An early warning method and system for refrigerant leakage of an air conditioner, and the air conditioner

Abstract

The application discloses an early warning method and system for air conditioner refrigerant leakage and an air conditioner. The early warning method comprises the following steps: collecting current operation parameters of the air conditioner according to a preset sampling period when the air conditioner enters a stable operation state; pre-processing the collected current operation parameters, and calculating current characteristic parameters according to the pre-processed operation parameters; comparing each current characteristic parameter with a corresponding calculation reference band, and determining that the current characteristic parameter is abnormal when at least one current characteristic parameter deviates from the corresponding calculation reference band continuously in a continuous preset number of sampling periods; calculating a current refrigerant risk score based on each abnormal current characteristic parameter, and comparing the current refrigerant risk score with a preset score; and determining that the air conditioner refrigerant has an early leakage fault when the current refrigerant risk score exceeds the preset score. The early warning method does not need to add detection hardware, does not increase the BOM cost of the whole machine, and is compatible with the existing air conditioner hardware architecture.

Description

Technical Field

[0001] This invention belongs to the field of air conditioning fault early warning technology, and particularly relates to an early warning method, early warning system and air conditioner for refrigerant leakage in air conditioners. Background Technology

[0002] Refrigerant leakage is a common fault affecting the cooling / heating performance, energy efficiency, and operational safety of air conditioners. It not only reduces the cooling / heating capacity and increases energy consumption, but also hinders timely and accurate identification and early warning of refrigerant leaks. Therefore, timely and accurate identification of refrigerant leaks and early warning systems have always been an important research direction in this field. With the development of intelligent air conditioning technology, various data-driven refrigerant leak detection schemes have emerged, enabling leak identification without relying on additional dedicated refrigerant detection sensors. However, several limitations still exist. First, the existing technical solutions only disclose the concepts of dynamic benchmark modeling and environmental adaptive compensation, but do not disclose the specific compensation calculation methods for the current indoor and outdoor temperature difference. Second, most existing technologies use a single parameter value to compare with a threshold for judgment, without setting a secondary filtering mechanism for abnormal sampling multiple times. This makes it impossible to effectively filter out false alarms caused by occasional fluctuations in operating conditions, and the false alarm rate remains at a high level. Third, the existing early warning capabilities are insufficient, unable to identify risks at the stage of minute leaks, and cannot meet the safety requirements of air conditioners containing flammable refrigerants; therefore, further improvements are still needed to the existing air conditioner refrigerant leak detection solutions. Summary of the Invention

[0003] This invention addresses the following problems with existing air conditioning refrigerant leak detection technologies: detection lag, inability to provide early warning of minor leaks; high false alarm rate, susceptibility to fluctuations in operating conditions; reliance on additional dedicated sensors, leading to high hardware costs; and lack of leak trend prediction capabilities. Therefore, this invention provides an air conditioning refrigerant leak early warning solution that requires no additional hardware, has a low false alarm rate, and enables early warning.

[0004] This invention provides an early warning method for refrigerant leakage in an air conditioner, the air conditioner including a compressor and an indoor fan; after the air conditioner is started, the early warning method includes: Determine whether the air conditioner has entered a stable operating state; Once the air conditioner enters a stable operating state, its current operating parameters are collected according to a preset sampling period. The current operating parameters include at least the current return air temperature, current exhaust air temperature, and current operating frequency related to the compressor, as well as the current indoor temperature, current outdoor temperature, and current speed of the indoor fan. The collected current operating parameters are preprocessed, and the current characteristic parameters are calculated based on the preprocessed operating parameters; wherein, the current characteristic parameters include at least superheat, unit frequency energy efficiency ratio change rate, and return gas temperature rise slope; Each of the current feature parameters is compared with its corresponding calculation reference band. If at least one current feature parameter deviates from its corresponding calculation reference band within a consecutive preset number of sampling periods, the current feature parameter is determined to be abnormal. The calculation reference band is a calculation reference interval and is the fluctuation range of normal operating parameters determined based on the historical normal operating parameters of the air conditioner and the current indoor temperature and current outdoor temperature. Based on the current characteristic parameters of each anomaly, it is determined whether there is an early leakage fault of the air conditioning refrigerant.

[0005] Further optionally, the method for determining the calculation reference band corresponding to each of the current feature parameters includes: Based on the historical normal operating parameters of the air conditioner, a dynamic reference band is constructed that adapts to different operating conditions and corresponds to each of the current characteristic parameters. Calculate the current indoor and outdoor temperature difference based on the current indoor and outdoor temperatures; Based on the current indoor-outdoor temperature difference and the reference temperature difference, calculate the compensation amount corresponding to each of the current characteristic parameters; Calculate the calculation reference band corresponding to each of the current feature parameters based on the dynamic reference band and compensation amount.

[0006] Further optionally, the step of calculating the compensation amount corresponding to each of the current characteristic parameters based on the current indoor-outdoor temperature difference and the reference temperature difference includes: Subtract the reference temperature difference from the current indoor-outdoor temperature difference to obtain the compensated temperature difference; The compensation amount corresponding to each of the current characteristic parameters is equal to the compensation constant corresponding to each of the current characteristic parameters multiplied by the compensation temperature difference.

[0007] Further optionally, determining whether the air conditioning refrigerant has an early leakage fault based on the current characteristic parameters of each anomaly includes: Calculate the current refrigerant risk score based on the current characteristic parameters of each anomaly; Compare the current refrigerant risk score with the preset score; Based on the comparison between the current refrigerant risk score and the preset score, it is determined whether there is an early leakage fault in the air conditioning refrigerant.

[0008] Further, optionally, the calculation of the current refrigerant risk score based on the current characteristic parameters of each anomaly includes: Determine the current operating mode of the air conditioner; Based on the current operating mode of the air conditioner, the current indoor temperature, and the current outdoor temperature, determine the weights corresponding to the current characteristic parameters of each anomaly. The current refrigerant risk score is obtained by weighted calculation based on the current characteristic parameters and corresponding weights of each anomaly; wherein the value range of the current refrigerant risk score is 0 to 100.

