Expansion valve control method and device for multi-connected air conditioner and multi-connected air conditioner

Abstract

The application provides a multi-connected air conditioner, an expansion valve control method and device thereof; in the expansion valve control method of the multi-connected air conditioner, learning data is obtained by clustering detection data when the multi-connected air conditioner is stably running, and a prediction model is updated according to the learning data, so that the deviation problem caused by the detection data is avoided, the learning deviation is reduced, and the updated prediction model can be applied to any working condition. For the scenes of multi-connected air conditioner equipment difference, environmental difference, system characteristic change after the multi-connected air conditioner is used for many years and the like, adaptive adjustment of the prediction model can be realized, and the general requirement is met.

Description

Technical Field

[0001] This invention relates to the field of air conditioning technology, and in particular to a method and apparatus for controlling the expansion valve of a multi-split air conditioner, and to a multi-split air conditioner. Background Technology

[0002] For multi-split air conditioners with multiple indoor and outdoor units, practical applications require corresponding control of the outdoor compressor speed and the expansion valve opening of each indoor unit based on load changes. Especially in scenarios with significant load variations, such as changes in the number of operating indoor units, existing control methods employ feedback control to regulate the expansion valve opening and prevent instability during transitional phases. This control can, to some extent, suppress transitional instability caused by rapid load changes, such as variations in the number of operating units. However, the predictive models are generally based on test results under representative operating conditions, limiting their versatility and accuracy.

[0003] Based on this, patent CN112268348A employs an expansion valve opening prediction method, which predicts the corresponding expansion valve opening by mapping environmental parameters before and after changes. However, the data volume increases exponentially with the number of parameters, making it impractical to encompass all real-world application scenarios. Patent CN115077053A uses a deep learning model for expansion valve opening control, but because it learns parameters in real-time under stable conditions, its learning results are affected by the frequency of actual environmental events, resulting in a biased view. Therefore, existing expansion valve opening control schemes are only applicable to representative operating conditions or frequently occurring conditions, lacking versatility and failing to meet practical application needs. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide an expansion valve control method, device and multi-split air conditioner to alleviate the above-mentioned technical problems.

[0005] In a first aspect, embodiments of the present invention provide a method for controlling the expansion valve of a multi-split air conditioner. The method includes: acquiring detection data during stable operation of the multi-split air conditioner; wherein the detection data includes: compressor frequency, outdoor ambient temperature, indoor ambient temperature, and expansion valve opening degree; clustering the detection data based on a learning dataset to obtain learning data; wherein the learning dataset includes at least one learning data cluster, each learning data cluster corresponding to a representative point, and the learning data includes the representative point of each learning data cluster; updating a pre-constructed prediction model based on the learning data, and controlling the opening degree of each expansion valve according to the updated prediction model.

[0006] The expansion valve control method for multi-split air conditioners described above obtains learning data by clustering the detection data during stable operation of the multi-split air conditioner, and updates the prediction model based on the learning data. This not only avoids the deviation problem caused by the detection data and reduces the learning deviation, but also makes the updated prediction model applicable to any operating condition. It can adaptively adjust the prediction model for scenarios such as differences in multi-split air conditioner equipment, environmental differences, and changes in system characteristics after years of use and aging of multi-split air conditioners, thus meeting the requirements for universality.

[0007] Preferably, the above-mentioned clustering of detection data based on the learning dataset includes: calculating the distance between the detection data and the representative points of each learning data cluster; taking the representative points whose distance is less than a preset distance threshold as target representative points, and merging the detection data into the learning data cluster in which the target representative points are located.

[0008] Preferably, the above method further includes: updating the target representative point based on a preset moving distance and detection data to obtain the updated target representative point.

[0009] Preferably, the above method further includes: if the distances are all not less than a preset distance threshold, then a new learning data cluster is created in the learning dataset, and the detection data is incorporated into the new learning data cluster.

[0010] Preferably, the above method further includes: obtaining the total number of learning data clusters in the learning dataset; if the total number is greater than a preset number, deleting the learning data clusters one by one according to the creation time of the learning data clusters in descending order of creation time, until the total number is not greater than the preset number.

