Refrigerator and operation energy consumption estimation method thereof

Abstract

The invention discloses a refrigerator and an operation energy consumption estimation method thereof, and the method comprises the steps: obtaining the current operation state data of the refrigerator in the operation process of the refrigerator, determining the current operation mode of the refrigerator, constructing a feature vector corresponding to the operation mode according to the operation state data, and calculating the operation energy consumption of the refrigerator; and inputting the feature vector into an electric energy reasoning model which corresponds to the operation mode and is obtained through deep learning pre-training so as to estimate the electric energy consumption of the refrigerator in the previous electric energy detection period. By adopting the embodiment of the invention, the corresponding feature vector and the electric energy reasoning model can be determined according to the operation mode of the refrigerator, reasoning calculation of the operation energy consumption is realized by using the operation state data generated in the operation process of the refrigerator, and the accuracy of energy consumption estimation is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of refrigerators, and in particular to a refrigerator and a method for estimating operating energy consumption thereof. Background Art

[0002] Refrigerators provide a good low-temperature environment for food. Usually, refrigerators are powered on 24 hours a day, and the electricity they consume accounts for a large proportion of household electricity consumption, so users are also very concerned about the energy consumption of refrigerators. In the prior art, power consumption is measured and calculated by an external power measurement module or by adding voltage and current detection elements to the refrigerator circuit, but this increases the hardware cost. In another scheme, the power is calculated by collecting the instantaneous or average voltage and current of the variable frequency compressor, and then the energy consumption at a certain time interval is calculated. This scheme has the disadvantage of using a point to represent the whole, which will lead to large data deviations, and there is also the problem of missing the energy consumption of components such as fan motors and dew heaters. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a refrigerator and a method for estimating its operating energy consumption, which can determine the corresponding feature vector and electric energy inference model according to the operating mode of the refrigerator, use the operating status data generated during the operation of the refrigerator to realize the inference calculation of the operating energy consumption, and improve the accuracy of energy consumption estimation.

[0004] To achieve the above object, an embodiment of the present invention provides a refrigerator, comprising:

[0005] A box door, arranged at the opening of the box body, for opening and closing the storage room;

[0006] A controller configured to:

[0007] During the operation of the refrigerator, obtaining current operation status data of the refrigerator and determining the current operation mode of the refrigerator;

[0008] Constructing a feature vector corresponding to the operation mode according to the operation status data; wherein the feature vector includes the operation status data of the refrigerator related to the power consumption in the operation mode;

[0009] Acquire a preset electric energy inference model corresponding to the operation mode;

[0010] The feature vector is input into the power inference model to estimate the power consumption of the refrigerator in a previous power detection cycle.

[0011] As an improvement of the above solution, the operating mode is a refrigeration mode, a defrosting mode or a standby mode.

[0012] As an improvement of the above solution, the refrigerator further comprises a refrigeration system and a defrosting system; wherein the refrigeration system comprises a compressor and an evaporator, and the defrosting system comprises a defrosting heater;

[0013] Then determining the current operating mode of the refrigerator includes:

[0014] When the compressor is in a working state, determining that the current operating mode of the refrigerator is a cooling mode;

[0015] When the compressor is not in the working state and the defrost heater is in the working state, determining that the current operating mode of the refrigerator is the defrost mode;

[0016] When both the compressor and the defrost heater are not in the working state, it is determined that the current operation mode of the refrigerator is the standby mode.

[0017] As an improvement of the above solution, the current operating status data of the refrigerator includes: ambient temperature, ambient humidity, set temperature of the storage chamber, actual temperature of the storage chamber and evaporator temperature.

[0018] As an improvement of the above solution, the step of constructing a feature vector corresponding to the operation mode according to the operation status data includes:

[0019] When the refrigerator is in a cooling mode or a standby mode, a first feature vector is constructed, wherein a vector value of the first feature vector includes the ambient temperature, the set temperature of the storage chamber, the actual temperature of the storage chamber, the temperature difference between the set temperature and the actual temperature of the storage chamber, and the operation mode of the refrigerator;

[0020] When the refrigerator is in the defrosting mode, a second feature vector is constructed, and a vector value of the second feature vector includes the ambient temperature, the ambient humidity, the evaporator temperature, and the refrigerator operation mode.

[0021] As an improvement of the above solution, the preset power reasoning model includes a first power reasoning model and a second power reasoning model, the first power reasoning model is trained to estimate the power consumption of the refrigerator in a refrigeration mode or a standby mode, and the second power reasoning model is trained to estimate the power consumption of the refrigerator in a defrosting mode;

[0022] Then, inputting the feature vector into the power inference model to estimate the power consumption of the refrigerator in the previous power detection cycle includes:

[0023] When the refrigerator is in a cooling mode or a standby mode, inputting the constructed first feature vector into the first power inference model to estimate the power consumption of the refrigerator in a previous power detection cycle;

[0024] When the refrigerator is in the defrosting mode, the constructed second feature vector is input into the second power inference model to estimate the power consumption of the refrigerator in the previous power detection cycle.

