Optimization method for improving operating efficiency of refrigeration system based on digital twinning technology

Abstract

The application discloses an optimization method for improving the operation efficiency of a refrigeration system based on digital twin technology. It relates to the field of air conditioning technology. Through preliminary screening of the cold machine combination, only key parameters are selected as input, and interference factors are excluded, which significantly improves the model prediction accuracy. The energy efficiency ratio and current ratio of the preliminary screened cold machine combination are predicted using the model, providing a scientific basis for subsequent rounds of optimization. Through two iterations of optimization, the efficiency potential of the cold machine is fully tapped, and a higher energy efficiency cold machine combination is found. Through double stability judgment rules, the optimal cold machine combination is further selected, ensuring the stability of the refrigeration system during start-stop process, and balancing stability and energy efficiency. Through dynamic calculation of current ratio limit, the current ratio limit can be forcibly set under the unique working condition of the refrigeration system start-up stage, significantly improving the energy saving effect. The technical problems of single and extensive optimization method of refrigeration system control mode and inability to balance energy efficiency and stability in the prior art are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of air conditioning technology, in particular to an optimization method for improving the operating efficiency of a refrigeration system based on digital twin technology. BACKGROUND

[0002] A common refrigeration system is usually composed of a refrigeration host (hereinafter referred to as "cold machine"), a chilled water pump, a cooling water pump, a cooling tower, etc. The cold machine, as the core equipment of the refrigeration system, its operating efficiency directly affects the energy efficiency performance of the whole system. The operating efficiency of the cold machine is usually measured by the coefficient of performance (COP), i.e. the ratio of the cooling capacity output by the cold machine to the power consumption. Since the power consumption of the cold machine accounts for 60%~90% of the total power consumption of the refrigeration system, optimizing the operating efficiency of the cold machine is the key to improving the overall energy efficiency of the refrigeration system.

[0003] The operating efficiency of the cold machine is affected by many working parameters, mainly including load rate, supply water temperature and cooling water temperature, etc. Under a given refrigeration demand, the load rate and supply water temperature of the cold machine are the key parameters that can be controlled. By reasonably adjusting these parameters, the operating efficiency of the cold machine can be effectively improved, thereby improving the energy efficiency of the whole refrigeration system. However, the existing technology has the following defects in the optimization of the operating efficiency of the cold machine:

[0004] (1) The optimization method is single and extensive: the existing technology usually adopts a feedback control strategy based on current ratio to adjust the operating state of the cold machine. Specifically, when the current ratio of the cold machine in operation is higher than a certain set value (such as 95%), the system will start a new cold machine; on the contrary, when the current ratio is lower than a certain set value (such as 60%), the system will reduce the operation of a cold machine. Although this control method can adjust the load rate of the cold machine to a certain extent, its control range is extensive, and it can only avoid the cold machine running in the most unfavorable working condition, but it cannot achieve fine energy efficiency optimization.

[0005] (2) Stability and energy efficiency are difficult to balance: the existing technology usually only controls the air conditioning equipment of the air conditioning system according to the energy efficiency ratio predicted by the model in the control of the number of cold machine start and stop. This control strategy is easy to cause stability problems in the start and stop process of the refrigeration system, and it is difficult to maximize the energy efficiency while ensuring stability.

[0006] In view of the above technical problems in the prior art that the optimization method of the refrigeration system control mode is single and extensive and cannot balance the energy efficiency and stability, no effective solution has been proposed so far. SUMMARY

[0007] Embodiments of the present disclosure provide an optimization method for improving the operation efficiency of a refrigeration system based on digital twin technology. At least the technical problems of the optimization method of the existing refrigeration system control mode being single and extensive and being unable to balance energy efficiency and stability are solved.

[0008] According to an aspect of an embodiment of the present disclosure, an optimization method for improving the operation efficiency of a refrigeration system based on digital twin technology is provided, including determining a current cooling load demand, screening m initial cooling unit combinations that meet the current cooling load demand from all cooling unit combinations of the refrigeration system according to the maximum cooling capacity and the minimum cooling capacity of each cooling machine in the refrigeration system, and performing initial decomposition on the current cooling load demand to determine the initial cooling load of each cooling machine in each initial cooling unit combination; wherein the principle of the initial decomposition is that the load rate of each cooling machine is the same; inputting the water supply temperature, the cooling water temperature and the initial cooling load under the current operation condition of each cooling machine in each initial cooling unit combination into a pre-trained digital twin model to output the energy efficiency ratio and the current ratio of the corresponding cooling machine under the current operation condition; determining the overall energy efficiency ratio of each initial cooling unit combination according to the current cooling load demand and the energy efficiency ratio of each cooling machine in each initial cooling unit combination; screening k candidate cooling unit combinations from the m initial cooling unit combinations according to the overall energy efficiency ratio of each initial cooling unit combination; using a preset input parameter adjustment rule to respectively perform two iterations of evaluation on each candidate cooling unit combination by using the digital twin model to determine the optimal overall energy efficiency ratio of each candidate cooling unit combination; wherein the input parameter adjustment rule is used to adjust the cooling load, the water supply temperature and the cooling water temperature of each cooling machine in each candidate cooling unit combination; screening d selectable cooling unit combinations from the k candidate cooling unit combinations according to the optimal overall energy efficiency ratio of each candidate cooling unit combination; performing stability judgment on each selectable cooling unit combination according to the number of start-stop cooling machines in each selectable cooling unit combination based on a preset first stability judgment rule and a second stability judgment rule to obtain a first stability judgment result and a second stability judgment result; determining b cooling unit combinations that meet the stability requirement from the d selectable cooling unit combinations according to the first stability judgment result and the second stability judgment result, and taking the cooling unit combination with the highest overall energy efficiency ratio in the b cooling unit combinations as the optimal cooling unit combination; and controlling the start-stop of each cooling machine in the refrigeration system and limiting the current ratio parameter of the started cooling machine according to the optimal cooling unit combination and the corresponding current ratio output by the digital twin model.

[0009] According to another aspect of an embodiment of the present disclosure, a storage medium is also provided, which includes a stored program, wherein the program is executed by a processor when the program is running to perform the above-mentioned method.

[0010] According to another aspect of the embodiments of the present disclosure, an optimization device for improving the operation efficiency of a refrigeration system based on digital twin technology is also provided, comprising: a first screening module configured to determine a current cooling load demand, screen m initial cooling unit combinations that meet the current cooling load demand from all cooling unit combinations of the refrigeration system according to the maximum cooling capacity and the minimum cooling capacity of each cooling machine in the refrigeration system, and initially decompose the current cooling load demand to determine the initial cooling load of each cooling machine in each initial cooling unit combination; wherein the principle of the initial decomposition is that the load rate of each cooling machine is the same; an energy efficiency ratio determination module configured to input the water supply temperature, the cooling water temperature and the initial cooling load under the current operating condition into a pre-trained digital twin model for each cooling machine in each initial cooling unit combination, output the energy efficiency ratio and the current ratio of the corresponding cooling machine under the current operating condition, and determine the overall energy efficiency ratio of each initial cooling unit combination according to the current cooling load demand and the energy efficiency ratio of each cooling machine in each initial cooling unit combination; a second screening module configured to screen k candidate cooling unit combinations from the m initial cooling unit combinations according to the overall energy efficiency ratio of each initial cooling unit combination; use a preset input parameter adjustment rule to respectively perform two iterations of evaluation on each candidate cooling unit combination by using the digital twin model, and determine the optimal overall energy efficiency ratio of each candidate cooling unit combination; wherein the input parameter adjustment rule is used to adjust the cooling load, the water supply temperature and the cooling water temperature of each cooling machine in each candidate cooling unit combination; a third screening module configured to screen d selectable cooling unit combinations from the k candidate cooling unit combinations according to the optimal overall energy efficiency ratio of each candidate cooling unit combination; perform stability judgment on each selectable cooling unit combination according to the number of start-stop cooling machines in each selectable cooling unit combination based on a preset first stability judgment rule and a second stability judgment rule, and obtain a first stability judgment result and a second stability judgment result; determine b cooling unit combinations that meet the stability requirement from the d selectable cooling unit combinations according to the first stability judgment result and the second stability judgment result, and take the cooling unit combination with the highest overall energy efficiency ratio in the b cooling unit combinations as the optimal cooling unit combination; and an adjustment control module configured to control the start-stop of each cooling machine in the refrigeration system and limit the current ratio parameter of the started cooling machine according to the optimal cooling unit combination and the corresponding current ratio output by the digital twin model.

