Stove

Abstract

The invention provides a kitchen range, relates to the technical field of intelligent household appliances, and can accurately estimate the temperature in a pot arranged on the kitchen range. A cooker is placed on the cooker, the cooker comprises a controller, and the controller is configured to obtain a first data set. The first data set comprises the volume of liquid in the cookware, the gear of the stove, the temperature value of the bottom of the cookware every first preset duration in the heating process and the heating time at the current moment. And processing the obtained first data set to obtain a second data set. The second data set comprises the normalized volume of the liquid in the cooker, the cooker bottom temperature value at the current moment, the cooker bottom temperature difference value, the heating time at the current moment and the total power corresponding to the current moment. Wherein the pot bottom temperature difference value is the difference value between the current pot bottom temperature difference value and the pot bottom temperature value before the set time. And inputting the second data into a pre-trained temperature prediction model of a full-connection neural network to obtain a predicted value of the temperature in the pot at the current moment.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent household appliances, and in particular to a stove. Background Art

[0002] As living standards continue to improve, people are paying more and more attention to the pursuit of health. When cooking, people often want to achieve precise control of temperature so as to cook healthy ingredients. However, in general, people judge the temperature in the pot by vision and touch, which largely depends on experience and is not accurate. Summary of the invention

[0003] An embodiment of the present invention provides a stove, which can estimate the temperature inside a pot of a pot arranged on the stove more accurately.

[0004] To achieve the above object, the embodiments of the present invention adopt the following technical solutions:

[0005] A cooker provided by the present invention is placed on the cooker, and the cooker includes a controller, which is configured to obtain a first data group. The first data group includes the volume of liquid inside the cooker, the gear position of the cooker, the temperature value of the bottom of the pot at intervals of a first preset time during the heating process, and the heating time at the current moment. The first data group obtained is processed to obtain a second data group. The second data group includes the normalized volume of liquid inside the cooker, the normalized bottom temperature value at the current moment, the normalized bottom temperature difference value, the normalized heating time at the current moment, and the normalized total power corresponding to the current moment. Among them, the bottom temperature difference value is the difference between the current bottom temperature difference value and the bottom temperature value before the set time. The second data is input into a temperature prediction model of a pre-trained fully connected neural network to obtain a predicted value of the temperature inside the cooker at the current moment.

[0006] By building and training a temperature prediction model in advance and writing it into the controller of the stove, the stove can estimate the temperature value in the pot corresponding to the current moment through the volume of liquid in the pot, the gear position of the stove, the temperature value of the bottom of the pot at every first preset time during the heating process, and the heating time at the current moment. Since the data of the volume of liquid in the pot, the gear position of the stove, the temperature value of the bottom of the pot at every first preset time during the heating process, and the heating time at the current moment are easy to obtain, the estimation of the temperature value in the pot is more convenient and accurate.

[0007] As a possible implementation, the controller is configured to process the acquired first data group to obtain a second data group. Specifically, it is configured to perform feature extraction on the first data group to obtain an intermediate data group. Each data in the intermediate data group is normalized to obtain a second data group. Feature extraction is performed on the first data group to obtain an intermediate data group, including obtaining the pot bottom temperature value at the current moment, the heating time at the current moment, and the liquid volume inside the pot. According to the pot bottom temperature value at the current moment and the pot bottom temperature value before the set time, the pot bottom temperature difference is obtained. According to the stove gear during the heating process, the total power corresponding to the current moment is calculated according to the following formula.

[0008]

[0009] Where w is the total power corresponding to the current moment, T is the heating time at the current moment, and w t is the heating power at time t, and γ is a constant between 0 and 1.

[0010] As a possible implementation, the temperature prediction model is trained according to the following steps: obtaining multiple third data groups, each of which includes the liquid volume inside the pot, the stove gear, the pot bottom temperature value every first preset time during the heating process, the pot temperature value every first preset time during the heating process, and the heating time at the current moment. The multiple third data groups are processed to obtain multiple fifth data groups. The multiple fifth data groups are used to train the temperature prediction model based on a fully connected neural network.

[0011] As a possible implementation, multiple third data groups are obtained, including: obtaining each third data group in sequence, each group of third data corresponds to a volume of liquid inside the pot, and when each volume of liquid inside the pot is heated, obtaining third data heated in different heating methods. Obtaining third data from the first volume to the second volume of the liquid inside the pot, at every preset volume. Obtaining third data when heating starts when there is residual temperature. Different heating methods include: using the same gear for heating from the start of heating to the liquid inside the pot reaching the boiling point. Using two gears to switch timedly for heating from the start of heating to the liquid inside the pot reaching the boiling point. Randomly switching gears for heating every 15 seconds from the start of heating to the liquid inside the pot reaching the boiling point.