[0025] As an improvement of the above solution, the refrigerator further comprises a human-computer interaction module, which is arranged outside the box and is used to receive a power check cycle set by a user and display the power consumption of the refrigerator during the power check cycle, wherein the power check cycle duration is an integer multiple of the first power detection cycle duration;

[0026] The controller is further configured to:

[0027] Acquire the electric energy checking cycle received by the human-computer interaction module;

[0028] Determining a plurality of power detection cycles included in the power checking cycle;

[0029] Accumulating the power consumption estimated in the multiple power detection cycles to obtain the total power consumption of the refrigerator in the power check cycle;

[0030] The total power consumption is sent to the human-computer interaction module for display.

[0031] To achieve the above object, an embodiment of the present invention further provides a method for estimating the operating energy consumption of a refrigerator, comprising:

[0032] During the operation of the refrigerator, obtaining current operation status data of the refrigerator and determining the current operation mode of the refrigerator;

[0033] Constructing a feature vector corresponding to the operation mode according to the operation status data; wherein the feature vector includes the operation status data of the refrigerator related to the power consumption in the operation mode;

[0034] Acquire a preset electric energy inference model corresponding to the operation mode;

[0035] The feature vector is input into the power inference model to estimate the power consumption of the refrigerator in a previous power detection cycle.

[0036] As an improvement of the above solution, the operation mode is a refrigeration mode, a defrost mode or a standby mode; the refrigerator further includes a refrigeration system and a defrost system; wherein the refrigeration system includes a compressor and an evaporator, and the defrost system includes a defrost heater; then determining the current operation mode of the refrigerator includes:

[0037] When the compressor is in a working state, determining that the current operating mode of the refrigerator is a cooling mode;

[0038] When the compressor is not in the working state and the defrost heater is in the working state, determining that the current operating mode of the refrigerator is the defrost mode;

[0039] When both the compressor and the defrost heater are not in the working state, it is determined that the current operation mode of the refrigerator is the standby mode.

[0040] As an improvement of the above solution, the step of constructing a feature vector corresponding to the operation mode according to the operation status data includes:

[0041] When the refrigerator is in a cooling mode or a standby mode, a first feature vector is constructed, wherein a vector value of the first feature vector includes the ambient temperature, the set temperature of the storage chamber, the actual temperature of the storage chamber, the temperature difference between the set temperature and the actual temperature of the storage chamber, and the operation mode of the refrigerator;

[0042] When the refrigerator is in the defrosting mode, a second feature vector is constructed, and a vector value of the second feature vector includes the ambient temperature, the ambient humidity, the evaporator temperature, and the refrigerator operation mode.

[0043] Compared with the prior art, the present invention discloses a refrigerator and a method for estimating its operating energy consumption. During the operation of the refrigerator, the current operating status data of the refrigerator is obtained, and the current operating mode of the refrigerator is determined. A feature vector corresponding to the operating mode is constructed according to the operating status data, and the feature vector is then input into an electric energy inference model obtained by deep learning pre-training corresponding to the operating mode, so as to estimate the electric energy consumption of the refrigerator in the previous electric energy detection cycle. By adopting the embodiment of the present invention, the corresponding feature vector and electric energy inference model can be determined according to the operating mode of the refrigerator, and the operating status data generated during the operation of the refrigerator can be used to realize the inference calculation of the operating energy consumption, thereby improving the accuracy of energy consumption estimation, and no external electric energy measurement module is required, thereby reducing hardware costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a schematic diagram of the external structure of a refrigerator provided by an embodiment of the present invention;

[0045] Figure 2 is a schematic diagram of the internal structure of a refrigerator provided by an embodiment of the present invention;

[0046] Figure 3 is a schematic structural diagram of a refrigeration system in a refrigerator provided by an embodiment of the present invention;

[0047] Figure 4 is a first working flow diagram of a controller provided by an embodiment of the present invention;

[0048] Figure 5is a flow chart of a controller determining a refrigerator operation mode provided by an embodiment of the present invention;

[0049] Figure 6 is a second working flow diagram of the controller provided by an embodiment of the present invention;

[0050] Figure 7 is a third working flow diagram of the controller provided in an embodiment of the present invention;

[0051] Figure 8 is a structural diagram of a neural network for training an electric energy inference model provided by an embodiment of the present invention;

[0052] Fig. 9 is a schematic diagram of connection between a controller and a hardware device provided by an embodiment of the present invention;

[0053] Fig.10 is a fourth working flow diagram of the controller provided by an embodiment of the present invention;

[0054] Fig.11 The present invention provides a flow chart of a refrigerator energy consumption estimation method. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0056] In the description of the present application, it should be understood that the terms "center", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0057] The terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, unless otherwise specified, "plurality" means two or more.