[0011] According to another aspect of the embodiments of the present disclosure, an optimization device for improving the operation efficiency of a refrigeration system based on digital twin technology is also provided, comprising: a processor; and a memory connected with the processor, used to provide the processor with instructions for processing the following processing steps: determining a current cooling load demand, screening m initial chiller combinations from all chiller combinations of the refrigeration system according to the maximum cooling capacity and the minimum cooling capacity of each chiller in the refrigeration system, and performing initial decomposition on the current cooling load demand to determine the initial cooling load of each chiller in each initial chiller combination; wherein the principle of the initial decomposition is that the load rate of each chiller is the same; inputting the water supply temperature, the cooling water temperature and the initial cooling load under the current operating condition of each chiller in each initial chiller combination into a pre-trained digital twin model to output the energy efficiency ratio and the current ratio of the corresponding chiller under the current operating condition; determining the overall energy efficiency ratio of each initial chiller combination according to the current cooling load demand and the energy efficiency ratio of each chiller in each initial chiller combination; screening k candidate chiller combinations from the m initial chiller combinations according to the overall energy efficiency ratio of each initial chiller combination; using a pre-set input parameter adjustment rule to respectively perform two iterations of evaluation on each candidate chiller combination by using the digital twin model to determine the optimal overall energy efficiency ratio of each candidate chiller combination; wherein the input parameter adjustment rule is used to adjust the cooling load, the water supply temperature and the cooling water temperature of each chiller in each candidate chiller combination; screening d selectable chiller combinations from the k candidate chiller combinations according to the optimal overall energy efficiency ratio of each candidate chiller combination; performing stability judgment on each selectable chiller combination based on a pre-set first stability judgment rule and a second stability judgment rule according to the number of started and stopped chillers in each selectable chiller combination to obtain first stability judgment results and second stability judgment results; determining b chiller combinations that meet the stability requirement from the d selectable chiller combinations according to the first stability judgment results and the second stability judgment results, and taking the chiller combination with the highest overall energy efficiency ratio from the b chiller combinations as the optimal chiller combination; and controlling the start and stop of each chiller in the refrigeration system and limiting the current ratio parameter of the started chiller according to the optimal chiller combination and the corresponding current ratio output by the digital twin model.

[0012] The technical solution of the present application first determines the current cooling load demand, screens the initial chiller combinations (m) that meet the demand in combination with the cooling capacity range of each chiller in the refrigeration system, and performs initial decomposition of the total load to each chiller using the equal load rate allocation principle to realize coarse-grained optimization of the chiller combination and lay a foundation for subsequent fine-tuning.

[0013] Next, the water supply temperature, cooling water temperature, and initial cooling load are input into the pre-trained digital twin model, and the COP (coefficient of performance) and current ratio of each chiller are output, and the overall energy efficiency ratio of each preliminary chiller combination is calculated. By selecting only the key parameters (water supply temperature, cooling water temperature, and load) as input, the model prediction accuracy is significantly improved, and the chiller efficiency is predicted based on real-time operating conditions, replacing the existing technology that relies only on the static threshold of current ratio, providing a scientific basis for subsequent rounds of optimization.

[0014] Further, k high-energy-efficiency candidate combinations are selected from the preliminary chiller combinations, and two iterations of evaluation are performed using input parameter adjustment rules (adjusting load, water supply temperature, and cooling water temperature) to obtain the optimal overall energy efficiency ratio. Through multi-parameter coordinated adjustment and two iterations of optimization, the limitations of existing technologies that only regulate load rate are overcome, and the efficiency potential of chillers is fully tapped. Through two iterations of optimization, higher-energy-efficiency chiller combinations are found, solving the problem of single regulation that cannot cover the optimal solution in existing technologies.

[0015] Subsequently, d selectable combinations are selected from the candidate combinations, and the number of start-stop units is combined to determine stability through a first stability determination rule (current operating condition stability) and a second stability determination rule (future load prediction stability), select b chiller combinations that meet the stability requirements, and the chiller combination with the highest overall energy efficiency ratio among the b chiller combinations is selected as the optimal chiller combination. Through the double stability determination rules, the refrigeration system remains stable during the start-stop process, balancing stability and energy efficiency, and avoiding the problem of oscillation caused by frequent switching.

[0016] Finally, according to the optimal chiller combination and the corresponding current ratio output by the digital twin model, the start-stop of each chiller in the refrigeration system is controlled, and the current ratio parameter of the started chiller is limited. By dynamically calculating the current ratio limit, the current ratio limit can be forcibly set under the unique operating conditions of the refrigeration system during the station opening stage, significantly improving energy saving effect. Thus, the technical problems of single and extensive optimization method of refrigeration system control mode and inability to balance energy efficiency and stability in existing technologies are solved. BRIEF DESCRIPTION OF DRAWINGS

[0017] The drawings described herein are used to provide further understanding of the present disclosure, and form a part of the present application. The illustrative embodiments of the present disclosure and their descriptions are used to explain the present disclosure, and do not constitute an improper limitation on the present disclosure. In the drawings:

[0018] Figure 1 is a flowchart of the optimization method for improving the operating efficiency of a refrigeration system based on digital twin technology according to embodiment 1 of the present disclosure;

[0019] Figure 2is a schematic diagram of an optimization device for improving the operating efficiency of a refrigeration system based on digital twin technology according to Embodiment 2 of the present disclosure; and

[0020] Figure 3 is a schematic diagram of an optimization device for improving the operating efficiency of a refrigeration system based on digital twin technology according to Embodiment 3 of the present disclosure. DETAILED DESCRIPTION

[0021] In order to enable persons skilled in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, not all the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by persons skilled in the art without creative labor should be within the scope of protection of the present disclosure.

[0022] It should be noted that the terms "first", "second", and the like in the specification and claims of the present disclosure and the above-described drawings are used to distinguish similar objects, and do not necessarily have to describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or units does not have to be limited to those steps or units clearly listed, but can include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0023] Embodiment 1: According to the present embodiment, a method embodiment of an optimization method for improving the operating efficiency of a refrigeration system based on digital twin technology is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described herein can be executed in an order different from that shown herein.

