Operation Prompt Method for Air Conditioner, Air Conditioner Equipment and Computer Storage Medium

Abstract

The present application discloses an operation prompt method for an air conditioner, an air conditioner device, and a computer storage medium. The method includes: detecting and confirming that an air conditioner operation instruction is received, and obtaining current characteristic parameters of the air conditioner; wherein, the characteristic parameters include air conditioner attribute parameters, air conditioner setting parameters, air conditioner operation parameters, environmental parameters, and historical derivative information, and the historical derivative information is determined by historical characteristic parameters; inputting the current characteristic parameters into a trained classification model, and obtaining a classification result output by the classification model; wherein, the classification model is trained based on historical characteristic parameters and corresponding classification results, and the classification results include necessary operations and non-necessary operations; if the classification result indicates that the air conditioner operation instruction is a non-necessary operation, a prompt is made. It can promptly give power consumption prompts to users and reduce energy waste.

Description

Technical Field

[0001] The present application relates to the technical field of air conditioners, and particularly to an operation prompt method for an air conditioner, an air conditioner device, and a computer-readable storage medium. Background Art

[0002] With the development of society, the three major household appliances, namely air conditioners, refrigerators, and washing machines, have gradually become essential appliances in people's homes. Among the three major household appliances, the power consumption ratio of air conditioners is relatively high. Therefore, the power consumption of air conditioners has always been a key issue that users are more concerned about. The traditional energy-saving and power-saving functions are mainly achieved by paying attention to the machine parameter settings of air conditioners during stable operation. However, it is still impossible to avoid or reduce the power consumption caused by users' non-essential operations in a short period of time. For example, going out to get takeout for a short time causes the air conditioner to be turned off and then turned on again. Such operations will cause the air conditioner to consume more power than when it is running continuously and stably, but users are unaware of this power consumption event. Summary of the Invention

[0003] To solve the above problems, the present application provides an operation prompt method for an air conditioner, an air conditioner device, and a computer-readable storage medium, which can timely prompt users of power consumption and reduce energy waste.

[0004] To solve the above technical problems, one technical solution adopted by the present application is: to provide an operation prompt method for an air conditioner, the method includes: detecting and confirming that an air conditioner operation instruction is received, and obtaining the current characteristic parameters of the air conditioner; wherein, the characteristic parameters include air conditioner attribute parameters, air conditioner setting parameters, air conditioner operation parameters, environmental parameters, and historical derivative information, and the historical derivative information is determined by historical characteristic parameters; inputting the current characteristic parameters into a trained classification model, and obtaining a classification result output by the classification model; wherein, the classification model is trained based on historical characteristic parameters and corresponding classification results, and the classification results include necessary operations and non-necessary operations; if the classification result indicates that the air conditioner operation instruction is a non-necessary operation, a prompt is made.

[0005] Wherein, the method further includes: establishing a classification model; detecting and confirming that a historical air conditioner operation instruction is received, and obtaining the first historical characteristic parameters of the air conditioner; wherein, the first historical characteristic parameters include air conditioner attribute parameters, historical air conditioner setting parameters, historical air conditioner operation parameters, and historical environmental parameters; determining the classification result corresponding to the historical air conditioner operation instruction according to the first historical characteristic parameters, and obtaining the second historical characteristic parameters corresponding to the classification result; inputting the classification result corresponding to the historical air conditioner operation instruction and the second historical characteristic parameters into the established classification model to train the classification model.

[0006] Among them, determine the classification result corresponding to the historical air conditioner operation instruction according to the first historical feature parameter, and obtain the second historical feature parameter corresponding to the classification result; calculate the power consumption information at the moment corresponding to the historical air conditioner operation instruction according to the air conditioner attribute parameter, the historical air conditioner setting parameter, the historical air conditioner operation parameter, and the historical environment parameter; if the power consumption information is greater than the preset power consumption information, determine that the classification result corresponding to the historical air conditioner operation instruction is an unnecessary operation, and calculate the historical derivative information of the air conditioner according to the air conditioner attribute parameter, the historical air conditioner setting parameter, the historical air conditioner operation parameter, and the historical environment parameter; determine the second historical feature parameter corresponding to the unnecessary operation according to the first historical feature parameter and the historical derivative information.

[0007] Among them, determining the second historical feature parameter corresponding to the unnecessary operation according to the first historical feature parameter and the historical derivative information includes: performing feature transformation on the first historical feature parameter and the historical derivative information to obtain a high-dimensional feature parameter set; performing feature extraction on the high-dimensional feature parameter set to determine the second historical feature parameter corresponding to the unnecessary operation.

[0008] Among them, the method further includes: if the power consumption information is less than or equal to the preset power consumption information, determine that the classification result corresponding to the historical air conditioner instruction operation is a necessary operation, and calculate the historical derivative information of the air conditioner according to the air conditioner attribute parameter, the historical air conditioner setting parameter, the historical air conditioner operation parameter, and the historical environment parameter; determine the second historical feature parameter corresponding to the necessary operation according to the first historical feature parameter and the historical derivative information.

[0009] Among them, inputting the classification result corresponding to the historical air conditioner operation instruction and the second historical feature parameter into the established classification model to train the classification model includes: using a part of the classification result corresponding to the historical air conditioner operation instruction and the second historical feature parameter as a training data set and inputting it into the established classification model to train the classification model; using another part of the classification result corresponding to the historical air conditioner operation instruction and the second historical feature parameter as a test data set to test the trained classification model to determine the target neighbor parameter of the classification model.

[0010] Among them, the classification result corresponding to the historical air conditioner operation instruction and another part of the second historical feature parameter are used as a test data set to test the trained classification model to determine the target neighbor parameter of the classification model, including: taking the classification result corresponding to the historical air conditioner operation instruction and another part of the second historical feature parameter as a test data set respectively according to multiple preset neighbor parameters as calculation criteria, and inputting them into the trained classification model to output the classification result corresponding to the test data set; detecting and confirming the classification result corresponding to the test data set to determine the multiple accuracy rates corresponding to the multiple preset neighbor parameters; sorting the multiple preset neighbor parameters corresponding to the multiple accuracy rates to take the preset neighbor parameter corresponding to the maximum accuracy rate as the target neighbor parameter of the classification model.

[0011] Among them, inputting the current feature parameter into the trained classification model and obtaining the classification result output by the classification model includes: inputting the current feature parameter into the trained classification model, and calculating multiple Euclidean distances between the current sample corresponding to the current feature parameter and multiple historical samples corresponding to the second historical feature parameter; sorting the multiple Euclidean distances and determining k classification results corresponding to the minimum Euclidean distance according to the target neighbor parameter; voting on the k classification results corresponding to the minimum Euclidean distance to determine the classification result of the current sample.

[0012] Among them, voting on the k classification results corresponding to the minimum Euclidean distance to determine the classification result of the current sample includes: respectively obtaining the weights of the k classification results; respectively calculating the sum of the weights of the same classification results among the k classification results; taking the classification result corresponding to the maximum sum of weights as the classification result corresponding to the current sample.

