Air conditioner refrigerant leakage early warning control method, air conditioner and storage medium

Abstract

The application provides an air conditioner refrigerant leakage early warning control method, an air conditioner and a storage medium. The method comprises the following steps: real-time acquisition of multi-modal data; construction of a multi-dimensional feature vector according to current multi-modal data, prediction of the multi-dimensional feature vector by using a preset model, and acquisition of a current refrigerant leakage probability; acquisition of historical multi-modal data within a first preset time length before the current time and a current refrigerant concentration early warning threshold, prediction of the historical multi-modal data by using a preset time sequence prediction model, acquisition of a refrigerant predicted concentration within a second preset time length after the current time, and determination of a leakage risk prediction level according to the refrigerant predicted concentration and the current refrigerant concentration early warning threshold; determination of an air conditioner control strategy according to the current refrigerant leakage probability and the leakage risk prediction level, and control of the air conditioner to operate according to the air conditioner control strategy. The air conditioner refrigerant leakage early warning control method provided by the application can improve the accuracy of refrigerant leakage determination and solve the problem of response lag.

Description

Technical Field

[0001] This invention relates to the field of air conditioning technology, specifically to an air conditioning refrigerant leakage early warning and control method, an air conditioner using the air conditioning refrigerant leakage early warning and control method, and a computer-readable storage medium using the air conditioning refrigerant leakage early warning and control method. Background Technology

[0002] With the increasing emphasis on environmental regulations regarding the use of flammable refrigerants such as R290 (propane) in the air conditioning industry, refrigerant leakage safety has become a focal point. Existing detection and control methods suffer from the following limitations: 1. Most existing technologies rely on fixed refrigerant concentration thresholds as alarm criteria. However, factors such as ambient temperature and humidity, background gas interference, and sensor drift can cause numerous false alarms or missed alarms under different operating conditions. 2. Existing solutions often separate refrigerant gas detection from system operating parameters. For example, relying solely on refrigerant gas sensor alarms ignores changes in physical quantities such as system pressure, current, and energy efficiency ratio that inevitably accompany refrigerant leakage; or relying solely on performance parameters for inference lacks direct gas evidence, leading to delayed response.

[0003] For example, one existing method for refrigerant leak detection and control determines whether a refrigerant leak has occurred in the air conditioner and the extent of the leak by detecting changes in the amount of refrigerant in the system. This allows for the execution of appropriate control measures during system operation, minimizing damage and mitigating safety hazards. However, this method still suffers from the aforementioned problems of false alarms, missed alarms, and delayed response.

[0004] Therefore, there is an urgent need for a more optimized method for early warning and control of air conditioning refrigerant leakage. Summary of the Invention

[0005] The primary objective of this invention is to provide an air conditioning refrigerant leak early warning and control method that can improve the accuracy of refrigerant leak detection while solving the problem of response lag.

[0006] The second objective of this invention is to provide an air conditioner that can improve the accuracy of refrigerant leakage detection while solving the problem of response lag.

[0007] A third objective of this invention is to provide a computer-readable storage medium that can improve the accuracy of refrigerant leak detection while solving the problem of response lag.

[0008] To achieve the aforementioned first objective, the air conditioning refrigerant leakage early warning and control method provided by the present invention includes: real-time acquisition of multimodal data, wherein the multimodal data includes refrigerant concentration collected by a refrigerant sensor, and the multimodal data also includes at least one of the thermodynamic parameters, electrical parameters, and indoor environmental parameters of the air conditioner; constructing a multidimensional feature vector based on the current multimodal data, predicting the multidimensional feature vector using a preset model to obtain the current refrigerant leakage probability; acquiring historical multimodal data within a first preset time period before the current moment and the current refrigerant concentration early warning threshold, predicting the historical multimodal data using a preset time-series prediction model to obtain the refrigerant predicted concentration within a second preset time period after the current moment, determining the leakage risk prediction level based on the refrigerant predicted concentration and the current refrigerant concentration early warning threshold; determining an air conditioning control strategy based on the current refrigerant leakage probability and the leakage risk prediction level, and controlling the operation of the air conditioner according to the air conditioning control strategy.

[0009] As can be seen from the above scheme, the air conditioning refrigerant leakage early warning and control method of the present invention uses a preset model to predict the current refrigerant leakage probability of multi-dimensional feature vectors, thereby fusing multi-modal data and improving the accuracy of refrigerant leakage judgment. Simultaneously, based on a time-series prediction model, historical multi-modal data is predicted to obtain the refrigerant predicted concentration within a second preset time period after the current moment, achieving the effect of early warning in the early stage of leakage and solving the problem of response lag.

[0010] In a further proposed solution, the multidimensional feature vector is: ;in, Current refrigerant concentration; pressure change rate , This represents the compressor suction pressure at the current moment. The compressor suction pressure at the previous moment; current deviation. , Input current to the compressor at the current moment. Preset target input current; energy efficiency deviation , This represents the energy efficiency value at the current moment. The preset target energy efficiency value.