[0009] Further optionally, when the current operating mode of the air conditioner is cooling mode, the weight allocation of each abnormal current characteristic parameter satisfies the following: the weight of the superheat is in the range of 0.15 to 0.25, the weight of the unit frequency energy efficiency ratio change rate is in the range of 0.05 to 0.15, and the weight of the return air temperature rise slope is in the range of 0.35 to 0.45.

[0010] Further optionally, determining the weights corresponding to the current characteristic parameters of each anomaly based on the current operating mode of the air conditioner, the current indoor temperature, and the current outdoor temperature includes: Input the current operating mode of the air conditioner, the current indoor temperature, and the current outdoor temperature into the pre-trained weighted model; The weighted model is controlled to output the weights corresponding to the current feature parameters of each anomaly.

[0011] Further optionally, determining whether the air conditioning refrigerant has an early leakage fault based on the comparison result of the current refrigerant risk score and the preset score includes: When the current refrigerant risk score exceeds the preset score, it is determined that the air conditioning refrigerant has an early leakage fault; If the current refrigerant risk score does not exceed the preset score, it is determined that there is no early leakage fault of the air conditioning refrigerant.

[0012] Further optionally, after determining that there is an early leakage fault in the air conditioning refrigerant, the early warning method further includes: Based on the current refrigerant risk score, determine the early leakage risk level of the air conditioning refrigerant; The early leakage risk level is transmitted to the user terminal.

[0013] Further optionally, determining the early leakage risk level of the air conditioning refrigerant based on the current refrigerant risk score includes: Determine the preset score range in which the current refrigerant risk score falls; When the current refrigerant risk score is within the first preset score range, the early leakage risk level of the air conditioning refrigerant is determined to be low risk. When the current refrigerant risk score is within the second preset score range, the early leakage risk level of the air conditioning refrigerant is determined to be medium risk. When the current refrigerant risk score is in the third preset score range, the early leakage risk level of the air conditioning refrigerant is determined to be high risk; The preset score intervals include a first preset score interval, a second preset score interval, and a third preset score interval; the first preset score interval is (75, 80], the second preset score interval is (80, 90], and the third preset score interval is (90, 100).

[0014] Further optionally, the method for determining whether the air conditioner has entered a stable operating state includes: Obtain the continuous running time and the fluctuation range of the operating frequency of the compressor; Determine whether the continuous running time of the compressor is greater than a preset time, whether the minimum threshold of the fluctuation range of the operating frequency is greater than a first preset frequency, and whether the maximum threshold of the fluctuation range of the operating frequency is less than a second preset frequency. When the continuous operating time of the compressor exceeds the preset duration, the minimum threshold of the fluctuation range of the operating frequency is greater than the first preset frequency, and the maximum threshold of the fluctuation range of the operating frequency is less than the second preset frequency, the air conditioner is determined to have entered a stable operating state.

[0015] Further optionally, the method for preprocessing the collected current operating parameters includes at least filtering and noise reduction, 3σ principle outlier removal, and Min-Max normalization.

[0016] Alternatively, the current operating parameters may further include the vibration signal of the compressor housing.

[0017] The present invention also provides an early warning system for air conditioning refrigerant leakage, characterized in that it comprises: The data acquisition unit is used to acquire the current operating parameters of the air conditioner according to a preset sampling period; wherein the current operating parameters include at least the current return air temperature, current exhaust air temperature and current operating frequency related to the compressor, as well as the current indoor temperature, current outdoor temperature and current speed of the indoor fan; The controller is communicatively connected to the acquisition unit; it is used to determine whether the air conditioner has entered a stable operating state, receive the operating parameters collected by the acquisition unit and preprocess the current operating parameters, calculate the current characteristic parameters based on the preprocessed operating parameters, compare each current characteristic parameter with the corresponding calculation reference band, and determine that the current characteristic parameter is abnormal when at least one current characteristic parameter continuously deviates from the corresponding calculation reference band within a consecutive preset number of sampling periods; based on each abnormal current characteristic parameter, determine whether the air conditioner refrigerant has an early leakage fault; wherein, the current characteristic parameters include at least superheat, unit frequency energy efficiency ratio change rate and return air temperature rise slope; the calculation reference band corresponding to each current characteristic parameter is determined by the historical normal operating parameters of the air conditioner and the current indoor temperature and current outdoor temperature.

[0018] Optionally, the early warning system further includes a user terminal and a cloud platform; The user terminal and the controller are connected in communication and the user terminal is used to receive the early refrigerant leakage fault information of the air conditioning refrigerant transmitted by the controller. The cloud platform and controller are connected in communication, and the cloud platform is used to receive the operating parameters transmitted by the controller and retrain and optimize the early warning model, and then remotely transmit them to each controller connected in communication with the cloud platform via OTA.

[0019] The present invention also provides an air conditioner, wherein the air conditioner employs any of the above-described methods for early warning of refrigerant leakage, or includes any of the above-described systems for early warning of refrigerant leakage. The air conditioner uses R290 refrigerant.

[0020] Compared with the prior art, the main advantages of the present invention are: (1) Compared with the traditional threshold method that relies on a single parameter or simple threshold judgment, this application integrates multi-dimensional operating parameters and constructs a dynamic benchmark model, which can capture subtle abnormal changes in operating characteristics when the refrigerant leakage is still in the trace stage, and can identify the performance degradation trend in the early stage of trace refrigerant leakage. (2) By introducing the current indoor and outdoor temperatures as reference band determining factors, environmental adaptive compensation is achieved; for example, in high temperature and high humidity environments, the temperature deviation tolerance can be automatically relaxed to avoid misjudgments caused by environmental interference. Compared with the traditional fixed threshold method, the false alarm rate can be significantly reduced; (3) Only the existing sensors and operating data in the air conditioner controller are used, including return air temperature sensor, exhaust air temperature sensor, indoor temperature sensor, outdoor temperature sensor (usually the outdoor unit has ambient temperature detection), fan speed detection and compressor operating frequency, etc.; the above sensors are all standard configurations of air conditioner products, and there is no need to configure additional refrigerant concentration sensors or high-precision pressure sensors. (4) The continuous judgment mechanism of “continuous preset quantity” enables the method to quantitatively assess the degree of leakage - the longer the duration of deviation and the greater the deviation, the higher the leakage risk level. (5) It can issue an early warning in the early stage of refrigerant leakage. For air conditioning products that use flammable refrigerants such as R290, this early detection mechanism can effectively prevent refrigerant from accumulating in a confined space and greatly reduce the risk of combustion or explosion. (6) The calculation baseline is determined by the "historical normal operation parameters", which means that the baseline can be continuously updated and optimized as air conditioning operation data continues to accumulate. The early warning model can be continuously optimized through cloud training, supports OTA remote updates, and realizes the self-evolution capability of "getting smarter with use", laying the technical foundation for building an air conditioning health management system in the future. Attached Figure Description

