Air conditioning system

Abstract

The present application discloses an air conditioning system, comprising: an outdoor unit; an indoor unit, wherein an air pipe and a liquid pipe are connected between the indoor unit and the outdoor unit; and a controller, which is configured as follows: a training step: obtaining a detection data set of several associated operating parameters as a training set; training a mathematical model according to the training set, taking any of the operating parameters as the output of the mathematical model, and taking other operating parameters other than the output as the input of the mathematical model for training; an online fault diagnosis step: judging the fault state of the detection element used to detect the output. The air conditioning system of the present application, which trains a mathematical model through a detection data set of several associated operating parameters, and uses a statistical hypothesis test method to judge the fault state, can eliminate the influence caused by factors such as actual application occasions, working conditions, pipe length and online schemes. After adding field data, the mathematical model will be more accurate, and can not only detect the fault of the detection element online, but also accurately judge the faulty detection element, which is convenient for personnel maintenance.

Description

Technical Field

[0001] The present application relates to the technical field of air conditioning, and in particular to an air conditioning system. Background Art

[0002] In order to achieve energy saving and improve indoor air quality, various optimization control strategies for air conditioning systems have become increasingly complex, but the implementation of these optimization strategies must have a prerequisite: the accuracy and reliability of the sensor. The temperature sensor of the air conditioner may drift in resistance during use, which will make the air conditioner control based on the data collected by the sensor invalid. Measurement failure will inevitably mislead the control system, resulting in the failure to achieve the goals of advanced control strategies. In addition, when the temperature deviation measured by the temperature sensor is too large, it will cause the air conditioner to malfunction.

[0003] The Chinese patent application with application number CN200410017182.X discloses an online fault diagnosis system for temperature and flow sensors in a central air-conditioning water system. The online fault diagnosis system for temperature and flow sensors in a central air-conditioning water system mainly includes data acquisition, noise filtering, steady-state data judgment, fault detection, storage, fault diagnosis, confidence judgment of diagnosis results, data recovery and output modules. Using statistical mathematical methods, based on the energy conservation and mass conservation of the system, online fault diagnosis is performed on the drift faults of the temperature and flow sensors in the air-conditioning water system. This patent uses different algorithms based on energy balance to determine that some of the measured values ​​in a group of sensors have a large deviation, but it cannot confirm which sensor has a problem. Summary of the invention

[0004] In order to solve the technical problem in the prior art that a detection element in an air-conditioning system cannot be detected and located when it fails, the present application provides an air-conditioning system that can solve the above problem.

[0005] An air conditioning system, comprising:

[0006] Outdoor unit;

[0007] An indoor unit, wherein a gas pipe and a liquid pipe are connected between the indoor unit and the outdoor unit;

[0008] The controller is configured as:

[0009] Training steps:

[0010] Obtaining detection data sets of several associated operating parameters as training sets;

[0011] Training a mathematical model according to the training set, taking any of the operating parameters as an output of the mathematical model, and taking other operating parameters other than the output as inputs of the mathematical model for training;

[0012] Online troubleshooting steps:

[0013] Get the test values ​​of all operating parameters online;

[0014] Inputting the detected value of the input quantity obtained online into a mathematical model, the mathematical model calculates and outputs the predicted value of the output quantity;

[0015] The statistical hypothesis testing method is used to judge the fault status, including:

[0016] Construct test statistics based on the predicted value of the output and the detected value of the output;

[0017] Determine critical values ​​for test statistics;

[0018] The fault state of the detection element used to detect the output quantity is determined according to the critical value and the test statistic.

[0019] In some embodiments, the step of obtaining the air conditioning operation mode corresponding to the operation parameter is further included.

[0020] In the training step, the air conditioner operation mode is input into the mathematical model for training;

[0021] In the online fault diagnosis step, when the detection values ​​of all operating parameters are obtained online, the air-conditioning operation mode corresponding to the detection value is obtained at the same time, and is input into the mathematical model together with the input quantity to obtain the predicted value of the output quantity. The air-conditioning operation mode includes a heating mode and a cooling mode, and the cooling mode includes a single cooling mode and / or a dehumidification mode.

[0022] In some embodiments, in the training step, each operating parameter is used as an output to train a mathematical model to obtain multiple mathematical models;

[0023] In the online fault diagnosis step, the input quantity is input into each mathematical model respectively, the fault state of the detection element corresponding to the corresponding output quantity is determined, and the fault state of all the detection elements used to detect each operating parameter is obtained.