[0013] Among them, inputting the current feature parameter into the trained classification model and obtaining the classification result output by the classification model includes: sending the current feature parameter to the server so that the server inputs the current feature parameter into the trained classification model and obtains the classification result output by the classification model; receiving the classification result sent by the server.

[0014] Among them, the method further includes: detecting and confirming that a feedback operation instruction is received, and then correcting the classification model by using the current feature parameter corresponding to the air conditioner operation instruction.

[0015] Among them, the air conditioner operation instruction refers to an instruction received by the air conditioner, sent by the air conditioner control terminal, and used to change the working state of the compressor of the air conditioner.

[0016] Among them, the air conditioner attribute parameters are the basic attribute information of the air conditioner obtained according to the model or identifier of the air conditioner; or, the air conditioner setting parameters are the working indicators that the air conditioner needs to achieve under the control of the air conditioner operation instruction; or, the air conditioner operation parameters are the working conditions information that the air conditioner actually reaches under the control of the air conditioner operation instruction; or, the environmental parameters are the environmental temperature information when the air conditioner is operating.

[0017] Among them, the air conditioner attribute parameters include at least one of the horsepower, cooling power, heating power, cooling capacity, heating capacity, and energy efficiency grade of the air conditioner; or,

[0018] The air conditioner setting parameters include at least one of the temperature setting parameter and the wind speed setting parameter of the air conditioner; or, the air conditioner operation parameters include at least one of the compressor operation frequency, indoor fan speed, outdoor fan speed, outdoor unit operation frequency, and power consumption value; or, the environmental parameters include at least one of the indoor environmental temperature, outdoor environmental temperature, indoor environmental humidity, and outdoor environmental humidity; or, the historical derivative information includes at least one of the non-essential operation time factor, non-essential operation probability, essential operation time factor, essential operation probability, and scenario characteristics at a historical moment.

[0019] To solve the above technical problems, another technical solution adopted by this application is: to provide an air conditioner device, which includes a processor and a memory. The memory is used to store a computer program, and when the computer program is executed by the processor, it is used to implement the above-mentioned operation prompt method for the air conditioner.

[0020] To solve the above technical problems, yet another technical solution adopted by this application is: to provide a computer-readable storage medium for storing a computer program, and when the computer program is executed by the processor, it is used to implement the above-mentioned operation prompt method for the air conditioner.

[0021] The beneficial effects of the embodiments of this application are: Different from the prior art, the operation prompt method for the air conditioner provided by this application obtains the current characteristic parameters of the air conditioner, such as air conditioner attribute parameters, air conditioner setting parameters, air conditioner operation parameters, environmental parameters, and historical derivative information, when detecting an air conditioner operation instruction. Further, this part of the characteristic parameters is input into the trained classification model, so as to obtain the classification result output by the classification model, and when the classification result indicates that the air conditioner operation instruction is a non-essential operation, a prompt is given to the user. In this way, on the one hand, adding historical derivative information to the classification model for classification prediction can improve the classification accuracy; on the other hand, when the classification result indicates a non-essential operation, it can timely give a power consumption prompt to the user, thereby reducing unnecessary energy waste and improving the energy-saving performance of the air conditioner. Description of the Drawings

[0022] To more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings. Among them:

[0023] Figure 1 is a schematic flowchart of an embodiment of the operation prompt method for an air conditioner provided by the present application;

[0024] Figure 2 is a schematic flowchart of another embodiment of the operation prompt method for an air conditioner provided by the present application;

[0025] Figure 3 is Figure 2 a specific flowchart of S23 in

[0026] Figure 4 is Figure 2 a specific flowchart of S24 in

[0027] Figure 5 is Figure 4 a specific flowchart of S242 in

[0028] Figure 6 is Figure 2 a specific flowchart of S26 in

[0029] Figure 7 is a schematic structural diagram of an embodiment of an air conditioner device provided by the present application;

[0030] Figure 8 is a schematic structural diagram of an embodiment of a computer-readable storage medium provided by the present application. Detailed implementation manners

[0031] The following will clearly and completely describe the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. It can be understood that the specific embodiments described herein are only used to explain the present application, rather than limiting the present application. Additionally, it should be noted that for the convenience of description, only the parts related to the present application are shown in the drawings, rather than all the structures. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present application.

[0032] Reference to "embodiment" in this text means that the specific features, structures, or characteristics described in connection with the embodiment may be included in at least one embodiment of the present application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment each time, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0033] Referring to Figure 1 , Figure 1 FIG. is a schematic flow chart of an embodiment of an air conditioner operation prompt method provided by the present application. The method steps of this embodiment specifically include:

[0034] S11: Detect and confirm that an air conditioner operation instruction is received, and obtain the current characteristic parameters of the air conditioner; wherein, the characteristic parameters include air conditioner attribute parameters, air conditioner setting parameters, air conditioner operation parameters, environmental parameters, and historical derivative information, and the historical derivative information is determined by historical characteristic parameters.

[0035] In this embodiment, the air conditioner operation instruction refers to an instruction received by the air conditioner, sent by an air conditioner control terminal, and used to change the working state of the air conditioner compressor. For example, the control operation instruction can be an air conditioner on / off instruction, a cooling / heating mode switching instruction, a temperature adjustment instruction, etc. In order to respond to these operation instructions, the air conditioner necessarily needs to adjust the working state of the air conditioner compressor to achieve. For example, a swing wind instruction will not change the working state of the air conditioner compressor, so the swing wind instruction is not included in the air conditioner operation instruction of this embodiment.

[0036] Among them, the air conditioner control terminal refers to a terminal device that can be used to send air conditioner operation instructions, such as an air conditioner remote control, a wall switch, and a user's mobile terminal, etc. The user sets the air conditioner control terminal to make the air conditioner control terminal send corresponding air conditioner operation instructions to the air conditioner end to achieve corresponding working state changes.

[0037] Among them, the air conditioner attribute parameters refer to the basic attribute information of the air conditioner that can be obtained according to the model or unique identifier of the air conditioner. For example, they include basic attribute information such as the horsepower of the air conditioner, cooling power, heating power, cooling capacity, heating capacity, and energy efficiency level. It can be understood that different values of the above parameters can cause different changes in the operating state and power consumption state of the air conditioner. Under the condition that other factors remain unchanged, for example, the difference in cooling power or cooling capacity will make the cooling ability of the air conditioner strong or weak, and the power consumption caused by different cooling abilities will surely be different; if combined with different air conditioner horsepowers, it will further affect the above differences. Therefore, the above air conditioner attribute parameters can be used to evaluate the performance of the air conditioner and thus serve as the basic parameters for subsequent model classification. Optionally, the air conditioner attribute parameters can be obtained by scanning a QR code or barcode in the server, or can also be obtained by querying the nameplate directly set on the indoor unit or outdoor unit.