[0021] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0022] The structures, proportions, sizes, etc. illustrated in this specification are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed herein, and are not intended to limit the conditions under which the present invention can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0023] Figure 1 This is a partial flowchart illustrating an embodiment of the early warning method for refrigerant leakage in air conditioning provided by the present invention. Figure 2 This is a schematic diagram of the entire process of an embodiment of the early warning method for air conditioning refrigerant leakage provided by the present invention; Figure 3 This is a schematic diagram of the multi-source data fusion and anomaly implementation process provided by the present invention; Figure 4A schematic diagram illustrating an embodiment of the dynamic benchmark modeling and environment adaptive compensation provided by this invention; Figure 5 This is a schematic diagram of an embodiment of the early warning system for air conditioning refrigerant leakage provided by the present invention. Detailed Implementation

[0024] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” used in the embodiments of this invention and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. “Multiple” generally includes at least two, but does not exclude the inclusion of at least one.

[0026] It should be understood that the term "and / or" used in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0027] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a product or system comprising a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a product or system. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the product or system that includes said element.

[0028] like Figures 1 to 4 As shown, this embodiment provides an early warning method for refrigerant leakage in an air conditioner. The air conditioner includes a compressor and an indoor fan. After the air conditioner is started, the early warning method includes: S1. Determine if the air conditioner has entered a stable operating state; S2. After the air conditioner enters a stable operating state, the current operating parameters of the air conditioner are collected according to the preset sampling period. The current operating parameters include at least the current return air temperature, current exhaust air temperature, and current operating frequency related to the compressor, as well as the current indoor temperature, current outdoor temperature, and current speed of the indoor fan. Specifically, the collection period is 30 seconds. S3. Preprocess the collected current operating parameters and calculate the current characteristic parameters based on the preprocessed operating parameters; among which, the current characteristic parameters include at least superheat, unit frequency energy efficiency ratio change rate and return gas temperature rise slope; S4. Compare each current feature parameter with its corresponding calculation reference band. If at least one current feature parameter continuously deviates from its corresponding calculation reference band within a preset number of sampling periods, then the current feature parameter is determined to be abnormal. The calculation reference band is a calculation reference interval and is the fluctuation range of normal operating parameters determined based on the historical normal operating parameters of the air conditioner and the current indoor temperature and current outdoor temperature. Preferably, the preset number is 5. S5. Based on the current characteristic parameters of each anomaly, determine whether there is an early leakage fault of the air conditioning refrigerant.

[0029] Thus, through the organic combination of "multi-dimensional parameter acquisition → multi-feature parameter calculation → dynamic benchmark comparison → persistent anomaly judgment", a synergistic effect of 1+1>2 is generated: Superheat reflects the heat exchange efficiency of the evaporator and is sensitive to changes in refrigerant flow rate; The rate of change of energy efficiency ratio per unit frequency reflects the cooling / heating efficiency of the compressor per unit frequency, and can eliminate the interference of changes in operating frequency on energy efficiency judgment. The slope of the return gas temperature rise reflects the trend of return gas temperature change and has an early indication effect on return gas overheating caused by refrigerant deficiency. These three elements characterize the system's thermodynamic state from different dimensions, complementing each other and enabling accurate judgments to be made in the very early stages of refrigerant leakage. Through multi-dimensional operating parameter fusion, dynamic benchmark modeling, and continuous anomaly judgment, early and accurate warnings of refrigerant leakage and a significant reduction in false alarm rates are achieved without increasing hardware costs. At the same time, it has good scalability and significant safety enhancement effects.

[0030] Specifically, the formula for calculating superheat is: ; in, This refers to the compressor return gas temperature. The saturation temperature corresponding to the suction pressure is estimated from the "system refrigerant pressure" (if present) or directly from a mapping model established between the "compressor operating frequency" and the "return gas temperature"; if the hardware does not have a pressure sensor, it is usually calculated using the "compressor operating frequency (f)" and the "return gas temperature (f)". The saturation temperature can be obtained by looking up a table or by back-calculating the theoretical evaporation pressure using empirical formulas. In low-cost models lacking high-pressure / low-pressure sensors, superheat is often indirectly characterized by the deviation between the return gas temperature and the reference return gas temperature after environmental compensation, or approximated by the deviation between the compressor's operating frequency and the return gas temperature. P_suction is the suction pressure. The unit frequency energy efficiency ratio change rate refers to the rate of change of the energy efficiency ratio relative to the compressor operating frequency. The energy efficiency ratio (COP) is equal to the cooling capacity (or heating capacity) divided by the input power; since the cooling capacity cannot be directly measured, this embodiment uses a relative change rate to indirectly characterize it. Among them, the estimated cooling / heating capacity is usually related to the fan speed (N). fan The indoor and outdoor temperature difference (ΔT) and the compressor's operating frequency (f) are positively correlated; the current theoretical cooling capacity (Q) can be estimated using a regression model based on historical normal operating conditions. est ; Input power estimate: mainly depends on the compressor's operating frequency (f) and voltage / current (if applicable); simplified models often assume a non-linear relationship between power and frequency, P∝f; Calculation method: Calculate the estimated COP value at the current moment and compare it with the standard COP value in the "dynamic reference band" to obtain the difference ΔCOP; then differentiate with respect to time or calculate the slope of change within the sliding window, which is the "rate of change".

[0031] The logical thresholds for triggering an early warning are "the current refrigerant risk score exceeds the preset score" or "within a preset number of consecutive sampling periods, at least one current characteristic parameter continuously deviates from the corresponding calculation baseline band". These thresholds are not fixed, but are related to the tolerance for false alarm rate and the detection sensitivity requirements.

