CN112303819A

CN112303819A - Air conditioner and control method

Info

Publication number: CN112303819A
Application number: CN202011019870.5A
Authority: CN
Inventors: 盛凯; 矫晓龙
Original assignee: Qingdao Hisense Hitachi Air Conditioning System Co Ltd
Current assignee: Qingdao Hisense Hitachi Air Conditioning System Co Ltd
Priority date: 2020-09-24
Filing date: 2020-09-24
Publication date: 2021-02-02

Abstract

The present invention provides an air conditioner and a control method, the air conditioner includes a compressor, an indoor heat exchanger, a sensor, and a controller configured to: acquiring operating data detected by the sensor; inputting the operation data into an SVM model to obtain a predicted energy efficiency ratio; and adjusting the control parameters of the air conditioner according to the predicted energy efficiency ratio. By applying the technical scheme, the energy efficiency of the air conditioner can be predicted more accurately, and then timely control is adopted, so that the energy consumption is saved on the premise of meeting the comfort.

Description

Air conditioner and control method

Technical Field

The present disclosure relates to the field of air conditioner control, and more particularly, to an air conditioner and a control method.

Background

With the improvement of living standard of people, the air conditioner has become a necessary product in each family, the opening frequency of the air conditioner also becomes high, and meanwhile, the intelligent level of the air conditioner also has higher requirements. The intelligent level of the air conditioner is not only embodied in the aspect of intelligent control, but also is an important aspect for controlling the energy consumption of the air conditioner. All large air conditioner manufacturers strive to improve the energy efficiency ratio of the air conditioner so that the air conditioner can operate under higher energy efficiency and meet the requirement of energy conservation. The method mainly comprises the following two aspects of research, wherein firstly, the method starts from the forward control of the air conditioner, obtains detailed operation parameters of the air conditioner through experiments, adjusts a control algorithm to achieve energy-saving control, and the other aspect is to predict the energy efficiency ratio of the air conditioner to be used as a feedback input of an air conditioner control system to complete more accurate control of the air conditioner. At present, various methods are used for predicting the air conditioner energy efficiency, such as linear regression, Bayesian estimation algorithm, genetic algorithm and the like, but the air conditioner energy efficiency is influenced by various factors and parameters, and is a very complex nonlinear system, and under certain conditions, the collection of air conditioner data is not very convenient, and a large amount of data cannot be obtained, so that the accurate prediction effect is difficult to obtain through the traditional method.

In summary, how to provide an air conditioner capable of accurately predicting the energy efficiency of the air conditioner is a technical problem to be solved urgently.

Disclosure of Invention

Because the problem that energy consumption control of the air conditioner is not accurate due to inaccurate energy efficiency prediction of the air conditioner exists in the prior art, the invention provides an air conditioner, which comprises:

the compressor is used for compressing low-temperature and low-pressure refrigerant gas into high-temperature and high-pressure refrigerant gas and discharging the high-temperature and high-pressure refrigerant gas to the condenser;

an indoor heat exchanger operating as a condenser or an evaporator;

a sensor for detecting operation data of the air conditioner;

the controller is configured to:

acquiring operating data detected by the sensor;

inputting the operation data into an SVM model to obtain a predicted energy efficiency ratio;

and adjusting the control parameters of the air conditioner according to the predicted energy efficiency ratio.

In some embodiments, the operational data specifically includes one or more of evaporator leaving water temperature, condenser entering water temperature, condenser leaving water temperature, heat exchanger entering water temperature, condenser loop temperature, heat exchanger leaving water temperature, evaporator loop temperature, building entering water temperature, building leaving water temperature, steam heating capacity, evaporator flow rate, condenser temperature differential characterization value, and refrigerant temperature.

In some embodiments, the SVM model is created by:

acquiring historical parameter variables of the air conditioner and establishing a data set, wherein the historical parameter variables comprise historical operating data detected by the sensor and corresponding historical energy efficiency ratio;

carrying out standardization processing on the data set, and randomly dividing the data set into a training set and a verification set;

establishing an initial SVM model and inputting the training set for training;

verifying the trained initial SVM model by using the verification set;

and taking the initial SVM model with the prediction precision reaching a preset value as the SVM model.

In some embodiments, the initial SVM model is built based on a radial basis kernel function;

the prediction accuracy is determined based on the root mean square error and the R-square score.

