Self-adaptive cold source system modeling method based on mechanism and data hybrid driving

Abstract

An adaptive cold source system modeling method based on mechanism and data hybrid driving comprises the following steps: S1, acquiring historical operation data of a cold source system, and preprocessing the historical operation data; s2, on the basis of the physical mechanism of the cold source system, constructing a mechanism model of equipment in the cold source system; s3, correcting key parameters of the mechanism model constructed in the step S2, and establishing a corrected model; and S4, based on the deviation between the model simulation output and the real system operation, adaptively updating the parameters of the correction model. According to the scheme, data dependence can be effectively relieved and reduced, and higher modeling precision and higher robustness are achieved.

Description

Technical Field

[0001] This invention relates to the field of cold source system modeling and simulation technology, and in particular to an adaptive cold source system modeling method based on a hybrid mechanism and data-driven approach. Background Technology

[0002] As a key component of buildings and industrial facilities, cooling systems play a crucial role in providing adequate cooling for indoor environments and ensuring stable equipment operation. With technological advancements and societal development, higher demands are being placed on the operational efficiency, stability, and energy conservation of cooling systems. Therefore, research into modeling methods for cooling systems has become a key approach to improving their performance. Accurate modeling allows for a deeper understanding of the operating characteristics of cooling systems, enabling the optimization of control strategies and achieving the goals of energy conservation, consumption reduction, and improved operational reliability.

[0003] Currently, research on cold source system modeling mainly focuses on combining artificial intelligence technology to improve the accuracy and adaptability of the models. However, in terms of adaptive adjustment, existing research largely relies on manually set thresholds or rules, making it difficult to automatically and accurately adjust based on the real-time operating status of the system. Therefore, there is an urgent need for a method to achieve high-precision, adaptive modeling of cold source systems, thereby providing a more effective solution for intelligent control and energy-saving operation of cold source systems. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an adaptive cold source system modeling method based on a hybrid mechanism and data-driven approach, which can effectively alleviate and reduce data dependence, and has higher modeling accuracy and stronger robustness.

[0005] The above-mentioned technical objective of the present invention is achieved through the following technical solution: An adaptive cold source system modeling method based on a hybrid mechanism and data-driven approach includes the following steps: S1, acquire historical operating data of the cold source system, and preprocess the historical operating data; S2, Based on the physical mechanism of the cold source system, construct the mechanism model of the equipment in the cold source system; S3, Modify the key parameters of the mechanism model constructed in step S2, and establish a modified model; S4, based on the deviation between the model simulation output and the actual system operation, adaptively updates the parameters of the corrected model.

[0006] Furthermore, the equipment in the cold source system in step S1 includes, but is not limited to: chiller units, cooling towers, chilled water pumps, and cooling water pumps; the historical operating data includes, but is not limited to: The chilled water outlet temperature, chilled water return temperature, cooling water outlet temperature, cooling water return temperature, chilled water flow rate, cooling water flow rate, and operating power of the chiller unit are specified. The flow rate, operating frequency, and operating power of the chilled water pump; The flow rate, operating frequency, and operating power of the cooling water pump; Cooling tower flow rate, inlet water temperature, outlet water temperature, number of fans in operation, power, and air volume; Indoor and outdoor ambient temperature and humidity.

[0007] Furthermore, when preprocessing the historical operating data, outliers are identified using the interquartile range method, isolated outliers are replaced with the mean of adjacent time periods, continuous outliers are replaced with historical data from the same period, and missing values ​​are filled in using linear interpolation.

[0008] Furthermore, in step S2, the mechanism model of the equipment in the cold source system is constructed, including at least: a chiller unit model, a variable frequency water pump model, and a cooling tower model.