[0038] Among them, the air conditioner setting parameters refer to the working indicators that the air conditioner needs to achieve under the control of the air conditioner operation instructions. The air conditioner setting parameters include, for example, temperature setting parameters and wind speed setting parameters, etc., and can be used to represent the temperature or wind speed effects that the air conditioner needs to achieve under different preset working states, such as the cooling effect in the cooling state and the degree of the cooling effect. It can be understood that when the preset working states are the same or different, different temperature or wind speed settings can bring different user experiences and can also be reflected in the power consumption state. For example, under the condition that the indoor and outdoor temperatures are constant, setting a lower temperature in the cooling state will surely cause more power consumption compared to a larger temperature parameter. Therefore, the above air conditioner setting parameters can serve as the basic parameters for model classification. Among them, the preset working states can include the cooling state, heating state, sleep state, dehumidification state, ventilation state, etc.

[0039] Among them, the air conditioner operation parameters refer to the working condition information that the air conditioner reaches in real time under the control of the air conditioner operation instructions. For example, in the cooling state, the air conditioner operation parameters can include information such as the compressor operation frequency of the air conditioner, indoor fan speed, outdoor fan speed, outdoor unit operation frequency, and power value, etc. This part of information can reflect the real-time operation parameters of the air conditioner in different working states, and this part of data is usually related to the air conditioner setting parameters and may also change due to objective factors. It can be understood that different air conditioner operation parameters may bring different user experiences and differences in power consumption state. For example, in the cooling state, with different compressor operation frequencies, the indoor temperatures obtained by cooling in the same time may be the same or different, but the power consumption of different compressor operation frequencies will surely be more. Therefore, the above air conditioner operation parameters can serve as the basic parameters for model classification.

[0040] Among them, the environmental parameters refer to the environmental temperature information during the operation of the air conditioner, such as the indoor environmental temperature, outdoor environmental temperature, indoor environmental humidity, and outdoor environmental humidity, etc. The indoor environmental temperature or humidity can be obtained through the sensor set at the air inlet position of the indoor unit panel, and the outdoor environmental temperature or humidity can be obtained by the server according to the corresponding date, regional information, and corresponding time period. It can be understood that since the magnitude of the indoor-outdoor temperature difference will directly affect the working state of the air conditioner, and thus affect the power consumption of the air conditioner. For example, on the premise of the same indoor temperature, the cooling demand of the air conditioner at an outdoor temperature of 30 degrees is greater than that at an outdoor temperature of 25 degrees, and the power consumption is also greater than that at an outdoor temperature of 25 degrees. Therefore, the above-mentioned current environmental parameters can be used as the basic parameters for model classification.

[0041] Among them, the historical derivative information is determined by historical characteristic parameters, that is, determined by the air conditioner attribute parameters, historical air conditioner setting parameters, historical air conditioner operation parameters, and historical environmental parameters. The historical derivative information refers to data that usually cannot be directly obtained and needs to be calculated, statistically analyzed, or transformed, such as the non-essential operation time factor, non-essential operation probability, essential operation time factor, essential operation probability, and scenario characteristics at a historical moment. It can be understood that the above-mentioned time or probability information usually represents a user's air conditioner usage habit. For example, the user is used to turning off the air conditioner at exactly 12:00 noon, and the probability of turning on the air conditioner at 12:15 noon after turning it off is 80%. Due to the existence of user habits, the operating state of the air conditioner may change, and thus the corresponding increase in air conditioner loss may occur. Therefore, the above-mentioned historical derivative information can be used as the basic parameters for model classification, and the historical derivative information will be further elaborated in subsequent embodiments.

[0042] It can be understood that the current characteristic parameters of the above-mentioned air conditioner are only examples in this embodiment and are not limited thereto. The characteristic parameters can be replaced according to the actual situation.

[0043] S12: Input the current characteristic parameters into the trained classification model and obtain the classification result output by the classification model.

[0044] Among them, the classification model is trained based on historical characteristic parameters and the corresponding classification results, and is mainly established using supervised classification algorithms, such as models including logistic regression, support vector machine classification, K-nearest neighbor classification, Gaussian naive Bayes, ensemble models (Xgboost classification, random forest classification), etc.; among them, the classification results include essential operations and non-essential operations.

[0045] In this embodiment, when it is detected and confirmed that an air conditioner operation instruction input or sent from the user terminal is received, it indicates that the user is controlling the air conditioner at this time, such as operations like increasing / decreasing the set temperature, changing the air conditioner mode, turning the air conditioner on or off, etc. As described above, different changes in the air conditioner characteristic parameters can result in different power consumptions. Therefore, when an air conditioner operation instruction is received, the current characteristic parameters of the air conditioner are obtained. Due to the relationship of the operation instruction, the characteristic parameters at the current moment will inevitably change. Therefore, the current characteristic parameters obtained after the change are input into the trained classification model so that the classification model outputs a classification result to indicate whether the air conditioner operation instruction currently input by the user is necessary, corresponding to whether the user's operation is a necessary operation.

[0046] Among them, a necessary operation refers to an air conditioner operation instruction input by the user. The change brought to the air conditioner operation compared with the power consumption during the normal operation of the air conditioner will not result in more power consumption, being less than or equal to the power consumption during the normal operation of the air conditioner. Then this is a necessary operation at this time. For example, when the air conditioner operates to make the indoor temperature not much different from the outdoor temperature, the temperature setting parameter is adjusted to reduce the frequency of the compressor, thereby reducing the power consumption.

[0047] A non - necessary operation refers to an air conditioner operation instruction input by the user. The change brought to the air conditioner operation compared with the power consumption during the normal operation of the air conditioner will result in more power consumption. Then this is a non - necessary operation at this time. For example, when the user goes out to get a takeout / express delivery for a short time at noon and turns the air conditioner off and then on again, this operation of turning off and on may cause the compressor to run at a higher frequency in order to restore the indoor state before the air conditioner was turned off after a period of time. Obviously, the power consumption of the compressor running at a high frequency during this period will be greater than the power consumption of the compressor running at a low frequency during the same period plus the time when the user is out, thus increasing the power consumption.

[0048] Moreover, in the classification model of this embodiment, in addition to adding the conventional characteristic parameters of the air conditioner, historical derivative information is also added for model training, so that when the classification model uses the current characteristic parameters for classification prediction, it can take into account the derivative phenomena, avoid classification prediction errors caused by accidental factors, and finally enable the classification model to output a classification result with a certain classification accuracy, improving the classification accuracy.

[0049] S13: If the classification result indicates that the air conditioner operation instruction is a non - necessary operation, a prompt is made.

[0050] In this embodiment, when the characteristic parameters of the air conditioner change due to the air conditioner operation instruction input by the user, the changed current characteristic parameters are obtained, and the current characteristic parameters are classified and predicted. According to the result of the classification and prediction, it is evaluated whether the corresponding air conditioner operation instruction is necessary. When the classification result indicates that it is not necessary, a prompt is given to the user, and when the classification result indicates that it is necessary, the air conditioner operation instruction is continued to be responded to.

[0051] When the user controls the air conditioner by inputting an operation instruction, they may not realize that this is an unnecessary power-consuming operation. Therefore, the user can reject the operation instruction according to the prompt or make other adjustments according to the prompt, so as to achieve the effect of energy conservation and power saving. Or the user may know that this is an unnecessary power-consuming operation but has to execute it, so the user can ignore the prompt.

