Energy-saving optimization method for refrigeration room system based on hybrid model

Abstract

A method for energy-saving optimization of a refrigeration room system based on a hybrid model, the method comprising: constructing an energy-saving optimization tree and a component relationship graph based on the historical operation data and connection relationship of the refrigeration room system through a thermodynamic model. Randomly traverse the nodes in the component relationship graph, obtain the random traversal path of each node in the component relationship graph as the first node, and use the random traversal path as the feature vector of the first node. Call the convolution layer in the convolutional neural network to convolve the feature vector, call the feature extraction layer to extract high-frequency feature data from the convolution result, and call the pooling layer to calculate the component relationship feature representation of the refrigeration room system from the high-frequency feature data. Input the component relationship feature representation into the graph neural network, and call the graph neural network to calculate the corresponding feature graph matrix. Obtain the current operation data of the refrigeration room system, and output the energy-saving optimization strategy based on the current operation data and the feature graph matrix.

Description

Technical Field

[0001] The present invention belongs to the technical field of cooling energy saving in refrigeration rooms, and more specifically, relates to an energy saving optimization method for a refrigeration room system based on a hybrid model. Background Art

[0002] In recent years, with the rapid development of new generation information technology, refrigeration rooms, as important infrastructure, are the carriers of various applications such as big data, artificial intelligence, AR / VR, industrial Internet of Things, smart cities, smart energy, smart finance, 5G, etc., and their scale has experienced explosive growth. Since the energy consumption density of servers and ancillary equipment and facilities in refrigeration rooms is high and they run 24 hours a day, the energy consumption intensity is dozens or even hundreds of times higher than that of conventional public buildings, so energy saving in refrigeration rooms is particularly important.

[0003] Refrigeration rooms usually use cold sources in conjunction with precision air conditioning systems to cool servers in the room. The cold source system provides cooling capacity, and after heat exchange with the precision air conditioning system, cold air is provided. The cold air is then used to exchange heat with the server equipment to remove heat. However, traditional cooling technology has been difficult to meet people's increasing cooling and energy-saving needs. The traditional cooling adjustment of refrigeration room equipment in the room is based on the return air temperature to determine the start and stop operation of the equipment. However, relying solely on the return air temperature to determine the start and stop operation of the equipment is often inaccurate, resulting in energy waste. Therefore, energy-saving optimization of the refrigeration room system has become a technical problem that needs to be urgently solved in the current refrigeration room cooling. Summary of the invention

[0004] In order to solve the deficiencies in the prior art, the purpose of the present invention is to solve the above-mentioned defects and further propose an energy-saving optimization method for a refrigeration room system based on a hybrid model.

[0005] The present invention adopts the following technical solution.

[0006] The first aspect of the present invention discloses a refrigeration room system energy-saving optimization method based on a hybrid model, the method comprising:

[0007] Obtain historical operation data of the refrigeration room system and the connection relationship of each component, and construct an energy-saving optimization tree and a component relationship diagram based on the historical operation data and the connection relationship through a thermodynamic model;

[0008] Randomly traverse the nodes in the component relationship graph to obtain a random traversal path with each node in the component relationship graph as the first node, and use the random traversal path as the feature vector of the first node;

[0009] Calling a convolutional layer in a convolutional neural network to convolve the feature vector, and calling a feature extraction layer to extract high-frequency feature data from the convolution result, so as to call a pooling layer to calculate a component relationship feature representation of the refrigeration room system from the high-frequency feature data;

[0010] Inputting the component relationship feature representation into a graph neural network to call the graph neural network to calculate a feature graph matrix corresponding to the component relationship feature representation;

[0011] Acquire current operating data of the refrigeration room system, and output an energy-saving optimization strategy based on the current operating data and the characteristic graph matrix;

[0012] Among them, the various components of the refrigeration room system include a chiller, a cooling tower, an air handling unit and a cold water pump. The chiller is used to cool the chilled water to a set temperature, and supply the chilled water cooled to the set temperature to the air handling unit for refrigeration; the cooling tower is used to release the waste heat generated by the chiller into the atmosphere by evaporative cooling; the air handling unit is used to adjust the air temperature and air humidity through a hot water coil, and transport the treated air to the refrigeration room; the cold water pump is used to drive the chilled water to circulate between the chiller, the cooling tower and the air handling unit.

[0013] Furthermore, the acquisition of historical operation data of the refrigeration room system and the connection relationship of each component, and the construction of an energy-saving optimization tree and a component relationship diagram based on the historical operation data and the connection relationship through a thermodynamic model, include:

[0014] Acquire the mechanism framework of the refrigeration room system, and construct a thermodynamic model of the refrigeration room system based on the mechanism framework, wherein the thermodynamic model includes a chiller model, a cooling tower model, and an air handling unit model;

[0015] The thermodynamic model expression of the chiller model is:

[0016] ;

[0017] In the formula, is the refrigerant flow rate, They represent the inlet specific enthalpy and outlet specific enthalpy of the refrigerant respectively, represents the chilled water flow rate, Respectively represent the inlet and outlet temperatures of chilled water. is the energy efficiency ratio of the chiller, is the specific heat capacity.

[0018] Furthermore, the thermodynamic model expression of the cooling tower model is:

[0019] ;

[0020] In the formula, represents the heat transfer coefficient of the cooling tower, represents the heat transfer area of ​​the cooling tower, Represent the cooling water temperature and air temperature respectively, Indicates the cooling water flow rate, Respectively represent the inlet temperature and outlet temperature of cooling water, is the energy efficiency ratio of the chiller;

[0021] The thermodynamic model expression of the air handling unit model is:

[0022] ;

[0023] In the formula, is the flow rate of the cooling water coil, are the water inlet temperature and water outlet temperature of the cooling water coil, Indicates the air flow rate, is the specific heat capacity of air, are the return air temperature and the supply air temperature, is the heat transfer coefficient of the coil, is the heat transfer area of ​​the coil, Indicates the cold water temperature in the coil.