[0009] Furthermore, a chiller unit model is constructed, specifically including: Based on the correction factor for rated cooling capacity and rated energy efficiency ratio correction factor Calculate the real-time cooling capacity of the chiller unit. and actual energy efficiency ratio The formula is:

[0010]

[0011] in, This refers to the cooling capacity under rated operating conditions. This refers to the energy efficiency ratio under rated operating conditions. Calculate the current load:

[0012] in, This represents the current load of the chiller unit. This refers to the chilled water flow rate. This refers to the inlet temperature of the chilled water. Set the outlet water temperature for the chilled water. The specific heat capacity of the fluid entering the chilled water pump; Compare current load and real-time cooling capacity , when ≤ Then determine the actual outlet temperature of the chilled water. It can achieve the set water temperature, that is = ; when > If the chiller unit cannot meet the required cooling capacity, then... = And recalculate the actual outlet temperature of the chilled water. and the actual outlet temperature of cooling water :

[0013]

[0014] This refers to the inlet temperature of the cooling water. For cooling water flow rate, Energy consumption of the chiller unit.

[0015] Furthermore, when constructing the variable frequency water pump model, The power consumption calculation formula for a variable frequency water pump is as follows:

[0016] in, This refers to the rated power of the variable frequency water pump. This is the power correction factor; The formula for calculating the pumping efficiency of a variable frequency water pump is:

[0017] in, This represents the overall efficiency of the variable frequency water pump. This represents the efficiency of the variable frequency motor; The formula for calculating the shaft power of a variable frequency water pump is:

[0018] The formula for calculating the energy transferred from the variable frequency water pump motor to the internal fluid is:

[0019] in, The proportion of waste heat generated due to low motor efficiency that is directly used to heat the fluid inside the pump. The formula for calculating the energy transferred from the variable frequency water pump motor to the surrounding air is:

[0020] The formula for calculating the outlet temperature of a variable frequency pump is:

[0021] in, The temperature of the fluid flowing out of the variable frequency water pump, The temperature of the fluid flowing into the variable frequency water pump. The mass of the fluid passing through the variable frequency water pump.

[0022] Furthermore, a cooling tower model is constructed, specifically including: Step 1: Calculate the capacity of the water flow. :

[0023] in, The specific heat capacity of the fluid entering the cooling tower, This refers to the flow rate of cooling water in the cooling tower. Calculate the simulated specific heat of air :

[0024] in, The enthalpy of the outlet saturated air. The enthalpy of the inlet saturated air. The inlet wet-bulb temperature, The outlet wet-bulb temperature; Calculate the volume of air :

[0025] in, The airflow rate in the cooling tower; Step Two: Determine parameters and :

[0026]

[0027]

[0028] Step 3: Calculate the simulated overall heat transfer coefficient of the cooling tower. :

[0029] in, To design the mass transfer coefficient, The specific heat of air; Step Four: Calculate the heat transfer unit coefficient : when When = 0, = 1×10¹⁵; otherwise

[0030] Step 5: Calculate heat transfer efficiency : when hour,

[0031] when hour, ,in,

[0032] Step Six: Calculate total heat transfer :

[0033] in, The inlet water temperature of the cooling tower. The inlet wet-bulb temperature of the cooling tower Calculated value of cooling tower outlet wet-bulb temperature

[0034]

[0035] The calculated value of the outlet wet-bulb temperature is compared with the value of the outlet wet-bulb temperature when calculating the simulated specific heat value of air. If the difference between the two is less than the preset value, the calculation is considered to have converged. Otherwise, the calculated value of the outlet wet-bulb temperature is used as the new value of the outlet wet-bulb temperature, and the process of steps one to six is ​​repeated until the calculation converges. Step Seven: Calculate the final cooling tower outlet water temperature :

[0036] Furthermore, in step S3, for the chiller unit model correction, a limit learning machine is used to adjust the rated cooling capacity correction coefficient. and rated energy efficiency ratio correction factor Regression prediction is performed with the input features being the chilled water outlet temperature and the cooling water inlet temperature, and the output targets being the energy efficiency ratio correction coefficient and the rated cooling capacity correction coefficient.

[0037] Furthermore, in step S3, the power correction coefficient of the variable frequency pump model is adjusted. Represented as actual traffic A polynomial function.

[0038] Furthermore, the adaptive update in step S4 includes the system checking the deviation between the model simulation output and the actual system every day. If the deviation rate is greater than the preset value, the parameters are updated and corrected again based on the running data.

