KR20240038184A - Wahsing machine and operating method thereof - Google Patents

Abstract

A washing machine is provided that can provide an optimized dosage of laundry detergent and fabric softener. A washing machine according to an embodiment of the present invention acquires laundry information about laundry, the total amount of laundry, washing machine information, and detergent information, and obtains the laundry information obtained using a detergent input amount inference model, the total amount of laundry, and The input amount of laundry detergent and fabric softener can be inferred from the washing machine information and the detergent information, and the inferred input amount of laundry detergent and fabric softener can be supplied.

Description

Washing machine and its operating method {WAHSING MACHINE AND OPERATING METHOD THEREOF}

The present invention relates to a washing machine, and more specifically, to a washing machine capable of providing information on the amount of detergent input.

In general, a washing machine is a device that removes contaminants from contaminated laundry through processes such as washing, rinsing, and dehydration according to a set algorithm.

Specifically, when the washing machine receives a washing command, it supplies washing water up to a preset water level and removes contaminants through the emulsification of the water flow and detergent generated when the washing machine or pulsator rotates.

When washing of the laundry is completed, the washing machine drains the washing water and then supplies the washing water again to rinse the laundry. When the rinsing cycle is completed, the washing machine drains the washing water again and then rotates the washing machine to dehydrate the laundry.

Previous patent literature related to this includes Korean Patent Publication No. 10-2015-0000628.

Previous patent literature discloses a washing machine equipped with a detergent supply device and a control method thereof that ensures that an appropriate amount of detergent is always supplied.

However, the washing machine disclosed in the prior patent literature may not be able to remove contamination well depending on the material of the laundry, and if the degree of contamination is severe, more detergent may need to be added.

In addition, in the prior patent literature, there are various factors that can affect the amount of detergent input, such as the characteristics of the laundry and the degree of contamination of the laundry, but these factors are not reflected in determining the amount of detergent input.

In addition, conventional prior patent literature assumes that turbidity changes depending on the amount of detergent input, but does not consider that the degree of contamination of laundry may affect turbidity, and the corrected amount of detergent input may not be accurate.

The purpose of the present invention is to provide a washing machine that can predict an accurate amount of detergent by reflecting all information on individual laundry.

The purpose of the present invention is to reduce the amount of calculation of the model by simplifying the information input to the model for inferring the input amount of detergent.

The purpose of the present invention is to utilize a machine learning algorithm to determine the amount of detergent input according to washing conditions.

The purpose of the present invention is to infer the detergent input amount and recommended washing course according to washing conditions using a machine learning algorithm.

A washing machine according to an embodiment of the present invention acquires laundry information about laundry, the total amount of laundry, washing machine information, and detergent information, and obtains the laundry information obtained using a detergent input amount inference model, the total amount of laundry, and The input amount of laundry detergent and fabric softener can be inferred from the washing machine information and the detergent information, and the inferred input amount of laundry detergent and fabric softener can be supplied.

The washing machine according to an embodiment of the present invention stores a clothing combination table in the memory that represents the combination of the type of clothing and the material of the clothing as a type value, and receives the type of clothing and the material of the clothing through user input. In one case, the type value can be extracted through the clothing combination table.

The training data set for supervised learning of the detergent input amount inference model includes learning data including laundry data for learning, data on the amount of laundry for learning, data on the washing machine, and detergent data, and detergent input amount data labeled in the learning data. And, the detergent input amount inference model may be supervised to minimize the difference between the inference result data of the detergent input amount inference model and the detergent input amount data.

The training data set for supervised learning of the tax input inference model is

Contains learning data including laundry data for learning, data about the amount of laundry for learning, data about a washing machine, and detergent data, and labeling data including detergent input amount data and recommended course data labeled in the learning data, and the detergent input amount. The inference model may be supervised to learn to minimize the difference between the inference result data of the detergent input amount inference model and the labeling data.

According to an embodiment of the present invention, the required amount of laundry detergent and fabric softener can be accurately predicted by reflecting the washing conditions that change each time of washing. Accordingly, users do not need to worry about the amount of laundry detergent and fabric softener used.

According to an embodiment of the present invention, the inference speed for the detergent input amount of the model can be improved by simplifying the data input to the model.

According to an embodiment of the present invention, the input amount of laundry detergent and fabric softener can be accurately predicted according to a machine learning algorithm. As a result, only the minimum amount of laundry detergent and fabric softener is used, which reduces waste of laundry detergent and fabric softener compared to before. In addition, since the optimal amount of laundry detergent and fabric softener are supplied, contamination of the washing machine can be prevented by preventing accumulation in the inner and outer tanks of the washing machine.

According to an embodiment of the present invention, an optimal amount of laundry detergent, an optimal amount of fabric softener, and an optimal washing course can be provided according to a machine learning algorithm.

