Complex air conditioning system annual energy consumption estimation method based on operation condition clustering and parameter optimization

Abstract

The invention provides a complex air conditioning system annual energy consumption estimation method based on operation condition clustering and parameter optimization, which belongs to the field of energy consumption calculation, and comprises the following steps: S1, establishing an air conditioning system simulation model; s2, annual hourly working condition data of the complex air conditioning system are obtained, and time sequence preprocessing is carried out; s3, performing clustering analysis by adopting an iterative clustering algorithm to obtain clustering working conditions, outdoor temperature and humidity and indoor heat and humidity load data corresponding to the clustering working conditions and the number of hours of a whole year; s4, performing layered optimization on the operation parameters of the air conditioning system simulation model to obtain an optimal operation parameter combination under each clustering working condition; and S5, inputting the optimal operation parameter combination into the air conditioning system simulation model, and calculating to obtain an annual energy consumption estimated value of the complex air conditioning system. According to the method, the air conditioner system simulation model is constructed, clustering analysis is carried out on the preprocessed data by adopting the iterative clustering algorithm, and layered parameter optimization is combined, so that the accuracy of air conditioner energy consumption analysis is improved.

Description

Technical Field

[0001] This invention relates to the field of energy consumption calculation technology, and in particular to a method for estimating the annual energy consumption of complex air conditioning systems based on operating condition clustering and parameter optimization. Background Technology

[0002] With economic growth and technological advancements, high-end industries such as electronics, pharmaceuticals, biotechnology, and medicine are flourishing. These industries demand high-precision temperature and humidity control, leading to more complex air conditioning systems and increased design challenges. Under the national goal of "peak carbon emissions and carbon neutrality," achieving low-energy consumption design for complex air conditioning systems geared towards high-end industries is crucial. Therefore, research on energy consumption estimation technology for complex air conditioning systems, as a direct basis for low-energy consumption design, is of great significance.

[0003] There are two main methods for estimating the annual energy consumption of conventional air conditioning systems: one is to make a rough estimate based on the air conditioning load and system energy efficiency, and the other is to use energy consumption simulation software to make a refined estimate of the annual energy consumption of the air conditioning system. However, for complex air conditioning systems, existing energy consumption estimation methods have the following technical shortcomings: First, due to the diverse influencing factors, the performance curve of complex air conditioning systems is difficult to determine, and the rough estimation method based on load and energy efficiency is difficult to obtain accurate results; second, complex air conditioning systems require independent control of temperature and humidity and have many adjustable parameters, making traditional temperature feedback control methods difficult to apply, and the system control logic is complex or unclear, resulting in greater difficulty in energy consumption estimation based on simulation calculations; in addition, different operating parameter configurations of complex air conditioning systems under the same operating conditions have a significant impact on system energy consumption, and there is a lack of effective operating parameter optimization methods to obtain the lowest energy consumption operating state.

[0004] Chinese invention patent CN115468278A discloses a two-layer control and management framework for cooling systems that can reduce energy consumption. This patent proposes a three-layer architecture comprising a controlled entity domain, a connection model domain, and a control algorithm domain. The technical solution employs a model-driven energy consumption optimization genetic algorithm and a model-independent energy consumption optimization reinforcement learning algorithm to reduce cooling system energy consumption, and uses the CLF index to evaluate the energy consumption optimization effect. However, it lacks an effective analysis method for complex air conditioning systems, and the optimal energy consumption is difficult to obtain. Summary of the Invention

[0005] In view of this, the present invention proposes a method for estimating the annual energy consumption of complex air conditioning systems based on operating condition clustering and parameter optimization. This method solves the problems in the prior art that it is impossible to differentiate complex air conditioning systems and that it is difficult to obtain the optimal energy consumption. It uses an iterative clustering algorithm to perform cluster analysis on the preprocessed operating condition data and combines it with hierarchical parameter optimization to improve the computational efficiency and the accuracy of air conditioning energy consumption analysis.