[0039] The present invention has the following beneficial effects: This invention, through the deep integration of mechanistic models and data-driven models and the introduction of an adaptive update mechanism, effectively overcomes the shortcomings of single methods, such as excessive reliance on data, poor robustness under complex working conditions, and limited modeling accuracy. As a result, it achieves reliable cold source system modeling that remains reliable even when data is insufficient, can automatically adjust when the system changes, and has higher overall accuracy, providing a more advanced solution for system optimization and control. Attached Figure Description

[0040] Figure 1 This is a comparison chart of the errors between the refrigeration mechanism data fusion model and the mechanism model of this invention; Figure 2 This is a comparison chart of the errors between the pump mechanism data fusion model and the mechanism model of this invention; Figure 3 This is a diagram illustrating the adaptive update process of the model in this invention. Detailed Implementation

[0041] The technical solutions of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0042] An adaptive cold source system modeling method based on a hybrid mechanism and data-driven approach mainly includes the following steps: S1: Obtain historical operating data of the target building's cooling system and preprocess the data.

[0043] The equipment in the cold source system includes: various chillers, cooling towers, chilled water pumps, and cooling water pumps. System operating history data includes, at various time points: chilled water outlet temperature, chilled water return temperature, cooling water outlet temperature, cooling water return temperature, chilled water flow rate, cooling water flow rate, and operating power for each chiller; flow rate, operating frequency, and operating power for each chilled water pump and cooling water pump; flow rate, inlet water temperature, outlet water temperature, number of fans operating, power, and air volume for the cooling tower; and indoor and outdoor ambient temperature and humidity.

[0044] The data preprocessing process consists of three steps: outlier identification, outlier handling, and missing value imputation. Outlier identification uses the quartile method for calculation.

[0045]

[0046]

[0047] The interquartile range (IQR) is defined as the difference between Q1 and Q3. Q1 refers to the first quartile, which is the value at the 25th percentile after sorting the dataset from smallest to largest. This means that 25% of the values ​​in the data are less than or equal to Q1. Q3 refers to the third quartile, which is the value at the 75th percentile after sorting the dataset from smallest to largest. This means that 75% of the values ​​in the data are less than or equal to Q3. k is an adjustment factor.

[0048] Outlier handling replaces isolated outliers with the mean of adjacent time periods, and consecutive outliers are replaced with historical data from the same period. Missing value imputation uses linear interpolation.

[0049] S2: Based on the physical mechanism of the cold source system, construct the mechanism model of the equipment in the cold source system, and construct the relationship between the main mechanism equipment and energy consumption.

[0050] (a) Constructing a chiller unit model.

[0051] The operating status of a chiller unit is affected by a variety of factors. The rated cooling capacity correction factor for the current operating condition can be obtained by consulting the operating data sheet. Correction factor for rated energy efficiency ratio The specific method for determining the status of the chiller unit is as follows: Calculate the real-time cooling capacity of the chiller unit and actual energy efficiency ratio The formula is:

[0052]

[0053] in, This refers to the cooling capacity under rated operating conditions. This refers to the energy efficiency ratio under rated operating conditions. The cooling capacity of a chiller unit can be expressed as:

[0054] in, This represents the current load of the chiller unit. The chilled water flow rate entering the chiller unit, This refers to the inlet temperature of the chilled water. Set the outlet water temperature for the chilled water. The specific heat capacity of the fluid entering the chilled water pump; Compare current load and real-time cooling capacity , when ≤ Then determine the actual outlet temperature of the chilled water. It can achieve the set water temperature, that is = ; when > If the chiller unit cannot meet the required cooling capacity, then... = And recalculate the actual outlet temperature of the chilled water. and the actual outlet temperature of cooling water :

[0055]

[0056] This refers to the inlet temperature of the cooling water. For cooling water flow rate, Energy consumption of the chiller unit.

[0057] (ii) Constructing a variable frequency water pump model.

[0058] The operating characteristics of a variable frequency pump are related to parameters such as flow rate, head, and efficiency. The power consumption of a variable frequency pump during startup can be calculated using the following formula. By normalizing the real-time operating frequency and using the performance curve provided by the manufacturer, a bilinear interpolation method is employed to obtain the power correction coefficient at the corresponding frequency. Combined with the rated power parameters, the actual energy consumption of the variable frequency pump can be determined.

