Energy-saving control method and device for refrigeration system, and refrigeration system

Abstract

The application relates to the technical field of refrigeration systems, and discloses an energy-saving control method for a refrigeration system, which comprises the following steps: obtaining a mechanism model associated with a water-cooling system and a comfort model associated with an air-cooling system; coupling the mechanism model and the comfort model to obtain a fitness function; determining boundary conditions and importing the boundary conditions into the fitness function to optimize parameters through a heuristic algorithm to obtain predicted parameters; and outputting the predicted parameters to an execution end. The application can improve the accuracy of parameter prediction of the refrigeration system and release the energy-saving potential of group control strategies. The application also discloses an energy-saving control device for the refrigeration system and the refrigeration system.

Description

Technical Field

[0001] This application relates to the field of refrigeration system technology, such as an energy-saving control method and device for refrigeration systems, and refrigeration systems. Background Technology

[0002] Currently, building energy conservation is a hot topic in the context of dual carbon emissions. Group control energy conservation of central air conditioning systems is a nonlinear optimization problem. Due to the variable operating conditions and uncertain health status of central air conditioning systems, it is difficult to obtain the optimal parameters using purely theoretical methods for group control energy conservation.

[0003] To address the above issues, the relevant technologies employ the following group control strategy: preset conditional logic, real-time parameter detection using high-precision sensors, execution of corresponding logical judgments according to a preset schedule, and triggering a standardized process for adding, removing, starting, and stopping machines.

[0004] In the process of implementing the embodiments of this disclosure, at least the following problems were found in the related art:

[0005] The group control strategy employed in related technologies relies heavily on the prediction of cooling load, which can lead to overcooling or overheating. The accuracy of predictive control is also poor, limiting the potential for energy savings.

[0006] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this application, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0007] To provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This summary is not intended as a general commentary, nor is it intended to identify key / important components or describe the scope of protection of these embodiments, but rather as a prelude to the detailed description that follows.

[0008] This disclosure provides an energy-saving control method, apparatus, and refrigeration system for a refrigeration system, to improve the accuracy of parameter prediction and unleash the energy-saving potential of group control strategies.

[0009] In some embodiments, the method includes: obtaining a mechanistic model associated with a water-cooled system and a comfort model associated with an air-cooled system; coupling the mechanistic model and the comfort model to obtain a fitness function; determining boundary conditions and importing the boundary conditions into the fitness function to optimize parameters through a heuristic algorithm to obtain predicted parameters; and outputting the predicted parameters to the execution end.

[0010] In some embodiments, the apparatus includes a processor and a memory storing program instructions, the processor being configured to execute, when running the program instructions, an energy-saving control method for a refrigeration system as described above.

[0011] In some embodiments, the refrigeration system includes: a refrigeration unit; a water-cooled system; an air-cooled system; and an energy-saving control device for the refrigeration system as described above, which is installed on the refrigeration unit.

[0012] The energy-saving control method, apparatus, and refrigeration system for refrigeration systems provided in this disclosure can achieve the following technical effects:

[0013] This disclosure provides embodiments for obtaining a mechanistic model associated with a water-cooled system and a comfort model associated with an air-cooled system. These models are then coupled to obtain a fitness function. This approach allows for a global consideration of both the energy-saving goals of the water-cooled system and the comfort goals of the air-cooled system, achieving global optimization. Next, this disclosure defines boundary conditions and imports these conditions into the fitness function. A heuristic algorithm is used to optimize the parameters to obtain predicted parameters, which are then output to the execution end. In this way, for two different types of systems, their respective models are coupled to obtain a fitness function that accurately reflects the operating conditions of the refrigeration system. The heuristic algorithm accurately obtains the predicted parameters, improving the accuracy of parameter prediction at the execution end and unlocking the energy-saving potential of the group control strategy.

[0014] The above general description and the description below are exemplary and illustrative only and are not intended to limit this application. Attached Figure Description

[0015] One or more embodiments are illustrated by way of example with reference to the accompanying drawings. These illustrations and drawings do not constitute a limitation on the embodiments. Elements having the same reference numerals in the drawings are shown as similar elements. The drawings are not to be scaled. And wherein:

[0016] Figure 1 This is a schematic diagram of an energy-saving control method for a refrigeration system provided in an embodiment of this disclosure;

[0017] Figure 2 This is a schematic diagram of another energy-saving control method for a refrigeration system provided in an embodiment of this disclosure;

[0018] Figure 3 This is a schematic diagram of another energy-saving control method for a refrigeration system provided in an embodiment of this disclosure;

[0019] Figure 4 This is a schematic diagram of another energy-saving control method for a refrigeration system provided in an embodiment of this disclosure;

[0020] Figure 5 This is a schematic diagram of another energy-saving control method for a refrigeration system provided in an embodiment of this disclosure;

[0021] Figure 6 This is an application illustration provided by an embodiment of the present disclosure;

[0022] Figure 7 This is a schematic diagram of an energy-saving control device for a refrigeration system provided in an embodiment of this disclosure. Detailed Implementation

[0023] To provide a more detailed understanding of the features and technical content of the embodiments of this disclosure, the implementation of the embodiments of this disclosure will be described in detail below with reference to the accompanying drawings. The accompanying drawings are for illustrative purposes only and are not intended to limit the embodiments of this disclosure. In the following technical description, for ease of explanation, several details are used to provide a full understanding of the disclosed embodiments. However, one or more embodiments may still be implemented without these details. In other cases, well-known structures and devices may be simplified in their depiction to simplify the drawings.

