Real-time Optimization Control Method, Storage Medium and Device for Central Air Conditioning System Based on Surrogate Model

Abstract

The present invention relates to a real-time optimization control method, storage medium and device for a central air-conditioning system based on a surrogate model. The present invention uses a machine learning algorithm to construct a surrogate model for predicting the results of an optimal control strategy. During actual application, by inputting real-time operating conditions into the surrogate model, the results of the optimal control strategy under these operating conditions are predicted. Compared with traditional model-based optimization control methods, the present invention reduces the computational complexity and significantly improves the computational efficiency of the optimal control strategy, meeting the real-time optimization control requirements of the central air-conditioning system. When training the model, based on the equipment composition and topological structure of the target central air-conditioning system, as well as the optimization objectives and constraints, a mathematical optimization model of the central air-conditioning system is established, so as to generate a large number of optimal control strategies corresponding to the operating conditions by randomly generated operating conditions, forming a training data set for training the surrogate model, so that the trained surrogate model can accurately predict the optimal control strategy.

Description

Technical Field

[0001] The present invention relates to a real-time optimization control method, storage medium and device for a central air-conditioning system based on a surrogate model, and is applicable to the field of operation control of integrated energy systems. Background Art

[0002] The central air-conditioning system is the most important energy-consuming system during the operation stage of public buildings, and its energy consumption accounts for more than 50% of the total energy consumption during the building operation stage. Conducting research on the optimization control method of the central air-conditioning system is of great significance for meeting the cooling load demand of building users and reducing the operation energy consumption of the system, and is a powerful means to achieve low-carbon energy conservation in public buildings.

[0003] The central air-conditioning system has characteristics such as high nonlinearity, complex topological structure, and strong thermal coupling. At present, the method of solving the system optimization model is generally used in this field to calculate the optimal control strategy of the system.

[0004] For example, in the Chinese invention patent with the patent publication number CN115577828A, a group control method for an air-conditioning chiller plant system based on data-driven modeling and optimization uses a differential evolution algorithm to solve the air-conditioning chiller plant model to achieve the optimization of the optimal cooling capacity of the chiller, and constructs an optimization database for operation and maintenance personnel to query; in the Chinese invention patent with the patent publication number CN115829093A, an optimization method for the set value of the chilled water outlet temperature based on data-driven uses a fish school algorithm to solve the indoor temperature and electric energy prediction models to achieve the optimization of the optimal chilled water outlet temperature of the chiller.

[0005] However, although the above methods have high optimization accuracy, their computational complexity and optimization time cost are usually high, and it is often difficult to meet the real-time optimization control requirements of the system in practical applications. How to reduce the computational complexity and improve the optimization efficiency while ensuring the optimization accuracy of the control strategy to achieve the efficient real-time control of the central air-conditioning system remains to be further studied. Summary of the Invention

[0006] The technical problem to be solved by the present invention is: in view of the above problems, to provide a real-time optimization control method, storage medium and device for a central air-conditioning system based on a surrogate model.

[0007] The technical solution adopted by the present invention is: a real-time optimization control method for a central air-conditioning system based on a surrogate model, including:

[0008] Obtain the real-time operating conditions of the target central air-conditioning system, and input the operating conditions into the trained surrogate model to predict the result of the optimal control strategy under the operating conditions;

[0009] The training of the surrogate model includes:

[0010] Determine the set of operating conditions of the target central air-conditioning system and the set of control parameters to be optimized for each device in the system; the set of operating conditions includes the numerical ranges of various types of operating condition parameters, and the set of control parameters to be optimized includes the numerical ranges obtained from various types of device operating parameters for each device in the system;

[0011] Based on the device composition and topological structure of the target central air-conditioning system, as well as the optimization objectives and constraints, establish a mathematical optimization model for this central air-conditioning system; the optimization objectives include minimizing the total operating energy consumption of the system, and the constraints include energy conservation constraints;

[0012] Randomly generate multiple operating condition samples based on the numerical ranges of the parameters in the set of operating conditions of the target central air-conditioning system, and solve the mathematical optimization model under each operating condition sample to obtain the optimal control strategies corresponding to the parameters in the set of control parameters to be optimized under each operating condition sample. Construct a training data set based on the operating condition samples and the corresponding optimal control strategies;

[0013] Train a surrogate model based on the training data set, and this surrogate model is constructed using machine learning algorithms.

