Multi-control object modeling method for air conditioning system

Abstract

The present invention discloses a multi-control object modeling method for an air-conditioning system, comprising: obtaining multi-source operating data collected by air-conditioning system sensors, performing principal component analysis and time-shift correlation analysis on the multi-source operating data, and screening out core input variables corresponding to target control objects; target control objects include virtual indoor temperature, energy consumption of air-source heat pump units, water pump energy consumption, and fan energy consumption; time-shift correlation analysis determines the lag step length of each variable through the maximum information coefficient, and screens out key time series features; based on the screened core input variables, respectively constructs a data-driven model for each control object, and a combined multi-control object simulation framework is used to predict the energy consumption and environmental status of the air-conditioning system. The present invention overcomes the dependence of the test method on the actual system and hardware, and reduces the test cost and risk. The source code can be tested before the system and hardware are configured, avoiding the frequent equipment control problem caused by testing in the actual system.

Description

Technical Field

[0001] The invention relates to a multi-control object modeling method for an air-conditioning system. Background Art

[0002] The full automation, informatization, and digitization of air-conditioning systems, along with the implementation of advanced control technologies, are key areas for the development of intelligent modern buildings. Software integrating relevant intelligent algorithms plays a crucial role in this process. However, traditional software testing methods, such as hardware-in-the-loop simulators, often rely on actual systems and hardware for testing. This approach has numerous limitations in practical applications.

[0003] Especially in the early stages of software and control algorithm development, using actual systems for testing is almost unrealistic. For one thing, large public building air conditioning systems are complex and massive, requiring frequent equipment control during testing, which can significantly disrupt the system's stable operation. Furthermore, actual testing sessions are typically short and discontinuous, making it difficult to fully verify the software's functionality and performance.

[0004] The invention patent with publication number CN110673585B proposes an innovative method for testing train air-conditioning systems. The method first constructs a train control system TCMS simulation device and an air-conditioning system simulation device, and designs corresponding air-conditioning operating conditions strategies. Subsequently, based on the input test configuration parameters, the system calls and integrates the configuration items in the preset air-conditioning operating conditions strategy to automatically generate test cases. These test cases will be transmitted to the train control system TCMS simulation device, which will test the air-conditioning system simulation device according to the test cases. The above-mentioned simulation device can be used to test the train air-conditioning system, providing ideas for testing the station air-conditioning system. Summary of the Invention

[0005] The purpose of the present invention is to overcome the shortcomings of the existing technology, overcome the dependence of traditional testing methods on actual systems and hardware, reduce testing costs and risks, and provide a multi-control object modeling method for air-conditioning systems.

[0006] The object of the present invention is achieved through the following technical solutions:

[0007] A multi-control object modeling method for an air-conditioning system includes: obtaining multi-source operating data collected by air-conditioning system sensors, wherein the multi-source operating data includes outdoor temperature (To), outdoor relative humidity (Ho), indoor temperature (Tn), supply water temperature (Tsup), return water temperature (Tret), supply water flow (msup), water pump frequency (npump), and the number of indoor occupants (Np); performing principal component analysis (PCA) and time-shift correlation analysis on the multi-source operating data to screen out core input variables corresponding to target control objects; wherein,

[0008] The target control objects include virtual indoor temperature, energy consumption of air source heat pump unit, water pump energy consumption and fan energy consumption;

[0009] The time-shift correlation analysis determines the lag step size of each variable through the maximum information coefficient (MIC) and screens key time series features; based on the screened core input variables, a data-driven model of each control object is constructed respectively, and the model is trained and verified to form a combined multi-control object simulation framework; wherein, the combined multi-control object simulation framework is used to predict the energy consumption and environmental status of the air-conditioning system, and outputs the results to the model predictive control (MPC) optimizer.

[0010] As a preferred method, the steps of constructing a virtual indoor temperature model include: determining the input variables through MIC analysis, including the 0-3 step lagged characteristics of the outdoor temperature (To), the 0-2 step lagged characteristics of the outdoor humidity (Ho) and the current number of people indoors (Np); performing PCA dimensionality reduction processing on the outdoor temperature and humidity, and taking the principal components with a cumulative variance contribution rate ≥98% as the input variables after dimensionality reduction, respectively denoted as Toa1, Toa2 and Hoa1, Hoa2; combining the reduced dimensionality variables with the historical water supply temperature (Tsup0-Tsup1), the historical water supply flow (msup0-msup1) and the current indoor temperature (Tn) to construct a training set with 10 input variables; using the Att-BiGRU algorithm to model the training set, and optimizing the network parameters through back propagation to obtain a virtual indoor temperature simulation model with a prediction error MAE≤0.07 and RMSE≤0.12.

[0011] As a preferred method, the steps of constructing the energy consumption model of the air source heat pump unit include: screening input variables, including the current supply water temperature (Tsup), return water temperature (Tret), flow rate (msup), outdoor temperature (To), outdoor humidity (Ho), indoor personnel (Np), indoor temperature (Tn) and historical cooling capacity (Qhis); the input variables are implanted into the stacked width learning system to establish an energy consumption prediction sub-model, the output of which is the coefficient of performance (COP) of the heat pump, and the calculation relationship satisfies: COP = f(T low ,T high ), where T low is the outlet water temperature, T high is the outdoor temperature; the integrated cooling quantum model and COP sub-model are used to calculate the ASHP =Q ASHP / COP*ΔT outputs the host energy consumption value, where Q is the system cooling capacity per unit time, kW; ΔT is the system operating time, h.

[0012] As a preferred method, when building a water pump energy consumption model, perform the following operations:

[0013] The flow rate of the water pump is proportional to the frequency, and the electric power is proportional to the cube of the frequency. The relationship between the water pump flow rate m, electric power Q and water pump frequency n is shown in formula (1-1) and formula (1-2):

[0014]

[0015] Where: m1 and m0 are the flow rates of the pump under actual working conditions and rated working conditions respectively, m 3 / h; n1 and n0 are the frequencies of the water pump under actual working conditions and rated working conditions, Hz; Q1 and Q0 are the electrical power of the water pump under actual working conditions and rated working conditions, kW;

[0016] For water systems where end-load water flow can be actively regulated, similar operating conditions no longer apply due to changes in the pipe network characteristic curve. Furthermore, as pump flow decreases, input power decreases, and pump efficiency plummets. Ultimately, changes in energy consumption are influenced by many factors, including pump model, valve characteristics, and control methods, making it difficult to accurately describe. Based on the literature, equations (1-3) and (1-4) can be used to express the relationship between flow rate, electrical power, and frequency under actual pump operating conditions.

[0017]

[0018] Where: a, b, c, d, e, and f are coefficients fitted by measured data.

[0019] As a preferred method, a combined air conditioning unit model is constructed, including a return water temperature prediction model:

[0020] For a single return constant air volume air conditioning system, the indoor air is mixed with the outdoor fresh air to obtain the initial state of air treatment with a dry bulb temperature of t1, an enthalpy value of h1, and a wet bulb temperature of t s1 , the amount of air being processed is G, the chilled water flow is m, and the initial temperature of the chilled water is T sup , then the final state t2 and return water temperature T that the surface cooler can reach when processing air ret The process is as follows:

[0021] A1. Calculate the wind speed V facing the surface cooler y And the water flow rate w, the calculation formula is shown in formula (2-1) and formula (2-2):

[0022]

[0023] Where: F y is the windward area of ​​the surface cooler, m 2 ; ρ is the density of water, kg / m 3 ;f w is the cross-sectional area of ​​water flow, m2 .

[0024] A2. Calculate the general heat exchange efficiency E' that a surface cooler can provide. This value can be obtained from the equipment sample.

[0025] A3. Assume that the dry-bulb temperature of the air at the final state is t2.

[0026] You can press t2=t w1 +(4~6)℃ to determine the final dry bulb temperature of the air, t w1 It is the measured value of the first stage water supply temperature in the air conditioning system. The final wet bulb temperature t s2 The calculation formula is shown in formula (2-3):

[0027] t s2 =t2-(t1-t s1 )(1-E') (2-3)

[0028] The calculation formula for the final state enthalpy h2 is shown in formula (2-4):

[0029]

[0030] A4. Calculate the wetting coefficient ξ using the formula shown in (2-5):

[0031]

[0032] A5. Calculate the heat transfer coefficient K s .