[0052] In a specific application scenario, the input parameters and output parameters of the classification model may include:

[0053]

[0054]

[0055] Different from the prior art, the operation prompt method of the air conditioner provided in this embodiment obtains the current characteristic parameters of the air conditioner, such as the air conditioner attribute parameters, air conditioner setting parameters, air conditioner operation parameters, environmental parameters, and historical derivative information, when detecting an air conditioner operation instruction. Further, this part of the characteristic parameters is input into the trained classification model, so as to obtain the classification result output by the classification model. When the classification result indicates that the air conditioner operation instruction is an unnecessary operation, a prompt is given to the user. In this way, on the one hand, adding historical derivative information to the classification model for classification and prediction can improve the classification accuracy; on the other hand, when the classification result indicates an unnecessary operation, a power consumption prompt can be given to the user in a timely manner, thereby reducing unnecessary energy waste and improving the energy-saving performance of the air conditioner.

[0056] Refer to Figure 2 , Figure 2 is a schematic flowchart of another embodiment of the operation prompt method of the air conditioner provided by this application. The method steps of this embodiment specifically include:

[0057] S21: Establish a classification model.

[0058] From models such as the logistic regression, support vector machine classification, K-nearest neighbor classification, Gaussian naive Bayes, and ensemble models (Xgboost classification, random forest classification) described above, a model with high prediction accuracy and low prediction deviation degree is comprehensively evaluated and selected as the classification model. In the subsequent embodiments, the nearest neighbor classification algorithm is taken as an example for illustration. In practical applications, the scores and risks of multiple classification algorithms will be comprehensively considered, and the selected one may not necessarily be the nearest neighbor classification algorithm. This embodiment is only for illustrative purposes.

[0059] S22: Detect and confirm the receipt of the historical air conditioner operation instruction, and obtain the first historical characteristic parameter of the air conditioner.

[0060] Among them, the historical air conditioner operation instruction refers to the air conditioner operation instruction input or sent from the user side received at a historical moment, corresponding to the air conditioner operation instruction in the above embodiment; and the first historical characteristic parameter also corresponds to some of the current characteristic parameters in the above embodiment, specifically including the air conditioner attribute parameter, historical air conditioner setting parameter, historical air conditioner operation parameter, and historical environment parameter.

[0061] S23: Determine the classification result corresponding to the historical air conditioner operation instruction according to the first historical characteristic parameter, and obtain the second historical characteristic parameter corresponding to the classification result.

[0062] Specifically, S23 can be implemented through Figure 3 the method steps shown, and the specific steps include:

[0063] S231: Calculate the power consumption information at the moment corresponding to the historical air conditioner operation instruction according to the air conditioner attribute parameter, historical air conditioner setting parameter, historical air conditioner operation parameter, and historical environment parameter.

[0064] Among them, the power consumption information refers to the power consumption obtained by converting the power according to the various characteristic parameters of the air conditioner per unit time. Generally speaking, under the condition that the outdoor temperature and the indoor space size are constant, the power consumption of the air conditioner per unit time is only related to the air conditioner attribute parameter and the air conditioner setting parameter. For example, in the cooling state, the larger the horsepower of the air conditioner, the greater the power consumption. If the temperature setting parameter is further considered, the greater the difference between the temperature setting parameter and the outdoor temperature, the greater the power consumption.

[0065] Optionally, for the calculation of power consumption information, on the one hand, power consumption data can also be directly obtained through a power sensor or socket ecological data, etc., and then the power consumption data is converted to obtain the power consumption. On the other hand, when there is no power sensor, a simple method can also be used to statistically calculate the power consumption. Specifically, the difference between the indoor and outdoor temperature of the current environment and the temperature setting parameter can be utilized, and after converting this difference into the frequency band where the air conditioner compressor is located, it is further converted into power consumption data. This conversion method may have a certain error compared to the direct acquisition of power, but it still has great reference value for statistically comparing the levels of relative power consumption.

[0066] S232: If the power consumption information is greater than the preset power consumption information, determine that the classification result corresponding to the historical air conditioner operation instruction is an unnecessary operation, and calculate the historical derivative information of the air conditioner according to the air conditioner attribute parameters, historical air conditioner setting parameters, historical air conditioner operation parameters, and historical environment parameters.

[0067] Among them, the preset power consumption information can be expressed as the power consumption in the normal operation state, that is, the power consumption when there is no obvious change in power consumption due to the unnecessary operation of the user. Therefore, in this embodiment, when the power loss caused by the historical air conditioner operation instruction is greater than the preset power consumption information, it indicates that the classification result of the operation instruction at that historical moment should be an unnecessary operation. That is to say, before the classification model is not / finished training, the method of classifying the air conditioner operation instruction at the historical moment is determined by comparing the power consumption information.

[0068] Among them, the historical derivative information includes the unnecessary operation time factor, unnecessary operation probability, necessary operation time factor, necessary operation probability, and scenario characteristics as described above. This part of the historical derivative information can be obtained through certain calculations, statistics, or conversions of the first historical feature parameters, and can be used to represent the habits or rules in the historical use process of the air conditioner.

[0069] In this embodiment, the time factor information can refer to the date or time when the user uses the air conditioner, such as December 12th, 8 pm on November 20th, and can also be a time interval such as 11 am to 13 pm on a certain day, 17 pm to 19 pm in the evening, 22 pm to 0 am at night, etc., and can also be holidays, solar terms, or time periods, such as the Spring Equinox, National Day. Therefore, the unnecessary operation time factor and the necessary operation time factor refer to the time factor expansion of the two classification results respectively after determining the corresponding classification results according to the power consumption information corresponding to the operation instruction, so as to find out the time rule of the user controlling the air conditioner.

[0070] The non-essential operation probability and the essential operation probability refer to the probabilities triggered by two classification results within a preset time period in the historical usage records. Using the time period to divide the interval, it represents the probability that the input operation instruction is a non-essential operation under the historical time period. Relatively, the probability of essential operations can also be directly obtained. For example, in the historical usage records of an air conditioner, within each working day from Monday to Friday, the probability that the input operation instruction is a non-essential operation is 5 / 9, which means that the user turns on the air conditioner N times (N is only an example) within the time period from Monday to Friday, and there are a total of 9 operations such as parameter adjustment or turning on / off, among which 5 are high-power consumption operations. Then the probability of non-essential operations within the preset time period can be determined. Similarly, the probability that the input operation instruction is an essential operation within each working day from Monday to Friday can be relatively obtained as 4 / 9. Another example is to break it down to the hour. From 12:00 to 13:00 on Monday to Friday, the probability that the input operation instruction is a non-essential operation is 7 / 8, which means that the user turns on the air conditioner N times within the time period from 12:00 to 13:00 on Monday to Friday, and there are a total of 8 operations such as parameter adjustment or turning on / off the lights, among which 7 are high-power consumption operations. Then the probability of non-essential operations within the preset time period is 7 / 8 and the probability of essential operations is 1 / 8 can be determined.

[0071] Among them, the scenario feature refers to the scenario capture feature when two classification results are triggered in the historical usage records of the air conditioner. For example, it includes how many minutes have passed since the start of the operation for this operation, the operating state or stage of the air conditioner to which this operation belongs, the attribute of this operation (switch / temperature adjustment), the interval duration of multiple historical operations, whether it is a continuous operation, and whether there have been corresponding non-essential operations for N consecutive days up to the current time.