[0024] Furthermore, the acquisition of historical operation data of the refrigeration room system and the connection relationship of each component, and the construction of an energy-saving optimization tree and a component relationship diagram based on the historical operation data and the connection relationship through a thermodynamic model, also includes:

[0025] The energy-saving optimization tree is constructed based on the historical operation data and the connection relationship between the components through the thermodynamic model, and each node in the energy-saving optimization tree is mapped to a physical quantity node;

[0026] Among them, the nodes in the energy-saving optimization tree include top-level nodes, intermediate nodes and bottom-level nodes. The top-level node is mapped to the total energy consumption of the refrigeration room system, which is used to characterize the energy-saving target of the refrigeration room system. The intermediate node is mapped to the energy consumption of the chiller, cooling tower and air handling unit, which is used to characterize the energy-saving target of each component system. The bottom-level node is mapped to the energy-saving strategy of each component system.

[0027] Furthermore, the acquisition of historical operation data of the refrigeration room system and the connection relationship of each component, and the construction of an energy-saving optimization tree and a component relationship diagram based on the historical operation data and the connection relationship through a thermodynamic model, also includes:

[0028] The component relationship diagram is constructed based on the historical operation data and the connection relationship of each component by the thermodynamic model, each node in the component relationship diagram is mapped to a physical quantity node, and the nodes in the component relationship diagram are the same as those in the energy-saving optimization tree;

[0029] The mechanism framework includes a fluid mechanics model and an air enthalpy change model, and the fluid mechanics model expression of the chiller is:

[0030] ;

[0031] In the formula, is the friction pressure drop of chilled water along the way, represents the friction coefficient on the chilled water side, is the length of the chilled water pipeline, is the chilled water pipe diameter, is the density of water, is the chilled water flow rate;

[0032] The air enthalpy change model expression of the air handling unit model is:

[0033] ;

[0034] In the formula, are the inlet and outlet specific enthalpies of the chiller coil, is the latent heat of vaporization of water, are the inlet air humidity and outlet air humidity of the chiller coil, respectively.

[0035] Furthermore, the randomly traversing nodes in the component relationship graph to obtain a random traversal path with each node in the component relationship graph as the first node, and using the random traversal path as the feature vector of the first node, includes:

[0036] Taking any node in the component relationship graph as the first node, randomly traversing the next node based on the first node, and traversing until the length of the nodes in the random traversal path reaches a first threshold, the random traversal path is used as the feature vector corresponding to the first node;

[0037] The probability of random traversal depends on the weight of the edge in the component relationship graph. The first node is taken as the first node, and the second node is taken as the next node of the random traversal of the first node. The probability expression of random traversal from the first node to the second node is:

[0038] ;

[0039] In the formula, a is the first node, b is the second node, is the probability of random traversal from the first node to the second node, and are the weights of the edge from the first node a to the second node b, and the weights of the edge from the first node a to the node i, respectively. Indicates the total number of nodes in the energy-saving optimization tree.

[0040] Furthermore, the calculation expression of the feature map matrix is:

[0041] ,

[0042] ;

[0043] In the formula, For the component relationship diagram The transposition of the component relationship characteristics corresponding to the physical quantity, is the transpose symbol, , is the quantity of all physical quantities in the component relationship diagram, is the adjacency matrix composed of adjacency vectors, is the diagonal matrix corresponding to the adjacency matrix, is an activation function, wherein the activation function includes a sigmoid function and a ReLu function, For the The feature map matrix of the layer, , Indicates The weight matrix of the layer, are the model parameters of the graph neural network;

[0044] The adjacency matrix is ​​calculated by the energy-saving optimization tree, and the expression of the adjacency matrix is:

[0045] ,

[0046] ,

[0047] ;

[0048] In the formula, and Respectively represent the The adjacent vectors corresponding to the nodes and the elements therein, Indicates the first The distance from a node to a vertex, Indicates the first The node is in the component relationship graph with the The distance between nodes, express The maximum value in .

[0049] The second aspect of the present invention discloses a refrigeration room system energy-saving optimization device based on a hybrid model, the device comprising:

[0050] A thermodynamic model building module is used to obtain historical operation data of the refrigeration room system and the connection relationship of each component, and to build an energy-saving optimization tree and a component relationship diagram based on the historical operation data and the connection relationship through a thermodynamic model;

[0051] A node traversal module, used for randomly traversing the nodes in the component relationship graph to obtain a random traversal path with each node in the component relationship graph as the first node, and using the random traversal path as the feature vector of the first node;

[0052] A component relationship feature representation module, used to call the convolution layer in the convolutional neural network to convolve the feature vector, and call the feature extraction layer to extract high-frequency feature data from the convolution result, so as to call the pooling layer to calculate the component relationship feature representation of the refrigeration room system from the high-frequency feature data;

[0053] A feature graph matrix calculation module, used to input the component relationship feature representation into a graph neural network, so as to call the graph neural network to calculate a feature graph matrix corresponding to the component relationship feature representation;

[0054] An energy-saving strategy output module, used to obtain the current operating data of the refrigeration room system, and output an energy-saving optimization strategy according to the current operating data and the characteristic map matrix;

[0055] Among them, the various components of the refrigeration room system include a chiller, a cooling tower, an air handling unit and a cold water pump. The chiller is used to cool the chilled water to a set temperature, and supply the chilled water cooled to the set temperature to the air handling unit for refrigeration; the cooling tower is used to release the waste heat generated by the chiller into the atmosphere by evaporative cooling; the air handling unit is used to adjust the air temperature and air humidity through a hot water coil, and transport the treated air to the refrigeration room; the cold water pump is used to drive the chilled water to circulate between the chiller, the cooling tower and the air handling unit.

[0056] The third aspect of the present invention discloses a terminal, including a processor and a storage medium; the characteristics are:

[0057] The storage medium is used to store instructions;

[0058] The processor is used to operate according to the instructions to execute the steps of the method described in the first aspect.

[0059] The fourth aspect of the present invention discloses a computer-readable storage medium having a computer program stored thereon, characterized in that when the program is executed by a processor, the steps of the method described in the first aspect are implemented.