[0012] As a possible implementation, the step of obtaining each third data group includes: obtaining the volume of the liquid inside the pot. According to the heating method, recording the pot gear during the heating process. Obtaining the pot bottom temperature value from the start of heating to the boiling point of the liquid inside the pot every second to obtain a plurality of pot bottom temperature values. Obtaining the pot temperature value from the start of heating to the boiling point of the liquid inside the pot every second to obtain a plurality of pot temperature values. Obtaining the time taken from the start of heating to the current moment to obtain the heating time.

[0013] As a possible implementation, the step of processing multiple third data groups to obtain a fifth data group includes: for each third data group, feature extraction is performed on the third data corresponding to each first preset time period after the third preset time period from the start of heating to obtain a fourth data group. The fourth data group includes: the pot bottom temperature value at the current moment, the liquid volume inside the pot, the pot bottom temperature difference, the heating time at the current moment, and the total power corresponding to the current moment. The pot bottom temperature difference is the difference between the current pot bottom temperature value and the pot bottom temperature value before the set time. The multiple fourth data groups are normalized to obtain multiple fifth data groups.

[0014] As a possible implementation manner, obtaining the total power corresponding to the current moment includes: calculating the total power corresponding to the current moment according to the following formula based on the stove gear position during the heating process.

[0015]

[0016] Where w is the total power corresponding to the current moment, T is the heating time at the current moment, and w t is the heating power at time t, and γ is a constant between 0 and 1.

[0017] As a possible implementation method, multiple fourth data groups are normalized to obtain multiple fifth data groups, including: for each characteristic data in the fourth data group, normalization is performed according to the following formula, and the characteristic data includes the bottom temperature value of the pot at the current moment, the bottom temperature difference of the pot, the heating time at the current moment, and the total power corresponding to the current moment.

[0018]

[0019] where x norm is the normalized data, x min 、x max are respectively the minimum value and the maximum value of the characteristic data in the plurality of fourth data groups, and x is the characteristic data.

[0020] As a possible implementation, the step of training the temperature prediction model based on the fully connected neural network using multiple fifth data groups includes: dividing the multiple fifth data groups into a training set and a validation set. The temperature prediction model is trained using the training set, and after a round of training is completed using the training set, the temperature prediction model is error analyzed using the validation set, and the training is terminated if the training termination condition is met, otherwise the temperature prediction model is trained for the next round using the training set.

[0021] As a possible implementation, the step of using the validation set to perform error analysis on the temperature prediction model includes: inputting the fifth data group in the validation set into the temperature prediction model, outputting a predicted value of the temperature inside the pot, and comparing the temperature inside the pot at the same time in the third data group corresponding to the fifth data group with the predicted value of the temperature inside the pot to obtain an error value.

[0022] As a possible implementation method, the termination condition is: average the error values ​​obtained from the error analysis of the temperature prediction model using the validation set in each round to obtain the average loss value, select the smallest average loss value in multiple rounds of training as the standard value, and terminate the training if the standard value does not decrease after multiple rounds of training.

[0023] As a possible implementation, each data in the intermediate data group is normalized to obtain a second data group, including:

[0024] For each characteristic data in the intermediate data group, normalization is performed according to the following formula. The characteristic data includes the current pot bottom temperature value, the pot bottom temperature difference, the current heating time and the total power corresponding to the current moment.

[0025]

[0026] where x norm is the normalized data, x min 、x max are respectively the minimum value and the maximum value of the characteristic data in the plurality of fourth data groups, and x is the characteristic data. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The accompanying drawings are used to provide a further understanding of the technical solution of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the technical solution of the present invention and do not constitute a limitation on the technical solution of the present invention.

[0028] Figure 1 A schematic diagram of a stove provided in an embodiment of the present application;

[0029] Figure 2 A schematic diagram of a first data group provided in an embodiment of the present application;

[0030] Figure 3 A schematic diagram of a stove provided in an embodiment of the present application;

[0031] Figure 4 A schematic diagram of a second data group provided in an embodiment of the present application;

[0032] Figure 5 A controller configuration flow chart provided in an embodiment of the present application;

[0033] Figure 6 A data processing flow chart provided for an embodiment of the present application;

[0034] Figure 7 A training flow chart of a temperature prediction model provided in an embodiment of the present application;

[0035] Figure 8 A schematic diagram of a third data group provided in an embodiment of the present application;

[0036] Fig. 9 A schematic diagram of another third data set provided in an embodiment of the present application;

[0037] Fig.10 A flowchart of data processing provided in an embodiment of the present application;

[0038] Fig.11 A schematic diagram of a fourth data group provided in an embodiment of the present application;

[0039] Fig.12 A flowchart of a model training provided in an embodiment of the present application;

[0040] Fig.13 A schematic diagram of dividing a fifth data group provided in an embodiment of the present application;

[0041] Fig.14 A flow chart for model error analysis provided in an embodiment of the present application. DETAILED DESCRIPTION

[0042] 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.