1 is a perspective view of a washing machine according to an embodiment of the present invention.
Figure 2 is a block diagram explaining the configuration of a washing machine according to an embodiment of the present invention.
Figure 3 is a diagram explaining the configuration of an artificial intelligence server according to an embodiment of the present invention.
Figure 4 is a flowchart for explaining a method of operating a washing machine according to an embodiment of the present invention.
Figure 5 is a diagram explaining a detergent input amount inference model according to an embodiment of the present invention.
Figure 6 is a diagram explaining a detergent input amount inference model according to another embodiment of the present invention.
7 to 10 are diagrams to specifically explain information input to the detergent input amount inference model of the present invention.
Figure 11 is a sequence diagram for explaining the operation method of an artificial intelligence system according to an embodiment of the present invention.

Figure 1 is a perspective view of a washing machine according to an embodiment of the present invention, and Figure 2 is a block diagram explaining the configuration of a washing machine according to an embodiment of the present invention.

Referring to FIG. 1, the washing machine 100 according to an embodiment of the present invention includes a cabinet 101 forming the exterior, a front cover 103 mounted on the front of the cabinet 101 and having a laundry entrance 102, and a laundry cloth. It may include a drum 107 and a tub (not shown) that accommodates the drum 107 and stores washing water therein.

Additionally, the washing machine 100 may further include a motor 162 that provides rotational power to the drum 107.

The drum 107 can be understood as an “inner tank,” and the tub can be understood as an “outer tank.” Alternatively, the drum 107 may be understood as a “washing tank” and the tub may be understood as a “dewatering tank.”

The cabinet 101 may be provided in a substantially hexahedral shape.

Inside the cabinet 101 of the washing machine 100, a space may be formed for installing a number of parts.

A number of parts include, for example, a drum 107, a tub, a motor 162, and a control part that controls the motor 162, a communication unit 110, an input unit 120, and a learning processor 130, which will be described in FIG. 2. , may include a sensing unit 140, a memory 170, and a processor 180.

A laundry entrance 102 may be formed in the front cover 103.

The laundry entrance 102 may be formed approximately at the center of the front cover 103. Additionally, a door 108 that opens and closes the laundry entrance 102 may be rotatably installed on the front cover 103.

A gasket (not shown) is provided between the door 108 and the tub to maintain airtightness.

The washing machine 100 may further include a control panel 104 provided at the top of the front of the cabinet 101.

The control panel 104 may include a display to display the operating status of the washing machine 100. The control panel 104 may be provided with a plurality of buttons or knobs for manipulating the operation of the washing machine 100.

The control panel 104 may be included in the user input unit 123 of FIG. 1 or the display unit 151 of FIG. 2.

The washing machine 100 may further include a detergent drawer 105 provided at the upper front of the cabinet 101.

The detergent drawer 105 may be referred to as a detergent supply unit.

The detergent drawer 105 may be provided on the side of the control panel 104.

The detergent drawer 105 may have a shape in which a part where detergent is introduced and stored and a part exposed to the front are integrated.

The detergent drawer 105 may be connected to a water supply unit 163 that supplies cold/hot water.

Cold/hot water may flow into the detergent drawer 105 from the water supply unit 163.

Water mixed with at least one of detergent and fabric softener in the detergent drawer 105 may be supplied to the inside of the drum 107 where laundry cloth is accommodated via the tub.

The washing machine 100 may further include a service cover 106 provided at the front bottom of the cabinet 101.

The service cover 106 is configured to be opened when the washing machine 100 is stopped to remove residual water present in the washing machine 100.

In the drum 107, there is a washing process in which dirt on clothes and bedding (hereinafter referred to as 'cloth') is separated by the action of detergent and water, a rinsing process in which the cloth is rinsed by the action of water, and a dehydration process in which the cloth is centrifugally dehydrated. It can be accomplished.

The drum 107 is provided in a cylindrical shape and is accommodated inside the tub. As an example, the drum 107 is formed in a cylindrical shape lying at a predetermined angle, and a hole (not shown) may be formed around it.

Accordingly, the washing water stored in the tub may flow into the drum 107 through a water hole. Alternatively, the washing water inside the drum 107 can be moved to the outside of the drum 107 through water.

The drum 107 may be equipped with a pulsator (not shown) that induces the flow of washing water inside the drum 107. The washing machine 100 may have a pulsator installed inside the drum 107 and a motor that directly rotates the drum 107 may be installed at the rear end of the tub.

The washing machine 100 is equipped with a pulsator inside the drum 107, and is also equipped with a motor and a power transmission mechanism such as a clutch that transmits the driving force of the motor to the pulsator or drum 107 at the rear end of the tub. possible.