[0059] The power consumption calculation formula for a variable frequency water pump is as follows:

[0060] in, This represents the power consumed by the pump at the current moment. This refers to the rated power of the variable frequency water pump. This is the power correction factor; The output flow rate of a variable frequency water pump is equal to the rated flow rate of the pump. The efficiency formula for a variable frequency water pump in driving water flow from the inlet to the outlet is as follows:

[0061] in, This represents the overall efficiency of the variable frequency water pump. This represents the efficiency of the variable frequency motor; the overall efficiency of the motor and the variable frequency water pump is determined by the factory parameters. The expression for the shaft power of the variable frequency water pump is derived as follows:

[0062] In the formula, This represents the shaft power required for the variable frequency pumping process.

[0063] Shaft power is the work done by the motor of a variable frequency water pump on the shaft per unit time. Shaft power transfers kinetic energy to the power components of the variable frequency pump, ultimately driving the fluid. The formula for calculating the energy transferred from the motor of the variable frequency water pump to the fluid flow is as follows:

[0064] In the formula, The energy transferred by the variable frequency pump motor to the fluid flow passing through the pump; The part of the variable frequency pump motor with low efficiency that causes the temperature of the fluid flowing through the variable frequency pump to rise.

[0065] The efficiency value of a water pump during water flow is between 0 and 1. This value determines whether the low efficiency of the variable frequency pump motor will cause the temperature of the fluid flowing through the pump to rise during flow. When the motor is operating inefficiently, all waste heat generated will affect the temperature of the fluid flow. Alternatively, the variable frequency pump motor may be installed externally, not in contact with the fluid, releasing waste heat from the operation into the environment.

[0066] During the water pumping process, its efficiency coefficient ( The ∈[0,1]) is used to characterize whether the variable frequency pump motor will cause fluid temperature rise when operating under inefficient conditions. When the motor is in an inefficient operating state, the heat loss generated during energy conversion has two transfer paths: if the motor is externally installed (isolated from the fluid medium), the heat loss is released to the environment through thermal radiation or convection; if the motor is in direct contact with the fluid, the heat loss will be completely converted into the fluid's internal energy, leading to an increase in water temperature. The energy transferred by the variable frequency pump motor to the environment can be quantified according to the first law of thermodynamics using the following formula. This model comprehensively considers parameters such as motor efficiency, input power, and heat dissipation coefficient. The specific expression is as follows:

[0067] In the formula, This represents the energy transferred from the pump's motor to the surrounding air.

[0068] The formula for calculating the outlet temperature of a variable frequency pump is:

[0069] in, The temperature of the fluid flowing out of the variable frequency water pump, The temperature of the fluid flowing into the variable frequency water pump. The mass of the fluid passing through the variable frequency water pump.

[0070] (III) Constructing a cooling tower model.

[0071] The saturated air enthalpy at the cooling tower inlet is calculated based on the inlet dry-bulb temperature and 100% relative humidity. The water flow capacity can then be determined as follows:

[0072] in, For the capacity of the water flow, The specific heat capacity of the fluid entering the cooling tower, This refers to the flow rate of cooling water in the cooling tower. The outlet wet-bulb temperature is estimated, and the cooling tower performance under these conditions is calculated. The outlet saturated air enthalpy is calculated based on the outlet wet-bulb temperature and 100% relative humidity. The simulated specific heat and volumetric rate are then calculated based on the saturated enthalpy and airflow rate, as shown in the following formulas: Calculate the simulated specific heat of air :

[0073] in, The enthalpy of the outlet saturated air. The enthalpy of the inlet saturated air. The inlet wet-bulb temperature, The outlet wet-bulb temperature; Calculate the volume of air :

[0074] in, The airflow rate in the cooling tower; Determine the maximum and minimum values ​​of the water-to-air capacity ratio, and calculate the capacity ratio.

