Air conditioning system, control method, computer readable storage medium and program product

Abstract

The application provides an air conditioning system and a control method, a computer readable storage medium and a program product. The control method comprises the following steps: determining all start-up combination modes of single or multiple water chillers according to the total number of the water chillers; for each mode, planning multiple value combination schemes of multiple characteristic parameters, inputting each scheme into each single-machine power model to obtain multiple single-machine powers and total power, and obtaining a total power data table; obtaining multiple characteristic parameters and total power measured values collected in real time, and updating the total power data table according to the multiple characteristic parameters and the total power measured values; for each start-up combination mode, constructing a total power model according to the total power data table; inputting the multiple characteristic parameters into each total power model to calculate corresponding total power, and determining a start-up combination mode with minimum total power as a target mode; and controlling the air conditioning system to operate in the target start-up combination mode. The application solves the problems that the water chiller number selection model has low generalization and low accuracy, and is difficult to operate stably for a long time.

Description

Technical Field

[0001] This invention relates to the field of central air conditioning technology, and in particular to an air conditioning system and control method, a computer-readable storage medium, and a program product. Background Technology

[0002] Central air conditioning systems typically consist of multiple chiller units connected in parallel. The number of chiller units in operation is adjusted according to the size of the terminal load to achieve energy-saving effects.

[0003] Chillers are the main energy-consuming equipment in air conditioning systems, accounting for a particularly large proportion of the system's energy consumption. Meanwhile, to meet the requirements of indoor thermal comfort in buildings, air conditioning systems are usually designed based on maximum cooling load, which means that chiller units operate at partial load most of the time. Therefore, optimizing the number of chiller units in operation based on the air conditioning system load, so that each chiller unit operates at its optimal overall performance, is crucial for building energy conservation.

[0004] Currently, chiller unit count control algorithms are typically based on chiller power models. However, these models mostly only consider the impact of load rate on chiller power, resulting in low model accuracy. Furthermore, models built based on manufacturer selection reports often differ significantly from actual field operating data. Models built based on historical field operating data exhibit poor generalization outside of the data conditions and fail to account for the performance degradation of chillers over time, leading to model failure. This results in low model accuracy and makes the algorithm unreliable for long-term use. Summary of the Invention

[0005] The purpose of this invention is to provide an air conditioning system and control method, a computer-readable storage medium, and a computer program product that can overcome or at least partially solve the above-mentioned problems.

[0006] The purpose of this invention is to solve the problems of low generalization and low accuracy of the chiller unit selection model, as well as the difficulty in long-term stable operation.

[0007] On one hand, the present invention provides a control method for an air conditioner, wherein the air conditioning system includes multiple chiller units, and the control method includes:

[0008] The total number of chiller units determines all possible start-up combinations for single or multiple chiller unit operation.

[0009] For each of the aforementioned start-up combinations, multiple value combination schemes of various characteristic parameters are planned and preset. The data of each value combination scheme is input into the single-unit power model of each chiller unit to obtain multiple single-unit power and sum them to obtain the total power, so as to obtain the total power data table of the start-up combination.

[0010] The system acquires the real-time collected multiple characteristic parameters and the measured value of the total power, and updates the total power data table accordingly.

[0011] For each of the aforementioned power-on combinations, a total power model is constructed based on the total power data table, using the various characteristic parameters as input parameters and the total power as the output parameter.

[0012] The current values ​​of the various feature parameters are input into the total power model of each of the power-on combinations to calculate the corresponding total power, and the power-on combination with the minimum total power is determined as the target power-on combination.

[0013] Control the air conditioning system to operate in the target start-up combination mode.

[0014] Optionally, "acquiring the measured values ​​of the various characteristic parameters and the total power collected in real time, and updating the total power data table accordingly" includes:

[0015] For each set of measured data collected, the set of original data that is closest to the value of that set of measured data in the total power data table is selected.

[0016] The original data that has been filtered out is placed at the front of the table and then deleted. The actual test data is then inserted at the back of the table to complete the queue-style update.

[0017] Optionally, the various characteristic parameters include chilled water supply temperature, cooling water return temperature, and total system load;

[0018] "For each set of measured data collected, the original data that is closest to the value of that set of measured data in the total power data table is selected," including:

[0019] In the total power data table, the data is filtered layer by layer in the order of the system total load closest, the chilled water supply temperature closest, and the cooling water return temperature closest, until the filtering result is a unique set. The original data of this set is used as the original data that is closest to the measured data value of this set.

[0020] Optionally, the various characteristic parameters include chilled water supply temperature, cooling water return temperature, and total system load;

[0021] "For each of the aforementioned power-on combinations, constructing a total power model based on its total power data table, using the various characteristic parameters as input parameters and the total power as the output parameter" includes:

[0022] The total power model is constructed as follows:

[0023]

[0024]

[0025]

[0026] Where y is the total power, c0~c5 are model parameters, and X1, X2, and X3 are the cooling water return temperature, chilled water supply temperature, and total system load, respectively.