[0011] Preferably, the above method further includes: if the total number is greater than a preset number, calculating the reliability of each learning data cluster, and deleting the learning data clusters with a reliability less than a preset reliability threshold.

[0012] Preferably, the above-mentioned control of the opening degree of each expansion valve according to the updated prediction model includes: when the number of operating indoor units in the multi-split air conditioner changes, obtaining a first number of operating indoor units before the change and a second number after the change, as well as detection parameters; wherein, the detection parameters include: compressor frequency, outdoor ambient temperature, and indoor ambient temperature; inputting the first number, the second number, and the detection parameters into the updated prediction model, so that the updated prediction model outputs the optimal opening degree of each expansion valve corresponding to the shortest fluctuation time of the opening degree of each expansion valve during the transition control stage after the change of the number of operating indoor units; using the optimal opening degree as the initial target opening degree of each expansion valve, and controlling the opening degree of each expansion valve based on the initial target opening degree.

[0013] Secondly, embodiments of the present invention also provide an expansion valve control device for a multi-split air conditioner, the device comprising: The data acquisition module is used to acquire detection data during the stable operation of the multi-split air conditioner; the detection data includes: compressor frequency, outdoor ambient temperature, indoor ambient temperature, and expansion valve opening. The data processing module is used to perform clustering processing on the detection data based on the learning dataset to obtain learning data; wherein, the learning dataset includes at least one learning data cluster, each learning data cluster corresponds to a representative point, and the learning data includes the representative point of each learning data cluster; The model update module is used to update the pre-built prediction model based on the learning data, and to control the opening degree of each expansion valve according to the updated prediction model.

[0014] Thirdly, embodiments of the present invention also provide a multi-split air conditioner, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the expansion valve control method of the multi-split air conditioner described in the first aspect.

[0015] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the expansion valve control method for a multi-split air conditioner described in the first aspect.

[0016] The embodiments of the present invention bring the following beneficial effects: This invention provides a method, apparatus, and multi-split air conditioner for controlling the expansion valves of a multi-split air conditioner. The method acquires detection data during stable operation of the multi-split air conditioner. The detection data includes compressor frequency, outdoor ambient temperature, indoor ambient temperature, and expansion valve opening. The detection data is clustered based on a learning dataset to obtain learning data. The learning dataset includes at least one learning data cluster, each cluster corresponding to a representative point. The learning data includes the representative point of each learning data cluster. A pre-built prediction model is updated based on the learning data, and the opening of each expansion valve is controlled according to the updated prediction model. This control method, by clustering the detection data during stable operation of the multi-split air conditioner to obtain learning data and updating the prediction model based on the learning data, not only avoids deviations caused by the detection data and reduces learning bias, but also makes the updated prediction model applicable to any operating condition. It allows for adaptive adjustment of the prediction model for scenarios such as differences in multi-split air conditioner equipment, environmental differences, and changes in system characteristics after years of use, thus meeting the requirement for versatility.

[0017] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained through the structures particularly pointed out in the description and the drawings.

[0018] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0019] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 A flowchart of an expansion valve control method for a multi-split air conditioner provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the clustering processing result of learning data provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the movement and updating of a target representative point provided in an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating the creation of a new learning data cluster according to an embodiment of the present invention; Figure 5 A schematic diagram of an expansion valve control device for a multi-split air conditioner provided in an embodiment of the present invention; Figure 6 This is a structural schematic diagram of a multi-split air conditioner provided in an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. 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.

[0022] To facilitate understanding of this embodiment, the embodiments of the present invention will be described in detail below.

[0023] This invention provides a method for controlling the expansion valve of a multi-split air conditioner. The executing entity is the controller of the multi-split air conditioner or a mobile control terminal communicatively connected to the multi-split air conditioner. For ease of explanation, this invention uses the controller of the multi-split air conditioner as an example to illustrate the control of the opening degree of the expansion valve on each indoor side of the multi-split air conditioner. Figure 1 As shown, the method includes the following steps: Step S102: Obtain detection data during stable operation of the multi-split air conditioner.