In some embodiments, the building of the SVM model further comprises: and adjusting a penalty factor and a gamma value based on the prediction precision of the training result and the verification result of the initial SVM model.

Accordingly, the present invention also provides a method for controlling an air conditioner, the method being applied to an air conditioner including a compressor and an indoor heat exchanger, the air conditioner further including a sensor and a controller, the method including:

acquiring operating data detected by the sensor;

In some embodiments, the SVM model is created by:

establishing an initial SVM model and inputting the training set for training;

verifying the trained initial SVM model by using the verification set;

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

Fig. 1 is a schematic structural view illustrating an air conditioner according to an embodiment of the present invention;

FIG. 2 is a schematic diagram illustrating a support vector machine model proposed by an embodiment of the present invention;

FIG. 3 is a graph illustrating the predicted effect of modeling using a linear kernel function according to an embodiment of the present invention;

FIG. 4 is a diagram illustrating the predicted effect of modeling using a polynomial kernel in an embodiment of the present invention;

FIG. 5 is a graph illustrating the predicted effect of modeling using radial basis kernel functions according to an embodiment of the present invention;

fig. 6 is a flowchart illustrating an air conditioner control method according to an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

In the description of the present application, it is to be understood that the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", and the like indicate orientations or positional relationships based on those shown in the drawings, and are only for convenience in describing the present application and simplifying the description, but do not indicate or imply that the referred device or element must have a particular orientation, be constructed in a particular orientation, and be operated, and thus should not be construed as limiting the present application.

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

In the description of the present application, it is to be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly, e.g., as meaning either a fixed connection, a removable connection, or an integral connection; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meaning of the above terms in the present application can be understood in a specific case by those of ordinary skill in the art.

To further describe the solution of the present application, fig. 1 is a schematic structural diagram of an air conditioner according to an embodiment of the present application, including:

and the compressor is used for compressing the low-temperature low-pressure refrigerant gas into high-temperature high-pressure refrigerant gas and discharging the high-temperature high-pressure refrigerant gas to the condenser.

The air conditioner performs a refrigeration cycle of the air conditioner by using a compressor, a condenser, an expansion valve, and an evaporator. The refrigeration cycle includes a series of processes involving compression, condensation, expansion, and evaporation, and supplies refrigerant to the air that has been conditioned and heat-exchanged.

The compressor compresses a refrigerant gas in a high-temperature and high-pressure state and discharges the compressed refrigerant gas. The discharged refrigerant gas flows into the condenser. The condenser condenses the compressed refrigerant into a liquid phase, and heat is released to the surrounding environment through the condensation process.

An indoor heat exchanger operating as a condenser or an evaporator.

The outdoor unit of the air conditioner includes a portion of the refrigeration cycle including the compressor and the outdoor heat exchanger, the indoor unit of the air conditioner includes the indoor heat exchanger, and the expansion valve may be provided in either the indoor unit or the outdoor unit.

The expansion valve expands the liquid-phase refrigerant in a high-temperature and high-pressure state condensed in the condenser into a low-pressure liquid-phase refrigerant. The evaporator evaporates the refrigerant expanded in the expansion valve and returns the refrigerant gas in a low-temperature and low-pressure state to the compressor. The evaporator can achieve a cooling effect by heat-exchanging with a material to be cooled using latent heat of evaporation of a refrigerant. The air conditioner can adjust the temperature of the indoor space throughout the cycle.

The indoor heat exchanger and the outdoor heat exchanger serve as a condenser or an evaporator. When the indoor heat exchanger is used as a condenser, the air conditioner is used as a heater in a heating mode, and when the indoor heat exchanger is used as an evaporator, the air conditioner is used as a cooler in a cooling mode.

And the sensor is used for detecting the operation data of the air conditioner.

Various sensors are arranged in the air conditioner, and the operation data of the air conditioner during operation can be obtained through the sensors.

The controller is configured to: acquiring operating data detected by the sensor; inputting the operation data into an SVM model to obtain a predicted energy efficiency ratio; and adjusting the control parameters of the air conditioner according to the predicted energy efficiency ratio.