[0033] The empirical formula for the heat transfer coefficient of the heat exchanger dehumidification cooling process is shown in formula (2-6):

[0034]

[0035] Where: A, P, B, m, n are coefficients and exponents obtained from experiments.

[0036] A6. Calculate the total heat exchange efficiency E' that the surface cooler can achieve g , the calculation formula is shown in formula (2-7) to formula (2-9):

[0037]

[0038] A7. Calculate the required total heat exchange efficiency E g And with E' g In comparison, E g The calculation formula is shown in formula (2-10):

[0039]

[0040] When|E g -E'g When |≤δ (usually 0.01), the assumed t2 is proved to be appropriate. g -E' g |>δ, t2 should be reset and the calculation should be performed again.

[0041] A8. Calculate the chilled water return temperature T ret , the calculation formula is shown in formula (2-11).

[0042]

[0043] As an optimal method, a combined air conditioning unit model is constructed, including: establishing a return water temperature prediction model: through MIC analysis, the input variables are determined to be To with a lag of 1 step, Ho with a lag of 1 step, Tn with a lag of 1 step, Tsup with a lag of 1 step, and msup with a lag of 1 step, and the Att-BiGRU algorithm is used to train a return water temperature prediction sub-model with MAE ≤ 0.11; establishing a fixed-frequency fan energy consumption model: the energy consumption of the combined air conditioning unit is the fan energy consumption. According to the similarity law, the power of the fan is proportional to the cube of its operating frequency. The fan in the combined air conditioning unit is a fixed-frequency fan, and its operating air volume and power remain unchanged; the fan energy consumption W fan The calculation formula is shown in formula (3-1):

[0044] W fan =Q fan *ΔT (3-1)

[0045] Where: Q fan is the actual operating power of the fan, kW.

[0046] As a preferred method, the model training and verification steps include: dividing the sample data into a training set and a validation set in a ratio of 7:3, using the first 10 time steps to predict the output of the next time step; using early stopping to monitor the validation set error, and terminating the training when the RMSE drops less than 1% after 10 consecutive iterations; calculating the mean absolute error (MAE), mean absolute percentage error (MAPE) and root mean square error (RMSE) of the validation set, selecting Att-BiGRU as the optimal basic algorithm, and optimizing its model weights by gradient descent.

[0047] As a preferred method, a model integration method based on a simulation platform is also included: building a building data model in the Matlab environment, establishing a Modbus / OPC communication link with the MPC controller through the Kepserverex software; integrating the virtual indoor temperature model, air source heat pump model, water pump model and air conditioning unit model in parallel, and ensuring the consistency of simulation steps through the time synchronization module; generating model output signals including: real-time temperature prediction value Tn+1, host energy consumption W ASHP, water pump energy consumption W pump , fan energy consumption W fan and maps it to the register address of kepserverex for the controller to read.

[0048] As a preferred method, the communication link construction specifically includes: creating a virtual device in kepserverex, binding the input and output variables of the Matlab model to the Modbus TCP protocol registers; setting the simulation step size to 5 minutes, and completing the closed-loop process of data writing → model calculation → register update → controller reading within each step size; configuring the OPC UA client on the controller side, regularly reading the predicted temperature and energy consumption data in the register, and feeding it back to the MPC optimization algorithm.

[0049] As a preferred method, it also includes a model online update method: deploying an operation data monitoring module to collect the actual temperature value T_real and the model prediction value T_pred in real time; when the average daily prediction error exceeds the threshold e (e=0.5℃), the model retraining process is triggered; using incremental learning, the latest 7 days of data are added to the training set, and the Att-BiGRU network weights are updated to maintain the model prediction accuracy.

[0050] The present invention has at least the following beneficial effects: the present invention overcomes the dependence of traditional testing methods on actual systems and hardware, and reduces testing costs and risks. By constructing a data-driven combined multi-control object simulation framework, the source code can be tested before the system and hardware are configured, avoiding the problems of frequent equipment control, interference with stable system operation, etc. caused by testing in the actual system. The present invention improves the efficiency and comprehensiveness of testing. Through principal component analysis and time-shift correlation analysis, the core input variables corresponding to the target control object are screened out, and an efficient data-driven model is constructed. The model can fully simulate the operating status of the air-conditioning system, provide strong support for the automatic generation and testing of test cases, thereby improving the efficiency and comprehensiveness of testing. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] To reveal the technical details of the embodiments of the present invention, the following is a brief introduction to the drawings involved in the embodiments. It should be emphasized that these drawings only illustrate several embodiments of the present invention and should not be considered as defining the scope of the invention. Those skilled in the art can deduce other relevant drawings based on these drawings without engaging in creative work.

[0052] Figure 1 Schematic diagram of the communication method between sensors and devices and software platform in the embodiment;

[0053] Figure 2 This is the simulation environment of the station air conditioning control system in the embodiment;

[0054] Figure 3 is the MIC value of each influencing factor and indoor temperature in the embodiment;

[0055] Figure 4 is the MIC value of each influencing factor and return water temperature in the embodiment. DETAILED DESCRIPTION

[0056] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings, but the protection scope of the present invention is not limited to the following.

[0057] In the following, embodiments of the present disclosure are described in detail with the aid of accompanying drawings. However, please be aware that the present disclosure is not limited to the specific forms shown herein. Rather, it should be understood to encompass various variations, equivalents, and / or alternatives to the embodiments of the present disclosure. In describing the drawings, the same reference numerals will be used to indicate similar components.

[0058] In the various embodiments of the present disclosure, expressions such as "first," "second," "the first," or "the second" are intended to modify different components, rather than to indicate order and / or importance, and do not limit the corresponding components. For example, a first user device and a second user device each represent different user devices, although they both fall within the scope of user devices. Similarly, a first component can be named a second component, and a second component can be named a first component, which does not change their essential attributes within the scope of the present disclosure.

[0059] In this disclosure, terms are used to illustrate specific embodiments and do not constitute limitations of this disclosure. In this context, the use of the singular also encompasses the plural, unless the text clearly indicates otherwise. In the process of explanation, it should be understood that terms such as "including" or "having" are intended to indicate the presence of a feature, quantity, step, operation, structural component, part, or combination thereof, and do not preclude the possibility or addition of one or more other features, quantities, steps, operations, structural components, parts, or combinations thereof.

[0060] It should be understood that while the following description provides extensive specific details intended to facilitate a comprehensive understanding of the example embodiments, those skilled in the art will appreciate that the example embodiments can be implemented without these specific details. For example, systems may be presented in block diagram form to avoid excessive detail that would obscure the clarity of the examples. In other cases, unnecessary details regarding well-known processes, structures, and techniques may be omitted to maintain clarity of the examples.

[0061] A multi-control object modeling method for an air-conditioning system includes: obtaining multi-source operating data collected by air-conditioning system sensors, wherein the multi-source operating data includes outdoor temperature (To), outdoor relative humidity (Ho), indoor temperature (Tn), supply water temperature (Tsup), return water temperature (Tret), supply water flow (msup), water pump frequency (npump), and the number of indoor occupants (Np); performing principal component analysis (PCA) and time-shift correlation analysis on the multi-source operating data to screen out core input variables corresponding to target control objects; wherein,

[0062] The target control objects include virtual indoor temperature, energy consumption of air source heat pump unit, water pump energy consumption and fan energy consumption;

[0063] The time-shift correlation analysis determines the lag step of each variable through the maximum information coefficient (MIC) and screens key time series features; based on the screened core input variables, a dynamic data-driven model of each control object is constructed respectively, and the model is trained and verified through a time series algorithm enhanced by an attention mechanism to form a combined multi-control object simulation framework; wherein, the combined multi-control object simulation framework is used to predict the energy consumption and environmental status of the air-conditioning system, and outputs the results to the model predictive control (MPC) optimizer.

[0064] A multi-control-object modeling approach for air-conditioning systems utilizes multi-source operational data collected by air-conditioning system sensors, including outdoor temperature, outdoor relative humidity, indoor temperature, supply water temperature, return water temperature, supply water flow rate, pump frequency, and indoor occupancy count, providing comprehensive and accurate data support for subsequent modeling. Principal component analysis (PCA) and time-shift correlation analysis are performed on the multi-source operational data to identify core input variables corresponding to the target control object, effectively reducing model complexity and improving operational efficiency. Furthermore, time-shift correlation analysis uses the maximum information coefficient (MIC) to determine the lag step size for each variable and identify key time series features, further enhancing the model's predictive capabilities. Based on the identified core input variables, dynamic data-driven models are constructed for each control object. These models are trained and validated using a time series algorithm enhanced with an attention mechanism, forming a combined multi-control-object simulation framework. This framework enables accurate prediction of air-conditioning system energy consumption and environmental conditions, providing strong support for subsequent model predictive control (MPC) optimizers. This approach not only improves air-conditioning system operational efficiency and reduces energy consumption, but also optimizes indoor environmental conditions and enhances user comfort. At the same time, this method has good versatility and scalability, and can be applied to different types of air-conditioning systems and control objects, providing new ideas and methods for the intelligent management and optimization of air-conditioning systems.

