Multi-split air conditioning system and compressor operation control method

Abstract

The embodiment of the invention belongs to the technical field of household appliances, and provides a multi-split air conditioning system and a compressor operation control method. At least two outdoor units; the compressor is arranged in the outdoor unit; the controller is configured to obtain operation data and environment parameters of the multi-split air conditioning system in the operation process of the multi-split air conditioning system; processing the operation data and the environmental parameters through a pre-constructed load prediction model to obtain the total load of the system in the next control period; and according to the total load, the performance parameters of the multiple compressors and the capacity of the multiple outdoor units, a target starting-up combination of the compressors is determined, and the compressors in the target starting-up combination are controlled to operate in the next control period. In this way, the accuracy of compressor control can be improved, and the operation efficiency of the compressor is improved.

Description

Technical Field

[0001] This application relates to the field of home appliance technology. More specifically, it relates to a multi-split air conditioning system and a compressor operation control method. Background Technology

[0002] Multi-split air conditioning systems are widely used in commercial buildings (such as office buildings, shopping malls, and hotels), high-end residences, and large industrial plants. This system combines multiple outdoor units with multiple indoor units to provide independent temperature control for different areas.

[0003] Currently, controlling the compressor in a multi-split air conditioning system involves determining the compressor's start / stop status by comparing the compressor's operating frequency with preset thresholds (such as starting a new compressor at the upper frequency limit and shutting down the compressor at the lower frequency limit) based on the system's current load demand.

[0004] However, controlling the compressor's start and stop based on its frequency may result in situations where the compressor's start and stop do not meet the actual needs of the current system, leading to poor accuracy in compressor start and stop control and consequently, poor compressor operating efficiency. Summary of the Invention

[0005] This application provides a multi-split air conditioning system and a compressor operation control method to improve the accuracy of compressor control and improve the operating efficiency of the compressor.

[0006] In a first aspect, embodiments of this application provide a multi-split air conditioning system, including:

[0007] At least one indoor unit;

[0008] At least two outdoor units;

[0009] The compressor is located inside the outdoor unit;

[0010] The controller is configured as follows:

[0011] During the operation of the multi-split air conditioning system, the operating data and environmental parameters of the multi-split air conditioning system are acquired;

[0012] By processing the operating data and environmental parameters using a pre-built load prediction model, the total load of the system in the next control cycle is obtained.

[0013] Based on the total load, the performance parameters of multiple compressors, and the capacity of multiple outdoor units, a target start-up combination of compressors is determined, and the compressors in the target start-up combination are controlled to operate in the next control cycle.

[0014] In this application, during system operation, the total system load for the next control cycle is predicted based on system operating data and environmental parameters. Then, based on the total load, the performance parameters of multiple compressors, and the capacity of multiple outdoor units, the starting combination of compressors meeting the load requirements is determined. This allows the compressors within that starting combination to be controlled in the next controller cycle. This ensures that the determined starting combination meets the actual needs of the system in the next control cycle, while also taking into account compressor performance, resulting in higher accuracy and effectively improving compressor operating efficiency.

[0015] In some embodiments of this application, the controller is configured as follows:

[0016] Based on the total load and the capacity of the multiple outdoor units, a plurality of first start-up combinations and the sub-load of the outdoor unit corresponding to each compressor in the first start-up combination are determined; the sum of the capacities of at least one outdoor unit included in the plurality of first start-up combinations is greater than or equal to the total load;

[0017] Based on the sub-loads corresponding to each outdoor unit and the performance parameters of multiple compressors, a target start-up combination is determined from the multiple first start-up combinations.

[0018] In this application, by allocating the total load to each outdoor unit within the first start-up combination, the sub-load of each outdoor unit is obtained. Then, based on the sub-load of the outdoor unit and the performance parameters of the compressor, the target start-up combination is determined, which makes the determination of the target start-up combination more accurate and can further improve the operating efficiency of the compressor.

[0019] In some embodiments of this application, the controller is configured as follows:

[0020] Based on the sub-load corresponding to the outdoor unit in each first start-up combination, determine the operating frequency of the compressor corresponding to the outdoor unit;

[0021] Based on the compressor's operating frequency and performance parameters, the compressor's energy efficiency is determined within its high-efficiency range.

[0022] Determine the total energy efficiency of each first start-up combination based on the energy efficiency of each compressor;

[0023] Among the plurality of first power-on combinations, at least one second power-on combination is determined based on the total energy efficiency of the first power-on combination; the at least one second power-on combination is a power-on combination whose total energy efficiency differs from that of the third power-on combination by a preset difference, and the third power-on combination is the first power-on combination with the highest total energy efficiency;

[0024] Among the at least one second power-on combination, a target power-on combination that meets preset conditions is determined.

[0025] In this application, by determining the maximum achievable energy efficiency of each compressor, the total energy efficiency of each first start-up combination is calculated. From this total energy efficiency, the start-up combination with the highest energy efficiency is selected as the target start-up combination. This ensures that the compressors in the determined target start-up combination all have high energy efficiency, effectively improving the system's energy efficiency ratio.

[0026] In some embodiments of this application, the performance parameters of the compressor are the correspondence between the compressor's energy efficiency and operating frequency; the performance parameters of the compressor are obtained by pre-correcting the initial performance parameters based on the compressor's operating parameters.

[0027] In this application, the performance parameters of the compressor are modified according to the actual operating conditions of the compressor, so that the performance parameters of the compressor are more in line with the actual situation, thereby improving the accuracy of the subsequently determined target start-up combination, and making the compressor in the target start-up combination run with higher energy efficiency.

