An optimized operation method of an air source heat pump-fan coil central air conditioning system

Abstract

The application belongs to the technical field of central air conditioning system optimal operation, and particularly discloses an optimal operation method of an air source heat pump-fan coil central air conditioning system, which comprises the following steps: firstly, obtaining the influencing factors of building load and outdoor meteorological parameters, constructing a central air conditioning load prediction model, an outdoor meteorological parameter prediction model and a data-driven model of an air source heat pump and a water pump, and predicting the total energy consumption of the system, the load demand at the next moment, the outdoor environment temperature and the outdoor relative humidity; secondly, setting an optimization objective function and constraint conditions; using a grid search optimization algorithm to solve the optimization objective function and the constraint conditions, obtaining the host set temperature at the next moment and the water pump set frequency at the next moment, and completing the optimization process. The application solves the problem of strong hysteresis of the central air conditioning water system, and realizes the optimization and solution of the host set temperature at the next moment and the water pump set frequency at the next moment according to the operation data at the current moment.

Description

Technical Field

[0001] This invention belongs to the field of air conditioning system optimization operation technology, specifically relating to an optimization operation method for an air source heat pump-fan coil central air conditioning system. Background Technology

[0002] The actual operating status of each device in a central air conditioning system is affected by factors such as outdoor environmental parameters and end-user demands. Failure to find the optimal operating point will impact energy consumption and even lead to unsatisfactory indoor thermal comfort. However, existing air conditioning systems suffer from problems such as oversized equipment and low equipment utilization, indicating significant room for energy-saving optimization. Therefore, finding the optimal operating state of the equipment under different outdoor meteorological parameters and load demands is crucial for optimizing the operation of central air conditioning systems. Current research on the optimized operation of central air conditioning chilled and hot water systems mainly focuses on two aspects: optimal control based on the number of units and optimal control based on set parameters.

[0003] There are two main modes for optimizing the number of units to be operated. One is start-stop control, which starts up the main unit, water pump, cooling tower and other equipment one by one according to the user's load demand. When a single unit cannot meet the user's environmental needs, additional equipment is turned on to provide cooling or heating. The other is load distribution control, which turns on multiple units at the same time and distributes the cooling and heating capacity according to the equipment performance, thereby solving the optimal load distribution state of each unit.

[0004] To address the issue of strong coupling in central air conditioning systems, this study focuses on optimizing the operation strategies of chilled and hot water systems. The goal is to calculate the parameter combination that minimizes energy consumption of the central air conditioning system while meeting indoor load requirements. The optimization path involves three main steps: establishing an equipment model, setting constraints, and finally, finding the optimal solution.

[0005] In addition, advancements in computer theory have provided technical support for the control of central air conditioning systems. The adjustment of chilled water system parameters no longer relies on local control but instead employs PID control. However, traditional PID control struggles to handle the complex coupling relationships between variables, making it difficult to achieve optimal system control. Furthermore, this control method is feedback control; when it's discovered that equipment energy efficiency or indoor thermal comfort does not meet set requirements, adjusting the operating parameters of the main unit or water pumps is too late, hindering its effectiveness in handling chilled water systems with significant lag. Summary of the Invention

[0006] The purpose of this invention is to solve the problem of strong lag in central air conditioning water systems, and to propose an optimized operation method for air source heat pump-fan coil central air conditioning systems.

[0007] The technical solution of this invention is: an optimized operation method for an air source heat pump-fan coil central air conditioning system, comprising the following steps:

[0008] S1. Obtain the influencing factors of building load and outdoor meteorological parameters;

[0009] S2. Construct a central air conditioning load forecasting model and an outdoor meteorological parameter forecasting model based on influencing factors, and predict the load demand Q at the next moment. t+1 The outdoor ambient temperature Tdb at the next moment t+1 and the outdoor relative humidity (RH) at the next moment t+1 ;

[0010] S3. Determine the main influencing factors on the operating performance of air source heat pumps and water pumps, construct data-driven models for air source heat pumps and water pumps, and conduct simulation operation based on the data-driven models to obtain the total energy consumption of the central air conditioning system.

[0011] S4. Based on the total energy consumption of the central air conditioning system and the load demand Q at the next moment. t+1 The outdoor ambient temperature Tdb at the next moment t+1 and the outdoor relative humidity (RH) at the next moment t+1 Based on the prediction results, set the optimization objective function and constraints;

[0012] S5. Solve the objective function and constraints using a grid search optimization algorithm to obtain the host setpoint temperature Tset for the next time step. t+1 And the water pump set frequency Nset for the next moment t+1 The optimization process is now complete.

[0013] The beneficial effects of this invention are:

[0014] This invention addresses the characteristics of system complexity, coupling, and strong hysteresis by establishing a predictive model and equipment model. Based on the current operating data, it optimizes the load demand, host set temperature, and pump set frequency for the next moment, thereby improving system operating efficiency and reducing operating energy consumption.

[0015] Preferably, step S2 specifically includes the following steps:

[0016] S21. Normalize the influencing factors. The formula for normalization is:

[0017]

[0018] Where v' is the normalized value, v is the value before normalization, and v max v is the maximum value of the parameters before normalization. minThis represents the minimum value of the parameters before normalization.

[0019] S22. Calculate the correlation coefficient between the normalized influencing factors and the cooling and heating loads. The formula for calculating the correlation coefficient is as follows:

[0020]

[0021] Where, r p Here, X is the correlation coefficient, q is the number of influencing factors, and X is the number of influencing factors. s For the s-th influencing factor, Y is the average value of the influencing factors. s For the s-th building load, This represents the average building load.