[0011] Therefore, it can be seen that the multidimensional feature vector, which integrates four parameters such as refrigerant concentration, pressure, current and energy efficiency, can better match the physical changes of refrigerant leakage and further improve the accuracy of leakage identification.

[0012] In a further proposed scheme, the current refrigerant concentration is the refrigerant concentration after compensation based on the original refrigerant concentration collected by the refrigerant sensor using indoor temperature and humidity.

[0013] Therefore, temperature and humidity compensation of the original refrigerant concentration can eliminate the sensor sensitivity shift caused by changes in ambient temperature and humidity, improve the detection accuracy of concentration data, and further reduce the false alarm rate.

[0014] In a further proposed solution, the current refrigerant concentration warning threshold is obtained using the following formula: ;in, , , These are preset weighting coefficients; The minimum volume concentration at which a refrigerant can burn or explode when it comes into contact with an ignition source; It is a nonlinear mapping function. + , , , It is a humidity contribution item. It is the relative humidity threshold. It is a contribution term for temperature fluctuations. It is the difference between the current indoor temperature and the previous temperature. It is the temperature change rate trigger threshold. It is the saturation threshold of the rate of temperature change; It is the standard deviation of refrigerant concentration within the first preset time period prior to the current moment. Average refrigerant concentration , This represents the total number of sampling points within the first preset time period. Indicates the first One sampling point, Representing the The time for each sampling point; and These are the weighting coefficients.

[0015] Therefore, the current refrigerant concentration warning threshold adopts a dynamic threshold, which is calculated and generated by comprehensively considering the refrigerant flammability safety concentration, ambient humidity, temperature fluctuations and long-term historical concentrations. It can automatically adjust the threshold size according to the sudden changes in indoor temperature and humidity, sensor baseline drift and other operating conditions, adapt to different usage environments and improve detection accuracy.

[0016] In a further scheme, the step of using a preset time-series prediction model to predict historical multimodal data and obtain the refrigerant predicted concentration within a second preset time period after the current moment includes: obtaining the historical compressor suction pressure change slope, historical compressor input current fluctuation amplitude, and historical refrigerant concentration change rate within a first preset time period based on historical multimodal data; and using the preset time-series prediction model to predict the historical compressor suction pressure change slope, historical compressor input current fluctuation amplitude, and historical refrigerant concentration change rate to obtain the refrigerant predicted concentration.

[0017] Therefore, using the slope of compressor suction pressure change, current fluctuation amplitude, and refrigerant concentration change rate as time-series input features is consistent with the evolution law of refrigerant leakage, which can improve the prediction accuracy of refrigerant concentration and ensure the accuracy of risk level determination.

[0018] In a further proposed solution, the steps for determining the leak risk prediction level based on the predicted refrigerant concentration and the current refrigerant concentration warning threshold include: obtaining the concentration difference between the predicted refrigerant concentration and the current refrigerant concentration warning threshold; when the concentration difference is less than a first preset difference, the leak risk prediction level is Level 1; when the concentration difference is greater than or equal to the first preset difference and less than or equal to the second preset difference, the leak risk prediction level is Level 2; and when the concentration difference is greater than the second preset difference, the leak risk prediction level is Level 3.

[0019] Therefore, dividing the leakage risk into three levels by concentration difference can facilitate matching corresponding control strategies and achieve refined hierarchical management.

[0020] In a further proposed solution, after the step of controlling the operation of the air conditioner according to the air conditioning control strategy, the solution further includes: obtaining the refrigerant concentration change rate after the air conditioner is controlled by the air conditioning control strategy; when the refrigerant concentration change rate is greater than zero, the leakage risk prediction level is increased, and the air conditioning control strategy is determined based on the increased leakage risk prediction level.

[0021] Therefore, by monitoring the refrigerant concentration change trend in real time after the control action is executed, when the concentration continues to rise, it indicates that the current control measures have failed to effectively suppress the leakage situation. The system automatically upgrades the risk level and switches to a higher level control strategy to solve the problem of insufficient control when the fixed control strategy is used to deal with the leakage escalation.

[0022] In a further proposed solution, there are two or more refrigerant sensors distributed in different areas of the current indoor space. After the step of real-time acquisition of multimodal data, the solution further includes: if the refrigerant concentration collected by two or more refrigerant sensors is greater than the current refrigerant concentration warning threshold, then the distance value and refrigerant concentration difference between any two refrigerant sensors are calculated, and the refrigerant concentration gradient index between the two refrigerant sensors is calculated based on the distance value and the refrigerant concentration difference; if the refrigerant concentration gradient index is greater than a preset gradient threshold, then the cross-correlation coefficient between the two refrigerant sensors is calculated, and the maximum time delay of the cross-correlation coefficient is calculated; the location of the leakage source between the two refrigerant sensors is determined based on the maximum time delay.

[0023] Therefore, by considering the spatial diffusion pattern of refrigerant and the logical correlation of sensor data at different locations, it is possible to distinguish between local leaks and global background interference. When a local leak occurs, the location of the leak source can be further confirmed so that users can investigate.