[0024] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this disclosure described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion.

[0025] Unless otherwise stated, the term "multiple" means two or more.

[0026] In this embodiment of the disclosure, the character " / " indicates that the objects before and after it are in an "or" relationship. For example, A / B means: A or B.

[0027] The term "and / or" describes an association between objects, indicating that three relationships can exist. For example, A and / or B means: A or B, or A and B.

[0028] The term "correspondence" can refer to an association or binding relationship. The correspondence between A and B means that there is an association or binding relationship between A and B.

[0029] This disclosure provides a refrigeration system, including a water-cooled system and an air-cooled system. The water-cooled system achieves refrigeration by exchanging heat with the input chilled water. The energy consumption of the water-cooled system is much higher than that of the air-cooled system. Therefore, this disclosure constructs a mechanism model for the water-cooled system with energy saving as the goal. As an example, the optimization objective of the water-cooled system is to minimize the power of the refrigeration unit while meeting preset cooling capacity conditions.

[0030] Air-cooled systems achieve cooling or heating by exchanging heat with air in the environment. The energy consumption of air-cooled systems is significantly lower than that of water-cooled systems. Therefore, this disclosure focuses on air-cooled systems and constructs a comfort model with comfort as the objective.

[0031] Based on the above refrigeration system, combined with Figure 1 As shown, this disclosure provides an energy-saving control method for a refrigeration system, including:

[0032] S11, the processor obtains the mechanism model associated with the water cooling system and the comfort model associated with the air cooling system.

[0033] S12, processor coupling mechanism model and comfort model, to obtain fitness function.

[0034] S13, the processor determines the boundary conditions and imports the boundary conditions into the fitness function to optimize the parameters through a heuristic algorithm to obtain the predicted parameters.

[0035] In this step, the heuristic algorithms include simulated annealing, genetic algorithms, or particle swarm optimization.

[0036] S14, the processor outputs the prediction parameters to the execution end.

[0037] The energy-saving control method for refrigeration systems provided in this disclosure obtains a mechanistic model associated with a water-cooled system and a comfort model associated with an air-cooled system. The mechanistic model and the comfort model are then coupled to obtain a fitness function. This allows for a global consideration of both the energy-saving goal of the water-cooled system and the comfort goal of the air-cooled system, achieving global optimization. Next, the disclosure determines boundary conditions and imports these boundary condition values ​​into the fitness function. A heuristic algorithm is used to optimize the parameters to obtain predicted parameters, which are then output to the execution end. Thus, for two different types of systems, their respective models are coupled to obtain a fitness function that accurately reflects the operating conditions of the refrigeration system. The heuristic algorithm accurately obtains the predicted parameters, improving the accuracy of parameter prediction at the execution end and unlocking the energy-saving potential of the group control strategy.

[0038] It should be noted that the execution entity of this embodiment can be configured in a cooling system or in a server that is communicatively connected to the cooling system. This embodiment does not impose specific limitations in this regard.

[0039] Optionally, the processor coupling mechanism model and the comfort model are used to obtain the fitness function, including:

[0040] calculate

[0041] Where TC and TP represent the comfort model and the mechanism model, respectively. qk TP qk Let represent the thermal comfort constant and the total power constant, respectively, and fitness represent the fitness function. δ represents the correction coefficient.

[0042] In order to achieve coordinated optimization between the water-cooling system and the air-cooling system, this embodiment of the disclosure couples the mechanistic model and the comfort model, thereby realizing coordinated energy saving between the water-cooling system and the air-cooling system. Furthermore, to prevent the fitness function from having a denominator of zero during the coupling process when no better value is obtained, this embodiment of the disclosure sets a correction coefficient δ. Thus, by coupling their respective models, this embodiment of the disclosure can obtain a fitness function that accurately reflects the operating conditions of the refrigeration system, which is beneficial to improving the accuracy of parameter prediction at the execution end and to unlocking the energy-saving potential of the refrigeration system.

[0043] Optionally, δ is 0.001%. Understandably, δ can also be any other value less than 0.01%.