[0027] Calculated using the recursive least squares method with a variable forgetting factor

[0028] Optionally, "the result is obtained by recursive least squares method with variable forgetting factor". "include:

[0029] The total power data table values ​​are recorded sequentially in ascending order of total system load as follows:

[0030]

[0031] Determine the loss function:

[0032]

[0033] Where m is the number of total power values ​​in the total power data table, and λ is the forgetting factor, 0 < λ ≤ 1;

[0034] Determine using the following steps To minimize the loss function J;

[0035] Step A, Parameter Initialization: Set initial values ​​for the following parameters;

[0036]

[0037]

[0038] Step B: Sample the current input and output data:

[0039]

[0040] During the first recursion, k = 1;

[0041] Step C: Update the gain matrix according to the following formula:

[0042]

[0043] Step D: Update the parameters according to the following formula:

[0044]

[0045] Step E: Update the covariance according to the following formula:

[0046]

[0047] Step F: Perform error verification according to the following formula:

[0048]

[0049] Step G: Update the forgetting factor according to the following formula:

[0050] λ = 1 - α × |tanh(ε(k))|, where 0 < α ≤ 0.1;

[0051] Increment the value of k by 1, and return to execute steps B through G until k = m.

[0052] Optionally, the control method further includes:

[0053] The database contains records of multiple sets of operating parameters for each chiller unit.

[0054] Based on the original database of each chiller unit, the single-unit power model of the chiller unit is trained and used to calculate the single-unit power of the chiller unit.

[0055] Optionally, the original database includes cooling water return temperature, chilled water supply temperature, single unit load, and single unit power;

[0056] "Based on the original database of each chiller unit, the single-unit power model of that chiller unit is trained" includes:

[0057] The single-machine power model is constructed as follows:

[0058]

[0059] Among them, P i Let a be the single-unit power of the i-th chiller unit. 0，i ~a 5，i Let X be the model parameters for the i-th chiller unit; 1，i X 2，i X 3，i These are the cooling water return temperature, chilled water supply temperature, and single unit load of the i-th chiller unit, respectively.

[0060] Based on the data from the original database, a is obtained through data fitting. 0，i ~a 5，i The value of .

[0061] Optionally, the various characteristic parameters include total system load x3;

[0062] The input parameters for each of the single-unit power models include the single-unit load X of the corresponding chiller unit. 3i ;

[0063] In each of the aforementioned power-on combinations, n represents the number of chiller units in this start-up combination.

[0064] Optionally, the various characteristic parameters include total system load x3;

[0065] In the total power data table calculated based on the preset single-unit power model, the total system load X3 takes multiple values ​​at equal intervals between the minimum allowable load and the maximum allowable load of all chiller units in the corresponding start-up combination mode;

[0066] The maximum allowable load is the sum of the maximum loads of all individual chiller units, and the minimum allowable load is t times the maximum value among the maximum loads of each individual unit, where t < 1.

[0067] Optionally, the control method further includes:

[0068] The database contains records of multiple sets of operating parameters for each chiller unit.

[0069] Based on the original database of each chiller unit, a single-unit maximum load model of the chiller unit is trained and used to calculate the single-unit maximum load of the chiller unit.

[0070] Optionally, the original database also includes the load rate of the chiller unit;

[0071] "Based on the original database for each chiller unit, the single-unit maximum load model for that chiller unit is trained" includes:

[0072] The single-machine maximum load model is constructed as follows:

[0073] Q max，i (X 1，i X 2，i )=b 0，i +b 1，i ×(X 1，i -X 2，i )+b 2，i ×(X 1，i -X 2，i ) 2 +b 3，i ×X 1，i +b 4，i ×X 2，i ;

[0074] Among them, Q max，i b is the maximum single-unit load of the i-th chiller unit. 0，i ~b 4，i Let X be the model parameters of the i-th chiller unit. 1，i X 2，iThese are the cooling water return temperature and chilled water supply temperature of the i-th chiller unit when the load rate is 100%, respectively.

[0075] Based on the data from the original database, b is obtained through data fitting. 0，i ~b 4，i The value of .

[0076] Optionally, the various characteristic parameters include chilled water supply temperature, cooling water return temperature, and total system load; or

[0077] The various characteristic parameters include chilled water supply temperature, cooling water outlet temperature, and total system load.

[0078] Optionally, the various characteristic parameters may also include chilled water flow rate and / or cooling water flow rate.

[0079] On the other hand, the present invention also provides a control device, including a memory and a processor, the memory storing a computer program, which, when executed by the processor, is used to implement the control method according to any one of the above.

[0080] On the other hand, the present invention also provides an air conditioning system that includes the control device as described above.