[0022] S23. Determine the input and output parameters of the XGBoost model based on the correlation coefficient, set the hyperparameters of the XGBoost model, and train the XGBoost model.

[0023] S24. The trained XGBoost model is optimized using the Sparrow Search algorithm to obtain the central air conditioning load prediction model and the outdoor meteorological parameter prediction model.

[0024] S25. Input the current time t, date, outdoor dry-bulb temperature at time t, outdoor wet-bulb temperature at time t, and building load at time t into the central air conditioning load prediction model to obtain the output data of the central air conditioning load prediction model;

[0025] S26. Input the current time t, the outdoor dry-bulb temperature at time t, and the outdoor wet-bulb temperature at time t into the outdoor meteorological parameter prediction model to obtain the output data of the outdoor meteorological parameter prediction model;

[0026] S27. Perform inverse normalization on the output data of the central air conditioning load forecasting model to obtain the load demand Q at the next time step. t+1 The output data of the outdoor meteorological parameter prediction model is then inversely normalized to obtain the outdoor ambient temperature Tdb at the next moment. t+1 and the outdoor relative humidity (RH) at the next moment t+1 The formula for the inverse normalization process is:

[0027] y = y'·(y max -y min )+y min

[0028] Where y is the output data after denormalization, and y' is the output data before denormalization. max y represents the maximum value of the load before normalization in the training data. minThis represents the minimum load value in the training data before normalization.

[0029] Preferably, the objective function of the XGBoost model during training is:

[0030]

[0031] Where Obj is the objective function of the XGBoost model, I j Let g be the set of all samples belonging to leaf node j. l h is the first-order partial derivative of the l-th sample in leaf node j. l Let λ be the second partial derivative of the l-th sample in leaf node j, λ be the regularization parameter, n be the total number of training data samples, γT be the regularization coefficient, γ be the adjustment coefficient, and T be the number of leaf nodes in the tree.

[0032] The beneficial effects of the above preferred solution are:

[0033] This preferred scheme uses the XGBoost model to construct a central air conditioning load prediction model and an outdoor meteorological parameter prediction model. It considers the current time t, date, outdoor dry-bulb temperature at time t, outdoor wet-bulb temperature at time t, and building load at time t as input parameters for the central air conditioning load prediction model. This avoids the situation where the load value at the next time moment is significantly different due to the strong heat storage capacity of the building body and the large changes in indoor thermal disturbance, thus improving the prediction accuracy.

[0034] Preferably, the data-driven models for the air source heat pump and water pump mentioned in step S3 include an air source hot water pump model and a water pump model.

[0035] Preferably, the objective function in step S4 is the total energy consumption of the system at the next time step, specifically expressed as follows:

[0036] P total(t+1) =P ashp(t+1) +P pump(t+1)

[0037] =f1(Tdb (t+1) ,RH (t+1) ,Tset (t+1) ,Nset (t+1) Q (t+1) )+f3(Nset (t+1) )

[0038] Among them, P total(t+1) Let be the total energy consumption of the central air conditioning system at time t+1, in kW and P. ashp(t+1) Let be the operating energy consumption of the heat pump at time t+1, in kW and P. pump(t+1) Let be the energy consumption of the water pump at time t+1, in kW and Tdb.(t+1) Let t+1 be the outdoor dry-bulb temperature, in °C and RH. (t+1) Let Tset be the outdoor relative humidity at time t+1, in % (t+1) Set the host temperature at time t+1, in °C, Nset. (t+1) Set the frequency Q of the water pump at time t+1. (t+1) Let f1(·) represent the load demand at time t+1, in kW, and let f1(·) represent the parameter Tdb. (t+1) RH (t+1) ,Tset (t+1) ,Nset (t+1) and Q (t+1) The functional relationship with the host is obtained from the air source hot water pump model, where f3(·) represents the parameter Nset. (t+1) The functional relationship with the water pump is obtained from the water pump model.

[0039] Preferably, the constraints in step S4 include host set temperature constraints, water pump set frequency constraints, and indoor thermal comfort constraints; the indoor thermal comfort constraint is the maximum heat exchange capacity of the terminal.

[0040] Preferably, the host computer sets the temperature constraint as follows:

[0041]

[0042] Among them, Tset (t+1) Set the host temperature at time t+1.

[0043] Preferably, the frequency constraint of the water pump is specifically as follows:

[0044] 30Hz≤Nset (t+1) ≤50Hz

[0045] Among them, Nset (t+1) Set the frequency of the water pump at time t+1.

[0046] Preferably, the formula for calculating the maximum heat exchange capacity at the terminal is:

[0047]

[0048] Among them, y totalc The sum of the summer heat exchange capacities of all fan coil units, i.e., the maximum summer terminal heat exchange capacity, y totalh This represents the sum of the winter heat exchange capacities of all fan coil units, i.e., the maximum winter terminal heat exchange capacity, where p is the number of different types of fan coil units, and y is the sum of the winter heat exchange capacities of all fan coil units. ic For the summer heat exchange capacity of a single fan coil unit, y ih For the winter heat exchange capacity of a single fan coil unit, k ic Let k be the slope of the summer heat transfer capacity equation.ih Let u be the slope of the winter heat exchange capacity equation, and b be the supply water temperature. ic b is the intercept in the summer heat exchange capacity. ih k is the intercept in the winter heat exchange capacity. totalc k is the slope of the equation for the maximum heat transfer capacity at the end of summer. totalh Let b be the slope of the equation for the maximum heat transfer capacity at the end of winter. totalc b is the intercept of the maximum heat transfer capacity at the summer terminal. totalh is the intercept of the maximum heat exchange capacity at the terminal in winter, i is the total number of fan coil unit types, i∈[1,7], and g and a represent the maximum and minimum values ​​of the number of different models, respectively.