[0024] As shown in Figure 1 According to the present embodiment, an optimization method for improving the operating efficiency of a refrigeration system based on digital twin technology is provided, comprising:

[0025] S102: Determine the current cooling load demand, according to the maximum cooling capacity and the minimum cooling capacity of each chiller in the refrigeration system, select m initial chiller combinations from all chiller combinations in the refrigeration system that meet the current cooling load demand, and initially decompose the current cooling load demand to determine the initial cooling load of each chiller in each initial chiller combination; wherein the principle of initial decomposition is that the load rate of each chiller is the same;

[0026] Specifically, the current cooling load demand Q can be obtained by manual input, sensor collection or historical data prediction. Assuming that the refrigeration system has N chillers, the number of all chiller combinations of the refrigeration system is 2 N . According to the maximum cooling capacity and the minimum cooling capacity of each chiller, m initial chiller combinations that can meet the current cooling load demand Q are extracted from 2 N chiller combinations; the initial load decomposition is performed for each initial chiller combination, and the initial decomposition principle is that the load rate of each chiller is the same, so as to determine the initial cooling load of each chiller in each initial chiller combination. Specifically, for each initial chiller combination, according to the maximum cooling capacity and the minimum cooling capacity of each chiller, the total cooling capacity range of the combination is calculated:

[0027]

[0028] Wherein, Q min,j and Q max,j are the minimum cooling capacity and the maximum cooling capacity of the jth chiller in each chiller combination, and n is the number of chillers in the initial chiller combination.

[0029] The chiller combination of Q min ≤Q≤Q max is retained, and m initial chiller combinations are obtained.

[0030] Then, the total load is initially decomposed to each chiller by using the equal load rate allocation principle. That is, for the chillers in each initial chiller combination, the same load rate LR is set:

[0031]

[0032] Then the initial cooling load Q j of a single chiller is determined by the following calculation formula:

[0033] In this way, coarse-grained optimization of chiller combination can be realized, laying a foundation for subsequent fine adjustment.

[0034] S104: For each chiller in each preliminary chiller combination, input the water supply temperature, cooling water temperature and initial cooling load under the current operating condition into the pre-trained digital twin model, output the energy efficiency ratio and current ratio of the corresponding chiller under the current operating condition; determine the overall energy efficiency ratio of each preliminary chiller combination according to the current cooling load demand and the energy efficiency ratio of each chiller in each preliminary chiller combination.

[0035] Optionally, the digital twin model of the refrigeration system is trained by the following steps: setting the optimization operation boundary condition of the refrigeration system; wherein the optimization operation boundary condition is used to limit the fixed number of chillers, the decision cycle and the locked operation parameters during modeling of the refrigeration system; determining the key parameters of each chiller in the refrigeration system under different operating conditions based on the operating characteristic data of each chiller under different operating conditions; wherein the key parameters include chilled water supply temperature, cooling water outlet temperature, load rate and energy efficiency ratio; and based on the optimization operation boundary condition and the key parameters, constructing and training the digital twin model of the refrigeration system; wherein the task of the digital twin model is to predict the energy efficiency ratio and current ratio of a chiller under a certain operating condition according to the water supply temperature, cooling water temperature and cooling load of the chiller under the operating condition.

[0036] Optionally, the operation of determining the overall energy efficiency ratio of each preliminary chiller combination according to the current cooling load demand and the energy efficiency ratio of each chiller in each preliminary chiller combination includes: calculating the power consumption of each chiller in each preliminary chiller combination according to the current cooling load demand and the energy efficiency ratio of each chiller in each preliminary chiller combination; and determining the overall energy efficiency ratio of each preliminary chiller combination according to the current cooling load demand and the power consumption of each chiller in each preliminary chiller combination.

[0037] Specifically, for a specified refrigeration system, the number of chillers is determined. A decision cycle of the refrigeration system is about 15 minutes to 60 minutes, during which the required cooling load and cooling temperature of the terminal are determined, and the cooling water temperature delivered from the cooling tower to the chiller is determined. Therefore, the number of chillers, cooling load, cooling temperature and cooling water temperature are set as boundary conditions.

[0038] Then, the operating characteristic data of each chiller under different operating parameters are collected, including chilled water supply temperature, chilled water return temperature, chilled water flow, cooling water outlet temperature, cooling water return temperature, cooling load, load rate, electric power, energy efficiency ratio, current ratio, and rated parameters of the chiller, etc., which are substituted into the thermodynamic formula to calculate the chilled water supply temperature, cooling water outlet temperature, load rate, and energy efficiency ratio as key parameters for subsequent model training. The historical data are subjected to data management, including standardization conversion of measurement units, cleaning of mutation data, interpolation of missing data, and timestamp alignment of various parameters. The data after management are randomly grouped, of which 70% are used as training data and 30% as test data. For the training data, the MLP (multilayer perception regression model) algorithm is used for modeling, and the test data are used to verify the model. After multiple rounds of parameter adjustment, when the model prediction accuracy reaches the preset accuracy of 97% (which can be adjusted), the digital twin modeling is completed, and the digital twin model of the refrigeration system is obtained. The model can predict the energy efficiency ratio and current ratio according to the input operating condition parameters. During the subsequent continuous operation of the system, the input data and the output data accurately calculated during actual operation are regularly used to check the model prediction results, and when the deviation is higher than the preset value, the next round of modeling iteration optimization is entered.

[0039] Then, after the coarse-grained optimization of the chiller combination is completed through the above step S102, for each chiller in the i-th preliminary chiller combination, the chilled water supply temperature and cooling water temperature under the current operating condition, the decomposed cooling load Q j are brought into the digital twin model to obtain the energy efficiency ratio COP j , current ratio, and other parameters of the j-th chiller under the operating condition. Wherein, i = 1 ~ m. Then, the electric power P j of the j-th chiller is calculated according to the formula . The energy efficiency ratio of each chiller in the i-th preliminary chiller combination is subjected to overall energy efficiency analysis, and the overall energy efficiency ratio COP i of the i-th preliminary chiller combination is obtained through the formula .

[0040] Thus, by selecting only the key parameters (chilled water supply temperature, cooling water temperature, and load) as input and excluding interference factors, the model prediction accuracy is significantly improved, and the chiller energy efficiency is predicted based on real-time operating conditions, replacing the static threshold judgment of current ratio in the prior art, and providing a scientific basis for subsequent multiple rounds of optimization.

[0041] S106: selecting k candidate chiller combinations from m preliminary chiller combinations according to the overall energy efficiency ratio of each preliminary chiller combination; using a preset input parameter adjustment rule, performing two iterative evaluations on each candidate chiller combination by using the digital twin model to determine the optimal overall energy efficiency ratio of each candidate chiller combination; wherein the input parameter adjustment rule is used to adjust the cooling load, supply water temperature and cooling water temperature of each chiller in each candidate chiller combination;

[0042] Optionally, the operation of using the preset input parameter adjustment rule, performing two iterative evaluations on each candidate chiller combination by using the digital twin model to determine the optimal overall energy efficiency ratio of each candidate chiller combination comprises: under a preset first constraint condition, redistributing the cooling load of each chiller in each candidate chiller combination by adjusting the chilled water flow, inputting the supply water temperature, cooling water temperature and redistributed cooling load under the current operating condition into the digital twin model, outputting the energy efficiency ratio and current ratio of the corresponding chiller under the current operating condition, and determining the first overall energy efficiency ratio of each candidate chiller combination according to the current cooling load demand and the energy efficiency ratio of each chiller in each candidate chiller combination; wherein the first constraint condition is that the total cooling load of all chillers in the candidate chiller combination is equal to the current cooling load demand; under the first constraint condition and a preset second constraint condition, redistributing the cooling load of each chiller in each candidate chiller combination by adjusting the supply water temperature and chilled water flow of each chiller, changing the supply water temperature of each chiller, inputting the changed supply water temperature, cooling water temperature under the current operating condition and adjusted cooling load into the digital twin model, outputting the energy efficiency ratio and current ratio of the corresponding chiller under the current operating condition, and determining the second overall energy efficiency ratio of each candidate chiller combination according to the current cooling load demand and the energy efficiency ratio of each chiller in each candidate chiller combination; wherein the second constraint condition is that: ;T 总供水温度 is the total supply water temperature of the refrigeration system, T 总管回水温度 is the total return water temperature of the refrigeration system, Q is the current cooling load demand, Flow j is the chilled water flow of the jth chiller in each candidate chiller combination; and determining the optimal overall energy efficiency ratio of each candidate chiller combination based on the first overall energy efficiency ratio and the second overall energy efficiency ratio.