Meanwhile, the tub may contain washing water for washing or rinsing. The tub is provided to accommodate the drum 107. As an example, the tub may be formed in a cylindrical shape.

The tub can be installed as if hanging on the cabinet 101 or the front cover 103. And the drum 107 disposed inside the tub can be rotated by the rotational force generated by the motor.

Additionally, the washing machine 100 may further include a water supply unit 163 for supplying wash water to the tub 161 and a drain unit 165 for draining wash water from the tub 161.

The water supply unit 163 may include a water supply hose connected to the outside (e.g., a faucet, etc.), and a water supply valve that controls water to be supplied or not supplied to the tub 161.

The drain unit 165 may include a drain valve that controls water in the tub 161 to be drained or not drained, and a drain hose that extends outward from the drain valve to drain water to the outside.

Figure 2 is a block diagram explaining the configuration of a washing machine according to an embodiment of the present invention.

Referring to FIG. 2, the washing machine 100 includes a communication unit 110, an input unit 120, a learning processor 130, a sensing unit 140, a memory 170, a washing module 160, and a processor 180. can do.

The communication unit 110 can transmit and receive data with external devices such as other terminals or AI servers using wired and wireless communication technology.

For example, the communication unit 110 may transmit and receive sensor information, user input, learning models, and control signals with external devices.

Communication technologies used by the communication unit 110 include Global System for Mobile communication (GSM), Code Division Multi Access (CDMA), Long Term Evolution (LTE), 5G, Wireless LAN (WLAN), Wireless-Fidelity (Wi-Fi), These include Bluetooth™, RFID (Radio Frequency Identification), Infrared Data Association (IrDA), ZigBee, and NFC (Near Field Communication).

The input unit 120 can acquire various types of data.

The input unit 120 may include a camera 121 for inputting video signals, a microphone 122 for receiving audio signals, and a user input unit 123 for receiving information from the user. there is.

The camera 121 processes image frames such as still images or moving images obtained by an image sensor in video call mode or shooting mode.

The processed image frame may be displayed on the display unit (151) or stored in the memory (170).

The microphone 122 processes external acoustic signals into electrical voice data. The processed voice data can be used in various ways depending on the function (or application program being executed) being performed by the washing machine 100.

Various noise removal algorithms may be applied to the microphone 122 to remove noise generated in the process of receiving an external acoustic signal.

The user input unit 123 is for receiving information from the user. When information is input through the user input unit 123, the processor 180 can control the operation of the washing machine 100 to correspond to the input information.

The user input unit 123 is a mechanical input means (or a mechanical key, such as a button, dome switch, jog wheel, jog switch, etc. located on the front/rear or side of the washing machine 100) and It may include a touch input means.

As an example, the touch input means consists of a virtual key, soft key, or visual key displayed on the touch screen through software processing, or on a part other than the touch screen. It can be done with a touch key placed.

The input unit 120 may acquire training data for model learning and input data to be used when obtaining an output using the learning model.

The input unit 120 may acquire unprocessed input data, and in this case, the processor 180 or the learning processor 130 may extract input features by preprocessing the input data.

The learning processor 130 can train a model composed of an artificial neural network using training data. Here, the learned artificial neural network may be referred to as a learning model.

A learning model can be used to infer a result value for new input data other than learning data, and the inferred value can be used as the basis for a decision to perform an operation.

The learning processor 130 may be implemented using the memory 170, an external memory directly coupled to the washing machine 100, or a memory maintained in an external device.

The sensing unit 140 may use various sensors to obtain at least one of internal information of the washing machine 100, information about the surrounding environment of the washing machine 100, and user information.

Sensors included in the sensing unit 140 include a proximity sensor, illuminance sensor, acceleration sensor, magnetic sensor, gyro sensor, inertial sensor, RGB sensor, IR sensor, fingerprint recognition sensor, ultrasonic sensor, optical sensor, microphone, lidar, and radar. etc.

The output unit 150 includes at least one of a display unit (151), a sound output unit (152), a haptic module (153), and an optical output unit (154). can do.

The display unit 151 displays (outputs) information processed by the washing machine 100. For example, the display unit 151 may display execution screen information of an application program running in the washing machine 100, or UI (User Interface) and GUI (Graphic User Interface) information according to this execution screen information.

The display unit 151 can implement a touch screen by forming a layered structure or being integrated with the touch sensor.

The touch screen functions as a user input unit 123 that provides an input interface between the washing machine 100 and the user, and can simultaneously provide an output interface between the washing machine 100 and the user. The control panel 33 in FIG. 1 may be implemented as a touch screen.

The audio output unit 152 may output audio data received from the communication unit 110 or stored in the memory 170 in call signal reception, call mode or recording mode, voice recognition mode, broadcast reception mode, etc.

The sound output unit 152 may include at least one of a receiver, a speaker, and a buzzer.