[0043] It should be noted that all directional indications in the embodiments of the present invention (such as up, down, left, right, front, back, etc.) are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0044] 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 the present invention, unless otherwise specified, "plurality" means two or more.

[0045] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "connected" and "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances. In addition, when describing the pipeline, the "connected" and "connection" used in the present invention have the meaning of conduction. The specific meaning needs to be understood in conjunction with the context.

[0046] In the embodiments of the present invention, words such as "exemplary" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0047] As living standards continue to improve, people are paying more and more attention to health. As an indispensable part of daily life, people naturally have higher requirements for diet. Whether the temperature can be accurately controlled is a key factor in whether delicious dishes can be cooked, because once the oil temperature exceeds 200 degrees Celsius, it is easy to produce oil smoke and other harmful substances. For low-temperature cooking methods, precise control of temperature is even more necessary. However, users often rely on their own experience to judge the temperature, which is not accurate.

[0048] Therefore, the present application embodiment provides a stove, such as Figure 1 As shown, the temperature inside the pot can be estimated more accurately. Figure 1 The stove 1 includes a controller 11 , and a pot 2 is placed on the stove 1 .

[0049] The controller 11 is configured to obtain a first data set. The first data set includes the liquid volume inside the pot, the stove gear, the temperature value of the pot bottom at a first preset time interval during the heating process, and the heating time at the current moment, wherein the first preset time interval may be 1 second. Figure 2 As shown, Figure 2 The composition of the first data set is shown. Figure 2 , Figure 2 The volume of the liquid inside the pot in the formula refers to the volume of water or oil. The controller 11 is also connected to the intelligent recipe for communication. Through the intelligent recipe, it can be known whether the volume of the liquid inside the pot is water or oil. The weight of the liquid inside the pot is calculated by subtracting the weight of the pot itself from the weight of the liquid added inside the pot. Since the controller also stores the density of water 1kg / l and the density of oil 0.92kg / l and the volume algorithm, that is, volume = weight / density, the volume of the liquid inside the pot can be calculated.

[0050] Heating time refers to the time from the start of heating to the current moment, recorded once every second. Figure 2 As shown in the example, the last column of the heating time is 80s, which means that the pot 2 is heated for a total of 80s.

[0051] The stove 1 often has different gears, and different gears correspond to different heating powers. For example, the stove has nine gears, starting from gear 1 to gear 9, and the heating power increases by 20w for each increase in gear. The controller 11 can obtain the gear information of the stove 1, and thus can also calculate the corresponding heating power. Figure 3 A temperature sensor 3 is provided at the bottom of the pot 2, and the temperature sensor 3 is communicatively connected with the controller 11 of the cooker 1, so that the controller 11 can obtain the temperature of the bottom of the pot 2, that is, the pot bottom temperature value, through the temperature sensor 3.

[0052] It should be noted that the data sets mentioned in the embodiments of the present application are a series of values ​​obtained when the liquid volume inside the cookware is fixed. Figure 2 For example, a series of values ​​includes the stove gear position and the temperature value of the bottom of the pot obtained every second, that is, when the volume of liquid inside the pot is constant, the stove gear position and the temperature value of the bottom of the pot at the nth second of heating are obtained.

[0053] After acquiring the first data group, the controller 2 is further configured to process the acquired first data group to obtain a second data group, such as Figure 4As shown. The second data set includes the normalized internal liquid volume of the pot, the normalized pot bottom temperature value at the current moment, the normalized pot bottom temperature difference, the normalized heating time at the current moment, and the normalized total power corresponding to the current moment. Among them, the pot bottom temperature difference is the difference between the current pot bottom temperature value and the pot bottom temperature value before the set time. The set time generally takes 3-6 values ​​within 30 seconds. For example, if the set time is 5 seconds, when the heating time is 40s, the pot bottom temperature value when the current pot bottom temperature difference is 40s is the difference between the pot bottom temperature value at the moment of 35s.

[0054] The volume of liquid inside the pot, the temperature value of the pot bottom at the current moment, the temperature difference of the pot bottom, the heating time at the current moment, and the total power corresponding to the current moment are normalized to obtain the normalized volume of liquid inside the pot, the normalized temperature value of the pot bottom at the current moment, the normalized temperature difference of the pot bottom, the normalized heating time at the current moment, and the normalized total power corresponding to the current moment. Normalization can convert multiple dimensional data into dimensionless data, and map the data within a specified range of values, usually between 0 and 1, which is convenient for processing model analysis under the combined influence of multiple factors.

[0055] After acquiring the second data set, the controller 11 is further configured to input the second data into a pre-trained temperature prediction model of a fully connected neural network to obtain a predicted value of the temperature in the pot at the current moment.