[0024] In some embodiments, in the online fault diagnosis step, the method for detecting the fault status of all detection elements includes:

[0025] Select a mathematical model and determine the current mathematical model in a set order or by random selection;

[0026] Determine the input and output of the selected mathematical model, and perform a fault state judgment step using a statistical hypothesis test method to judge the fault state of the detection element used to detect the current output, and when the fault state is normal, continue to select other mathematical models until the fault states of the detection elements of the outputs of all mathematical models are judged;

[0027] When the fault state is abnormal, in the step of selecting another mathematical model, the detection value of the detection element determined to be abnormal is no longer input into the other mathematical model as the detection value of the input quantity.

[0028] In some embodiments, when the fault state is abnormal, in the step of selecting other mathematical models, the predicted value of the detection element judged to be abnormal is input into the other mathematical model as the detection value of the input quantity.

[0029] In some embodiments, in the training step, the trained mathematical model is a neural network model, and the training set is detection data of normal operation of the air conditioner within a set time after installation.

[0030] In some embodiments, before the online fault diagnosis step, the steps of setting a detection cycle and determining the number of detections are also included. In the online fault diagnosis step, when the detection values ​​of all operating parameters are obtained online, the steps include:

[0031] The detection values ​​of all operating parameters within the set time after startup and shutdown are obtained online according to the detection cycle. When the detected number does not meet the detection number or the operating time does not reach the set time, no fault status judgment is performed.

[0032] In some embodiments, in the step of using a statistical hypothesis test method to determine the fault state, before constructing the test statistic, the step further includes:

[0033] Determine the null hypothesis, H0: μ1 = μ2, assuming that the mean of the predicted output value is equal to the mean of the detected output value;

[0034] In the step of constructing the test statistic:

[0035]

[0036] Where t is the test statistic, X1 is the mean of the test value of the input quantity, X2 is the mean of the predicted value of the input quantity, n1 is the sample size of the test value of the input quantity, and n2 is the sample size of the predicted value of the input quantity;

[0037]

[0038] is the sample variance of the input value, is the sample variance of the predicted value of the input quantity.

[0039] In some embodiments, in the step of determining the fault state of the detection element used to detect the output quantity according to the critical value and the test statistic, the test statistic t is compared with the critical value of the test statistic. If the absolute value of the test statistic t is greater than the critical value, the null hypothesis H0 is rejected, that is, the fault state of the corresponding detection element is an abnormal state. Otherwise, the null hypothesis H0 is accepted, that is, the fault state of the corresponding detection element is a normal state.

[0040] Alternatively, the P value corresponding to the observed value of the test statistic t is obtained and compared with the set significance level α, and the fault state of the detection element is determined based on the comparison result.

[0041] In some embodiments, in the training step, several associated operating parameters are obtained, namely: outlet air temperature To, return air temperature Ti, liquid pipe temperature Tl, and gas pipe temperature Tg.

[0042] The present application provides an air-conditioning system, which trains a mathematical model through a detection data set of several related operating parameters, and uses a statistical hypothesis testing method to judge the fault state. It can eliminate the influence of factors such as actual application occasions, working conditions, pipe lengths and online solutions. After adding field data, the mathematical model will be more accurate. It can not only detect the fault of the detection element online, but also accurately determine the faulty detection element, which is convenient for personnel to repair. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 is a system principle diagram of an embodiment of the air conditioning system proposed by the present invention;

[0044] Figure 2 is a schematic diagram of a refrigerant cycle in a refrigeration mode of the air-conditioning system proposed by the present invention;

[0045] Figure 3 is a schematic diagram of a heating agent circulation in a refrigeration mode of the air-conditioning system proposed by the present invention;

[0046] Figure 4 It is a schematic diagram of a mathematical model of an embodiment of the air conditioning system proposed by the present invention;

[0047] Figure 5 is a schematic diagram of a mathematical model of another embodiment of the air conditioning system proposed by the present invention;

[0048] Figure 6 is a control flow chart of an embodiment of the air conditioning system proposed by the present invention;

[0049] Figure 7It is a flow chart of fault state judgment of a detection element of an embodiment of an air conditioning system proposed by the present invention;

[0050] Figure 8 is a schematic diagram of a mathematical model of another embodiment of the air conditioning system proposed by the present invention;

[0051] Fig. 9 is a flow chart of a method for detecting the fault status of all detection elements of another embodiment of the air conditioning system proposed by the present invention;

[0052] Fig.10 It is a schematic diagram of the internal structure of the indoor unit of the air-conditioning system proposed by the present invention. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0054] In the description of the present invention, it should be understood that the terms "center", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0055] The terms "first", "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, unless otherwise specified, "plurality" means two or more.