[0032] (1) Logical threshold and related parameters: Safety level for application scenarios: For R290 flammable refrigerant, the threshold should be more stringent (lower) in order to provide earlier warnings; User group preferences: Home users prioritize comfort and low false alarms, so thresholds can be appropriately relaxed; Commercial users prioritize device protection, so thresholds are tightened. Seasonal / climate characteristics: Noise is high during the hot summer months, and the threshold may be dynamically adjusted to avoid frequent alarms.

[0033] (2) Methods for obtaining logical thresholds: ROC curve analysis: During the R&D phase, a large amount of historical data on "minor leaks", "moderate leaks", and "normal fluctuations" were collected. Determine the optimal cutoff point: Plot the ROC curve (Receiver Operating Characteristic) of the anomaly score to find the best balance between sensitivity (recall) and specificity (1 - false alarm rate); Hierarchical settings: Low-risk threshold (e.g., 75 points): corresponds to a higher false positive rate, but ensures that no potential risks are missed; High-risk threshold (e.g., 90 points): corresponds to an extremely high confidence level, confirming that a serious leak has occurred; Dynamic adjustment (optional): These thresholds can be fine-tuned through cloud reinforcement learning based on user feedback (e.g., users click "ignore alarm" or "there is indeed a fault").

[0034] The following explains the method for determining the calculation reference band corresponding to each current feature parameter in step S4. The method for determining the calculation reference band corresponding to each current feature parameter includes: Based on the historical normal operating parameters of the air conditioner, a dynamic reference band is constructed that adapts to different operating conditions and corresponds to each current characteristic parameter. Calculate the temperature difference between indoors and outdoors based on the current indoor and outdoor temperatures; Calculate the compensation amount corresponding to each current characteristic parameter based on the current indoor-outdoor temperature difference and the reference temperature difference; Calculate the calculation reference band corresponding to each feature parameter based on the dynamic reference band and compensation amount corresponding to each current feature parameter.

[0035] In this way, the following technical effects can be achieved: (1) By constructing a dynamic reference band that adapts to different operating conditions, a corresponding reference value is established for each possible operating condition point, thereby ensuring that the refrigerant leakage detection maintains high accuracy across the entire operating condition range; (2) By introducing temperature difference compensation, the dynamic reference band is further corrected: the judgment threshold is appropriately relaxed under high temperature difference conditions and the judgment threshold is narrowed under low temperature difference conditions, so as to effectively distinguish between "real leakage" and "normal parameter fluctuation caused by environmental temperature difference". (3) The dynamic reference band ensures the macroscopic accuracy of the reference value as the working conditions change, while the real-time compensation ensures the microscopic accuracy of the reference value as the environment changes. The calculation reference band generated by the combination of the two has both the universality based on big data training and the accuracy for the current environmental conditions.

[0036] Furthermore, based on the current indoor-outdoor temperature difference and the reference temperature difference, the compensation amount corresponding to each current characteristic parameter is calculated, including: Subtract the reference temperature difference from the current indoor and outdoor temperature difference to obtain the compensated temperature difference; The compensation amount corresponding to each current characteristic parameter is equal to the compensation constant corresponding to each current characteristic parameter multiplied by the compensation temperature difference.

[0037] Specifically, such as Figure 4 As shown, the dynamic benchmark modeling and environmental adaptive compensation mechanism includes: (1) Dynamic reference band construction: A dynamic reference band is generated by training with historical normal operation data to characterize the reasonable range of key parameters such as return gas temperature deviation and exhaust temperature under different indoor and outdoor temperature differences; for example, in the cooling mode, when the outdoor temperature is 35℃, the set temperature is 26℃, and the fan speed is medium, the model outputs the expected return gas temperature range under this condition (such as 12±2℃). (2) Environmental adaptive compensation: The system introduces an environmental adaptive compensation algorithm, and performs attribution analysis on temperature deviation through regression analysis or lightweight machine learning models (such as random forest) to eliminate fluctuations caused by non-leakage factors such as high temperature and humidity and load changes; when the indoor and outdoor temperature difference increases and the return air temperature rises naturally, the system automatically relaxes the judgment threshold, and the compensation amount is calculated as follows: compensation amount = 0.3 × (current indoor and outdoor temperature difference - reference temperature difference); The input parameters of the model are: ① Current operating parameters: return air temperature, exhaust air temperature, indoor ambient temperature, outdoor ambient temperature, indoor fan speed, compressor operating frequency, and (optional) air conditioning refrigerant pressure; ② Operating condition identification: operating mode code (cooling=1, heating=0, etc.) and set temperature; ③ Current characteristic parameters: superheat, return temperature difference, temperature change slope over the past N sampling periods, and cumulative runtime, etc.; ④ Environmental compensation factors: indoor and outdoor temperature difference and humidity (if applicable).

[0038] The model's output parameters are: ① Current refrigerant risk score: A continuous value (e.g., 0-100) indicates the degree to which the current state deviates from the healthy baseline; ② Early leakage risk level: discrete category labels, such as {Normal, Low_Risk, Medium_Risk, High_Risk}.

[0039] ③ (Optional advanced) Leakage probability: 0~1.

[0040] (3) Reference band comparison: The controller compares the current operating point with the reference band. If the measured return gas temperature is continuously higher than 1.5 times the standard deviation of the upper limit of the range, the abnormal scoring mechanism is triggered. The compensation mechanism should not be activated indefinitely, otherwise it may mask the true fault; the activation conditions should be based on "steady-state determination" and "operating condition validity". Start-up conditions: Stable system operation: The compressor's continuous running time > Tstable (e.g., 10-15 minutes), and the fluctuation range of the compressor's operating frequency < Δf (e.g., ±5Hz); This is to ensure that heat exchange reaches a balance and eliminate transient interference during start-up and shutdown; Non-defrosting / defrosting phase: In heating mode, the outdoor unit defrosting process must be avoided; because the airflow and temperature change drastically at this time, and the baseline model becomes invalid.

[0041] Environmental parameters must be within the effective range: for example, the outdoor temperature must be within the effective range of the sensor and it must not be in an extreme shutdown protection state.

[0042] Compensation action: Once the above conditions are met, the controller calls the "environmental adaptive compensation algorithm" to obtain the baseline compensation value from the preset three-dimensional lookup table (LUT) or regression equation based on the current indoor and outdoor temperature difference and fan speed, correct the original measurement value, or adjust the upper and lower limits of the dynamic reference band.