[0065] The present invention overcomes the dependence of traditional testing methods on actual systems and hardware, reducing testing costs and risks. By constructing a data-driven combined multi-control object simulation framework, the source code can be tested before the system and hardware are configured, avoiding problems such as frequent equipment control and interference with stable system operation caused by testing in the actual system. The present invention improves testing efficiency and comprehensiveness. Through principal component analysis and time-shift correlation analysis, the core input variables corresponding to the target control object are screened out, and an efficient data-driven model is constructed. The model can fully simulate the operating status of the air-conditioning system, providing strong support for the automatic generation and testing of test cases, thereby improving testing efficiency and comprehensiveness.

[0066] This embodiment enables effective verification of the predictive control algorithm for air conditioning systems. The proposed simulation platform can simulate a variety of operating conditions and verify and optimize the predictive control algorithm for air conditioning systems, improving the algorithm's stability and accuracy and providing strong support for the development of air conditioning system software. In summary, our proposed multi-control object modeling method for air conditioning systems effectively overcomes the limitations of traditional testing methods, improves testing efficiency and comprehensiveness, and enables effective verification of the predictive control algorithm for air conditioning systems, providing strong support for the development of air conditioning system software.

[0067] In one embodiment, the steps of constructing a virtual indoor temperature model include: determining input variables through MIC analysis, including the 0-3 step lagged feature of the outdoor temperature (To), the 0-2 step lagged feature of the outdoor humidity (Ho), and the current number of people indoors (Np); performing PCA dimensionality reduction processing on the outdoor temperature and humidity, and taking the principal components with a cumulative variance contribution rate ≥98% as the input variables after dimensionality reduction, respectively denoted as Toa1, Toa2 and Hoa1, Hoa2; combining the reduced dimensionality variables with the historical water supply temperature (Tsup0-Tsup1), the historical water supply flow (msup0-msup1) and the current indoor temperature (Tn) to construct a training set with 10 input variables; using the Att-BiGRU algorithm to model the training set, and optimizing the network parameters through back propagation to obtain a virtual indoor temperature simulation model with a prediction error MAE ≤ 0.07 and RMSE ≤ 0.12.

[0068] This example uses Maximum Information Coefficient (MIC) analysis to determine input variables. These variables include the lagged 0-3-step feature of the outdoor temperature (To), the lagged 0-2-step feature of the outdoor humidity (Ho), and the current number of people indoors (Np). These variables serve as the basic data for model construction. Dimensionality reduction is performed on the outdoor temperature and humidity using Principal Component Analysis (PCA). During the PCA process, principal components with a cumulative variance contribution greater than or equal to 98% are selected as the reduced input variables. These principal components are denoted as Toa1 and Toa2 (representing the two principal components of the outdoor temperature after dimensionality reduction) and Hoa1 and Hoa2 (representing the two principal components of the outdoor humidity after dimensionality reduction). The reduced variables are combined with the historical water supply temperature (Tsup0-Tsup1, i.e., the current and previous water supply temperatures), the historical water supply flow rate (msup0-msup1, i.e., the current and previous water supply flow rates), and the current indoor temperature (Tn). This constructs a training set with 10 input variables. The training set was modeled using the Attention Mechanism Bidirectional Gated Recurrent Unit (Att-BiGRU) algorithm. The network parameters were optimized using a backpropagation algorithm, enabling the model to accurately predict indoor temperature. After training and optimization, the resulting virtual indoor temperature simulation model should meet the requirements of a mean absolute error (MAE) of less than or equal to 0.07 and a root mean square error (RMSE) of less than or equal to 0.12. This model can accurately predict indoor temperature by comprehensively considering multiple factors, including outdoor temperature, humidity, number of people indoors, and water supply temperature and flow. This not only helps improve indoor environmental comfort but also provides strong support for building energy management and energy conservation and consumption reduction.

[0069] In one embodiment, the steps of constructing an energy consumption model of an air source heat pump unit include: screening input variables using a fuzzy clustering algorithm, including the current supply water temperature (Tsup), return water temperature (Tret), flow rate (msup), outdoor temperature (To), outdoor humidity (Ho), number of people indoors (Np), indoor temperature (Tn), and historical cooling capacity (Qhis); implanting the input variables into a stacked width learning system to establish an energy consumption prediction sub-model, the output of which is the coefficient of performance (COP) of the heat pump, and the calculation relationship satisfies: COP = f(T low ,T high ),Right now Where T low is the outlet water temperature, T high is the outdoor temperature; ASHP =Q ASHP / COP*ΔT outputs the output host energy consumption value, where Q ASHP is the system cooling capacity per unit time, kW; ΔT is the system operating time, h.

[0070] The energy consumption of the main unit is related to the cooling capacity and COP of the system. For air source heat pump units, under cooling conditions, the system energy consumption is related to the water supply temperature T sup , return water temperature T ret , flow m sup , outdoor ambient temperature T o , relative humidity H o 、N number of people indoors p , indoor temperature T N And historical cooling capacity Q his Therefore, the host energy consumption model can be expressed as shown in formula (1):

[0071] W ASHP =f ASHP (T sup ,T ret ,m sup ,T o ,H o ,N p ,T N ,Q his )*ΔT (1)

[0072] Where: f ASHP This is a data-driven model related to the energy consumption calculation of air source heat pump units. This embodiment uses a fuzzy clustering algorithm to screen input variables and uses a stacked broad learning system to establish a host energy consumption model.

[0073] Fuzzy clustering variable screening process: Data preprocessing: Standardize input variables (such as water supply temperature T sup , return water temperature T ret , flow m sup , outdoor ambient temperature T o , relative humidity H o 、N number of people indoors p , indoor temperature T N And historical cooling capacity Q his ). Algorithm selection: Fuzzy C-means clustering (FCM) is used, the membership function is defined as Gaussian, and the number of clusters is optimized by the silhouette index. Variable screening: Based on the membership degree of the variable in the cluster center (>0.7) and the correlation with the target variable Mutual Information (Mutual Information) of Q is used to select key variables. Validation: Cross-validation is used to compare model performance (e.g., RMSE, MAE) before and after the selection. This embodiment avoids interference from redundant variables, reduces model complexity, and improves generalization capabilities.

[0074] Modeling process of stacked width learning system: Model structure: Input layer: receives the filtered variables. Feature node layer: generates initial feature nodes (such as 100) through random mapping. Enhanced node layer: nonlinear transformation (such as Sigmoid) expands features and enhances expression capabilities. Stacking design: multi-layer feature-enhancement node stacking, extracting high-order features layer by layer (such as 3-layer stacking). Output layer: linear regression predicts energy consumption. Training strategy: incremental learning: dynamically expand nodes to adapt to changes in data distribution. Regularization: L2 regularization is used to prevent overfitting, and hyperparameters are optimized through grid search. The stacked structure enhances nonlinear fitting capabilities and is suitable for energy consumption prediction of multi-variable coupling.

[0075] In one embodiment, when constructing a water pump energy consumption model, the following operations are performed: the water pump flow rate is proportional to the frequency, and the electric power is proportional to the cube of the frequency; the relationship between the water pump flow rate m, the electric power Q, and the water pump frequency n is shown in Equations (1-1) and (1-2):

[0076]

[0077] Where: m1 and m0 are the flow rates of the pump under actual working conditions and rated working conditions respectively, m 3 / h; n1 and n0 are the frequencies of the water pump under actual working conditions and rated working conditions, Hz; Q1 and Q0 are the electrical power of the water pump under actual working conditions and rated working conditions, kW;

[0078] For water systems where end-load water flow can be actively regulated, similar operating conditions no longer apply due to changes in the pipe network characteristic curve. Furthermore, as pump flow decreases, input power decreases, and pump efficiency plummets. Ultimately, changes in energy consumption are influenced by many factors, including pump model, valve characteristics, and control methods, making it difficult to accurately describe. Based on the literature, equations (1-3) and (1-4) can be used to express the relationship between flow rate, electrical power, and frequency under actual pump operating conditions.