[0044] Optionally, combined Figure 2 As shown, the processor obtains the mechanistic model associated with the water cooling system, including:

[0045] S21, the processor constructs the mapping relationship between the water cooling parameters of the cooling device and the power of the cooling host, and obtains the baseline mechanism model pool.

[0046] S22, the processor inputs historical water-cooling data for each cooling device into the corresponding baseline mechanism model pool for model training, obtaining candidate mechanism models. In this step, the water-cooling data includes operating parameter data and control parameter data. Correspondingly, the historical water-cooling data includes historical operating parameter data and historical control parameter data. The operating parameter data is obtained by collecting water-cooling parameters from sensors.

[0047] S23, the processor selects the candidate mechanism model as the mechanism model if the candidate mechanism model meets the error accuracy requirements.

[0048] In this way, the embodiments of this disclosure construct a mapping relationship between the water-cooling parameters of the cooling devices and the power of the chiller. After obtaining a reference mechanism model pool, the historical water-cooling data of each cooling device is input into the reference mechanism model pool for training to obtain candidate mechanism models, thus achieving sufficient learning from the historical water-cooling data. When a candidate mechanism model meets the error accuracy requirements, it is selected as the mechanism model. Therefore, the embodiments of this disclosure, for water-cooling systems, achieve the goal of constructing a mechanism model with energy saving as the objective, which is beneficial to improving the accuracy of subsequent parameter prediction.

[0049] Optionally, the processor determines whether the candidate mechanism model meets the error accuracy in the following manner:

[0050] The processor obtains the current error of the alternative mechanism model.

[0051] The current error is less than or equal to the error precision.

[0052] Optionally, the processor inputs historical water-cooling data of each cooling device into the baseline mechanism model pool for model training, to obtain candidate mechanism models, including:

[0053] When the model training is not incremental, the processor extracts full data and test data from historical water-cooling data, and uses the benchmark mechanism model corresponding to the benchmark mechanism model pool to train the model on the full data and test data respectively.

[0054] When the model training is incremental, the processor randomly selects a target number of historical water cooling data from the historical water cooling data and inputs them into the benchmark mechanism model pool corresponding to the benchmark mechanism model for training.

[0055] Thus, for the mechanistic model, after the data volume accumulates to a certain level, full data is used for model training. However, new data will be continuously generated, carrying the latest operating information of the water cooling system. Therefore, it is necessary to incorporate the new data into the model training. However, if new data is input into the benchmark mechanistic model pool every time for training, it will result in a waste of computing power and poor applicability in practical engineering. To address this, in this embodiment, when the model training is not incremental (i.e., the first time model training is performed), the processor extracts full data and test data from the historical water cooling data, and inputs the full data and test data into the benchmark mechanistic model for training, respectively. During the iteration process of model training, the full data and test data are shuffled to ensure sufficient learning from the historical water cooling data. In this embodiment, when the model training is incremental (i.e., the historical water cooling data includes the latest data), the processor randomly selects a target number of target historical water cooling data from the historical water cooling data and inputs them into the benchmark mechanistic model pool for training. Thus, the embodiments of this disclosure perform differentiated model training for full and incremental data, saving computing resources and improving deployability.

[0056] As an example, when the model training is not incremental, the total amount of data (M) that the processor extracts from the historical water-cooled data, including both full data and test data, is M.

[0057] In the case of incremental model training, the processor randomly selects a target number of historical water-cooling data points from the historical water-cooling data and inputs them into the benchmark mechanism model pool for training, including:

[0058] The target number is N×M. N≥3. Preferably, N is 5, 7, or 10. Understandably, the specific value of N can also be determined according to the needs of model training. This disclosure does not impose specific limitations on this.

[0059] Optionally, the processor randomly selects a target number of historical water-cooling data points from the historical water-cooling data and inputs them into the benchmark mechanism model pool for model training. This includes: the processor training the model on the target historical water-cooling data using K-Fold validation based on the benchmark mechanism model. This improves the reliability of model training.

[0060] Optionally, the cooling devices include a chiller, a chilled pump, a cooling pump, and a cooling tower. The processor constructs a mapping relationship between the water-cooling parameters of the cooling devices and the power of the chiller, obtaining a baseline mechanistic model pool, including:

[0061] The processor constructs the correspondence between the chilled water outlet temperature, the cooling water inlet temperature, and the cooling capacity and power of the chiller, thus obtaining the first baseline mechanism model.

[0062] The processor constructs the operating frequency of the chilled pump, the operating frequency of the cooling pump, the wind speed of the cooling tower, the operating frequency of the cooling tower, and the correspondence between the outdoor ambient temperature and humidity and the power of the refrigeration unit, thus obtaining a second baseline mechanism model.

[0063] The processor constructs the correspondence between the chilled flow rate of the chilled pump and the chilled water temperature difference and the cooling capacity of the chiller, thus obtaining the third benchmark mechanism model.

