) in that the estimation unit 4 as a learned machine learning unit is provided in the remote control 2 of each home instead of the management computer CC. . By giving the machine-learned data, which is the result of the machine learning process performed using the complex information from the human-sensitive remote controller 3, to the remote controller 2 of the home-use fuel cell cogeneration system CS installed in each home, An estimating unit 4 is constructed in the remote controller 2 . Although it is convenient to use the data communication line DN to transfer the machine-learned data from the management computer CC to the remote controller 2, a data transfer method that mediates a smart phone, a portable memory, or the like may be employed. Also, the remote controller 2 pre-installed with the estimation unit 4 constructed using the machine-learned data may be installed in each home. In the method of transferring the machine-learned data via the data communication line DN, the machine-learned data can be generated periodically and the estimation unit 4 of the remote controller 2 can be updated to the latest version.

(2) FIG. 6 shows a functional block diagram of yet another embodiment of a power mode management system. In this alternative embodiment, the human information required for machine learning is not sent from the domestic fuel cell cogeneration system CS equipped with the human remote controller 3 to the management computer CC, but is installed in a specific home. This differs from the previous embodiment (see FIG. 3) in that it is sent from a human sensor 30 that is independent of the domestic fuel cell cogeneration system CS. The human information sent from this independent human sensor 30 also incorporates time information. Therefore, in the management computer CC, the state information and environment information sent from the remote controller 2 of the specific household are related to the human information sent from the human sensor 30 of the specific household based on the time information, and combined. Information can be generated. Although the estimation unit 4 is constructed in the management computer CC in FIG. 6, it may be constructed in the remote controller 2 as shown in FIG.

(3) In the above-described embodiment, the power mode setting unit 23 sets the power mode to the normal power mode or the power saving mode using the result of estimation of the presence or absence of people around the operation device by the estimation unit 4 as the setting condition. set. Alternatively, additional setting conditions may be set for switching the power mode other than human presence or human absence. For example, in the embodiment shown in FIG. 7, the remote controller 2 is provided with a no-operation determination unit 24 that determines a state in which the remote controller 2 has not been operated for a predetermined time or longer (non-operation state). The no-operation determination unit 24 sets a no-operation flag when determining the no-operation state. The power mode setting unit 23 sets the power mode to the power saving mode on condition that the no-operation flag is set and the unattended signal is output.

(4) In the above-described embodiment, the household fuel cell cogeneration equipment 1 was taken up as a household appliance operated by the remote controller 2 or the human-sensitive remote controller 3 as an operating device. In addition to this, the present invention can be applied to various domestic appliances such as air conditioners and water heaters. The operating device may be of a voice-operated type in addition to a manually operated type using a touch panel or buttons.

It should be noted that the configurations disclosed in the above embodiments (including other embodiments, the same shall apply hereinafter) can be applied in combination with configurations disclosed in other embodiments as long as there is no contradiction. The embodiments disclosed in this specification are exemplifications, and the embodiments of the present invention are not limited thereto, and can be modified as appropriate without departing from the object of the present invention.

INDUSTRIAL APPLICABILITY The present invention is applied to a power mode management system that manages the power mode of a controller that operates household appliances.

1: household fuel cell cogeneration equipment 2: remote controller 3: human-sensing remote controller 4: estimation unit 11: hot water storage tank 12: fuel cell unit 13: heat source device 21: status information generator 22: environment information generator 23: power mode Setting unit 24: No operation determination unit 30: Human sensor 51: Information storage unit 52: Data preprocessing unit 53: Machine learning unit CC: Management computer CS: Home fuel cell cogeneration system DN: Data communication line

Claims (8)

A power mode management system for operating a household appliance and managing the power mode of an operation device without a human sensor ,
a state information generating unit that generates state information regarding the operating state of the home appliance over time caused by the operation of the operation device;
an environmental information generating unit that generates environmental information about the environmental state of the home appliance over time;
Estimation having an estimation operation function that outputs a human presence signal when the presence of a person around the operating device is estimated, and outputs a human absence signal when the presence of a person around the operating device is not estimated. a unit;
a power mode setting unit that sets a power saving mode as the power mode when the human absence signal is output, and sets a normal power mode as the power mode when the human presence signal is output,
The state information includes at least one of time-series usage pattern data for hot water supply, time-series usage pattern data for hot water heating, and time-series usage pattern data for electricity,
The environmental information includes at least one of time-series pattern data of water temperature, time-series pattern data of air temperature, date data, and time data,
Machine learning in which the estimating unit outputs the human presence signal or the human absence signal using the state information and the environment information generated by the operation device having no human sensor as input data. configured as a unit
Presence of a person around another operating device, wherein the machine learning unit generates supervised learning information for the machine learning unit by an operating device other than the operating device that does not have a human sensor . Alternatively, output machine-learned data using composite information indicating the relationship between human information , which is time-series data indicating the absence of a person , and the state information and the environment information,
A power mode management system , wherein the estimating unit is constructed using the machine-learned data . 2. The power mode management system according to claim 1, wherein the human information is generated based on a detection result of a human sensor that detects the presence or absence of a person around the another operation device. 3. The power mode management system according to claim 2, wherein said another operating device incorporating said human sensor is provided as a human operating device, said human operating device generating said human information and generating said composite information. . The different manipulators or the human-sensing manipulators located in various places are connected to a management computer through a data communication line so as to be able to exchange data, and the management computer includes a machine that performs learning of the machine learning unit. 4. The power mode management system of claim 3, wherein a learning unit is constructed and the composite information is sent from the human operator to the machine learning unit. The estimating unit is provided in the management computer, the power mode setting section is provided in the separate operation device, and the human presence signal or the human absence signal is transmitted from the management computer to the power mode setting section through the data communication line. 5. The power mode management system of claim 4, wherein the power mode management system is sent to. 5. The power mode management system according to claim 4, wherein the estimating unit and the power mode setting section are provided in the separate operation device. 7. The power mode management system according to any one of claims 1 to 6, wherein as a condition for setting the power saving mode, no operation is added to the another operation device for a predetermined time or longer. 8. A power mode management system according to any one of claims 1 to 7, wherein said domestic appliance is a domestic fuel cell cogeneration appliance.

Images