The invention provides a method based on a Support Vector Machine (SVM) to predict energy efficiency, and then control parameters (such as set temperature, wind speed and the like) of an air conditioner are adjusted according to the predicted energy efficiency ratio (COP) so as to achieve the effect of saving energy consumption. COP refers to the ratio of the cooling capacity (or heating capacity) of the air conditioner to the effective input power, which cannot be obtained by a sensor, and historical COP can obtain a more accurate value by calculating the formula COP as the cooling capacity/loss power, but transient COP is difficult to obtain, which is also an advantage for energy efficiency prediction. The SVM method can be used for classification research, and can classify samples in a high-dimensional space so as to solve the problem that the nonlinear problem cannot be classified in a low-latitude space. The SVM is also suitable for regression prediction analysis, regression prediction can be carried out by using data of small samples, the algorithm is called Support Vector Regression (SVR), in some cases, the collection of air conditioner data is not very convenient, a large amount of data cannot be obtained, and the prediction advantage of the SVR is very obvious in the case. It should be noted that, because SVR belongs to one of SVMs, hereinafter referred to as SVM, and the SVM model of the present invention may also be operated in a PC computer or a cloud platform system, and perform data transmission with an air conditioner through network communication or wireless communication, so as to complete energy efficiency prediction.

The operation data detected by the sensor is an internal operation parameter variable and an environmental parameter variable of the air conditioner, and in some embodiments, the operation data specifically includes one or more of evaporator outlet water Temperature (TEI), evaporator outlet water Temperature (TEO), condenser inlet water Temperature (TCI), condenser outlet water Temperature (TCO), heat exchanger inlet water Temperature (TSI), condenser loop Temperature (TSO), heat exchanger outlet water Temperature (TBI), evaporator loop Temperature (TBO), building inlet water temperature (Cond Tons), building outlet water temperature (Cooling Tons), steam heating amount (kW), evaporator flow rate (TEA), condenser flow rate (TCA), condenser temperature difference characterization value (TRE), and refrigerant Temperature (TRC). The embodiment of the present invention adopts all types of operation data for explanation, and it should be noted that the embodiment of the present invention is only an implementation example, and all required data cannot be acquired in actual data acquisition, but the processing mode is the same.

In order to facilitate the input of the operation data, the operation data is usually required to be established into an input parameter vector and input into the SVM model after being subjected to a normalization process.

The input parameter vector established based on the 15 kinds of operation data is as follows:

x ═ TEI, TEO, TCI, TCO, TSI, TSO, TBI, TBO, CondTons, CoolingTons, kW, TEA, TCA, TRE, TRC. The normalization process can be a normalization process, and various methods can be used, such as one-hot encoding the data, then converting the data into a range interval of (0-1), removing unit limitation of the data, and converting the data into a dimensionless value.

In some embodiments of the present invention, the SVM model is built by:

acquiring historical parameter variables of the air conditioner and establishing a data set; carrying out standardization processing on the data set, and randomly dividing the data set into a training set and a verification set; establishing an initial SVM model and inputting the training set for training; verifying the trained initial SVM model by using the verification set; and taking the initial SVM model with the prediction precision reaching a preset value as the SVM model.

(1) Parameter variable acquisition

The historical parameter variables selected in the invention comprise historical operation data detected by the sensor and corresponding historical energy efficiency ratio. The historical operation data is obtained by a sensor, and the corresponding historical COP can be obtained by calculating the COP as the cooling capacity/loss power.

(2) Acquisition of a data set

The input data set is established based on the 15 kinds of operation data, and 15-dimensional vectors established by the 15 kinds of operation data are as follows:

X＝[TEI,TEO,TCI,TCO,TSI,TSO,TBI,TBO,CondTons,CoolingTons,kW,TEA,TCA,TRE,TRC]；

the historical data set of the air conditioner can be obtained by detecting the states of the air conditioner at different time points, and the input data set is X ═[ X1, X2 … Xi ], wherein i ═ N and N represents the total collection times. Based on the fact that the dimension of the input vector is 15 dimensions and the same historical COP data is used as a one-dimensional output variable, the algorithm model established by the embodiment of the invention predicts the value of the one-dimensional COP by inputting the 15-dimensional data set.

(3) Data processing

The obtained data set is subjected to normalization processing, normalization processing can be adopted, and various methods can be used, such as one-hot encoding is carried out on the data, then the data is converted into a range interval of (0-1), unit limitation of the data is removed, and the data is converted into a dimensionless numerical value.