[0024] To achieve the second objective of the present invention, the present invention provides an air conditioner including a processor and a memory, the memory storing a computer program, and when the computer program is executed by the processor, it implements the steps of the above-described air conditioning refrigerant leakage early warning control method.

[0025] To achieve the third objective of the present invention, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a controller, implements the steps of the above-described air conditioning refrigerant leakage early warning control method. Attached Figure Description

[0026] Figure 1 This is a structural block diagram of an air conditioner that applies the air conditioning refrigerant leakage early warning and control method of the present invention.

[0027] Figure 2 This is a flowchart of an embodiment of the air conditioning refrigerant leakage early warning and control method of the present invention.

[0028] Figure 3 This is a flowchart illustrating how a preset time-series prediction model is used to predict historical multimodal data in an embodiment of the air conditioning refrigerant leakage early warning and control method of the present invention, thereby obtaining the predicted refrigerant concentration within a second preset time period after the current moment.

[0029] Figure 4 This is a schematic diagram of a preset time-series prediction model used in an embodiment of the air conditioning refrigerant leakage early warning and control method of the present invention.

[0030] Figure 5 This is a flowchart following the steps of controlling the operation of the air conditioner according to the air conditioner control strategy in an embodiment of the air conditioner refrigerant leakage early warning control method of the present invention.

[0031] Figure 6 This is a flowchart of real-time multimodal data acquisition in an embodiment of the air conditioning refrigerant leakage early warning and control method of the present invention.

[0032] The present invention will be further described below with reference to the accompanying drawings and embodiments. Detailed Implementation

[0033] Various exemplary embodiments of the invention will now be described in detail with reference to the accompanying drawings. The descriptions of the exemplary embodiments are merely illustrative and are in no way intended to limit the invention or its application or use. The invention can be embodied in many different forms and is not limited to the embodiments described herein. These embodiments are provided to make the invention thorough and complete, and to fully express the scope of the invention to those skilled in the art. It should be noted that, unless otherwise specifically stated, the relative arrangement of components and steps, the composition of materials, numerical expressions, and values ​​set forth in these embodiments should be interpreted as merely exemplary and not as limiting.

[0034] The terms "first," "second," and similar words used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different parts. Words such as "including" or "comprising" mean that the element preceding the word encompasses the element listed after it, without excluding the possibility of encompassing other elements. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0035] In this invention, when a specific device is described as being located between a first device and a second device, an intermediary device may or may not be present between the specific device and the first or second device. When a specific device is described as being connected to other devices, the specific device may be directly connected to the other devices without an intermediary device, or it may be not directly connected to the other devices but have an intermediary device.

[0036] All terms used in this invention (including technical or scientific terms) have the same meaning as understood by one of ordinary skill in the art, unless otherwise specifically defined. It should also be understood that terms defined in general dictionaries should be interpreted as having the meaning consistent with their meaning in the context of the relevant art, and not as having an idealized or highly formalized meaning, unless expressly defined herein.

[0037] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and equipment should be considered part of the specification.

[0038] Example of an air conditioning refrigerant leakage early warning and control method: The air conditioning refrigerant leakage early warning control method of the present invention is an application program used in air conditioners for refrigerant leakage early warning control.

[0039] In this embodiment, see Figure 1 The air conditioner includes a main control chip 1, a sensor group 2, and an actuator 3. Both the sensor group 2 and the actuator 3 are electrically connected to the main control chip 1. The sensor group 2 includes a refrigerant sensor, a thermodynamic sensor, an electrical sensor, and a temperature and humidity sensor. The actuator 3 includes a compressor, an indoor fan, an outdoor fan, and an electromagnetic expansion valve, among other actuators.

[0040] See Figure 2In this embodiment, the air conditioning refrigerant leakage early warning control method first executes step S1 to collect multimodal data in real time. This multimodal data includes refrigerant concentration collected by the refrigerant sensor, thermodynamic parameters of the air conditioner, electrical parameters of the air conditioner, and indoor environmental parameters. During air conditioner operation, multimodal data collection is synchronized with time and space, collecting the following multimodal data at a frequency of 1Hz: Refrigerant concentration includes refrigerant sensor concentration readings at locations such as the indoor unit return air vent, outdoor unit compressor compartment, or the ground. Thermodynamic parameters include compressor suction pressure, compressor discharge pressure, and inlet and outlet air temperatures of the evaporator and condenser. Electrical parameters include compressor input current, compressor input voltage, and fan speed. Environmental parameters include indoor ambient temperature and relative humidity. During data collection, all data can be timestamped and smoothed using a sliding window to eliminate transient pulse noise.

[0041] After acquiring the multimodal data, step S2 is executed. A multidimensional feature vector is constructed based on the current multimodal data, and a preset model is used to predict the multidimensional feature vector to obtain the current refrigerant leakage probability. In order to better reflect the physical laws of refrigerant leakage and improve the detection accuracy of refrigerant leakage, it is necessary to predict the current refrigerant leakage probability based on the current multimodal data.