[0043] "Reference temperature difference" refers to "the relationship between the return gas temperature under theoretical normal conditions and the ambient temperature / set temperature", or "the reference reference when constructing the dynamic reference zone". (1) Relevant parameters: ① Outdoor ambient temperature (Tout): determines the condensation heat dissipation capacity; ② Indoor set temperature (Tset) or indoor return air temperature (Troom_returnT): determines the evaporation requirements; ③ Compressor frequency (f): determines the amount of refrigerant circulating; ④ Indoor fan speed (n): determines air volume and heat transfer coefficient; (2) How to obtain (construction method): ① Offline training / calibration: Collect return gas temperature data of "healthy machines" under different operating conditions (different combinations of Tout, Tset, f, n) at factory test benches or a large number of user homes; ② Establish a mapping model: Use multiple linear regression or a lightweight neural network to establish the function Tsuction_baseline=f(Tout, Tset, f, n).

[0044] ③ Online updates: For new models, a general benchmark is used initially; as data accumulates in the cloud, a more accurate benchmark model for specific models is released via OTA.

[0045] The standard deviation is explained below; the standard deviation represents "the normal fluctuation range of air conditioner operation under a specific operating condition", which includes natural fluctuations caused by non-fault factors such as sensor error, minor load fluctuations and air flow disturbances.

[0046] If the measured value exceeds "benchmark value + k * standard deviation", it indicates a systematic deviation that exceeds normal random fluctuations, which is highly likely to be a fault (such as a leak).

[0047] Standard deviation refers to "the standard deviation of the distribution of historical health data under the same working conditions"; Grouping criteria: Historical health data are clustered or binned according to (outdoor temperature range, indoor set temperature range, frequency level, fan level). Calculation object: Calculate the standard deviation σ of the return gas temperature (or other characteristics) of all historical healthy samples falling into the same "condition chamber"; Dynamic nature: This standard deviation is not globally uniform, but varies with operating conditions; for example, in high temperature and high humidity environments, due to factors such as condensation, the temperature fluctuation itself is large, and σ will become larger, thus avoiding false alarms; under ideal operating conditions of constant temperature and humidity, σ is smaller, and the detection is more sensitive.

[0048] The following provides further explanation of step S5, which includes: S51. Calculate the current refrigerant risk score based on the current characteristic parameters of each anomaly. S52. Compare the current refrigerant risk score with the preset score; S53. Based on the comparison results between the current refrigerant risk score and the preset score, determine whether there is an early leakage fault of the air conditioning refrigerant; Furthermore, S51 includes: S511. Determine the current operating mode of the air conditioner; S512. Based on the current operating mode of the air conditioner, the current indoor temperature, and the current outdoor temperature, determine the weights corresponding to the current characteristic parameters of each anomaly. S513. Based on the current characteristic parameters and corresponding weights of each anomaly, a weighted calculation is performed to obtain the current refrigerant risk score; wherein the current refrigerant risk score ranges from 0 to 100.

[0049] In this way, the following technical effects can be achieved: (1) By dynamically adjusting the weights according to the current operating mode, the risk score is always dominated by the most sensitive and indicative feature parameter under different working modes, avoiding the scoring distortion caused by "too low weight of sensitive parameters and too high weight of insensitive parameters" in non-design working conditions under the fixed weight system. From the perspective of algorithm mechanism, the scientificity and accuracy of the scoring results are ensured in the full range of working conditions. (2) In addition to determining the weights based on the operating mode, the weights are further adjusted based on the "current indoor temperature and current outdoor temperature", thus constructing a dual adaptive weight adjustment mechanism of "mode level + operating condition level". The indoor and outdoor temperature difference is a key environmental variable affecting the normal fluctuation range of each characteristic parameter. When the indoor and outdoor temperature difference is large, the system heat load increases and the normal fluctuation range of each operating parameter expands accordingly. At this time, the weight of parameters sensitive to fluctuation (such as superheat) should be appropriately reduced to avoid normal load fluctuations being misjudged as leakage. When the temperature difference is small, the system operation tends to be stable and the parameter fluctuations are more due to changes in the system's own state. At this time, the weights of each abnormal parameter should be increased accordingly to improve the ability to capture weak leakage signals. (3) The weights are set as variables related to the operating mode and temperature, rather than fixed constants. This feature provides a key adjustable parameter space for continuous optimization of models driven by big data in the cloud.

[0050] The following explains the method for determining the weights corresponding to the current feature parameters of each anomaly. S512 includes: Input the current operating mode of the air conditioner, the current indoor temperature, and the current outdoor temperature into the pre-trained weighted model; The weighted model outputs the weights corresponding to the current feature parameters of each anomaly.

[0051] Specifically, when the current operating mode of the air conditioner is cooling mode, the weight allocation of each abnormal current characteristic parameter satisfies the following: the weight of superheat ranges from 0.15 to 0.25, the weight of the unit frequency energy efficiency ratio change rate ranges from 0.05 to 0.15, and the weight of the return air temperature rise slope ranges from 0.35 to 0.45; preferably, the weight of superheat is 0.2, the weight of the unit frequency energy efficiency ratio change rate is 0.1, and the weight of the return air temperature rise slope is 0.4.

[0052] Depending on the current operating mode (cooling / heating), indoor and outdoor temperature difference, and other operating conditions, the weights of the current characteristic parameters of each anomaly have a certain variation pattern, following the principle of "sensitivity priority"; Impact of Cooling Mode: The most significant characteristic of refrigerant leakage is insufficient liquid supply to the evaporator, leading to increased return gas temperature and superheat. Therefore, in cooling mode, return gas temperature and its derived superheat have the highest weight, followed by exhaust temperature. Impact of heating mode: Refrigerant leakage can lead to insufficient liquid supply to the condenser, causing an abnormal increase in exhaust temperature (because the compression ratio may change and gas overheating increases), while the return gas temperature may decrease or fluctuate due to poor indoor heat exchange; therefore, in heating mode, the weight of exhaust temperature and its rate of change should be higher than that of return gas temperature. Temperature difference impact: When the temperature difference between indoors and outdoors is extremely large (extreme operating conditions), the system is in a high load area, with high noise and drastic fluctuations in single parameters. At this time, the weight of single-point temperature should be appropriately reduced and the weight of trend features (such as slope and rate of change) should be increased to enhance anti-interference ability.