[0058] In the description of this application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0059] See also Figure 1 , Figure 1 1 is a schematic diagram of the external structure of a refrigerator 100 provided in an embodiment of the present invention. The refrigerator 100 of this embodiment is approximately rectangular in shape. The refrigerator includes a box body defining a storage space and a plurality of door bodies arranged at the opening of the box body, wherein the door body includes a door body shell located outside the box body, a door body liner located inside the box body, an upper end cover, a lower end cover, and an insulation layer located between the door body shell, the door body liner, the upper end cover, and the lower end cover; usually, the insulation layer is filled with foam. The box body is provided with a chamber, wherein the chamber includes a component storage chamber for placing components in the refrigerator, such as a press cabin, etc., and also includes a storage space for storing food, etc.

[0060] See also Figure 2 , Figure 2 : is a schematic diagram of the internal structure of a refrigerator provided by an embodiment of the present invention. The storage space can be divided into multiple storage rooms. The storage rooms can be configured as refrigerators and freezers according to different uses. They can also include variable temperature rooms, vacuum drawers, moisturizing drawers, etc. Each storage room corresponds to one or more door bodies, for example, Figure 1 The upper storage room is provided with a double-opening door body. The door body can be pivotally arranged at the opening of the box body, and can also be opened in a drawer-like manner to achieve drawer-like storage. The refrigerator door is provided with a display screen, which is used to display prompt information and receive user touch operations.

[0061] See also Figure 3 , Figure 3The schematic diagram of the structure of the refrigeration system in the refrigerator provided by the embodiment of the present invention, the refrigeration system includes a compressor 1, an evaporator 2, a drying filter (not shown in the figure), a capillary tube 3, a condenser 4 and a gas-liquid separator (not shown in the figure). The working process of the refrigeration system includes a compression process, a condensation process, a throttling process and an evaporation process. The compression process is: plug in the power cord of the refrigerator, when the contacts of the thermostat are connected, the compressor 1 starts to work, and the low-temperature and low-pressure refrigerant is sucked into the compressor 1, and is compressed into a high-temperature and high-pressure superheated gas in the cylinder of the compressor 1 and then discharged into the condenser 4; the condensation process is: the high-temperature and high-pressure refrigerant gas dissipates heat through the condenser 4, the temperature continues to drop, and is gradually cooled to a saturated vapor of normal temperature and high pressure, and is further cooled to a saturated liquid, and the temperature no longer drops. The temperature at this time is called the condensation temperature, and the pressure of the refrigerant remains almost unchanged during the entire condensation process; the throttling process is as follows ... The process is as follows: the condensed refrigerant saturated liquid is filtered through a drying filter to remove moisture and impurities and then flows into the capillary 3, through which it is throttled and depressurized, and the refrigerant becomes wet steam at room temperature and low pressure; the evaporation process is as follows: the wet steam at room temperature and low pressure begins to absorb heat and vaporize in the evaporator 2, which not only reduces the temperature of the evaporator 2 and its surroundings, but also turns the refrigerant into a low-temperature, low-pressure gas. The refrigerant coming out of the evaporator 2 passes through the gas-liquid separator and returns to the compressor 1 again, repeating the above process to transfer the heat in the refrigerator to the air outside the box, thereby achieving the purpose of refrigeration.

[0062] The refrigerator provided in the embodiment of the present invention also includes a defrost system, which includes a defrost heater, and the defrost heater is arranged at the lower end of the evaporator 2. After the compressor 1 has been working for a period of time (about 8-10 hours), frost will form on the surface of the evaporator 2. If defrosting is not performed, the frost will become thicker and thicker, which will cause the cold air to be unable to circulate and the refrigeration effect to decrease. After the refrigerator has been working for a period of time (about 8-10 hours), the defrost system automatically cuts off the power supply of the compressor 1 and turns on the defrost heater. The defrost heater heats and melts the frost formed on the evaporator 2. After defrosting is completed, the defrost system cuts off the power supply of the defrost heater and turns on the power supply of the compressor 1 to resume refrigeration, and the cycle repeats itself.

[0063] Specifically, the controller of the refrigerator is configured to: during the operation of the refrigerator, obtain the current operating status data of the refrigerator and determine the current operating mode of the refrigerator; construct a feature vector corresponding to the operating mode based on the operating status data; wherein the feature vector includes operating status data of the refrigerator related to power consumption in the operating mode; obtain a preset power inference model corresponding to the operating mode; and input the feature vector into the power inference model to estimate the power consumption of the refrigerator in the previous power detection cycle.

[0064] For example, see Figure 4 , Figure 4 is a first working flow chart of a controller provided in an embodiment of the present invention, wherein the controller is configured to execute steps S11 to S14:

[0065] S11, obtaining the current operating status data of the refrigerator and determining the current operating mode of the refrigerator, and then proceeding to step S12;

[0066] S12, constructing a feature vector corresponding to the operation mode, and then proceeding to step S13;

[0067] S13, obtaining a preset electric energy inference model corresponding to the operation mode, and then proceeding to step S14;

[0068] S14. Input the feature vector into the power inference model to estimate the power consumption in the previous power detection cycle.