[0079]

[0080] Where: a, b, c, d, e, and f are coefficients fitted by measured data;

[0081] According to the collected data, the relationship between the flow rate, electric power and frequency of the water pump under actual working conditions is calculated as shown in Equations (1-5) and (1-6).

[0082] m pump =6.5001n pump -40.5961 (1-5)

[0083] Q pump =-0.0083n pump3 +1.0829n pump 2 -45.6339n pump +641.4388 (1-6)

[0084] Where: m pump is the flow rate of the water pump in this project under actual working conditions, m 3 / h;n pump is the frequency of the water pump under actual working conditions, Hz; Q pump is the electric power of the water pump under actual working conditions, kW.

[0085] Energy consumption of water pump W pump The calculation formula is shown in formula (1-7):

[0086] W pump =Q pump *ΔT(1-7).

[0087] In another embodiment, constructing a combined air conditioning unit model includes establishing a return water temperature prediction model:

[0088] Combined air handling units are mainly composed of surface coolers, humidifiers, heaters, fans and other equipment. In the cooling season, surface coolers are mainly used to cool and dehumidify the air. For a single return air constant air volume air conditioning system, the indoor air is mixed with the outdoor fresh air to obtain the initial state of air treatment with a dry bulb temperature of t1, an enthalpy value of h1, and a wet bulb temperature of t s1 , the amount of air being processed is G, the chilled water flow is m, and the initial temperature of the chilled water is T sup , then the final state t2 and return water temperature T that the surface cooler can reach when processing air ret The process is as follows:

[0089] A1. Calculate the wind speed V facing the surface cooler y And the water flow rate w, the calculation formula is shown in formula (2-1) and formula (2-2):

[0090]

[0091] Where: F y is the windward area of ​​the surface cooler, m 2 ; ρ is the density of water, kg / m 3 ;f w is the cross-sectional area of ​​water flow, m 2 .

[0092] A2. Calculate the general heat exchange efficiency E' that a surface cooler can provide. This value can be obtained from equipment samples.

[0093] A3. Assume that the dry-bulb temperature of the air at the final state is t2.

[0094] You can press t2=t w1 +(4~6)℃ to determine the final dry bulb temperature of the air, t w1 It is the measured value of the first stage water supply temperature in the air conditioning system. The final wet bulb temperature t s2 The calculation formula is shown in formula (2-3):

[0095] t s2 =t2-(t1-t s1 )(1-E') (2-3)

[0096] The calculation formula for the final state enthalpy h2 is shown in formula (2-4):

[0097]

[0098] A4. Calculate the wetting coefficient ξ using the formula shown in (2-5):

[0099]

[0100] A5. Calculate the heat transfer coefficient K s .

[0101] The empirical formula for the heat transfer coefficient of the heat exchanger dehumidification cooling process is shown in formula (2-6):

[0102]

[0103] Where: A, P, B, m, n are coefficients and exponents obtained from experiments.

[0104] A6. Calculate the total heat exchange efficiency E' that the surface cooler can achieve g , the calculation formula is shown in formula (2-7) to formula (2-9):

[0105]

[0106] A7. Calculate the required total heat exchange efficiency E g And with E' g In comparison, E g The calculation formula is shown in formula (2-10):

[0107]

[0108] When|E g -E' g When |≤δ (usually 0.01), the assumed t2 is proved to be appropriate. g -E' g |>δ, t2 should be reset and the calculation should be performed again.

[0109] A8. Calculate the chilled water return temperature T ret , the calculation formula is shown in formula (2-11).

[0110]

[0111] Through the above steps, we can calculate the final dry-bulb temperature t2 and chilled water return temperature Tret that the air processed by the air cooler can achieve. This method not only improves the accuracy of air treatment but also optimizes energy efficiency. Through detailed calculations, it is possible to ensure that the air reaches the desired dry-bulb temperature t2, wet-bulb temperature ts2, and enthalpy h2 after treatment, meeting the specific requirements of the air conditioning system. The calculation process takes into account multiple factors such as the heat exchange efficiency, heat transfer coefficient, and moisture separation coefficient of the air cooler, thereby enabling more rational configuration of the chilled water flow rate and temperature, reducing unnecessary energy consumption. By assuming and verifying the final dry-bulb temperature of the air, the stability of the system can be ensured during actual operation. If the assumed t2 deviates from the actual requirements, timely adjustments can be made to avoid excessive or insufficient fluctuations in the system. This embodiment enables system maintenance personnel to more clearly understand the system's operating status, promptly identify and resolve problems, and reduce system maintenance costs and management difficulties.

[0112] In one embodiment, a modular air conditioning unit model is constructed, including: establishing a return water temperature prediction model by another method: determining the input variables as To lag 1 step, Ho lag 1 step, Tn lag 1 step, Tsup lag 1 step, and msup lag 1 step through MIC analysis, and using the Att-BiGRU algorithm to train a return water temperature prediction sub-model with MAE ≤ 0.11; establishing a fixed-frequency fan energy consumption model: the energy consumption of the modular air conditioning unit is the fan energy consumption. According to the similarity law, the power of the fan is proportional to the cube of its operating frequency. The fan in the modular air conditioning unit is a fixed-frequency fan, and its operating air volume and power remain unchanged; the fan energy consumption W fan The calculation formula is shown in formula (3-1):

[0113] W fan =Q fan *ΔT (3-1)

[0114] Where: Q fan is the actual operating power of the fan, kW.

[0115] The total energy consumption calculation formula of the air conditioning system is shown in formula (3-2):

[0116] W total =(f ASHP +Q pump +Q fan )*ΔT (3-2)

[0117] In this example, through MIC analysis, we determined that the input variables of the return water temperature prediction model are outdoor temperature (To) lagged by 1 step, indoor humidity (Ho) lagged by 1 step, indoor temperature (Tn) lagged by 1 step, supply water temperature (Tsup) lagged by 1 step, and supply water flow (msup) lagged by 1 step. Using the Att-BiGRU algorithm for training, we obtained a return water temperature prediction sub-model with a mean absolute error (MAE) less than or equal to 0.11. This model can accurately predict return water temperature, providing strong support for energy efficiency management and optimization of air conditioning systems. When establishing the fixed-frequency fan energy consumption model, based on the similarity law and the actual situation of the fan in the combined air conditioning unit, it was determined that the fan power is proportional to the cube of its operating frequency. Since the fan operates at a fixed frequency, its operating air volume and power remain unchanged. Through calculation, the actual operating power of the fan can be obtained, and then the total energy consumption of the air conditioning system can be obtained. This model provides an important basis for evaluating and optimizing the energy consumption of air conditioning systems. The constructed combined air-conditioning unit model can accurately predict the return water temperature and effectively evaluate the energy consumption of the air-conditioning system, providing strong support for the efficient operation and energy-saving management of the air-conditioning system.

[0118] In one embodiment, when constructing a virtual indoor temperature model or establishing a return water temperature prediction model, the model training and verification steps include: dividing the sample data into a training set and a validation set in a ratio of 7:3, using the first 10 time steps of input to predict the next time step output; using early stopping to monitor the validation set error, and terminating the training when the RMSE decreases by less than 1% after 10 consecutive iterations; calculating the mean absolute error (MAE), mean absolute percentage error (MAPE), and root mean square error (RMSE) of the validation set, selecting Att-BiGRU as the optimal basic algorithm, and optimizing its model weights by gradient descent. By dividing the sample data into a training set and a validation set in a ratio of 7:3, the present invention ensures that the model can fully learn the inherent laws of the data during training and effectively evaluate the generalization ability of the model through the validation set, thus avoiding the occurrence of overfitting. The method of using the first 10 time steps of input to predict the next time step output fully utilizes the autocorrelation of time series data and improves the model's prediction accuracy for temperature change trends. The early stopping method is used to monitor the validation set error, and training is terminated when the RMSE drops by less than 1% after 10 consecutive iterations. This strategy effectively prevents overtraining of the model during training, saves computing resources, and ensures the generalization performance of the model. By calculating the mean absolute error (MAE), mean absolute percentage error (MAPE) and root mean square error (RMSE) of the validation set, the prediction performance of the model can be comprehensively evaluated, providing a reliable basis for model selection and optimization. Att-BiGRU is selected as the optimal basic algorithm, which fully utilizes its powerful modeling ability of bidirectional gated recurrent unit for time series data and the ability of the attention mechanism to capture key information, thereby improving the prediction accuracy and stability of the model. The model weights are optimized by gradient descent, ensuring that the model can gradually approach the optimal solution during training, further improving the prediction performance of the model. The present invention constructs a training and verification step for a virtual indoor temperature model or a return water temperature prediction model. By scientifically and rationally dividing the data set, selecting an effective prediction strategy, and adopting advanced algorithms and optimization methods, the prediction accuracy and generalization ability of the model are significantly improved, providing strong support for research and application in the field of temperature prediction.