[0053] The following provides further explanation of step S53, which includes: When the current refrigerant risk score exceeds the preset score, it is determined that there is an early leakage fault in the air conditioning refrigerant. If the current refrigerant risk score does not exceed the preset score, it is determined that there is no early leakage fault of the air conditioning refrigerant.

[0054] After confirming an early refrigerant leak in the air conditioner, early warning methods also include: S6. Determine the early leakage risk level of the air conditioning refrigerant based on the current refrigerant risk score; the higher the current refrigerant risk score, the greater the early leakage risk level of the air conditioning refrigerant. S7. Transmit the early leakage risk level to the user terminal; specifically, push the warning information to the user terminal through communication modules such as Wi-Fi or 4G.

[0055] S8. Upload the log of this operation to the cloud platform for iterative optimization of the early warning model; The S9 cloud platform retrains a more accurate early warning model based on massive amounts of data and pushes it to all networked devices via OTA, enabling the system to become more accurate with use.

[0056] In this way, the following technical effects can be achieved: (1) A complete data closed loop and model evolution link has been established. After each confirmed leakage fault, the operation log containing key information such as abnormal feature parameters, operating conditions, and risk scores is uploaded to the cloud platform, so that the cloud can accumulate a large number of "real leakage samples" and "normal operation samples". The cloud platform can continuously optimize the core algorithm parameters such as baseline parameters, preset score thresholds, and weight matrices by retraining the model based on these incremental data, and push them to all networked devices via OTA. (2) The rating is mapped to risk levels with clear semantics, such as "low risk, medium risk and high risk", and presented in an intuitive way on user terminals (mobile APP, smart speaker, display screen, etc.), which greatly reduces the user's understanding threshold and makes the warning information truly readable and actionable; different risk levels can trigger differentiated follow-up actions: low risk can only prompt the user to pay attention, medium risk can suggest contacting after-sales inspection, and high risk can actively push a work order to the after-sales service platform; this graded response mechanism realizes the precise connection between user notification and after-sales service, avoids the user paralysis effect caused by "alarming all", and improves the practical value and user trust of the warning system; (3) By aggregating the operation logs of actual leakage cases distributed around the world and in various usage environments to the cloud, a real leakage database covering multiple working conditions, multiple regions, multiple seasons, and multiple usage habits can be built.

[0057] Furthermore, S6 includes: Determine the preset score range within which the current refrigerant risk score falls; When the current refrigerant risk score is within the first preset score range, the early leakage risk level of the air conditioning refrigerant is determined to be low risk. When the current refrigerant risk score is in the second preset score range, the early leakage risk level of the air conditioning refrigerant is determined to be medium risk. When the current refrigerant risk score is in the third preset score range, the early leakage risk level of the air conditioning refrigerant is determined to be high risk; The preset score intervals include a first preset score interval, a second preset score interval, and a third preset score interval; the first preset score interval is (75, 80], the second preset score interval is (80, 90], and the third preset score interval is (90, 100).

[0058] Alternatively, based on the early leakage risk level, it can be divided into three categories: "low risk (leakage amount <5%, inspection recommended)," "medium risk (5%-10%, performance degradation)," and "high risk (>10%, immediate shutdown)." The following explains the specific methods for determining whether the air conditioner has entered a stable operating state in step S1. These methods include: Obtain the continuous running time and fluctuation range of the compressor's operating frequency; Determine whether the continuous running time of the compressor is greater than the preset time, whether the minimum threshold of the fluctuation range of the operating frequency is greater than the first preset frequency, and whether the maximum threshold of the fluctuation range of the operating frequency is less than the second preset frequency. When the continuous running time of the compressor exceeds the preset duration, the minimum threshold of the fluctuation range of the operating frequency is greater than the first preset frequency, and the maximum threshold of the fluctuation range of the operating frequency is less than the second preset frequency, the air conditioner is determined to have entered a stable operating state.

[0059] The preset duration is 10 minutes, the first preset frequency is -5Hz, and the second preset frequency is 5Hz.

[0060] The following describes the method for preprocessing the collected current operating parameters in step S3. The method for preprocessing the collected current operating parameters includes at least filtering and noise reduction, 3σ principle outlier removal, and Min-Max normalization.

[0061] In addition, the current operating parameters also include vibration signals from the compressor housing.

[0062] like Figure 5 As shown, this embodiment also provides an early warning system for air conditioning refrigerant leakage, including: The data acquisition unit is used to collect the current operating parameters of the air conditioner in real time according to a preset sampling period; the current operating parameters include at least the current return air temperature, current exhaust air temperature and current operating frequency related to the compressor, as well as the current indoor temperature, current outdoor temperature and current speed of the indoor fan; The controller, which is communicatively connected to the acquisition unit, is used to determine whether the air conditioner has entered a stable operating state. It receives operating parameters collected by the acquisition unit and preprocesses the current operating parameters, calculating current characteristic parameters based on the preprocessed parameters. Each current characteristic parameter is compared with its corresponding calculation reference band. If at least one current characteristic parameter continuously deviates from its corresponding calculation reference band within a preset number of sampling periods, the current characteristic parameter is determined to be abnormal. Based on these abnormal current characteristic parameters, it determines whether there is an early leakage fault in the air conditioner refrigerant. The current characteristic parameters include at least superheat, unit frequency energy efficiency ratio change rate, and return gas temperature rise slope. The calculation reference band corresponding to each current characteristic parameter is determined by the air conditioner's historical normal operating parameters and the current indoor and outdoor temperatures. The user terminal is connected to the controller and is used to receive early refrigerant leakage fault information transmitted by the controller. The cloud platform communicates with the controller to receive operating parameters transmitted by the controller, retrain and optimize the early warning model, and then remotely transmits them to each controller that communicates with the cloud platform via OTA.