[0014] The types of operating condition parameters in the set of operating conditions include terminal cooling load, outdoor temperature, and outdoor humidity; the types of parameters in the set of control parameters to be optimized include the setpoint parameters of each device in the target central air-conditioning system.

[0015] The setpoint parameters of each device in the target central air-conditioning system include the chilled water outlet temperature of the chiller, the operating frequency of the chilled water pump, the operating frequency of the cooling water pump, and the operating frequency of the cooling tower fan.

[0016] The constraints include:

[0017] The energy conservation constraint between the cooling water side and the chilled water side, that is, the total heat rejection of the cooling tower should be equal to the sum of the cooling load and the operating power of each chiller;

[0018] The energy conservation constraint between the cooling capacity and the cooling load, that is, the cooling load should be equal to the sum of the cooling capacities of each chiller;

[0019] The energy conservation constraint of the chilled water loop; the energy conservation constraint of the cooling water loop.

[0020] The optimization algorithms for solving the mathematical optimization model include genetic algorithms, particle swarm algorithms, and grey wolf algorithms and their variants.

[0021] The machine learning algorithms used to construct the surrogate model include extreme gradient boosting, recurrent neural networks, artificial neural networks, and support vector machine regression.

[0022] A storage medium stores a computer program executable by a processor. When the computer program is executed, the steps of the real-time optimization control method for the central air-conditioning system are implemented.

[0023] A real-time optimization control device for a central air-conditioning system has a memory and a processor. The memory stores a computer program executable by the processor. When the computer program is executed, the steps of the real-time optimization control method for the central air-conditioning system are implemented.

[0024] The beneficial effects of the present invention are as follows: The present invention constructs a surrogate model for predicting the results of the optimal control strategy by using a machine learning algorithm. In actual application, by inputting the real-time operating conditions into the surrogate model, the results of the optimal control strategy under the operating conditions are predicted. Compared with the traditional model-based optimization control method, the present invention reduces the computational complexity, significantly improves the computational efficiency of the optimal control strategy, and meets the real-time optimization control requirements of the central air-conditioning system.

[0025] When training the model, based on the equipment composition, topological structure, optimization objectives, and constraint conditions of the target central air-conditioning system, a mathematical optimization model of the central air-conditioning system is established, so as to generate a large number of optimal control strategies corresponding to the operating conditions by randomly generated operating conditions, constituting the training data set for training the surrogate model, overcoming the technical problem that the amount of data of the target central air-conditioning system is small and cannot be effectively trained, so that the trained surrogate model can accurately predict the optimal control strategy.

[0026] In the present invention, model training can be carried out on a dedicated model training device. Only the trained surrogate model needs to be deployed on the specific controller. Compared with the traditional method of deploying a mathematical optimization model in the controller, the computing power requirement for the controller is low, the controller overhead is small, and the operation efficiency is extremely high, which can significantly improve the real-time control performance of the system.

[0027] The present invention selects the terminal cooling load, outdoor temperature, and outdoor humidity as inputs for model prediction. Among them, the terminal cooling load mainly affects the operation on the chilled water side and has an important impact on the outlet water temperature set point of the chiller and the frequency of the chilled water pump; the outdoor temperature and outdoor humidity mainly affect the operation on the cooling water side and have an important impact on the heat rejection performance of the cooling tower, and further affect the frequency of the cooling water pump. Therefore, selecting the terminal cooling load, outdoor temperature, and outdoor humidity as inputs helps the surrogate model to accurately predict the optimal control strategy. Description of the Drawings

[0028] Figure 1 It is a schematic structural diagram of the target central air-conditioning system in the embodiment of the present invention.