[0081] On the other hand, the present invention also provides a computer program that, when executed by a processor, implements the control method as described in any of the preceding claims.

[0082] On the other hand, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the control method as described in any of the above claims.

[0083] The control method of the air conditioning system of the present invention constructs a total power data table for each possible start-up combination of the chiller unit. The table includes characteristic parameters and total power. A total power model is constructed based on the total power data table so that when the air conditioning system is running, the current value of the characteristic parameter is input into the total power model of each start-up combination to calculate the corresponding total power. The start-up combination with the minimum total power is determined as the target start-up combination.

[0084] Furthermore, and importantly, in the control method of the present invention, the total power data table is updated based on the actual operating data of the chiller unit, so that the data in the total power data table is not limited to theoretical calculation values, but can better reflect the actual operating status and capacity of the chiller unit under the current working environment. This makes the total power model calculated based on the total power data table more accurate and reliable, so that the start-up combination selected according to each total power model is indeed the current optimal start-up combination.

[0085] Furthermore, in the control method of the present invention, when updating the total power data table, the various characteristic parameters in each set of real-time data collected are compared with the same parameters in the total power data table. The original data in the total power data table that is closest to the set of real-time data in terms of various characteristic parameters is deleted and the set of real-time data is added to the table. Through gradual updates and replacements, the data in the total power data table becomes more accurate and reliable, and can better reflect the actual operating status and capacity of the chiller unit under the current working environment.

[0086] Furthermore, in the control method of the present invention, the data in the total power data table is updated in a queue-style update manner, and the total power model is generated by the recursive least squares method with a variable forgetting factor. The introduction of the forgetting factor makes the newly added data in the table have a larger weight value when training parameters, so as to improve the model accuracy and tracking ability, enable the algorithm to run stably for a long time, and make it more practical.

[0087] The above and other objects, advantages and features of the present invention will become more apparent to those skilled in the art from the following detailed description of specific embodiments of the invention in conjunction with the accompanying drawings. Attached Figure Description

[0088] The following sections will describe some specific embodiments of the invention in detail by way of example and not limitation, with reference to the accompanying drawings. The same reference numerals in the drawings denote the same or similar parts or portions. Those skilled in the art should understand that these drawings are not necessarily drawn to scale. In the drawings:

[0089] Figure 1 This is a schematic diagram of a control method for an air conditioning system according to an embodiment of the present invention;

[0090] Figure 2 This is a flowchart illustrating a control method for an air conditioning system according to an embodiment of the present invention;

[0091] Figure 3 This is a schematic diagram of a control device for an air conditioning system according to an embodiment of the present invention;

[0092] Figure 4 This is a schematic diagram of an air conditioning system according to an embodiment of the present invention;

[0093] Figure 5 This is a schematic block diagram of a computer-readable storage medium according to an embodiment of the present invention;

[0094] Figure 6 This is a schematic block diagram of a computer program product according to an embodiment of the present invention. Detailed Implementation

[0095] The flowchart provided in this embodiment is not intended to indicate that the operations of the method will be performed in any particular order, or that all operations of the method are included in every case. Furthermore, the method may include additional operations. Within the scope of the technical concept provided by the method in this embodiment, additional variations can be made to the above method.

[0096] It should be noted here that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus or device (such as a computer-based system, a processor-included system or other system that can fetch and execute instructions from, an instruction execution system, apparatus or device).

[0097] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of an instruction, which contains one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0098] One embodiment of the present invention provides a control method for an air conditioning system. The air conditioning system includes multiple chiller units. The chiller units can be connected in parallel. Each chiller unit includes a vapor compression refrigeration cycle system consisting of a compressor, a condenser, a throttling device, an evaporator, and other components. The condenser dissipates heat through cooling water, the evaporator produces chilled water, and a water pump delivers the chilled water to the air conditioning terminals to achieve cooling of the indoor environment.

[0099] Figure 1 This is a schematic diagram of a control method for an air conditioning system according to an embodiment of the present invention.

[0100] Step S102: Determine the total number of chiller units and the complete set of start-up combinations for single or multiple chiller units based on the total number of chiller units.

[0101] Step S104: For each start-up combination mode, plan multiple combinations of preset characteristic parameters, input the data of each combination mode into the single-unit power model of each chiller unit, obtain multiple single-unit power and sum them to obtain the total power, so as to obtain the total power data table of the start-up combination mode.

[0102] Step S106: Obtain the measured values ​​of various characteristic parameters and total power collected in real time, and update the total power data table accordingly.

[0103] Step S108: For each power-on combination, construct a total power model with multiple characteristic parameters as input parameters and total power as output parameter based on its total power data table.

[0104] Step S110: Input the current values ​​of various characteristic parameters into the total power model of each power-on combination, calculate the corresponding total power, and determine the power-on combination with the minimum total power as the target power-on combination.