The haptic module 153 generates various tactile effects that the user can feel. A representative example of a tactile effect generated by the haptic module 153 may be vibration.

The optical output unit 154 uses light from the light source of the washing machine 100 to output a signal to notify that an event has occurred.

Examples of events that occur in the washing machine 100 may include receiving a message, receiving a call signal, a missed call, an alarm, a schedule notification, receiving an email, receiving information through an application, etc.

The memory 170 may store data supporting various functions of the washing machine 100.

The memory 170 may store input data, learning data, learning models, learning history, etc. obtained from the input unit 120.

The laundry module 160 includes a tub 161, a drum 107, a motor 162, a water supply unit 163, a pump 164, and a drain unit 165.

The tub 161 can store washing water.

The drum 107 is rotatably installed in the tub 161 so that laundry can be loaded into it.

The motor 162 may transmit power to rotate the drum 107 to the drum 107 .

The water supply unit 163 can supply washing water.

The pump 164 can circulate or discharge washing water.

The drain unit 165 can drain discharged washing water.

The processor 180 may determine at least one executable operation of the washing machine 100 based on information determined or generated using a data analysis algorithm or a machine learning algorithm.

The processor 180 may control the components of the washing machine 100 to perform the determined operation.

To this end, the processor 180 may request, retrieve, receive, or utilize data from the learning processor 130 or the memory 170, and execute a predicted operation or an operation determined to be desirable among at least one executable operation. Components of the washing machine 100 can be controlled to do so.

If linkage with an external device is necessary to perform a determined operation, the processor 180 may generate a control signal to control the external device and transmit the generated control signal to the external device.

The processor 180 may obtain intent information regarding user input and determine the user's request based on the obtained intent information.

The processor 180 uses at least one of a STT (Speech To Text) engine for converting voice input into a character string or a Natural Language Processing (NLP) engine for acquiring intent information of natural language, to input user input. Corresponding intent information can be obtained.

At this time, at least one of the STT engine or the NLP engine may be composed of at least a portion of an artificial neural network learned according to a machine learning algorithm. And, at least one of the STT engine or the NLP engine is learned by the learning processor 130, learned by the learning processor 240 of the AI server 200, or learned by distributed processing thereof. It could be.

The processor 180 collects history information including the user's feedback on the operation of the washing machine 100 and stores it in the memory 170 or the learning processor 130, or stores it in an external device such as the AI server 200. It can be sent to the device. The collected historical information can be used to update the learning model.

The processor 180 may control at least some of the components of the washing machine 100 to run an application program stored in the memory 170. Furthermore, the processor 180 may operate two or more of the components included in the washing machine 100 in combination with each other to run an application program.

Artificial Neural Network (ANN) is a model used in machine learning. It can refer to an overall model with problem-solving capabilities that is composed of artificial neurons (nodes) that form a network through the combination of synapses. Artificial neural networks can be defined by connection patterns between neurons in different layers, a learning process that updates model parameters, and an activation function that generates output values.

An artificial neural network may include an input layer, an output layer, and optionally one or more hidden layers. Each layer includes one or more neurons, and the artificial neural network may include synapses connecting neurons.

In an artificial neural network, each neuron can output the activation function value for the input signals, weight, and bias input through the synapse.

Model parameters refer to parameters determined through learning and include the weight of synaptic connections and the bias of neurons.

Hyperparameters refer to parameters that must be set before learning in a machine learning algorithm and include learning rate, number of repetitions, mini-batch size, initialization function, etc.

The purpose of artificial neural network learning can be seen as determining model parameters that minimize the loss function. The loss function can be used as an indicator to determine optimal model parameters in the learning process of an artificial neural network.

Machine learning can be classified into supervised learning, unsupervised learning, and reinforcement learning depending on the learning method.

Supervised learning refers to a method of training an artificial neural network with a given label for the learning data. A label refers to the correct answer (or result value) that the artificial neural network must infer when learning data is input to the artificial neural network. It can mean.

Unsupervised learning can refer to a method of training an artificial neural network in a state where no labels for training data are given.

Reinforcement learning can refer to a learning method in which an agent defined within an environment learns to select an action or action sequence that maximizes the cumulative reward in each state.

Among artificial neural networks, machine learning implemented with a deep neural network (DNN) that includes multiple hidden layers is also called deep learning, and deep learning is a part of machine learning. Hereinafter, machine learning is used to include deep learning.

Figure 3 is a diagram explaining the configuration of an artificial intelligence server according to an embodiment of the present invention.

The AI server 200 may refer to a device that trains an artificial neural network using a machine learning algorithm or uses a learned artificial neural network.

The AI server 200 may include a communication unit 210, a memory 230, a learning processor 240, and a processor 260.