[0024] The detection data includes: compressor frequency, outdoor ambient temperature, indoor ambient temperature, and expansion valve opening. In practical applications, detection data is acquired at preset intervals, such as every 30 minutes. The number of data points may be one or more, such as the detection data for each indoor unit. In this case, the detection data includes compressor frequency, outdoor ambient temperature, indoor ambient temperature corresponding to that indoor unit, and expansion valve opening. For example, the outdoor ambient temperature is obtained through a temperature sensor located on the outdoor side, and the indoor ambient temperature is obtained through a temperature sensor located on the indoor side. The specific settings can be configured according to actual conditions.

[0025] Step S104: Cluster the detection data based on the learning dataset to obtain the learning data.

[0026] After acquiring the detection data, the controller clusters the detection data based on the learning dataset to obtain learning data. The learning dataset includes at least one learning data cluster, each cluster corresponding to a representative point, and the learning data includes the representative point of each cluster. In practical applications, the detection data can also be referred to as learning data. Multiple similar learning data are grouped into a single learning data cluster, and the entire cluster is replaced by a representative point for predictive model updates. To distinguish between the learning data within a learning data cluster and the most recently acquired learning data, this embodiment of the invention refers to the learning data acquired during the stable operation of the multi-split air conditioner as the detection data.

[0027] In practical applications, as the stable operating time of multi-split air conditioners increases, a large amount of similar learning data will inevitably be collected. The prediction model obtained by machine learning based on similar learning data will have reduced prediction accuracy for other learning data that deviates from the similar learning data. If similar learning data is removed from the dataset through data filtering, it will be unable to cope with predictions in scenarios of slow changes caused by years of use and aging.

[0028] Taking seven learning data as an example, such as Figure 2 As shown, in (a), seven learning data points (1-7) are directly used to update the prediction model. Since learning data points 3-7 are quite similar, the updated prediction model is biased towards the lower left learning data points 3-7. In (b), the similar learning data points 3-7 are clustered, and representative points are used to replace them. The prediction model is then updated using learning data points 1, 2, and the representative points. Therefore, clustering the detection data using the learning dataset avoids the bias caused by the detection data, thus improving the prediction accuracy of the prediction model.

[0029] Step S106: Update the pre-built prediction model based on the learning data, and control the opening degree of each expansion valve according to the updated prediction model.

[0030] The expansion valve control method for multi-split air conditioners provided in this invention clusters the detection data from the stable operation of the multi-split air conditioner to obtain learning data, and updates the prediction model based on the learning data. This not only avoids the deviation problem caused by the detection data and reduces the learning deviation, but also makes the updated prediction model applicable to any operating condition. It can adaptively adjust the prediction model for scenarios such as differences in multi-split air conditioner equipment, environmental differences, and changes in system characteristics after years of use and aging of multi-split air conditioners, thus meeting the requirements of universality.

[0031] In one implementation, clustering of the detection data based on the learning dataset includes: calculating the distance between the detection data and the representative points of each learning data cluster; taking the representative points whose distance is less than a preset distance threshold as target representative points, and merging the detection data into the learning data cluster in which the target representative points are located.

[0032] Specifically, for newly acquired detection data, the controller calculates the distance D between the detection data and the representative point R of each learning data cluster. Here is an example of the formula for calculating D: (1) Among them, P i Represents the detection data, R i The representative point is represented by n, which represents the dimension of the detection data. Since the detection data includes compressor frequency, outdoor ambient temperature, indoor ambient temperature, and expansion valve opening, n=4 here. For example... i =1 indicates the compressor frequency. i =2 represents the outdoor ambient temperature. i =3 indicates the indoor ambient temperature. i =4 indicates the opening degree of the expansion valve.

[0033] Therefore, by calculating the distance D between the detection data and the representative point of each learning data cluster using the above formula (1), and taking the representative point whose distance D is less than the preset distance threshold (such as 0.1) as the target representative point, it is determined that the detection data and the target representative point are similar data. In this case, the detection data is merged into the learning data cluster in which the target representative point is located, thereby realizing the clustering of the detection data and avoiding the bias problem caused by the detection data. It should be noted that the value of the preset distance threshold can be set according to the actual situation.

[0034] In one implementation, the method further includes: updating the target representative point based on a preset moving distance and detection data to obtain the updated target representative point.