[0029] Figure 2 It is a training flow chart of the surrogate model in the embodiment of the present invention.

[0030] Figure 3 This is a comparison of the optimization accuracy of the optimal control strategies under different methods in the embodiments of the present invention.

[0031] Figure 4 This is a comparison of the calculation speeds of the optimal control strategies under different methods in the embodiments of the present invention. Detailed implementation manners

[0032] To better understand the technical solutions of this application, the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0033] It should be clear that the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in this application without creative efforts belong to the scope of protection of this application.

[0034] The terms used in the embodiments of this application are only for the purpose of describing specific embodiments, and are not intended to limit this application. The singular forms of "a", "the" and "said" used in the embodiments of this application and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise.

[0035] Embodiment 1: Figure 1 This is a schematic structural diagram of the target central air-conditioning system in this embodiment. The central air-conditioning system includes three chillers, three chilled water pumps, three cooling water pumps and three cooling towers. On the chilled water side, the low-temperature chilled water generated by the chiller is transmitted to the end for cooling, and the chilled water return water returns to the chiller through the chilled water pump for circulation; on the cooling water side, the high-temperature cooling water return water is transmitted to the cooling tower through the cooling water pump, and after being cooled by the cooling tower, it returns to the chiller.

[0036] This embodiment is a real-time optimization control method for a central air-conditioning system based on a surrogate model, which specifically includes the following steps:

[0037] A. Obtain the real-time operating conditions of the target central air-conditioning system, and the operating conditions include the specific values of parameters such as the end cooling load , outdoor temperature , outdoor humidity , etc.

[0038] B. Input the operating conditions obtained in step A into the trained surrogate model, and predict the results of the optimal control strategy under the operating conditions through the surrogate model.

[0039] In this example, the optimal control strategy includes the chilled water outlet temperature of each chiller (where the subscripts 1-3 correspond to the three chillers in this example respectively), the operating frequency of each chilled water pump The operating frequency of each cooling water pump The operating frequency of the fans of each cooling tower The specific values of parameters such as

[0040] In this embodiment, the surrogate model is used to predict the optimal control strategy based on the operating conditions. As Figure 2 shown, the training of the surrogate model includes:

[0041] S100. Determine the set of operating conditions of the target central air-conditioning system and the set of control parameters to be optimized for each device in the system.

[0042] In this embodiment, the set of operating conditions includes the terminal cooling load The outdoor temperature And the outdoor humidity The numerical ranges of operating condition parameters such as

[0043] The set of operating conditions Is shown as follows:

[0044] (1)

[0045] In this example, the set of control parameters to be optimized includes the chilled water outlet temperature of each chiller The operating frequency of each chilled water pump The operating frequency of each cooling water pump The operating frequency of the fans of each cooling tower The numerical ranges of equipment operating parameters such as

[0046] The set of control parameters to be optimized Is shown as follows:

[0047] (2)

[0048] S200. Based on the equipment composition and topological structure of the target central air-conditioning system, as well as the optimization objectives and constraints, establish a mathematical optimization model for the central air-conditioning system.

[0049] In this embodiment, the optimization objective function is to minimize the total operating energy consumption of the system; the constraints include the energy conservation constraint between the cooling water side and the chilled water side, the energy conservation constraint between the cooling capacity and the cooling load, the energy conservation constraint of the chilled water loop and the energy conservation constraint of the cooling water loop, as well as the numerical range constraint of the operating parameters of each device in the system.