[0028] In some embodiments of this application, the controller is configured as follows:

[0029] If the number of second power-on combinations is one, then the second power-on combination is determined to be the target power-on combination.

[0030] In this application, when there is only one second start-up combination, it may be because the difference between the other first start-up combinations and the first start-up combination with the highest total energy efficiency is greater than the preset difference. In other words, the total energy efficiency of the other first start-up combinations is relatively small. Therefore, the unique second start-up combination is determined as the target start-up combination, so that the system has higher energy efficiency when the compressor in the target start-up combination is running.

[0031] In some embodiments of this application, the controller is configured as follows:

[0032] If the number of second start-up combinations is at least two, then determine the total cumulative runtime of the compressor in each second start-up combination;

[0033] The second startup combination with the smallest total cumulative runtime is determined as the target startup combination.

[0034] In this application, given the same energy efficiency ratio, the combination with shorter operating time is preferentially selected by comparing the cumulative operating time of the compressors. Optimizing the cumulative operating time extends equipment life and reduces mechanical wear. For example, given the same energy efficiency ratio, compressors with shorter operating times are preferentially selected to avoid damage to the equipment caused by frequent start-stop cycles.

[0035] In some embodiments of this application, the controller is further configured to:

[0036] Determine whether all compressors in the target start-up combination meet their respective start-up and stop time thresholds;

[0037] If at least one compressor does not meet the corresponding start-stop time threshold, then among the other start-up combinations, a new target start-up combination that meets the preset conditions is determined; the other start-up combinations are the start-up combinations other than the target start-up combination among the plurality of first start-up combinations.

[0038] In this application, the need to re-determine a new target start-up combination is determined based on the start-up and stop time thresholds of each compressor, so that the compressors in the final target start-up combination all meet their respective start-up and stop time threshold requirements, thereby avoiding frequent compressor start-up and stop, reducing mechanical wear and improving system stability.

[0039] In some embodiments of this application, the load prediction model is a hierarchical bidirectional long short-term memory network model;

[0040] The controller is also configured to:

[0041] Historical operating data and the environmental parameters are input into a hierarchical bidirectional long short-term memory network model to generate short-term load forecast results and long-term load forecast results.

[0042] The total load of the system in the next control cycle is generated by weighted fusion of the short-term load forecast results and the long-term load forecast results.

[0043] In this application, a hierarchical bidirectional long short-term memory network model is used to predict the total load in the next control cycle, which makes the predicted total load more accurate, thereby improving the accuracy of the subsequently determined target start-up combination.

[0044] In some embodiments of this application, the controller is further configured to:

[0045] The current load forecasting model is optimized based on historical data to obtain a new load forecasting model, which is then used for load forecasting. The historical data includes the historical operating data, environmental parameters, and compressor start-up combinations of the multi-split air conditioning system during the model update cycle.

[0046] In this application, the current prediction model is iteratively optimized using historical operating data to generate a new model for subsequent predictions. For example, during the model update cycle, the system adjusts the parameters of the Bi-LSTM model based on the latest operating data to adapt to performance drift caused by equipment aging or environmental changes.

[0047] Secondly, this application provides a compressor operation control method applied to a multi-split air conditioning system, the multi-split air conditioning system comprising:

[0048] At least one indoor unit;

[0049] At least two outdoor units;

[0050] The compressor is located inside the outdoor unit;

[0051] Controller;

[0052] The method includes:

[0053] During the operation of the multi-split air conditioning system, the operating data and environmental parameters of the multi-split air conditioning system are acquired;

[0054] By processing the operating data and environmental parameters using a pre-built load prediction model, the total load of the system in the next control cycle is obtained.

[0055] Based on the total load, the performance parameters of multiple compressors, and the capacity of multiple outdoor units, a target start-up combination of compressors is determined, and the compressors in the target start-up combination are controlled to operate in the next control cycle.

[0056] Thirdly, this application provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a computer, are used to implement the method described in the second aspect.

[0057] The computer-readable storage medium provided in this application embodiment can execute the technical solutions in the above method embodiments, and its beneficial effects are similar, so they will not be described again here.

[0058] Fourthly, this application provides a computer program product, including a computer program that, when executed by a computer, is used to implement the method described in the second aspect.

[0059] The computer program product provided in this application embodiment can execute the technical solutions in the above method embodiments, and its beneficial effects are similar, so they will not be described again here. Attached Figure Description

[0060] To more clearly illustrate the implementation methods in the embodiments of this application or related technologies, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings.

[0061] Figure 1 This is a schematic diagram of a multi-split air conditioning system according to some embodiments;

[0062] Figure 2This is a flowchart illustrating a compressor operation control method according to some embodiments;

[0063] Figure 3 This is a flowchart illustrating a method for determining a target power-on combination according to some embodiments;

[0064] Figure 4 This is a schematic diagram of the performance curves of a compressor according to some embodiments;

[0065] Figure 5 This is a flowchart illustrating a method for creating a load forecasting model according to some embodiments;

[0066] Figure 6 This is a schematic diagram of a Bayesian optimization process according to some embodiments;

[0067] Figure 7 This is a schematic diagram illustrating the entire training and prediction process of a Bayesian-optimized bidirectional long short-term memory network prediction model according to some embodiments.

[0068] Figure 8 This is a schematic diagram comparing a predicted power with the actual power according to some embodiments. Detailed Implementation

[0069] To make the objectives, implementation methods and advantages of this application clearer, the exemplary implementation methods of this application will be clearly and completely described below with reference to the accompanying drawings of the exemplary embodiments of this application. Obviously, the described exemplary embodiments are only some embodiments of this application, and not all embodiments.