[0049] In this preferred solution, the sum of the main unit energy consumption and the water pump energy consumption at the next moment is used as the optimization objective function, and the main unit set temperature, the water pump set frequency, and the maximum heat exchange capacity of the terminal are used as constraints, which improves the system operating efficiency and reduces operating energy consumption.

[0050] Preferably, step S5 specifically includes the following steps:

[0051] S51. Set the optimization step size and optimization range for the host set temperature and the water pump set frequency respectively;

[0052] S52. Set the parameter grid and traverse all grid nodes according to the optimization step size and optimization interval;

[0053] S53. Identify all nodes whose heat exchange capacity meets the load demand, and substitute all nodes into the objective function to calculate the node corresponding to the minimum objective function. This node is the host setpoint Tset for the next time step. t+1 And the water pump set frequency Nset for the next moment t+1 The optimal combination is found to complete the optimization process.

[0054] The beneficial effects of the above preferred solution are:

[0055] When solving for the objective function and constraints, the grid search optimization algorithm systematically traverses all possible combinations in the parameter space using an exhaustive method. This ensures that all potential optimal solutions are found within a finite time, thus accurately solving for the objective function and constraints, effectively improving system efficiency and reducing energy consumption. Attached Figure Description

[0056] Figure 1 The diagram shows a flowchart of an optimized operation method for an air source heat pump-fan coil central air conditioning system.

[0057] Figure 2The diagram shown is a schematic diagram of the optimized energy consumption analysis for summer operating condition 1 provided in Embodiment 2 of the present invention.

[0058] Figure 3 The diagram shown is a schematic diagram of the optimized energy consumption analysis for summer operating condition 2 provided in Embodiment 2 of the present invention.

[0059] Figure 4 The diagram shown is a schematic diagram of the optimized energy consumption analysis for summer operating condition 3 provided in Embodiment 2 of the present invention.

[0060] Figure 5 The diagram shown is a schematic diagram of the optimized energy consumption analysis for summer operating condition 4 provided in Embodiment 2 of the present invention.

[0061] Figure 6 The diagram shown is a schematic diagram of the optimized energy consumption analysis for winter operating condition 1 provided in Embodiment 2 of the present invention.

[0062] Figure 7 The diagram shown is a schematic diagram of the optimized energy consumption analysis for winter operating condition 2 provided in Embodiment 2 of the present invention.

[0063] Figure 8 The diagram shown is a schematic diagram of the optimized energy consumption analysis for winter operating condition 3 provided in Embodiment 2 of the present invention.

[0064] Figure 9 The diagram shown is a schematic diagram of the optimized energy consumption analysis for winter operating condition 4 provided in Embodiment 2 of the present invention.

[0065] Figure 10 The figure shown is a summer optimization result diagram provided in Embodiment 3 of the present invention.

[0066] Figure 11 The figure shown is a winter optimization result diagram provided in Embodiment 3 of the present invention.

[0067] Figure 12 The figure shown is a comparison of energy consumption before and after summer optimization provided in Embodiment 3 of the present invention.

[0068] Figure 13 The figure shown is a comparison of energy consumption before and after winter optimization provided in Embodiment 3 of the present invention. Detailed Implementation

[0069] Exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be understood that the embodiments shown and described in the drawings are merely exemplary and are intended to illustrate the principles and spirit of the invention, and are not intended to limit the scope of the invention.

[0070] Example 1:

[0071] like Figure 1As shown, an optimized operation method for an air source heat pump-fan coil central air conditioning system includes the following steps:

[0072] S1. Obtain the influencing factors of building load and outdoor meteorological parameters;

[0073] S2. Construct a central air conditioning load forecasting model and an outdoor meteorological parameter forecasting model based on influencing factors, and predict the load demand Q at the next moment. t+1 The outdoor ambient temperature Tdb at the next moment t+1 and the outdoor relative humidity (RH) at the next moment t+1 ;

[0074] S3. Determine the main influencing factors on the operating performance of air source heat pumps and water pumps, construct data-driven models for air source heat pumps and water pumps, and use the Type155 component in TRNSYS software to simulate the energy consumption of the entire cooling and heating seasons based on the data-driven models, and obtain the total energy consumption of the central air conditioning system. The set temperature of chilled water is 10℃, the pump frequency is 50Hz, the set temperature of hot water in winter is 42℃, and the pump frequency is 50Hz.

[0075] S4. Based on the total energy consumption of the central air conditioning system and the load demand Q at the next moment. t+1 The outdoor ambient temperature Tdb at the next moment t+1 and the outdoor relative humidity (RH) at the next moment t+1 Based on the prediction results, set the optimization objective function and constraints;

[0076] S5. Solve the objective function and constraints using a grid search optimization algorithm to obtain the host setpoint temperature Tset for the next time step. t+1 And the water pump set frequency Nset for the next moment t+1 The optimization process is now complete.