[0056] By building and training the temperature prediction model in advance and writing it into the controller 11 of the cooker 1, the cooker 1 can estimate the pot temperature value corresponding to the current moment through the volume of the liquid inside the pot, the gear position of the cooker, the temperature value of the bottom of the pot every first preset time during the heating process, and the heating time at the current moment. Since the data of the volume of the liquid inside the pot, the gear position of the cooker, the temperature value of the bottom of the pot every 1s during the heating process, and the heating time at the current moment are easy to obtain, there will be no serious delay in obtaining the above data, so the estimation of the temperature value inside the pot is more convenient and accurate.

[0057] The stove provided in this application is usually used in conjunction with other smart home appliances. For example, when a user wants to cook scallion steak, the smart recipe recognizes that the cooking method of scallion steak is to bake at 80 degrees Celsius for 5 minutes, and then bake at 100 degrees Celsius for 15 seconds. The stove provided in this application will estimate the temperature value in the pot corresponding to the current moment based on the liquid volume inside the pot, that is, the volume of oil, the stove gear, the temperature value of the bottom of the pot every 1 second during the heating process, and the heating time at the current moment. When it is estimated that the temperature in the pot reaches 80 degrees Celsius, it cooperates with other algorithm programs to control the gear to remain unchanged and record the time. After reaching 5 minutes, it starts to increase the gear until the temperature value in the pot is estimated to be 100 degrees Celsius, and the heat is maintained, and the baking time of 15 seconds is recorded.

[0058] The stove provided by the present application can estimate the temperature inside the pot more accurately, and can achieve accurate temperature control in conjunction with other smart home appliances. It will not cause the temperature inside the pot to be too high due to experience problems, thereby preventing the generation of a large amount of oil smoke and harmful substances, thereby improving the user experience.

[0059] It should be noted that Figure 2 and Figure 4 The table shown in the figure is to better display the information obtained by the controller 11, so all the data from the start of heating to the stop of heating are listed. In actual application, the controller 11 records the data group every second, and the user wants to know the temperature inside the pot at a certain moment. The controller 11 can estimate the temperature inside the pot by using only the corresponding moment data in the second data group, that is, the temperature inside the pot can be estimated by a certain column of data in the second data group.

[0060] As a possible implementation, the controller 11 processes the acquired first data group to obtain a second data group. The specific configuration process is as follows: Figure 5 As shown:

[0061] S-100: Extract features from the first data group to obtain an intermediate data group.

[0062] The specific steps of extracting features from the first data group to obtain the intermediate data group are as follows: Figure 6 As shown, including:

[0063] S-101: Get the current temperature of the pot bottom, the current heating time and the volume of liquid inside the pot.

[0064] S-102: Obtain a pot bottom temperature difference according to a pot bottom temperature value at the current moment and a pot bottom temperature value before a set time.

[0065] S-103: Calculate the total power corresponding to the current moment according to the stove gear position during the heating process.

[0066] The total power calculation formula corresponding to the current moment is as follows:

[0067]

[0068] Where w is the total power corresponding to the current moment, T is the heating time at the current moment, and w t is the heating power at time t, the controller 11 can obtain the heating power w at the current moment according to the gear position at the current moment t Since in the cooking process, the earlier the heating is, the smaller the contribution to the current temperature, a coefficient needs to be set, that is, γ as a discount factor, which is a constant between 0 and 1, for example, 0.95.

[0069] S-110: Normalize each data in the intermediate data group to obtain a second data group.

[0070] For each characteristic data in the intermediate data group, normalization is performed according to the following formula. The characteristic data includes the current pot bottom temperature value, the pot bottom temperature difference, the current heating time and the total power corresponding to the current moment.

[0071]

[0072] where x norm is the normalized data, x min 、x max are respectively the minimum value and the maximum value of the characteristic data in the plurality of fourth data groups, and x is the characteristic data.

[0073] The second data set is Figure 4 The second data is input into the temperature prediction model of the pre-trained fully connected neural network to obtain the predicted value of the temperature in the pot at the current moment.

[0074] For example, Figure 7 As shown, Figure 7 The training process of the temperature prediction model is shown. Figure 7 , the temperature prediction model is trained according to the following steps:

[0075] S-120: Acquire a plurality of third data groups.

[0076] Wherein, each third data group is as follows Figure 8 As shown, it includes the liquid volume inside the pot, the stove gear, the temperature value of the bottom of the pot every 1 second during the heating process, the temperature value inside the pot every 1 second during the heating process, and the heating time at the current moment.