[0070] It should be noted that the brief descriptions of terms in this application are only for the convenience of understanding the embodiments described below, and are not intended to limit the embodiments of this application. Unless otherwise stated, these terms should be understood in their ordinary and common meaning.

[0071] Furthermore, the terms “comprising” and “having”, and any variations thereof, are intended to cover but not exclusively include, for example, a product or device that includes a series of components is not necessarily limited to those that are explicitly listed, but may include other components that are not explicitly listed or that are inherent to such product or device.

[0072] Multi-split air conditioning systems are widely used in commercial buildings (such as office buildings, shopping malls, and hotels), high-end residences, and large industrial plants, and are favored for their flexible zoning control and ability to adapt to complex spatial layouts. In actual operation, this system provides independent temperature control for different areas through a combination of multiple outdoor units and multiple indoor units.

[0073] Traditional control methods often rely on fixed frequency thresholds to regulate compressor start-up and shutdown. Specifically, the system monitors the current compressor's operating frequency, triggering the start of a new compressor when the frequency reaches a preset upper limit, and triggering the compressor's shutdown when the frequency drops to a lower limit.

[0074] However, static control methods cannot respond to changes in system load in real time. There may be situations where the compressor start-up and shutdown do not meet the actual needs of the current system, resulting in poor accuracy of compressor start-up and shutdown control. This can easily lead to frequent compressor start-up and shutdown or operation in inefficient ranges (such as multiple compressors operating at low frequency under high load, or a single compressor operating at overfrequency under low load), resulting in low compressor operating efficiency.

[0075] Based on this, this application provides a multi-split air conditioning system and a compressor operation control method. During system operation, the total load of the system in the next control cycle is predicted based on the system's operating data and environmental parameters. Then, based on the total load, the performance parameters of multiple compressors, and the capacity of multiple outdoor units, the operating combination of compressors meeting the load requirements is determined. The compressors within this operating combination are then controlled to operate within the next controller. This ensures that the determined operating combination meets the actual needs of the system in the next control cycle and takes into account compressor performance, resulting in higher accuracy and effectively improving compressor operating efficiency.

[0076] The technical solutions of this application will be described in detail below with reference to specific embodiments. The following specific embodiments can be combined with each other or exist independently. The same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will be described below with reference to the accompanying drawings.

[0077] First, the structure of the multi-split air conditioning system provided in some embodiments of this application will be described.

[0078] In one possible implementation, Figure 1 This is a schematic diagram of a multi-split air conditioning system according to some embodiments, such as Figure 1 As shown, the multi-split air conditioning system 10 includes at least one indoor unit 101, at least two outdoor units 102, and a controller 103.

[0079] The outdoor unit 102 is equipped with a compressor 1021, therefore, the multi-split air conditioning system 10 includes at least two compressors.

[0080] In this embodiment, the controller 103 can be a micro controller unit (MCU) or other types of controllers. This embodiment does not specifically limit the controller.

[0081] During the operation of the multi-split air conditioning system 10, it is necessary to control the start and stop of the compressor in order to achieve the operation of different compressor combinations.

[0082] In one possible implementation, the multi-split air conditioning system 10 also includes a refrigerant circulation loop ( Figure 1 (Not shown in the diagram) The refrigerant circulation loop allows the refrigerant to circulate in the compressor, condenser, expansion valve, and evaporator; the indoor unit is connected to the outdoor unit via refrigerant connection pipes.

[0083] The controller 103 is configured to: predict the total load of the system in the next control cycle by using a pre-built load prediction model based on the operating data and environmental parameters of the multi-split air conditioning system. Then, based on this total load, the performance parameters of multiple compressors, and the capacity of multiple outdoor units, a preferred compressor start-up combination is determined, and the compressors in this start-up combination are controlled to operate in the next control cycle.

[0084] Below, in conjunction with the above Figure 1 The structure of the multi-split air conditioning system is shown, and the compressor operation control method of the multi-split air conditioning system is described. This method is applied to the controller 103 in the above embodiment.

[0085] Figure 2 This is a flowchart illustrating a compressor operation control method according to some embodiments.

[0086] like Figure 2 As shown, the compressor operation control method includes:

[0087] S201. During the operation of the multi-split air conditioning system, acquire the operating data and environmental parameters of the multi-split air conditioning system.

[0088] Operational data refers to parameters collected during the operation of a multi-split air conditioning system, including but not limited to compressor frequency, capacity of each indoor unit, start / stop status of indoor units, operating frequency of each compressor, total system load, and total system power.

[0089] For example, operating data might include the operating frequency of all compressors at a certain moment being 60Hz and the total system load being 800kW.

[0090] The operating parameters can be obtained from sensors in the multi-split air conditioning system or stored in advance in the controller. This application embodiment does not limit the method of obtaining the operating parameters.

[0091] Environmental parameters refer to parameters related to the operating environment of the air conditioning system, including but not limited to ambient temperature, ambient humidity, solar radiation intensity, indoor temperature of each air-conditioned room, and weather conditions. Ambient temperature can include the indoor temperature and / or outdoor temperature, and ambient humidity can include the indoor humidity and / or outdoor humidity.

[0092] For example, outdoor temperature and humidity can be collected by a temperature sensor installed in the outdoor unit of a multi-split air conditioning system. Weather conditions can be obtained via a network connection. This application does not limit the method of acquiring environmental parameters.

[0093] For example, environmental parameters might include an outdoor temperature of 35°C, an outdoor humidity of 60%, and a solar radiation intensity of 800 W / m².

[0094] This application does not limit the operating data and environmental parameters in its embodiments.

[0095] S202. By processing the operating data and environmental parameters through a pre-built load prediction model, the total load of the system in the next control cycle is obtained.