[0075] Determine parameters and :

[0076]

[0077]

[0078] The simulated total heat transfer coefficient of the cooling tower is adjusted based on the simulated specific heat. The simulated total heat transfer coefficient of the cooling tower is expressed by the formula:

[0079] The simulated overall heat transfer coefficient value representing the cooling tower. To design the mass transfer coefficient, It is the specific heat of air.

[0080] Calculate the heat transfer unit coefficient : when When = 0, = 1×10¹⁵; otherwise .

[0081] Calculate heat transfer efficiency : when hour,

[0082] when hour, ,in, ; The total heat transfer can be calculated by assessing the effectiveness under different conditions. And the wet-bulb temperature leaving the cooling tower.

[0083] Calculate total heat transfer :

[0084] in, The inlet water temperature of the cooling tower. The inlet wet-bulb temperature of the cooling tower; Calculated value of cooling tower outlet wet-bulb temperature :

[0085] The calculated value of the outlet wet-bulb temperature is compared with the value of the outlet wet-bulb temperature when calculating the simulated specific heat value of air. If the difference between the two is less than a preset value, the calculation is considered to have converged. Otherwise, the calculated value of the outlet wet-bulb temperature is used as the new value of the outlet wet-bulb temperature. The process is iterated until the calculated value of the outlet wet-bulb temperature and the assumed value of the outlet wet-bulb temperature tend to be consistent.

[0086] Calculate the final cooling tower outlet water temperature :

[0087] S3: Using a data-driven approach, the parameters of the mechanistic model constructed in step S2 are modified to establish a modified model.

[0088] (a) Correction of the characteristic curve of the chiller unit.

[0089] Using Extreme Learning Machine and Regression prediction is performed. The input features have two dimensions: chilled water outlet temperature and cooling water inlet temperature. The output targets have two dimensions: COP correction coefficient and rated cooling capacity correction coefficient. The entire process consists of four steps: The first step is data preprocessing: Outlier removal: Remove outlier data using the interquartile range method.

[0090] Data normalization: Mapping input data and output to a [-1,1] distribution to avoid the model being biased towards a certain feature due to differences in units.

[0091] Dataset partitioning: The data is randomly divided into training and test sets in a 7:3 ratio.

[0092] The next step is to build an Extreme Learning Machine model: Hidden layer node selection: Select values ​​between 20 and 100, and choose the optimal value through 5-fold cross-validation.

[0093] Activation function selection: For the nonlinear relationship of the refrigeration correction coefficient, the Sigmoid function is used. Its output range is [0,1], which has a high degree of matching with the common range of the correction coefficient (0-1.5) and the fitting is more stable.

[0094] Then comes model training: Then comes model training: The weights (W, with a dimension of "number of input features × number of hidden nodes", i.e., 2×N) and the hidden layer biases (b, with a dimension of 1×N, where N is the number of hidden nodes) from the input layer to the hidden layer are randomly initialized, usually sampled from a uniform distribution in [-1,1].

[0095] Calculate the hidden layer output matrix H: Substitute the activation function, the formula is H = activation function(W×X+b), where X is the input of the training set (dimension is "number of samples × 2"), and the final dimension of H is "number of samples × N".

[0096] Analytical solution for output layer weights β: Calculate the generalized inverse using the least squares method, with the formula β = H' × Y_train, where H' is the Moore-Penrose generalized inverse of H, Y_train is the training set output (dimension is "number of samples × 2"), and β has a final dimension of "N×2", corresponding to the weights of the two outputs.

[0097] The error changes before and after the refrigeration characteristics correction are shown in the attached figure. Figure 1 As shown.

[0098] (II) Variable frequency pump characteristic correction model.

[0099] In the water pump energy-saving technology system, variable frequency speed control technology dynamically regulates the input power of the motor through a frequency converter, thereby achieving precise adjustment of the water pump speed. Its operating principle can be characterized by the following formula:

[0100] This is the synchronous speed of the motor; The frequency of the power supply; This represents the number of pole pairs in the motor windings. Let be the slip of the asynchronous motor. Based on the theoretical analysis of the above equation, the pump speed and operating frequency exhibit a significant linear positive correlation. By changing the power supply frequency, the pump speed will adjust accordingly, thereby causing dynamic changes in the fluid flow rate. Based on the similarity law, the quantitative relationship between pump speed and key operating parameters can be expressed by the following formula:

[0101]

[0102]

[0103] In the formula: , These are the rated flow rate and actual flow rate of the water pump, respectively. , The rated speed and actual speed of the water pump; , These are the actual head and rated head of the water pump, respectively. and This refers to the rated power and actual power of the water pump.