[0077] In this embodiment, step S2 specifically includes the following steps:

[0078] S21. Normalize the influencing factors. The formula for normalization is:

[0079]

[0080] Where v' is the normalized value, v is the value before normalization, and v max v is the maximum value of the parameters before normalization. min This represents the minimum value of the parameters before normalization.

[0081] S22. Calculate the correlation coefficient between the normalized influencing factors and the cooling and heating loads. The formula for calculating the correlation coefficient is as follows:

[0082]

[0083] Where, r p Here, X is the correlation coefficient, q is the number of influencing factors, and X is the number of influencing factors. s For the s-th influencing factor, Y is the average value of the influencing factors. s For the s-th building load, This represents the average building load.

[0084] S23. Determine the input and output parameters of the XGBoost model based on the correlation coefficient, set the hyperparameters of the XGBoost model, and train the XGBoost model; the hyperparameters include: maximum tree depth, learning rate, minimum child weight, subsample of base learning tree dataset, number of base learning trees (n_estimators), colsample of base learning tree features, and regularization coefficient γT;

[0085] The maximum tree depth represents the maximum number of layers that a single decision tree can reach during learning. The greater the tree depth, the better the model can reflect the complex relationships between features, but the computation time will be longer, and overfitting is more likely to occur. However, if the tree depth is too small, the accuracy of the model will be reduced, so it is set to 6.

[0086] A higher learning rate setting will give the decision tree the largest weight, but it is easy to cause model instability; while a lower setting will affect the calculation speed and suppress the weight of each tree in the whole model, so the value is set to 0.3.

[0087] The minimum node weight represents the minimum number of samples in each node. When the number of samples in a node is less than this value, the decision tree will stop splitting. Similarly, a larger value can easily lead to underfitting of the model, while a smaller value can easily lead to overfitting. Therefore, this value is set to 1.

[0088] The proportion of the base learning tree dataset controls the proportion of random sampling for each tree, so this value is set to 1;

[0089] The number of base learning trees represents the number of decision trees generated, also known as the number of iterations, and this value is set to 50;

[0090] The base learning tree feature ratio is the proportion of sample features selected by each decision tree, and this value is set to 1;

[0091] The default value of the regularization coefficient γT is set to 1;

[0092] S24. The maximum tree depth, the number of base learning trees, and the learning rate in the trained XGBoost model are optimized using the sparrow search algorithm. The optimization ranges are [1,50], [1,200], and [0.01,0.3], respectively, to obtain the central air conditioning load prediction model and the outdoor meteorological parameter prediction model.

[0093] The central air conditioning load prediction model is divided into a summer cooling load prediction model and a winter heating load prediction model. The maximum tree depth, number of base learning trees, and learning rate of the summer cooling load prediction model are 5, 143, and 0.0955, respectively. The maximum tree depth, number of base learning trees, and learning rate of the winter heating load prediction model are 6, 121, and 0.124, respectively.

[0094] The outdoor meteorological parameter prediction model is divided into a summer outdoor meteorological parameter prediction model and a winter outdoor meteorological parameter prediction model. For the summer outdoor meteorological parameter prediction model, the maximum tree depth, number of base learning trees, and learning rate for outdoor dry-bulb temperature are 42, 200, and 0.29, respectively; and for outdoor relative humidity, the maximum tree depth, number of base learning trees, and learning rate are 48, 193, and 0.2758, respectively. For the winter outdoor meteorological parameter prediction model, the maximum tree depth, number of base learning trees, and learning rate for outdoor dry-bulb temperature are 29, 124, and 0.2466, respectively; and for outdoor relative humidity, the maximum tree depth, number of base learning trees, and learning rate are 35, 139, and 0.2105, respectively.

[0095] S25. Input the current time t, date, outdoor dry-bulb temperature at time t, outdoor wet-bulb temperature at time t, and building load at time t into the central air conditioning load prediction model to obtain the output data of the central air conditioning load prediction model;

[0096] S26. Input the current time t, the outdoor dry-bulb temperature at time t, and the outdoor wet-bulb temperature at time t into the outdoor meteorological parameter prediction model to obtain the output data of the outdoor meteorological parameter prediction model;

[0097] S27. Perform inverse normalization on the output data of the central air conditioning load forecasting model to obtain the load demand Q at the next time step. t+1 The output data of the outdoor meteorological parameter prediction model is then inversely normalized to obtain the outdoor ambient temperature Tdb at the next moment. t+1 and the outdoor relative humidity (RH) at the next moment t+1 The formula for the inverse normalization process is:

[0098] y = y'·(y max -y min )+y min

[0099] Where y is the output data after denormalization, and y' is the output data before denormalization. max y represents the maximum value of the load before normalization in the training data. min This represents the minimum load value in the training data before normalization.

[0100] In this embodiment, during the training of the XGBoost model in step S23, multiple decision trees are constructed sequentially. During training, the model parameters of each tree are determined by combining the prediction results of the previous tree. Through continuous iteration, accurate prediction results are obtained. Simultaneously, in each iteration, gradient descent is used to assign prediction weights to erroneous samples, ensuring that the error in each iteration is smaller than the previous one. The objective function of the XGBoost model during training is:

[0101]

[0102] Where Obj is the objective function of the XGBoost model, I j Let g be the set of all samples belonging to leaf node j. l h is the first-order partial derivative of the l-th sample in leaf node j. l Let λ be the second partial derivative of the l-th sample in leaf node j, λ be the regularization parameter, n be the total number of training data samples, γT be the regularization coefficient, γ be the adjustment coefficient, and T be the number of leaf nodes in the tree.