[0105] Step S112: Control the air conditioning system to operate in the target start-up combination mode.

[0106] Specifically, in step S102, the total number of chiller units is defined as N. All possible start-up combinations include: starting any 1 unit, any 2 units, and so on up to starting any N-1 units, as well as starting all N units. For example, if the air conditioning system includes two chiller units A and B, then all possible start-up combinations include three types: starting chiller unit A, starting chiller unit B, and starting both chiller units A and B.

[0107] In step S104, "multiple characteristic parameters" can be related parameters such as environmental operating conditions and the chiller's own performance parameters. For example, multiple characteristic parameters include chilled water supply temperature, cooling water return temperature, and total system load. Alternatively, multiple characteristic parameters may include chilled water supply temperature, cooling water outlet temperature, and total system load. In addition, multiple characteristic parameters may also include chilled water flow rate and / or cooling water flow rate. In planning multiple combinations of values ​​for various characteristic parameters, taking the characteristic parameters including chilled water supply temperature, cooling water return temperature, and total system load as an example, the range of chilled water supply temperature can be [T1, T2] (e.g., [5℃, 18℃]), with one temperature value taken every T0 (e.g., 0.5℃), resulting in 'a' temperature values; the range of cooling water return temperature can be [T3, T4] (e.g., [20℃, 35℃]), with one temperature value taken every T0 (e.g., 0.5℃), resulting in 'b' temperature values; similarly, one point value is taken for the total system load at preset intervals, resulting in 'c' total load values. By arranging and combining the above data for the three characteristic parameters, a maximum of a*b*c value combinations can be arranged, each including one value for chilled water supply temperature, one value for cooling water return temperature, and one value for total system load.

[0108] Table 1 illustrates a "startup combination" with two units in operation, and provides a total power data table with chilled water supply temperature, cooling water return temperature, and total system load as characteristic parameters.

[0109] Table 1

[0110] <![CDATA[Q1]]> 29 7 <![CDATA[P1+P2]]> <![CDATA[Q2]]> 29.5 7 <![CDATA[P1+P2]]> … … … … <![CDATA[Q n ]]> 30 7 <![CDATA[P1+P2]]>

[0111] In Table 1, X1, X2, and X3 represent the chilled water supply temperature, cooling water return temperature, and total system load, respectively. sum This represents the total power. The values ​​of X1, X2, and X3 in each row of Table 1 constitute a so-called "value combination scheme." For example, in the first row, P1 represents the single-unit power calculated based on the single-unit power model of chiller unit A when X3 = Q1, X1 = 29℃, and X2 = 7℃. P2 represents the single-unit power calculated based on the single-unit power model of chiller unit B when X3 = Q, X1 = 29℃, and X2 = 7℃. P1 + P2 represents the total power of chiller units A and B. The single-unit power model is either pre-stored in the system or calculated in advance, which will be discussed in detail later.

[0112] In step S106, the characteristic parameters and total power are collected in real time during the operation of the air conditioning system to obtain the actual data of the chiller unit during operation, and the total power data table is updated accordingly.

[0113] In step S108, for each power-on combination method, a total power model is constructed based on its total power data table, using multiple characteristic parameters as input parameters and total power as the output parameter. It can be seen that the number of total power models is the same as the number of power-on combinations and corresponds one-to-one.

[0114] In step S110, the current values ​​of various characteristic parameters are input into the total power model of each power-on combination mode, the corresponding total power is calculated, and the power-on combination mode with the minimum total power is determined as the target power-on combination mode.

[0115] Step S112: Control the air conditioning system to operate in the target start-up combination mode, switch the start-up combination scheme of the chiller unit to achieve the lowest energy consumption effect.

[0116] The inventors recognized that the raw data in the total power data sheet comes from theoretical calculations, which often deviate from actual field operation data. Furthermore, the data does not account for the performance degradation of the chiller unit over time, resulting in compromised accuracy. If a total power model is generated solely based on this data, the model's accuracy will be low, and the algorithm will be difficult to use reliably in the long term.

[0117] To address this deficiency, this embodiment of the invention updates the total power data table based on real operating data of the chiller unit. This ensures that the data in the total power data table is not limited to theoretical calculations, but rather better reflects the actual operating status and capacity of the chiller unit under the current working environment. This makes the subsequent total power model calculated based on the total power data table more accurate and reliable, better reflecting the actual operating status and capacity of the chiller unit under the current working environment. Consequently, the start-up combination selected based on each total power model is indeed the optimal start-up combination at present.

[0118] In some embodiments, the aforementioned step S106, namely "acquiring the measured values ​​of various characteristic parameters and total power collected in real time, and updating the total power data table accordingly", specifically includes:

[0119] For each set of measured data collected, the set of original data in the total power data table that is closest to the value of that set of measured data is selected.