[0043] Specifically, k candidate chiller combinations are selected from m preliminary combinations according to the overall energy efficiency ratio of each preliminary chiller combination. The selection rule is, for example, to sort all preliminary chiller combinations according to the overall energy efficiency ratio, and retain the preliminary chiller combinations ranked in the top k positions, thereby obtaining the k candidate chiller combinations. Then, for each candidate chiller combination selected, two iterative evaluations are performed using a preset input parameter adjustment rule to further determine the optimal overall energy efficiency ratio of the candidate chiller combination. The input parameter adjustment rule is mainly used to adjust the cooling load, supply water temperature and cooling water temperature of each chiller. The following are the detailed steps of the two iterative evaluations:

[0044] (1) The first iterative evaluation determines the first overall energy efficiency ratio of each candidate chiller combination:

[0045] 1) Under the preset first constraint condition (i.e., the total cooling load of all chillers in the candidate chiller combination is equal to the current cooling load demand), the cooling load of each chiller is redistributed by adjusting the chilled water flow. The adjustment strategy is, for example, to preferentially increase the flow of high-efficiency chillers and reduce the flow of low-efficiency chillers, and to ensure that the load of each chiller is within the allowed range, i.e. min,j ≤Q j ≤Q max,j ;

[0046] 2) The supply water temperature, cooling water temperature and redistributed cooling load under the current operating condition are input into the digital twin model;

[0047] 3) The model outputs the energy efficiency ratio and current ratio of the corresponding chiller under the current operating condition;

[0048] 4) According to the current cooling load demand and the energy efficiency ratio of each chiller in each candidate chiller combination, the overall energy efficiency ratio of each candidate chiller combination is re-determined according to the corresponding process in step S104 (corresponding to the first overall energy efficiency ratio).

[0049] (2) The second iterative evaluation determines the second overall energy efficiency ratio of each candidate chiller combination:

[0050] 1) Under the first constraint condition and a preset second constraint condition, the cooling load is redistributed by adjusting the supply water temperature and chilled water flow of each chiller, and the supply water temperature is changed at the same time. The second constraint condition is: ;T 总供水温度 is the total supply water temperature of the refrigeration system, T 总管回水温度 is the total return water temperature of the refrigeration system, Q is the current cooling load demand, Flow j is the chilled water flow of the jth chiller in each candidate chiller combination.

[0051] wherein the adjustment strategy is: according to the target total water supply temperature Backstepping total chilled water flow rate required

[0052] wherein p is the density of water, and c is the specific heat capacity of water. Then, the chilled water flow rate of each chiller is allocated according to the total flow rate backstepped , and the corresponding water supply temperature is calculated by the following formula :

[0053] 2) input the changed water supply temperature, the cooling water temperature under the current operating condition, and the adjusted cooling load into the digital twin model;

[0054] 3) the model also outputs the energy efficiency ratio and the current ratio of the corresponding chiller under the current operating condition;

[0055] 4) according to the current cooling load demand and the energy efficiency ratio of each chiller in each candidate chiller combination, the overall energy efficiency ratio of each candidate chiller combination is re-determined according to the corresponding process in the above step S104 (corresponding to the second overall energy efficiency ratio).

[0056] (3) based on the first overall energy efficiency ratio and the second overall energy efficiency ratio, the optimal overall energy efficiency ratio of each candidate chiller combination is determined. For each candidate chiller combination, the maximum value is taken from the first overall energy efficiency ratio and the second overall energy efficiency ratio as the optimal overall energy efficiency ratio of the candidate chiller combination.

[0057] Thus, through multi-parameter coordinated adjustment and twice iterative optimization, the limitations of the prior art of only regulating the load rate are broken through, the efficiency potential of the chiller is fully tapped, and through twice iterative optimization, a chiller combination with higher energy efficiency is found, solving the problem that single regulation of the prior art cannot cover the optimal solution.

[0058] S108: according to the optimal overall energy efficiency ratio of each candidate chiller combination, d kinds of optional chiller combinations are selected from the k kinds of candidate chiller combinations; according to the number of start-stop chillers in each optional chiller combination, the stability of each optional chiller combination is judged based on the first stability judgment rule and the second stability judgment rule, to obtain the first stability judgment result and the second stability judgment result; according to the first stability judgment result and the second stability judgment result, b kinds of chiller combinations that meet the stability requirement are determined from the d kinds of optional chiller combinations, and the chiller combination with the highest overall energy efficiency ratio in the b kinds of chiller combinations is taken as the optimal chiller combination.

[0059] Alternatively, the first stability judgment rule is: comparing each optional chiller combination with the currently running chiller combination, and selecting the chiller combination from the d kinds of optional chiller combinations, whose difference in the number of start-stop chillers and the number of start-stop chillers of the currently running chiller combination is less than a first preset threshold.​

[0060] Optionally, the second stability determination rule is: determining a future cooling load demand of the refrigeration system in a future preset time period, selecting a plurality of target cold machine combinations satisfying the future cooling load demand from all cold machine combinations of the refrigeration system based on the future cooling load demand and the digital twin model, and comparing the d optional cold machine combinations with the plurality of target cold machine combinations respectively in pairs, and selecting a cold machine combination from the d optional cold machine combinations, whose difference in cold machine start-stop number and the cold machine start-stop number of the plurality of target cold machine combinations is less than a second preset threshold.

[0061] Specifically, the principle of screening d optional cold machine combinations from k candidate cold machine combinations is: taking the candidate cold machine combination with the highest energy efficiency as the benchmark, all combinations not lower than 98% of the energy efficiency of this combination (this value can be adjusted). Considering that the cold machine should not be frequently started and stopped, the developed cold machine combination should pursue high energy efficiency ratio while taking into account the stability of operation and reducing the start-stop number of the cold machine. Therefore, two stability determination screenings need to be performed on the d optional cold machine combinations screened in step S106, and the detailed process is as follows:

[0062] (1) First stability determination:

[0063] 1) Taking the currently running cold machine combination as the comparison target, the stability determination condition is that the change in the start-stop number of the candidate cold machine combination does not exceed a first preset threshold (for example, 1, which can be adjusted), so as to avoid frequent start-stop leading to equipment wear and system shock;

[0064] 2) Obtain the currently running cold machine combination, and calculate the difference between the number of cold machines in each optional cold machine combination and the number of cold machines in the currently running cold machine combination;

[0065] 3) Determine whether the difference is less than the first preset threshold, if yes, pass the first stability determination, otherwise, no.

[0066] (2) Second stability determination:

[0067] 1) Taking the cooling load demand in a future preset time period (such as the next cycle or the next 1 hour) as the prediction target, using a time series model (ARIMA / LSTM) to predict the future cooling load demand, and then using the same method as steps S102-S106 above to generate a plurality of target cold machine combinations satisfying the future cooling load demand;

[0068] 2) For each optional cold machine combination, calculate the difference in cold machine start-stop number between it and the plurality of target cold machine combinations;

[0069] 3) Determine whether the difference is less than a second preset threshold, if yes, pass the second stability determination, otherwise, no.

[0070] (3) Further screen the two stability judgment results to meet the best of two stability judgment conditions, and preferentially meet the first stability judgment condition as the screening principle, and determine b kinds of cold machine combinations meeting the stability requirement from the d kinds of selectable cold machine combinations under the premise of considering the operation stability.