[0077] Obtain a variety of different pot internal liquid volumes, stove gears, pot bottom temperature values ​​every 1 second during the heating process, and heating time. As long as the liquid inside the pot is different, the liquid volume inside the pot is different, and the stove gear is different, it is a different situation. Since the temperature prediction model envisioned to complete the training estimates the temperature value inside the pot based on the liquid volume inside the pot, the stove gear, the pot bottom temperature value, and the heating time, it is necessary to obtain a large number of pot internal liquid volumes, stove gears, pot bottom temperature values ​​every 1 second during the heating process, and heating time as preliminary inputs for training. The purpose of obtaining the temperature value inside the pot is to obtain the actual temperature value inside the pot at the corresponding moment, so as to compare it with the estimated temperature value inside the pot during the training process and verify the accuracy of the temperature prediction model.

[0078] It should be noted that the third data set is obtained from the start of heating until a fixed volume of water or oil in the pot reaches the boiling point. Figure 8 It only shows that under standard atmospheric pressure, when the liquid inside the pot is water, after heating for 80 seconds, when the temperature inside the pot is 100 degrees Celsius, the water reaches the boiling point and the recording of this set of data stops.

[0079] S-130: Process the plurality of third data groups to obtain a plurality of fifth data groups.

[0080] After the obtained plurality of third data groups are processed, a plurality of fifth data groups are obtained, and there is a correspondence between the fifth data groups and the third data groups. For example, if two data groups with the same internal liquid volume of the pot, stove gear position and heating time can be found from the plurality of fifth data groups and the plurality of third data groups, then the two data groups correspond to each other.

[0081] S-140: Training a temperature prediction model based on a fully connected neural network using the plurality of fifth data groups.

[0082] A fully connected neural network is also called a multi-layer perceptron. It is a type of feedforward neural network, consisting of an input layer, a hidden layer, and an output layer. Each hidden layer can have multiple neurons. A fully connected neural network can be used as a versatile learning method for almost all tasks, such as classification, regression, and even unsupervised learning. Therefore, using a fully connected neural network to train the temperature prediction model can make the output of the model, that is, the predicted temperature inside the pot, closer to the actual temperature inside the pot.

[0083] In the process of acquiring multiple third data groups in step S-120, each third data group needs to be acquired in sequence. Each third data group corresponds to a liquid volume inside the pot, and when each liquid volume inside the pot is heated, the third data heated in different heating modes is acquired.

[0084] Different heating methods include at least the following three forms: First, the same gear is used for heating from the start of heating to the time when the liquid inside the pot reaches the boiling point. For example, gear 1 is continuously used for heating from the start of heating to the time when the liquid inside the pot reaches the boiling point. Second, two gears are used for heating in a timed switching manner from the start of heating to the time when the liquid inside the pot reaches the boiling point. For example, gears 1 and 2 are used from the start of heating to the time when the liquid inside the pot reaches the boiling point, and the gear is switched every 10 seconds for heating. Third, the gear is randomly switched for heating every second preset time period, for example, 15 seconds, from the start of heating to the time when the liquid inside the pot reaches the boiling point.

[0085] Each set of third data corresponds to the volume of liquid inside a pot. Specifically, the third data is obtained from the first volume to the second volume of the liquid inside the pot, at every preset volume. For example, the first volume is 0.6l, the second volume is 3l, and the preset volume is 0.3l. That is to say, when the volume of liquid inside the pot is 0.6l, the temperature value of the bottom of the pot, the gear position of the stove, the temperature value inside the pot, and the heating time are recorded every second from the start of heating to the end of boiling of the liquid in the pot. Then, when the volume of liquid inside the pot is measured to be 0.9l, the temperature value of the bottom of the pot, the gear position of the stove, the temperature value inside the pot, and the heating time are recorded every second from the start of heating to the end of boiling of the liquid in the pot, until the liquid in the pot is 3l.

[0086] It should be noted that when obtaining the third data from the first volume to the second volume, at every preset volume, when the volume of liquid inside the pot is fixed, the stove gear needs to cover at least three heating methods. The embodiment of the present application only provides three basic heating methods. In fact, there are more possible heating methods. The more comprehensive heating methods are used for data acquisition, the more accurate the training model will be. In other words, when the volume of liquid inside the pot is fixed, it is necessary to test if the same gear is always used, and it is necessary to test if the two gears are switched at regular intervals, and it is also necessary to test if the gears are switched at regular intervals randomly. And each gear of the stove needs to be covered. For example, it is necessary to test if the 1st gear is always used for heating, and it is also necessary to test if the 2nd gear is always used.

[0087] In order to train the temperature prediction model to be more stable and accurate and applicable to a wider range of scenarios, it is also necessary to obtain the third data when heating starts from residual temperature. The difference between heating from residual temperature and heating until the liquid inside the pot reaches the boiling point is that the former has a certain initial temperature when heating starts. The data to be obtained and the stove gears to be covered are the same as above.