[0120] The original data that has been filtered out is placed at the front of the table and then deleted. The measured data is then inserted at the back of the table to complete the queued update. The following is an example of a queued update, referring to Table 1: Assume a set of real-time collected data is X3 = Q1 + x, X1 = 29℃, X2 = 7℃, P1 + P2 = P a The data in row n is calculated to be the closest to the corresponding data in row n of the table. Therefore, the data in row n constitutes the "original data closest to the measured data value in this set," and the data in row n is deleted. The measured data set is X3 = Q1 + x, X1 = 29℃, X2 = 7℃, P1 + P2 = P... aAdd it to the table.

[0121] Furthermore, in some embodiments, various characteristic parameters include chilled water supply temperature, cooling water return temperature, and total system load. "For each set of collected measured data, the original data in the total power data table that is closest to the value of that set of measured data" specifically includes:

[0122] In the total power data table, the data is filtered layer by layer in the order of the system total load closest, the chilled water supply temperature closest, and the cooling water return temperature closest, until the filtering result is a unique set. The original data of this set is used as the original data that is closest to the measured data value of this set.

[0123] Specifically, first filter the total power data table to find the system total load X3 that is closest to the measured data of that set. If the result of the filtering is a set, then the set of data is the "original data that is closest to the measured data value of that set", and the filtering is completed.

[0124] If the results are multiple sets, continue to filter the results to find the data whose chilled water supply temperature is closest to the measured data of that set. If the results are only one set, then that set of data is the "original data that is closest to the measured data value of that set", and the filtering is complete.

[0125] If multiple sets of results are obtained, continue to filter the results and select the set of data that is closest to the cooling water return temperature of the set of measured data. This set of data is the "original data that is closest to the value of the set of measured data". The filtering is then complete.

[0126] In other embodiments, the total power data table can also be updated using methods such as sliding window update or stacked update, which will not be described in detail here.

[0127] In some embodiments, the total power model is generated using a recursive least squares method with a variable forgetting factor. The steps for generating the total power model using the recursive least squares method with a variable forgetting factor are described below, taking various characteristic parameters including chilled water supply temperature, cooling water return temperature, and total system load as examples.

[0128] First, the total power model is constructed as follows:

[0129]

[0130]

[0131]

[0132] Where y is the total power, c0~c5 are model parameters, and X1, X2, and X3 are the cooling water return temperature, chilled water supply temperature, and total system load, respectively.

[0133] Calculated using the recursive least squares method with a variable forgetting factor Specifically, it includes:

[0134] The total power data table values ​​are recorded sequentially in ascending order of total system load as follows:

[0135]

[0136] Determine the loss function:

[0137]

[0138] Where m is the number of total power data in the total power data table, which is also the number of rows in Table 1, and λ is the forgetting factor, 0<λ≤1. The value of λ is dynamically adjusted according to the model error to improve the model accuracy and tracking capability.

[0139] Determine using the following steps To minimize the loss function J;

[0140] Step A, Parameter Initialization: Set initial values ​​for the following parameters;

[0141]

[0142] Specifically, the initial values ​​are as follows:

[0143]

[0144] Step B: Sample the current input and output data:

[0145]

[0146] In the first iteration, k = 1.

[0147] Step C: Update the gain matrix according to the following formula:

[0148]

[0149] Step D: Update the parameters according to the following formula:

[0150]

[0151] Step E: Update the covariance according to the following formula:

[0152]

[0153] Step F: Perform error verification according to the following formula:

[0154]

[0155] Step G: Update the forgetting factor according to the following formula:

[0156] λ = 1 - α × |tanh(ε(k))|, where 0 < α ≤ 0.1, and is set according to the site conditions.

[0157] Increment the value of k by 1, and return to execute steps B through G until k = m.

[0158] In the control method of this invention, the data in the total power data table is updated using a queue-style update method, and the total power model is generated using a variable forgetting factor recursive least squares method. The introduction of the forgetting factor gives the newly added data to the table a larger weight value when training parameters, thereby improving the model accuracy and tracking ability, enabling the algorithm to run stably for a long time and making it more practical.

[0159] In some embodiments, the air conditioning control method calculates the single-unit power model through the following steps:

[0160] The database contains records of multiple sets of operating parameters for each chiller unit.

[0161] Based on the original database of each chiller unit, a single-unit power model of the chiller unit is trained and used to calculate the single-unit power of the chiller unit.

[0162] Specifically, the data sources for the original database may include selection data stored by the manufacturer in the selection software, laboratory test data of the chiller units, and operating data of the same model of chiller units installed at other sites.

[0163] For example, the data in the original database may include cooling water return temperature, chilled water supply temperature, unit load, and unit power. "Training the unit power model for each chiller based on its original database" specifically includes:

[0164] The single-machine power model is constructed as follows:

[0165]

[0166] Among them, P i Let a be the single-unit power of the i-th chiller unit. 0，i ~a 5，i Let X be the model parameters for the i-th chiller unit; 1，i X 2，i X 3，i These are the cooling water return temperature, chilled water supply temperature, and single unit load of the i-th chiller unit, respectively.