[0072] S233: Determine the second historical feature parameter corresponding to the non-essential operation according to the first historical feature parameter and the historical derivative information.

[0073] It can be known that in order to improve the accuracy of the classification model, the model can be trained by constructing more and better features. Therefore, the second historical feature parameter refers to a set of more features constructed on the basis of the first historical feature parameter and the historical derivative information.

[0074] Specifically, S233 can be implemented through the following steps: perform feature transformation on the first historical feature parameter and the historical derivative information to obtain a high-dimensional feature parameter set; perform feature extraction on the high-dimensional feature parameter set to determine the second historical feature parameter corresponding to the non-essential operation.

[0075] In many machine learning competitions, the training set (features + categories) is usually directly provided. For this, "transformation" operations can be performed on the given features to construct more features. In this embodiment, the directly provided training set is equivalent to the first historical feature parameters, historical derivative information, and the corresponding classification results. Therefore, by performing feature transformation on the first historical feature parameters and historical derivative information, a high-dimensional feature parameter set composed of more features is constructed.

[0076] Among them, feature transformation can adopt methods such as discrete feature transformation, non-linear transformation, and polynomial transformation. Since the feature parameters or derivative information obtained in the foregoing steps belong to different types, different methods can be used for transformation for different types of features. For example, the feature data set composed of the first historical feature parameters and historical derivative information includes (x 1 , x 2 ). After feature transformation, the obtained feature set may include (x 1 , x 2 , x 1 2 , x 1 x 2 , x 2 2 , 1), etc., so as to construct more features.

[0077] It can be known that due to different transformation methods, the features after the above transformation may have a relatively high dimension, contain more redundancy, and have a relatively low interpretability in the feature set. Therefore, it is also necessary to perform certain high-interpretability feature extraction on the transformed features, obtain features with a high variance interpretation degree by reducing the high dimension, such as using methods such as PCA (Principal Component Analysis), SVD (Singular Value Decomposition), FA (factor analysis), and ICA (Independent Component Correlation Algorithm). Finally, some of the most important features are retained from the high-dimensional data, which corresponds to determining the second historical feature parameters corresponding to non-essential operations in this embodiment.

[0078] S234: If the power consumption information is less than or equal to the preset power consumption information, determine that the classification result corresponding to the historical air conditioner instruction operation is an essential operation, and calculate the historical derivative information of the air conditioner according to the air conditioner attribute parameters, historical air conditioner setting parameters, historical air conditioner operation parameters, and historical environment parameters.

[0079] Among them, the preset power consumption information can be expressed as the power consumption in the normal operation state, that is, the power consumption when there is no obvious change in power consumption caused by the user's non-essential operations. Therefore, in this embodiment, when the power loss caused by the historical air conditioner operation instruction is less than or equal to the preset power consumption information, it indicates that the classification result of the operation instruction at that historical moment should be an essential operation.

[0080] S235: Determine the second historical feature parameter corresponding to the essential operation according to the first historical feature parameter and the historical derivative information.

[0081] As can be seen from the above, essential operations and non-essential operations are two relative classification results. For an air conditioner operation instruction, it is either an essential operation or a non-essential operation. Therefore, the historical derivative information and the second historical feature parameter under essential operations should correspond to the various information parameters under non-essential operations, and can be calculated by the method described above, which will not be elaborated here.

[0082] S24: Input the classification result corresponding to the historical air conditioner operation instruction and the second historical feature parameter into the established classification model to train the classification model.

[0083] Among them, the classification model used in this embodiment is a model established based on the K-Nearest Neighbors (KNN) classification algorithm, also known as K-Nearest Neighbours, the K-value nearest neighbor classification algorithm. Different from other supervised algorithms, KNN is a non-parametric and instance-based algorithm. Non-parametric means that it does not make any speculation on the underlying data distribution, and instance-based means that it does not explicitly learn a model, but selects to memorize the training instances. Therefore, KNN is often referred to as a lazy algorithm, that is, KNN does not pre-generate a classification or prediction model for the prediction of new samples, but performs the construction of the model and the prediction of unknown data simultaneously.

[0084] Specifically, S24 can be implemented through Figure 4 the method steps shown, and the specific steps include:

[0085] S241: Use a part of the classification result corresponding to the historical air conditioner operation instruction and the second historical feature parameter as the training data set, and input it into the established classification model to train the classification model.

[0086] Among them, 70% of the second historical feature parameter can be used as the training data set. According to the above, due to the nature of the KNN algorithm, inputting the training data set and the corresponding classification result into the established model for training is actually memorizing and storing these "instances". When classifying and predicting samples, the training data set stored in memory is directly used for synchronous calculation.

[0087] S242: Use the classification result corresponding to the historical air conditioner operation instruction and the other part of the second historical feature parameter as a test data set to test the trained classification model to determine the target neighbor parameter of the classification model.

[0088] Among them, 30% of the second historical feature parameter can be used as the test data set. It can be known that the KNN algorithm classifies by majority voting. Therefore, it is necessary to determine the number of "people" voting in the "majority voting", that is, the target neighbor parameter k of the classification model to be determined in this embodiment, that is, the k that makes the classification prediction accuracy of the model the highest.

[0089] Among them, when training and testing the model, the second historical feature parameter corresponding to the historical air conditioner operation instruction should be used as a sample for input. In this embodiment, for the convenience of understanding and explanation, the feature parameter is directly used as the input object for description.

[0090] Optionally, S242 can be implemented through Figure 5 the method steps shown, and the specific steps include:

[0091] S2421: Respectively use multiple preset neighbor parameters as calculation criteria, use the classification result corresponding to the historical air conditioner operation instruction and the other part of the second historical feature parameter as a test data set, and input it into the trained classification model to output the classification result corresponding to the test data set.

[0092] Among them, the preset neighbor parameters can be 1, 3, 5, 7... X. In this embodiment, 1 to X are used as neighbor parameters respectively, and the trained classification model is used to output the classification prediction results of all test data sets.

[0093] S2422: Detect and confirm the classification result corresponding to the test data set to determine the multiple accuracies corresponding to the multiple preset neighbor parameters.

[0094] In this embodiment, since both the test data set and the training data set come from the second historical feature parameter, and the classification result of the historical air conditioner operation instruction corresponding to the second historical feature parameter is also known, it is possible to compare the classification prediction results of all test data sets output by the classification model with the known and correct classification results, so as to determine the prediction accuracy under each preset neighbor parameter.

[0095] In an application scenario, for example, when 1 is used as the nearest neighbor parameter k, 80 of the classification prediction results of the model output test data set are the same as the known classification results, and 20 are different from the known classification results. Then the accuracy of the preset nearest neighbor parameter of 1 is 80%. Similarly, the accuracy when 3, 5, 7...X are used as the nearest neighbor parameter k is calculated.

[0096] S2423: Sort the multiple preset neighbor parameters corresponding to the multiple accuracy rates, so as to use the preset neighbor parameter corresponding to the maximum accuracy rate as the target neighbor parameter of the classification model.