[0042] In this embodiment, the multidimensional feature vector is: ;in, Current refrigerant concentration; pressure change rate , This represents the compressor suction pressure at the current moment. The compressor suction pressure at the previous moment; current deviation. , Input current to the compressor at the current moment. Preset target input current; energy efficiency deviation , This represents the energy efficiency value at the current moment. The preset target energy efficiency value is used. The current energy efficiency value can be calculated through current and voltage detection, which is a technique known to those skilled in the art and will not be elaborated here. The multi-dimensional feature vector integrates four parameters: refrigerant concentration, pressure, current, and energy efficiency, which can better reflect the physical changes of refrigerant leakage and further improve the accuracy of leak identification.

[0043] In this embodiment, the current refrigerant concentration is the refrigerant concentration after compensation based on the original refrigerant concentration collected by the refrigerant sensor using indoor temperature and humidity. When compensating for the original refrigerant concentration collected by the refrigerant sensor using indoor temperature and humidity, the corrected refrigerant concentration is obtained using a lookup table method or empirical formula. The refrigerant sensor's response sensitivity to the target gas is not constant but depends on the ambient temperature and relative humidity. Therefore, to eliminate nonlinear interference caused by environmental factors, temperature and humidity compensation is performed on the original refrigerant concentration to eliminate sensor sensitivity shifts caused by changes in ambient temperature and humidity, improve the detection accuracy of concentration data, and further reduce the false alarm rate.

[0044] In this embodiment, the preset model can be a pre-trained support vector machine (SVM) or random forest model. The multi-dimensional feature vector V(t) is input into the preset model, which outputs the current refrigerant leakage probability at the current moment. Support vector machine (SVM) or random forest models are well-known techniques to those skilled in the art and will not be elaborated upon here.

[0045] After obtaining the current refrigerant leakage probability, step S3 is executed to acquire historical multimodal data within a first preset time period before the current moment and the current refrigerant concentration warning threshold. A preset time-series prediction model is used to predict the historical multimodal data to obtain the predicted refrigerant concentration within a second preset time period after the current moment. The leakage risk prediction level is determined based on the predicted refrigerant concentration and the current refrigerant concentration warning threshold. The first and second preset time periods can be preset based on experimental data; for example, the first preset time period is 30 minutes and the second preset time period is 5 minutes. The preset time-series prediction model uses a pre-trained LSTM prediction model. The training method for the LSTM prediction model is a well-known technique to those skilled in the art and will not be described further here.

[0046] In this embodiment, the current refrigerant concentration warning threshold is obtained by the following formula: ;in, , , These are preset weighting coefficients; The minimum volume concentration at which a refrigerant can burn or explode when it comes into contact with an ignition source; It is a nonlinear mapping function. + , , , It is a humidity contribution item. It is the relative humidity threshold. It is a contribution term for temperature fluctuations. It is the difference between the current indoor temperature and the previous temperature. It is the temperature change rate trigger threshold (e.g., 1.0℃ / min to 2.0℃ / min). It is the saturation threshold of the rate of temperature change (e.g., 5.0 °C / min to 10.0 °C / min). It is the standard deviation of refrigerant concentration within the first preset time period prior to the current moment. Average refrigerant concentration , This represents the total number of sampling points within the first preset time period. Indicates the first One sampling point, Representing the The time for each sampling point; and These are the weighting coefficients, and The current refrigerant concentration warning threshold is a dynamic threshold, calculated by comprehensively considering the refrigerant's flammable safety concentration, ambient humidity, temperature fluctuations, and long-term historical concentrations. It can automatically adjust the threshold value according to changes in indoor temperature and humidity, sensor baseline drift, and other operating conditions, adapting to different usage environments and improving detection accuracy.

[0047] In this embodiment, see Figure 3 and Figure 4 When using a preset time-series prediction model to predict historical multimodal data and obtain the predicted refrigerant concentration for the second preset time period after the current moment, step S11 is first executed to obtain the historical compressor suction pressure change slope, historical compressor input current fluctuation amplitude, and historical refrigerant concentration change rate for the first preset time period based on the historical multimodal data. The historical compressor suction pressure change slope, historical compressor input current fluctuation amplitude, and historical refrigerant concentration change rate can be calculated from the historical multimodal data.

[0048] For example, the slope of historical compressor suction pressure changes can be obtained by linearly fitting the compressor suction pressure over a first preset time period, reflecting the pressure change trend. Assume there are M sampling points within the first preset time period. slope From the least squares formula: , in, This represents the time of the j-th sampling point. The average over time. , Let j be the compressor suction pressure at time j. The average value of the compressor suction pressure within the first preset time period.

[0049] The historical compressor input current fluctuation amplitude indicates the severity of input current fluctuations within a first preset time period, which can be measured by the standard deviation. , for The compressor input current at any given moment, The average value of the compressor input current within the first preset time period.

[0050] The historical refrigerant concentration change rate can be calculated using the least squares method: , for The refrigerant concentration at any given time The average refrigerant concentration within the first preset time period.

[0051] After obtaining the historical compressor suction pressure change slope, historical compressor input current fluctuation amplitude, and historical refrigerant concentration change rate, step S12 is executed. A preset time-series prediction model is used to predict these parameters to obtain the predicted refrigerant concentration. Using the compressor suction pressure change slope, current fluctuation amplitude, and refrigerant concentration change rate as time-series input features aligns with the evolution of refrigerant leakage, improving the accuracy of refrigerant concentration prediction and ensuring the accuracy of risk level assessment.