[0056] In the description of this application, it should be noted that, unless otherwise clearly specified and limited, the terms "connected" and "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances. In addition, when describing pipelines, the terms "connected" and "connection" used in this application have the meaning of conduction. The specific meaning needs to be understood in conjunction with the context.

[0057] In the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0058] To facilitate understanding, some basic concepts of terms or technologies involved in the embodiments of the present invention are first briefly introduced and explained.

[0059] Refrigeration mode: The compressor of the air conditioning system draws the low-temperature and low-pressure gaseous refrigerant after evaporation in the evaporator into the compressor cavity, compresses it into a high-temperature and high-pressure gaseous refrigerant, and enters the condenser. The high-temperature and high-pressure gaseous refrigerant is condensed into a high-temperature and high-pressure liquid refrigerant in the condenser, and then passes through a throttling element such as a capillary throttling, and becomes a low-temperature and low-pressure liquid refrigerant. After evaporation in the evaporator, it finally returns to the compressor, thus completing the entire refrigeration cycle. Among them, the outdoor heat exchanger in the refrigeration mode is used as a condenser, and the indoor heat exchanger is used as an evaporator.

[0060] Refrigerant: A substance that easily absorbs heat and turns into gas, and easily releases heat and turns into liquid. In the air conditioning system, the refrigerant transfers heat energy through evaporation and condensation, producing a cooling effect.

[0061] Expansion valve: It consists of two parts: valve body and coil, used for throttling, reducing pressure and regulating flow. The expansion valve in the air conditioning system can make the medium-temperature and high-pressure liquid refrigerant become low-temperature and low-pressure wet steam through its throttling, and then the refrigerant absorbs heat in the evaporator to achieve the cooling effect, and the valve flow is controlled by the change of superheat at the evaporator outlet.

[0062] Temperature sensor: It is the core part of the temperature measuring instrument. It can sense the temperature and convert it into a usable output signal. It sends the detection signal to the controller, which can control the actions of other components according to the detection signal. There are usually 4 indoor temperature sensors in a multi-split unit, namely the outlet air temperature To, the return air temperature Ti, the liquid pipe temperature Tl and the gas pipe temperature Tg.

[0063] Pressure sensor: A device that can convert the detected pressure signal into an electronic signal. The pressure sensor is generally installed on the air conditioner outdoor unit pipeline to perform pressure detection and send the detection signal to the controller, which controls the operation of other components.

[0064] The various optimization control strategies for air conditioning systems have become increasingly complex, but the implementation of these optimization strategies must have a premise: the accuracy and reliability of the sensor. For example, the temperature sensor of the air conditioner may drift in resistance during use, which will make the air conditioner control based on the data collected by the sensor invalid. Measurement failure will inevitably mislead the control system, resulting in the failure to achieve the goals of advanced control strategies. In addition, when the temperature deviation measured by the temperature sensor is too large, it will cause the air conditioner to malfunction.

[0065] At present, after various detection component failures occur in the air-conditioning system, the component failure can be diagnosed, but it is impossible to locate which detection component in the air-conditioning system has failed. In particular, when multiple temperature sensors are set, only the temperature detection failure can be known, but it is impossible to locate the position of the temperature sensor failure. Manual observation and analysis by staff is required, which not only causes a large workload for the staff, but also fails to quickly locate the failed detection component, resulting in increased maintenance costs and affecting the user experience.

[0066] Based on this, the embodiment of the present application provides a mathematical model for training through a detection data set of several associated operating parameters, and adopts a statistical hypothesis testing method to judge the fault status, which can eliminate the influence caused by factors such as actual application occasions, working conditions, pipe length and online scheme. After adding field data, the mathematical model will be more accurate, not only can the fault of the detection element be detected online, but also the faulty detection element can be accurately determined, which is convenient for personnel to repair.

[0067] To further describe the solution of this application, please refer to Figure 1 , Figure 1 It is a structural schematic diagram of an air conditioning system provided in accordance with an exemplary embodiment of the present application.