[0050] In this example, the mathematical optimization model of the target central air-conditioning system includes:

[0051] (3)

[0052] (4)

[0053] (5)

[0054] (6)

[0055] (7)

[0056] Wherein, , , and respectively represent the numbers of the chiller, chilled water pump, cooling water pump and cooling tower, all of which are 3 in this embodiment. , , and respectively represent the operating energy consumptions of the th chiller, chilled water pump, cooling water pump and cooling tower. is the terminal cooling load. is the heat rejection of the th cooling tower. is the cooling capacity of the th chiller. is the constant-pressure specific heat capacity of water. and respectively represent the chilled water and cooling water flow rates of the rd loop. and respectively represent the supply water temperature and return water temperature of the chilled water of the th chiller. and respectively represent the outlet water temperature and inlet water temperature of the th cooling tower.

[0057] In this optimization model, Equation (3) represents the optimization objective function, that is, to minimize the total operating energy consumption of the system; Equation (4) represents the energy conservation constraint between the cooling water side and the chilled water side, that is, the total heat rejection of the cooling tower should be equal to the sum of the cooling load and the operating powers of each chiller; Equation (5) represents the energy conservation constraint between the cooling capacity and the cooling load, that is, the cooling load should be equal to the sum of the cooling capacities of each chiller; Equation (6) represents the energy conservation constraint of the chilled water loop; Equation (7) represents the energy conservation constraint of the cooling water loop.

[0058] In this embodiment, the energy consumption model of the chiller is shown as the following equation:

[0059] (8)

[0060] Wherein, is the performance coefficient of the chiller, is the rated cooling capacity, is the rated performance coefficient. The energy consumption models of the chilled water pump and the cooling water pump are shown in the following formula:

[0061] (9)

[0062] where, is the performance coefficient of the pump, is the pump efficiency, is the rated frequency of the pump. The energy consumption model of the cooling tower is shown in the following formula:

[0063] (10)

[0064] where, is the performance coefficient of the cooling tower, is the rated power of the cooling tower.

[0065] S300. Randomly generate multiple operating condition samples based on the numerical ranges of the parameters in the set of operating conditions of the target central air-conditioning system, and solve the mathematical optimization model under each operating condition sample to obtain the optimal control strategy corresponding to the set of control parameters to be optimized under each operating condition sample. Construct a training data set based on the operating condition samples and the corresponding optimal control strategies.

[0066] S310. In this embodiment, the terminal cooling load ranges from [500, 4000] kW, the outdoor temperature ranges from [15, 40] °C, and the outdoor humidity ranges from [30, 90] %. Randomly sample from the above operating condition ranges to generate 8000 groups of operating condition samples , as shown in the following formula:

[0067] (11)

[0068] where, represents the th operating condition sample; is the terminal cooling load in the th operating condition sample; is the outdoor temperature in the th operating condition sample; is the outdoor humidity in the th operating condition sample.

[0069] S320. For each operating condition sample , substitute it into the mathematical optimization model and solve to obtain the optimal control strategy under this condition , as shown in the following formula:

[0070] (12)

[0071] where, represents the optimal control strategy corresponding to the th operating condition sample; In the above formula, the superscript i represents the th operating condition sample.

[0072] In this embodiment, the optimization algorithms for solving the mathematical optimization model of the target central air-conditioning system include but are not limited to heuristic algorithms such as genetic algorithms, particle swarm algorithms, and grey wolf algorithms and their variants.

[0073] In some specific embodiments, the particle swarm algorithm is used to solve the mathematical optimization model, and the specific steps are as follows:

[0074] S321. Set the particle swarm size , the inertia weight , the acceleration factors , , the maximum number of iterations and the iteration counter ;

[0075] S322. Substitute the terminal cooling load , the outdoor temperature and the outdoor humidity into the optimization model to complete the model initialization;

[0076] S323. Randomly initialize the position vector and the velocity vector of each particle in the search space, as shown in the following formulas respectively:

[0077] (13)

[0078] (14)

[0079] In this embodiment, the search space of is [6, 10] °C, , and the search spaces of and are [20, 50] Hz;

[0080] S324. Each particle calculates the corresponding fitness value by substituting the control parameters represented by its own position variable into the optimization model. The calculation formula of the fitness is as shown in the following formula:

[0081] (15)