[0071] Therefore, by the first stability judgment rule, the number of start-stop machines can be limited to avoid equipment impact and ensure the short-term stability of the refrigeration system. By the second stability judgment rule, the combination can ensure smooth transition under future load, ensuring the long-term adaptability of the refrigeration system. Therefore, the optimal combination considering efficiency and stability is finally screened out, providing reliable input for control instruction issuing (step S110), so that the refrigeration system remains stable during start-stop process, and the stability and energy efficiency are considered, avoiding the problem of frequent switching leading to shock.

[0072] S110: According to the optimal cold machine combination and the corresponding current ratio output by the digital twin model, control the start-stop of each cold machine in the refrigeration system and limit the current ratio parameters of the started cold machine.

[0073] Specifically, the refrigeration system usually needs to be started in advance before work to remove the high temperature load accumulated in the building at night, and at this time, the circulating water temperature (25°C~35°C) is much higher than the water temperature (about 12°C) during normal work, so the cold machine will work directly in the working condition of extremely high load rate, and various conventional cold machine adding and subtracting strategies are invalid. This working condition generally needs to run for 30 to 120 minutes before entering the normal working mode, which consumes a lot of energy. Through the above steps S102~S108, the optimal cold machine combination and the corresponding working current ratio parameters can be evaluated, and these current ratio parameters of the cold machine can be limited according to the optimization results, so as to ensure that the refrigeration system can still operate in the high energy efficiency working condition in this specific scenario.

[0074] Therefore, by dynamically calculating the current ratio limit value, the current ratio limit value can be forcibly set in the unique working condition of the refrigeration system during the station opening stage, significantly improving the energy saving effect.

[0075] As in the background art, the prior art has the following defects in cold machine operation efficiency optimization: (1) The optimization method is single and rough: the prior art usually adopts a feedback control strategy based on current ratio to adjust the running state of the cold machine. Specifically, when the current ratio of the running cold machine is higher than a certain set value (such as 95%), the system will start a new cold machine; on the contrary, when the current ratio is lower than a certain set value (such as 60%), the system will reduce the operation of a cold machine. Although this control method can adjust the load rate of the cold machine to a certain extent, its control range is relatively rough, and it can only avoid the cold machine running in the most unfavorable working condition, but cannot realize fine energy efficiency optimization. (2) Stability and energy efficiency are difficult to balance: the prior art usually only controls the air conditioning equipment of the air conditioning system according to the energy efficiency ratio predicted by the model. This control strategy is easy to cause stability problems in the start-stop process of the refrigeration system, and it is difficult to maximize energy efficiency while ensuring stability.

[0076] Therefore, the technical scheme of the present application first determines the current cooling load demand, combines the cooling capacity range of each cold machine in the refrigeration system, selects the initial cold machine combination (m kinds) that meets the demand, and uses the equal load rate allocation principle to initially decompose the total load to each cold machine, realizing the coarse-grained optimization of the cold machine combination and laying a foundation for subsequent fine adjustment.

[0077] Then, the water supply temperature, cooling water temperature and initial cooling load are input into the pre-trained digital twin model, and the COP (COP) and current ratio of each cold machine are output, and the overall energy efficiency ratio of each initial cold machine combination is calculated. By selecting only the key parameters (water supply temperature, cooling water temperature, load) as input, the model prediction accuracy is significantly improved, and the cold machine efficiency is predicted based on real-time working conditions, replacing the static threshold judgment of current ratio in the prior art, providing a scientific basis for subsequent multi-round optimization.

[0078] Further, k high-energy-efficiency candidate combinations are selected from the initial cold machine combination, and two iterations are evaluated using input parameter adjustment rules (adjusting load, water supply temperature, cooling water temperature) to obtain the optimal overall energy efficiency ratio. Through multi-parameter collaborative adjustment and two iterations of optimization, the limitations of the prior art of only adjusting the load rate are broken, the efficiency potential of the cold machine is fully tapped, and through two iterations of optimization, a cold machine combination with higher energy efficiency is found, solving the problem that the single control of the prior art cannot cover the optimal solution.

[0079] Subsequently, d optional combinations are screened from the candidate combinations, combined with the number of start-stop changes, the stability is judged through the first stability judgment rule (current working condition stability) and the second stability judgment rule (future load prediction stability), b cold machine combinations meeting the stability requirement are selected, and the cold machine combination with the highest overall energy efficiency ratio in the b cold machine combinations is taken as the optimal cold machine combination. Through the double stability judgment rules, the refrigeration system remains stable during the start-stop process, and the stability and energy efficiency are considered, avoiding the shock problem caused by frequent switching.

[0080] Finally, according to the optimal cold machine combination and the corresponding current ratio output by the digital twin model, the start-stop of each cold machine in the refrigeration system is controlled, and the current ratio parameter of the started cold machine is limited. By dynamically calculating the current ratio limit, the current ratio limit can be forcibly set under the unique working condition of the refrigeration system in the station opening stage, significantly improving the energy saving effect. Thus, the technical problems of single and extensive optimization method of refrigeration system control mode and inability to consider energy efficiency and stability in the prior art are solved.

[0081] Hereinafter, taking the refrigeration system of a certain large industrial project as an example, a specific application scenario of the embodiment is given:

[0082] The refrigeration system of a certain large industrial project has 10 refrigeration units, specifically 2 fixed-frequency units (rated refrigeration capacity 6822kW, rated COP 5.67, referred to as A-type units), 3 variable-frequency units (rated refrigeration capacity 8051kW, rated COP 5.0, referred to as B-type units), 4 variable-frequency units (rated refrigeration capacity 10021kW, rated COP 5.13, referred to as C-type units), and 1 variable-frequency unit (rated refrigeration capacity 8051kW, rated COP 6.58, referred to as D-type unit). The current operating condition is: the current cooling demand is 15000kW, the chilled water supply temperature is 7 degrees Celsius, and the cooling water outlet temperature is 33 degrees Celsius.

[0083] Under the current operating condition, the optimization process of the refrigeration system is as follows:

[0084] (1) Initial cold machine combination screening: according to the maximum cooling capacity and the minimum cooling capacity of each cold machine in the refrigeration system, m initial cold machine combinations meeting the current cooling load demand are screened from all cold machine combinations in the refrigeration system, and the number of start machines of each initial cold machine combination is 2, 3, or 4. Subsequently, each cold machine of each initial cold machine combination is decomposed, and the initial cooling load of each cold machine is determined;

[0085] (2) Primary evaluation of energy efficiency ratio: According to the pre-established digital twin model of the refrigeration system, the operating energy efficiency (i.e., overall energy efficiency ratio) of each primary selected chiller combination is obtained. The parameters of some primary selected chiller combinations are as follows: 1) 2 B-type units, single unit cold load allocation 7500 kW, load rate 93.2%, overall energy efficiency 5.96; 2) 1 A-type unit + 3 B-type units, single A-type unit cold load 3304 kW, single B-type unit cold load 3899 kW, load rate 48.4%, overall energy efficiency 5.96; 3) 3 C-type units, single unit cold load allocation 5000 kW, load rate 49.9%, overall energy efficiency 6.29; 4) 2 A-type units + 1 D-type unit, single A-type unit cold load 4717 kW, single D-type unit cold load 5566 kW, load rate 69.1%, overall energy efficiency 6.30; 5) 3 B-type units, single unit cold load allocation 5000 kW, load rate 62.1%, overall energy efficiency 6.31; and the like;