[0060] The beneficial effects of the present invention are that, compared with the prior art, the present invention has the following advantages:

[0061] By obtaining the historical operation data of the refrigeration room system and the connection relationship of each component, and constructing an energy-saving optimization tree and a component relationship graph based on the historical operation data and the connection relationship through a thermodynamic model. Then randomly traverse the nodes in the component relationship graph to obtain the random traversal path of each node in the component relationship graph as the first node, and use the random traversal path as the feature vector of the first node. Subsequently, call the convolution layer in the convolutional neural network to convolve the feature vector, call the feature extraction layer to extract high-frequency feature data from the convolution result, and then call the pooling layer to calculate the component relationship feature representation of the refrigeration room system from the high-frequency feature data. Input the component relationship feature representation into the graph neural network, and call the graph neural network to calculate the feature graph matrix corresponding to the component relationship feature representation. Finally, obtain the current operation data of the refrigeration room system, and output the energy-saving optimization strategy based on the current operation data combined with the feature graph matrix. This method can flexibly perform intelligent regulation of the energy-saving optimization strategy according to the operation data of the refrigeration room system, improve the accuracy of the start-stop operation judgment of the refrigeration room equipment, and reduce energy waste. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 It is a flow chart of a refrigeration room system energy-saving optimization method based on a hybrid model provided by the present invention;

[0063] Figure 2 It is a schematic diagram of the overall architecture of energy-saving optimization of the refrigeration room system in the specific embodiment provided by the present invention. DETAILED DESCRIPTION

[0064] The present application is further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present application.

[0065] like Figure 1 As shown, in one embodiment, a refrigeration room system energy-saving optimization method based on a hybrid model includes the following steps:

[0066] Step S110, obtaining historical operation data of the refrigeration room system and the connection relationship of each component, and constructing an energy-saving optimization tree and a component relationship diagram based on the historical operation data and the connection relationship through a thermodynamic model.

[0067] Among them, the components of the refrigeration room system include chillers, cooling towers, air handling units and chilled water pumps. The chillers are used to cool the chilled water to a set temperature and supply the chilled water to the air handling unit for refrigeration. The cooling tower is used to release the waste heat generated by the chiller into the atmosphere by evaporative cooling. The air handling unit is used to adjust the air temperature and air humidity through the hot water coil and transport the treated air to the refrigeration room. The chilled water pump is used to drive the chilled water to circulate between the chiller, cooling tower and air handling unit.

[0068] It should be noted that historical operation data refers to various data recorded by the refrigeration room system during its past operation, including physical quantity data such as temperature, flow, specific enthalpy, energy efficiency ratio, etc. involved in the thermodynamic model.

[0069] In some embodiments, the energy-saving optimization method for a refrigeration room system based on a hybrid model provided by the present invention obtains historical operation data of the refrigeration room system and the connection relationship of each component, and constructs an energy-saving optimization tree and a component relationship diagram based on the historical operation data and the connection relationship through a thermodynamic model, specifically including the following steps:

[0070] Step S111, obtaining a mechanism framework of a refrigeration room system, and constructing a thermodynamic model of the refrigeration room system based on the mechanism framework, wherein the thermodynamic model includes a chiller model, a cooling tower model, and an air handling unit model.

[0071] Among them, the thermodynamic model expression of the chiller model is:

[0072] .

[0073] In the formula, is the refrigerant flow rate, They represent the inlet specific enthalpy and outlet specific enthalpy of the refrigerant respectively, represents the chilled water flow rate, Respectively represent the inlet and outlet temperatures of chilled water. is the energy efficiency ratio of the chiller, is the specific heat capacity.

[0074] In some embodiments, the energy-saving optimization method for a refrigeration room system based on a hybrid model provided by the present invention, the thermodynamic model expression of the cooling tower model is:

[0075] .

[0076] In the formula, represents the heat transfer coefficient of the cooling tower, represents the heat transfer area of ​​the cooling tower, Represent the cooling water temperature and air temperature respectively, Indicates the cooling water flow rate, Respectively represent the inlet temperature and outlet temperature of cooling water, is the energy efficiency ratio of the chiller.

[0077] The thermodynamic model expression of the air handling unit model is:

[0078] .

[0079] In the formula, is the flow rate of the cooling water coil, are the water inlet temperature and water outlet temperature of the cooling water coil, Indicates the air flow rate, is the specific heat capacity of air, are the return air temperature and the supply air temperature, is the heat transfer coefficient of the coil, is the heat transfer area of ​​the coil, Indicates the cold water temperature in the coil.

[0080] In some embodiments, the energy-saving optimization method for a refrigeration room system based on a hybrid model provided by the present invention obtains historical operation data of the refrigeration room system and the connection relationship of each component, and constructs an energy-saving optimization tree and a component relationship diagram based on the historical operation data and the connection relationship through a thermodynamic model, and specifically includes the following steps:

[0081] Step S112, constructing an energy-saving optimization tree based on historical operation data and the connection relationship between each component through a thermodynamic model, and each node in the energy-saving optimization tree is mapped to a physical quantity node.

[0082] Among them, the nodes in the energy-saving optimization tree include top-level nodes, middle nodes and bottom-level nodes. The top-level node is mapped to the total energy consumption of the refrigeration room system, which is used to characterize the energy-saving target of the refrigeration room system. The middle node is mapped to the energy consumption of the chiller, cooling tower and air handling unit, which is used to characterize the energy-saving target of each component system. The bottom-level node is mapped to the energy-saving strategy of each component system.

[0083] In some embodiments, the energy-saving optimization method for a refrigeration room system based on a hybrid model provided by the present invention obtains historical operation data of the refrigeration room system and the connection relationship of each component, and constructs an energy-saving optimization tree and a component relationship diagram based on the historical operation data and the connection relationship through a thermodynamic model, and specifically includes the following steps:

[0084] Step S113, constructing a component relationship diagram based on historical operation data and the connection relationship of each component through a thermodynamic model, each node in the component relationship diagram is mapped to a physical quantity node, and the nodes in the component relationship diagram are the same as those in the energy-saving optimization tree.

[0085] Among them, the mechanism framework includes the fluid mechanics model and the air enthalpy change model. The fluid mechanics model expression of the chiller is:

[0086] .

[0087] In the formula, is the friction pressure drop of chilled water along the way, represents the friction coefficient on the chilled water side, is the length of the chilled water pipeline, is the chilled water pipe diameter, is the density of water, is the chilled water flow rate.