[0119] The present invention also includes a model integration method based on a simulation platform: building a building data model in a Matlab environment, establishing a Modbus / OPC communication link with an MPC controller through the Kepserverex software; integrating a virtual indoor temperature model, an air source heat pump model, a water pump model, and an air conditioning unit model in parallel, and ensuring the consistency of simulation steps through a time synchronization module; generating model output signals including: real-time temperature prediction value Tn+1, host energy consumption W ASHP , water pump energy consumption W pump , fan energy consumption W fanand maps it to the register address of kepserverex for the controller to read in real time.

[0120] The communication link construction specifically includes: creating a virtual device in Kepserverex and binding the input and output variables of the Matlab model to Modbus TCP protocol registers; setting the simulation step size to 5 minutes, and completing the closed-loop process of data writing → model calculation → register update → controller reading within each step; configuring the OPC UA client on the controller side, regularly reading the predicted temperature and energy consumption data in the registers, and feeding it back to the MPC optimization algorithm.

[0121] The present invention also includes an online model update method: deploying an operational data monitoring module to collect actual temperature values ​​T_real and model prediction values ​​T_pred in real time; triggering a model retraining process when the daily average prediction error exceeds a threshold e (e = 0.5°C); using incremental learning, adding the latest seven days of data to the training set, updating the Att-BiGRU network weights, and maintaining model prediction accuracy. Application of this method to the central air-conditioning system of a railway passenger station includes: dividing the waiting hall into multiple thermal zones and performing virtual temperature modeling in parallel for each zone; dynamically modifying the Np input parameter based on the zone's passenger density and simultaneously adjusting the pump frequency setpoint; and calculating the optimal energy consumption strategy through an MPC controller, outputting water supply temperature, flow rate, and equipment start / stop commands to an on-site PLC actuator.

[0122] Corresponding to the method, the present invention also discloses a multi-control object modeling system for an air-conditioning system, including: a sensor data acquisition module, used to obtain multi-source data of outdoor temperature, humidity, indoor temperature and humidity, and equipment operating frequency in real time; a feature processing module, configured to: perform principal component analysis (PCA) dimensionality reduction processing on the original data; screen key lag step features through a time-shifted MIC algorithm; a multi-model construction module, including: a virtual temperature prediction unit based on Att-BiGRU; a host energy consumption prediction unit based on stacked width learning; a water pump energy consumption unit based on nonlinear fitting; a fan energy consumption unit based on fixed parameters; a simulation integration platform, used to integrate sub-models and establish an OPC communication interface through kepserverex; an optimization control module, receiving the predicted temperature and energy consumption data output by the model, and executing the MPC algorithm to generate optimization control instructions.

[0123] The feature processing module includes: a MIC analysis subunit, which calculates the maximum information coefficient between variables and selects the lag step; a PCA dimensionality reduction subunit, which performs feature decomposition on To and Ho and outputs the principal components after dimensionality reduction; and a dynamic variable selection unit, which associates Toa1, Toa2, Hoa1, and Hoa2 with different device model inputs based on thermodynamic rules.

[0124] The multi-model construction module further includes: a model verification unit, configured to: divide the training set and the verification set; calculate the MAE, MAPE, and RMSE indicators; compare the performance of Att-LSTM, Att-GRU, and Att-BiLSTM, and select Att-BiGRU as the basic algorithm.

[0125] The simulation integration platform includes: a Matlab model container that runs virtual temperature prediction and equipment energy consumption calculation programs; Kepserverex communication middleware that configures the Modbus register mapping table; and a time synchronizer that triggers model recalculation and updates register data in a 5-minute cycle.

[0126] The present invention also includes an online update module, which includes: a data quality detection unit, which identifies abnormal sensor data and triggers a cleaning process; an incremental learning unit, which loads the latest data for retraining when the model prediction error exceeds the limit; and a version management unit, which saves historical model parameters and supports rollback to a stable version.

[0127] The present invention discloses a multi-control object modeling method for an air-conditioning system, comprising the following steps: collecting building environment data and air-conditioning equipment operation data, wherein the building environment data includes indoor temperature, outdoor temperature, humidity, occupant density, and solar radiation intensity, and the air-conditioning equipment operation data includes host water supply temperature, return water temperature, flow rate, and equipment frequency; performing time-shift correlation analysis on the collected building environment data to determine the lag effect step length of each environmental parameter on the target variable, and performing dimensionality reduction processing on the high-dimensional environmental parameters through principal component analysis to generate a feature input variable set; constructing a bidirectional gated recurrent unit network model based on an attention mechanism, inputting the feature input variable set and the air-conditioning equipment operation data into the model, training to obtain a virtual indoor temperature prediction model, and outputting a temperature prediction value and a confidence interval for a preset time period in the future; establishing a compressor dynamic response model based on the operation characteristics of the air-conditioning host, and constructing a host energy consumption prediction function in combination with the equipment frequency and water supply flow data, wherein the function includes a product relationship between cooling capacity, performance coefficient, and time variable;

[0128] A cubic frequency-power correlation model for variable frequency water pumps was established. Based on the pipe network characteristic curve, the flow-frequency correction coefficient under actual working conditions was fitted to generate a dynamic water pump energy consumption calculation equation.

[0129] A hybrid modeling framework was constructed for modular air handling units, including a physics-based heat transfer equation for the surface cooler and a data-driven return water temperature prediction model. The physics equations included parameters such as headwind velocity, water flow rate, and moisture extraction coefficient.

[0130] A software-in-the-loop simulation platform was constructed, connecting the building data model with the controller algorithm via a communication protocol. A virtual indoor temperature model, host model, water pump model, and air handling unit model were integrated within the Matlab environment to form a co-simulation system. The discrete models were connected in parallel into a unified framework using the Matlab integration platform. A time synchronization module was used to ensure consistent prediction steps across all models (e.g., a 5-minute cycle). Virtual temperature and energy consumption predictions were then transmitted to the controller in real time via Modbus / OPC communication. For example, the heat pump energy consumption model predicted the host load based on the current supply water temperature and outdoor temperature, while the water pump model dynamically adjusted flow and power based on frequency fluctuations. The output of the return water temperature model was fed back to the heat pump model to form a closed loop. The controller integrated all model outputs (e.g., temperature deviations from predicted values, energy consumption surge warnings), dynamically adjusted the air conditioning system operating parameters (e.g., reducing frequency to save energy), ultimately achieving the combined optimization of temperature control and energy consumption.

[0131] Injecting actual collected disturbance signals into the simulation environment, collaboratively verifying multiple control object models, calculating the root mean square error between the predicted and actual temperatures, and triggering the model parameter adaptive adjustment mechanism when the error exceeds the threshold;

[0132] Based on the verified multi-object model, a predictive control optimization objective function is constructed, which includes temperature comfort weight, equipment energy consumption weight and control action smoothness constraints; the optimized control instructions are transmitted to the physical equipment through the KepserverEX middleware to realize closed-loop adjustment of the air-conditioning system operating parameters.

[0133] The time-shift correlation analysis specifically includes: calculating the maximum information coefficient between each environmental parameter and the target temperature variable to determine the lag time window with significant correlation; performing sliding window sampling on the outdoor temperature and humidity parameters to construct a multidimensional time series matrix; calculating the cumulative variance explanation rate of the eigenvalues ​​of each dimension through principal component analysis, and selecting the principal component with an explanation rate exceeding 95% as the feature after dimensionality reduction.

[0134] The construction of the bidirectional gated recurrent unit network model of the attention mechanism includes: introducing a temporal attention module after the bidirectional GRU layer to calculate the attention weights of the hidden state at each time step; designing a spatial attention subnet to dynamically assign weights to multivariate input features; and using residual connections to fuse the original features with the attention-weighted features to prevent gradient vanishing.