[0064] The processor constructs the correspondence between the cooling pump's cooling flow rate and the cooling water temperature difference and the cooling capacity, thus obtaining the fourth baseline mechanism model.

[0065] The baseline mechanism model pool includes a first baseline mechanism model, a second baseline mechanism model, a third baseline mechanism model, and a fourth baseline mechanism model. Cooling capacity is the sum of the cooling capacity and power of the refrigeration unit.

[0066] Thus, for a chiller, the main factors affecting its power include the chilled water outlet temperature, the cooling water inlet temperature, and the chiller's cooling capacity. Therefore, this embodiment establishes a correspondence between the chilled water outlet temperature, the cooling water inlet temperature, and the chiller's cooling capacity and power. In addition, since the cooling tower is affected by the intensity of external convective heat transfer, besides the operating frequencies of the chilled water pump, cooling water pump, and cooling tower, the cooling tower's wind speed and the outdoor ambient temperature and humidity also affect the chiller's power. Therefore, this embodiment establishes a correspondence between the operating frequency of the chilled water pump, the operating frequency of the cooling water pump, the cooling tower's wind speed, the cooling tower's operating frequency, and the outdoor ambient temperature and humidity and the chiller's power. Meanwhile, from the perspective of energy conservation, the chiller's cooling capacity is energy-related to the chilled water flow rate and the chilled water temperature difference. Simultaneously, the cooling capacity is also energy-related to the cooling pump's cooling flow rate and the cooling water temperature difference. Therefore, this embodiment of the disclosure constructs the correspondence between the chilled flow rate and chilled water temperature difference of the chilled pump and the cooling capacity of the refrigeration unit, and constructs the correspondence between the cooling flow rate and cooling water temperature difference of the cooling pump and the cooling capacity. In summary, this embodiment of the disclosure establishes a mapping relationship between the water cooling parameters of the cooling devices and the power of the refrigeration unit based on the correlation between different types of water cooling parameters and the correlation between different cooling devices, in order to obtain a baseline mechanism model pool.

[0067] Optionally, the processor constructs a mapping relationship between the water-cooling parameters of the cooling devices and the power of the cooling host to obtain a baseline mechanism model pool, which also includes:

[0068] The processor constructs the correspondence between the operating frequency of the chilled pump, the switching combination state of the chilled pump, and the chilled flow rate of the chilled pump, thus obtaining the fifth baseline mechanism model.

[0069] The processor constructs the correspondence between the operating frequency of the cooling pump, the switching combination state of the cooling pump, and the cooling flow rate of the cooling pump, and obtains the sixth baseline mechanism model.

[0070] The benchmark mechanism model pool includes the first benchmark mechanism model, the second benchmark mechanism model, the third benchmark mechanism model, the fourth benchmark mechanism model, the fifth benchmark mechanism model, and the sixth benchmark mechanism model.

[0071] Thus, in addition to the correlations between different types of water-cooling parameters and between different cooling devices, there are also constraints between the parameters of the chilled pump and the cooling pump. Specifically, the chilled flow rate of the chilled pump is related to its operating frequency and its on / off state. The cooling flow rate of the cooling pump is related to its operating frequency and its on / off state. Therefore, this embodiment of the disclosure establishes the correspondence between the operating frequency, on / off state, and chilled flow rate of the chilled pump, and the corresponding correspondence between the operating frequency, on / off state, and cooling flow rate of the cooling pump, respectively. This further optimizes the baseline mechanism model.

[0072] In one embodiment, when the historical water cooling data includes the historical outlet temperature of chilled water, the historical inlet temperature of cooling water, and the historical cooling capacity of the refrigeration unit, the processor inputs the historical water cooling data of each cooling device to the benchmark mechanism model corresponding to the benchmark mechanism model pool for model training to obtain candidate mechanism models. This includes: the processor inputting the historical outlet temperature of chilled water, the historical inlet temperature of cooling water, and the historical cooling capacity of the refrigeration unit to the first benchmark mechanism model for model training to obtain candidate mechanism models corresponding to the aforementioned historical water cooling data.

[0073] In another embodiment, when the historical water-cooling data includes the historical operating frequency of the chilled pump, the historical operating frequency of the cooling pump, the historical wind speed of the cooling tower, the historical operating frequency of the cooling tower, and the historical outdoor temperature and humidity, the processor inputs the historical water-cooling data of each cooling device to the benchmark mechanism model corresponding to the benchmark mechanism model pool for model training to obtain alternative mechanism models. This includes: the processor inputting the historical operating frequency of the chilled pump, the historical operating frequency of the cooling pump, the historical wind speed of the cooling tower, the historical operating frequency of the cooling tower, and the historical outdoor temperature and humidity to the second benchmark mechanism model to obtain alternative mechanism models corresponding to the aforementioned historical water-cooling data.