[0096] A predictive model is a machine learning model used to predict the total load of a system within a future control cycle. Its inputs are operating data and environmental parameters, and its output is the predicted total load value.

[0097] For example, the prediction model may be trained based on historical operational data and meteorological data, such as a model built using a Bi-LSTM network combined with a Bayesian optimization algorithm.

[0098] In this application, taking the hierarchical bidirectional long short-term memory network (Bi-LSTM) model as an example, the total load of the system in the next control cycle can be obtained by processing the operating data and environmental parameters through the pre-built load prediction model. This can include: inputting historical operating data and environmental parameters into the hierarchical bidirectional long short-term memory network model to generate short-term load prediction results and long-term load prediction results; and generating the total load of the system in the next control cycle by weighted fusion of the short-term load prediction results and long-term load prediction results.

[0099] The hierarchical Bi-LSTM model is a Bi-LSTM model consisting of a short-term prediction layer (minute level) and a long-term prediction layer (hour level / day level).

[0100] For example, a hierarchical Bi-LSTM model is used to process short-term (minute-level) high-frequency fluctuation data and long-term (hour-level / day-level) low-frequency trend data respectively to generate short-term and long-term load forecast results. By weighted fusion of the two forecast results, a forecast result of the total system load within the future control cycle is generated.

[0101] In this way, by using a hierarchical bidirectional long short-term memory network model to predict the total load in the next control cycle, the accuracy of the predicted total load is higher, thereby improving the accuracy of the subsequently determined target start-up combination.

[0102] S203. Based on the total load, the performance parameters of multiple compressors, and the capacity of multiple outdoor units, determine the target start-up combination of the compressors, and control the operation of the compressors in the target start-up combination in the next control cycle.

[0103] The target start-up combination may include one compressor or multiple compressors. In this application embodiment, the number of compressors in the target start-up combination is not limited.

[0104] The control cycle can be a pre-set cycle for re-determining the compressor start-up combination. For example, the control cycle can be 5 minutes, 10 minutes, or 30 minutes. This application embodiment does not limit the control cycle.

[0105] In this way, during system operation, the total system load for the next control cycle is predicted based on system operating data and environmental parameters. Then, based on the total load, the performance parameters of multiple compressors, and the capacity of multiple outdoor units, the operating combination of compressors meeting the load requirements is determined. This allows the compressors within that operating combination to be controlled in the next controller cycle. This ensures that the determined operating combination meets the actual needs of the system in the next control cycle, while also taking into account compressor performance, resulting in higher accuracy and effectively improving compressor operating efficiency.

[0106] The following is a detailed explanation of the method for determining the target power-on combination.

[0107] Figure 3 This is a flowchart illustrating a method for determining a target power-on combination according to some embodiments.

[0108] like Figure 3 As shown, the method includes:

[0109] S301. Based on the total load and the capacity of multiple outdoor units, determine multiple first start-up combinations and the sub-loads corresponding to each outdoor unit in the first start-up combination.

[0110] The sum of the capacities of at least one outdoor unit included in the multiple first-start combinations is greater than or equal to the total load.

[0111] The capacity of the outdoor unit can be pre-stored, and the capacities of different outdoor units can be the same or different. This application does not specifically limit this.

[0112] For example, for any first start-up combination, the controller can use an average distribution or a capacity-weighted distribution load distribution method to distribute the total load to the outdoor units corresponding to the compressors in the start-up combination, thereby determining the sub-load of the outdoor units corresponding to each compressor in the first start-up combination.

[0113] Specifically, average distribution can include: distributing the total load equally among the outdoor units; when the average load exceeds the capacity of a particular outdoor unit, that outdoor unit is assigned the maximum load, and the remaining outdoor units share the excess load equally.

[0114] Capacity-weighted allocation: The total load is allocated according to the outdoor unit capacity using the following formula:

[0115]

[0116] Where Q is the total load, Q i The sub-load allocated to the i-th outdoor unit, G i The capacity of each outdoor unit.

[0117] For example, a multi-split air conditioning system includes four outdoor units, numbered a, b, c, and d, with capacities of 448, 560, 560, and 672 respectively.

[0118] When the total demand load is 400, the first start-up combination includes the following Table 1:

[0119] Table 1

[0120]

[0121] When the demand load is 600, the first start-up combination includes the following Table 2:

[0122] Table 2

[0123]

[0124] When the demand load is 1000, the first start-up combination includes the following Table 3:

[0125] Table 3

[0126]

[0127] When the demand load is 1500, the first start-up combination includes the following Table 4:

[0128] Table 4

[0129]

[0130] When the demand load is 1650, the first start-up combination includes the following Table 5:

[0131] Table 5

[0132]

[0133] S302. Based on the sub-loads corresponding to each outdoor unit and the performance parameters of multiple compressors, determine the target start-up combination among multiple first start-up combinations.

[0134] In this application, determining the target start-up combination among multiple first start-up combinations based on the sub-loads corresponding to each outdoor unit and the performance parameters of multiple compressors may include: determining the operating frequency of the compressor corresponding to the outdoor unit based on the sub-loads corresponding to the outdoor unit in each first start-up combination; determining the energy efficiency of the compressor within the high energy efficiency range of the compressor based on the operating frequency of the compressor and the performance parameters of the compressor; and determining the total energy efficiency of each first start-up combination based on the energy efficiency of each compressor.

[0135] For example, the performance parameters of a compressor are its performance curves, which represent the relationship between the compressor's energy efficiency and its operating frequency.

[0136] The performance curves of different compressors may be the same or different, depending on the actual situation of the compressor. This application does not limit this.