[0088] For example, Fig. 9 As shown, Fig. 9The table shown is the third data set with residual heat. Fig. 9 It can be seen that when the heating time is 0 seconds, that is, when there is no heating, the temperature of the pot bottom measured by the pot bottom temperature sensor is 34 degrees Celsius, and the temperature inside the pot is 26 degrees Celsius. For the third data set that starts heating without residual heat, the temperature measured without starting heating is basically consistent with the indoor temperature.

[0089] When measuring the third data of heating started in the residual heat state, it is also necessary to measure the third data from the first volume to the second volume of the liquid inside the pot, and every preset volume. In the case of a fixed volume of liquid inside the pot, each fixed volume of each liquid must also cover at least three heating methods, that is, using the same gear for heating from the beginning of heating to the boiling point of the liquid inside the pot; using two gears to switch timedly for heating from the beginning of heating to the boiling point of the liquid inside the pot; and randomly switching gears every 15 seconds from the beginning of heating to the boiling point of the liquid inside the pot. There are many situations in this, such as different gears selected, different intervals selected, as long as there is a different variable, it belongs to a different set of third data groups. The above is that when the initial temperature of the residual heat is fixed, multiple third data groups also need to include the situation where heating starts from residual heat with different initial temperatures.

[0090] As a possible implementation, the step of obtaining each third data group includes: obtaining the volume of the liquid inside the pot. According to the heating method, recording the pot gear during the heating process. Obtaining the pot bottom temperature value from the start of heating to the boiling point of the liquid inside the pot every second to obtain a plurality of pot bottom temperature values. Obtaining the pot temperature value from the start of heating to the boiling point of the liquid inside the pot every second to obtain a plurality of pot temperature values. Obtaining the time taken from the start of heating to the current moment to obtain the heating time.

[0091] The plurality of third data sets cover heating conditions under various circumstances, and after data processing, they can be used as training sets for the temperature prediction model. For example, Fig. 9 As shown, Fig. 9 The specific flow of step S-130 of processing the plurality of third data groups to obtain the plurality of fifth data groups is shown.

[0092] Reference Fig.10 The step of processing the plurality of third data groups to obtain a fifth data group includes:

[0093] S-131: For each third data group, perform feature extraction on the third data corresponding to every 1 second after the third preset time period from the start of heating to obtain a fourth data group.

[0094] The third preset time length may be 30 seconds, such as Fig.11 As shown, Fig.11 The composition of the fourth data group is shown, and the fourth data group includes: the current pot bottom temperature value, the liquid volume inside the pot, the pot bottom temperature difference, the current heating time, and the total power corresponding to the current time. The pot bottom temperature difference is the difference between the current pot bottom temperature value and the pot bottom temperature value before the set time.

[0095] The setting time is generally 3-6 values ​​within 30 seconds. For example, if the setting time is 5 seconds, then the pot bottom temperature difference at the 31st second in the fourth data group is the pot bottom temperature value at the 31st second minus the pot bottom temperature value at the 26th second in the third data group with the same volume of liquid inside the pot, the same stove gear and the same initial pot bottom temperature value. The requirement that the initial pot bottom temperature is the same ensures that one data group does not have residual heat while the other data group is heated from the beginning.

[0096] As a possible implementation manner, obtaining the total power corresponding to the current moment includes: calculating the total power corresponding to the current moment according to the following formula based on the stove gear position during the heating process.

[0097]

[0098] Where w is the total power corresponding to the current moment, T is the heating time at the current moment, and w t is the heating power at time t, the controller 11 can obtain the heating power w at the current moment according to the gear position at the current moment t Since in the cooking process, the earlier the heating is, the smaller the contribution to the current temperature, a coefficient needs to be set, that is, γ as a discount factor, which is a constant between 0 and 1, for example, 0.95.

[0099] S-132: Normalize the multiple fourth data groups to obtain multiple fifth data groups.

[0100] As a possible implementation method, multiple fourth data groups are normalized to obtain multiple fifth data groups, including: for each characteristic data in the fourth data group, normalization is performed according to the following formula, and the characteristic data includes the bottom temperature value of the pot at the current moment, the bottom temperature difference of the pot, the heating time at the current moment, and the total power corresponding to the current moment.

[0101]

[0102] where x norm is the normalized data, x min 、x maxare respectively the minimum value and the maximum value of the characteristic data in the plurality of fourth data groups, and x is the characteristic data.

[0103] It should be noted that x min 、x max They are respectively the minimum and maximum values ​​of the characteristic data in multiple fourth data groups. The maximum and minimum values ​​here are obtained by comparing all data of the same category. That is to say, when calculating the maximum and minimum values ​​of the bottom temperature of the pot, it is necessary to compare all the fourth data groups and select the maximum and minimum values ​​of the bottom temperature of the pot at the current moment.