The obtained data sets are randomly distributed according to the ratio of 8:2 to be used as training sets and verification sets, and the data sets can be distributed according to other ratios as required.

(4) Construction of SVM model

The idea of the SVM is to find a hyperplane to separate as many positive and negative examples as possible for a given training sample, and to select the optimal hyperplane based on the positive and negative examples being as far away from the hyperplane as possible, as shown in fig. 2. The distance of a point from the hyperplane can be expressed as the confidence or accuracy of the classification prediction. The SVM is to maximize the value of this interval. And the point on the dotted line is called the support vector (Supprot vector).

To take the geometric spacing that maximizes the support vector to the hyperplane, the objective function can be expressed as:

s.t.y_i(w^Tx_i+b)≥γ′(i＝1,2,...,m)

when the non-linear problem needs to be solved, a kernel function needs to be introduced. Some linearly indivisible datasets in a low dimensional space are mapped into a high dimensional space to be linearly separated with a greater probability. Usually, three kernel functions are available for modeling, which are a linear function, a polynomial function and a radial basis function, and the kernel functions are different, and the prediction effect of the model is also different. In some embodiments of the invention, the initial SVM model is built based on a radial basis kernel function.

And the initial SVM model is also provided with a penalty factor C and a gamma, wherein the C represents a penalty coefficient of the model to errors, and the gamma reflects the distribution of the data after being mapped to a high-dimensional feature space.

C characterizes how much you value outliers, the larger C the more important, the less likely you want to discard them. And when the C value is large, the punishment on the error classification is increased, and when the C value is small, the punishment on the error classification is reduced. The larger C is, the smaller margin is, and when C approaches infinity, the existence of classification errors is not allowed, so that overfitting is easy; the smaller C, the larger margin, and when C goes to 0, it means we do not pay attention to whether the classification is correct any more, and it is easy to under-fit. gamma is the kernel coefficient of the radial basis kernel function and must have a value greater than 0. With the increase of gamma, the classification effect on the test set is poor, and the classification effect on the training set is good, and overfitting easily occurs in generalization errors. Thus, in some embodiments, an initial SVM model is established and input to a training set for training; verifying the trained initial SVM model by using a verification set; and adjusting punishment factors and gamma values based on the training results and the prediction accuracy of the initial SVM model, wherein the initial SVM model with the prediction accuracy reaching a preset value is used as a final SVM model, and the specific preset value can be set according to actual needs.

The model prediction accuracy measurement indexes comprise the following indexes:

mean absolute error (mean absolute error), which is the mean of the absolute errors of all data points of a given data set;

mean squared error (mean squared error), the mean of the squares of the errors for all data points of a given data set;

median absolute error (mean absolute error): given the median of the errors of all data points of the data set, the interference of outliers can be eliminated;

explained square difference (extended variance score): the model is used for measuring the interpretability of the model to the fluctuation of the data set, and if the score is 1.0, the model is perfect.

R-side score (R2 score): the read R-square refers to a deterministic correlation coefficient used for measuring the prediction effect of the model on unknown samples, the best score is 1.0, and the value can also be a negative number.

In some embodiments of the invention, the prediction accuracy is determined based on the root mean square error and the R-square score.

The following description is based on a specific implementation scenario, and 5000 sets of data are collected for an air conditioner to obtain an initial data set. The partial samples of the data set collected are as follows:

the data set was randomly sampled, 20% as test and 80% as training.

All data sets were normalized.

And establishing an SVM model, training to obtain a trained model, and verifying by using a test set.

As shown in fig. 3-5, the prediction effect graphs using different kernel functions are obtained by modeling using linear kernel functions, polynomial kernel functions, and radial basis kernel functions, respectively. The data shown on the data map are normalized data, not true values, only to verify the predicted effect.

Model accuracy indexes under three kernel functions are as follows:

the default evaluation values for the linear kernel support vector machine are: 0.9777971050401596, respectively;

the R _ squared value of the linear kernel support vector machine is: 0.9777971050401596, respectively;

the mean square error of the linear kernel function support vector machine is 1.3668512599057803 e-06;

the average absolute error of the linear kernel function support vector machine is 0.0008922184055729438;

the default evaluation value for the polynomial kernel is: 0.9449249268521741, respectively;

the R _ squared value of the polynomial kernel is: 0.9449249268521741, respectively;

the mean square error of the polynomial kernel is 3.3905233194892276 e-06;

the mean absolute error of the polynomial kernel is 0.0011859746909909379;

the default evaluation values for the radial basis kernel function are: 0.9967230827566224, respectively;

the R _ squared value of the radial basis kernel is: 0.9967230827566224, respectively;

the mean square error of the radial basis kernel function is 2.0173308349290764 e-07;

the mean absolute error of the radial basis kernel function is 0.0003399103788662626.