[0103] In this embodiment, the data-driven models for the air source heat pump and water pump mentioned in step S3 include an air source hot water pump model and a water pump model.

[0104] In this embodiment, the main objective of optimization is to reduce the total energy consumption of the air conditioning water system. The objective in this embodiment is to determine the optimal host set temperature and water pump set frequency for the next time step t+1 based on the operating parameters at the current time t. Therefore, the objective function in step S4 is the total energy consumption of the system at the next time step, specifically expressed as:

[0105] P total(t+1) =P ashp(t+1) +P pump(t+1)

[0106] =f1(Tdb (t+1) ,RH (t+1) ,Tset (t+1) ,Nset (t+1) Q (t+1) )+f3(Nset (t+1) )

[0107] Among them, P total(t+1)Let be the total energy consumption of the central air conditioning system at time t+1, in kW and P. ashp(t+1) Let be the operating energy consumption of the heat pump at time t+1, in kW and P. pump(t+1) Let be the energy consumption of the water pump at time t+1, in kW and Tdb. (t+1) Let t+1 be the outdoor dry-bulb temperature, in °C and RH. (t+1) Let Tset be the outdoor relative humidity at time t+1, in % (t+1) Set the host temperature at time t+1, in °C, Nset. (t+1) Set the frequency Q of the water pump at time t+1. (t+1) Let f1(·) represent the load demand at time t+1, in kW, and let f1(·) represent the parameter Tdb. (t+1) RH (t+1) ,Tset (t+1) ,Nset (t+1) and Q (t+1) The functional relationship with the host is obtained from the air source hot water pump model, where f3(·) represents the parameter Nset. (t+1) The functional relationship with the water pump is obtained from the water pump model.

[0108] In this embodiment, the constraints in step S4 include host set temperature constraints, water pump set frequency constraints, and indoor thermal comfort constraints; the indoor thermal comfort constraint is the maximum heat exchange capacity of the terminal.

[0109] In this embodiment, the host computer sets a temperature constraint specifically as follows:

[0110]

[0111] Among them, Tset (t+1) Set the host temperature at time t+1.

[0112] In this embodiment, the water pump frequency setting constraint is specifically as follows:

[0113] 30Hz≤Nset (t+1) ≤50Hz

[0114] Among them, Nset (t+1) Set the frequency of the water pump at time t+1.

[0115] In this embodiment, the cold and hot water supply temperatures and the pump frequency affect the heat exchange capacity of the fan coil unit, thus affecting the final state of the treated air. During summer cooling, raising the chilled water setpoint temperature or lowering the pump frequency increases the enthalpy of the supply air state point in the fan coil unit's surface cooler. This reduces the enthalpy difference between the supply and return air states, leading to a decrease in the fan coil unit's heat exchange capacity and consequently, unsatisfactory room thermal comfort. During winter heating, lowering the hot water setpoint temperature or reducing the pump frequency lowers the supply air state point temperature in the fan coil unit's surface cooler. This reduces the temperature difference between the supply and return air states, further decreasing the fan coil unit's heat exchange capacity and resulting in unsatisfactory room thermal comfort. Therefore, when the fan coil unit's heat exchange capacity under different supply water temperatures and pump frequencies is not lower than the indoor load requirement, it can be determined that the indoor thermal comfort requirement is met. This embodiment uses the maximum heat exchange capacity at the terminal as the indoor constraint condition. The formula for calculating the maximum heat exchange capacity at the terminal is:

[0116]

[0117] Among them, y totalc The sum of the summer heat exchange capacities of all fan coil units, i.e., the maximum summer terminal heat exchange capacity, y totalh The sum of the winter heat exchange capacities of all fan coil units, i.e., the maximum winter terminal heat exchange capacity, is expressed in kW; p represents the number of different types of fan coil units, expressed in y. ic For the summer heat exchange capacity of a single fan coil unit, y ih For the winter heat exchange capacity of a single fan coil unit, k ic Let k be the slope of the summer heat transfer capacity equation. ih Let be the slope of the winter heat exchange capacity equation, u be the supply water temperature (°C), and b be the slope. ic b is the intercept in the summer heat exchange capacity. ih k is the intercept in the winter heat exchange capacity. totalc k is the slope of the equation for the maximum heat transfer capacity at the end of summer. totalh Let b be the slope of the equation for the maximum heat transfer capacity at the end of winter. totalc b is the intercept of the maximum heat transfer capacity at the summer terminal. totalh is the intercept of the maximum heat exchange capacity at the terminal in winter, i∈[1,7] and is a positive integer, the number of the 7 types of fan coil units are a, b, c, d, e, f, g respectively, and p is the number of different models of fan coil units.

[0118] In this embodiment, step S5 specifically includes the following steps:

[0119] S51. Set the optimization step size and optimization range for the main unit set temperature and the water pump set frequency respectively. Since frequent adjustments to the main unit outlet water temperature and water pump frequency will affect the service life of the equipment, in this embodiment, the optimization range for the main unit set temperature in summer and winter is set to 7-15℃ and 36-45℃ respectively, and the optimization step size is set to 1℃. The optimization range for the water pump frequency is 30-50Hz, and the optimization range is 5Hz.