[0167] Based on the data from the original database, a was obtained through data fitting. 0，i ~a 5，iThe value can be determined using methods such as least squares, polynomial models, and neural network models.

[0168] In some embodiments, the aforementioned multiple characteristic parameters include the total system load X3. The input parameters for each unit power model include the unit load X of the corresponding chiller unit. 3i In each boot combination, n represents the number of chiller units in this start-up configuration. That is, when multiple chiller units are started simultaneously, the total load is distributed equally among each chiller unit.

[0169] In some embodiments, the aforementioned multiple characteristic parameters include the total system load X3. In the total power data table calculated based on a preset single-unit power model, the total system load X3 takes multiple values ​​at equal intervals between the minimum and maximum allowable loads of all chillers in the corresponding operating combination, preferably at equal intervals. The minimum allowable load is the load that all chillers can bear when operating at their lowest capacity in the operating combination, and the maximum allowable load is the maximum load that all chillers can bear. For example, Q1 in Table 1 is taken as the minimum allowable load, Q... n The value is taken as the maximum allowable load. Q1, Q2, up to Q... n In this context, the difference between two adjacent values ​​is equal.

[0170] The maximum allowable load is the sum of the maximum loads of all individual chiller units, and the minimum allowable load is t times the maximum value among the maximum loads of each individual unit, where t < 1. Each chiller unit has its own t value; for example, t can be 1 / 10.

[0171] In some embodiments, the control method further includes:

[0172] The database contains records of multiple sets of operating parameters for each chiller unit.

[0173] Based on the original database of each chiller unit, a single-unit maximum load model for that chiller unit is trained and used to calculate the single-unit maximum load of that chiller unit.

[0174] Specifically, the original database also includes the load rate of the chiller units;

[0175] "Based on the original database of each chiller unit, the single-unit maximum load model of that chiller unit is trained" includes:

[0176] The maximum load model for a single machine is constructed as follows:

[0177] Q max，i (X 1，i X 2，i )=b 0，i +b 1，i ×(X 1，i-X 2，i )+b 2，i ×(X 1，i -X 2，i ) 2 +b 3，i ×X 1，i +b 4，i ×X 2，i ;

[0178] Among them, Q max，i b is the maximum single-unit load of the i-th chiller unit. 0，i ~b 4，i Let X be the model parameters of the i-th chiller unit. 1，i X 2，i These are the cooling water return temperature and chilled water supply temperature of the i-th chiller unit when the load rate is 100%, respectively.

[0179] Based on the data from the original database, b was obtained through data fitting. 0，i →b 4，i The value can be determined using methods such as least squares, polynomial models, and neural network models.

[0180] Figure 2 This is a flowchart illustrating a control method for an air conditioning system according to an embodiment of the present invention.

[0181] In some optional embodiments, higher technical effects can be achieved by further optimizing and configuring the above steps. The control method of the present invention will be described in detail below with reference to an optional execution flow of the present invention. This embodiment is only an example of the execution flow. In specific implementation, the execution order and running conditions of some steps can be modified according to specific implementation requirements.

[0182] like Figure 2 As shown, the air conditioner control method of a preferred embodiment of the present invention performs the following steps in sequence:

[0183] Step S202: Obtain the original database containing multiple sets of operating parameters for each chiller unit.

[0184] Step S204: Based on the original database of each chiller unit, train its single-unit power model and single-unit maximum load model.

[0185] Step S206: Determine all possible start-up combinations for single or multiple chiller units.

[0186] Step S208: For each start-up combination, calculate the maximum load of each chiller unit based on the single unit maximum load model, sum them up to obtain the maximum allowable total load, and calculate the minimum allowable total load accordingly.

[0187] Step S210: For each start-up combination mode, plan multiple combinations of values ​​for various characteristic parameters, input the data of each value combination mode into the power model of each unit, obtain multiple unit power values, and sum them to obtain the total power, so as to obtain the total power data table of the start-up combination mode. In the table, the total system load takes multiple values ​​at equal intervals between the minimum allowable medium load and the maximum allowable total load of all chillers in the corresponding start-up combination mode.

[0188] Step S212: Obtain the measured values ​​of various characteristic parameters and total power collected in real time, and update the total power data table in a queue-style update manner.

[0189] Step S214: For each power-on combination, construct the total power model using the recursive least squares method with a variable forgetting factor based on its total power data table.

[0190] Step S216: Input the current values ​​of various characteristic parameters into the total power model of each power-on combination, calculate the corresponding total power, and determine the power-on combination with the minimum total power as the target power-on combination.