[0035] After the detection data is incorporated into the learning data cluster where the target representative point is located, the data in the learning data cluster changes. At this time, the original target representative point may not be able to replace all the data in the learning data cluster, or it may replace all the data in the learning data cluster, but some data are far away from the target representative point. In order to ensure the accuracy of the target representative point, the target representative point in the learning data cluster needs to be updated.

[0036] Specifically, the target representative point is moved in the direction of the newly incorporated detection data, and updated based on a preset moving distance and detection data to obtain the updated target representative point. The update formula is as follows: (2) in, R represents the updated target point. i This represents the target point before the update. P represents the preset travel distance. i Let represent the detection data, and n represent the dimension of the detection data. In practical applications, the preset moving distance is preferably 0.1, but it can be adjusted according to the actual situation.

[0037] For example, such as Figure 3 As shown, the target representative point in the learning data cluster before the update is located at point A. After incorporating the detection data P into the learning data cluster, the target representative point is moved towards the detection data P to obtain the updated target representative point, which is now located at point B. By updating the target representative point, the scenario of the learning data cluster changing over time is realized, ensuring that the accuracy of the learning data is adapted to the operating conditions of the multi-split air conditioner. This allows the updated prediction model to be applicable to any operating condition. For scenarios such as differences in multi-split air conditioner equipment, environmental differences, and changes in system characteristics after years of use and aging of multi-split air conditioners, the prediction model can be adaptively adjusted, meeting the requirements for versatility.

[0038] In one implementation, the method further includes: if the distances are all not less than a preset distance threshold, then creating a new learning data cluster in the learning dataset and incorporating the detection data into the new learning data cluster.

[0039] After the controller calculates the distance between the detection data and the representative point of each learning data cluster, if each distance is greater than or equal to the preset distance threshold, the detection data cannot be incorporated into all existing learning data clusters in the learning dataset. Therefore, the controller creates a new learning data cluster in the learning dataset and incorporates the detection data into the new learning data cluster, thereby further ensuring the accuracy of the learning data used for updating the prediction model and thus improving the control accuracy of the expansion valve.

[0040] For scenarios where there are multiple learning data clusters in the learning dataset, such as Figure 4 As shown, there are three learning data clusters, with representative points R1, R2 and R3 respectively. For new data (i.e. newly acquired detection data), if the distance between it and representative points R1, R2 and R3 is not less than the preset distance threshold, a new learning data cluster is created and the new data is merged into the new learning data cluster. The representative point of the new learning data cluster is representative point R4.

[0041] It should be noted that since the newly added learning data cluster contains only new data, the new data can be used as the representative point R4 at this time. When new data is added to the new learning data cluster in the subsequent process, the representative point R4 will be updated according to the newly added data.

[0042] In one implementation, the method further includes: obtaining the total number of learning data clusters in the learning dataset; if the total number is greater than a preset number, deleting learning data clusters one by one according to their creation time in descending order of creation time, until the total number is not greater than the preset number.

[0043] To prevent outdated or interfering data from negatively impacting the accuracy of the prediction model, the learning data clusters in the learning dataset need to be updated. Specifically, the controller obtains the total number of learning data clusters in the learning dataset. When the total number exceeds a preset number, it deletes clusters sequentially based on their creation time, from oldest to youngest, until the total number is no greater than the preset number. This discards older learning data clusters, further ensuring the accuracy of the learning data used for updating the prediction model, thus guaranteeing the accuracy of the prediction model and improving the opening of the expansion valve. The preset number is preferably 50, but can be set according to actual conditions.

[0044] In one embodiment, the method further includes: if the total number is greater than a preset number, calculating the reliability of each learning data cluster, and deleting learning data clusters with a reliability less than a preset reliability threshold.

[0045] In addition to updating learning data clusters based on their creation time, updates can also be made based on their reliability. Specifically, when the total number exceeds a preset limit (e.g., 50), the reliability of each learning data cluster is calculated, and clusters with reliability below a preset threshold are deleted. The formula for calculating reliability is as follows: (3) in, Cr Indicates reliability. This indicates the current distance from the last update of the representative point, in days. This represents the forgetting time constant, which can be preset in the controller, such as 20. N represents the number of times the point is updated, in units of times.