[0082] Among them, is the objective function value of the optimized model under this set of control parameters, is the violation degree value of each constraint of the optimized model under this set of control parameters, is the penalty coefficient. By adding a penalty term for constraint violation on the basis of the original objective function value, it is ensured that the particles can satisfy the model constraints during the optimization;

[0083] S325. Update the velocity vector of each particle , as shown in the following formula:

[0084] (16)

[0085] Among them, and are random numbers in the interval [0, 1], is the particle The position vector corresponding to the minimum objective function value during the iteration, is the position vector corresponding to the minimum objective function value of all particles during the iteration;

[0086] S326. Update the position vector of each particle , as shown in the following formula:

[0087] (17)

[0088] S327. If the number of iterations reaches or remains unchanged within a certain number of consecutive iterations , the algorithm stops, and the control parameters characterized by are output to obtain the optimal . Otherwise, return to step S323.

[0089] S330. Using each operating condition sample as the input and the corresponding optimal control strategy as the output, construct the training data set of the surrogate model.

[0090] S400. Based on the training data set, train the surrogate model, and this surrogate model is constructed using a machine learning algorithm.

[0091] In this example, the machine learning algorithms for constructing the surrogate model include but are not limited to extreme gradient boosting, recurrent neural network, artificial neural network, support vector machine regression, etc.

[0092] In some specific embodiments, an artificial neural network algorithm is used to construct a surrogate model for predicting the optimal control strategy. The specific steps include:

[0093] S410. Set the network structure parameters as shown in the following table. Randomly initialize the weight and bias parameters of the network according to the structure parameters;

[0094] Table 1 Artificial Neural Network Structure Parameters

[0095]

[0096] S420. Divide the data set into a training set and a validation set in a ratio of 7:3, which are used to train the surrogate model and verify the model accuracy respectively;

[0097] S430. Use min-max normalization to uniformly transform the data into the range of [0, 1] as shown in the following formula:

[0098] (18)

[0099] where, is the original sample data, including input data and output data, and are the maximum and minimum values of this type of data respectively, is the data after the normalization operation.

[0100] S440. Perform batch processing update on the weights and biases of the neural network as shown in the following formula:

[0101] (19)

[0102] where, is the learning rate, which is taken as 0.01 in this embodiment, is the batch size, which is taken as 256 in this embodiment, is the gradient function with respect to the loss function. In this embodiment, the loss function is the root mean square error as shown in the following formula:

[0103] (20)

[0104] where, is the number of samples, and are respectively the actual value and predicted value of the

[0105] In this embodiment, for the real-time operating conditions of the system , a surrogate model is used to predict the optimal control strategy result under this condition as shown in the following formula:

[0106] (21)

[0107] Among them, and are the weights and biases of the first hidden layer, and are the weights and biases of the second hidden layer, is the activation function.

[0108] The computational time results of the optimal control strategies under different methods are compared as Figure 3 shown. The traditional model-based optimization method needs to solve the system optimization model every time the optimal control strategy is calculated. Due to the large number of parameters to be optimized in the central air-conditioning system optimization model and the high degree of non-linearity between parameters, the time required for model solving is relatively high, and the average computational time reaches 176.6 s. In contrast, this method does not need to solve the system optimization model and can directly predict the optimal control strategy through the surrogate model. Its average computational time is only 0.1 s, which is 99.9% lower than the traditional method.

[0109] The operating energy consumption results of the central air-conditioning system under different methods are compared as Figure 4 shown. It can be seen from the figure that the prediction result of the optimal control strategy of this method comes from model optimization. Compared with the traditional control rule-based method (the chilled water outlet temperature of the chiller is always 7°C, and the frequencies of the water pump and cooling tower fan are always at the power frequency), the obtained optimal control strategy has a better optimization effect, and its daily operating energy consumption is reduced from 11149.8 kWh to 9695.2 kWh, and the energy saving rate reaches 13.0%.