[0120] S52. Set the parameter grid, and traverse all grid nodes according to the optimization step size and optimization interval. Calculate a total of 35 groups of nodes to be optimized in summer and 50 groups of nodes to be optimized in winter according to the optimization interval set in step S51.

[0121] S53. Identify all nodes whose heat exchange capacity meets the load demand, and substitute all nodes into the objective function to calculate the node corresponding to the minimum objective function. This node is the host setpoint Tset for the next time step. t+1 And the water pump set frequency Nset for the next moment t+1 The optimal combination is found to complete the optimization process.

[0122] Example 2:

[0123] Based on Example 1, data from four different operating conditions in summer and winter were selected for pre- and post-optimization analysis to compare the system performance before and after optimization. There are five sets of operating data for each operating condition.

[0124] In this embodiment, the comparison results of summer operation data are shown in Table 1. Energy consumption analysis was performed on the fifth set of data from each of the four operating conditions. Figure 2 , Figure 3 , Figure 4 and Figure 5 As shown. Since a water temperature of 10℃ and a water supply temperature of 50Hz are sufficient to meet indoor comfort requirements during summer operation, the analysis is mainly based on these set parameters.

[0125] Table 1 Comparison results before and after summer operating condition optimization

[0126]

[0127]

[0128] like Figure 2 As shown, in operating condition 1, due to the low outdoor temperature and low indoor load demand, raising the chilled water supply temperature to 15℃ can meet the indoor comfort requirements. Therefore, the optimal chilled water supply temperature is calculated to be 15℃, and the pump frequency is set to 30Hz. At this point, the total energy consumption of the central air conditioning system reaches its minimum at 51.4kW. Furthermore, according to... Figure 2It can be seen that under this operating condition, the total energy consumption of the central air conditioning system does not decrease significantly when the chilled water supply temperature is increased, but the energy saving can be achieved by reducing the pump frequency. Among them, the central air conditioning water system under this set of operating conditions can achieve an energy saving rate of 33.3% after optimization.

[0129] like Figure 3 As shown, for operating condition 2, as the outdoor temperature rises, the indoor load demand increases. At this time, the heat exchange capacity of the terminal cannot meet the thermal comfort requirements when the chilled water supply temperature is set to 14℃, 15℃, and 13℃, and the pump frequency is 30Hz, 35Hz, and 40Hz. Increasing the pump frequency can effectively reduce the energy consumption of the main unit, and the reduction in energy consumption of the main unit is higher than the increase in energy consumption of the pump, thereby improving the overall operating energy efficiency of the central air conditioning system. However, when the pump frequency is increased to 40Hz, the increase in energy consumption of the pump is too large, which offsets the reduction in energy consumption of the main unit. Therefore, the overall energy consumption of the central air conditioning system is too high when the pump frequency is set to 40Hz and above. At this time, according to the optimization algorithm, the optimal chilled water setting temperature is 12℃, the pump frequency is 35Hz, and the total operating energy consumption is 118.4kW. After optimization, an energy saving of 16.8% can be achieved in this group of operating conditions.

[0130] like Figure 4 As shown, the load demand in operating condition 3 is higher. At this time, the setting parameters that meet the indoor comfort requirements are only 11℃ and 40Hz, 45Hz, 50Hz and 10℃. Similar to the operating performance of operating condition 2, the phenomenon of the total energy consumption of the central air conditioning system increases when the water pump frequency increases to 40Hz and above also occurs. At this time, the optimal chilled water setting temperature is calculated to be 10℃, the water pump setting frequency is 35Hz, and the total operating energy consumption is 159.3kW. After optimization, an energy saving of 11.9% can be achieved in this group of operating conditions.

[0131] like Figure 5 As shown, the outdoor temperature is the highest in operating condition 4, and the load demand is also the greatest. At this time, only 40Hz, 45Hz and 50Hz at a chilled water set temperature of 10℃ meet the indoor thermal comfort requirements. The calculated minimum energy consumption is 221.8kW, the optimal chilled water set temperature is 10℃ and the water pump set frequency is 40Hz. After optimization, 6.4% energy saving can be achieved in this group of operating conditions.

[0132] In this embodiment, the comparison results of winter operation data are shown in Table 3. In each of the four operating conditions, the fifth set of data was also selected for energy consumption analysis. Figure 6 , Figure 7 , Figure 8 and Figure 9 As shown.

[0133] Table 2 Comparison results before and after summer working condition optimization

[0134]

[0135]

[0136] like Figure 6 As shown, in operating condition 1, due to the high outdoor temperature and low indoor load demand, a hot water temperature setting of 36℃ and a water pump frequency setting of 30Hz are sufficient to meet indoor comfort requirements. Calculations show that the central air conditioning system energy consumption reaches its lowest level of 52.5kW under these parameters. Similarly, under this operating condition, reducing the hot water supply temperature does not significantly reduce the overall energy consumption of the central air conditioning system, while reducing the water pump frequency can achieve significant energy savings. After optimization, the central air conditioning water system in this operating condition can achieve an energy saving rate of 36.1%.

[0137] like Figure 7 As shown, for operating condition 2, as the outdoor temperature decreases, the indoor load demand increases. At this time, the pump frequencies of 30Hz, 35Hz and 40Hz at 36℃ cannot meet the indoor comfort requirements. The optimal hot water setting temperature is calculated to be 37℃. At the same time, since the load is still low, the overall energy consumption of the central air conditioning system will also increase after increasing the pump frequency. Therefore, the optimal pump frequency is calculated to be 30Hz. After optimization, 25.9% energy saving can be achieved in this group of operating conditions.