[0046] Therefore, when the total number of learning data clusters is greater than the preset number, such as 50, the reliability of each learning data cluster is calculated according to the above formula (3). Cr and reliability Cr Learning data clusters with reliability values ​​below a preset reliability threshold are deleted, i.e., learning data clusters with low reliability in the learning dataset are discarded. This further ensures the accuracy of the learning data used for updating the prediction model, thereby ensuring the accuracy of the prediction model and thus improving the opening degree of the expansion valve.

[0047] It should be noted that updates to the learning data clusters in the learning dataset can be performed at preset intervals, when the total number exceeds a preset limit, or when the number of learning data clusters in the learning dataset increases. The specific settings can be configured according to the actual situation. In particular, besides updating learning data clusters based on creation time or reliability as mentioned above, in some scenarios, updates can also be performed based on creation time and reliability. For example, weights can be set for creation time and reliability for each learning data cluster, and then calculations can be performed based on creation time, reliability, and their corresponding weights. Learning data clusters whose calculation results are less than a preset value can be deleted.

[0048] In one implementation, controlling the opening degree of each expansion valve according to the updated prediction model includes: when the number of operating indoor units in a multi-split air conditioner changes, acquiring a first number of operating indoor units before the change and a second number after the change, as well as detection parameters; wherein, the detection parameters include: compressor frequency, outdoor ambient temperature, and indoor ambient temperature; inputting the first number, the second number, and the detection parameters into the updated prediction model, so that the updated prediction model outputs the optimal opening degree of each expansion valve corresponding to the shortest fluctuation duration of the opening degree of each expansion valve during the transition control phase after the change in the number of operating indoor units; using the optimal opening degree as the initial target opening degree of each expansion valve, and controlling the opening degree of each expansion valve based on the initial target opening degree.

[0049] Specifically, when the load of the multi-split air conditioner changes drastically (i.e., the number of indoor units in operation changes), the first quantity, the second quantity, and the detection parameters are acquired and input into the updated prediction model. This allows the updated prediction model to output the optimal opening of each expansion valve during the transition control phase after the change in the number of indoor units in operation, corresponding to the shortest fluctuation duration of the opening of each expansion valve. The optimal opening is used as the initial target opening of each expansion valve, and the opening of each expansion valve is controlled based on the initial target opening. After the opening is adjusted, the superheat PID (Proportional Integral Derivative) control is entered. The specific superheat PID control can be referred to in the prior art, and will not be described in detail in this embodiment of the invention.

[0050] Therefore, by predicting the optimal opening of the expansion valve using the updated prediction model, the time of the unstable transition control phase of the expansion valve opening can be minimized when the load fluctuates greatly. At the same time, only the initial target opening of the transition phase needs to be predicted, which can significantly reduce the computational load caused by frequent prediction of the expansion valve opening, thereby reducing the computational load, improving the stability of air conditioner operation, and improving the reliability of expansion valve opening control.

[0051] It should be noted that when the first quantity and the second quantity are different, the updated predictive model is used to predictively control the opening of each expansion valve. Alternatively, in some scenarios, the quantity change value is determined based on the first and second quantities. When the quantity change value is greater than or equal to the preset difference, the updated predictive model is used to predictively control the opening of each expansion valve. When the quantity change value is less than the preset difference, the number of indoor units in operation changes less, and the load fluctuation is smaller. In this case, traditional feedback control methods such as PID control can suppress the load fluctuation. Thus, different expansion valve control schemes are used for different scenarios, improving the comfort and energy efficiency of multi-split air conditioners.

[0052] To facilitate understanding, the expansion valve control process of a multi-split air conditioner is illustrated here with an example. Specifically, it includes the following steps: (1) Prediction model construction.

[0053] Based on operational data acquired in the laboratory beforehand, machine learning is used to build a predictive model. Each data point includes: the number of indoor units before and after the change in operating quantity, as well as detection parameters: compressor frequency, outdoor ambient temperature, and indoor ambient temperature. Since the predictive model is built based on laboratory data, although the initial learning data after actual installation may be insufficient, it still possesses a certain level of accuracy. Therefore, the laboratory learning data and the predictive model are stored in the controller's microcontroller memory and EEPROM (Electrically Erasable Programmable Read-Only Memory) for subsequent updates to the predictive model.