[0069] Since the power and electric energy consumption of refrigerators in different operating modes are different, the embodiment of the present invention periodically obtains refrigerator operating status data during the operation of the refrigerator, determines the refrigerator operating mode, filters out data related to electric energy consumption that matches the current operating mode from the refrigerator operating status data, constructs a feature vector, and inputs the feature vector into the corresponding pre-trained electric energy inference model to complete the estimation of the electric energy consumption of the previous electric energy detection cycle. The present invention achieves accurate estimation of electric energy consumption in different operating modes by constructing different feature vectors for different operating modes and inputting them into the corresponding electric energy inference model.

[0070] As a preferred implementation, the operation mode is a refrigeration mode, a defrosting mode or a standby mode.

[0071] For example, see Figure 5 , Figure 5 It is a flowchart of a controller provided in an embodiment of the present invention to determine the operation mode of a refrigerator, and the controller is configured to execute steps S111 to S116. The operation mode of the refrigerator is divided into three categories according to the load operation condition of the refrigerator, namely, refrigeration mode, defrost mode and standby mode. First, determine whether the refrigerator compressor 1 is in working state. If so, it can be determined that the current operation mode of the refrigerator is refrigeration mode. If not, continue to determine whether the refrigerator defrost heater is in working state. If the defrost heater is in working state, it can be determined that the current operation mode of the refrigerator is defrost mode. If not, it means that the compressor 1 and the defrost heater are both on standby at this time, and the refrigerator operation mode is standby mode at this time; after the judgment is completed, save the current refrigerator operation mode for subsequent data processing and calculation.

[0072] As a preferred embodiment, the current operating state data of the refrigerator is obtained by obtaining sensor data provided inside or outside the refrigerator and obtaining the set temperature of the storage chamber set by the user. Specifically, a temperature sensor and a humidity sensor are provided outside the cabinet to obtain the ambient temperature and the ambient humidity respectively; a temperature sensor is provided inside the storage chamber to obtain the actual temperature of the storage chamber; and a temperature sensor is provided on the evaporator to obtain the temperature of the evaporator. Furthermore, when a plurality of storage chambers are formed in the cabinet, a temperature sensor is provided in each storage chamber.

[0073] It should be noted that, in actual applications, the current operating status data of the refrigerator can be obtained by reading the data of the refrigerator main control board without installing additional hardware equipment such as sensors on the refrigerator, thereby reducing the hardware cost in the process of estimating the refrigerator energy consumption. It is only necessary to deploy an electric energy inference model pre-trained by deep learning in the controller, and the operating status data generated during the operation of the refrigerator can be used to more accurately estimate the operating energy consumption.

[0074] For example, see Figure 6 , Figure 6 is a second working flow chart of the controller provided by an embodiment of the present invention. After executing step S11, the controller is also used to execute steps S121 to S123, and after executing steps S121 to S123, continue to execute steps S13 to S14. When constructing the feature vector corresponding to the operating mode according to the refrigerator operating state data, it is determined whether the refrigerator is in the cooling mode or the standby mode. If so, a first feature vector is constructed; if not, a second feature vector is constructed. Specifically, the vector value of the first feature vector includes the refrigerator operating mode, the set temperature of the storage room, the ambient temperature obtained by the sensor, the actual temperature of the storage room, and the calculated temperature difference between the set temperature of the storage room and the actual temperature of the storage room. Further, when the refrigerator is provided with multiple storage rooms, the vector value of the first feature vector includes the set temperature, actual temperature and temperature difference data between the set temperature and the actual temperature of each of the multiple storage rooms. Specifically, the vector value of the second feature vector includes the ambient temperature, ambient humidity, evaporator temperature and the refrigerator operating mode obtained by the sensor.

[0075] It should be noted that, in the embodiment of the present invention, since the operating status data related to the refrigerator operating energy consumption in the refrigeration mode and the standby mode are the same, the feature vectors corresponding to the refrigeration mode and the standby mode are both first feature vectors, but in these two modes, the degree of influence of the operating status data on the refrigerator operating energy consumption is different, so the refrigerator operating mode is included in the vector value of the first feature vector. In practical applications, since the vector values ​​in the feature vector are all digital values, different refrigerator operating modes can be added to the feature vector by assigning different digital values ​​to them.

[0076] For example, see Figure 7 , Figure 7 is the third working flow chart of the controller provided by the embodiment of the present invention, after executing step S11, the controller executes the judgment step S121, and according to the judgment result, sequentially executes step S122, step S131, step S141 or sequentially executes step S123, step S132, step S142, and then executes step S143. Specifically, the first power reasoning model is trained to estimate the power consumption of the refrigerator in the cooling mode or the standby mode, and the second power reasoning model is trained to estimate the power consumption of the refrigerator in the defrosting mode. Furthermore, when it is determined that the operating mode of the refrigerator is a cooling mode or a standby mode, the controller obtains a first power inference model from the preset power inference model, and inputs a first feature vector into the first power inference model to obtain the power consumption of the refrigerator in the previous power detection cycle in the cooling mode or the standby mode; when it is determined that the operating mode of the refrigerator is a standby mode (that is, not a cooling mode or a defrost mode), the controller obtains a second power inference model from the preset power inference model, and inputs a second feature vector into the second power inference model to obtain the power consumption of the refrigerator in the previous power detection cycle in the standby mode.