[0104] The pump shaft power ratio exhibits a significant positive correlation with the cube of the ratio of real-time flow rate to rated operating flow rate. In central air conditioning refrigeration systems, adjusting the pump's operating frequency allows for speed control, thereby affecting the flow rate. Based on the fluid machinery similarity law, the power-flow characteristic model of a variable frequency pump can be expressed as:

[0105]

[0106] In the formula: This refers to the actual power of the cooling water pump; This is for partial load factor; , , , The coefficient of performance (COP) of the cooling water pump; Rated flow rate; This represents the actual flow rate of chilled water. This is the rated frequency of the water pump; since the flow rate of a variable frequency water pump is linearly related to the frequency of the frequency converter, therefore... This can be expressed as the actual flow rate of the water pump. linear functions ; For the coefficient, since , , , and All of these are constants, thus the energy consumption of the cooling water pump can be transformed into a function of the flow rate.

[0107]

[0108]

[0109] in, These are the performance coefficients determined through fitting.

[0110] The error changes before and after the pump characteristic correction are shown in the attached figure. Figure 2 As shown.

[0111] S4: Model parameter update.

[0112] The model parameter update process is shown in the appendix. Figure 3 The system checks the deviation between the model simulation output and the real system daily. If the deviation rate is greater than 5%, the parameters are updated and corrected based on the running data.

[0113] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A modeling method for adaptive cold source systems based on a hybrid mechanism and data-driven approach, characterized in that, Includes the following steps: S1, acquire historical operating data of the cold source system, and preprocess the historical operating data; S2, Based on the physical mechanism of the cold source system, construct the mechanism model of the equipment in the cold source system; S3, Modify the key parameters of the mechanism model constructed in step S2, and establish a modified model; S4, based on the deviation between the model simulation output and the actual system operation, adaptively updates the parameters of the corrected model.

2. The adaptive cold source system modeling method based on mechanism and data hybrid driving as described in claim 1, characterized in that, The equipment in the cold source system in step S1 includes, but is not limited to: chiller units, cooling towers, chilled water pumps, and cooling water pumps; the historical operating data includes, but is not limited to: The chilled water outlet temperature, chilled water return temperature, cooling water outlet temperature, cooling water return temperature, chilled water flow rate, cooling water flow rate, and operating power of the chiller unit are specified. The flow rate, operating frequency, and operating power of the chilled water pump; The flow rate, operating frequency, and operating power of the cooling water pump; Cooling tower flow rate, inlet water temperature, outlet water temperature, number of fans in operation, power, and air volume; Indoor and outdoor ambient temperature and humidity.

3. The adaptive cold source system modeling method based on mechanism and data hybrid driving as described in claim 1 or 2, characterized in that, When preprocessing the historical operating data, outliers are identified using the interquartile range method. Isolated outliers are replaced with the mean of adjacent time periods, and continuous outliers are replaced with historical data from the same period. Missing values ​​are filled in using linear interpolation.

4. The adaptive cold source system modeling method based on mechanism and data hybrid driving as described in claim 1, characterized in that, In step S2, a mechanism model of the equipment in the cold source system is constructed, including at least: a chiller unit model, a variable frequency water pump model, and a cooling tower model.

5. The adaptive cold source system modeling method based on mechanism and data hybrid driving as described in claim 4, characterized in that, The construction of a chiller unit model specifically includes: Based on the correction factor for rated cooling capacity and rated energy efficiency ratio correction factor Calculate the real-time cooling capacity of the chiller unit. and actual energy efficiency ratio The formula is: in, This refers to the cooling capacity under rated operating conditions. This refers to the energy efficiency ratio under rated operating conditions. Calculate the current load: in, This represents the current load of the chiller unit. This refers to the chilled water flow rate. This refers to the inlet temperature of the chilled water. Set the outlet water temperature for the chilled water. The specific heat capacity of the fluid entering the chilled water pump; Compare current load and real-time cooling capacity , when ≤ Then determine the actual outlet temperature of the chilled water. It can achieve the set water temperature, that is = ; when > If the chiller unit cannot meet the required cooling capacity, then... = And recalculate the actual outlet temperature of the chilled water. and the actual outlet temperature of cooling water : This refers to the inlet temperature of the cooling water. For cooling water flow rate, Energy consumption of the chiller unit.