[0074] In another embodiment, when the historical water cooling data includes the historical chilled flow rate of the chilled pump and the historical temperature difference of the chilled water, the processor inputs the historical water cooling data of each cooling device to the benchmark mechanism model corresponding to the benchmark mechanism model pool for model training to obtain alternative mechanism models, including: the processor inputs the historical chilled flow rate of the chilled pump and the historical temperature difference of the chilled water to the third benchmark mechanism model to obtain the alternative mechanism models corresponding to the aforementioned historical water cooling data.

[0075] When the historical water cooling data includes the historical cooling flow rate of the cooling pump and the historical temperature difference of the cooling water, or when the historical water cooling data includes the historical operating frequency of the chilled pump and the historical switching combination state of the chilled pump, or when the historical water cooling data includes the historical operating frequency of the cooling pump and the historical switching combination state of the cooling pump, the method for obtaining the alternative mechanism model can be referred to the foregoing description, and will not be repeated in this embodiment.

[0076] Optionally, combined Figure 3 As shown, the processor obtains a comfort model associated with the air-cooling system, including:

[0077] S31, the processor obtains the air cooling data of the air-cooled system and obtains comfort rating information. The air cooling data includes some or all of the following: unit set temperature, fresh air temperature, fresh air humidity, fresh air carbon dioxide concentration, indoor temperature, indoor humidity, and indoor carbon dioxide concentration.

[0078] S32, the processor inputs air-cooling data and comfort rating information into the baseline comfort model to obtain the comfort model.

[0079] In this step, the baseline comfort model can be either a white-box model or a black-box model.

[0080] In this way, achieving the goal of constructing a comfort model with comfort as the objective will help improve the accuracy of subsequent parameter predictions.

[0081] Optionally, the processor obtains comfort rating information, including: the processor accesses a database storing multiple evaluation metrics and their corresponding weights. As an example, the evaluation metrics include airflow sensitivity coefficient, temporal uniformity, spatial uniformity, vertical temperature difference, and PMV (Predicted Mean Vote). The weights for airflow sensitivity, temporal uniformity, spatial uniformity, and vertical temperature difference are all 0.15, while the weight for PMV is 0.4.

[0082] Optionally, the processor obtains historical water-cooling data or air-cooling data, including:

[0083] The processor acquires data from different types of sensors and the data transmission types of each type of sensor.

[0084] When the data transmission type matches the standard type, the processor parses the data collected by the sensor to obtain data information and time information.

[0085] Thus, for a central air conditioning refrigeration system, the system is equipped with various sensors, such as pipe temperature sensors, pressure sensors, airflow sensors, return air temperature sensors, and return air humidity sensors. With prolonged use, the refrigeration system may replace or add sensors. These replacements or additions may differ from the original sensors in model, data transmission method, and parsing method. The information collected by these replacements or additions, along with the original sensors, constitutes multi-source information. To reduce the difficulty of subsequent data retrieval and processing, it is necessary to unify and integrate this multi-source information. Therefore, this embodiment obtains data collected by different types of sensors and their respective data transmission types. When the data transmission type matches a standard type, a data parsing operation is initiated to obtain data information and actual information, and the data is stored. This facilitates subsequent data retrieval and processing, improving the speed of generating subsequent mechanistic and comfort models.

[0086] Optionally, if the data transmission type does not match the standard type, the processor stores the data transmission type in the standard rule base. The standard rule base includes the sensor type, data transmission type information, and manufacturer information. Understandably, the standard rule base may also include other information identifying sensor differences.

[0087] Optionally, the processor stores data and time information, including storing the data and time information in the time-series database according to the principle of proximity. As an example, for data returned at 12:45:41, the processor stores the time information as 12:45:30. For data returned at 12:45:52, the processor stores the time information as 12:46:00. Simultaneously, the processor's storage of data and time information also includes determining whether the data contains outliers or missing values. If outliers or missing values ​​are determined, the processor performs interpolation based on statistical principles. This ensures the integrity of the data in the time-series database, facilitating subsequent data retrieval.

[0088] Optionally, combined Figure 4 As shown, the processor determines the boundary conditions and imports them into the fitness function to optimize the parameters through a heuristic algorithm, obtaining the predicted parameters, including:

[0089] S41, the processor obtains the predicted load for the next moment based on the control parameters input by the user.

[0090] S42, the processor determines the boundary conditions.

[0091] S43, the processor optimizes the fitness function based on the predicted load and boundary conditions at the next time step to obtain the prediction parameters.

[0092] Thus, when using heuristic algorithms to predict parameters of a refrigeration system, the principle is to optimize the parameters using an objective function, optimization parameters, and boundary conditions. Based on this, the embodiments of this disclosure first obtain the predicted load for the next time step based on the control parameters input by the user. Then, the boundary conditions are determined. Finally, the fitness function is optimized based on the predicted load for the next time step and the boundary conditions to obtain the predicted parameters. In this way, the embodiments of this disclosure use the predicted load for the next time step as an input condition to achieve high-frequency regulation, which is more conducive to tapping the energy-saving potential of the group control strategy.