[0077] As an example, the refrigerator body has two storage rooms, namely the first storage room and the second storage room. For an electric energy detection cycle starting at the i-th moment, the operating status data of the refrigerator is obtained and the operating mode of the refrigerator is determined at the i-th moment. If the refrigerator compressor is working, the refrigerator operating mode Si at the i-th moment is 1. If the refrigerator compressor is not working but the refrigerator defrost heater is working, the refrigerator operating mode Si at the i-th moment is 2. If neither the refrigerator compressor nor the defrost heater is working, the refrigerator operating mode Si at the i-th moment is 0. The controller judges Si, and when Si=1 or Si=0, constructs the first feature vector at the i-th moment. The vector value of the first eigenvector includes the ambient temperature, the set temperature of the first storage chamber, the actual temperature of the first storage chamber, the set temperature of the second storage chamber, the actual temperature of the second storage chamber, the temperature difference between the set temperature and the actual temperature of the first storage chamber, the temperature difference between the set temperature and the actual temperature of the second storage chamber, and the refrigerator operation mode Si. Then, the first eigenvector at the i-th moment is The output value of the first power reasoning model is the estimated value of the running energy consumption ΔE in the power detection cycle at the starting time being the i-th time. i ; When Si = 2, construct the second eigenvector at the i-th moment The vector value of the second eigenvector includes the ambient temperature, ambient humidity, evaporator temperature and refrigerator operation mode Si. Then, the second eigenvector at the i-th moment The output value of the second power inference model is the estimated value of the running energy consumption ΔE in the power detection cycle at the starting time being the i-th time. i .

[0078] Furthermore, in one embodiment, the training method of the electric energy inference model includes:

[0079] Collect training data, and collect refrigerator operating status data and energy consumption ΔE consumed by the refrigerator during the Δt period according to the power detection cycle Δt=60s. The collected operating status data includes ambient temperature, ambient humidity, set temperature of the storage room, actual temperature of the storage room and evaporator temperature.

[0080] Determine the working state Si of the refrigerator at the i-th moment, and assign values ​​according to the following rules: when the refrigerator is in the refrigeration state, Si=1; when the refrigerator is in the defrosting state, Si=2; when the refrigerator is in the standby state, Si=0.

[0081] The feature vector is constructed according to the state of the refrigerator. When the refrigerator working state Si=0 or Si=1, the first feature vector is constructed; when the refrigerator working state Si=2, the second feature vector is constructed. The first feature vector X (1) The vector value of includes the ambient temperature collected at the current moment, the set temperature of the storage room, the actual temperature of the storage room, the temperature difference between the set temperature and the actual temperature of the storage room, and the refrigerator operation mode. The second eigenvector X (2) The vector value includes the ambient temperature, ambient humidity, evaporator temperature and refrigerator operation mode collected at the current moment.

[0082] The constructed feature vector is paired with the power consumption ΔE detected in the current power detection cycle to construct a set of training data {X, ΔE}, and this is repeated to continuously collect N sets of data. In one embodiment, N=20,000.

[0083] The N groups of training data are divided according to the refrigerator operation mode. The data in the refrigeration mode or standby mode is divided into D1, and the data in the defrosting mode is divided into D2. In one embodiment, N1=19000, data N2=1000, data

[0084] Normalize the training data. For each data in D1 and D2 above, normalize it according to the following formula:

[0085]

[0086] Among them, xmin represents the minimum value of data x, and xmax represents the maximum value of data x.

[0087] To build a neural network, see Figure 8 , Figure 8 is a structural diagram of a neural network for training an electric energy inference model provided by an embodiment of the present invention, wherein the neural network consists of an input layer (vector A [0] ), 3 hidden layers (vector A [1] ~A [3] ), output layer (vector A [4] ); Specifically, in one embodiment, the feature vector of the input layer of the first power inference model is the first feature vector X (1) Hidden layer A [1] Contains 10 neurons (a [1] 1 ,a [1] 2 ,a [1] 3 ,…a [1] 10 ,), hidden layer A [2] Contains 36 neurons (a [2] 1 ,a [2] 2 ,a [2] 3 ,…a [2] 36 ,), hidden layer A [3] Contains 10 neurons (a [3] 1 ,a [3] 2 ,a [3] 3 ,…a [3] 10 ,); the output layer contains 1 neuron (a [4]), the energy consumed by the neuron and refrigerator during the power detection cycle Δt Corresponding to: the equation of the first electric energy reasoning model The eigenvector of the input layer of the second power inference model is the second eigenvector X (2) Hidden layer A [1] Contains 10 neurons (a [1] 1 ,a [1] 2 ,a [1] 3 ,…a [1] 10 ,), hidden layer A [2] Contains 36 neurons (a [2] 1 ,a [2] 2 ,a [2] 3 ,…a [2] 36 ,), hidden layer A [3] Contains 10 neurons (a [3] 1 ,a [3] 2 ,a [3] 3 ,…a [3] 10 ,); the output layer contains 1 neuron (a [4] ), the energy consumed by the neuron and refrigerator during the power detection period Δt Corresponding to; the equation of the model Q2 express.