[0006] The technical solution of this invention is implemented as follows: This invention provides a method for estimating the annual energy consumption of complex air conditioning systems based on operating condition clustering and parameter optimization, comprising the following steps: S1. Use the MATLAB-Simulink platform to establish a simulation model of a complex air conditioning system. S2. Obtain the hourly operating data of the complex air conditioning system throughout the year, and perform time-series preprocessing on the hourly operating data throughout the year to obtain the preprocessed operating data, which includes outdoor temperature, outdoor relative humidity, indoor sensible heat load and indoor moisture dissipation data. S3. The preprocessed working condition data is clustered using an iterative clustering algorithm to obtain clustered working conditions, corresponding outdoor temperature and humidity and indoor heat and humidity load data, and the number of hours throughout the year. S4. Based on the clustered operating conditions, the corresponding outdoor temperature and humidity and indoor heat and humidity load data of the clustered operating conditions, and the number of hours throughout the year, the operating parameters of the air conditioning system simulation model are optimized in layers to obtain the optimal combination of operating parameters under each clustered operating condition. S5. Input the optimal combination of operating parameters under each cluster condition into the air conditioning system simulation model, and obtain the optimized energy consumption value under each cluster condition through simulation calculation. Based on the optimized energy consumption value and corresponding number of hours under each cluster condition, calculate the estimated annual energy consumption value of the complex air conditioning system.

[0007] Based on the above technical solutions, preferably, the simulation model of the complex air conditioning system includes a heat pump system, a dehumidifier impeller, a fan module, an electric heater module, and an air mixing and indoor air parameter calculation module. The heat pump system is modeled using the System-Level Refrigeration Cycle module in Simulink, the fan module is modeled using the Flow Rate Source module in Simulink, the electric heater module is modeled using the Pipe, Heat Flow Rate Source, and Thermal Reference modules in Simulink, the air mixing and indoor air parameter calculation module is modeled using the Constant Volume Chamber module in Simulink, and the dehumidifier impeller is modeled using the potential function method, and the potential function efficiency is corrected based on the structural performance data of the dehumidifier impeller.

[0008] Based on the above technical solutions, preferably, step S3 specifically includes: S31. Extract the operating characteristic parameters of the complex air conditioning system from the preprocessed operating condition data, and calculate the Pearson correlation coefficient between each characteristic parameter and the air conditioning energy consumption. Divide the absolute value of the Pearson correlation coefficient of each characteristic parameter by the sum of the absolute values ​​of the Pearson correlation coefficients of all characteristic parameters to obtain the normalized basic weight coefficient of the characteristic parameter. The characteristic parameters include outdoor temperature, outdoor relative humidity, indoor sensible heat load and indoor moisture dissipation. S32. Adjust the basic weight coefficients based on seasonal factors to obtain seasonal weights; S33. The Calinski-Harabasz index is used to determine the optimal number of clusters by traversing different numbers of clusters. S34. Based on the optimal number of clusters and seasonal weights, an iterative clustering method is used to cluster the preprocessed operating data, generating clustered operating conditions, corresponding outdoor temperature and humidity and indoor heat and humidity load data, and the number of hours throughout the year.

[0009] Based on the above technical solutions, preferably, step S32 specifically includes: The season type is determined based on the outdoor temperature, with the following logic: ; in, Represents a seasonal type variable. Indicates the heating season. Indicates the transitional season. Indicates the cooling season. Indicates the outdoor temperature; The seasonal adjustment coefficients for each characteristic parameter are calculated using the following formula: ; in, This represents the seasonal adjustment coefficient for the i-th characteristic parameter. Indicates the feature parameter index; Based on the seasonal type and seasonal adjustment coefficient, the seasonal weight is determined using the following formula: ; in, This represents the seasonal weighting coefficient adjusted for the i-th feature parameter. This represents the basic weight coefficient of the i-th feature parameter.