[0063] Specifically, the data acquisition unit is a sensor; such as Figure 3As shown, if refrigerant pressure exists in the air conditioner, the current operating parameters also include refrigerant pressure; the controller has a built-in early warning model for refrigerant leakage, which is used to perform fusion analysis on the collected current operating parameters and output early warning information; the user terminal includes a mobile APP or an indoor unit display screen, and the user terminal and the controller are connected via Wi-Fi, and the cloud platform and the controller are connected via Wi-Fi; the controller calculates the current refrigerant risk score based on the current characteristic parameters of each anomaly, and compares the current refrigerant risk score with a preset score; if the current refrigerant risk score exceeds the preset score, it is determined that there is an early refrigerant leakage fault in the air conditioner; if the current refrigerant risk score does not exceed the preset score, it is determined that there is no early refrigerant leakage fault in the air conditioner.

[0064] This embodiment also provides an air conditioner, which employs any of the above-described methods for early warning of refrigerant leakage, or includes any of the above-described systems for early warning of refrigerant leakage; the air conditioner is an air conditioner using R290 refrigerant.

[0065] In summary, the early warning method and system for air conditioning refrigerant leakage provided by this invention have the following significant technical advantages and practical benefits compared to existing technologies: (1) To achieve early and accurate warning of refrigerant leakage; By integrating multi-dimensional operating parameters such as compressor return gas temperature, exhaust temperature, indoor and outdoor ambient temperature, fan speed, operating frequency, and refrigerant pressure (if any), a dynamic benchmark model and anomaly scoring mechanism are constructed. This can identify the performance degradation trend in the early stage of minor refrigerant leakage and issue early warnings several hours to several days earlier than the traditional threshold method, greatly improving detection sensitivity and response timeliness. (2) Significantly reduces false alarm rate and improves system robustness; An environmental adaptive compensation algorithm is introduced to dynamically adjust the judgment threshold based on the current indoor and outdoor temperature difference, set temperature, operating mode and other conditions, effectively distinguishing between "real leakage" and "normal load fluctuation". For example, in high temperature and high humidity environments, the temperature deviation tolerance is automatically relaxed to avoid false judgments caused by environmental interference, and the false alarm rate is reduced by more than 60% compared with traditional methods. (3) Zero new hardware required, strong compatibility, cost reduction and efficiency improvement; By making full use of the existing sensors and operation log data in the air conditioning controller, there is no need to configure additional refrigerant concentration sensors or high-precision pressure sensors, and no increase in BOM cost, which has good industrialization and promotion value.

[0066] (4) Supports leakage trend prediction and risk level assessment; Based on time series analysis and machine learning models (such as LSTM or random forest), the leakage level is quantitatively assessed and a three-level warning level of "low risk, medium risk, and high risk" is output. It supports remote cloud platform push of maintenance suggestions, providing scientific decision-making basis for users and after-sales service, and improving the intelligent service level of products. (5) Enhanced safety, especially suitable for models using flammable refrigerants; For air conditioning products that use flammable refrigerants such as R290, early detection of minor leaks can effectively prevent them from accumulating in enclosed spaces, significantly reducing the risk of combustion or explosion, meeting the increasingly stringent global market requirements for home appliance safety, and enhancing the brand's international competitiveness. (6) It has strong scalability and supports OTA remote upgrades and model optimization; The early warning model can be continuously optimized through cloud training, supports OTA remote updates, and achieves the self-evolution capability of "getting smarter with use", laying the technical foundation for building an air conditioning health management system in the future.

[0067] Exemplary embodiments of this disclosure have been specifically shown and described above. It should be understood that this disclosure is not limited to the detailed structures, arrangements, or implementations described herein; rather, this disclosure is intended to cover various modifications and equivalent arrangements contained within the spirit and scope of the appended claims.

Claims

1. A method for early warning of refrigerant leakage in an air conditioner, the air conditioner comprising a compressor and an indoor fan; characterized in that, After the air conditioner is turned on, the early warning method includes: Determine whether the air conditioner has entered a stable operating state; Once the air conditioner enters a stable operating state, its current operating parameters are collected according to a preset sampling period. The current operating parameters include at least the current return air temperature, current exhaust air temperature, and current operating frequency related to the compressor, as well as the current indoor temperature, current outdoor temperature, and current speed of the indoor fan. The collected current operating parameters are preprocessed, and the current characteristic parameters are calculated based on the preprocessed operating parameters; wherein, the current characteristic parameters include at least superheat, unit frequency energy efficiency ratio change rate, and return gas temperature rise slope; Each of the current feature parameters is compared with its corresponding calculation reference band. If at least one current feature parameter deviates from its corresponding calculation reference band within a consecutive preset number of sampling periods, the current feature parameter is determined to be abnormal. The calculation reference band is a calculation reference interval and is the fluctuation range of normal operating parameters determined based on the historical normal operating parameters of the air conditioner and the current indoor temperature and current outdoor temperature. Based on the current characteristic parameters of each anomaly, it is determined whether there is an early leakage fault of the air conditioning refrigerant.

2. The early warning method for air conditioning refrigerant leakage according to claim 1, characterized in that, The method for determining the calculation reference band corresponding to each of the current feature parameters includes: Based on the historical normal operating parameters of the air conditioner, a dynamic reference band is constructed that adapts to different operating conditions and corresponds to each of the current characteristic parameters. Calculate the current indoor and outdoor temperature difference based on the current indoor and outdoor temperatures; Based on the current indoor-outdoor temperature difference and the reference temperature difference, calculate the compensation amount corresponding to each of the current characteristic parameters; Calculate the calculation reference band corresponding to each of the current feature parameters based on the dynamic reference band and compensation amount.

3. The early warning method for air conditioning refrigerant leakage according to claim 2, characterized in that, The calculation of the compensation amount corresponding to each of the current characteristic parameters based on the current indoor-outdoor temperature difference and the reference temperature difference includes: Subtracting the reference temperature difference from the current indoor-outdoor temperature difference yields the compensated temperature difference; The compensation amount corresponding to each of the current feature parameters is equal to the compensation constant corresponding to each of the current feature parameters multiplied by the compensation temperature difference.

4. The early warning method for air conditioning refrigerant leakage according to claim 1, characterized in that, Determining whether the air conditioning refrigerant has an early leakage fault based on the current characteristic parameters of each anomaly includes: Calculate the current refrigerant risk score based on the current characteristic parameters of each anomaly; Compare the current refrigerant risk score with the preset score; Based on the comparison between the current refrigerant risk score and the preset score, it is determined whether there is an early leakage fault in the air conditioning refrigerant.