[0135] The process of constructing the host energy consumption prediction function includes: collecting the measured COP data of the host at different load rates, establishing an operating condition classification model based on fuzzy C-means clustering; constructing a stacked width learning system prediction network in various operating condition subspaces, with input parameters including supply water temperature, return water temperature difference, ambient temperature and humidity, and historical energy consumption trends; and cascading the operating condition classification model and the prediction network to form a hybrid prediction architecture for host energy consumption.

[0136] The construction of the variable frequency water pump model includes: fitting a cubic relationship at the rated frequency reference point to obtain a theoretical flow-frequency curve; collecting actual pipe network characteristic data to calculate the flow attenuation coefficient at each frequency point; and constructing a modified cubic equation including the pipe network damping coefficient.

[0137] A differential equation model for heat transfer in a surface cooler was established, and a return water temperature data-driven model based on Att-BiGRU was constructed. The input features included two-order lagged outdoor temperature and humidity, indoor temperature, and water supply parameters.

[0138] Design a model fusion strategy to enable the data-driven model to perform compensation prediction when the physical model calculation residual exceeds the threshold.

[0139] The construction of the software-in-the-loop simulation platform specifically includes: constructing a building thermodynamic equation module in Matlab / Simulink and integrating the EnergyPlus core calculation engine; establishing a real-time data channel between KepserverEX and the simulation model through the OPC UA protocol; and embedding the control logic code library of actual equipment in the simulation environment to achieve lossless migration of control algorithms.

[0140] The present invention also discloses the following content to help understand the technical solution.

[0141] The full automation, informatization, and digitization of air-conditioning systems, as well as the implementation of advanced control technologies, often rely on software that integrates relevant intelligent algorithms. Traditional software testing methods typically rely on actual systems and hardware, such as hardware-in-the-loop simulators. For early software and control algorithm development, testing using actual systems is almost impractical. Especially for large public buildings, frequent equipment control during testing makes it difficult to ensure stable system operation. Consequently, actual testing time is typically short and discontinuous. To reduce testing costs and risks, a software-in-the-loop (SIL) simulation platform for air-conditioning control systems was established. This platform is used to test source code before system and hardware configuration, thereby verifying the effectiveness of the proposed predictive control algorithm for air-conditioning systems at an early stage.

[0142] Simulation platform construction: Communication methods between various sensors and equipment in the station and the software platform are as follows: Figure 1As shown. Sensor and equipment signals are uploaded to the automation equipment data connection and acquisition software kepserverex through wired, wireless networks, edge nodes, etc. The kepserverex software converts the signals into common data formats and then uploads them to the software platform through communication protocols such as Modbus and OPC. The software platform uses intelligent algorithms to process and calculate the collected data, sends out control signals, and issues control instructions to the equipment through the kepserverex software. In order to fully simulate the communication method between the station sensors and equipment and the software platform, the following is built Figure 2 The station air-conditioning control system simulation platform shown in the figure uses Matlab software to establish the building data model and controller algorithm, and uses KepServerEx software to achieve communication between the building data model and the controller. In previous studies, software such as EnergyPlus and Trnsys were often used to simulate the building environment, but the simulated building model may differ significantly from the actual project. Therefore, the actual station environment and equipment data collected were used to establish a digital building and equipment model to simulate the actual building and equipment status. At the same time, it can also transmit the relevant internal and external interference simulation signals established by the real data. After the MPC controller is tested, it can be used directly to control the actual system, or it can be integrated with other software platforms for intelligent optimization control of the air-conditioning system.

[0143] Building data model establishment

[0144] A digital virtual indoor temperature and equipment simulation model was established based on the actual collected station environment and equipment data to simulate the actual indoor and equipment conditions.

[0145] (1) Virtual indoor temperature simulation model

[0146] A virtual indoor temperature simulation model is constructed based on the digital twin modeling method of air conditioning load based on the time-shift correlation of characteristic variables and deep learning. First, the outdoor dry bulb temperature (T o ), relative humidity (H o ), wind speed (W f ), Wind Direction (W d ), solar radiation intensity (R o ) and indoor personnel (N p ) was used as the initial environmental influencing factor for correlation analysis. The MIC values ​​of each influencing factor and indoor temperature are as follows Figure 3 shown.

[0147] from Figure 3 It can be seen that the indoor temperature and T o , H with a hysteresis step of 0-2 o and N with a lag step of 0 p High correlation with Wf 、W d and R o The correlation is low. o 、H o and N p The maximum MIC values ​​of T were 0.70, 0.63 and 0.64 respectively. o and H o Perform dimensionality reduction processing. The results of different dimensional parameters are shown in Table 1 and Table 2. Among them, T o The cumulative variance explained by the first two principal components is 98.35%, and H o The cumulative variance explanation rate of the first two principal components of T is 99.08%. Therefore, two principal components are selected as o The relevant input variables are denoted as T oa1 and T oa2 ; Select two principal components as o The relevant input variables are denoted as H oa1 and H oa2 ; The number of people in the room with a lag step of 0 is recorded as N p .

[0148] Table 1 Outdoor temperature (T o )PCA dimensionality reduction results

[0149]

[0150] Table 2 Outdoor relative humidity (H o )PCA dimensionality reduction results

[0151]

[0152] The water supply temperature T is set to 0-1 with a hysteresis step size of 0-1 sup0 、T sup1 and the water flow rate m with a hysteresis step of 0-1 sup0 、m sup1 and the indoor temperature T at the predicted time n It is also used as the input variable of the prediction model for subsequent modeling. The indoor temperature prediction model includes T oa1 、T oa2 、H oa1 、H oa2 、N p 、T n 、T sup0 、T sup1 、m sup0 and m sup1 There are 10 input variables in total, and the above variables represent variables related to outdoor dry-bulb temperature, relative humidity, number of people indoors, indoor temperature, water supply temperature, and water supply flow.

[0153] Four algorithms, Att-LSTM, Att-BiLSTM, Att-GRU, and Att-BiGRU, were used to construct a virtual indoor temperature simulation model. The simulation accuracy on the model validation set is shown in Table 3.

[0154] Table 3 Comparison of prediction performance of different algorithms

[0155]

[0156] As can be seen in Table 3, the Att-BiGRU algorithm achieves superior simulation accuracy on the model validation set to the other three algorithms, with MAE, MAPE, and RMSE values ​​of 0.069, 0.002, and 0.112, respectively. This demonstrates that the virtual indoor temperature simulation model built with the Att-BiGRU model has extremely high accuracy and can be used to simulate real building environments.

[0157] (2) Virtual air conditioning equipment model

[0158] Air source heat pump model: The air source heat pump is mainly composed of components such as a compressor, an evaporator, a condenser and a throttling device. Taking the currently commonly used compression air source heat pump as an example, its working principle is as follows: First, the compressor works on the refrigerant to change the refrigerant vapor in a low-temperature and low-pressure state into a high-temperature and high-pressure state and discharge it into the condenser, where it is converted into a liquid refrigerant through condensation and heat release. After passing through the throttling device, the pressure and temperature of the liquid refrigerant are reduced at the same time, and it is converted into a gas-liquid two-phase state and enters the evaporator. In the evaporator, it absorbs heat from the ambient medium and converts it into refrigerant gas to enter the compressor, and the above process is continuously repeated in the system to complete the heat transfer and achieve the purpose of cooling or heating. We have established a host energy consumption model for the air source heat pump, and the establishment process is as follows:

[0159] The coefficient of performance (COP) of the heat pump is calculated as shown in formula (4-1):

[0160]

[0161] Where: T low is the low-temperature heat source temperature, and is the outlet water temperature of the heat pump unit under cooling conditions, ℃; T high is the high temperature heat source temperature, and under cooling conditions it is the outdoor ambient temperature, ℃.

[0162] In addition, we also use fuzzy clustering and stacked width learning system to establish the COP prediction model of air source heat pump unit.

[0163] Host energy consumption W ASHP The calculation formula is shown in formula (4-2):

[0164] W ASHP =Q ASHP / COP*ΔT (4-2)

[0165] Where: Q ASHP is the system cooling capacity per unit time, kW; ΔT is the system operating time, h.

[0166] From formula (6-2), we can see that the host energy consumption is related to the system cooling capacity and COP. System energy consumption and water supply temperature T sup , return water temperature T ret , flow m sup , outdoor ambient temperature T o , relative humidity H o 、N number of people indoors p , indoor temperature T N And historical cooling capacity Q his Therefore, the host energy consumption model can be expressed as shown in formula (4-3):

[0167] W ASHP =f ASHP (T sup ,T ret ,m sup ,T o ,H o ,N p ,T N ,Q his )*ΔT (4-3)

[0168] Where: f ASHP It is a data-driven model related to the energy consumption calculation of air source heat pump units.