[0010] Based on the above technical solutions, the preferred formula for calculating the Calinski-Harabasz index is as follows: ; in, This represents the Calinski-Harabasz index value. Represents the sum of squares between clusters. Represents the sum of squares within a cluster. This represents the total number of clustered operating conditions. This indicates the total number of preprocessed operating condition data points. This represents the adjustment coefficient. This represents the clustering condition index. This represents the average energy consumption value for the j-th cluster condition. This represents the overall average energy consumption value across all data points.

[0011] Based on the above technical solutions, preferably, step S34 specifically includes: Step a: Set the control parameters for iterative clustering, including the maximum number of iterations, convergence accuracy threshold, and outlier removal ratio; initialize the cluster centers using the K-means++ algorithm. Step b: Perform outlier detection on the preprocessed working condition data, calculate the weighted distance from each working condition data point to its respective cluster center, determine the outlier threshold based on the statistical characteristics of the distance distribution, mark the working condition data points that exceed the outlier threshold as outliers and remove them from the current dataset to form a new dataset; Step c: Update the cluster center locations based on the new dataset, and assign each data point to the nearest cluster center to form the dataset after cluster assignment; Step d, repeat steps b and c until the maximum number of iterations is reached, to obtain the iteratively optimized cluster centers and the final data allocation. Based on the iteratively optimized cluster centers and the final data allocation, generate clustering conditions, corresponding outdoor temperature and humidity and indoor heat and humidity load data, and the number of hours throughout the year.

[0012] Based on the above technical solutions, preferably, step S4 specifically includes: The adjustable operating parameters of the air conditioning system are divided into primary parameters and secondary parameters according to their impact on system energy consumption. The primary parameters include heat pump frequency, air supply volume and primary and secondary return air ratio, while the secondary parameters include electric heating power and fresh air volume. A mixed-integer nonlinear substitution optimization algorithm is used to perform the first-level coarse optimization of the main parameters, with the objective function being to minimize system energy consumption, to obtain the first-level optimal parameter combination, where the objective function is: ; The constraints are: ; in, This represents the minimum energy consumption target for optimizing the main parameters. This represents the energy consumption function with key parameters as variables and operating conditions as parameters. Indicates the heat pump frequency. Indicates the air supply volume. This indicates the ratio of primary to secondary return air. Indicates the outdoor dry-bulb temperature. Indicates outdoor relative humidity. Indicates the indoor sensible heat load. Indicates indoor moisture dissipation. Represents the set of positive integers; Based on the first-level optimal parameter combination, a second-level fine-tuning is performed on the secondary parameters to obtain the optimal operating parameter combination for each clustering condition. The objective function of the second-level fine-tuning is: The constraints are: in, This represents the minimum energy consumption target for optimizing secondary parameters. This represents the energy consumption function with secondary parameters as decision variables, the first-level optimization result, and operating conditions as given parameters. Indicates the electric heating power. Indicates the fresh air volume. This represents the optimal heat pump frequency obtained from the first layer of optimization. This represents the optimal air supply volume obtained from the first layer of optimization. This represents the optimal primary and secondary return air ratio obtained from the first layer of optimization. This indicates the minimum fresh air volume. This indicates the moisture content of the return air. This indicates the return air temperature.

[0013] Based on the above technical solutions, the preferred logic for calculating the estimated annual energy consumption of a complex air conditioning system is as follows: ; in, This represents the estimated annual energy consumption of a complex air conditioning system. This represents the clustering condition index. This represents the total number of clustered operating conditions. This represents the optimized energy consumption value for the j-th clustering condition. This represents the total number of hours throughout the year for the j-th cluster condition.

[0014] Based on the above technical solutions, preferably, the calculation formula for establishing the dehumidification rotor model using the potential function method is as follows: ; ; in, This represents the first potential function of the dehumidifier rotor. This indicates the temperature of the air entering the dehumidifier rotor. This indicates the moisture content of the air entering the dehumidification rotor. This represents the second potential function of the dehumidification rotor.