The communication unit 210 can transmit and receive data with an external device such as the washing machine 100.

Memory 230 may include a model storage unit 231. The model storage unit 231 may store a model (or artificial neural network, 231a) that is being trained or has been learned through the learning processor 240.

The learning processor 240 can train the artificial neural network 231a using training data.

The learning model may be used while mounted on the AI server 200 of the artificial neural network, or may be mounted and used on an external device such as a washing machine 100.

Learning models can be implemented in hardware, software, or a combination of hardware and software.

When part or all of the learning model is implemented as software, one or more instructions constituting the learning model may be stored in the memory 230.

The processor 260 may infer a result value for new input data using a learning model and generate a response or control command based on the inferred result value.

Figure 4 is a flowchart for explaining a method of operating a washing machine according to an embodiment of the present invention.

Referring to FIG. 4, the processor 180 of the washing machine 100 obtains laundry information for each of a plurality of laundry items (S401).

In one embodiment, when the laundry is clothing, the laundry information may include clothing information.

Clothing information may include one or more of the type of clothing, the material of the clothing, the degree of clothing contamination, and the source of clothing contamination.

In one embodiment, the processor 180 may obtain clothing information through the control panel 104. That is, the processor 180 may receive user input through the control panel 104 and obtain clothing information corresponding to the received user input.

In another embodiment, the processor 180 may obtain clothing information through a mobile terminal such as a user's smart phone. A device control application for controlling the washing machine 100 may be installed on the mobile terminal.

The mobile terminal may receive clothing information through a device control application and transmit the received clothing information to the washing machine 100 through a wireless communication interface.

The processor 180 obtains the total amount of laundry (S403).

When a plurality of laundry items are placed in the drum 107, the processor 180 can measure the total amount of laundry using a weight sensor (not shown) provided in the drum 107.

The processor 180 obtains washing machine information (S405).

Washing machine information may include one or more of the type of washing machine 100, capacity of washing machine 100, washing course, and wash water temperature.

The type of washing machine 100 may be either a bucket type or a drum type.

The capacity of the washing machine 100 may be the maximum amount of laundry that can be accommodated in the drum 107 of the washing machine 100.

The washing course may be a course selected by user input.

In one implementation, the processor 180 may obtain washing machine information through the control panel 104.

In another embodiment, the processor 180 may receive washing machine information from a mobile terminal. The mobile terminal may receive user information through a device control application and transmit the received information to the communication unit 110 of the washing machine 100.

The processor 180 obtains laundry detergent information and fabric softener information (S407).

Detergent information may include laundry detergent information and fabric softener information.

Laundry detergent information may be a standard input amount of laundry detergent.

The standard dosage of laundry detergent may represent the recommended standard dosage provided by the manufacturer of the laundry detergent.

The fabric softener information may be the standard dosage of fabric softener.

The standard dosage of fabric softener may be the recommended standard usage information provided by the manufacturer of the fabric softener.

The processor 180 uses the detergent input amount inference model to infer the input amount of laundry detergent and fabric softener from laundry information, weight of all laundry, washing machine information, and detergent information (S409).

In one embodiment, the detergent input amount inference model may be an artificial neural network-based model supervised through machine learning.

The detergent input amount inference model can be modeled as a deep neural network (DNN) consisting of multiple hidden layers between the input layer and the output layer.

The detergent input amount inference model may be a model that infers the input amount of laundry detergent and fabric softener to be input into the washing machine 100 from laundry information, total amount of laundry, washing machine information, and detergent information.

The training data set for learning the detergent input amount inference model may include learning data and labeling data.

Learning data may include laundry data, total amount of laundry data, washing machine data, and detergent data.

In one embodiment, the labeling data may include input amount data of laundry detergent and input amount data of fabric softener.

In another embodiment, the labeling data may include laundry detergent dosage data, fabric softener dosage data, and recommended course data.

The learning method of the tax input inference model will be described in detail later.

In one embodiment, the detergent input amount inference model may be a model stored in the memory 170 of the washing machine 100.

In another embodiment, the detergent input amount inference model may be stored in the model storage unit 231 of the AI server 200.

If the detergent input amount inference model is stored in the model storage unit 231 of the AI server 200, the washing machine 100 can transmit laundry information, total laundry amount, washing machine information, and detergent information to the AI server 200. there is.

The AI server 200 may obtain the inferred detergent input amount information by inputting laundry information, total amount of laundry, washing machine information, and detergent information into the detergent input amount inference model.

The AI server 200 may transmit the obtained detergent input amount information to the washing machine 100.

The processor 180 controls the detergent drawer 105 to supply the inferred amount of laundry detergent and fabric softener to the drum 107 (S411).

The detergent drawer 105 may include a dispenser (not shown) that can supply each of laundry detergent and fabric softener to the drum 107.