[0086] (3) Iterative evaluation of candidate chiller combinations: According to the overall energy efficiency ratio of each primary selected chiller combination, k candidate chiller combinations are selected from the m primary selected chiller combinations. Then, each candidate chiller combination is evaluated twice, and after adjusting the iteration parameters, the new operating conditions of each combination are obtained, and the digital twin model is used to evaluate the energy efficiency to select high-energy-efficiency combinations, and the optimal overall energy efficiency ratio of each candidate chiller combination is obtained. The parameters of the k candidate chiller combinations are as follows: 1) 1 B-type unit, single unit cold load 4834 kW, load rate 60.0%; 2 C-type units, single unit cold load 5083 kW, load rate 50.7%, overall energy efficiency 6.27; 2) 2 B-type units, single unit cold load 4350 kW, load rate 54.0%; 1 D-type unit, single unit cold load 6301 kW, load rate 78.2%, overall energy efficiency 6.40; 3) 1 D-type unit, single unit cold load 5589 kW, load rate 69.4%, chilled water supply temperature 5.5 degrees Celsius; 2 C-type units, single unit cold load 4706 kW, load rate 47.0%, chilled water supply temperature 7.7 degrees Celsius, overall energy efficiency 6.48;

[0087] (4) Selection of optional chiller combinations and stability judgment: According to the optimal overall energy efficiency ratio of each candidate chiller combination, d optional chiller combinations are selected from the k candidate chiller combinations. Then, the stability of each optional chiller combination is judged: compared with the currently running chiller combination, compared with the chiller combination running in the future for a period of time, and according to the set stability boundary conditions, the optimal or gradual switching is performed;

[0088] (5) Starting scene optimization adjustment: The refrigeration system needs to be started in advance before work to remove the high temperature load accumulated in the building at night. At this time, the circulating water temperature (25°C~35°C) is much higher than the normal water temperature (about 12°C), so the chiller will work directly in the condition of extremely high load rate, and the current ratio is near 100%. Through this method, the optimal chiller combination and corresponding working current ratio parameters (such as 50%~60%) can be evaluated, and the current ratio parameters of the chiller are limited, so that the refrigeration system can still run in high energy efficiency condition in this specific scene; and in the process of circulating water temperature transition to normal water temperature, multiple rounds of current ratio parameter optimization and reset are carried out.

[0089] Through the above optimization process, the refrigeration system can meet the current cooling demand while realizing efficient and stable operation and bringing significant energy-saving benefits (5%~15% energy saving).

[0090] In addition, according to the embodiment, a storage medium is also provided. The storage medium includes a stored program, wherein when the program is executed by a processor, the method described in any one of the above embodiments is executed.

[0091] It should be noted that for the above-mentioned method embodiments, in order to simply describe, they are all expressed as a series of action combinations, but those skilled in the art should know that the present application is not limited by the described action sequence, because according to the present application, certain steps can be performed in other order or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification all belong to preferred embodiments, and the actions and modules involved are not necessarily necessary for the present application.

[0092] Example 2: Figure 2A structure diagram of the optimization device 200 for improving the operation efficiency of the refrigeration system based on the digital twin technology according to the embodiment is shown. The optimization device 200 comprises: a first screening module 210, configured to determine a current cooling load demand, screen m initial cooling unit combinations satisfying the current cooling load demand from all cooling unit combinations of the refrigeration system according to the maximum cooling capacity and the minimum cooling capacity of each cooling machine in the refrigeration system, and initially decompose the current cooling load demand to determine the initial cooling load of each cooling machine in each initial cooling unit combination; wherein the principle of the initial decomposition is that the load rate of each cooling machine is the same; an energy efficiency ratio determination module 220, configured to input the water supply temperature, the cooling water temperature and the initial cooling load under the current operation condition into a pre-trained digital twin model for each cooling machine in each initial cooling unit combination, output the energy efficiency ratio and the current ratio of the corresponding cooling machine under the current operation condition, and determine the overall energy efficiency ratio of each initial cooling unit combination according to the current cooling load demand and the energy efficiency ratio of each cooling machine in each initial cooling unit combination; a second screening module 230, configured to screen k candidate cooling unit combinations from the m initial cooling unit combinations according to the overall energy efficiency ratio of each initial cooling unit combination; use a preset input parameter adjustment rule to respectively perform two iterations of evaluation on each candidate cooling unit combination by using the digital twin model, and determine the optimal overall energy efficiency ratio of each candidate cooling unit combination; wherein the input parameter adjustment rule is used to adjust the cooling load, the water supply temperature and the cooling water temperature of each cooling machine in each candidate cooling unit combination; a third screening module 240, configured to screen d selectable cooling unit combinations from the k candidate cooling unit combinations according to the optimal overall energy efficiency ratio of each candidate cooling unit combination; perform stability judgment on each selectable cooling unit combination according to the number of start-stop cooling machines in each selectable cooling unit combination based on a preset first stability judgment rule and a second stability judgment rule, and obtain first stability judgment results and second stability judgment results; determine b cooling unit combinations satisfying the stability requirement from the d selectable cooling unit combinations according to the first stability judgment results and the second stability judgment results, and take the cooling unit combination with the highest overall energy efficiency ratio in the b cooling unit combinations as the optimal cooling unit combination; and an adjustment control module 250, configured to control the start-stop of each cooling machine in the refrigeration system and limit the current ratio parameter of the started cooling machine according to the optimal cooling unit combination and the corresponding current ratio output by the digital twin model.

[0093] Optionally, the optimization device 200 further comprises a model building and training module for training the digital twin model of the refrigeration system by the following steps: setting an optimized operation boundary condition of the refrigeration system; wherein the optimized operation boundary condition is used to limit the fixed number of cold machines, the decision cycle and the locked operation parameters when the refrigeration system is modeled; determining the key parameters of each cold machine in the corresponding working condition based on the operation characteristic data of each cold machine in the refrigeration system under different operation conditions; wherein the key parameters include chilled water supply temperature, cooling water outlet temperature, load rate and energy efficiency ratio; and building and training the digital twin model of the refrigeration system based on the optimized operation boundary condition and the key parameters; wherein the task of the digital twin model is to predict the energy efficiency ratio and current ratio of a cold machine under a certain operation condition according to the supply water temperature, cooling water temperature and cooling load of the cold machine under the condition.

[0094] Therefore, according to the present embodiment, by selecting only the key parameters (supply water temperature, cooling water temperature, load) as input, excluding interference factors, the model prediction accuracy is significantly improved, and the cold machine efficiency is predicted based on real-time working conditions, replacing the static threshold judgment of current ratio in the prior art, providing a scientific basis for subsequent multi-round optimization. Through multi-parameter collaborative adjustment and twice iterative optimization, the limitations of only adjusting the load rate in the prior art are broken through, the efficiency potential of the cold machine is fully tapped, and through twice iterative optimization, a cold machine combination with higher energy efficiency is found, solving the problem that single adjustment in the prior art cannot cover the optimal solution. Through the double stability judgment rule, the refrigeration system remains stable during the start-stop process, balancing stability and energy efficiency, and avoiding the problem of shock caused by frequent switching. Through dynamic calculation of the current ratio limit value, the current ratio limit value can be forcibly set under the unique working condition of the refrigeration system in the station opening stage, significantly improving the energy saving effect. Therefore, the technical problems of single and extensive optimization method of the refrigeration system control mode and the inability to balance energy efficiency and stability in the prior art are solved.