[0093] Optionally, the processor obtains the predicted load for the next time step based on the control parameters input by the user, including:

[0094] Provided that the amount of cold data meets the data integrity requirements, the processor uses a single-index time-series forecasting algorithm to predict the load at the next moment. Optionally, the single-index time-series forecasting algorithm includes AR (Auto-Regressive), MA (Moving Average), and ARIMA (Autoregressive Integrated Moving Average model) algorithms.

[0095] If the amount of cold data does not meet the data integrity requirements, the processor uses a deep neural network learning algorithm to predict the load at the next moment. Optionally, the deep neural network learning algorithm includes the LSTM (Long Short-Term Memory) algorithm or the Informer algorithm.

[0096] When the amount of cold data is less than the upper limit threshold but greater than the lower limit threshold and there is missing data, the processor uses data clustering to predict the load at the next moment.

[0097] The cooling data includes historical water-cooled data and air-cooled data.

[0098] Thus, the accuracy of the predicted load directly affects the subsequent group control operation of the water-cooled and air-cooled systems. Therefore, this disclosure employs different prediction strategies for different cooling load data. When the amount of cooling load data meets the data integrity condition, it indicates high data integrity; in this case, a deep neural network learning algorithm can be used for load prediction. When the amount of cooling load data does not meet the data integrity condition, it indicates poor data integrity; in this case, a single-index time-series prediction algorithm can be used for load prediction. When the amount of cooling load data is less than the upper limit threshold but greater than the lower limit threshold and contains missing data, data clustering can be used for load prediction. This helps improve the accuracy of load prediction.

[0099] Optionally, the processor determines whether the amount of cold energy data satisfies the data integrity requirements in the following manner.

[0100] Within the most recent H years, the amount of cooling data is greater than or equal to the upper limit threshold and is complete with no missing data. Furthermore, each cooling data point has corresponding environmental or device parameters. Where H is greater than or equal to 3 and less than or equal to 5.

[0101] Optionally, the processor determines that the amount of cold energy data does not meet the data integrity requirements in the following manner.

[0102] In the most recent H years, the amount of cold energy data is less than the lower limit threshold and is incomplete.

[0103] Optionally, the processor determines whether the amount of cold energy data is less than the upper limit threshold and greater than the lower limit threshold, and whether there is missing data, in the following manner:

[0104] In the past W years, the amount of cold data is less than the upper limit threshold but greater than the lower limit threshold, and there are missing data.

[0105] Where W is greater than or equal to 1 and less than or equal to 3.

[0106] This helps improve the accuracy of load forecasting.

[0107] Optionally, combined Figure 5 As shown, the processor uses data clustering to predict the load at the next time step, including:

[0108] S51, the processor performs clustering processing on the daily cooling load sequences to obtain the daily reference load function. Here, the cooling load sequence represents time-series continuous cooling data.

[0109] S52, the processor obtains the daily aggregated values ​​of environmental parameters based on the daily reference load function.

[0110] The S53 processor builds an environmental model based on the daily aggregated values ​​of environmental parameters.

[0111] The S54 processor predicts the load for the next moment based on an environmental model.

[0112] In this way, for cold data with a moderate amount of data and missing data, the load at the next moment can be accurately predicted.

[0113] Optionally, the processor predicts the load for the next moment based on an environmental model, including:

[0114] The processor predicts the target load function for the day based on the environmental model.

[0115] The processor predicts the load for the next moment based on the target load function for the day.

[0116] Optionally, the execution end includes an edge actuator. The processor outputs predicted parameters to the execution end, including: the processor sending control commands to the edge actuator to cause the edge actuator to perform corresponding control operations. The control commands include the predicted parameters. The edge actuator includes an electronic device with execution capabilities. As an example, the edge actuator is a host computer.

[0117] It should be noted that, along with sending control commands to the edge actuator, the processor also receives correction schemes from experts. The processor then adjusts the prediction parameters according to these correction schemes to obtain updated prediction parameters. In this way, by incorporating correction strategies from experts, situations where obviously abnormal prediction parameter values ​​are input to the execution end are avoided.

[0118] In practical applications, such as Figure 6 As shown, the energy-saving control method for a refrigeration system specifically implements the following steps:

[0119] S101, the processor obtains historical operating parameter data and historical control parameter data, integrates the historical operating parameter data using a standard rule base, and stores the integrated historical operating parameter data and historical control parameter data in a time-series database.

[0120] S102, the processor constructs the mapping relationship between the water cooling parameters of the cooling device and the power of the cooling host, and obtains the baseline mechanism model pool.