[0137] The compressor's performance parameters can be found in [reference]. Figure 4 As shown, Figure 4 This is a schematic diagram of the performance curve of a compressor according to some embodiments.

[0138] like Figure 4 As shown in the compressor performance curve, the horizontal axis represents the compressor's operating frequency (in Hertz, Hz), and the compressor's output power can be changed by adjusting the frequency. The vertical axis represents the Coefficient of Performance (COP), which is the ratio of cooling capacity to input power and directly reflects the compressor's energy efficiency level. A higher COP value indicates better energy efficiency.

[0139] In this application, the performance coefficient of the compressor can be used to represent energy efficiency, so that the performance coefficient is equal to the energy efficiency.

[0140] exist Figure 4 In this context, the range where the system's performance exceeds a preset value is defined as the high-performance range.

[0141] For example, when the controller determines the compressor's energy efficiency in the high energy efficiency range, it can determine the energy efficiency closest to the maximum energy efficiency value as the compressor's energy efficiency.

[0142] In this application, for each first start-up combination, the energy efficiency of each compressor included can be determined, and the sum of the energy efficiencies of each compressor can be determined as the total energy efficiency of the first start-up combination, thereby obtaining the total energy efficiency of each first start-up combination.

[0143] Furthermore, among the multiple first power-on combinations, at least one second power-on combination is determined based on the total energy efficiency of the first power-on combinations; among the at least one second power-on combination, a target power-on combination that meets preset conditions is determined.

[0144] Among them, at least one second start-up combination is a start-up combination whose total energy efficiency and the total energy efficiency of the third start-up combination are less than a preset difference, and the third start-up combination is the first start-up combination with the largest total energy efficiency.

[0145] The preset difference can be an empirical value or a value obtained through prior simulation; this application does not limit this.

[0146] It should be understood that the second start-up combination determined by the above method is the start-up combination that is close to the first start-up combination with the greatest total energy efficiency.

[0147] In this way, by determining the maximum energy efficiency that each compressor can achieve, the total energy efficiency of each first start-up combination is calculated. Then, the start-up combination with the higher energy efficiency is selected from the total energy efficiency as the target start-up combination. This ensures that the compressors in the determined target start-up combination all have high energy efficiency, which can effectively improve the system's energy efficiency ratio.

[0148] In this application, the performance parameters of the compressor are obtained by pre-correcting the initial performance parameters based on the operating parameters of the compressor.

[0149] For example, based on historical operating data such as compressor frequency, total system load, and total system power, the energy efficiency of each compressor at a certain frequency can be calculated. Then, based on the actual operating energy efficiency, the original compressor performance curve can be corrected to obtain the corrected performance curve, i.e., the performance parameters. For instance, the least squares method can be used for fitting and correction.

[0150] For example, the performance curve of the compressor at the factory can be corrected by using the least squares method, and historical operating data (such as compressor frequency, total system load, and total power) can be fitted with the initial performance curve at the time of manufacture to generate a performance model that is closer to reality.

[0151] In this way, the compressor's performance parameters are corrected based on the actual operating conditions of the compressor, making the compressor's performance parameters more consistent with the actual situation. This improves the accuracy of the subsequently determined target start-up combinations, resulting in higher energy efficiency when the compressors in the target start-up combinations are operated.

[0152] In this application, determining the target power-on combination that satisfies the preset conditions in at least one second power-on combination may include the following two possible implementations:

[0153] One possible implementation is to determine the second power-on combination as the target power-on combination if the number of second power-on combinations is one.

[0154] When there is only one second start-up combination, it may be because the difference between the other first start-up combinations and the first start-up combination with the highest total energy efficiency is greater than the preset difference. In other words, the total energy efficiency of the other first start-up combinations is relatively small. Therefore, the unique second start-up combination is determined as the target start-up combination, so that the system has higher energy efficiency when the compressor in the target start-up combination is running.

[0155] Another possible implementation is to determine the total cumulative runtime of the compressor in each second start-up combination if there are at least two second start-up combinations; and to determine the second start-up combination with the smallest total cumulative runtime as the target start-up combination.

[0156] Cumulative runtime refers to the total runtime of the compressor in its historical operation, and is used to measure the degree of equipment wear.

[0157] If the cumulative operating time of the two compressors is 1000 hours and 1500 hours respectively, the compressor with the shorter operating time shall be selected.

[0158] When energy efficiency ratios are the same, the system prioritizes the combination with shorter cumulative operating time by comparing the cumulative operating time of the compressors. For example, when two combinations have the same energy efficiency ratio, the system will choose the combination with the shorter cumulative operating time to balance equipment wear and energy efficiency optimization.

[0159] In this way, by optimizing the cumulative runtime, equipment lifespan can be extended and mechanical wear reduced. For example, given the same energy efficiency ratio, compressors with shorter operating times can be prioritized to avoid damage to the equipment caused by frequent start-stop cycles.

[0160] In this application, after determining the target start-up combination, it can be determined whether all compressors within the target start-up combination meet their respective start-stop time thresholds. If at least one compressor does not meet its corresponding start-stop time threshold, then among the other start-up combinations, a new target start-up combination that meets the preset conditions is determined. The other start-up combinations are the start-up combinations other than the target start-up combination among the multiple first start-up combinations.

[0161] The start-stop time threshold requirement is the minimum critical time interval between two start-stop operations of the compressor.

[0162] The method for determining the target power-on combination that meets the preset conditions among other power-on combinations is the same as that for determining the target power-on combination among multiple first power-on combinations as described above. Please refer to the description of the above embodiments, which will not be repeated here.