[0104] After obtaining the fifth data set, the temperature model can be trained. The embodiment of the present application uses a fully connected neural network as the basic model, uses a four-layer neural network, including three hidden layers and one output layer, and uses about 300 neurons in each layer for learning and training, and finally obtains a temperature prediction model.

[0105] Exemplarily, the steps of training the temperature prediction model based on a fully connected neural network using multiple fifth data sets are as follows: Fig.12 As shown, including:

[0106] S-150: Divide the plurality of fifth data sets into a training set and a validation set.

[0107] As a possible implementation, Fig.13 As shown, Fig.13 The division method of multiple fifth data groups is shown. Multiple fifth data groups can also be divided into a training set, a validation set and a test set. The training set is used to learn and train the temperature prediction model. The trained temperature prediction model is verified by the validation set to determine whether the output of the temperature prediction model is accurate and stable. The test set is used to further test the trained temperature prediction model.

[0108] S-160: Use the training set to train the temperature prediction model. After completing a round of training using the training set, use the validation set to perform error analysis on the temperature prediction model.

[0109] As a possible implementation method, the steps of using the validation set to perform error analysis on the temperature prediction model are as follows: Fig.14 As shown, including:

[0110] S-161: Input the fifth data group in the validation set into the temperature prediction model and output the predicted value of the temperature in the pot.

[0111] That is, the normalized pot bottom temperature value at the current moment, the normalized liquid volume inside the pot, the normalized pot bottom temperature difference, the normalized heating time at the current moment, and the normalized total power corresponding to the current moment are input to obtain the predicted value of the pot temperature at the current moment.

[0112] S-162: Compare the temperature value inside the pot at the same time in the third data group corresponding to the fifth data group with the predicted temperature value inside the pot to obtain an error value.

[0113] The temperature value inside the pot in the third data group is accurate data obtained after temperature measurement. The output of the temperature prediction model is the temperature inside the pot estimated by the trained model. When the liquid volume inside the pot, the stove gear and the heating time are the same, the size of the error can be judged based on the difference between the temperature value inside the pot in the third data group and the predicted temperature value inside the pot.

[0114] S-170: Determine whether the termination condition has been met.

[0115] If the training termination condition is met, the training is terminated, otherwise the training set is used to perform the next round of training on the temperature prediction model.

[0116] In the original plan, the temperature prediction model based on the fully connected neural network needs to be trained for 1000 rounds to ensure the final accuracy of the model. However, too many training times are prone to overfitting. In order to prevent the occurrence of overfitting and the impact on the accuracy of the temperature prediction model, the embodiment of the present application adopts early termination technology to prevent overfitting. The early termination technology used in the embodiment of the present application is as follows.

[0117] As a possible implementation method, the termination condition is: average the error values ​​obtained by error analysis of the temperature prediction model using the validation set in each round to obtain the average loss value, select the smallest average loss value in multiple rounds of training as the standard value, and terminate the training if the standard value does not decrease after multiple rounds of training. By adopting such a termination condition, the temperature prediction model can be made more accurate, with an average error within 4 degrees Celsius.

[0118] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A cooking appliance, characterized in that: A pot is placed on the stove; the stove comprises: A controller, the controller being configured to: Acquire a first data group, the first data group including the volume of liquid inside the pot, the stove gear position, the temperature value of the bottom of the pot at intervals of a first preset time during the heating process, and the heating time at a current moment; The first data group is processed to obtain a second data group, wherein the second data group includes a normalized volume of liquid inside the pot, a normalized pot bottom temperature value at the current moment, a normalized pot bottom temperature difference value, a normalized heating time at the current moment, and a normalized total power corresponding to the current moment; wherein the pot bottom temperature difference value is a difference between the current pot bottom temperature difference value and the pot bottom temperature value before a set time; The second data is input into a pre-trained temperature prediction model to obtain a predicted value of the temperature in the pot at the current moment.

2. The cooker according to claim 1, characterized in that: The controller is configured to process the acquired first data group to obtain a second data group, and is specifically configured to: Extracting features from the first data group to obtain an intermediate data group; Normalizing each data in the intermediate data group to obtain a second data group; Extracting features from the first data group to obtain an intermediate data group includes: Get the current temperature of the pot bottom, the current heating time and the volume of liquid inside the pot; According to the current temperature value of the pot bottom and the temperature value of the pot bottom before the set time, the temperature difference of the pot bottom is obtained; According to the stove gear position during the heating process, the total power corresponding to the current moment is calculated according to the following formula; Where w is the total power corresponding to the current moment, T is the heating time at the current moment, and w t is the heating power at time t, and γ is a constant between 0 and 1.