[0088] Initialize the parameters of the neural network. In one embodiment, the neural network model architecture torch.nn.Module of the pytorch architecture is adopted but not limited to, and its loss function loss = torch.nn.MSELoss() is defined, and its optimizer optimizer = torch.optim.SGD() is defined.

[0089] Input data X into the model to calculate the predicted value Calculate the deviation value using the function Calculate the deviation between the predicted value and the actual value, and reversely derive the deviation value loss.backward(); perform reverse iterative optimization, define its optimizer optimizer.step(), and then reset the optimizer's gradient optimizer.Zero_grad().

[0090] The above process of inputting the measured data into the model to generate a predicted value and calculating the deviation value for reverse iterative optimization is continuously repeated until the deviation value is less than a preset value, and the training of the electric energy inference model is completed.

[0091] It should be noted that during the training of the power reasoning model, an external power measurement module is connected to the power input terminal of the refrigerator to measure the accurate energy consumption of the refrigerator during operation, in order to obtain a more accurate power reasoning model. After the power reasoning model training is completed, the model can be deployed to the refrigerator controller for the controller to obtain and use. In other words, during the actual operation of the refrigerator, there is no need to connect an external power measurement module, and the operating energy consumption can be estimated by using the existing operating status data of the refrigerator.

[0092] Furthermore, the refrigerator is also provided with a human-computer interaction module 101, and the human-computer interaction module is arranged outside the refrigerator. Fig. 9 , Fig. 9 1 is a schematic diagram of the connection between the controller and the hardware device provided in an embodiment of the present invention. The controller in the embodiment of the present invention is connected to the human-computer interaction module 101. The controller can receive the power viewing cycle set by the user, and can also control the display of the power consumption of the refrigerator in the power viewing cycle on the human-computer interaction module. The power viewing cycle duration is an integer multiple of the first power detection cycle duration. Specifically, see Fig.10 , Fig.10 It is the fourth working flow chart of the controller provided in an embodiment of the present invention. The controller is also configured to execute steps S21 to S24. When receiving an energy viewing cycle from a human-computer interaction module, the total energy consumption of the refrigerator in the energy viewing cycle can be calculated based on multiple energy detection cycles included in the energy viewing cycle and the estimated energy consumption in each energy detection cycle is accumulated. The total energy consumption is then sent to the human-computer interaction module for display.

[0093] As an example, the controller receives from the human-computer interaction module an energy viewing cycle of one day, and the energy detection cycle of the refrigerator Δt=60s, that is, the energy viewing cycle includes 1440 consecutive energy detection cycles. The controller then accumulates the estimated energy consumption under these 1440 energy detection cycles, thereby obtaining the total energy consumption of the refrigerator in one day, and then sends the total energy consumption to the human-computer interaction module for display.

[0094] Compared with the prior art, the refrigerator disclosed in the present invention collects the operating status data of the refrigerator to construct a feature vector corresponding to its operating mode, and then inputs the feature vector into a pre-trained power inference model to estimate the power consumption of the previous power detection cycle. By adopting the embodiment of the present invention, the corresponding feature vector and power inference model can be determined according to the operating mode of the refrigerator, and the operating status data generated during the operation of the refrigerator can be used to realize the inference calculation of the operating energy consumption, thereby improving the accuracy of energy consumption estimation, and no external power measurement module is required, thereby reducing hardware costs.

[0095] See also Fig.11 , Fig.11 : is a flow chart of a refrigerator operation energy consumption estimation method provided by an embodiment of the present invention. The refrigerator operation energy consumption estimation method according to the embodiment of the present invention is executed by a controller in the refrigerator. The refrigerator operation energy consumption estimation method includes steps S31 to S34:

[0096] S31, during the operation of the refrigerator, obtaining current operation status data of the refrigerator and determining the current operation mode of the refrigerator;

[0097] S32, constructing a feature vector corresponding to the operation mode according to the operation status data; wherein the feature vector includes the operation status data of the refrigerator related to the power consumption in the operation mode;

[0098] S33, obtaining a preset electric energy inference model corresponding to the operation mode;

[0099] S34, inputting the feature vector into the power inference model to estimate the power consumption of the refrigerator in the previous power detection cycle.

[0100] It should be noted that, in actual applications, the current operating status data of the refrigerator can be obtained by reading the data of the refrigerator main control board without installing additional hardware equipment such as sensors on the refrigerator, thereby reducing the hardware cost in the process of estimating the refrigerator energy consumption. It is only necessary to deploy an electric energy inference model pre-trained by deep learning in the controller, and the operating status data generated during the operation of the refrigerator can be used to more accurately estimate the operating energy consumption.

[0101] Exemplarily, the refrigerator operation mode includes a refrigeration mode, a defrost mode, and a standby mode. When the refrigerator compressor is in operation, it is determined that the current operation mode of the refrigerator is the refrigeration mode; when the refrigerator compressor is not in operation and the refrigerator defrost heater is in operation, it is determined that the current operation mode of the refrigerator is the defrost mode; when both the refrigerator compressor and the refrigerator defrost heater are not in operation, it is determined that the current operation mode of the refrigerator is the standby mode.