[0052] In this embodiment, the step of determining the leakage risk prediction level based on the predicted refrigerant concentration and the current refrigerant concentration warning threshold includes: obtaining the concentration difference between the predicted refrigerant concentration and the current refrigerant concentration warning threshold; when the concentration difference is less than a first preset difference, the leakage risk prediction level is Level 1 leakage risk level; when the concentration difference is greater than or equal to the first preset difference and less than or equal to a second preset difference, the leakage risk prediction level is Level 2 leakage risk level; when the concentration difference is greater than the second preset difference, the leakage risk prediction level is Level 3 leakage risk level. The first and second preset differences can be preset based on experimental data. Dividing the leakage risk level into three levels by segmenting the concentration difference facilitates matching corresponding control strategies and achieves refined hierarchical management.

[0053] After determining the leak risk prediction level, step S4 is executed. Based on the current refrigerant leak probability and the leak risk prediction level, an air conditioning control strategy is determined, and the air conditioner is operated according to this strategy. The current refrigerant leak probability indicates the reliability of the current leak, while the leak risk prediction level indicates the short-term trend of the refrigerant leak. Combining these two factors improves the accuracy of refrigerant leak detection and addresses the issue of response lag.

[0054] In this embodiment, the step of determining the air conditioning control strategy based on the current refrigerant leakage probability and leakage risk prediction level includes: if the current refrigerant leakage probability is within a first probability range and the leakage risk prediction level is a level one leakage risk level, maintain the current control strategy; if the current refrigerant leakage probability is within a first probability range and the leakage risk prediction level is a level two leakage risk level, maintain the current control strategy and issue a risk warning; if the current refrigerant leakage probability is within a first probability range and the leakage risk prediction level is a level three leakage risk level, maintain the current control strategy and issue a sensor fault warning; if the current refrigerant leakage probability is within a second probability range and the leakage risk prediction level is a level one leakage risk level, reduce the current compressor frequency by a first preset frequency range and control the indoor fan to operate at a first preset fan speed; if the current refrigerant leakage probability is within a second probability range and the leakage risk prediction level is a level two leakage risk level, reduce the current compressor frequency by a second preset frequency range and control the indoor fan to operate at a second preset fan speed. The indoor fan is controlled with a second preset frequency amplitude greater than the first preset frequency amplitude, and a second preset airflow speed greater than the first preset airflow speed. If the current refrigerant leakage probability is within the second probability range and the leakage risk prediction level is level three, the current compressor frequency is reduced to a third preset frequency amplitude, and the indoor fan is controlled with a third preset airflow speed, where the third preset frequency amplitude is greater than the second preset frequency amplitude, and the third preset airflow speed is greater than the second preset airflow speed. If the current refrigerant leakage probability is within the third probability range and the leakage risk prediction level is level one, the first preset forced ventilation mode is activated. If the current refrigerant leakage probability is within the third probability range and the leakage risk prediction level is level two, the second preset forced ventilation mode is activated, with a ventilation intensity greater than the first preset forced ventilation mode. If the current refrigerant leakage probability is within the third probability range and the leakage risk prediction level is level three, the compressor is powered off and shut down, and all electric exhaust valves are opened. The lower limit of the second probability range is greater than the upper limit of the first probability range, and the lower limit of the third probability range is greater than the upper limit of the second probability range. The first probability range, second probability range, third probability range, first preset frequency amplitude, first preset wind speed, second preset frequency amplitude, second preset wind speed, third preset frequency amplitude, and third preset wind speed can be preset based on experimental data. For example, the first probability range is greater than or equal to 0 and less than or equal to 0.3, the second probability range is greater than 0.3 and less than 0.8, the third probability range is greater than or equal to 0.8 and less than or equal to 1, the preset low speed is the lowest wind speed, the first preset frequency amplitude is 10%-30% of the current compressor frequency, the second preset frequency amplitude is 30%-50% of the current compressor frequency, the third preset frequency amplitude is 50%-70% of the current compressor frequency, the first preset wind speed is the low wind speed, the second preset wind speed is the medium wind speed, and the third preset wind speed is the high wind speed.

[0055] When the current refrigerant leakage probability is within the first probability range and the leakage risk prediction level is the first level of leakage risk, the system is in normal operating condition and running as usual. The current control strategy remains unchanged, and the current feature vector is incorporated into the background model for adaptive updating without warning or intervention.

[0056] When the current refrigerant leakage probability is within the first probability range and the leakage risk prediction level is the second level, a typical early warning or slight sensor drift will occur. During this period, no audible or visual alarm will be triggered, and only a risk warning will be displayed on the screen.

[0057] When the current refrigerant leakage probability is within the highest probability range and the leakage risk prediction level is level three, indicating almost no leakage at present but a high risk prediction for the future, it is highly likely due to a momentary sensor failure or rapid transient interference. In this case, all automatic protection actions should be forcibly rejected, a sensor fault should be reported, and manual intervention should be requested to maintain the current operating status.