[0095] Embodiment 3: The embodiment provides an optimization device for improving the operation efficiency of a refrigeration system based on digital twin technology, comprising: a processor 310; and a memory 320 connected with the processor 310, used to provide the processor 310 with instructions for processing the following processing steps: determining a current cooling load demand, screening m initial cooling unit combinations that meet the current cooling load demand from all cooling unit combinations of the refrigeration system according to the maximum cooling capacity and the minimum cooling capacity of each cooling machine in the refrigeration system, and initially decomposing the current cooling load demand to determine the initial cooling load of each cooling machine in each initial cooling unit combination; wherein the principle of the initial decomposition is that the load rate of each cooling machine is the same; for each cooling machine in each initial cooling unit combination, inputting the water supply temperature, the cooling water temperature and the initial cooling load under the current operation condition into a pre-trained digital twin model, and outputting the energy efficiency ratio and the current ratio of the corresponding cooling machine under the current operation condition; determining the overall energy efficiency ratio of each initial cooling unit combination according to the current cooling load demand and the energy efficiency ratio of each cooling machine in each initial cooling unit combination; screening k candidate cooling unit combinations from the m initial cooling unit combinations according to the overall energy efficiency ratio of each initial cooling unit combination; using a pre-set input parameter adjustment rule to respectively perform two iterations of evaluation on each candidate cooling unit combination by using the digital twin model to determine the optimal overall energy efficiency ratio of each candidate cooling unit combination; wherein the input parameter adjustment rule is used to adjust the cooling load, the water supply temperature and the cooling water temperature of each cooling machine in each candidate cooling unit combination; screening d selectable cooling unit combinations from the k candidate cooling unit combinations according to the optimal overall energy efficiency ratio of each candidate cooling unit combination; performing stability judgment on each selectable cooling unit combination according to the number of start-stop cooling machines in each selectable cooling unit combination based on a pre-set first stability judgment rule and a second stability judgment rule to obtain first stability judgment results and second stability judgment results; determining b cooling unit combinations that meet the stability requirement from the d selectable cooling unit combinations according to the first stability judgment results and the second stability judgment results, and taking the cooling unit combination with the highest overall energy efficiency ratio in the b cooling unit combinations as the optimal cooling unit combination; and controlling the start-stop of each cooling machine in the refrigeration system and limiting the current ratio parameter of the started cooling machine according to the optimal cooling unit combination and the corresponding current ratio output by the digital twin model.

[0096] Thus, according to the present embodiment, by selecting only key parameters (water supply temperature, cooling water temperature, load) as inputs, excluding interference factors, the model prediction accuracy is significantly improved, and the real-time working condition is predicted based on the cooling machine efficiency, replacing the static threshold judgment in the prior art which only relies on the current ratio, providing a scientific basis for subsequent multi-round optimization. Through multi-parameter collaborative adjustment and twice iterative optimization, the limitations of the prior art of only regulating the load rate are broken through, the cooling machine efficiency potential is fully tapped, and through twice iterative optimization, a higher energy-efficient cooling machine combination is found, solving the problem that the single regulation in the prior art cannot cover the optimal solution. Through the double stability judgment rule, the refrigeration system remains stable during the start-stop process, balancing stability and energy efficiency, and avoiding the problem of oscillation caused by frequent switching. Through dynamic calculation of the current ratio limit, the current ratio limit can be forcibly set under the unique working condition of the refrigeration system start-up stage, significantly improving the energy-saving effect. Thus, the technical problems of single and extensive optimization method of the refrigeration system regulation mode and the inability to balance energy efficiency and stability in the prior art are solved.

[0097] The above-mentioned serial numbers of the embodiments of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0098] In the above-mentioned embodiments of the present application, the description of each embodiment has its own emphasis, and the parts not described in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0099] The above-mentioned only is the preferred embodiment of the present application, it should be pointed out, for the ordinary skilled in the art, without departing from the principles of the present application, can make a number of improvements and refinements, these improvements and refinements should also be considered as the protection scope of the present application.

Claims

1. An optimization method for improving the operating efficiency of a refrigeration system based on digital twin technology, characterized in that, The method comprises the following steps: determining a current cooling load demand, screening m initial cooling unit combinations from all cooling unit combinations of a refrigeration system according to the maximum cooling capacity and the minimum cooling capacity of each cooling unit in the refrigeration system, and performing initial decomposition on the current cooling load demand to determine the initial cooling load of each cooling unit in each initial cooling unit combination; wherein the principle of the initial decomposition is that the load rate of each cooling unit is the same; for each cooling unit in each initial cooling unit combination, inputting the water supply temperature, the cooling water temperature and the initial cooling load under the current operating condition into a pre-trained digital twin model to output the energy efficiency ratio and the current ratio of the corresponding cooling unit under the current operating condition; and determining the overall energy efficiency ratio of each initial cooling unit combination according to the current cooling load demand and the energy efficiency ratio of each cooling unit in each initial cooling unit combination; screening k candidate cooling unit combinations from the m initial cooling unit combinations according to the overall energy efficiency ratio of each initial cooling unit combination; using a pre-set input parameter adjustment rule to perform two iterations of evaluation on each candidate cooling unit combination by using the digital twin model to determine the optimal overall energy efficiency ratio of each candidate cooling unit combination; wherein the input parameter adjustment rule is used to adjust the cooling load, the water supply temperature and the cooling water temperature of each cooling unit in each candidate cooling unit combination; screening d selectable cooling unit combinations from the k candidate cooling unit combinations according to the optimal overall energy efficiency ratio of each candidate cooling unit combination; performing stability judgment on each selectable cooling unit combination according to the number of started and stopped cooling units in each selectable cooling unit combination based on a pre-set first stability judgment rule and a second stability judgment rule to obtain first stability judgment results and second stability judgment results; determining b cooling unit combinations that meet the stability requirement from the d selectable cooling unit combinations according to the first stability judgment results and the second stability judgment results, and taking the cooling unit combination with the highest overall energy efficiency ratio among the b cooling unit combinations as the optimal cooling unit combination; and controlling the start and stop of each cooling unit in the refrigeration system and limiting the current ratio parameter of the started cooling unit according to the optimal cooling unit combination and the corresponding current ratio output by the digital twin model.

2. The method of claim 1, wherein, The digital twin model of the refrigeration system is trained by the following steps: setting the optimization operation boundary conditions of the refrigeration system; wherein the optimization operation boundary conditions are used to limit the fixed number of cooling units, the decision cycle and the locked operation parameters during modeling of the refrigeration system; determining the key parameters of each cooling unit under different operating conditions based on the operating characteristic data of each cooling unit in the refrigeration system under different operating conditions; wherein the key parameters include the chilled water supply temperature, the cooling water outlet temperature, the load rate and the energy efficiency ratio; and constructing and training the digital twin model of the refrigeration system based on the optimization operation boundary conditions and the key parameters; wherein the task of the digital twin model is to predict the energy efficiency ratio and the current ratio of a certain cooling unit under a certain operating condition according to the water supply temperature, the cooling water temperature and the cooling load of the cooling unit under the operating condition.

3. The method of claim 1, wherein, The operation of determining the overall energy efficiency ratio of each preliminary cooling unit combination according to the current cooling load demand and the energy efficiency ratio of each cooling unit in each preliminary cooling unit combination comprises: According to the current cooling load demand and the energy efficiency ratio of each cooling unit in each preliminary cooling unit combination, the power consumption of each cooling unit in each preliminary cooling unit combination is calculated; and According to the current cooling load demand and the power consumption of each cooling unit in each preliminary cooling unit combination, the overall energy efficiency ratio of each preliminary cooling unit combination is determined.