[0102] As a preferred embodiment, the current operating state data of the refrigerator is obtained by obtaining sensor data provided inside or outside the refrigerator and obtaining the set temperature of the storage room set by the user. Specifically, a temperature sensor and a humidity sensor are provided outside the refrigerator body, respectively for obtaining the ambient temperature and the ambient humidity; a temperature sensor is provided inside the storage room, for obtaining the actual temperature of the storage room; and a temperature sensor is provided on the evaporator, for obtaining the temperature of the evaporator. Furthermore, when a plurality of storage rooms are formed in the box body, a temperature sensor is provided in each storage room.

[0103] It should be noted that, in actual applications, the current operating status data of the refrigerator can be obtained by reading the data of the refrigerator main control board without installing additional hardware equipment such as sensors on the refrigerator, thereby reducing the hardware cost in the process of estimating the refrigerator energy consumption. It is only necessary to deploy an electric energy inference model pre-trained by deep learning in the controller, and the operating status data generated during the operation of the refrigerator can be used to more accurately estimate the operating energy consumption.

[0104] Exemplarily, constructing a feature vector corresponding to the operating mode based on the operating status data includes: when the refrigerator is in a refrigeration mode or a standby mode, constructing a first feature vector, the vector value of which includes the ambient temperature, the set temperature of the storage chamber, the actual temperature of the storage chamber, the temperature difference between the set temperature and the actual temperature of the storage chamber, and the refrigerator operating mode; when the refrigerator is in a defrost mode, constructing a second feature vector, the vector value of which includes the ambient temperature, the ambient humidity, the evaporator temperature, and the refrigerator operating mode.

[0105] It should be noted that, in the embodiment of the present invention, since the operating status data related to the refrigerator operating energy consumption in the refrigeration mode and the standby mode are the same, the feature vectors corresponding to the refrigeration mode and the standby mode are both first feature vectors, but in these two modes, the degree of influence of the operating status data on the refrigerator operating energy consumption is different, so the refrigerator operating mode is included in the vector value of the first feature vector. In practical applications, since the vector values ​​in the feature vector are all digital values, different refrigerator operating modes can be added to the feature vector by assigning different digital values ​​to them.

[0106] Specifically, the first power reasoning model is trained to estimate the power consumption of the refrigerator in the refrigeration mode or the standby mode, and the second power reasoning model is trained to estimate the power consumption of the refrigerator in the defrost mode. Further, when it is determined that the operation mode of the refrigerator is the refrigeration mode or the standby mode, the controller obtains the first power reasoning model from the preset power reasoning model, and inputs the first feature vector into the first power reasoning model to obtain the power consumption of the refrigerator in the previous power detection cycle in the refrigeration mode or the standby mode; when it is determined that the operation mode of the refrigerator is the standby mode (that is, not the refrigeration mode or the defrost mode), the controller obtains the second power reasoning model from the preset power reasoning model, and inputs the second feature vector into the second power reasoning model to obtain the power consumption of the refrigerator in the previous power detection cycle in the standby mode.

[0107] As a preferred embodiment, a human-computer interaction module is also provided on the outside of the refrigerator body, and the method for estimating the energy consumption of the refrigerator operation also includes: when receiving an energy viewing cycle from the human-computer interaction module, based on multiple energy detection cycles contained in the energy viewing cycle, and accumulating the estimated energy consumption in each energy detection cycle, the total energy consumption of the refrigerator in the energy viewing cycle can be calculated, and then the total energy consumption is sent to the human-computer interaction module for display.

[0108] As an example, the controller receives from the human-computer interaction module an energy viewing cycle of one day, and the energy detection cycle of the refrigerator Δt=60s, that is, the energy viewing cycle includes 1440 consecutive energy detection cycles. The controller then accumulates the estimated energy consumption under these 1440 energy detection cycles, thereby obtaining the total energy consumption of the refrigerator in one day, and then sends the total energy consumption to the human-computer interaction module for display.

[0109] Compared with the prior art, the refrigerator operation energy consumption estimation method disclosed in the present invention collects the operation status data of the refrigerator to construct a feature vector corresponding to its operation mode, and then inputs the feature vector into a pre-trained power inference model to estimate the power consumption of the previous power detection cycle. By adopting the embodiment of the present invention, the corresponding feature vector and power inference model can be determined according to the operation mode of the refrigerator, and the operation status data generated during the operation of the refrigerator can be used to realize the inference calculation of the operation energy consumption, thereby improving the accuracy of energy consumption estimation, and no external power measurement module is required, thereby reducing hardware costs.

[0110] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A refrigerator, characterized in that: include: a box body, in which at least one storage chamber is formed for storing articles; A box door, arranged at the opening of the box body, for opening and closing the storage room; A controller configured to: During the operation of the refrigerator, obtaining current operation status data of the refrigerator and determining the current operation mode of the refrigerator; Constructing a feature vector corresponding to the operation mode according to the operation status data; wherein the feature vector includes the operation status data of the refrigerator related to the power consumption in the operation mode; Acquire a preset electric energy inference model corresponding to the operation mode; The feature vector is input into the power inference model to estimate the power consumption of the refrigerator in a previous power detection cycle.