[0058] When the current refrigerant leakage probability is in the second probability range and the leakage risk prediction level is the first level, maintain the operation status, reduce the current compressor frequency by 10%-30%, reduce the refrigerant circulation volume, and at the same time, turn on the indoor fan to run at low speed to promote air mixing and dilution, without warning.

[0059] When the current refrigerant leakage probability is in the second probability range and the leakage risk prediction level is level two, maintain the operation status, reduce the current compressor frequency by 30%-50%, reduce the refrigerant circulation volume, and at the same time, turn on the indoor fan to run at medium speed to promote air mixing and dilution, triggering a yellow warning notification.

[0060] When the current refrigerant leakage probability is in the second probability range and the leakage risk prediction level is the third leakage risk level, maintain the operation status, reduce the current compressor frequency by 50%-70%, reduce the refrigerant circulation volume, and at the same time, turn on the indoor fan to run at high speed to promote air mixing and dilution, triggering a yellow warning audible and visual notification.

[0061] When the current refrigerant leakage probability is in the third probability range and the leakage risk prediction level is the first level, a leak has indeed occurred, but the predicted hazard will dissipate quickly (e.g., the trace leak is immediately diluted by ventilation), posing no short-term threat. In this case, the first preset forced ventilation mode should be activated, triggering a yellow alert.

[0062] If the current refrigerant leakage probability is in the third probability range and the leakage risk prediction level is level two, a leakage is currently occurring and is predicted to reach a medium risk level, showing a tendency to worsen but not yet requiring immediate attention. The second preset forced ventilation mode should be activated, closing the liquid line solenoid valve on the condensate pipe, turning the outdoor fan to maximum speed, and simultaneously turning the indoor fan in reverse or at maximum forward speed to create directional airflow, triggering a yellow warning audible and visual notification. If the risk remains medium, it will be upgraded to high risk.

[0063] When the current refrigerant leakage probability is in the third probability range and the leakage risk prediction level is level three, the leakage is confirmed and is predicted to rapidly escalate to a high risk, requiring immediate intervention. Power off and shut down the compressor, open all electric exhaust valves, and continue alarming until manually reset and a red alert audible and visual notification is triggered.

[0064] In this embodiment, see Figure 5 After the step of controlling the air conditioner's operation according to the air conditioning control strategy, step S6 is also executed to obtain the refrigerant concentration change rate after the air conditioner is controlled by the air conditioning control strategy. In order to optimize the control strategy, it is necessary to detect the refrigerant concentration change rate after the air conditioner is controlled by the air conditioning control strategy for feedback adjustment.

[0065] After obtaining the refrigerant concentration change rate, proceed to step S7 to determine whether the refrigerant concentration change rate is greater than zero.

[0066] When the refrigerant concentration change rate is greater than zero, step S8 is executed to upgrade the leakage risk prediction level. The air conditioning control strategy is then determined based on the upgraded leakage risk prediction level. If the refrigerant concentration change rate is greater than zero, it indicates that the current strategy is ineffective, and the leakage risk prediction level can be automatically upgraded; for example, the leakage risk prediction level can jump directly from level one to level three.

[0067] When the refrigerant concentration change rate is less than or equal to zero, execute step S9 to lower the leak risk prediction level. Determine the air conditioning control strategy based on the lowered leak risk prediction level. A refrigerant concentration change rate less than or equal to zero indicates that the current strategy is effective. After confirming safety, the system gradually resumes normal operation. Simultaneously, this event can be recorded for model self-learning optimization.

[0068] In one embodiment, there are two or more refrigerant sensors, distributed in different areas of the current indoor space. See also Figure 6 After performing step S1, which involves real-time acquisition of multimodal data, step S31 is executed to determine whether there are two or more refrigerant sensors whose refrigerant concentrations exceed the current refrigerant concentration warning threshold. This can be achieved by reading the refrigerant concentrations collected by all refrigerant sensors and filtering out the set of sensors whose current readings exceed the current refrigerant concentration warning threshold; this set is referred to as: .in, This represents the total number of refrigerant sensors in the current indoor space.

[0069] If two or more refrigerant sensors collect refrigerant concentrations exceeding the current refrigerant concentration warning threshold, then step S32 is executed to calculate the distance and refrigerant concentration difference between any two refrigerant sensors. Based on the distance and refrigerant concentration difference, the refrigerant concentration gradient index between the two sensors is calculated. To distinguish between localized leakage and global interference, it is necessary to calculate the relationship between the concentration gradient and distance between each pair of refrigerant sensors in set A.

[0070] In this embodiment, the distance between any two refrigerant sensors is obtained by the following formula: Among them, sensors Spatial coordinates are The refrigerant concentration difference between any two refrigerant sensors is obtained by the following formula: The refrigerant concentration gradient exponent is obtained by the following formula: .

[0071] After obtaining the refrigerant concentration gradient index, step S33 is executed to determine whether the refrigerant concentration gradient index is greater than a preset gradient threshold. The preset gradient threshold can be set in advance based on experimental data.