[0088] The air enthalpy change model expression of the air handling unit model is:

[0089] .

[0090] In the formula, are the inlet and outlet specific enthalpies of the chiller coil, is the latent heat of vaporization of water, are the inlet air humidity and outlet air humidity of the chiller coil, respectively.

[0091] Step S120, randomly traversing the nodes in the component relationship graph to obtain a random traversal path with each node in the component relationship graph as the first node, and using the random traversal path as the feature vector of the first node.

[0092] In some embodiments, the energy-saving optimization method for a refrigeration room system based on a hybrid model provided by the present invention randomly traverses the nodes in the component relationship graph to obtain a random traversal path with each node in the component relationship graph as the first node, and uses the random traversal path as the feature vector of the first node, specifically including the following steps:

[0093] Step S121, taking any node in the component relationship graph as the first node, randomly traversing the next node based on the first node, and when the length of the nodes in the random traversal path reaches a first threshold, the random traversal path is used as the feature vector corresponding to the first node.

[0094] The probability of random traversal depends on the weight of the edge in the component relationship graph. Taking the first node as the first node and the second node as the next node of the random traversal of the first node, the probability expression of random traversal from the first node to the second node is:

[0095] .

[0096] In the formula, a is the first node, b is the second node, is the probability of random traversal from the first node to the second node, and are the weights of the edge from the first node a to the second node b, and the weights of the edge from the first node a to the node i, respectively. Indicates the total number of nodes in the energy-saving optimization tree.

[0097] Step S130, calling the convolution layer in the convolutional neural network to convolve the feature vector, and calling the feature extraction layer to extract high-frequency feature data from the convolution result, and calling the pooling layer to calculate the component relationship feature representation of the refrigeration room system from the high-frequency feature data.

[0098] Step S140, input the component relationship feature representation into the graph neural network to call the graph neural network to calculate the feature graph matrix corresponding to the component relationship feature representation.

[0099] In some embodiments, in the refrigeration room system energy-saving optimization method based on the hybrid model provided by the present invention, the calculation expression of the characteristic graph matrix is:

[0100] ,

[0101] .

[0102] In the formula, For the component relationship diagram The transposition of the component relationship characteristics corresponding to the physical quantity, is the transpose symbol, , is the quantity of all physical quantities in the component relationship diagram, is the adjacency matrix composed of adjacency vectors, is the diagonal matrix corresponding to the adjacency matrix, is the activation function, which includes sigmoid function and ReLu function. For the The feature map matrix of the layer, , Indicates The weight matrix of the layer, are the model parameters of the graph neural network.

[0103] Among them, the adjacency matrix is ​​calculated by the energy-saving optimization tree, and the expression of the adjacency matrix is:

[0104] ,

[0105] ,

[0106] .

[0107] In the formula, and Respectively represent the The adjacent vectors corresponding to the nodes and the elements therein, Indicates the first The distance from a node to a vertex, Indicates the first The node is in the component relationship graph with the The distance between nodes, express The maximum value in .

[0108] Step S150, obtaining the current operating data of the refrigeration room system, and outputting an energy-saving optimization strategy according to the current operating data and the characteristic graph matrix.

[0109] Combination Figure 2 As shown, in a specific embodiment, the present invention provides a refrigeration room system energy-saving optimization method based on a hybrid model. The refrigeration room is an important part of the modern information infrastructure, and its energy consumption problem is becoming increasingly prominent. In order to solve the high energy consumption problem of the refrigeration room, its refrigeration room system (including air conditioning, servers, power management, etc.) must be energy-efficiently optimized.

[0110] In this embodiment, steps 1 to 4 are included:

[0111] Step 1: Based on the historical operation data of the refrigeration room system and the connection relationship of components, an energy-saving optimization tree and a component relationship diagram are constructed.

[0112] Among them, the refrigeration room system mainly includes chillers, cooling towers, air handling units and cold water pumps.

[0113] Chiller: Cools the chilled water to the set temperature and supplies it to the air handling unit for cooling.

[0114] Cooling tower: Releases waste heat from the chiller into the atmosphere through evaporative cooling.

[0115] Air handling unit: adjusts the air temperature and humidity through hot and cold water coils, and sends the treated air to the refrigeration room.

[0116] Chilled water pump: drives the chilled water to circulate between the chiller, cooling tower and air handling unit.

[0117] Specifically, according to the thermodynamic model, an energy-saving optimization tree and a component relationship diagram are constructed, including steps S11 to S13:

[0118] Step S11, based on the mechanism framework of the refrigeration room system, construct a corresponding thermodynamic model, including a chiller model, a cooling tower model and an air handling unit model.

[0119] Among them, the thermodynamic model expression of the chiller model is:

[0120] .

[0121] In the formula, is the refrigerant flow rate, They represent the inlet specific enthalpy and outlet specific enthalpy of the refrigerant respectively, represents the chilled water flow rate, Respectively represent the inlet and outlet temperatures of chilled water. is the energy efficiency ratio of the chiller, is the specific heat capacity.

[0122] The thermodynamic model expression of the cooling tower model is:

[0123] .

[0124] In the formula, represents the heat transfer coefficient of the cooling tower, represents the heat transfer area of ​​the cooling tower, Represent the cooling water temperature and air temperature respectively, Indicates the cooling water flow rate, Respectively represent the inlet temperature and outlet temperature of cooling water, is the energy efficiency ratio of the chiller.

[0125] The thermodynamic model expression of the air handling unit model is:

[0126] .

[0127] In the formula, is the flow rate of the cooling water coil, are the water inlet temperature and water outlet temperature of the cooling water coil, Indicates the air flow rate, is the specific heat capacity of air, are the return air temperature and the supply air temperature, is the heat transfer coefficient of the coil, is the heat transfer area of ​​the coil, Indicates the cold water temperature in the coil.

[0128] Therefore, the total energy consumption expression is:

[0129] .

[0130] In the formula, Respectively represent the energy consumption of chillers, cooling towers and air units, is the total energy consumption.