The processor 180 may transmit a control signal to the dispenser of the detergent drawer 105 to supply the input amount of laundry detergent and fabric softener to the drum 107, which are inferred through the detergent input amount inference model.

In one embodiment, the processor 180 may display information about the amount of detergent input through the display 13 provided in the control panel 104. Information about the amount of detergent input may include the amount of laundry detergent and fabric softener to be input in the course selected by the user, and the predicted contamination level of the washing machine.

In another embodiment, the processor 180 may output information about the amount of detergent input as a voice through the audio output unit 152.

Figure 5 is a diagram explaining a detergent input amount inference model according to an embodiment of the present invention.

Referring to FIG. 5, an artificial neural network-based detergent input amount inference model 500 is shown.

The detergent input amount inference model 500 may include an input layer, a hidden layer, and an output layer.

Each of the input layer, hidden layer, and output layer may be composed of a plurality of units, and each unit may be composed of one perceptron.

Each unit can have input values and weights and output output values through an activation function. Each unit on two adjacent floors is connected to each other through a weighted connection line and has a one-to-one correspondence.

The artificial neural network of the detergent input amount inference model 500 can multiply input values with weights and add all the results. The weighted sum value can be input into an activation function and converted into the final output. The activation function can convert a mathematical linear combination of input values into various types of non-linear (or linear) combinations.

Since the output result of the detergent input amount inference model 500 can be expressed in numbers, the activation function used may be a linear function.

Laundry information, total amount of laundry, washing machine information, and detergent information may be input to each of the plurality of units included in the input layer of the detergent input amount inference model 500.

The amount of laundry detergent and the amount of fabric softener may be output through each of the plurality of units included in the output layer of the detergent input amount inference model 500.

The training data set for learning the detergent input amount inference model 500 includes learning data in the same format as the input information, including laundry information, total amount of laundry, washing machine information, and detergent information, and labeling data in the same format as the detergent information. can do.

Learning data used for supervised learning of the detergent input amount inference model 500 may be labeled with detergent input amount data (laundry detergent input amount and fabric softener input amount).

The detergent input amount inference model 500 may be trained for the purpose of accurately inferring detergent input amount data from learning data in the same format as input information including laundry information, total laundry amount, washing machine information, and detergent information.

The loss function (cost function) of the detergent input amount inference model 500 is expressed as the square mean of the difference between labeling data representing detergent input amount data corresponding to each learning data and detergent input amount inference data inferred from the learning data. You can.

In the detergent input amount inference model 500, model parameters (eg, weights) included in the artificial neural network may be determined to minimize the loss function through learning.

When an input feature vector is extracted from the learning data and input, the decision result for the detergent input amount data is output as a target feature vector, and the detergent input amount inference model 500 determines the difference between the output target feature vector and the labeled detergent input amount data. It may be learned to minimize the loss function corresponding to .

The processor 180 inputs laundry information, total amount of laundry, washing machine information, and detergent information into the learned detergent input amount inference model 500, and detergent input information including the inferred laundry detergent input amount and fabric softener input amount. can be obtained.

Meanwhile, the user does not need to input all values of independent variables constituting laundry information, total amount of laundry, washing machine information, and detergent information. Users can selectively enter only the values of key variables they want.

The processor 180 may replace the value of an uninput variable with a standard value (or average value).

Figure 6 is a diagram explaining a detergent input amount inference model according to another embodiment of the present invention.

Referring to FIG. 6, an artificial neural network-based detergent input amount inference model 600 according to another embodiment of the present invention is shown.

The detergent input amount inference model 600 of FIG. 6 may be a model in which the input layer unit is reduced by one and the output layer unit is increased by one compared to the detergent input amount inference model 500 of FIG. 5 .

Specifically, the washing machine information of the detergent input amount inference model 600 of FIG. 6 may include only the washing machine type, washing machine capacity, and wash water temperature. In other words, the number of units in the input layer that receives the laundry course input may be reduced.

Additionally, a unit representing a recommended course may be added to the output layer of the detergent input amount inference model 600 compared to the output layer of FIG. 5 .

The recommended course may represent a recommended washing course for a plurality of laundry items to be placed in the drum 107.

Learning data other than the course data used for supervised learning of the detergent input amount inference model 600 may be labeled with detergent input amount data and recommended course data.

The detergent input amount inference model 600 is instructed to accurately infer the amount of laundry detergent, amount of fabric softener, and recommended course from learning data in the same format as laundry information, total amount of laundry, washing machine information (excluding course information), and detergent information. It can be learned.

The detergent input amount inference model 600 of FIG. 6 can output a recommended course to further optimize the amount of laundry detergent and fabric softener.

7 to 10 are diagrams to specifically explain information input to the detergent input amount inference model of the present invention.