[0191] Step S218: Control the air conditioning system to operate in the target start-up combination mode.

[0192] Further details of the above steps have been described above and will not be repeated here.

[0193] On the other hand, embodiments of the present invention also provide a control device for an air conditioning system.

[0194] Figure 3 This is a schematic diagram of a control device for an air conditioning system according to an embodiment of the present invention.

[0195] like Figure 3 As shown, the control device 800 includes a processor 810 and a memory 820. The memory 820 stores a computer program 821. When the processor 810 executes the computer program 821, it is used to implement the air conditioner control method of any of the following embodiments of the present invention. The control device 800 may be a component of the air conditioning system or a separate device independent of the air conditioning system.

[0196] The processor 810 can be a central processing unit (CPU), a digital processing unit, etc. The processor 810 sends and receives data via a communication interface. The memory 820 stores the program executed by the processor 810. The memory 820 can be any medium capable of carrying or storing desired program code in the form of instructions or data structures, accessible by a computer, or a combination of multiple memories. The computer program 821 can be downloaded from a computer-readable storage medium to the corresponding computing / processing device or downloaded and installed to the control device 800 via a network (e.g., the Internet, a local area network, a wide area network, and / or a wireless network). The computer program 821 can be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages.

[0197] On the other hand, embodiments of the present invention also provide an air conditioning system.

[0198] Figure 4 This is a schematic diagram of an air conditioning system according to an embodiment of the present invention.

[0199] like Figure 4 As shown, the air conditioning system of this embodiment includes multiple chiller units A, B...C and a control device 800. Furthermore, the air conditioning system may also include multiple air conditioning terminals, water pumps, piping, and other components, which are well known to those skilled in the art and will not be described in detail here.

[0200] On the other hand, embodiments of the present invention also provide a computer-readable storage medium.

[0201] Figure 5 This is a schematic block diagram of a computer-readable storage medium according to an embodiment of the present invention.

[0202] like Figure 5 As shown, the computer-readable storage medium 700 of this embodiment stores a computer program 821, and when the computer program 821 is executed, it causes the device where the computer-readable storage medium 700 is located to perform the air conditioning control method of any embodiment of this invention.

[0203] The computer-readable storage medium 700 may be an electronic memory such as flash memory, EEPROM (Electrically Erasable Programmable Read-Only Memory), EPROM, hard disk, or ROM. The computer-readable storage medium 700 has storage space for a computer program 821 for performing any method steps of the air conditioner control method of the embodiments of the present invention. These computer programs 821 can be read from or written to one or more computer program products. These computer program products include program code carriers such as hard disks, CDs, memory cards, or floppy disks. When the device containing the computer-readable storage medium 700 executes the aforementioned computer program 821, it can perform the various steps of the method described in the present invention.

[0204] On the other hand, embodiments of the present invention also provide a computer program product.

[0205] Figure 6 This is a schematic block diagram of a computer program product according to an embodiment of the present invention.

[0206] like Figure 5 As shown, the computer program product 600 of this embodiment includes a computer program 821, and when the computer program 821 is run, it causes the device containing the computer-readable storage medium 700 to execute the air conditioning control method of any embodiment of this invention.

[0207] Therefore, those skilled in the art should recognize that although numerous exemplary embodiments of the present invention have been shown and described in detail herein, many other variations or modifications conforming to the principles of the present invention can be directly determined or derived from the disclosure of the present invention without departing from the spirit and scope of the invention. Thus, the scope of the present invention should be understood and construed as covering all such other variations or modifications.

Claims

1. A control method for an air conditioning system, the air conditioning system comprising multiple chiller units, the control method comprising: The total number of chiller units determines all possible start-up combinations for single or multiple chiller unit operation. For each of the aforementioned start-up combinations, multiple value combination schemes of various characteristic parameters are planned and preset. The data of each value combination scheme is input into the single-unit power model of each chiller unit to obtain multiple single-unit power and sum them to obtain the total power, so as to obtain the total power data table of the start-up combination. The system acquires the real-time collected multiple characteristic parameters and the measured value of the total power, and updates the total power data table accordingly. For each of the aforementioned power-on combinations, a total power model is constructed based on the total power data table, using the various characteristic parameters as input parameters and the total power as the output parameter. The current values ​​of the various feature parameters are input into the total power model of each of the power-on combinations to calculate the corresponding total power, and the power-on combination with the minimum total power is determined as the target power-on combination. Control the air conditioning system to operate in the target start-up combination mode; The phrase "acquiring the real-time collected multiple characteristic parameters and the measured value of the total power, and updating the total power data table accordingly" includes: For each set of measured data collected, the set of original data that is closest to the value of that set of measured data in the total power data table is selected. The original data that has been filtered out is placed at the front of the table and then deleted. The actual test data is then inserted at the back of the table to complete the queue-style update. The various characteristic parameters include chilled water supply temperature, cooling water return temperature, and total system load; "For each set of measured data collected, the original data in the total power data table that is closest to the value of that set of measured data is selected," including: In the total power data table, the data is filtered layer by layer in the order of the system total load closest, the chilled water supply temperature closest, and the cooling water return temperature closest, until the filtering result is a unique set. The original data of this set is used as the original data that is closest to the measured data value of this set.