[0131] It should be noted that historical operation data refers to various data recorded by the refrigeration room system during its past operation, including physical quantity data such as temperature, flow, specific enthalpy, energy efficiency ratio, etc. involved in the thermodynamic model.

[0132] Step S12, constructing an energy-saving optimization tree, in which each node is mapped to a physical quantity node, such as the outlet specific enthalpy of the refrigerant, the chilled water flow rate, the return air temperature, etc.

[0133] Specifically, all nodes on the energy-saving optimization tree are divided into three layers, that is, the top-level node is mapped to the total energy consumption, indicating the overall energy-saving target of the refrigeration room; the middle node represents the optimization target of each subsystem, mapped to the energy consumption of the chiller, cooling tower and air unit; the bottom-level node represents the specific optimization strategy for each subsystem, such as adjusting the temperature of the chilled water, optimizing the cooling water flow, adjusting the air temperature of the air handling unit, etc.

[0134] Step S13, constructing a component relationship diagram, where each node in the component relationship diagram is mapped to a physical quantity node.

[0135] The mechanism framework also includes indirect factors that affect energy consumption, such as fluid mechanics models, air enthalpy change models, etc.

[0136] Among them, the fluid mechanics model expression of the chiller is:

[0137] .

[0138] In the formula, is the friction pressure drop of chilled water along the way, represents the friction coefficient on the chilled water side, is the length of the chilled water pipeline, is the chilled water pipe diameter, is the density of water, is the chilled water flow rate.

[0139] The air enthalpy change model expression of the air handling unit model is:

[0140] .

[0141] In the formula, are the inlet and outlet specific enthalpies of the chiller coil, is the latent heat of vaporization of water, are the inlet air humidity and outlet air humidity of the chiller coil, respectively.

[0142] It should be noted that the nodes in the energy-saving optimization tree are a subset of the nodes in the component relationship diagram. The nodes in the energy-saving optimization tree only contain physical quantities that directly affect energy-saving optimization, that is, the physical quantities involved in the thermodynamic model, while the component relationship diagram also includes physical quantities involved in indirect factors affecting energy consumption.

[0143] In this embodiment, the bottom nodes can be further limited based on the expressions in the mechanism framework. Taking the thermodynamic model of the air handling unit model as an example, the middle nodes are mapped to , i.e., the energy consumption of the air unit, and the underlying nodes can be limited based on expressions, such as Figure 2 As shown in the dashed box in The weight of , while the edge The weight of ; and the edge The weights are all 1.

[0144] It should be noted that the above models (e.g. thermodynamic model, fluid dynamics model, air enthalpy change model) are all idealized models. In practical applications, the energy-saving optimization process of the refrigeration room system must comprehensively consider a series of practical factors, including the specifications and performance of the hardware equipment, the aging degree of the equipment, the failure rate and the operating status. In addition, it is also necessary to consider the impact of external environmental factors such as temperature, humidity, wind speed, etc. on system performance. These variables may affect the prediction accuracy and optimization effect of the model.

[0145] Therefore, in order to more comprehensively evaluate and optimize the energy efficiency of the system, a deep learning network can be combined to analyze these actual data, extract key features, and build a more accurate prediction model. This can not only enhance the model's adaptability to complex real-world scenarios, but also achieve more efficient resource allocation and fault warning, thereby maximizing the energy-saving optimization effect of the refrigeration room in actual operation.

[0146] In the component relationship diagram, the weight from node a to node b represents the association between the two, which can be obtained by simulating the finite element model of the refrigeration room system. For example, in the finite element model, the physical quantities corresponding to the above nodes a and b are set, and then the physical quantity corresponding to node a is adjusted, and it is observed whether the physical quantities corresponding to other nodes will change accordingly. For example, the change rate of node b can be used as the weight from node a to node b.

[0147] Step 2: Determine the component relationship feature representation by randomly traversing the nodes in the component relationship graph.

[0148] Specifically, it includes steps 2.1~2.2:

[0149] Step 2.1, for each node in the component relationship graph at each moment, execute: take any node as the first node of the path, and continue to randomly traverse the next node until the total node length in the path is End, and use the path as the feature vector corresponding to the first node.

[0150] The probability of random traversal depends on the weight of the edge in the component relationship graph. The probability of random traversal from node a to node b is The expression is:

[0151] .

[0152] In the formula, and are the weights of the edge from node a to node b, and the weights of the edge from node a to node i, respectively. Indicates the total number of nodes in the energy-saving optimization tree.

[0153] It should be noted that each node It can be expressed as , Indicates time Next Node Historical operating data, , The number of times in the historical data are collected.

[0154] Step 2.2, input the feature vector into the convolutional neural network for training.

[0155] Specifically, the convolutional neural network includes: a convolution layer, a feature extraction layer, and a pooling layer. The convolution layer is used to convolve the feature vector, the feature extraction layer is used to extract high-frequency feature data from the convolution result, and the pooling layer is used to calculate the component relationship feature representation from the high-frequency feature data.

[0156] Step 3: Substitute the component relationship feature representation into the graph neural network and calculate the feature graph matrix.

[0157] Among them, graph neural network can be used as a The convolutional model of the layer, is the number of graph convolutional layers, if The larger the value of , the better the training effect and the longer the iteration time.

[0158] The expression for calculating the feature map matrix is:

[0159] ,

[0160] .

[0161] In the formula, For the component relationship diagram The transposition of the component relationship characteristics corresponding to the physical quantity, is the transpose symbol, , It is the quantity of all physical quantities in the component relationship diagram. is the adjacency matrix composed of adjacency vectors, is the diagonal matrix corresponding to the adjacency matrix, and its diagonal values express Middle The number of elements in the row whose value is greater than 0.5, is the activation function, which can be a sigmoid function or a ReLu function. For the The feature map matrix of the layer, , Indicates The weight matrix of the layer is obtained through graph neural network training. As the model parameters of the graph neural network, they can be continuously optimized through, for example, the back-propagation algorithm.

[0162] Among them, the adjacency matrix is ​​calculated by the energy-saving optimization tree, and the expression of the adjacency matrix is:

[0163] ,

[0164] ,

[0165] .