Information input to the detergent input amount inference model of FIGS. 7 to 10 may be input through the control panel 104 of the washing machine 100 or through a device control application provided on a mobile terminal.

The input information can be normalized and converted into a vector.

The number of input units constituting the input layer of the detergent input quantity inference model may be set sufficiently high to correspond to the maximum number of clothes that can actually be input into the drum 107. For example, the maximum number of clothes that can be input may be 100, but this is only an example.

Referring to FIG. 7, each piece of detailed information constituting information about one piece of clothing may correspond to an input unit. Clothing information in FIG. 7 may include the type of clothing and the material of the clothing.

If there is no input value in the input layer, the processor 180 replaces the input value with 0 and performs dummy processing.

Regardless of the number and combination of input clothing, one algorithm design can be used, and as the number of input clothing decreases, the amount of algorithm calculation can be reduced.

For example, when two pieces of clothing are input, information about the two pieces of clothing is normally input into the input units of the detergent input quantity inference model, and dummy values (e.g., 0) can be input into the remaining input units. there is.

According to another embodiment of the present invention, clothing information may be obtained by photographing a clothing tag using the camera 121.

The processor 180 may obtain clothing information through analysis of images for clothing tags.

The processor 180 can recognize text through optical character recognition (OCR) from the image of the tag and obtain clothing information from the text.

Next, FIG. 8 will be described.

Referring to FIG. 8, a clothing combination table 800 is shown showing combinations between clothing types and materials of each clothing type.

The type of clothing may be any of T-shirts, dress shirts, sweaters, or jeans.

The clothing material may be any one of cotton, linen, silk, or wool.

The clothing combination table 800 can be used to reduce the number of input units per individual clothing by integrating the type of clothing and the material of the clothing into one type value.

The clothing combination table 800 may be displayed through the control panel 104. In this case, the washing machine 100 may receive a user input for selecting one type value included in the clothing combination table 800.

The number of input units is reduced by the number of clothes put into the washing machine 100 through the clothing combination table 800, and the number of weights connected to the hidden layer is reduced, thereby significantly reducing the computational amount of the detergent input amount inference model.

In another embodiment, when the processor 180 receives information about the type of clothing and the material of the clothing through a user input, the processor 180 combines the type of clothing and the material through the clothing combination table 800 stored in the memory 170. The corresponding type value can be automatically extracted.

Next, FIG. 9 will be described.

Referring to FIG. 9, it is a diagram illustrating washing machine information in detail.

The type of washing machine 100 may be either a bucket type or a drum type.

The capacity of the washing machine 100 can be entered in units of weight.

Each type and capacity of the washing machine 100 can be received through the control panel 104.

The processor 180 may receive a laundry course through the control panel 104.

Processor 180 may receive wash water temperature through a temperature selection table displayed on control panel 104.

The type of washing machine 100, the capacity of the washing machine 100, the washing course, and the washing water temperature included in the washing machine information can each be converted into vector form and input into the input unit of the detergent input amount inference model.

Next, FIG. 10 will be described.

Referring to FIG. 10 , the control panel 104 may receive a standard input amount of laundry detergent or a standard input amount of fabric softener through user input.

The standard dosage of laundry detergent may represent the recommended standard dosage provided by the manufacturer of the laundry detergent.

The standard dosage of fabric softener may represent the recommended standard dosage provided by the manufacturer of the fabric softener.

The standard input amount of laundry detergent and the standard input amount of fabric softener may each be input into the input unit of the input layer.

Figure 11 is a sequence diagram for explaining the operation method of an artificial intelligence system according to an embodiment of the present invention.

The artificial intelligence system may include a washing machine 100 and an AI server 200.

The embodiment of FIG. 11 is an example where the detergent input amount inference model is stored in the AI server 200, unlike the embodiment of FIG. 5 or FIG. 6.

The processor 180 of the washing machine 100 obtains laundry information, total amount of laundry, washing machine information, and detergent information (S1101).

The description and acquisition process of each information is replaced with the embodiment of FIG. 4.

The processor 180 of the washing machine 100 transmits laundry information, total amount of laundry, washing machine information, and detergent information to the AI server 200 through the communication unit 110 (S1103).

The processor 260 of the AI server 200 determines the input amount of laundry detergent and fabric softener from laundry information, total amount of laundry, washing machine information, and detergent information through the detergent input amount inference model stored in the model storage unit 231. Obtain (S1105).

The detergent input amount inference model may be a model according to the embodiment of FIG. 5.

The processor 260 of the AI server 200 transmits the obtained input amount of laundry detergent and fabric softener to the washing machine 100 through the communication unit 210 (S1107).

The degree of clothing shrinkage may include one or more of the size of the clothing after drying and the area ratio compared to the area of the clothing before and after drying.