[0097] For example, the accuracy of the preset neighbor parameter of 1 is 80%, the accuracy of the preset neighbor parameter of 3 is 75%, the accuracy of the preset neighbor parameter of 5 is 70%, the accuracy of the preset neighbor parameter of 7 is 85%, the accuracy of the preset neighbor parameter of 9 is 70%, etc., and the preset neighbor parameters corresponding to all X accuracy rates are further sorted, so that the maximum accuracy rate is used as the target neighbor parameter of the classification model. For example, the above 85% is obviously ranked at the front, so 7 is determined as the target neighbor parameter k. The above is only an example, and the specific needs need to be determined according to the actual situation.

[0098] S25: Detect and confirm that the air-conditioning operation instruction is received, and obtain the current characteristic parameters of the air-conditioning.

[0099] The characteristic parameters include air conditioning attribute parameters, air conditioning setting parameters, air conditioning operation parameters, environmental parameters and historical derivative information; wherein the historical derivative information is determined by the historical characteristic parameters.

[0100] S26: Input the current feature parameters into the trained classification model, and obtain the classification result output by the classification model.

[0101] Specifically, S26 can be Figure 6 The method steps shown are implemented, and the specific steps include:

[0102] S261: Input the current feature parameter into the trained classification model, and calculate multiple Euclidean distances between the current sample corresponding to the current feature parameter and multiple historical samples corresponding to the second historical feature parameter.

[0103] Specifically, the calculation formula of Euclidean distance is: Among them, x i Indicates the current sample corresponding to the current feature parameter. Since there may be multiple current feature parameters, the Euclidean distance calculation needs to be performed in units of samples; x i represents multiple historical samples corresponding to the second historical feature parameter, so this distance function is to calculate x i With other samples x j The distance between.

[0104] Optionally, in addition to calculating the distance between two feature parameters using the Euclidean distance function, functions such as cosine distance, Hamming distance, and Manhattan distance can also be used for calculation. It should be noted that all feature parameters in this embodiment need to be subjected to comparable quantization processing so that all feature parameters can perform distance calculation, and normalization processing is also required to make the influence of all feature parameters on distance calculation smaller.

[0105] S262: Sort multiple Euclidean distances and determine k classification results corresponding to the minimum Euclidean distance according to the target nearest neighbor parameter.

[0106] In an application scenario, for example, still taking k = 7 as an example, sort X Euclidean distances, and select 7 historical samples from the end arranged from large to small to obtain 7 corresponding classification results.

[0107] S263: Vote on the k classification results corresponding to the minimum Euclidean distance to determine the classification result of the current sample.

[0108] In this embodiment, according to the classification decision rule (such as the majority voting method), the classification result of the current sample is determined from the classification results corresponding to 7 historical samples. For example, among the 7 classification results, 4 are non-essential operations and 3 are essential operations. According to the majority voting method, since 4 > 3, it can be determined that the classification result of the current sample is a non-essential operation, that is, the classification result of the current feature parameter is a non-essential operation, which means that the air conditioner operation instruction at the current moment may cause greater power consumption. Another example is that among the 7 classification results, 2 are non-essential operations and 5 are essential operations. According to the majority voting method, since 5 > 2, it can be determined that the classification result of the current sample is an essential operation, that is, the classification result of the current feature parameter is an essential operation, which means that the air conditioner at the current moment will not cause greater power consumption.

[0109] In some embodiments, in order to weaken the influence of the target nearest neighbor parameter k on classification, a weighted voting method can also be used for calculation, introducing a time weight to the classification result corresponding to each historical sample. Specifically, S263 can be implemented through the following steps: respectively obtain the weights of k classification results; respectively calculate the sum of the weights of the same classification results among k classification results; use the classification result corresponding to the largest weight sum as the classification result corresponding to the current sample.

[0110] Among them, the time difference is used as the weight of each classification result, specifically the reciprocal of the difference between the date corresponding to the historical sample and the date corresponding to the current sample. For example, it is calculated that the nearest historical sample is A, and the time when A occurred is Tuesday obtained from the second historical feature parameter corresponding to A, while the current time is Friday obtained from the current feature parameter corresponding to the current sample. Therefore, the weight of the classification result corresponding to the historical sample A is 1 / (5 - 2) = 1 / 3, and the time weights of k classification results are calculated in this way. In some embodiments, the calculation of the time weight can also be performed in hours and can be specifically set according to the actual situation.

[0111] It should be noted that when the difference between the date corresponding to the historical sample and the date corresponding to the current sample is 1 day, it can be selected not to include it in the calculation of the sum of the weights of subsequent identical classification results, or it can be selected to set a specific time weight for subsequent calculation, such as setting it to be greater than 1 / 2.

[0112] Further, the time weights corresponding to all non-essential operations and essential operations are respectively summed, and the classification result with the largest sum of weights is used as the classification result of the current sample; for example, the sum of the time weights of the classification result of non-essential operations is 1 / 10, and the sum of the time weights of the classification result of essential operations is 1 / 15. Since 1 / 10 > 1 / 15, it can be determined that the classification result of the current sample is non-essential operations. In this way, the influence of the target nearest neighbor function k on the classification result is weakened, so that the subsequent obtained classification result can better meet the habits and needs of users.

[0113] Optionally, the influence of the target nearest neighbor function k on the classification result can also be weakened by introducing distance weights. By setting weights for the k smallest Euclidean distances, the closer the distance, the greater the weight. The distance weight can be specifically set as the reciprocal of the square of the Euclidean distance. For example, the Euclidean distance between the current sample and the nearest historical sample B is 5 (units), then the distance weight of the historical sample B is 1 / 25, and the distance weights of k classification results are calculated in this way.

[0114] Further, the distance weights corresponding to non-essential operations and essential operations are also respectively summed, and the classification result with the largest sum of weights is used as the classification result of the current sample; for example, the sum of the distance weights of the classification result of non-essential operations of historical samples is 1 / 100, and the sum of the distance weights of the classification result of essential operations of historical samples is 1 / 150. Since 1 / 100 > 1 / 150, it can be determined that the classification result of the current sample is non-essential operations. Therefore, by introducing time weights or distance weights, the influence of k on classification can be weakened, and when the number of samples with non-essential operations is significantly less than the number of samples with essential operations, it can be changed due to the addition of weights, thereby reducing the error rate of classification prediction.

[0115] Optionally, the classification model can be saved on the server corresponding to the air conditioner or locally on the air conditioner. In this embodiment, the classification model is saved on the server. At this time, step S26 can be implemented through the following steps: sending the current feature parameters to the server so that the server inputs the current feature parameters into the trained classification model and obtains the classification result output by the classification model; receiving the classification result sent by the server.

[0116] Specifically, after obtaining the current feature parameters of the air conditioner, the air conditioner can send the current feature parameters to the server so that the server inputs the current feature parameters into the trained classification model. Further, the server obtains the classification result output by the classification model and sends the classification result to the air conditioner side so that the air conditioner can perform corresponding control and use.

[0117] S27: If the classification result indicates that the air conditioner operation instruction is a non-essential operation, a prompt is given.