[0166] In the formula, and Respectively represent the The adjacent vectors corresponding to the nodes and the elements therein, Indicates the first The distance from a node to a vertex, Indicates the first The node is in the component relationship graph with the The distance between nodes, express The maximum value in .

[0167] Step 4: Based on the feature map matrix and combined with the current operating data, output the energy-saving optimization strategy.

[0168] Specifically, the energy-saving optimization strategy is the operating data corresponding to the refrigeration room system after energy-saving optimization, including physical quantity data such as temperature, flow, specific enthalpy, energy efficiency ratio, etc. involved in the thermodynamic model, and also includes specific optimization strategies for each subsystem in the underlying node, such as adjusting the temperature of chilled water, optimizing the cooling water flow, and adjusting the air temperature delivered by the air handling unit.

[0169] The following is a description of the refrigeration room system energy-saving optimization device based on the hybrid model provided by the present invention. The refrigeration room system energy-saving optimization device based on the hybrid model described below and the refrigeration room system energy-saving optimization method based on the hybrid model described above can refer to each other.

[0170] In one embodiment, a hybrid model-based refrigeration room system energy-saving optimization device includes a thermodynamic model building module, a node traversal module, a component relationship feature representation module, a feature graph matrix calculation module, and an energy-saving strategy output module.

[0171] The thermodynamic model building module is used to obtain the historical operation data of the refrigeration room system and the connection relationship of each component, and to build an energy-saving optimization tree and a component relationship diagram based on the historical operation data and connection relationship through the thermodynamic model.

[0172] The node traversal module is used to randomly traverse the nodes in the component relationship graph to obtain a random traversal path with each node in the component relationship graph as the first node, and use the random traversal path as the feature vector of the first node.

[0173] The component relationship feature representation module is used to call the convolution layer in the convolutional neural network to convolve the feature vector, and call the feature extraction layer to extract high-frequency feature data from the convolution result, so as to call the pooling layer to calculate the component relationship feature representation of the refrigeration room system from the high-frequency feature data.

[0174] The feature graph matrix calculation module is used to input the component relationship feature representation into the graph neural network, so as to call the graph neural network to calculate the feature graph matrix corresponding to the component relationship feature representation.

[0175] The energy-saving strategy output module is used to obtain the current operating data of the refrigeration room system, and output the energy-saving optimization strategy based on the current operating data and the characteristic map matrix.

[0176] Among them, the components of the refrigeration room system include chillers, cooling towers, air handling units and chilled water pumps. The chillers are used to cool the chilled water to a set temperature and supply the chilled water to the air handling unit for refrigeration. The cooling tower is used to release the waste heat generated by the chiller into the atmosphere by evaporative cooling. The air handling unit is used to adjust the air temperature and air humidity through the hot water coil and transport the treated air to the refrigeration room. The chilled water pump is used to drive the chilled water to circulate between the chiller, cooling tower and air handling unit.

[0177] In this embodiment, the energy-saving optimization device for a refrigeration room system based on a hybrid model provided by the present invention, the thermodynamic model building module is specifically used for:

[0178] The mechanism framework of the refrigeration room system is obtained, and a thermodynamic model of the refrigeration room system is constructed based on the mechanism framework. The thermodynamic model includes a chiller model, a cooling tower model, and an air handling unit model.

[0179] Among them, the thermodynamic model expression of the chiller model is:

[0180] .

[0181] In the formula, is the refrigerant flow rate, They represent the inlet specific enthalpy and outlet specific enthalpy of the refrigerant respectively, represents the chilled water flow rate, Respectively represent the inlet and outlet temperatures of chilled water. is the energy efficiency ratio of the chiller, is the specific heat capacity.

[0182] In this embodiment, the energy-saving optimization device for a refrigeration room system based on a hybrid model provided by the present invention has a thermodynamic model expression of a cooling tower model as follows:

[0183] .

[0184] In the formula, represents the heat transfer coefficient of the cooling tower, represents the heat transfer area of ​​the cooling tower, Represent the cooling water temperature and air temperature respectively, Indicates the cooling water flow rate, Respectively represent the inlet temperature and outlet temperature of cooling water, is the energy efficiency ratio of the chiller.

[0185] The thermodynamic model expression of the air handling unit model is:

[0186] .

[0187] In the formula, is the flow rate of the cooling water coil, are the water inlet temperature and water outlet temperature of the cooling water coil, Indicates the air flow rate, is the specific heat capacity of air, are the return air temperature and the supply air temperature, is the heat transfer coefficient of the coil, is the heat transfer area of ​​the coil, Indicates the cold water temperature in the coil.

[0188] In this embodiment, the energy-saving optimization device for a refrigeration room system based on a hybrid model provided by the present invention, the thermodynamic model building module is specifically used for:

[0189] An energy-saving optimization tree is constructed through a thermodynamic model based on historical operating data and the connection relationship between each component. Each node in the energy-saving optimization tree is mapped to a physical quantity node.

[0190] Among them, the nodes in the energy-saving optimization tree include top-level nodes, middle nodes and bottom-level nodes. The top-level node is mapped to the total energy consumption of the refrigeration room system, which is used to characterize the energy-saving target of the refrigeration room system. The middle node is mapped to the energy consumption of the chiller, cooling tower and air handling unit, which is used to characterize the energy-saving target of each component system. The bottom-level node is mapped to the energy-saving strategy of each component system.

[0191] In this embodiment, the energy-saving optimization device for a refrigeration room system based on a hybrid model provided by the present invention, the thermodynamic model building module is specifically used for:

[0192] A component relationship diagram is constructed based on the historical operating data and the connection relationship of each component through a thermodynamic model. Each node in the component relationship diagram is mapped to a physical quantity node, and the nodes in the component relationship diagram are the same as those in the energy-saving optimization tree.

[0193] Among them, the mechanism framework includes the fluid mechanics model and the air enthalpy change model. The fluid mechanics model expression of the chiller is:

[0194] .

[0195] In the formula, is the friction pressure drop of chilled water along the way, represents the friction coefficient on the chilled water side, is the length of the chilled water pipeline, is the chilled water pipe diameter, is the density of water, is the chilled water flow rate.

[0196] The air enthalpy change model expression of the air handling unit model is:

[0197] .