5. The early warning method for air conditioning refrigerant leakage according to claim 4, characterized in that, The calculation of the current refrigerant risk score based on the current characteristic parameters of each anomaly includes: Determine the current operating mode of the air conditioner; Based on the current operating mode of the air conditioner, the current indoor temperature, and the current outdoor temperature, determine the weights corresponding to the current characteristic parameters of each anomaly. The current refrigerant risk score is obtained by weighted calculation based on the current characteristic parameters and corresponding weights of each anomaly; wherein the value range of the current refrigerant risk score is 0 to 100.

6. The early warning method for air conditioning refrigerant leakage according to claim 5, characterized in that, When the current operating mode of the air conditioner is cooling mode, the weight allocation of each abnormal current characteristic parameter satisfies the following: the weight of the superheat is in the range of 0.15 to 0.25, the weight of the unit frequency energy efficiency ratio change rate is in the range of 0.05 to 0.15, and the weight of the return air temperature rise slope is in the range of 0.35 to 0.

45.

7. The early warning method for air conditioning refrigerant leakage according to claim 5, characterized in that, The step of determining the weights corresponding to the current characteristic parameters of each anomaly based on the current operating mode of the air conditioner, the current indoor temperature, and the current outdoor temperature includes: Input the current operating mode of the air conditioner, the current indoor temperature, and the current outdoor temperature into the pre-trained weighted model; The weighted model is controlled to output the weights corresponding to the current feature parameters of each anomaly.

8. The early warning method for air conditioning refrigerant leakage according to claim 4, characterized in that, The step of determining whether the air conditioning refrigerant has an early leakage fault based on the comparison result of the current refrigerant risk score and the preset score includes: When the current refrigerant risk score exceeds the preset score, it is determined that the air conditioning refrigerant has an early leakage fault; If the current refrigerant risk score does not exceed the preset score, it is determined that there is no early leakage fault of the air conditioning refrigerant.

9. The early warning method for air conditioning refrigerant leakage according to claim 8, characterized in that, After determining that the air conditioning refrigerant has an early leakage fault, the early warning method further includes: Based on the current refrigerant risk score, determine the early leakage risk level of the air conditioning refrigerant; The early leakage risk level is transmitted to the user terminal.

10. The early warning method for air conditioning refrigerant leakage according to claim 9, characterized in that, The determination of the early leakage risk level of the air conditioning refrigerant based on the current refrigerant risk score includes: Determine the preset score range in which the current refrigerant risk score falls; When the current refrigerant risk score is within the first preset score range, the early leakage risk level of the air conditioning refrigerant is determined to be low risk. When the current refrigerant risk score is within the second preset score range, the early leakage risk level of the air conditioning refrigerant is determined to be medium risk. When the current refrigerant risk score is in the third preset score range, the early leakage risk level of the air conditioning refrigerant is determined to be high risk; The preset score intervals include a first preset score interval, a second preset score interval, and a third preset score interval; the first preset score interval is (75, 80], the second preset score interval is (80, 90], and the third preset score interval is (90, 100).

11. The early warning method for air conditioning refrigerant leakage according to claim 1, characterized in that, The method for determining whether the air conditioner has entered a stable operating state includes: Obtain the continuous running time and the fluctuation range of the operating frequency of the compressor; Determine whether the continuous running time of the compressor is greater than a preset time, whether the minimum threshold of the fluctuation range of the operating frequency is greater than a first preset frequency, and whether the maximum threshold of the fluctuation range of the operating frequency is less than a second preset frequency. When the continuous operating time of the compressor exceeds the preset duration, the minimum threshold of the fluctuation range of the operating frequency is greater than the first preset frequency, and the maximum threshold of the fluctuation range of the operating frequency is less than the second preset frequency, the air conditioner is determined to have entered a stable operating state.

12. The early warning method for air conditioning refrigerant leakage according to claim 1, characterized in that, The method for preprocessing the collected current operating parameters includes at least filtering and noise reduction, 3σ principle outlier removal, and Min-Max normalization.

13. The early warning method for air conditioning refrigerant leakage according to claim 1, characterized in that, The current operating parameters also include the vibration signal of the compressor housing.

14. An early warning system for air conditioning refrigerant leakage, characterized in that, include: The data acquisition unit is used to acquire the current operating parameters of the air conditioner according to a preset sampling period; wherein the current operating parameters include at least the current return air temperature, current exhaust air temperature and current operating frequency related to the compressor, as well as the current indoor temperature, current outdoor temperature and current speed of the indoor fan; The controller is communicatively connected to the acquisition unit; it is used to determine whether the air conditioner has entered a stable operating state, receive the operating parameters collected by the acquisition unit and preprocess the current operating parameters, calculate the current characteristic parameters based on the preprocessed operating parameters, compare each current characteristic parameter with the corresponding calculation reference band, and determine that the current characteristic parameter is abnormal when at least one current characteristic parameter continuously deviates from the corresponding calculation reference band within a consecutive preset number of sampling periods; based on each abnormal current characteristic parameter, determine whether the air conditioner refrigerant has an early leakage fault; wherein, the current characteristic parameters include at least superheat, unit frequency energy efficiency ratio change rate and return air temperature rise slope; the calculation reference band corresponding to each current characteristic parameter is determined by the historical normal operating parameters of the air conditioner and the current indoor temperature and current outdoor temperature.

15. The early warning system for air conditioning refrigerant leakage according to claim 14, characterized in that, It also includes user terminals and cloud platforms; The user terminal and the controller are connected in communication and the user terminal is used to receive the early refrigerant leakage fault information of the air conditioning refrigerant transmitted by the controller. The cloud platform and controller are connected in communication, and the cloud platform is used to receive the operating parameters transmitted by the controller and retrain and optimize the early warning model, and then remotely transmit them to each controller connected in communication with the cloud platform via OTA.

16. An air conditioner, characterized in that, The air conditioner employs the early warning method for air conditioner refrigerant leakage as described in any one of claims 1 to 13, or includes the early warning system for air conditioner refrigerant leakage as described in any one of claims 14 to 15. The air conditioner uses R290 refrigerant.

Images