[0163] In this way, by determining whether a new target start-up combination needs to be redefined based on the start-up and stop time thresholds of each compressor, the compressors in the final target start-up combination can meet their respective start-up and stop time threshold requirements, thereby avoiding frequent compressor start-ups and stops, reducing mechanical wear, and improving system stability.

[0164] It should be noted that in this application, the target start-up combination can be determined by calculating the energy efficiency ratio (EER) of the start-up combination. The EER can be calculated using the following formula:

[0165]

[0166] Where EE is the energy efficiency ratio of the power-on combination. Let be the energy efficiency ratio of the i-th compressor, i.e., the COP of a single compressor. Let Q be the load corresponding to the i-th compressor, and Q be the total load.

[0167] The method for determining the target start-up combination based on the energy efficiency ratio is similar to the method for determining the target start-up combination based on energy efficiency described above, and can be found in the description of the above embodiments, which will not be repeated here.

[0168] Therefore, the method for determining the target start-up combination in this application embodiment obtains the sub-load of each outdoor unit by allocating the total load to each outdoor unit in the first start-up combination, and then determines the target start-up combination based on the sub-load of the outdoor unit and the performance parameters of the compressor. This makes the accuracy of the determined target start-up combination higher and can further improve the operating efficiency of the compressor.

[0169] The following describes the method for creating a load forecasting model.

[0170] Figure 5 This is a schematic flowchart of a method for creating a load forecasting model according to some embodiments.

[0171] like Figure 5 As shown, methods for creating load forecasting models may include:

[0172] S501, Obtain historical data.

[0173] The types of historical data include operational data and environmental parameters as described in the above embodiments.

[0174] For example, historical data may include ambient temperature, ambient humidity, solar radiation intensity, indoor temperature of each air-conditioned room, capacity of each indoor unit, start / stop status of indoor units, operating frequency of each compressor, total system load, total system power, etc. This application embodiment does not limit the use of historical data.

[0175] S502. Preprocess the historical data to obtain the processed data.

[0176] The historical data is normalized using the following formula to obtain the processed data.

[0177]

[0178] Where x is the normalized data, and its value range is usually [-1, 1]. ori For historical data; x min The minimum value in the historical data set; x max This represents the maximum value in the historical dataset.

[0179] S503. Based on the processed data, construct a Bayesian-optimized bidirectional long short-term memory network energy consumption prediction model.

[0180] The Bi-LSTM model construction includes: constructing a bidirectional long short-term memory network model. The input parameters of this model include ambient temperature, ambient humidity, solar radiation intensity, average indoor temperature, total capacity of indoor units, and total capacity of operating indoor units. The output parameter is the total system load.

[0181] Perform Bayesian optimization; the process can be found in [link to relevant documentation]. Figure 6 As shown. Figure 6 This is a schematic diagram of a Bayesian optimization process according to some embodiments.

[0182] like Figure 6 As shown, Bayesian optimization includes:

[0183] (1) Initialize the hyperparameters of the Bi-LSTM model: Set the range of values ​​for the hyperparameters of the Bi-LSTM model, and randomly select hyperparameter observations to form a hyperparameter set D, where It is a combination of hyperparameter observations, and y is the objective function value.

[0184] (2) Constructing a Gaussian process model: Constructing a Bayesian optimized Gaussian model.

[0185] (3) Obtain the acquisition function based on the posterior distribution of the Gaussian process: The acquisition function (the desired improvement function) is determined by using the posterior distribution of the Gaussian process model. This function is used to guide the selection of the next hyperparameter sampling point in order to efficiently find the optimal hyperparameter.

[0186] (4) Obtain the next point and calculate Select the next hyperparameter sampling point based on the acquisition function. Input it into the Bi-LSTM model and calculate the corresponding objective function value. .

[0187] (5) Has the number of iterations been reached?: Determine if the current iteration has reached the preset number of iterations. If not, proceed to add a new point. Add the new hyperparameter samples to the hyperparameter set D, and then return to the step of building the Gaussian process model to update the Gaussian model.

[0188] (6) Output optimal parameters: Output the optimal combination of hyperparameters obtained by optimization, substitute it into the Bi-LSTM model, and finally obtain the Bi-LSTM energy consumption prediction model based on BO optimization.

[0189] In this application, the current load forecasting model can be optimized based on historical data to obtain a new load forecasting model, which can then be used for load forecasting.

[0190] Historical data includes the historical operating data, environmental parameters, and compressor start-up combinations of the multi-split air conditioning system during the model update cycle.

[0191] The model update cycle can be one month, 15 days, or other times; this application does not limit this.

[0192] In this way, the current prediction model is iteratively optimized using historical operating data to generate a new model for subsequent predictions. For example, during the model update cycle, the system adjusts the parameters of the Bi-LSTM model based on the latest operating data to adapt to performance drift caused by equipment aging or environmental changes. For instance, when compressor performance declines due to aging, the model can dynamically adjust parameters so that the prediction results still reflect the actual operating state, thereby maintaining the long-term optimization of the system's energy efficiency ratio.

[0193] Based on the content described in the above embodiments, the entire training and prediction process of the Bidirectional Long Short-Term Memory Network (Bi-LSTM) prediction model based on Bayesian optimization (BO) can be found in [link to documentation]. Figure 7 As shown. Figure 7 This is a schematic diagram illustrating the entire training and prediction process of a Bayesian-optimized bidirectional long short-term memory network prediction model according to some embodiments.

[0194] like Figure 7 As shown, the entire training and prediction process of the Bidirectional Long Short-Term Memory (Bi-LSTM) prediction model based on Bayesian optimization (BO) includes:

[0195] Input training set: The training set is the source of input data for model training. It contains historical data (such as ambient temperature, humidity, air conditioning operating parameters, system load, etc.) used to learn data patterns and is the basic data support for the model to learn time-series correlation features.