[0121] S103, the processor inputs the historical water-cooling data of each cooling device into the baseline mechanism model pool, and uses the model training module to train the baseline mechanism model corresponding to the above water-cooling data to obtain candidate mechanism models. If the candidate mechanism model meets the error accuracy requirements, the candidate mechanism model is selected as the mechanism model. The historical water-cooling data includes historical operating parameter data and historical control parameter data.

[0122] S104, the processor obtains the air cooling data of the air-cooled system and obtains comfort rating information. The air cooling data includes part or all of the following: unit set temperature, fresh air temperature, fresh air humidity, fresh air carbon dioxide concentration, indoor temperature, indoor humidity, and indoor carbon dioxide concentration.

[0123] S105: The processor inputs air-cooling data and comfort rating information into the baseline comfort model to obtain the comfort model.

[0124] S106, processor coupling mechanism model and comfort model, to obtain fitness function (i.e. objective function).

[0125] S107, the processor determines the boundary conditions and obtains the predicted load for the next time step. Then, based on the optimizer, the predicted load for the next time step, and the boundary conditions, the fitness function is optimized to obtain the prediction parameters.

[0126] S108, the processor outputs the predicted parameters to the execution end.

[0127] Combination Figure 7 As shown, this disclosure provides an energy-saving control device for a refrigeration system, including a processor 400 and a memory 401. Optionally, the device may further include a communication interface 402 and a bus 403. The processor 400, communication interface 402, and memory 401 can communicate with each other via the bus 403. The communication interface 402 can be used for information transmission. The processor 400 can call logical instructions in the memory 401 to execute the energy-saving control method for the refrigeration system described in the above embodiment.

[0128] Furthermore, the logic instructions in the aforementioned memory 401 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium.

[0129] The memory 401, as a computer-readable storage medium, can be used to store software programs and computer-executable programs, such as program instructions / modules corresponding to the methods in the embodiments of this disclosure. The processor 400 executes functional applications and data processing by running the program instructions / modules stored in the memory 401, thereby implementing the energy-saving control method for the refrigeration system described above.

[0130] The memory 401 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the terminal device. Furthermore, the memory 401 may include high-speed random access memory and may also include non-volatile memory.

[0131] This disclosure provides a refrigeration system, including: a refrigeration unit, a water-cooling system, an air-cooling system, and the aforementioned energy-saving control device for the refrigeration system. The energy-saving control device for the refrigeration system is installed in the refrigeration unit. The installation relationship described herein is not limited to placement inside the product, but also includes installation connections with other components of the product, including but not limited to physical connections, electrical connections, or signal transmission connections. Those skilled in the art will understand that the energy-saving control device for the refrigeration system can be adapted to feasible product bodies to achieve other feasible embodiments.

[0132] This disclosure provides a computer-readable storage medium storing computer-executable instructions configured to execute the above-described energy-saving control method for a refrigeration system.

[0133] The aforementioned computer-readable storage medium may be a transient computer-readable storage medium or a non-transitory computer-readable storage medium.

[0134] The technical solutions of this disclosure can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes one or more instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in this disclosure. The aforementioned storage medium can be a non-transitory storage medium, including: a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and other media capable of storing program code; it can also be a transient storage medium.

[0135] The foregoing description and accompanying drawings fully illustrate embodiments of this disclosure to enable those skilled in the art to practice them. Other embodiments may include structural, logical, electrical, procedural, and other changes. The embodiments represent only possible variations. Individual components and functions are optional unless explicitly required, and the order of operation may vary. Parts and features of some embodiments may be included in or replace parts and features of other embodiments. Moreover, the terminology used in this application is for describing embodiments only and is not intended to limit the claims. As used in the description of embodiments and claims, the singular forms “a,” “an,” and “the” are intended to equally include the plural forms unless the context clearly indicates otherwise. Similarly, the term “and / or” as used in this application means including one or more of the associated listed items and all possible combinations thereof. Additionally, when used in this application, the term "comprise" and its variations "comprises" and / or "comprising" refer to the presence of stated features, integrals, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof. Without further limitations, an element defined by the phrase "comprises a..." does not exclude the presence of other identical elements in the process, method, or apparatus that includes said element. In this document, each embodiment may focus on the differences from other embodiments, and similar or identical parts between embodiments can be referred to mutually. For methods, products, etc., disclosed in the embodiments, if they correspond to the method section disclosed in the embodiments, the relevant parts can be referred to the description of the method section.

[0136] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the embodiments of this disclosure. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0137] The methods and products (including but not limited to devices and equipment) disclosed in the embodiments herein can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of units may be merely a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the shown or discussed units may be through some interfaces, and the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected to implement this embodiment according to actual needs. Furthermore, the functional units in the embodiments of this disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0138] 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 embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing 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 that 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. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. Each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, can 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.