[0015] Furthermore, step S2 specifically includes: Calculate the statistical characteristics of the data distribution for hourly operating data throughout the year, including the first quartile, the third quartile, and the interquartile range; Based on the statistical principle of box plots, the criteria for identifying outlier data are determined, and data points that exceed the normal data distribution range are marked as outliers. The identified abnormal data is marked and temporarily removed, and an abnormal data index list is established, which includes the time location, original value and abnormal type of the abnormal data. The data points marked as anomalous are numerically reconstructed using cubic spline interpolation, and smoothed by combining the correlation between time series data to obtain preprocessed operating condition data.

[0016] The annual energy consumption estimation method for complex air conditioning systems based on operating condition clustering and parameter optimization of the present invention has the following advantages over the prior art: (1) By constructing a simulation model of the air conditioning system, time-series preprocessing is used to eliminate abnormal interference in the hourly operating data throughout the year. Iterative clustering algorithm is used to perform cluster analysis on the preprocessed operating data and combined with hierarchical parameter optimization to improve computational efficiency and the accuracy of air conditioning energy consumption analysis. (2) By calculating the quartiles and interquartile ranges, the outlier discrimination criteria are determined, and outlier data points that deviate from the normal distribution are accurately identified. The cubic spline interpolation method is used for numerical reconstruction, which effectively eliminates outlier interference and maintains the continuity of time series data. (3) The basic weights of each feature parameter are calculated by Pearson correlation coefficient, and seasonal weights are obtained by adjusting seasonal factors. The optimal number of clusters is automatically determined by the Calinski-Harabasz index. Outlier removal and cluster center optimization are achieved by iterative clustering method to ensure steady-state calculation of typical clustering conditions, which greatly reduces the difficulty of energy consumption estimation. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1This is a flowchart of a method for estimating the annual energy consumption of a complex air conditioning system based on operating condition clustering and parameter optimization, according to the present invention. Figure 2 This is a block diagram of a method for estimating the annual energy consumption of a complex air conditioning system based on operating condition clustering and parameter optimization according to the present invention. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0020] like Figure 1 and Figure 2 As shown, this invention provides a method for estimating the annual energy consumption of a complex air conditioning system based on operating condition clustering and parameter optimization, including the following steps: S1. A simulation model of a complex air conditioning system is established using the MATLAB-Simulink platform. The simulation model includes a heat pump system, a dehumidifier impeller, a fan module, an electric heater module, and modules for air mixing and indoor air parameter calculation. The heat pump system is modeled using the System-Level Refrigeration Cycle module in Simulink, the fan module is modeled using the Flow Rate Source module in Simulink, the electric heater module is modeled using the Pipe (MA), Heat Flow Rate Source, and Thermal Reference modules in Simulink, the air mixing and indoor air parameter calculation module is modeled using the Constant Volume Chamber module in Simulink, and the dehumidifier impeller is modeled using the potential function method, with the potential function efficiency corrected based on the structural performance data of the dehumidifier impeller.

[0021] Specifically, the calculation formula for establishing the dehumidification rotor model using the potential function method is as follows: in, This represents the first potential function of the dehumidifier rotor. This indicates the temperature of the air entering the dehumidifier rotor. This indicates the moisture content of the air entering the dehumidification rotor. This represents the second potential function of the dehumidification rotor.

[0022] Furthermore, the actual heat and mass transfer efficiency of the dehumidification impeller model is: in, This represents the efficiency coefficient of the first potential function. This represents the calculated value of the first potential function under the air conditions at the dehumidifier rotor outlet. This represents the calculated value of the first potential function under the air handling conditions at the inlet of the dehumidifier rotor. This represents the calculated value of the first potential function under the condition of regenerated air in the dehumidifier rotor. This represents the efficiency coefficient of the second potential function. This represents the calculated value of the second potential function under the air conditions at the dehumidifier rotor outlet. This represents the calculated value of the second potential function under the air handling conditions at the inlet of the dehumidifier rotor. This represents the calculated value of the second potential function under the condition of regenerating air in the dehumidifier rotor.