[0196] Data preprocessing: The original data of the training set is standardized (usually by normalization, mapping the data to the [-1,1] interval) to eliminate the scale difference of data in different dimensions, avoid the interference of data volume on model training, and obtain standardized data suitable for Bi-LSTM training.

[0197] Initialize Bi-LSTM model hyperparameters: Start the Bayesian optimization process, first set the value range of Bi-LSTM model hyperparameters (such as the number of hidden layer units, the number of LST layers, etc.), and then randomly select initial hyperparameter combinations to form an initial set containing hyperparameters and corresponding model performance, providing a starting point for subsequent optimization.

[0198] Construct a Gaussian model and obtain the acquisition function: Construct a Bayesian optimized Gaussian process model to fit the relationship between hyperparameters and model performance; then, based on the posterior distribution of the Gaussian model, determine the acquisition function (such as the desired improvement function), which is used to efficiently select the next combination of hyperparameters to be verified.

[0199] The number of hidden layer units, the number of LSTM layers, the number of training iterations, and the learning rate are the target hyperparameters for this Bayesian optimization. They directly affect the model structure and training efficiency of Bi-LSTM. Different combinations of parameters will be input into the Bi-LSTM algorithm, resulting in different model performances. They are the core adjustment objects in the optimization process.

[0200] Bi-LSTM algorithm: Substitute the currently selected hyperparameter combination into the Bi-LSTM model, use the preprocessed training set data, and learn the correlation patterns in the data through bidirectional time-series memory (such as the time-series correspondence between environmental parameters and system load) to complete a single model training.

[0201] The next sampling point is calculated as follows: Based on the sampling function, the next hyperparameter combination to be verified is selected from the hyperparameter space, and after inputting it into the Bi-LSTM algorithm for training, the corresponding model performance index is obtained.

[0202] Determine if the iteration count has been reached: Determine if the current optimization iteration has reached the preset number. If not, feed the new sampling points back to the Gaussian model construction step, update the surrogate model and sampling function, and continue to the next round of hyperparameter selection; if the count has been reached, terminate the iteration.

[0203] Optimal hyperparameter combination: After the iteration terminates, select the hyperparameter combination that optimizes the performance of the Bi-LSTM model from all the hyperparameter sampling results. This is the final output of Bayesian optimization and is used to build a high-performance prediction model.

[0204] Input test set: An independent dataset used to verify the generalization ability of the final model. It contains data of the same type as the training set but without overlap, to avoid the model overfitting the training set and to ensure the reliability of the prediction results.

[0205] Data preprocessing: Perform preprocessing operations (such as normalization) on the original test set data that are exactly the same as those on the training set to ensure that the format and scale of the test data match those of the training data and avoid data differences from affecting the accuracy of the prediction results.

[0206] The final Bi-LSTM model: By substituting the optimal combination of hyperparameters into the Bi-LSTM model and combining it with the temporal patterns learned from the training set, the final and optimized prediction model is constructed, which is the core tool for completing the prediction task.

[0207] Output prediction results: Input the preprocessed test set data into the final Bi-LSTM model. The model outputs prediction results (such as the predicted value of system load) based on the learned rules, thus completing the entire prediction process.

[0208] Based on the content described in the above embodiments, actual test data is used as the training set and detection set to verify the performance and prediction accuracy of the method proposed in this application.

[0209] First, the Bayesian optimization algorithm was used to optimize the hyperparameters of the Bi-LSTM. The upper and lower bounds for the hyperparameters in Bi-LSTM—number of hidden units, number of LSTM layers, number of training iterations, and learning rate—were [10, 100], [1, 5], [10, 100], and [0.0001, 0.01], respectively. After Bayesian optimization, the optimal hyperparameters for the training set were obtained: 19 hidden units, 3 LSTM layers, 23 training iterations, and a learning rate of 0.0187. The trained model was then used to validate its performance on the validation set. The results are as follows: Figure 8 As shown. Figure 8 This is a schematic diagram comparing a predicted power with the actual power according to some embodiments.

[0210] like Figure 8 The energy consumption prediction accuracy rate shown is over 85%. Table 6 shows a comparison of some predicted and actual power values.

[0211] Table 6

[0212]

[0213]

[0214] This application also provides a computer-readable storage medium, which may include various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a disk, or an optical disk. Specifically, the computer-readable storage medium stores computer-executable instructions, which are executed by a computer to implement the technical solutions shown in the above-described method embodiments.

[0215] This application also provides a program product, which includes executable instructions stored in a readable storage medium. When the computer program is executed by a computer, the technical solution shown in the above method embodiments is executed. The specific implementation method and technical effect are similar, and will not be described again here.

[0216] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0217] For ease of explanation, the above description has been provided in conjunction with specific embodiments. However, the above exemplary discussion is not intended to be exhaustive or to limit the embodiments to the specific forms disclosed above. Various modifications and variations can be obtained based on the above teachings. The selection and description of the above embodiments are for the purpose of better explaining the principles and practical applications, thereby enabling those skilled in the art to better utilize the embodiments and various different variations of embodiments suitable for specific application considerations.

[0218] In this application, "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent three cases: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document indicates that the preceding and following related objects have an "or" relationship.

[0219] "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or multiple items. For example, at least one of a, b, or c can mean: a, b, c, a and b, a and c, b and c, or a, b, and c, where each of a, b, and c can be an element itself or a set containing one or more elements.