[0068] In some embodiments, Figure 1 As shown, the air conditioning system includes an outdoor unit 11, an indoor unit 12, an expansion valve 13, a four-way valve 14, a compressor 15, an indoor heat exchanger 16, an outdoor heat exchanger 17 and a controller (not shown in the figure). During the refrigerant circulation process, one of the indoor heat exchanger 16 and the outdoor heat exchanger 17 serves as an evaporator and the other serves as a condenser.

[0069] The refrigeration cycle of the air conditioner is performed by using the compressor 15, the condenser, the expansion valve 13 and the 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.

[0070] The compressor 15 compresses the 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.

[0071] The expansion valve 13 expands the high-temperature and high-pressure liquid refrigerant condensed in the condenser into a low-pressure liquid refrigerant. The evaporator evaporates the refrigerant expanded in the expansion valve 13 and returns the refrigerant gas in a low-temperature and low-pressure state to the compressor 15. The evaporator can achieve a refrigeration effect by utilizing the latent heat of evaporation of the refrigerant to exchange heat with the material to be cooled. In the whole cycle, the air conditioner can adjust the temperature of the indoor space.

[0072] like Figure 2 , Figure 3 As shown, by controlling the communication direction of the four-way valve 14, the refrigeration cycle or heating cycle of the indoor unit 12 is realized.

[0073] The outdoor unit of the air conditioner refers to a part of a refrigeration cycle including a compressor 15 and an outdoor heat exchanger 17 , the indoor unit 12 of the air conditioner includes an indoor heat exchanger 16 , and the expansion valve 13 may be provided in the indoor unit 12 or the outdoor unit 11 .

[0074] The indoor heat exchanger 16 and the outdoor heat exchanger 17 function as a condenser or an evaporator. When the indoor heat exchanger 16 functions as a condenser, the air conditioner functions as a heater in a heating mode, and when the indoor heat exchanger 16 functions as an evaporator, the air conditioner functions as a cooler in a cooling mode.

[0075] The indoor unit 12 may be provided with a gas pipe 19 and a liquid pipe 18 for guiding the flow of refrigerant. The gas pipe 19 and the liquid pipe 18 are connected to the indoor heat exchanger 16 so as to extend to the outside of the housing through the internal space of the housing for connection with the outdoor unit 11.

[0076] The gas pipe 19 is used to guide the flow of gas refrigerant, and the liquid pipe 18 is used to guide the flow of liquid refrigerant. The gas pipe 19 and the liquid pipe 18 may extend parallel to each other.

[0077] The indoor unit 12 may be a wall-mounted indoor unit 12 installed on a wall of an indoor space, or may be a cabinet-type indoor unit 12 .

[0078] The indoor unit 12 includes a housing in which a plurality of components constituting a refrigeration cycle are installed. The housing includes a front surface that is at least partially open, a rear surface that is mounted on a wall of an indoor space and provided with a mounting plate, a bottom surface that defines a bottom structure, side surfaces that are provided on both sides of the bottom surface, and a top surface that defines a top appearance.

[0079] The indoor unit 12 is provided with a return air port and an air outlet. The air flows into the indoor unit 12 from the return air port, exchanges heat in the indoor heat exchanger 16, and is discharged to the indoor space through the air outlet. Figure 1 As shown, a return air temperature sensor 20 is provided at the return air inlet, and the return air temperature sensor 20 is used to detect the return air temperature. An outlet air temperature sensor 21 is provided at the outlet air outlet, and is used to detect the outlet air temperature.

[0080] like Fig.10 As shown, a gas pipe temperature sensor 22 is provided in the gas pipe 19 for detecting the temperature of the refrigerant in the gas pipe 19 , and a liquid pipe temperature sensor 23 is provided in the liquid pipe 18 for detecting the temperature of the refrigerant in the liquid pipe 18 .

[0081] The controller is configured as Figure 6 As shown, including:

[0082] Training steps:

[0083] Obtaining detection data sets of several associated operating parameters as training sets;

[0084] Training a mathematical model according to a training set, taking any operating parameter as an output of the mathematical model, and taking other operating parameters other than the output as inputs of the mathematical model for training;

[0085] Online troubleshooting steps:

[0086] Get the test values ​​of all operating parameters online;

[0087] The detected value of the input quantity obtained online is input into the mathematical model, and the mathematical model calculates the predicted value of the output quantity and outputs it;

[0088] The statistical hypothesis testing method is used to judge the fault status, including:

[0089] Construct test statistics based on the predicted value of the output and the detected value of the output;

[0090] Determine critical values ​​for test statistics;

[0091] The fault state of the detection element used to detect the output quantity is determined based on the critical value and the test statistic.