[0169] Water pump model

[0170] For variable frequency water pumps, according to the similarity law, when the fluid flow process satisfies geometric similarity, motion similarity, and dynamic similarity at the same time, the water pump flow rate is proportional to the frequency, and the electric power is proportional to the cube of the frequency. The relationship between the water pump flow rate m, the electric power Q, and the water pump frequency n is shown in equations (4-4) and (4-5):

[0171]

[0172] Where: m1 and m0 are the flow rates of the pump under actual working conditions and rated working conditions respectively, m 3 / h; n1 and n0 are the frequencies of the water pump under actual working conditions and rated working conditions, Hz; Q1 and Q0 are the electrical power of the water pump under actual working conditions and rated working conditions, kW.

[0173] For water systems where the end-load water flow can be actively regulated, equations (4-6) and (4-7) can be used to express the relationship between the flow rate, electric power, and frequency of the water pump under actual operating conditions.

[0174]

[0175] Where: a, b, c, d, e, and f are fitting coefficients.

[0176] Based on the collected data, it can be calculated that the relationship between the flow rate, electric power and frequency of the water pump under actual working conditions is shown in Equations (4-8) and (4-9).

[0177] m pump =6.5001n pump -40.5961 (4-8)

[0178] Q pump =-0.0083n pump 3 +1.0829n pump 2 -45.6339n pump +641.4388 (4-9)

[0179] Where: m pump is the flow rate of the water pump in this project under actual working conditions, m 3 / h;n pump is the frequency of the water pump under actual working conditions, Hz; Q pump is the electric power of the water pump under actual working conditions, kW.

[0180] Energy consumption of water pump W pump The calculation formula is shown in formula (4-10):

[0181] W pump =Q pump *ΔT (4-10)

[0182] Combined air handling unit model

[0183] The combined air handling unit simulation model established by the present invention includes a return water temperature model and a fan energy consumption model.

[0184] A. Return water temperature model

[0185] Combined air handling units are mainly composed of surface coolers, humidifiers, heaters, fans and other equipment. In the cooling season, surface coolers are mainly used to cool and dehumidify the air. For a single return air constant air volume air conditioning system, the indoor air is mixed with the outdoor fresh air to obtain the initial state of air treatment with a dry bulb temperature of t1, an enthalpy value of h1, and a wet bulb temperature of t s1 , the amount of air being processed is G, the chilled water flow is m, and the initial temperature of the chilled water is T sup , then the final state t2 and return water temperature T that the surface cooler can reach when processing air retThe calculation process is as follows:

[0186] (1) Calculate the wind speed V facing the surface cooler y And the water flow rate w, the calculation formula is shown in formula (4-11) and formula (4-12):

[0187]

[0188] Where: F y is the windward area of ​​the surface cooler, m 2 ; ρ is the density of water, kg / m 3 ;f w is the cross-sectional area of ​​water flow, m 2 .

[0189] (2) Calculate the general heat exchange efficiency E' that the surface cooler can provide. This value can be obtained from equipment samples.

[0190] (3) Assume that the dry-bulb temperature of the air at the final state is t2.

[0191] Generally, t2=t w1 +(4~6)℃ to determine the final dry bulb temperature of the air. The final wet bulb temperature t s2 The calculation formula is shown in formula (4-13):

[0192] t s2 =t2-(t1-t s1 )(1-E') (4-13)

[0193] The calculation formula for the final state enthalpy h2 is shown in formula (4-14):

[0194] h2=0.0707t s 2 2+0.6452t s2 +16.18 (4-14)

[0195] (4) Calculate the wetting coefficient ξ, as shown in (4-15):

[0196]

[0197] Among them, c p is the constant pressure specific heat capacity of water, take 4.2×10 3 J / (kg·℃);

[0198] (5) Calculate the heat transfer coefficient K s .

[0199] The empirical formula for the heat transfer coefficient of the heat exchanger dehumidification cooling process is shown in formula (4-16):

[0200]

[0201] Where: A, P, B, m, n are coefficients and exponents obtained from experiments.

[0202] (6) Calculate the total heat exchange efficiency E' that the surface cooler can achieve g , the calculation formula is shown in formula (4-17) to formula (4-19):

[0203]

[0204] Where, F represents the windward area of ​​the surface cooler;

[0205]

[0206] Where c represents the specific heat capacity of water;

[0207] (7) Calculate the required total heat exchange efficiency E g And with E' g In comparison, E g The calculation formula is shown in formula (4-20):

[0208]

[0209] When|E g -E' g When |≤δ (δ is the allowable error, generally 0.01), it proves that the assumed t2 is appropriate. g -E' g |>δ, t2 should be reset and the calculation should be performed again.

[0210] (8) Calculate the chilled water return temperature T ret , the calculation formula is shown in formula (4-21).

[0211]

[0212] From the above analysis, it can be seen that for air handling units, a data-driven approach is appropriate for establishing a black box model of the unit. This embodiment uses a digital twin modeling method based on time-shifted correlation of characteristic variables and deep learning to establish a return water temperature prediction model after heat exchange. The modeling process is as follows:

[0213] (1) Determine the input variables

[0214] By analyzing the above return water temperature calculation process, it can be found that the return water temperature is related to the fresh air dry-bulb temperature, fresh air wet-bulb temperature, return air dry-bulb temperature, return air wet-bulb temperature, supply water temperature, and supply water flow. The fresh air dry-bulb temperature can be replaced by the outdoor temperature, and the return air dry-bulb temperature can be replaced by the indoor temperature. Since the wet-bulb temperature is difficult to measure, relative humidity is used as the input variable instead of the wet-bulb temperature. The initial influencing factor of the return water temperature prediction model can be determined as the outdoor temperature T o , Outdoor relative humidity H o , indoor dry bulb temperature T n , indoor relative humidity H n , water supply temperature T sup And water supply flow m sup The MIC values ​​of various influencing factors and return water temperature are as follows: Figure 4 shown.

[0215] from Figure 4 It can be seen that the return water temperature is related to the T with a hysteresis step of 0-1. o , H with a hysteresis step size of 0-1 o , T with a hysteresis step size of 0-1 n , T with a hysteresis step size of 0-1 sup and m with a lag step of 0-1 sup The input variables are denoted as T o1 、T o2 、H o1 、H o2 、T n1 、T n2 、T sup1 、T sup2 、m sup1 and m sup2 , the above variables represent variables related to outdoor dry-bulb temperature, outdoor relative humidity, indoor dry-bulb temperature, water supply temperature and water supply flow rate, respectively.

[0216] (2) Model establishment

[0217] Four algorithms, Att-LSTM, Att-BiLSTM, Att-GRU, and Att-BiGRU, were used to construct the return water temperature simulation model. The simulation accuracy on the model validation set is shown in Table 4.

[0218] As can be seen in Table 4, the Att-BiGRU algorithm achieves superior simulation accuracy on the model validation set to the other three algorithms, with MAE, MAPE, and RMSE values ​​of 0.102, 0.006, and 0.189, respectively. This indicates that the return water temperature simulation model established by the Att-BiGRU model has extremely high accuracy and can be used to simulate real return water temperatures.

[0219] Table 4 Comparison of prediction performance of different algorithms

[0220]

[0221] B. Fan Energy Consumption Model

[0222] The energy consumption of the combined air conditioning unit is the fan energy consumption. According to the similarity law, the fan power is proportional to the cube of its operating frequency, that is:

[0223]

[0224] Where: n fan1 and n fan2 are the frequencies of the fan under actual working conditions and rated working conditions, Hz; Q fan1 and Q fan0 They are the electric power of the fan under actual working conditions and rated working conditions, kW.

[0225] The fan in the combined air conditioning unit is a fixed frequency fan, and its operating air volume and power remain unchanged. Therefore, the fan energy consumption W fan The calculation formula is shown in formula (4-23):

[0226] W fan =Q fan *ΔT (4-23)

[0227] Where: Q fan is the actual operating power of the fan, kW.