[0118] Among them, prompting the user can include methods such as reminding, recommending, or specifying suggestions. For example, through intelligent interaction, messages are pushed on the user-side app for reminder or suggestion. The user can install a home appliance app corresponding to the air conditioner on the mobile device and bind it to the air conditioner. When the classification result indicates a non-essential operation, a confirmation message can be pushed to the terminal display interface through this home appliance app. The confirmation message includes the consequences of this air conditioner operation instruction and the operation suggestions for this air conditioner operation instruction. The user can directly view and confirm this classification result through the home appliance app to determine whether to continue to execute according to the operation instruction or accept the suggestions in the confirmation message and execute. Among them, the recommendation or specified suggestion can be calculated based on the feature parameters at the current moment, so as to recommend a relatively reasonable adjustment method to the user. For example, one or more relatively power-saving set temperatures are recommended for the user to choose.

[0119] It is also possible to set the home appliance trusteeship mode for the air conditioner, which is turned on through the air conditioner display panel or the home appliance app. In this mode, all input operation instructions are invalid, and the air conditioner continues to run with the original set parameters, which can prevent unnecessary power consumption caused by children in the family frequently adjusting the remote control or wall buttons, and avoid waste of electricity.

[0120] Optionally, since the user may not be able to check the mobile phone in time, the method of app message push is not timely at this time. Therefore, the user's current operation instruction can also be prompted on-site through voice broadcast or air conditioner screen display, so that the user can confirm this operation in time. At the same time, voice broadcast or air conditioner screen display can also be used to give operation suggestions to the user to provide a more comfortable and energy-saving choice.

[0121] Further, on the basis of prompting the user through voice broadcast, air conditioner screen display, or app push, the display of the electricity bill can be set. For example, on the home appliance app or the air conditioner screen display, the unnecessary power consumption and the necessary power consumption are displayed to form an obvious contrast, so as to specifically inform the user of the electricity cost loss caused by this unnecessary operation and prompt its importance. A threshold for unnecessary power consumption can also be set. For example, taking one week as a time period, the threshold for unnecessary power consumption represents the part that is allowed to exceed the normal power consumption within one week. Within the loss threshold, only display without prompting to ensure the smooth execution of unnecessary operations at some specific moments and improve the user experience.

[0122] In some other embodiments, when the classification result indicates that the air conditioner operation instruction is a necessary operation, that is, the current user operation theoretically will not cause more power consumption, then no action needs to be taken, and the acquisition of the air conditioner operation instruction is continued to continuously judge subsequent control events.

[0123] Optionally, after S27, it may further include: detecting and confirming the receipt of the feedback operation instruction, and then using the current characteristic parameters corresponding to the air conditioner operation instruction to correct the classification model.

[0124] Among them, the feedback operation instruction refers to the operation feedback made by the user according to the prompt of the air conditioner. For example, when the user executes the prompt, the feedback operation instruction at this time is the air conditioner operation instruction corresponding to invalidity to reduce energy waste. It can also be, for example, that the user ignores the prompt or refuses the prompt. At this time, although the user already knows the loss consequence brought by the air conditioner operation instruction, due to actual needs, the user has to perform this operation. At this time, the feedback operation instruction is to continue to execute the corresponding air conditioner operation instruction to meet the user's needs.

[0125] Further, after the user makes feedback such as execution, refusal, or ignoring the prompt, the air conditioner operation instruction can be classified again according to the results of different feedback operation instructions. For example, according to the air conditioner operation instruction input by the user, the current characteristic parameters are obtained for classification prediction, and the classification result is an unnecessary operation. However, because the user ignores or refuses the prompt subsequently, the classification result of the air conditioner operation instruction can be determined again, and it is determined to be a necessary operation according to the feedback instruction. After the secondary determination, the current characteristic parameters and the secondary classification result after this feedback operation instruction are incorporated into the training data set to correct the model, so that the model more conforms to the user's usage habits and improves the accuracy of classification prediction.

[0126] In the practical application of this embodiment, it may be to receive an instruction from the user to monitor non - energy - saving operations, thereby starting the steps of S21 - S27. For example, the user turns on the air conditioner and enables the function of the "Energy - saving Cloud Butler" so that the air conditioner starts to execute the above - mentioned steps.

[0127] Refer to Figure 7 , Figure 7 FIG. is a schematic structural diagram of an embodiment of an air - conditioning device provided by the present application. The air - conditioning device 70 of this embodiment includes a processor 71 and a memory 72. The processor 71 is coupled to the memory 72. Among them, the memory 72 is used to store a computer program executed by the processor 71, and the processor 71 is used to execute the computer program to implement the following method steps:

[0128] Detect and confirm that an air - conditioning operation instruction is received, and obtain the current characteristic parameters of the air conditioner; among them, the characteristic parameters include air - conditioner attribute parameters, air - conditioner setting parameters, air - conditioner operation parameters, environmental parameters, and historical derivative information, and the historical derivative information is determined by historical characteristic parameters; input the current characteristic parameters into a trained classification model, and obtain a classification result output by the classification model; among them, the classification model is trained based on historical characteristic parameters and corresponding classification results, and the classification results include necessary operations and non - necessary operations; if the classification result indicates that the air - conditioning operation instruction is a non - necessary operation, then give a prompt.

[0129] Refer to Figure 8 , Figure 8 FIG. is a schematic structural diagram of an embodiment of a computer - readable storage medium provided by the present application. The computer - readable storage medium 80 of this embodiment is used to store a computer program 81. When the computer program 81 is executed by a processor, it is used to implement the following method steps:

[0130] Detect and confirm that an air - conditioning operation instruction is received, and obtain the current characteristic parameters of the air conditioner; among them, the characteristic parameters include air - conditioner attribute parameters, air - conditioner setting parameters, air - conditioner operation parameters, environmental parameters, and historical derivative information, and the historical derivative information is determined by historical characteristic parameters; input the current characteristic parameters into a trained classification model, and obtain a classification result output by the classification model; among them, the classification model is trained based on historical characteristic parameters and corresponding classification results, and the classification results include necessary operations and non - necessary operations; if the classification result indicates that the air - conditioning operation instruction is a non - necessary operation, then give a prompt.

[0131] It should be noted that the method steps executed by the computer program 81 of this embodiment are based on the above - mentioned method embodiment, and their implementation principles and steps are similar. Therefore, when the computer program 81 is executed by a processor, it can also implement other method steps in any of the above - mentioned embodiments, which will not be elaborated here.

[0132] When the embodiments of the present application are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the methods described in various embodiments of the present application. The foregoing storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs.

[0133] The above are only the embodiments of the present application, and do not limit the patent scope of the present application. Any equivalent structural or equivalent process transformation made according to the content of the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present application.

Claims

1. An operation prompt method for an air conditioner, characterized in that, the method includes: detecting and confirming receipt of an air conditioner operation instruction, and obtaining the current characteristic parameters of the air conditioner; wherein, the characteristic parameters include air conditioner attribute parameters, air conditioner setting parameters, air conditioner operation parameters, environmental parameters, and historical derivative information, and the historical derivative information is determined by historical characteristic parameters; the historical derivative information includes at least one of a non-essential operation time factor, a non-essential operation probability, an essential operation time factor, an essential operation probability, and a scenario characteristic at a historical moment; inputting the current characteristic parameters into a trained classification model, and obtaining a classification result output by the classification model; wherein, the classification model is trained based on the historical characteristic parameters and the corresponding classification results, and the classification results include essential operations and non-essential operations; if the classification result indicates that the air conditioner operation instruction is a non-essential operation, a prompt is given.