[0198] In the formula, are the inlet and outlet specific enthalpies of the chiller coil, is the latent heat of vaporization of water, are the inlet air humidity and outlet air humidity of the chiller coil, respectively.

[0199] In this embodiment, the refrigeration room system energy-saving optimization device based on the hybrid model provided by the present invention, the node traversal module is specifically used for:

[0200] Take any node in the component relationship graph as the first node, randomly traverse the next node based on the first node, and when the length of the nodes in the random traversal path reaches a first threshold, take the random traversal path as the feature vector corresponding to the first node.

[0201] The probability of random traversal depends on the weight of the edge in the component relationship graph. Taking the first node as the first node and the second node as the next node of the random traversal of the first node, the probability expression of random traversal from the first node to the second node is:

[0202] .

[0203] In the formula, a is the first node, b is the second node, is the probability of random traversal from the first node to the second node, and are the weights of the edge from the first node a to the second node b, and the weights of the edge from the first node a to the node i, respectively. Indicates the total number of nodes in the energy-saving optimization tree.

[0204] In this embodiment, the energy-saving optimization device for a refrigeration room system based on a hybrid model provided by the present invention has a calculation expression of a characteristic graph matrix as follows:

[0205] ,

[0206] .

[0207] In the formula, For the component relationship diagram The transposition of the component relationship characteristics corresponding to the physical quantity, is the transpose symbol, , is the quantity of all physical quantities in the component relationship diagram, is the adjacency matrix composed of adjacency vectors, is the diagonal matrix corresponding to the adjacency matrix, is the activation function, which includes sigmoid function and ReLu function. For the The feature map matrix of the layer, , Indicates The weight matrix of the layer, are the model parameters of the graph neural network.

[0208] Among them, the adjacency matrix is ​​calculated by the energy-saving optimization tree, and the expression of the adjacency matrix is:

[0209] ,

[0210] ,

[0211] .

[0212] In the formula, and Respectively represent the The adjacent vectors corresponding to the nodes and the elements therein, Indicates the first The distance from a node to a vertex, Indicates the first The node is in the component relationship graph with the The distance between nodes, express The maximum value in .

[0213] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0214] A computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the above. More specific examples (a non-exhaustive list) of computer-readable storage media include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination of the above. The computer-readable storage medium used herein is not to be interpreted as a transient signal itself, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through a wire.

[0215] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0216] The computer program instructions for performing the operations of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer, partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect through the Internet). In some embodiments, by using the state information of the computer-readable program instructions to personalize an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), the electronic circuit may execute the computer-readable program instructions, thereby implementing various aspects of the present disclosure.

[0217] Various aspects of the present disclosure are described herein with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of boxes in the flowchart and / or block diagram can be implemented by computer-readable program instructions.

[0218] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, thereby producing a machine, so that when these instructions are executed by the processor of the computer or other programmable data processing device, a device that implements the functions / actions specified in one or more boxes in the flowchart and / or block diagram is generated. These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions cause the computer, programmable data processing device, and / or other equipment to work in a specific manner, so that the computer-readable medium storing the instructions includes a manufactured product, which includes instructions for implementing various aspects of the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0219] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operating steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to implement the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0220] The flow chart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to multiple embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a part of a module, program segment or instruction, and a part of the module, program segment or instruction includes one or more executable instructions for realizing the specified logical function. In some alternative implementations, the function marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two continuous square boxes can actually be executed substantially in parallel, and they can sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or action, or can be implemented with a combination of special hardware and computer instructions.

[0221] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A refrigeration room system energy-saving optimization method based on a hybrid model, characterized in that: The method comprises: Obtain historical operation data of the refrigeration room system and the connection relationship of each component, and construct an energy-saving optimization tree and a component relationship diagram based on the historical operation data and the connection relationship through a thermodynamic model; Randomly traverse the nodes in the component relationship graph to obtain a random traversal path with each node in the component relationship graph as the first node, and use the random traversal path as the feature vector of the first node; Calling a convolutional layer in a convolutional neural network to convolve the feature vector, and calling a feature extraction layer to extract high-frequency feature data from the convolution result, so as to call a pooling layer to calculate a component relationship feature representation of the refrigeration room system from the high-frequency feature data; Inputting the component relationship feature representation into a graph neural network to call the graph neural network to calculate a feature graph matrix corresponding to the component relationship feature representation; Acquire current operating data of the refrigeration room system, and output an energy-saving optimization strategy based on the current operating data and the characteristic graph matrix; The components of the refrigeration room system include a water chiller, a cooling tower, an air handling unit and a cold water pump. The water chiller is used to cool the chilled water to a set temperature and supply the chilled water cooled to the set temperature to the air handling unit for refrigeration; the cooling tower is used to release the waste heat generated by the water chiller into the atmosphere by evaporative cooling; the air handling unit is used to adjust the air temperature and air humidity through a hot water coil and transport the treated air to the refrigeration room; the cold water pump is used to drive the chilled water to circulate among the water chiller, the cooling tower and the air handling unit; The randomly traversing nodes in the component relationship graph to obtain a random traversal path with each node in the component relationship graph as the first node, and using the random traversal path as the feature vector of the first node, includes: Taking any node in the component relationship graph as the first node, randomly traversing the next node based on the first node, and traversing until the length of the nodes in the random traversal path reaches a first threshold, the random traversal path is used as the feature vector corresponding to the first node; The probability of random traversal depends on the weight of the edge in the component relationship graph. The first node is taken as the first node, and the second node is taken as the next node of the random traversal of the first node. The probability expression of random traversal from the first node to the second node is: ； In the formula, a is the first node, b is the second node, is the probability of random traversal from the first node to the second node, and are the weights of the edge from the first node a to the second node b, and the weights of the edge from the first node a to the node i, respectively. Represents the total number of nodes in the energy-saving optimization tree; The calculation expression of the feature map matrix is: ， ； In the formula, For the component relationship diagram The transposition of the component relationship characteristics corresponding to the physical quantity, is the transpose symbol, , is the quantity of all physical quantities in the component relationship diagram, is the adjacency matrix composed of adjacency vectors, is the diagonal matrix corresponding to the adjacency matrix, is an activation function, wherein the activation function includes a sigmoid function and a ReLu function, For the The feature map matrix of the layer, , Indicates The weight matrix of the layer, are the model parameters of the graph neural network; The adjacency matrix is ​​calculated by the energy-saving optimization tree, and the expression of the adjacency matrix is: ， ， ； In the formula, and Respectively represent the The adjacent vectors corresponding to the nodes and the elements therein, Indicates the first The distance from a node to a vertex, Indicates the first The node is in the component relationship graph with the The distance between nodes, express The maximum value in .