[0138] like Figure 8 As shown, the load demand is higher in operating condition 3. At this time, the hot water set temperature of 37℃ and the pump frequency of 45Hz, 50Hz and the hot water set temperature of 38℃-45℃ can meet the indoor demand. The operating performance is similar to that of operating conditions 2 and 3 in summer cooling mode. The phenomenon of the total energy consumption of the central air conditioning system increases when the pump frequency increases to 40Hz and above also occurs. At this time, the optimal hot water set temperature is calculated to be 38℃, the pump set frequency is 35Hz, and the total operating energy consumption is 113.2kW. After optimization, an energy saving rate of about 18.5% can be achieved.

[0139] like Figure 9 As shown, the load demand is the highest in operating condition 4, around 460kW. At this time, the indoor comfort requirements can be met with hot water set temperatures of 41℃-45℃ and 40℃ and a pump frequency of 50Hz. The calculated minimum energy consumption is 145kW, the optimal hot water set temperature is 41℃, and the pump set frequency is 35Hz, which can achieve an energy saving of about 14.1%.

[0140] Example 3:

[0141] Based on Example 1, an optimized operation method for the air source heat pump-fan coil central air conditioning system as provided in Example 1 was developed using MATLAB and imported into the TRNSYS module for simulation analysis.

[0142] The actual start time of the central air conditioning system is around 8 a.m. on weekdays. However, the optimization algorithm optimizes the best operating parameters at 9 a.m. based on the operating data at 8 a.m. and calculates them hourly in that order. Therefore, it is not possible to directly optimize the operating data at 8 a.m. and it is necessary to set the initial operating state at that time.

[0143] This embodiment statistically analyzes the load data at 8 AM in summer and winter. It finds that the maximum cooling load at this time is around 500kW in summer and around 400kW in winter. Combining the changes in terminal cooling and heating capacity under different combinations of main unit set water temperature and water pump frequency, it is found that a water supply temperature of 12℃ and a water pump frequency of 50Hz in summer, and a water supply temperature of 38℃ and a water pump frequency of 50Hz in winter are sufficient to meet the requirements. Therefore, the initial setting parameters at 8 AM on weekdays are determined to be: a chilled water set temperature of 12℃ and a water pump set frequency of 50Hz in summer, and a hot water set temperature of 38℃ and a water pump frequency of 50Hz in winter.

[0144] Summer and winter optimization results and comparison of operating energy consumption are as follows: Figure 10 , Figure 11 , Figure 12 and Figure 13 As shown in Table 3, the energy consumption, COP of the main unit, and COP of the water system of the central air conditioning system before and after optimization in summer and winter are as follows. It can be seen that under summer cooling conditions, the total energy consumption of the main unit decreased from 69476 kWh to 65845.67 kWh after optimization, a reduction of only about 5.23%, while the energy consumption of the water pump decreased from 18244.61 kWh to 5868.04 kWh, achieving an energy saving rate as high as 67.84%. The COP of the main unit increased from 3.15 to 3. The energy consumption of the main unit decreased from 2.29 kWh to 29059.28 kWh, a reduction of 17.82%, while the energy consumption of the water pumps decreased from 11009.68 kWh to 3444.32 kWh, with an energy saving rate still above 65%. Under these conditions, the energy consumption of the main unit increased from 2.91 to 3.47, and the energy consumption of the water system increased from 2.15 to 3.09. Analysis of the annual operating data shows that the main unit's annual energy savings were 9506.83 kWh, with an energy saving rate of approximately 9%; the water pumps' annual energy savings were 19941.93 kWh, with an energy saving rate of approximately 68%; and the system's annual energy savings were 29448.76 kWh, with an energy saving rate of approximately 22%. The data above reveals that the energy savings achieved by adjusting the water pump frequency are significantly higher than those achieved by adjusting the hot and cold water supply temperatures. Furthermore, the lower energy savings of the main unit under summer cooling conditions is due to the fact that the initial summer simulation calculations were based on a chilled water setpoint of 10°C, which already resulted in some energy savings compared to the standard requirement of a 7°C supply water temperature. Overall, the central air conditioning system demonstrates significant energy-saving effects in both summer and winter.

[0145] Table 3 Comparison of annual air conditioning system energy consumption results

[0146]

[0147]

[0148] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of this invention.