4. The method of claim 1, wherein, The operation of adjusting the input parameters according to the preset input parameter adjustment rule, and using the digital twin model to perform two iterations of evaluation on each candidate cooling unit combination to determine the optimal overall energy efficiency ratio of each candidate cooling unit combination comprises: Under a preset first constraint condition, the cooling load of each cooling unit in each candidate cooling unit combination is redistributed by adjusting the chilled water flow rate, the supply water temperature, the cooling water temperature and the redistributed cooling load under the current operating condition are input into the digital twin model, the energy efficiency ratio and the current ratio of the corresponding cooling unit under the current operating condition are output, and the first overall energy efficiency ratio of each candidate cooling unit combination is determined according to the current cooling load demand and the energy efficiency ratio of each cooling unit in each candidate cooling unit combination; wherein the first constraint condition is that the total cooling load of all cooling units in the candidate cooling unit combination is equal to the current cooling load demand; Under the first constraint condition and a preset second constraint condition, by adjusting the chilled water temperature and the chilled water flow rate of each chiller, the cooling load of each chiller in each candidate chiller combination is redistributed, and the chilled water temperature of each chiller is changed, the changed chilled water temperature, the cooling water temperature under the current operating condition and the adjusted cooling load are input into the digital twin model, the energy efficiency ratio and the current ratio of the corresponding chiller under the current operating condition are output, and the second overall energy efficiency ratio of each candidate chiller combination is determined according to the current cooling load demand and the energy efficiency ratio of each chiller in each candidate chiller combination; wherein the second constraint condition is: ;T 总供水温度 is the total chilled water temperature of the refrigeration system, T 总管回水温度 is the total pipe return water temperature of the refrigeration system, Q is the current cooling load demand, Flow j is the chilled water flow rate of the jth chiller in each candidate chiller combination; and Based on the first overall energy efficiency ratio and the second overall energy efficiency ratio, the optimal overall energy efficiency ratio of each candidate cooling unit combination is determined.

5. The method of claim 1, wherein, The first stability judgment rule is to compare each selectable cooling unit combination with the currently running cooling unit combination, and select a cooling unit combination from the d selectable cooling unit combinations whose difference in the number of cooling unit starts and stops from the currently running cooling unit combination is less than a first preset threshold.

6. The method of claim 1, wherein, The second stability judgment rule is to determine the future cooling load demand of the refrigeration system in a future preset time period, select a plurality of target cooling unit combinations from all cooling unit combinations of the refrigeration system based on the future cooling load demand and the digital twin model, compare each of the d selectable cooling unit combinations with each of the plurality of target cooling unit combinations, and select a cooling unit combination from the d selectable cooling unit combinations whose difference in the number of cooling unit starts and stops from the plurality of target cooling unit combinations is less than a second preset threshold.

7. A storage medium, characterized by The storage medium comprises a stored program, wherein the program is executed by a processor when the program is run to perform the method of any one of claims 1 to 6.

8. An optimization device for improving the operating efficiency of a refrigeration system based on digital twin technology, comprising: A first screening module is configured to determine a current cooling load demand, screen m preliminary cooling unit combinations that meet the current cooling load demand from all cooling unit combinations of a refrigeration system according to the maximum cooling capacity and the minimum cooling capacity of each cooling unit in the refrigeration system, and determine an initial cooling load of each cooling unit in each preliminary cooling unit combination by initially decomposing the current cooling load demand; wherein the principle of the initial decomposition is that the load rate of each cooling unit is the same. The energy efficiency ratio determination module is configured to input the water supply temperature, the cooling water temperature and the initial cooling load in a current operation condition into a pre-trained digital twin model for each chiller in each preliminary chiller combination, and output the energy efficiency ratio and the current ratio of the corresponding chiller in the current operation condition; and determine the overall energy efficiency ratio of each preliminary chiller combination according to the current cooling load demand and the energy efficiency ratio of each chiller in each preliminary chiller combination. The second screening module is configured to screen k candidate chiller combinations from the m preliminary chiller combinations according to the overall energy efficiency ratio of each preliminary chiller combination; and determine the optimal overall energy efficiency ratio of each candidate chiller combination by performing two iterations of evaluation on each candidate chiller combination using the digital twin model and a preset input parameter adjustment rule, wherein the input parameter adjustment rule is used to adjust the cooling load, the water supply temperature and the cooling water temperature of each chiller in each candidate chiller combination. The third screening module is configured to screen d selectable chiller combinations from the k candidate chiller combinations according to the optimal overall energy efficiency ratio of each candidate chiller combination; perform stability judgment on each selectable chiller combination based on a preset first stability judgment rule and a second stability judgment rule according to the number of started and stopped chillers in each selectable chiller combination, and obtain first and second stability judgment results; determine b chiller combinations that meet the stability requirement from the d selectable chiller combinations according to the first and second stability judgment results, and select the chiller combination with the highest overall energy efficiency ratio from the b chiller combinations as the optimal chiller combination; and The adjustment control module is configured to control the start and stop of each chiller in the refrigeration system and limit the current ratio of the started chiller according to the optimal chiller combination and the corresponding current ratio output by the digital twin model.

9. The apparatus of claim 8, wherein, The model construction and training module is further configured to train the digital twin model of the refrigeration system by the following steps: setting an optimized operation boundary condition of the refrigeration system, wherein the optimized operation boundary condition is used to limit the fixed number of chillers, the decision cycle and the locked operation parameters during modeling of the refrigeration system; determining the key parameters of each chiller in different operation conditions based on the operation characteristic data of each chiller in the refrigeration system, wherein the key parameters include the chilled water supply temperature, the cooling water outlet temperature, the load rate and the energy efficiency ratio; and constructing and training the digital twin model of the refrigeration system based on the optimized operation boundary condition and the key parameters, wherein the task of the digital twin model is to predict the energy efficiency ratio and the current ratio of a chiller in a certain operation condition according to the water supply temperature, the cooling water temperature and the cooling load of the chiller in the operation condition.

10. An optimization device for improving the operating efficiency of a refrigeration system based on digital twin technology, characterized in that, comprise: a processor; and a memory connected with the processor, configured to provide the processor with instructions for processing the following processing steps: Determine the current cooling load demand, according to the maximum cooling capacity and the minimum cooling capacity of each chiller in the refrigeration system, screen m initial chiller combinations from all chiller combinations in the refrigeration system to meet the current cooling load demand, and perform initial decomposition on the current cooling load demand to determine the initial cooling load of each chiller in each initial chiller combination; wherein the principle of initial decomposition is that the load rate of each chiller is the same; For each chiller in each initial chiller combination, input the water supply temperature, cooling water temperature and initial cooling load under the current operating condition into the pre-trained digital twin model, output the energy efficiency ratio and current ratio of the corresponding chiller under the current operating condition; according to the current cooling load demand and the energy efficiency ratio of each chiller in each initial chiller combination, determine the overall energy efficiency ratio of each initial chiller combination; According to the overall energy efficiency ratio of each initial chiller combination, screen k candidate chiller combinations from the m initial chiller combinations; use the preset input parameter adjustment rule to respectively perform two iterations on each candidate chiller combination using the digital twin model, and determine the optimal overall energy efficiency ratio of each candidate chiller combination; wherein the input parameter adjustment rule is used to adjust the cooling load, water supply temperature and cooling water temperature of each chiller in each candidate chiller combination; According to the optimal overall energy efficiency ratio of each candidate chiller combination, screen d optional chiller combinations from the k candidate chiller combinations; according to the number of start-stop chillers in each optional chiller combination, perform stability judgment on each optional chiller combination based on the preset first stability judgment rule and second stability judgment rule, and obtain the first stability judgment result and the second stability judgment result; according to the first stability judgment result and the second stability judgment result, determine b chiller combinations that meet the stability requirement from the d optional chiller combinations, and take the chiller combination with the highest overall energy efficiency ratio in the b chiller combinations as the optimal chiller combination; and According to the optimal chiller combination and the corresponding current ratio output by the digital twin model, control the start-stop of each chiller in the refrigeration system and limit the current ratio parameter of the started chiller.

Images