2. The control method according to claim 1, wherein The various characteristic parameters include chilled water supply temperature, cooling water return temperature, and total system load; "For each of the aforementioned power-on combinations, constructing a total power model based on its total power data table, using the various characteristic parameters as input parameters and the total power as the output parameter," includes: The total power model is constructed as follows: ； ； ； Where y is the total power. For the model parameters, X1, X2, and X3 are the cooling water return temperature, chilled water supply temperature, and total system load, respectively. Calculated using the recursive least squares method with a variable forgetting factor .

3. The control method according to claim 2, wherein "the variable forgetting factor is calculated using the recursive least squares method" "include: The total power data table values ​​are recorded sequentially in ascending order of total system load as follows: ； Determine the loss function: ； Where m is the number of total power values ​​in the total power data table, and λ is the forgetting factor, 0 < λ ≤ 1; Determine using the following steps To minimize the loss function J; Step A, Parameter Initialization: Set initial values ​​for the following parameters; ； ； Step B, sample the current input and output data: ， In the first iteration, k=1; Step C: Update the gain matrix according to the following formula: ； Step D: Update the parameters according to the following formula: ； Step E: Update the covariance according to the following formula: ； Step F: Perform error verification according to the following formula: ； Step G: Update the forgetting factor according to the following formula: , where 0 < α ≤ 0.1; Increment the value of k by 1, and return to execute steps B through G until k = m.

4. The control method according to claim 1 further includes: The database contains records of multiple sets of operating parameters for each chiller unit. Based on the original database for each chiller unit, a single-unit power model for that chiller unit is trained to calculate its single-unit power.

5. The control method according to claim 4, wherein The original database includes cooling water return temperature, chilled water supply temperature, single unit load, and single unit power. "Based on the original database of each chiller unit, the single-unit power model of that chiller unit is trained" includes: The single-machine power model is constructed as follows: ； in, Let be the single-unit power of the i-th chiller unit. Let be the model parameters for the i-th chiller unit; These are the cooling water return temperature, chilled water supply temperature, and single unit load of the i-th chiller unit, respectively. Based on the data from the original database, the following method was used to obtain... The value of .

6. The control method according to claim 4, wherein The various characteristic parameters include total system load. ; The input parameters for each of the single-unit power models include the single-unit load of the corresponding chiller unit. ; In each of the aforementioned power-on combinations, n represents the number of chiller units in this start-up combination.

7. The control method according to claim 1, wherein The various characteristic parameters include total system load. ; In the total power data table calculated based on the preset single-unit power model, the total system load is... In the corresponding start-up combination, multiple values ​​are taken at equal intervals between the minimum allowable load and the maximum allowable load of all chiller units; The maximum allowable load is the sum of the maximum loads of all individual chiller units, and the minimum allowable load is t times the maximum value among the maximum loads of each individual unit, where t < 1.

8. The control method according to claim 7, further comprising: The database contains records of multiple sets of operating parameters for each chiller unit. Based on the original database of each chiller unit, a single-unit maximum load model of the chiller unit is trained and used to calculate the single-unit maximum load of the chiller unit.

9. The control method according to claim 8, wherein The original database also includes the load rate of the chiller units; "Based on the original database for each chiller unit, a single-unit maximum load model for that chiller unit is trained," including: The single-machine maximum load model is constructed as follows: ； in, The maximum load of the i-th chiller unit is given by [the maximum load of the unit]. Let the model parameters be those of the i-th chiller unit. These are the cooling water return temperature and chilled water supply temperature of the i-th chiller unit when the load rate is 100%, respectively. Based on the data from the original database, the following method was used to obtain... The value of .

10. The control method according to claim 1, wherein The various characteristic parameters include chilled water supply temperature, cooling water return temperature, and total system load; or The various characteristic parameters include chilled water supply temperature, cooling water outlet temperature, and total system load.

11. The control method according to claim 10, wherein The various characteristic parameters also include chilled water flow rate and / or cooling water flow rate.

12. A control device comprising a memory and a processor, the memory storing a computer program which, when executed by the processor, is used to implement the control method according to any one of claims 1 to 11.

13. An air conditioning system comprising a plurality of chiller units and a control device as described in claim 12.

14. A computer-readable storage medium storing a computer program that, when executed by a processor, implements the control method according to any one of claims 1 to 11.

15. A computer program product comprising a computer program that, when executed by a processor, implements the control method according to any one of claims 1 to 11.

Images