[0054] (2) Prediction of expansion valve opening.

[0055] When the number of indoor units in a multi-split air conditioner changes, the number of indoor units before and after the change, as well as the detection parameters, are obtained. The optimal opening degree of each expansion valve is then predicted based on the prediction model. The specific prediction process can be referred to the aforementioned embodiments, and the embodiments of the present invention will not be described in detail here.

[0056] (3) Expansion valve opening control.

[0057] After the above prediction model outputs the optimal opening degree of each expansion valve, the optimal opening degree is used as the initial target opening degree of each expansion valve, and the opening degree of each expansion valve is controlled based on the initial target opening degree; and after adjusting and fixing the opening degree of the expansion valve for a certain period of time (such as 80s), it enters the superheat PID control. The specific control process can be referred to the aforementioned embodiment, and the embodiment of the present invention will not be described in detail here.

[0058] (4) Data acquisition.

[0059] After the expansion valve opening is adjusted and the multi-split air conditioner is running stably, the detection data of the multi-split air conditioner during stable operation is obtained. The detection data includes: compressor frequency, outdoor ambient temperature, indoor ambient temperature and expansion valve opening; for example, the compressor frequency is 55Hz, the outdoor ambient temperature is 7℃, the indoor ambient temperature is 21℃ and the expansion valve opening is 110pls.

[0060] The detection data is clustered based on the learning dataset to obtain the learning data. The learning dataset consists of multiple learning data clusters, each corresponding to a representative point. Taking 50 learning data clusters as an example, the representative points and reliability of the 50 learning data clusters are shown in Table 1 below: Table 1

[0061] In addition, the detection data was added to the learning dataset as new data. The representative points of the 50 learning data clusters and the new data were standardized, and the distance between the new data and the representative points of the 50 learning data clusters was calculated, as shown in Table 2 below: Table 2

[0062] Assuming a preset distance threshold of 0.1, only when the distance between the new data and the representative point of learning data cluster 1 is less than 0.1, will the new data (i.e., the detection data) be incorporated into learning data cluster 1, and the representative point of learning data cluster 1 be updated using formula (2). Specifically, assuming a preset moving distance of 0.1, the update process is as follows: Compressor frequency: .

[0063] Outdoor ambient temperature: .

[0064] Indoor ambient temperature: .

[0065] Expansion valve opening: .

[0066] In addition, the reliability of learning data cluster 1 is updated. Assuming that the current time since the latest update of the representative point is 0 days, the number of updates of the representative point is 10, and the forgetting time constant is 20, the reliability of learning data cluster 1 after the update can be calculated as 1.04 according to formula (4). Since the new data is incorporated into learning data cluster 1 this time, no new learning data cluster is added. If new learning data clusters are added later, when the total number of learning data clusters reaches more than the preset number of 50, the learning data clusters with reliability lower than the preset reliability threshold will be discarded. In some scenarios, the learning data cluster with the longest creation time can also be deleted according to the creation time of the learning data cluster.

[0067] (5) Prediction model update.

[0068] For the updated learning data clusters, representative points from each cluster are combined to form learning data. The pre-built prediction model is then updated based on this learning data to control the opening of each expansion valve according to the updated prediction model. The prediction model, using NN (Neural Networks) models or regression analysis models, can achieve normal operation without affecting the actual control time due to its computational complexity. Furthermore, this approach is easily applicable to existing multi-split air conditioners.

[0069] The expansion valve control method for multi-split air conditioners provided in this invention clusters the detection data from the stable operation of the multi-split air conditioner to obtain learning data, and updates the prediction model based on the learning data. This not only avoids the deviation problem caused by the detection data and reduces the learning deviation, but also makes the updated prediction model applicable to any operating condition. It can adaptively adjust the prediction model for scenarios such as differences in multi-split air conditioner equipment, environmental differences, and changes in system characteristics after years of use and aging of multi-split air conditioners, thus meeting the requirements of universality.