3. The cooker according to claim 1, characterized in that: The temperature prediction model is trained according to the following steps: Acquire a plurality of third data groups, wherein the third data groups include the volume of liquid inside the pot, the range position, the temperature value of the bottom of the pot at intervals of a first preset time during the heating process, the temperature value of the inside of the pot at intervals of a first preset time during the heating process, and the heating time at a current moment; Processing the plurality of third data groups to obtain a plurality of fifth data groups; The temperature prediction model is trained based on a fully connected neural network using multiple fifth data sets.

4. The cooker according to claim 3, characterized in that: The obtaining of a plurality of third data groups comprises: Acquire multiple third data groups, each group of third data corresponds to a volume of liquid inside the pot, and when each volume of liquid inside the pot is heated, acquire third data heated in different heating modes; Acquire third data when the volume of the liquid inside the cookware increases from the first volume to the second volume at intervals of a preset volume; Acquire the third data when heating starts from the time when there is residual temperature; The different heating methods include: Use the same setting from the beginning of heating until the liquid inside the pot reaches boiling point; From the beginning of heating until the liquid inside the pot reaches boiling point, heating is performed by switching between two gears at a fixed time. From the beginning of heating until the liquid inside the pot reaches the boiling point, the gear is randomly switched for heating every second preset time period.

5. The cooker according to claim 4, characterized in that: The step of acquiring each third data group comprises: Get the liquid volume inside the pot; According to the heating method, record the pot position during the heating process; Obtain the temperature value of the bottom of the pot from the beginning of heating to the time when the liquid inside the pot reaches the boiling point, and obtain multiple temperature values ​​of the bottom of the pot; Obtain the temperature value of the pot from the beginning of heating to the time when the liquid inside the pot reaches the boiling point, and obtain multiple temperature values ​​of the pot; Get the time from the start of heating to the current moment to get the heating time.

6. The cooker according to claim 3, characterized in that: The step of performing data processing on the plurality of third data groups to obtain a fifth data group comprises: For each third data group, feature extraction is performed on the third data corresponding to each first preset time period after the third preset time period of heating starts, to obtain a fourth data group; the fourth data group includes: the pot bottom temperature value at the current moment, the liquid volume inside the pot, the pot bottom temperature difference, the heating time at the current moment, and the total power corresponding to the current moment; wherein the pot bottom temperature difference is the difference between the current pot bottom temperature value and the pot bottom temperature value before the set time; The multiple fourth data groups are normalized to obtain multiple fifth data groups.

7. The cooker according to claim 6, characterized in that: Obtaining the total power corresponding to the current moment includes: According to the stove gear position during the heating process, the total power corresponding to the current moment is calculated according to the following formula; Where w is the total power corresponding to the current moment, T is the heating time at the current moment, and w t is the heating power at time t, and γ is a constant between 0 and 1.

8. The cooker according to claim 6, characterized in that: The step of normalizing the plurality of fourth data groups to obtain a plurality of fifth data groups comprises: The characteristic data in the fourth data group are normalized according to the following formula, wherein the characteristic data include the pot bottom temperature value at the current moment, the pot bottom temperature difference, the heating time at the current moment, and the total power corresponding to the current moment, where x norm is the normalized data, x min 、x max are respectively the minimum value and the maximum value of the characteristic data in the plurality of fourth data groups, and x is the characteristic data.

9. The cooker according to claim 3, characterized in that: The step of training the temperature prediction model using a plurality of fifth data groups comprises: Dividing the plurality of fifth data sets into a training set and a validation set; The temperature prediction model is trained using the training set. After one round of training is completed using the training set, the temperature prediction model is subjected to error analysis using the validation set. If the training termination condition is met, the training is terminated. Otherwise, the temperature prediction model is subjected to the next round of training using the training set.

10. The cooker according to claim 9, characterized in that: The step of using the validation set to perform error analysis on the temperature prediction model comprises: Input the fifth data group in the validation set into the temperature prediction model and output the predicted value of the temperature in the pot; The temperature value in the pot at the same time in the third data group corresponding to the fifth data group is compared with the predicted value of the temperature in the pot to obtain an error value.

11. The cooker according to claim 9, characterized in that: The termination conditions are: The error values ​​obtained by performing error analysis on the temperature prediction model using the validation set in each round are averaged to obtain the average loss value. The smallest average loss value in multiple rounds of training is selected as the standard value. If the standard value does not decrease after multiple rounds of training, the training is terminated.

12. The cooker according to claim 2, characterized in that: The step of normalizing the data in the intermediate data group to obtain the second data group includes: The characteristic data in the intermediate data group are normalized according to the following formula, wherein the characteristic data includes the pot bottom temperature value at the current moment, the pot bottom temperature difference, the heating time at the current moment, and the total power corresponding to the current moment. where x norm is the normalized data, x min 、x max are respectively the minimum value and the maximum value of the characteristic data in the plurality of fourth data groups, and x is the characteristic data.

Images