[0092] The air-conditioning system trains a mathematical model through detection data sets of several related operating parameters, and uses statistical hypothesis testing methods to judge the fault status. It can eliminate the influence of factors such as actual application occasions, working conditions, pipe length and online solutions. After adding field data, the mathematical model will be more accurate. It can not only detect the fault of the detection component online, but also accurately determine the faulty detection component, which is convenient for personnel to repair.

[0093] During the operation of the air conditioner, some operating parameters change according to certain rules, and usually the changes in cooling and heating are different. Based on this, in some embodiments, the step of obtaining the air conditioner operation mode corresponding to the operating parameter is also included, which is used to associate the operating parameter with the air conditioner operation mode, so that the trained mathematical model has higher accuracy.

[0094] In some embodiments, in the training step, the air conditioner operation mode is input into the mathematical model for training;

[0095] Accordingly, since the trained mathematical model incorporates the air conditioning operation mode, in the online fault diagnosis step, such as Figure 8 As shown, when the detection values ​​of all operating parameters are obtained online, the air-conditioning operation mode corresponding to the detection values ​​is obtained at the same time, and is input into the trained mathematical model together with the input quantity to obtain the predicted value of the output quantity.

[0096] The air conditioner operation mode includes a heating mode and a cooling mode, and the cooling mode includes a cooling mode and / or a dehumidification mode. That is, by training a mathematical model for each air conditioner operation mode, so that the mathematical model is more consistent with the actual operation mode, the trained output prediction value is closer to the actual value.

[0097] In some embodiments, in the training step, each operating parameter is used as an output to train the mathematical model to obtain multiple mathematical models.

[0098] In some embodiments, Figure 4 As shown, the training set has parameter 1, parameter 2, ... parameter n, a total of n operating parameters. Parameter 1 is used as the output of the mathematical model and the other parameters are used as the input of the mathematical model to train a mathematical model. Parameter 2 is used as the output of the mathematical model and the other parameters are used as the input of the mathematical model to train a mathematical model. ..., parameter n is used as the output of the mathematical model and the other parameters are used as the input of the mathematical model to train a mathematical model. A total of n mathematical models are trained.

[0099] Although we only know that the operating parameters have certain correlations, the correlation coefficients between each operating parameter and other operating parameters are different. Therefore, Figure 4 , Figure 5 As shown, each operating parameter needs to be trained as an output to obtain a mathematical model that reflects the relationship between each operating parameter and other operating parameters.

[0100] In the online fault diagnosis step, the input quantity is input into each mathematical model respectively, the fault state of the detection element corresponding to the corresponding output quantity is determined, and the fault state of all the detection elements used to detect each operating parameter is obtained.

[0101] In some embodiments, the trained mathematical model uses the regression function of a neural network machine learning algorithm, and the value of any one operating parameter can be accurately inferred from the values ​​of the other three operating parameters.

[0102] In some embodiments, in the online fault diagnosis step, the method for detecting the fault status of all detection elements includes:

[0103] The mathematical model is selected, and the current mathematical model is determined in a set order or randomly selected. As mentioned above, when there are n associated operating parameters, n mathematical models are trained accordingly. Therefore, in the online fault diagnosis step, the fault state of each detection element needs to be judged in a certain order or manner.

[0104] In some embodiments, Fig. 9 As shown, the input and output of the selected mathematical model are determined, and a fault state judgment step is performed using a statistical hypothesis test method to judge the fault state of the detection element used to detect the current output. When the fault state is normal, other mathematical models are continuously selected until the fault states of the detection elements of the outputs of all mathematical models are judged.

[0105] As can be seen from the foregoing, after each trained mathematical model is trained, its input and output are uniquely determined. Therefore, for a selected mathematical model, the input and output of the mathematical model can be determined. The fault diagnosis method in this scheme can use the mathematical model to perform fault diagnosis on the detection element used to detect the output of the mathematical model.

[0106] When the fault state is abnormal, in the step of selecting another mathematical model, the detection value of the detection element determined to be abnormal is no longer input into the other mathematical model as the detection value of the input quantity.