The processor 180 of the washing machine 100 controls the detergent drawer 105 to supply the input amount of laundry detergent and fabric softener to the drum 107 (S1109).

The present disclosure described above can be implemented as computer-readable code on a program-recorded medium. Computer-readable media includes all types of recording devices that store data that can be read by a computer system. Examples of computer-readable media include HDD (Hard Disk Drive), SSD (Solid State Disk), SDD (Silicon Disk Drive), ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc. There is. Additionally, the computer may include the processor 180 of the washing machine.

Claims (15)

In the washing machine,
detergent supply department;
A laundry module that washes laundry;
A memory that stores a supervised learning detergent input amount inference model based on an artificial neural network; and
Obtain laundry information, total amount of laundry, washing machine information, and detergent information for laundry, and obtain laundry detergent from the laundry information, amount of total laundry, washing machine information, and detergent information obtained using the detergent input amount inference model. Including a processor for inferring the input amount of laundry detergent and the input amount of fabric softener, and controlling the detergent supply unit so that the inferred input amount of laundry detergent and the input amount of fabric softener are supplied.
washing machine. According to paragraph 1,
The above laundry information is
Includes one or more of the following: type of clothing, material of clothing, degree of contamination of clothing, or source of contamination of clothing
washing machine. According to paragraph 2,
The memory is
Storing a clothing combination table representing a combination of the type of clothing and the material of the clothing as a type value,
The processor is
When the type of clothing and the material of the clothing are received through user input, the type value is extracted through the clothing combination table.
washing machine. According to paragraph 1,
The washing machine information is
Containing one or more of the type of the washing machine, the capacity of the washing machine, the washing course selected by the user, and the washing water temperature.
washing machine. According to paragraph 1,
The tax information above is
Including the recommended standard usage amount provided by the manufacturer of the laundry detergent and the recommended standard usage amount provided by the manufacturer of the fabric softener.
washing machine. According to paragraph 1,
The training data set for supervised learning of the tax input inference model is
Contains learning data including laundry data for learning, data on the amount of laundry for learning, data on the washing machine, and detergent data, and detergent input amount data labeled in the learning data,
The detergent input amount inference model is supervised to minimize the difference between the inference result data of the detergent input amount inference model and the detergent input amount data.
washing machine. According to paragraph 1,
The training data set for supervised learning of the tax input inference model is
Contains learning data including laundry data for learning, data on the amount of laundry for learning, data on the washing machine, and detergent data, and labeling data including detergent input amount data and recommended course data labeled in the learning data,
The detergent input amount inference model is supervised to minimize the difference between the inference result data of the detergent input amount inference model and the labeling data.
washing machine. According to paragraph 1,
Further comprising a control panel for inputting the laundry information, the total amount of laundry, the washing machine information, and the detergent information.
washing machine. In the operating method of the washing machine,
Obtaining laundry information, total amount of laundry, washing machine information, and detergent information for laundry;
Inferring the input amount of laundry detergent and fabric softener from the laundry information, the total amount of laundry, the washing machine information, and the detergent information obtained using a detergent input amount inference model supervised and learned based on an artificial neural network; and
Comprising the step of supplying the inferred amount of laundry detergent and the amount of fabric softener.
How a washing machine works. According to clause 9,
The above laundry information is
Includes one or more of the following: type of clothing, material of clothing, degree of contamination of clothing, or source of contamination of clothing
How a washing machine works. According to clause 9,
storing a clothing combination table representing a combination of the type of clothing and the material of the clothing as a type value;
Receiving the type of clothing and the material of the clothing through user input; and
Further comprising extracting the type value through the clothing combination table.
How a washing machine works. According to clause 9,
The washing machine information is
Containing one or more of the type of the washing machine, the capacity of the washing machine, the washing course selected by the user, and the washing water temperature.
How a washing machine works. According to clause 9,
The tax information above is
Including the recommended standard usage amount provided by the manufacturer of the laundry detergent and the recommended standard usage amount provided by the manufacturer of the fabric softener.
How a washing machine works. According to clause 9,
The training data set for supervised learning of the tax input inference model is
Contains learning data including laundry data for learning, data on the amount of laundry for learning, data on the washing machine, and detergent data, and detergent input amount data labeled in the learning data,
The detergent input amount inference model is supervised to minimize the difference between the inference result data of the detergent input amount inference model and the detergent input amount data.
How a washing machine works. According to clause 14,
The training data set for supervised learning of the tax input inference model is
Contains learning data including laundry data for learning, data on the amount of laundry for learning, data on the washing machine, and detergent data, and labeling data including detergent input amount data and recommended course data labeled in the learning data,
The detergent input amount inference model is supervised to minimize the difference between the inference result data of the detergent input amount inference model and the labeling data.
How a washing machine works.