Claims

1. An energy saving control method for a refrigeration system, characterized by, The method comprises the following steps: obtaining a mechanism model associated with a water cooling system and a comfort model associated with an air cooling system; coupling the mechanism model and the comfort model to obtain a fitness function; determining boundary conditions and importing the boundary conditions into the fitness function to optimize parameters by a heuristic algorithm to obtain predicted parameters; outputting the predicted parameters to an execution end; wherein the obtaining of the mechanism model associated with the water cooling system comprises: constructing a mapping relationship between water cooling parameters of a cooling device and power of a refrigeration host to obtain a pool of benchmark mechanism models; inputting historical water cooling data of the cooling device into a benchmark mechanism model corresponding to the pool of benchmark mechanism models for training to obtain a candidate mechanism model; selecting the candidate mechanism model as the mechanism model if the candidate mechanism model meets an error accuracy; the obtaining of the comfort model associated with the air cooling system comprises: obtaining air cooling data of the air cooling system and obtaining comfort rating information, the air cooling data comprising part or all of set temperature of a unit, fresh air temperature, fresh air humidity, carbon dioxide concentration of fresh air, indoor temperature, indoor humidity and indoor carbon dioxide concentration; inputting the air cooling data and the comfort rating information into a benchmark comfort model to obtain a comfort model.

2. The method of claim 1, wherein, the coupling of the mechanism model and the comfort model to obtain the fitness function comprises: Computing ; wherein, respectively represent a comfort model and a mechanism model, respectively represent a thermal comfort constant and a total power constant, represents a fitness function, represents a correction coefficient.

3. The method of claim 1, wherein, the inputting of the historical water cooling data of the cooling device into the benchmark mechanism model corresponding to the pool of benchmark mechanism models for training to obtain the candidate mechanism model comprises: if the model training is not incremental training, extracting full data and test data from the historical water cooling data, and respectively training the full data and the test data by using the benchmark mechanism model corresponding to the pool of benchmark mechanism models; if the model training is incremental training, randomly selecting a target number of target historical water cooling data from the historical water cooling data and inputting the target historical water cooling data into the benchmark mechanism model corresponding to the pool of benchmark mechanism models for training.

4. The method of claim 1, wherein, the cooling device comprises a refrigeration host, a freezing pump, a cooling pump and a cooling tower, and the constructing of the mapping relationship between the water cooling parameters of the cooling device and the power of the refrigeration host to obtain the pool of benchmark mechanism models comprises: constructing a corresponding relationship between freezing water outlet temperature, cooling water inlet temperature and refrigerating capacity of the refrigeration host and power of the refrigeration host to obtain a first benchmark mechanism model; constructing a corresponding relationship between operating frequency of the freezing pump, operating frequency of the cooling pump, wind speed of the cooling tower, operating frequency of the cooling tower and outdoor environment temperature and humidity and power of the refrigeration host to obtain a second benchmark mechanism model; constructing a corresponding relationship between freezing flow of the freezing pump and freezing water temperature difference and refrigerating capacity of the refrigeration host to obtain a third benchmark mechanism model; constructing a corresponding relationship between cooling flow of the cooling pump and cooling water temperature difference and cooling capacity to obtain a fourth benchmark mechanism model.

5. The method of claim 4, wherein, the constructing of the mapping relationship between the water cooling parameters of the cooling device and the power of the refrigeration host to obtain the pool of benchmark mechanism models further comprises: constructing a corresponding relationship between operating frequency of the freezing pump, switch combination state of the freezing pump and freezing flow of the freezing pump to obtain a fifth benchmark mechanism model; constructing a corresponding relationship between operating frequency of the cooling pump, switch combination state of the cooling pump and cooling flow of the cooling pump to obtain a sixth benchmark mechanism model.

6. The method according to any one of claims 1 to 5, characterized in that, The determining boundary conditions and importing the boundary conditions to the fitness function to optimize the parameters by the heuristic algorithm to obtain the prediction parameters, comprising: According to the control parameters input by the user, obtaining the predicted load at the next time; Determining boundary conditions; According to the predicted load at the next time and the boundary conditions, optimizing the fitness function to obtain the prediction parameters.

7. The method of claim 6, wherein, The method comprises: In the case that the data volume of the cold data meets the data integrity condition, using a single-index time series prediction algorithm to predict the load at the next time; In the case that the data volume of the cold data does not meet the data integrity condition, using a deep neural network learning algorithm to predict the load at the next time; In the case that the data volume of the cold data is less than the upper limit threshold and greater than the lower limit threshold, and there is missing data, using data clustering to predict the load at the next time.

8. An energy saving control device for a refrigeration system comprising a processor and a memory having stored therein program instructions, wherein, The processor is configured to execute the energy-saving control method for the refrigeration system as claimed in any one of claims 1 to 7 when running the program instructions.

9. A refrigeration system characterized by, Comprise: A refrigeration host; A water cooling system; An air cooling system; And, The energy-saving control device for the refrigeration system as claimed in claim 8 is installed in the refrigeration host.

Images