Claims

1. An optimized operation method for an air source heat pump-fan coil central air conditioning system, characterized in that, The method includes the following steps: S1. Obtain the influencing factors of building load and outdoor meteorological parameters; S2. Construct a central air conditioning load forecasting model and an outdoor meteorological parameter forecasting model based on influencing factors, and predict the load demand at the next moment. The outdoor ambient temperature at the next moment and the outdoor relative humidity at the next moment ; S3. Determine the main influencing factors on the operating performance of air source heat pumps and water pumps, construct data-driven models for air source heat pumps and water pumps, and conduct simulation operation based on the data-driven models to obtain the total energy consumption of the central air conditioning system. S4. Based on the total energy consumption of the central air conditioning system and the load demand at the next moment. The outdoor ambient temperature at the next moment and the outdoor relative humidity at the next moment Based on the prediction results, set the optimization objective function and constraints; The constraints include host set temperature constraints, water pump set frequency constraints, and indoor thermal comfort constraints; the indoor thermal comfort constraint is the maximum heat exchange capacity of the terminal. The formula for calculating the maximum heat exchange capacity at the terminal is: in, This represents the sum of the summer heat exchange capacities of all fan coil units, i.e., the maximum heat exchange capacity at the terminal in summer. This represents the sum of the winter heat exchange capacities of all fan coil units, i.e., the maximum winter heat exchange capacity at the terminal. This refers to the number of fan coil units of different models. This refers to the summer heat exchange capacity of a single fan coil unit. This refers to the winter heat exchange capacity of a single fan coil unit. The slope of the summer heat transfer capacity equation is denoted as . The slope of the winter heat transfer capacity equation is given. For water supply temperature, This is the intercept in the summer heat exchange capacity. This is the intercept in the winter heat exchange capacity. The slope of the equation for the maximum heat transfer capacity at the end of summer is given. The slope of the equation for the maximum heat transfer capacity at the end of winter. This is the intercept of the maximum heat exchange capacity at the end of summer. This is the intercept in the maximum heat exchange capacity at the winter terminal. This represents the total number of fan coil unit types. , and These represent the maximum and minimum number of different models, respectively. S5. Solve the objective function and constraints using a grid search optimization algorithm to obtain the host setpoint temperature at the next time step. And the water pump set frequency at the next moment The optimization process is now complete.

2. The optimized operation method of the air source heat pump-fan coil central air conditioning system according to claim 1, characterized in that, Step S2 specifically includes the following steps: S21. Normalize the influencing factors. The formula for normalization is: in, The value is the normalized value. The values ​​before normalization. The maximum value of the parameters before normalization. This represents the minimum value of the parameters before normalization. S22. Calculate the correlation coefficient between the normalized influencing factors and the cooling and heating loads. The formula for calculating the correlation coefficient is as follows: in, The correlation coefficient is... The number of influencing factors. For the first One influencing factor, This is the average value of the influencing factors. For the first Building load, This represents the average building load. S23. Determine the input and output parameters of the XGBoost model based on the correlation coefficient, set the hyperparameters of the XGBoost model, and train the XGBoost model. S24. The trained XGBoost model is optimized using the Sparrow Search algorithm to obtain the central air conditioning load prediction model and the outdoor meteorological parameter prediction model. S25. Input the current time t, date, outdoor dry-bulb temperature at time t, outdoor wet-bulb temperature at time t, and building load at time t into the central air conditioning load prediction model to obtain the output data of the central air conditioning load prediction model; S26. Input the current time t, the outdoor dry-bulb temperature at time t, and the outdoor wet-bulb temperature at time t into the outdoor meteorological parameter prediction model to obtain the output data of the outdoor meteorological parameter prediction model; S27. Perform inverse normalization on the output data of the central air conditioning load forecasting model to obtain the load demand at the next time step. The outdoor meteorological parameter prediction model's output data is then inversely normalized to obtain the outdoor ambient temperature at the next moment. and the outdoor relative humidity at the next moment The formula for the inverse normalization process is: in, To output the denormalized values ​​of the data, To output the values ​​of the data before denormalization, The maximum value of the load before normalization in the training data. This represents the minimum load value in the training data before normalization.

3. The optimized operation method of the air source heat pump-fan coil central air conditioning system according to claim 2, characterized in that, The objective function of the XGBoost model during training is: in, Let XGBoost be the objective function. Belongs to the leaf node All sample sets, leaf node The Middle First-order partial derivatives of each sample leaf node The Middle The second-order partial derivatives of each sample, For regularization parameters, The total number of training data samples, The regularization coefficient is . For adjustment coefficients, This represents the number of leaf nodes in the tree.

4. The optimized operation method of the air source heat pump-fan coil central air conditioning system according to claim 1, characterized in that, The data-driven models for the air source heat pump and water pump mentioned in step S3 include an air source hot water pump model and a water pump model.

5. The optimized operation method of the air source heat pump-fan coil central air conditioning system according to claim 4, characterized in that, The objective function mentioned in step S4 is the total energy consumption of the system at the next time step, and its specific formula is as follows: in, for Total energy consumption of the central air conditioning system at any given time. for The energy consumption of the heat pump at any given time. for Energy consumption of the water pump at any given time. for outdoor dry-bulb temperature at any given time for outdoor relative humidity at any given time for The host's set temperature at any given time. for The water pump's set frequency at any given time, for The load demand at any time, Indicates parameters , , , and The functional relationship with the host is obtained from the air source hot water pump model. Indicates parameters The functional relationship with the water pump is obtained from the water pump model.

6. The optimized operation method of the air source heat pump-fan coil central air conditioning system according to claim 1, characterized in that, The specific temperature constraint set for the host is as follows: in, for The host's set temperature at any given time.

7. The optimized operation method of the air source heat pump-fan coil central air conditioning system according to claim 1, characterized in that, The specific frequency constraint for the water pump is as follows: in, for The water pump's set frequency at any given time.

8. The optimized operation method of the air source heat pump-fan coil central air conditioning system according to claim 1, characterized in that, Step S5 specifically includes the following steps: S51. Set the optimization step size and optimization range for the host set temperature and the water pump set frequency respectively; S52. Set the parameter grid and traverse all grid nodes according to the optimization step size and optimization interval; S53. Identify all nodes whose heat exchange capacity meets the load demand, and substitute all nodes into the objective function to calculate the node corresponding to the minimum objective function. This node is the setpoint temperature for the main unit at the next time step. And the water pump set frequency at the next moment The optimal combination is found to complete the optimization process.

Images