[0110] In summary, the method proposed by the present invention realizes the rapid and accurate calculation of the system optimal control strategy by constructing a surrogate model of the control strategy, can realize the real-time optimal control of the central air-conditioning system on the premise of ensuring the optimization accuracy of the control strategy, and can be effectively applied to the efficient and real-time operation control scenario of the central air-conditioning system.

[0111] Embodiment 2: This embodiment is a storage medium on which a computer program executable by a processor is stored. When the computer program is executed, the steps of the real-time optimal control method for the central air-conditioning system in Embodiment 1 are realized.

[0112] Embodiment 3: A real-time optimal control device for a central air-conditioning system has a memory and a processor. A computer program executable by the processor is stored on the memory. When the computer program is executed, the steps of the real-time optimal control method for the central air-conditioning system in Embodiment 1 are realized.

Claims

1. A real-time optimization control method for a central air-conditioning system based on an agent model, characterized in that: include: Obtain the real-time operating conditions of the target central air-conditioning system, input the operating conditions into the trained proxy model, and predict the optimal control strategy results under the operating conditions; The training of the proxy model includes: Determine an operating condition set of a target central air-conditioning system and a set of control parameters to be optimized for each device in the system; the operating condition set includes a numerical range of multiple types of operating condition parameters, and the set of control parameters to be optimized includes a numerical range of multiple types of device operating parameters for each device in the system; Based on the equipment composition and topological structure of the target central air-conditioning system, as well as the optimization objectives and constraints, a mathematical optimization model of the central air-conditioning system is established; the optimization objectives include minimizing the total operating energy consumption of the system, and the constraints include energy conservation constraints; Based on the numerical range of each parameter in the target central air-conditioning system operating condition set, a plurality of operating condition samples are randomly generated, and the mathematical optimization model under each operating condition sample is solved to obtain the optimal control strategy corresponding to the parameters in the control parameter set to be optimized under each operating condition sample, and a training data set is constructed based on the operating condition samples and the corresponding optimal control strategy; Based on the training data set, train a proxy model, which is built using a machine learning algorithm; The operating condition parameter categories in the operating condition set include terminal cooling load, outdoor temperature and outdoor humidity; the parameter categories in the control parameter set to be optimized include set value parameters of each device of the target central air-conditioning system; The machine learning algorithms used to construct the proxy model include extreme gradient boosting, recursive neural network, artificial neural network, and support vector machine regression.

2. The real-time optimization control method for a central air-conditioning system based on an agent model according to claim 1 is characterized in that: The set value parameters of each device of the target central air-conditioning system include the chilled water outlet temperature of the chiller, the operating frequency of the chilled water pump, the operating frequency of the cooling water pump, and the operating frequency of the cooling tower fan.

3. The real-time optimization control method for a central air-conditioning system based on an agent model according to claim 1 is characterized in that: The constraints include: The energy conservation constraint between the cooling water side and the chilled water side, that is, the total heat rejection of the cooling tower should be equal to the sum of the cooling load and the operating power of each chiller; The energy conservation constraint between cooling capacity and cooling load, that is, the cooling load should be equal to the sum of the cooling capacity of each chiller; Energy conservation constraints for the chilled water loop; Energy conservation constraints for the cooling water loop.

4. The real-time optimization control method for a central air-conditioning system based on an agent model according to claim 1, characterized in that: The optimization algorithms for solving the mathematical optimization model include genetic algorithm, particle swarm algorithm and grey wolf algorithm and their variants.

5. A storage medium having stored thereon a computer program executable by a processor, characterized in that: When the computer program is executed, the steps of the real-time optimization control method for the central air-conditioning system according to any one of claims 1 to 4 are implemented.

6. A central air conditioning system real-time optimization control device, comprising a memory and a processor, wherein the memory stores a computer program executable by the processor, characterized in that: When the computer program is executed, the steps of the real-time optimization control method for the central air-conditioning system according to any one of claims 1 to 4 are implemented.

Images