[0228] The total energy consumption of the air conditioning system can be calculated by using formula (6-3), formula (6-10) and formula (6-23), as shown in formula (4-24):

[0229] W total =(f ASHP +Q pump +Q fan )*ΔT (4-24)

[0230] Although the preferred embodiments of the present invention have been described in detail, those skilled in the art will be able to make other changes and modifications to these embodiments after understanding the basic inventive concepts. Therefore, the appended claims are intended to cover these preferred embodiments and all changes and modifications that fall within the scope of the present invention. The foregoing is only a preferred embodiment of the present invention and is not intended to limit its scope. It should be understood that any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A multi-control object modeling method for air conditioning systems, characterized in that: include: Acquire multi-source operating data collected by air conditioning system sensors, the multi-source operating data including outdoor temperature To, outdoor relative humidity Ho, indoor temperature Tn, supply water temperature Tsup, return water temperature Tret, supply water flow rate msup, water pump frequency npump, and number of indoor occupants Np; perform principal component analysis and time-shift correlation analysis on the multi-source operating data to screen out core input variables corresponding to target control objects; wherein the target control objects include virtual indoor temperature, energy consumption of air source heat pump units, water pump energy consumption, and fan energy consumption; The time-shift correlation analysis determines the lag step size of each variable through the maximum information coefficient and screens key time series features. Based on the screened core input variables, a data-driven model for each control object is constructed, and the model is trained and verified to form a combined multi-control object simulation framework; wherein, the combined multi-control object simulation framework is used to predict the energy consumption and environmental status of the air conditioning system; Build a modular air conditioning unit model, including a return water temperature prediction model: For a single return constant air volume air conditioning system, the indoor air is mixed with the outdoor fresh air to obtain the initial dry bulb temperature of the air treatment. , the enthalpy is , the wet bulb temperature is , the amount of air being processed is G, the chilled water flow is m, and the initial temperature of the chilled water is , then the final state that the air can reach after being treated by the surface cooler and return water temperature The process is as follows: A1. Calculate the wind speed facing the surface cooler And the water flow rate w, the calculation formula is shown in formula (2-1) and formula (2-2): Where: is the windward area of ​​the surface cooler, m 2 ; is the density of water, kg / m 3 ; is the cross-sectional area of ​​water flow, m 2 ; A2. Calculate the general heat exchange efficiency that a surface cooler can provide , this value is obtained through equipment samples; A3. Assume that the dry bulb temperature of the air at the final state is ; according to Determine the final dry-bulb temperature of the air; the final wet-bulb temperature The calculation formula is shown in formula (2-3): Enthalpy of the final state The calculation formula is shown in formula (2-4): A4. Calculate the wet coefficient , the calculation formula is shown in (2-5): A5. Calculate the heat transfer coefficient ; The empirical formula for the heat transfer coefficient of the heat exchanger dehumidification cooling process is shown in formula (2-6): Where: A , P , B , m , n are the coefficients and exponents obtained from the experiment; A6. Calculate the total heat exchange efficiency that the surface cooler can achieve , the calculation formula is shown in formula (2-7) to formula (2-9): A7. Calculate the required total heat exchange efficiency and with In comparison, The calculation formula is shown in formula (2-10): when When proving the hypothesis Suitable; if , you should reset Perform the calculation again; A8. Calculate the chilled water return temperature , the calculation formula is shown in formula (2-11); 。 2. The multi-control object modeling method for air conditioning system according to claim 1, characterized in that: The steps of constructing a virtual indoor temperature model include: determining input variables through MIC analysis, including the 0-3 step lagged characteristics of the outdoor temperature To, the 0-2 step lagged characteristics of the outdoor humidity Ho, and the current number of people indoors Np; performing PCA dimensionality reduction processing on the outdoor temperature and humidity, and taking the principal components with cumulative variance contribution rates ≥ 98% as the input variables after dimensionality reduction, denoted as Toa1, Toa2 and Hoa1, Hoa2 respectively; combining the reduced dimensionality variables with the historical water supply temperature Tsup0-Tsup1, the historical water supply flow msup0-msup1, and the current indoor temperature Tn to construct a training set with 10 input variables; using the Att-BiGRU algorithm to model the training set, and optimizing the network parameters through back propagation to obtain a virtual indoor temperature simulation model with a prediction error MAE ≤ 0.07 and RMSE ≤ 0.

12.

3. The multi-control object modeling method for air conditioning system according to claim 1, characterized in that: The steps of constructing the energy consumption model of the air source heat pump unit include: screening input variables, including the current supply water temperature Tsup, return water temperature Tret, flow rate msup, outdoor temperature To, outdoor humidity Ho, indoor occupancy Np, indoor temperature Tn, and historical cooling capacity Qhis; implanting the input variables into the stacked width learning system to establish the energy consumption prediction sub-model, whose output is the heat pump performance coefficient COP, and the calculation relationship satisfies: ,in is the outlet water temperature, is the outdoor temperature; Output the output host energy consumption value, where is the system cooling capacity per unit time, ; is the system running time, h.

4. The multi-control object modeling method for air conditioning system according to claim 1, characterized in that: When building a water pump energy consumption model, perform the following operations: The flow rate of the water pump is proportional to the frequency, and the electric power is proportional to the cube of the frequency; the flow rate of the water pump is proportional to the frequency. m Electric power Q and pump frequency n The relationship between them is shown in formula (1-1) and formula (1-2): Where: and are the flow rates of the pump under actual working conditions and rated working conditions, m 3 / h; and are the frequencies of the pump under actual working conditions and rated working conditions, Hz; are the electric power of the water pump under actual working conditions and rated working conditions, kW; Formulas (1-3) and (1-4) are used to express the relationship between the flow rate, electric power and frequency of the water pump under actual working conditions; Where: a 、 b 、 c 、 d 、 e 、 f are the coefficients fitted by the measured data.

5. The multi-control object modeling method for air conditioning system according to claim 1, characterized in that: Construct a combined air-conditioning unit model, including: establishing a return water temperature prediction model: through MIC analysis, determine the input variables as To lag 1 step, Ho lag 1 step, Tn or Tn lag 1 step, Tsup lag 1 step and msup lag 1 step, and use the Att-BiGRU algorithm to train to obtain a return water temperature prediction sub-model with MAE ≤ 0.11; establish a fixed-frequency fan energy consumption model: the energy consumption of the combined air-conditioning unit is the fan energy consumption; according to the similarity law, the fan power is proportional to the cube of its operating frequency. The fan in the combined air-conditioning unit is a fixed-frequency fan, and its operating air volume and power remain unchanged; fan energy consumption The calculation formula is shown in formula (3-1): Where: is the actual operating power of the fan, kW.

6. The multi-control object modeling method for air conditioning system according to claim 5, characterized in that: When establishing the return water temperature prediction model, the model training and verification steps include: dividing the sample data into a training set and a validation set in a ratio of 7:3, using the first 10 time steps as input to predict the output of the next time step; using the early stopping method to monitor the validation set error, and terminating the training when the RMSE drops by less than 1% after 10 consecutive iterations; calculating the mean absolute error, mean absolute percentage error, and root mean square error of the validation set, selecting Att-BiGRU as the optimal basic algorithm, and optimizing its model weights through gradient descent.

7. The multi-control object modeling method for air conditioning system according to claim 1, characterized in that: It also includes a model integration method based on the simulation platform: building a building data model in the Matlab environment, establishing a Modbus / OPC communication link with the MPC controller through the Kepserverex software; integrating the virtual indoor temperature model, air source heat pump model, water pump model and air conditioning unit model in parallel, and ensuring the consistency of simulation steps through the time synchronization module; generating model output signals including: real-time temperature prediction value Tn+1, host energy consumption , water pump energy consumption , fan energy consumption and maps it to the register address of kepserverex for the controller to read.

8. The multi-control object modeling method for air conditioning system according to claim 7, characterized in that: The communication link construction specifically includes: creating a virtual device in Kepserverex and binding the input and output variables of the Matlab model to the Modbus TCP protocol registers; setting the simulation step size to 5 minutes, and completing the closed-loop process of data writing → model calculation → register update → controller reading within each step; configuring the OPC UA client on the controller side to regularly read the predicted temperature and energy consumption data in the registers.

9. The multi-control object modeling method for air conditioning system according to claim 1, characterized in that: It also includes a model online update method: deploying an operation data monitoring module to collect the actual temperature value T_real and the model prediction value T_pred in real time; when the daily average prediction error exceeds the threshold e, the model retraining process is triggered; using incremental learning, the latest 7 days of data are added to the training set, and the Att-BiGRU network weights are updated to maintain the model prediction accuracy.

Images