[0107] If the fault state of a detection element is judged to be abnormal, it means that there is a serious deviation between the detection result and the actual value. Therefore, it is no longer suitable to be input into other mathematical models as an input quantity for fault judgment of gas operating parameters, otherwise it will affect the detection results of other operating parameters.

[0108] In order to avoid misjudgment when diagnosing other components that have not yet been diagnosed as faulty, in some embodiments, when the fault state is abnormal, in the step of selecting other mathematical models, the predicted value of the detection element judged to be abnormal is input into the other mathematical model as the detection value of the input quantity.

[0109] That is, since the mathematical model is trained based on actual data on site, the predicted value outputted by it should be very close to the actual value. It is understandable that if the detection element is normal, the predicted value, the detection value and the actual value should all be very close. If the detection element is abnormal, the detection value deviates from the predicted value and the actual value, but the predicted value and the actual value are still very close. Therefore, in this solution, when the fault state is abnormal, in the step of selecting other mathematical models, the predicted value of the detection element judged to be abnormal is input as the detection value of the input quantity into the other mathematical model to improve the diagnostic accuracy of other operating parameters.

[0110] In some embodiments, in the training step, the trained mathematical model is a neural network model, and the training set is detection data of normal operation of the air conditioner within a set time after installation.

[0111] For example, the data of the normal operation of the air conditioner in the first year of installation can be recorded and added to the training set database to revise the neural network model. Because the actual application occasions, working conditions, pipe lengths, and online solutions will be different from those in the laboratory, the model regression value will be more accurate after adding field data.

[0112] In some embodiments, before the online fault diagnosis step, the steps of setting a detection cycle and determining the number of detections are also included. In the online fault diagnosis step, when the detection values ​​of all operating parameters are obtained online, the steps include:

[0113] The detection values ​​of all operating parameters within the set time after startup and shutdown are obtained online according to the detection cycle. When the detected quantity does not meet the detection quantity or the running time does not reach the set time, no fault status judgment is performed.

[0114] In some embodiments, in the step of using a statistical hypothesis test method to determine the fault state, before constructing the test statistic, the following step is further included:

[0115] Determine the null hypothesis, H0: μ1 = μ2, assuming that the mean of the predicted output value is equal to the mean of the detected output value;

[0116] The task of hypothesis testing is to confirm whether the null hypothesis is true. The approach is: first assume that the null hypothesis is true, and then use samples to judge the truth or falsity. Since the information in the samples is relatively scattered, it is necessary to construct a statistic to make a judgment. This statistic is called a test statistic.

[0117] In the step of constructing the test statistic:

[0118]

[0119] Where t is the test statistic, X1 is the mean of the test value of the input quantity, X2 is the mean of the predicted value of the input quantity, n1 is the sample size of the test value of the input quantity, and n2 is the sample size of the predicted value of the input quantity;

[0120]

[0121] is the sample variance of the input value, is the sample variance of the predicted value of the input quantity.

[0122] Before determining the critical value of the test statistic, it also includes determining the significance level α, including:

[0123] When trying to judge whether the null hypothesis H0 is true, there is a possibility of making mistakes, which means taking risks. In order to control this risk, we first need to use a probability to represent this risk. This probability is the probability of the event "H0 is true but rejected", and this probability is the significance level that needs to be determined. Due to the randomness of the sample, it is impossible to completely avoid making the "error of rejecting the true", so the probability of this event can only be controlled within a very small range, usually α is selected as 1%, 5% and 10%.

[0124] In some embodiments, Figure 7 As shown, in the step of judging the fault state of the detection element used to detect the output quantity according to the critical value and the test statistic, the test statistic t is compared with the critical value of the test statistic. If the absolute value of the test statistic t is greater than the critical value, the null hypothesis H0 is rejected, that is, the fault state of the corresponding detection element is an abnormal state. Otherwise, the null hypothesis H0 is accepted, that is, the fault state of the corresponding detection element is a normal state.

[0125] Alternatively, the P value corresponding to the observed value of the test statistic t is obtained and compared with the set significance level α, and the fault state of the detection element is determined based on the comparison result.

[0126] That is, compare the statistical value calculated using the sample with the critical value at a given significance level. If the absolute value of the statistical value calculated is greater than the critical value, the null hypothesis H0 is rejected; otherwise, the null hypothesis can only be accepted.

[0127] For example: P value = 0.21484> α = 0.05; therefore, the null hypothesis H0 is not rejected, and the difference is not statistically significant. There is no sufficient reason to explain that the means of the two populations are not equal, that is, the sensor is not faulty, otherwise it is judged that the indoor sensor is faulty.