[0220] In this application, "at least one" means one or more. "More than one" means two or more. The descriptions of "first," "second," etc., appearing in the embodiments of this application are only for illustration and to distinguish the described objects, and have no order, nor do they indicate a special limitation on the number of devices in the embodiments of this application, and cannot constitute any limitation on the embodiments of this application. For example, "first threshold" and "second threshold" are only used to distinguish different thresholds, and do not indicate that the size, priority, or importance of these two thresholds are different.

[0221] In this application, terms such as "exemplary," "in some embodiments," and "in other embodiments" are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in this application should not be construed as being more preferred or advantageous than other embodiments or designs. Rather, the term "exemplary" is used to present the concept in a specific manner.

[0222] In this application, the terms "of," "corresponding (relevant)," "corresponding," and "related" may sometimes be used interchangeably. It should be noted that, unless a distinction is emphasized, their intended meanings are consistent. Similarly, in the embodiments of this application, "communication" and "transmission" may sometimes be used interchangeably. It should be noted that, unless a distinction is emphasized, their intended meanings are consistent. For example, transmission can include sending and / or receiving, and can be a noun or a verb.

[0223] In this application, "equal to" can be used with "less than" or "greater than", but not simultaneously with both. When "equal to" is used with "less than", it applies to the technical solution adopted by "less than". When "equal to" is used with "greater than", it applies to the technical solution adopted by "greater than".

Claims

1. A multi-split air conditioning system, characterized in that, include: At least one indoor unit; At least two outdoor units; The compressor is located inside the outdoor unit; The controller is configured as follows: During the operation of the multi-split air conditioning system, the operating data and environmental parameters of the multi-split air conditioning system are acquired; By processing the operating data and environmental parameters using a pre-built load prediction model, the total load of the system in the next control cycle is obtained. Based on the total load, the performance parameters of multiple compressors, and the capacity of multiple outdoor units, a target start-up combination of compressors is determined, and the compressors in the target start-up combination are controlled to operate in the next control cycle.

2. The multi-split air conditioning system according to claim 1, characterized in that, The controller is configured as follows: Based on the total load and the capacity of the multiple outdoor units, a plurality of first start-up combinations and the sub-load of the outdoor unit corresponding to each compressor in the first start-up combination are determined; the sum of the capacities of at least one outdoor unit included in the plurality of first start-up combinations is greater than or equal to the total load; Based on the sub-loads corresponding to each outdoor unit and the performance parameters of multiple compressors, a target start-up combination is determined from the multiple first start-up combinations.

3. The multi-split air conditioning system according to claim 1, characterized in that, The controller is configured as follows: Based on the sub-load corresponding to the outdoor unit in each first start-up combination, determine the operating frequency of the compressor corresponding to the outdoor unit; Based on the compressor's operating frequency and performance parameters, the compressor's energy efficiency is determined within its high-efficiency range. Determine the total energy efficiency of each first start-up combination based on the energy efficiency of each compressor; Among the plurality of first power-on combinations, at least one second power-on combination is determined based on the total energy efficiency of the first power-on combination; the at least one second power-on combination is a power-on combination whose total energy efficiency differs from that of the third power-on combination by a preset difference, and the third power-on combination is the first power-on combination with the highest total energy efficiency; Among the at least one second power-on combination, a target power-on combination that meets preset conditions is determined.

4. The multi-split air conditioning system according to claim 3, characterized in that, The performance parameters of the compressor are the correspondence between the compressor's energy efficiency and operating frequency; the performance parameters of the compressor are obtained by pre-correcting the initial performance parameters based on the compressor's operating parameters.

5. The multi-split air conditioning system according to claim 3, characterized in that, The controller is configured as follows: If the number of second power-on combinations is one, then the second power-on combination is determined to be the target power-on combination.

6. The multi-split air conditioning system according to claim 3, characterized in that, The controller is configured as follows: If the number of second start-up combinations is at least two, then determine the total cumulative runtime of the compressor in each second start-up combination; The second startup combination with the smallest total cumulative runtime is determined as the target startup combination.

7. The multi-split air conditioning system according to claim 5 or 6, characterized in that, The controller is also configured to: Determine whether the compressors in the target start-up combination all meet their respective start-up and stop time threshold requirements; If at least one compressor does not meet the corresponding start-stop time threshold requirement, then among the other start-up combinations, a new target start-up combination that meets the preset conditions is determined; the other start-up combinations are the start-up combinations other than the target start-up combination among the plurality of first start-up combinations.

8. The multi-split air conditioning system according to claim 1, characterized in that, The load prediction model is a hierarchical bidirectional long short-term memory network model. The controller is also configured to: Historical operating data and the environmental parameters are input into a hierarchical bidirectional long short-term memory network model to generate short-term load forecast results and long-term load forecast results. The total load of the system in the next control cycle is generated by weighted fusion of the short-term load forecast results and the long-term load forecast results.

9. The multi-split air conditioning system according to claim 8, characterized in that, The controller is also configured to: The current load forecasting model is optimized based on historical data to obtain a new load forecasting model, which is then used for load forecasting. The historical data includes the historical operating data, environmental parameters, and compressor start-up combinations of the multi-split air conditioning system during the model update cycle.

10. A compressor operation control method, characterized in that, Applied to multi-split air conditioning systems, the multi-split air conditioning system includes: At least one indoor unit; At least two outdoor units; The compressor is located inside the outdoor unit; Controller; The method includes: During the operation of the multi-split air conditioning system, the operating data and environmental parameters of the multi-split air conditioning system are acquired; By processing the operating data and environmental parameters using a pre-built load prediction model, the total load of the system in the next control cycle is obtained. Based on the total load, the performance parameters of multiple compressors, and the capacity of multiple outdoor units, a target start-up combination of compressors is determined, and the compressors in the target start-up combination are controlled to operate in the next control cycle.

Images