6. The adaptive cold source system modeling method based on mechanism and data hybrid driving as described in claim 4, characterized in that, When constructing a variable frequency water pump model, The power consumption calculation formula for a variable frequency water pump is: in, This refers to the rated power of the variable frequency water pump. This is the power correction factor; The formula for calculating the pumping efficiency of a variable frequency water pump is: in, This represents the overall efficiency of the variable frequency water pump. This represents the efficiency of the variable frequency motor; The formula for calculating the shaft power of a variable frequency water pump is: The formula for calculating the energy transferred from the variable frequency water pump motor to the internal fluid is: in, The proportion of waste heat generated due to low motor efficiency that is directly used to heat the fluid inside the pump. The formula for calculating the energy transferred from the variable frequency water pump motor to the surrounding air is: The formula for calculating the outlet temperature of a variable frequency pump is: in, The temperature of the fluid flowing out of the variable frequency water pump, The temperature of the fluid flowing into the variable frequency water pump. The mass of the fluid passing through the variable frequency water pump.

7. The adaptive cold source system modeling method based on mechanism and data hybrid driving as described in claim 4, characterized in that, Constructing a cooling tower model specifically includes: Step 1: Calculate the capacity of the water flow. : in, The specific heat capacity of the fluid entering the cooling tower, This refers to the flow rate of cooling water in the cooling tower. Calculate the simulated specific heat of air : in, The enthalpy of the outlet saturated air. The enthalpy of the inlet saturated air. The inlet wet-bulb temperature, The outlet wet-bulb temperature; Calculate the volume of air : in, The airflow rate in the cooling tower; Step Two: Determine parameters and : Step 3: Calculate the simulated overall heat transfer coefficient of the cooling tower. : in, To design the mass transfer coefficient, The specific heat of air; Step Four: Calculate the heat transfer unit coefficient : when When = 0, = 1×10¹⁵; otherwise Step 5: Calculate heat transfer efficiency : when hour, when hour, ,in, ; Step Six: Calculate total heat transfer : in, The inlet water temperature of the cooling tower. The inlet wet-bulb temperature of the cooling tower; Calculated value of cooling tower outlet wet-bulb temperature : The calculated value of the outlet wet-bulb temperature is compared with the value of the outlet wet-bulb temperature when calculating the simulated specific heat value of air. If the difference between the two is less than the preset value, the calculation is considered to have converged. Otherwise, the calculated value of the outlet wet-bulb temperature is used as the new value of the outlet wet-bulb temperature, and the process of steps one to six is ​​repeated until the calculation converges. Step Seven: Calculate the final cooling tower outlet water temperature : 。 8. The adaptive cold source system modeling method based on mechanism and data hybrid driving as described in claim 5, characterized in that, In step S3, for the chiller unit model correction, a limit learning machine is used to adjust the rated cooling capacity correction coefficient. and rated energy efficiency ratio correction factor Regression prediction is performed with the input features being the chilled water outlet temperature and the cooling water inlet temperature, and the output targets being the energy efficiency ratio correction coefficient and the rated cooling capacity correction coefficient.

9. The adaptive cold source system modeling method based on mechanism and data hybrid driving as described in claim 6, characterized in that, In step S3, the power correction coefficient of the variable frequency pump model is adjusted. Represented as actual traffic A polynomial function.

10. The adaptive cold source system modeling method based on mechanism and data hybrid driving as described in claim 1, characterized in that, The adaptive update in step S4 includes daily checks on the deviation between the model simulation output and the actual system. If the deviation rate is greater than the preset value, the parameters are updated and corrected again based on the running data.

Images