From the above comparative analysis, it can be seen that the prediction accuracy of the SVM model established based on the radial basis function is the best.

Corresponding to the air conditioner in the embodiment of the present application, the embodiment of the present application further provides a control method of an air conditioner, the method is applied to an air conditioner including a compressor and an indoor heat exchanger, the air conditioner further includes a sensor and a controller, as shown in fig. 6, the method includes:

s201, acquiring the operation data detected by the sensor.

In some embodiments of the present invention, the operational data specifically includes one or more of an evaporator leaving water temperature, a condenser entering water temperature, a condenser leaving water temperature, a heat exchanger entering water temperature, a condenser loop temperature, a heat exchanger leaving water temperature, an evaporator loop temperature, a building entering water temperature, a building leaving water temperature, a steam heating amount, an evaporator flow rate, a condenser temperature difference characterization value, and a refrigerant temperature.

S202, inputting the operation data into an SVM model to obtain a predicted energy efficiency ratio.

In some embodiments of the present invention, the SVM model is built by:

establishing an initial SVM model and inputting the training set for training;

verifying the trained initial SVM model by using the verification set;

In some embodiments of the invention, the initial SVM model is built based on a radial basis kernel function;

In some embodiments of the present invention, the building of the SVM model further comprises: and adjusting a penalty factor and a gamma value based on the prediction precision of the training result and the verification result of the initial SVM model.

And S203, adjusting control parameters of the air conditioner according to the predicted energy efficiency ratio.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit the same; although the present application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not necessarily depart from the spirit and scope of the corresponding technical solutions in the embodiments of the present application.

Claims

1. An air conditioner comprising:

an indoor heat exchanger operating as a condenser or an evaporator;

characterized in that, the air conditioner still includes:

a sensor for detecting operation data of the air conditioner;

a controller configured to:

acquiring operating data detected by the sensor;

2. The air conditioner of claim 1, wherein the operational data includes one or more of evaporator leaving water temperature, condenser entering water temperature, condenser leaving water temperature, heat exchanger entering water temperature, condenser loop temperature, heat exchanger leaving water temperature, evaporator loop temperature, building entering water temperature, building leaving water temperature, steam heating capacity, evaporator flow, condenser temperature differential characterization value, and refrigerant temperature.

3. The air conditioner as claimed in claim 1, wherein the SVM model is established by:

establishing an initial SVM model and inputting the training set for training;

verifying the trained initial SVM model by using the verification set;

4. The air conditioner according to claim 3, wherein:

the initial SVM model is established based on a radial basis kernel function;

5. The air conditioner according to claim 3, wherein the establishment of the SVM model further comprises:

and adjusting a penalty factor and a gamma value based on the prediction precision of the training result and the verification result of the initial SVM model.

6. A control method of an air conditioner, the method being applied to an air conditioner including a compressor and an indoor heat exchanger, wherein the air conditioner further includes a sensor and a controller, the method comprising:

acquiring operating data detected by the sensor;

7. The control method of claim 6, wherein the operational data includes one or more of evaporator leaving water temperature, condenser entering water temperature, condenser leaving water temperature, heat exchanger entering water temperature, condenser loop temperature, heat exchanger leaving water temperature, evaporator loop temperature, building entering water temperature, building leaving water temperature, steam heating, evaporator flow, condenser temperature differential characterization value, and refrigerant temperature.

8. The control method of claim 6, wherein the SVM model is created by:

establishing an initial SVM model and inputting the training set for training;

verifying the trained initial SVM model by using the verification set;

9. The control method according to claim 8, characterized in that:

the initial SVM model is established based on a radial basis kernel function;

10. The control method of claim 8, wherein the establishing of the SVM model further comprises: and adjusting a penalty factor and a gamma value based on the prediction precision of the training result and the verification result of the initial SVM model.