[0070] Corresponding to the above-described expansion valve control method for multi-split air conditioners, this embodiment of the invention also provides an expansion valve control device for multi-split air conditioners, such as... Figure 5 As shown, the device includes: a data acquisition module 51, a data processing module 52, and a model update module 53; the functions of each module are as follows: The data acquisition module 51 is used to acquire detection data during the stable operation of the multi-split air conditioner; the detection data includes: compressor frequency, outdoor ambient temperature, indoor ambient temperature and expansion valve opening. The data processing module 52 is used to perform clustering processing on the detection data based on the learning dataset to obtain learning data; wherein, the learning dataset includes at least one learning data cluster, each learning data cluster corresponds to a representative point, and the learning data includes the representative point of each learning data cluster; The model update module 53 is used to update the pre-built prediction model based on the learning data, and to control the opening degree of each expansion valve according to the updated prediction model.

[0071] The expansion valve control device for multi-split air conditioners provided in this invention clusters the detection data from the stable operation of the multi-split air conditioner to obtain learning data, and updates the prediction model based on the learning data. This not only avoids the deviation problem caused by the detection data and reduces the learning deviation, but also makes the updated prediction model applicable to any operating condition. It can adaptively adjust the prediction model for scenarios such as differences in multi-split air conditioner equipment, environmental differences, and changes in system characteristics after years of use and aging of multi-split air conditioners, thus meeting the requirements for universality.

[0072] Preferably, the data processing module 52 is further configured to: calculate the distance between the detection data and the representative point of each learning data cluster; take the representative point whose distance is less than a preset distance threshold as the target representative point, and incorporate the detection data into the learning data cluster in which the target representative point is located.

[0073] Preferably, the above-mentioned device further includes: updating the target representative point based on a preset moving distance and detection data to obtain the updated target representative point.

[0074] Preferably, the above-mentioned device further includes: if the distances are not less than a preset distance threshold, then creating a new learning data cluster in the learning dataset and incorporating the detection data into the new learning data cluster.

[0075] Preferably, the above-mentioned device further includes: obtaining the total number of learning data clusters in the learning dataset; if the total number is greater than a preset number, deleting the learning data clusters one by one according to the creation time of the learning data clusters in descending order of creation time, until the total number is not greater than the preset number.

[0076] Preferably, the above-mentioned device further includes: if the total number is greater than a preset number, calculating the reliability of each learning data cluster, and deleting the learning data clusters whose reliability is less than a preset reliability threshold.

[0077] Preferably, the above-mentioned control of the opening degree of each expansion valve according to the updated prediction model includes: when the number of operating indoor units in the multi-split air conditioner changes, obtaining a first number of operating indoor units before the change and a second number after the change, as well as detection parameters; wherein, the detection parameters include: compressor frequency, outdoor ambient temperature, and indoor ambient temperature; inputting the first number, the second number, and the detection parameters into the updated prediction model, so that the updated prediction model outputs the optimal opening degree of each expansion valve corresponding to the shortest fluctuation time of the opening degree of each expansion valve during the transition control stage after the change of the number of operating indoor units; using the optimal opening degree as the initial target opening degree of each expansion valve, and controlling the opening degree of each expansion valve based on the initial target opening degree.

[0078] The expansion valve control device for multi-split air conditioners provided in this embodiment of the invention has the same technical features as the expansion valve control method for multi-split air conditioners provided in the above embodiments, so it can also solve the same technical problems and achieve the same technical effects.

[0079] This invention also provides a multi-split air conditioner, including a processor and a memory. The memory stores a computer program that can be executed by the processor, and the processor executes the computer program to implement the expansion valve control method of the multi-split air conditioner described above.

[0080] See Figure 6 As shown, the multi-split air conditioner includes a processor 100 and a memory 101. The memory 101 stores a computer program that can be executed by the processor 100. The processor 100 executes the computer program to implement the expansion valve control method of the multi-split air conditioner described above.

[0081] Furthermore, Figure 6 The multi-split air conditioner shown also includes a bus 102 and a communication interface 103. The processor 100, the communication interface 103 and the memory 101 are connected through the bus 102.