[0072] If the refrigerant concentration gradient index is less than or equal to the preset gradient threshold, all sensor readings will increase synchronously, and the readings of each sensor will increase inversely. The lack of spatial diffusion directionality indicates that the leak is due to ambient background gas fluctuations and the source cannot be confirmed.

[0073] If the refrigerant concentration gradient exponent is greater than a preset gradient threshold, then step S34 is executed to calculate the cross-correlation coefficient between the two refrigerant sensors and the maximum time delay of the cross-correlation coefficient. If at least one pair of sensors (i, j) satisfies... , If a preset gradient threshold is set, a significant spatial concentration gradient is confirmed, initially ruling out global background interference, such as a synchronous increase caused by the entire room being filled with smoke. At this point, the sensor on the high-concentration side points towards the potential leak direction.

[0074] Since it takes time for gas to diffuse from the leak point, the concentration change of sensors closer to the leak source should lead that of sensors further away. Therefore, the time lag between adjacent sensor sequences can be calculated using a cross-correlation function.

[0075] In this embodiment, the cross-correlation coefficients of the two refrigerant sensors are obtained by the following formula: ;in, This is the time offset (lag steps). The mean within the sliding window. The maximum time delay of the cross-correlation coefficient is obtained by the following formula: .

[0076] After obtaining the maximum time delay of the cross-correlation coefficient, step S35 is executed to determine the location of the leak source between the two refrigerant sensors based on the maximum time delay. When determining the location of the leak source between the two refrigerant sensors, if... If the change in refrigerant sensor i is greater than that in refrigerant sensor j, it indicates that the gas diffusion direction is from refrigerant sensor i to refrigerant sensor j; otherwise, the diffusion direction is opposite. If the absolute value of the current concentration of sensor i is simultaneously satisfied... If the value is maximized, then the physical area where sensor i is located is the estimated location of the leak source. By considering the spatial diffusion pattern of the refrigerant and the logical correlation between sensor data from different locations, it is possible to distinguish between local leaks and global background interference. In the event of a local leak, the location of the leak source can be further confirmed to facilitate troubleshooting by the user.

[0077] Furthermore, when there are two or more refrigerant sensors, the refrigerant concentration with the highest concentration reading among all sensors can be used as the final current refrigerant concentration. Alternatively, the arithmetic mean of the concentration data from all refrigerant sensors can be used as the final current refrigerant concentration. Based on this current concentration, the probability of refrigerant leakage can be predicted to obtain an air conditioning control strategy.

[0078] As described above, the air conditioning refrigerant leakage early warning and control method of the present invention uses a preset model to predict the current refrigerant leakage probability from multi-dimensional feature vectors, thereby fusing multi-modal data and improving the accuracy of refrigerant leakage determination. Simultaneously, based on a time-series prediction model, historical multi-modal data is predicted to obtain the refrigerant predicted concentration within a second preset time period after the current moment, achieving early warning at the early stage of leakage and solving the problem of response lag.

[0079] Air conditioner example: The air conditioner in this embodiment includes a controller, which executes the steps in the above-described air conditioning refrigerant leakage early warning control method embodiment when executing a computer program.

[0080] For example, a computer program can be divided into one or more modules, one or more of which are stored in memory and executed by a controller to complete the present invention. One or more modules can be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in an air conditioner.

[0081] An air conditioner may include, but is not limited to, a controller and a memory. Those skilled in the art will understand that an air conditioner may include more or fewer components, or a combination of certain components, or different components; for example, an air conditioner may also include input / output devices, network access devices, buses, etc.

[0082] For example, the controller can be a Central Processing Unit (CPU), or other general-purpose controllers, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose controller can be a microcontroller or any conventional controller. The controller is the control center of the air conditioner, connecting all parts of the air conditioner through various interfaces and lines.

[0083] The memory can be used to store computer programs and / or modules. The controller implements various functions of the air conditioner by running or executing the computer programs and / or modules stored in the memory, and by calling the data stored in the memory. For example, the memory may mainly include a program storage area and a data storage area, wherein the program storage area may store the operating system, at least one application program required for a function, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, RAM, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0084] Examples of computer-readable storage media: If the modules integrated into the air conditioner in the above embodiments are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the air conditioner refrigerant leak early warning control method can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by the controller, it can implement the steps of the above embodiments of the air conditioner refrigerant leak early warning control method. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The storage medium can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content contained in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0085] It should be noted that the above are only preferred embodiments of the present invention, but the design concept of the invention is not limited thereto. Any non-substantial modifications made to the present invention using this concept also fall within the protection scope of the present invention.