[0128] In some embodiments, in the training step, several associated operating parameters are obtained, namely, outlet air temperature To, return air temperature Ti, liquid pipe temperature Tl, and gas pipe temperature Tg. During the operation of the air conditioner, the four temperatures change according to a certain rule, so the four temperature values ​​can be used as associated operating parameters for training.

[0129] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0130] For the convenience of explanation, the above description has been made in conjunction with specific embodiments. However, the above exemplary discussion is not intended to be exhaustive or limit the embodiments to the specific forms disclosed above. Based on the above teachings, various modifications and variations can be obtained. The selection and description of the above embodiments are to better explain the principles and practical applications, so that those skilled in the art can better use the embodiments and various different variations of the embodiments suitable for specific use considerations.

Claims

1. An air conditioning system, characterized in that: include: Outdoor unit; An indoor unit, wherein a gas pipe and a liquid pipe are connected between the indoor unit and the outdoor unit; The controller is configured as: Training steps: Obtaining detection data sets of several associated operating parameters as training sets; Training a mathematical model according to the training set, taking any of the operating parameters as an output of the mathematical model, and taking other operating parameters other than the output as inputs of the mathematical model for training; Online troubleshooting steps: Obtain the test values ​​of all operating parameters online; Inputting the detected value of the input quantity obtained online into a mathematical model, the mathematical model calculates and outputs the predicted value of the output quantity; The statistical hypothesis testing method is used to judge the fault status, including: Construct test statistics based on the predicted value of the output and the detected value of the output; Determine critical values ​​for test statistics; Determine the fault state of the detection element used to detect the output quantity according to the critical value and the test statistic; In the training step, each operating parameter is used as an output to train a mathematical model to obtain multiple mathematical models; In the online fault diagnosis step, the input quantity is input into each mathematical model respectively, the fault state of the detection element corresponding to the corresponding output quantity is determined, and the fault state of all the detection elements used to detect each operating parameter is obtained.

2. The air conditioning system according to claim 1, characterized in that: The step also includes obtaining the air conditioning operation mode corresponding to the operation parameter. In the training step, the air conditioner operation mode is input into the mathematical model for training; In the online fault diagnosis step, when the detection values ​​of all operating parameters are obtained online, the air-conditioning operation mode corresponding to the detection value is obtained at the same time, and is input into the mathematical model together with the input quantity to obtain the predicted value of the output quantity. The air-conditioning operation mode includes a heating mode and a cooling mode, and the cooling mode includes a single cooling mode and a dehumidification mode.

3. The air conditioning system according to claim 1, characterized in that: In the online fault diagnosis step, the method for detecting the fault status of all detection elements includes: Select a mathematical model and determine the current mathematical model in a set order or by random selection; Determine the input and output of the selected mathematical model, and perform a fault state judgment step using a statistical hypothesis test method to judge the fault state of the detection element used to detect the current output, and when the fault state is normal, continue to select other mathematical models until the fault states of the detection elements of the outputs of all mathematical models are judged; When the fault state is abnormal, in the step of selecting another mathematical model, the detection value of the detection element determined to be abnormal is no longer input into the other mathematical model as the detection value of the input quantity.

4. The air conditioning system according to claim 3, characterized in that: When the fault state is abnormal, in the step of selecting another mathematical model, the predicted value of the detection element judged to be abnormal is input into the other mathematical model as the detection value of the input quantity.

5. The air conditioning system according to any one of claims 1 to 4, characterized in that: In the training step, the trained mathematical model is a neural network model, and the training set is the detection data of the air conditioner during normal operation within a set time after installation.

6. The air conditioning system according to any one of claims 1 to 4, characterized in that: Before the online fault diagnosis step, the steps of setting the detection cycle and determining the number of detections are also included. In the online fault diagnosis step, when the detection values ​​of all operating parameters are obtained online, the following steps are included: The detection values ​​of all operating parameters within the set time after startup and shutdown are obtained online according to the detection cycle. When the detected number does not meet the detection number or the operating time does not reach the set time, no fault status judgment is performed.

7. The air conditioning system according to any one of claims 1 to 4, characterized in that: In the training step, several associated operating parameters are obtained, namely, outlet air temperature To, return air temperature Ti, liquid pipe temperature Tl, and gas pipe temperature Tg.

Images