2. The refrigerator according to claim 1, characterized in that: The operation mode is a refrigeration mode, a defrosting mode or a standby mode.

3. The refrigerator according to claim 2, characterized in that: The refrigerator further comprises a refrigeration system and a defrosting system; wherein the refrigeration system comprises a compressor and an evaporator, and the defrosting system comprises a defrosting heater; Then determining the current operating mode of the refrigerator includes: When the compressor is in a working state, determining that the current operating mode of the refrigerator is a cooling mode; When the compressor is not in the working state and the defrost heater is in the working state, determining that the current operating mode of the refrigerator is the defrost mode; When both the compressor and the defrost heater are not in the working state, it is determined that the current operation mode of the refrigerator is the standby mode.

4. The refrigerator according to claim 3, characterized in that: The current operating status data of the refrigerator includes: ambient temperature, ambient humidity, set temperature of the storage chamber, actual temperature of the storage chamber and evaporator temperature.

5. The refrigerator according to claim 4, characterized in that: The constructing a feature vector corresponding to the operation mode according to the operation status data includes: When the refrigerator is in a cooling mode or a standby mode, a first feature vector is constructed, wherein a vector value of the first feature vector includes the ambient temperature, the set temperature of the storage chamber, the actual temperature of the storage chamber, the temperature difference between the set temperature and the actual temperature of the storage chamber, and the operation mode of the refrigerator; When the refrigerator is in the defrosting mode, a second feature vector is constructed, and a vector value of the second feature vector includes the ambient temperature, the ambient humidity, the evaporator temperature, and the refrigerator operation mode.

6. The refrigerator according to claim 5, characterized in that: The preset power reasoning model includes a first power reasoning model and a second power reasoning model, wherein the first power reasoning model is trained to estimate the power consumption of the refrigerator in a refrigeration mode or a standby mode, and the second power reasoning model is trained to estimate the power consumption of the refrigerator in a defrosting mode; Then, inputting the feature vector into the power inference model to estimate the power consumption of the refrigerator in the previous power detection cycle includes: When the refrigerator is in a cooling mode or a standby mode, inputting the constructed first feature vector into the first power inference model to estimate the power consumption of the refrigerator in a previous power detection cycle; When the refrigerator is in the defrosting mode, the constructed second feature vector is input into the second power inference model to estimate the power consumption of the refrigerator in the previous power detection cycle.

7. The refrigerator according to claim 1, characterized in that: The refrigerator further comprises a human-computer interaction module, which is arranged outside the box and is used to receive a power check cycle set by a user and display the power consumption of the refrigerator during the power check cycle, wherein the power check cycle duration is an integer multiple of the first power detection cycle duration; The controller is further configured to: Acquire the electric energy checking cycle received by the human-computer interaction module; Determining a plurality of power detection cycles included in the power checking cycle; Accumulating the power consumption estimated in the multiple power detection cycles to obtain the total power consumption of the refrigerator in the power check cycle; The total power consumption is sent to the human-computer interaction module for display.

8. A method for estimating the operating energy consumption of a refrigerator, characterized in that: include: During the operation of the refrigerator, obtaining current operation status data of the refrigerator and determining the current operation mode of the refrigerator; Constructing a feature vector corresponding to the operation mode according to the operation status data; wherein the feature vector includes the operation status data of the refrigerator related to the power consumption in the operation mode; Acquire a preset electric energy inference model corresponding to the operation mode; The feature vector is input into the power inference model to estimate the power consumption of the refrigerator in a previous power detection cycle.

9. A method for estimating the operating energy consumption of a refrigerator as claimed in claim 8, characterized in that: The operation mode is a refrigeration mode, a defrost mode or a standby mode; the refrigerator further includes a refrigeration system and a defrost system; wherein the refrigeration system includes a compressor and an evaporator, and the defrost system includes a defrost heater; then determining the current operation mode of the refrigerator includes: When the compressor is in a working state, determining that the current operating mode of the refrigerator is a cooling mode; When the compressor is not in the working state and the defrost heater is in the working state, determining that the current operating mode of the refrigerator is the defrost mode; When both the compressor and the defrost heater are not in the working state, it is determined that the current operation mode of the refrigerator is the standby mode.

10. A refrigerator operation energy consumption estimation method according to claim 9, characterized in that: The constructing a feature vector corresponding to the operation mode according to the operation status data includes: When the refrigerator is in a cooling mode or a standby mode, a first feature vector is constructed, wherein a vector value of the first feature vector includes the ambient temperature, the set temperature of the storage chamber, the actual temperature of the storage chamber, the temperature difference between the set temperature and the actual temperature of the storage chamber, and the operation mode of the refrigerator; When the refrigerator is in the defrosting mode, a second feature vector is constructed, and a vector value of the second feature vector includes the ambient temperature, the ambient humidity, the evaporator temperature, and the refrigerator operation mode.

Images