[0082] The memory 101 may include high-speed random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 103 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc. The bus 102 may be an ISA (Industrial Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Enhanced Industry Standard Architecture) bus, etc. These buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 6 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0083] Processor 100 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 100 or by instructions in software form. Processor 100 may be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this invention. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this invention can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software module can reside in a readily available storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory 101, and the processor 100 reads the information from memory 101 and, in conjunction with its hardware, completes the steps of the method described in the foregoing embodiments.

[0084] This embodiment also provides a computer-readable storage medium storing a computer program. When the computer program is called and executed by a processor, the computer program causes the processor to implement the above-described expansion valve control method for a multi-split air conditioner.

[0085] The expansion valve control method, apparatus, and computer program product for multi-split air conditioners provided in this invention include a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the preceding method embodiments. For specific implementation details, please refer to the method embodiments, which will not be repeated here.

[0086] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system and apparatus described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0087] Furthermore, in the description of the embodiments of the present invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in the present invention based on the specific circumstances.

[0088] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0089] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0090] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for controlling the expansion valve of a multi-split air conditioner, characterized in that, The method includes: Acquire detection data during stable operation of a multi-split air conditioner; wherein, the detection data includes: compressor frequency, outdoor ambient temperature, indoor ambient temperature, and expansion valve opening; The detection data is clustered based on the learning dataset to obtain learning data; wherein, the learning dataset includes at least one learning data cluster, each learning data cluster corresponds to a representative point, and the learning data includes the representative point of each learning data cluster; The pre-built prediction model is updated based on the learning data, and the opening degree of each expansion valve is controlled according to the updated prediction model.

2. The method according to claim 1, characterized in that, The clustering process of the detection data based on the learning dataset includes: Calculate the distance between the detected data and the representative point of each of the learned data clusters; Representative points whose distance is less than a preset distance threshold are taken as target representative points, and the detection data is incorporated into the learning data cluster in which the target representative points are located.

3. The method according to claim 2, characterized in that, The method further includes: Based on the preset moving distance and the detection data, the target representative point is updated to obtain the updated target representative point.

4. The method according to claim 2, characterized in that, The method further includes: If all distances are not less than the preset distance threshold, a new learning data cluster is created in the learning dataset, and the detection data is incorporated into the new learning data cluster.

5. The method according to claim 1, characterized in that, The method further includes: Obtain the total number of learning data clusters in the learning dataset. If the total number is greater than a preset number, delete the learning data clusters one by one according to their creation time in descending order of creation time, until the total number is not greater than the preset number.

6. The method according to claim 5, characterized in that, The method further includes: If the total number is greater than a preset number, the reliability of each learning data cluster is calculated, and the learning data clusters with a reliability less than a preset reliability threshold are deleted.

7. The method according to claim 1, characterized in that, The step of controlling the opening degree of each expansion valve according to the updated prediction model includes: When the number of indoor units in the multi-split air conditioner changes, the system acquires a first number of indoor units before the change and a second number after the change, as well as detection parameters; wherein, the detection parameters include: compressor frequency, outdoor ambient temperature, and indoor ambient temperature; The first quantity, the second quantity, and the detection parameters are input into the updated prediction model so that the updated prediction model outputs the optimal opening of each expansion valve corresponding to the shortest fluctuation time of each expansion valve opening during the transition control phase after the change in the number of indoor units in operation. The optimal opening degree is used as the initial target opening degree for each of the expansion valves, and the opening degree of each of the expansion valves is controlled based on the initial target opening degree.

8. An expansion valve control device for a multi-split air conditioner, characterized in that, The device includes: The data acquisition module is used to acquire detection data during the stable operation of the multi-split air conditioner; wherein, the detection data includes: compressor frequency, outdoor ambient temperature, indoor ambient temperature, and expansion valve opening. The data processing module is used to perform clustering processing on the detection data based on the learning dataset to obtain learning data; wherein, the learning dataset includes at least one learning data cluster, each learning data cluster corresponds to a representative point, and the learning data includes the representative point of each learning data cluster; The model update module is used to update the pre-built prediction model based on the learning data, and to control the opening degree of each expansion valve according to the updated prediction model.

9. A multi-split air conditioner, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the expansion valve control method for a multi-split air conditioner according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, performs the steps of the expansion valve control method for a multi-split air conditioner according to any one of claims 1-7.

Images