Claims

1. A method for early warning and control of refrigerant leakage in air conditioning systems, characterized in that: include: Real-time acquisition of multimodal data is performed, wherein the multimodal data includes the refrigerant concentration collected by the refrigerant sensor, and the multimodal data also includes at least one of the thermodynamic parameters of the air conditioner, the electrical parameters of the air conditioner, and the indoor environmental parameters; A multidimensional feature vector is constructed based on the current multimodal data, and the multidimensional feature vector is predicted using a preset model to obtain the current refrigerant leakage probability. The system acquires historical multimodal data within a first preset time period before the current moment and the current refrigerant concentration warning threshold. It then uses a preset time-series prediction model to predict the historical multimodal data, obtains the refrigerant predicted concentration within a second preset time period after the current moment, and determines the leakage risk prediction level based on the refrigerant predicted concentration and the current refrigerant concentration warning threshold. An air conditioning control strategy is determined based on the current refrigerant leakage probability and the leakage risk prediction level, and the air conditioner is controlled to operate according to the air conditioning control strategy.

2. The air conditioning refrigerant leakage early warning and control method according to claim 1, characterized in that: The multidimensional feature vector is: ； in, This represents the current refrigerant concentration; Pressure change rate , This represents the compressor suction pressure at the current moment. This represents the compressor suction pressure at the previous moment; Current deviation , Input current to the compressor at the current moment. Set the target input current; Energy efficiency deviation , This represents the energy efficiency value at the current moment. The preset target energy efficiency value.

3. The air conditioning refrigerant leakage early warning and control method according to claim 2, characterized in that: The current refrigerant concentration is the refrigerant concentration after compensation based on the original refrigerant concentration collected by the refrigerant sensor using indoor temperature and humidity.

4. The air conditioning refrigerant leakage early warning and control method according to claim 1, characterized in that: The current refrigerant concentration warning threshold is obtained by the following formula: ； in, , , These are preset weighting coefficients; The minimum volume concentration at which a refrigerant can burn or explode when it comes into contact with an ignition source; It is a nonlinear mapping function. + , , , It is a humidity contribution item. It is the relative humidity threshold. It is a contribution term for temperature fluctuations. It is the difference between the current indoor temperature and the previous temperature. It is the temperature change rate trigger threshold. It is the saturation threshold of the rate of temperature change; It is the standard deviation of refrigerant concentration within the first preset time period prior to the current moment. Average refrigerant concentration , This represents the total number of sampling points within the first preset time period. Indicates the first One sampling point, Representing the The time for each sampling point; and These are the weighting coefficients.

5. The air conditioning refrigerant leakage early warning and control method according to any one of claims 1 to 4, characterized in that: The step of using a preset time-series prediction model to predict the historical multimodal data and obtain the predicted refrigerant concentration within a second preset time period after the current moment includes: Based on the historical multimodal data, the historical compressor suction pressure change slope, historical compressor input current fluctuation amplitude, and historical refrigerant concentration change rate within the first preset time period are obtained. The predicted refrigerant concentration is obtained by using a preset time-series prediction model to predict the slope of historical compressor suction pressure change, the amplitude of historical compressor input current fluctuation, and the rate of change of historical refrigerant concentration.

6. The air conditioning refrigerant leakage early warning and control method according to any one of claims 1 to 4, characterized in that: The steps for determining the leak risk prediction level based on the predicted refrigerant concentration and the current refrigerant concentration warning threshold include: Obtain the concentration difference between the predicted refrigerant concentration and the current refrigerant concentration warning threshold; When the concentration difference is less than the first preset difference, the leakage risk prediction level is a level one leakage risk level; When the concentration difference is greater than or equal to a first preset difference and less than or equal to a second preset difference, the leakage risk prediction level is a level two leakage risk level. When the concentration difference is greater than the second preset difference, the leakage risk prediction level is a level three leakage risk level.

7. The air conditioning refrigerant leakage early warning and control method according to any one of claims 1 to 4, characterized in that: After the step of controlling the operation of the air conditioner according to the air conditioning control strategy, the method further includes: Obtain the rate of change of refrigerant concentration after the air conditioner is operated under the aforementioned air conditioning control strategy; When the refrigerant concentration change rate is greater than zero, the leakage risk prediction level is increased, and the air conditioning control strategy is determined based on the increased leakage risk prediction level.

8. The air conditioning refrigerant leakage early warning and control method according to any one of claims 1 to 4, characterized in that: The number of refrigerant sensors is two or more, and the two or more refrigerant sensors are distributed in different areas of the current indoor space; After the steps of real-time acquisition of multimodal data, the following are also included: If the refrigerant concentration collected by two or more of the refrigerant sensors is greater than the current refrigerant concentration warning threshold, then the distance value and the refrigerant concentration difference between any two of the refrigerant sensors are calculated, and the refrigerant concentration gradient index between the two refrigerant sensors is calculated based on the distance value and the refrigerant concentration difference. If the refrigerant concentration gradient index is greater than the preset gradient threshold, then the cross-correlation coefficient between the two refrigerant sensors is calculated, and the maximum time delay of the cross-correlation coefficient is calculated. The location of the leak source between the two refrigerant sensors is determined based on the maximum time delay.

9. An air conditioner, comprising a processor and a memory, characterized in that: The memory stores a computer program, which, when executed by the processor, implements the steps of the air conditioning refrigerant leakage early warning and control method as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the controller, it implements the steps of the air conditioning refrigerant leakage early warning control method as described in any one of claims 1 to 8.

Images