2. The method according to claim 1, characterized in that, the classification model is trained based on the historical characteristic parameters and the corresponding classification results, and further includes: establishing a classification model; detecting and confirming receipt of a historical air conditioner operation instruction, and obtaining the first historical characteristic parameters of the air conditioner; wherein, the first historical characteristic parameters include air conditioner attribute parameters, historical air conditioner setting parameters, historical air conditioner operation parameters, and historical environmental parameters; determining the classification result corresponding to the historical air conditioner operation instruction according to the first historical characteristic parameters, and obtaining the second historical characteristic parameters corresponding to the classification result; inputting the classification result corresponding to the historical air conditioner operation instruction and the second historical characteristic parameters into the established classification model to train the classification model.

3. The method according to claim 2, characterized in that, determining the classification result corresponding to the historical air conditioner operation instruction according to the first historical characteristic parameters, and obtaining the second historical characteristic parameters corresponding to the classification result; calculating the power consumption information at the moment corresponding to the historical air conditioner operation instruction according to the air conditioner attribute parameters, historical air conditioner setting parameters, historical air conditioner operation parameters, and historical environmental parameters; if the power consumption information is greater than the preset power consumption information, determining that the classification result corresponding to the historical air conditioner operation instruction is a non-essential operation, and calculating the historical derivative information of the air conditioner according to the air conditioner attribute parameters, historical air conditioner setting parameters, historical air conditioner operation parameters, and historical environmental parameters; determining the second historical characteristic parameters corresponding to the non-essential operation according to the first historical characteristic parameters and the historical derivative information.

4. The method according to claim 3, characterized in that, determining the second historical characteristic parameters corresponding to the non-essential operation according to the first historical characteristic parameters and the historical derivative information, includes: performing feature transformation on the first historical characteristic parameters and the historical derivative information to obtain a high-dimensional feature parameter set; performing feature extraction on the high-dimensional feature parameter set to determine the second historical characteristic parameters corresponding to the non-essential operation.

5. The method according to claim 3, characterized in that, the method further includes: If the power consumption information is less than or equal to the preset power consumption information, determine that the classification result corresponding to the historical air conditioner instruction operation is a necessary operation, and calculate the historical derivative information of the air conditioner according to the air conditioner attribute parameters, historical air conditioner setting parameters, historical air conditioner operation parameters, and historical environment parameters; Determine the second historical feature parameter corresponding to the necessary operation according to the first historical feature parameter and the historical derivative information.

6. The method according to claim 2, wherein, The inputting the classification result corresponding to the historical air conditioner operation instruction and the second historical feature parameter into the established classification model to train the classification model includes: Taking the classification result corresponding to the historical air conditioner operation instruction and a part of the second historical feature parameter as a training data set, and inputting it into the established classification model to train the classification model; Taking the classification result corresponding to the historical air conditioner operation instruction and the other part of the second historical feature parameter as a test data set to test the trained classification model to determine the target neighbor parameter of the classification model.

7. The method according to claim 6, wherein, The inputting the classification result corresponding to the historical air conditioner operation instruction and the other part of the second historical feature parameter as a test data set to test the trained classification model to determine the target neighbor parameter of the classification model includes: Taking the classification result corresponding to the historical air conditioner operation instruction and the other part of the second historical feature parameter as a test data set, and inputting it into the trained classification model respectively with multiple preset neighbor parameters as calculation criteria to output the classification result corresponding to the test data set; Detect and confirm the classification result corresponding to the test data set to determine the multiple accuracies corresponding to the multiple preset neighbor parameters; Sort the multiple preset neighbor parameters corresponding to the multiple accuracies, and take the preset neighbor parameter corresponding to the maximum accuracy as the target neighbor parameter of the classification model.

8. The method according to claim 6, wherein, The inputting the current feature parameter into the trained classification model and obtaining the classification result output by the classification model includes: Inputting the current feature parameter into the trained classification model, and calculating multiple Euclidean distances between the current sample corresponding to the current feature parameter and multiple historical samples corresponding to the second historical feature parameter; Sort the multiple Euclidean distances, and determine k classification results corresponding to the minimum Euclidean distance according to the target neighbor parameter; Vote on the k classification results corresponding to the minimum Euclidean distance to determine the classification result of the current sample.

9. The method according to claim 8, wherein, The voting on the k classification results corresponding to the minimum Euclidean distance to determine the classification result of the current sample includes: Respectively obtain the weights of the k classification results; Respectively calculate the sum of the weights of the same classification results among the k classification results; Use the classification result corresponding to the sum of the maximum weights as the classification result corresponding to the current sample.

10. The method according to claim 1, wherein, the step of inputting the current feature parameter into the trained classification model and obtaining the classification result output by the classification model includes: sending the current feature parameter to a server, so that the server inputs the current feature parameter into the trained classification model and obtains the classification result output by the classification model; receiving the classification result sent by the server.

11. The method according to claim 1, wherein, the method further includes: detecting and confirming that a feedback operation instruction is received, and then correcting the classification model by using the current feature parameter corresponding to the air conditioner operation instruction.

12. The method according to claim 1, wherein, the air conditioner operation instruction refers to an instruction received by the air conditioner, sent by an air conditioner control terminal, and used to change the operating state of the compressor of the air conditioner.

13. The method according to claim 1, wherein, the air conditioner attribute parameter is the basic attribute information of the air conditioner obtained according to the model or identification of the air conditioner; or, the air conditioner setting parameter is the working index that the air conditioner needs to reach under the control of the air conditioner operation instruction; or, the air conditioner operation parameter is the working condition information that the air conditioner reaches in real time under the control of the air conditioner operation instruction; or, the environmental parameter is the environmental temperature information when the air conditioner is operating.

14. The method according to claim 13, wherein, the air conditioner attribute parameter includes at least one of the number of horsepower, cooling power, heating power, cooling capacity, heating capacity, and energy efficiency grade of the air conditioner; or, the air conditioner setting parameter includes at least one of the temperature setting parameter and the wind speed setting parameter of the air conditioner; or, the air conditioner operation parameter includes at least one of the compressor operation frequency, indoor fan speed, outdoor fan speed, outdoor unit operation frequency, and power value; or, the environmental parameter includes at least one of the indoor environmental temperature, outdoor environmental temperature, indoor environmental humidity, and outdoor environmental humidity.

15. An air conditioner device, wherein, it includes a processor and a memory, wherein the memory is used to store a computer program, and when the computer program is executed by the processor, it is used to implement the operation prompting method of the air conditioner according to any one of claims 1-14.

16. A computer-readable storage medium, wherein, it is used to store a computer program, and when the computer program is executed by a processor, it is used to implement the operation prompting method of the air conditioner according to any one of claims 1-14.

Images