2. The energy-saving optimization method for a refrigeration room system based on a hybrid model according to claim 1 is characterized in that: The method of obtaining historical operation data of the refrigeration room system and the connection relationship between each component, and constructing an energy-saving optimization tree and a component relationship diagram based on the historical operation data and the connection relationship through a thermodynamic model, includes: Acquire the mechanism framework of the refrigeration room system, and construct a thermodynamic model of the refrigeration room system based on the mechanism framework, wherein the thermodynamic model includes a chiller model, a cooling tower model, and an air handling unit model; The thermodynamic model expression of the chiller model is: ； In the formula, is the refrigerant flow rate, They represent the inlet specific enthalpy and outlet specific enthalpy of the refrigerant respectively, represents the chilled water flow rate, Respectively represent the inlet and outlet temperatures of chilled water. is the energy efficiency ratio of the chiller, is the specific heat capacity.

3. The energy-saving optimization method for a refrigeration room system based on a hybrid model according to claim 2 is characterized in that: The thermodynamic model expression of the cooling tower model is: ； In the formula, represents the heat transfer coefficient of the cooling tower, represents the heat transfer area of ​​the cooling tower, Represent the cooling water temperature and air temperature respectively, Indicates the cooling water flow rate, Respectively represent the inlet temperature and outlet temperature of cooling water, is the energy efficiency ratio of the chiller; The thermodynamic model expression of the air handling unit model is: ； In the formula, is the flow rate of the cooling water coil, are the water inlet temperature and water outlet temperature of the cooling water coil, Indicates the air flow rate, is the specific heat capacity of air, are the return air temperature and the supply air temperature, is the heat transfer coefficient of the coil, is the heat transfer area of ​​the coil, Indicates the cold water temperature in the coil.

4. The energy-saving optimization method for a refrigeration room system based on a hybrid model according to claim 3 is characterized in that: The method of obtaining historical operation data of the refrigeration room system and the connection relationship of each component, and constructing an energy-saving optimization tree and a component relationship diagram based on the historical operation data and the connection relationship through a thermodynamic model, further includes: The energy-saving optimization tree is constructed based on the historical operation data and the connection relationship between the components through the thermodynamic model, and each node in the energy-saving optimization tree is mapped to a physical quantity node; Among them, the nodes in the energy-saving optimization tree include top-level nodes, intermediate nodes and bottom-level nodes. The top-level node is mapped to the total energy consumption of the refrigeration room system, which is used to characterize the energy-saving target of the refrigeration room system. The intermediate node is mapped to the energy consumption of the chiller, cooling tower and air handling unit, which is used to characterize the energy-saving target of each component system. The bottom-level node is mapped to the energy-saving strategy of each component system.

5. The energy-saving optimization method for a refrigeration room system based on a hybrid model according to claim 4 is characterized in that: The method of obtaining historical operation data of the refrigeration room system and the connection relationship of each component, and constructing an energy-saving optimization tree and a component relationship diagram based on the historical operation data and the connection relationship through a thermodynamic model, further includes: The component relationship diagram is constructed based on the historical operation data and the connection relationship of each component by the thermodynamic model, each node in the component relationship diagram is mapped to a physical quantity node, and the nodes in the component relationship diagram are the same as those in the energy-saving optimization tree; The mechanism framework includes a fluid mechanics model and an air enthalpy change model, and the fluid mechanics model expression of the chiller is: ； In the formula, is the friction pressure drop of chilled water along the way, represents the friction coefficient on the chilled water side, is the length of the chilled water pipeline, is the chilled water pipe diameter, is the density of water, is the chilled water flow rate; The air enthalpy change model expression of the air handling unit model is: ； In the formula, are the inlet and outlet specific enthalpies of the chiller coil, is the latent heat of vaporization of water, are the inlet air humidity and outlet air humidity of the chiller coil, respectively.

6. A refrigeration room system energy-saving optimization device based on a hybrid model, applied to any of the methods described in claims 1-5, characterized in that: The device comprises: A thermodynamic model building module is used to obtain historical operation data of the refrigeration room system and the connection relationship of each component, and to build an energy-saving optimization tree and a component relationship diagram based on the historical operation data and the connection relationship through a thermodynamic model; A node traversal module, used for randomly traversing the nodes in the component relationship graph to obtain a random traversal path with each node in the component relationship graph as the first node, and using the random traversal path as the feature vector of the first node; A component relationship feature representation module, used to call the convolution layer in the convolutional neural network to convolve the feature vector, and call the feature extraction layer to extract high-frequency feature data from the convolution result, so as to call the pooling layer to calculate the component relationship feature representation of the refrigeration room system from the high-frequency feature data; A feature graph matrix calculation module, used to input the component relationship feature representation into a graph neural network, so as to call the graph neural network to calculate a feature graph matrix corresponding to the component relationship feature representation; An energy-saving strategy output module, used to obtain the current operating data of the refrigeration room system, and output an energy-saving optimization strategy according to the current operating data and the characteristic map matrix; Among them, the various components of the refrigeration room system include a chiller, a cooling tower, an air handling unit and a cold water pump. The chiller is used to cool the chilled water to a set temperature, and supply the chilled water cooled to the set temperature to the air handling unit for refrigeration; the cooling tower is used to release the waste heat generated by the chiller into the atmosphere by evaporative cooling; the air handling unit is used to adjust the air temperature and air humidity through a hot water coil, and transport the treated air to the refrigeration room; the cold water pump is used to drive the chilled water to circulate between the chiller, the cooling tower and the air handling unit.

7. A terminal comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to execute the steps of the method according to any one of claims 1-5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images