[0023] Understandable. and This reflects the actual efficiency of the dehumidification rotor model in both temperature and humidity regulation. The efficiency coefficient value is determined through actual equipment performance test data, thereby correcting the performance of the theoretical potential function model.

[0024] S2. Obtain the hourly operating data of the complex air conditioning system throughout the year, and perform time-series preprocessing on the hourly operating data throughout the year to obtain the preprocessed operating data, which includes outdoor temperature, outdoor relative humidity, indoor sensible heat load and indoor moisture dissipation data.

[0025] In one embodiment of the present invention, step S2 specifically includes: Calculate the statistical characteristics of the data distribution for hourly operating data throughout the year, including the first quartile, third quartile, and interquartile range; the calculation formula is: in, This represents the first quartile, or 25th quartile, of the data sequence. This represents the third quartile, or 75th quartile, of the data sequence. Indicates the interquartile range. This represents the value at the k-th position in the data sequence. Indicates the total length of the data sequence; Based on the statistical principles of box plots, the criteria for identifying outliers are determined, and data points that exceed the normal data distribution range are marked as outliers. The discrimination formula is as follows: in, This represents the operating condition data value at time g. Data points that satisfy any of the above conditions are marked as outliers, with the lower bound of outliers being [value missing]. The upper bound for outliers is ; The identified abnormal data is marked and temporarily removed, and an abnormal data index list is established, which includes the time location, original value and abnormal type of the abnormal data. For data points marked as outliers, cubic spline interpolation is used for numerical reconstruction. Smoothing is then applied based on the correlation between time series data to obtain preprocessed operating condition data. The calculation formula is as follows: in, This represents the operating condition data value at time g after interpolation and smoothing. Represents the coefficients of cubic spline interpolation. This represents the time coordinate at the g-th moment. This represents the degree exponent in the cubic spline interpolation formula.

[0026] Understandably, the interpolation coefficients are determined by the values ​​of adjacent normal data points and slope constraints.

[0027] In one embodiment of the present invention, for the case of multiple consecutive abnormal data points, a method of piecewise linear interpolation combined with weight decay is used for processing: in, and These represent the nearest normal data values ​​on the left and right sides of the abnormal data segment, respectively. Indicates the weighting coefficient on the left. Indicates the weighting coefficient on the right. and These represent the time distances to the left and right normal data points, respectively. The attenuation coefficient is set to 0.1.

[0028] Understandably, the hourly operating data throughout the year is obtained through building energy consumption analysis software or meteorological databases, containing 8,760 hours of continuous data points, covering an outdoor temperature range of -10°C to 40°C and an outdoor relative humidity range of 15% to 98%. The indoor sensible heat load and moisture dissipation are determined based on the building characteristics and usage functions.

[0029] This invention determines the outlier discrimination criteria by calculating quartiles and interquartile ranges, accurately identifies outlier data points that deviate from the normal distribution, and uses cubic spline interpolation to reconstruct the numerical values, effectively eliminating outlier interference, maintaining the continuity of time series data, and significantly improving the accuracy and reliability of energy consumption estimation results.

[0030] S3. The preprocessed working condition data is clustered using an iterative clustering algorithm to obtain clustered working conditions, corresponding outdoor temperature and humidity and indoor heat and humidity load data, and the number of hours throughout the year. Specifically, step S3 includes: S31. Extract the operating characteristic parameters of the complex air conditioning system from the preprocessed operating condition data, and calculate the Pearson correlation coefficient between each characteristic parameter and the air conditioning energy consumption. Divide the absolute value of the Pearson correlation coefficient of each characteristic parameter by the sum of the absolute values ​​of the Pearson correlation coefficients of all characteristic parameters to obtain the normalized basic weight coefficient of that characteristic parameter; where the characteristic parameters include outdoor temperature. outdoor relative humidity Indoor sensible heat load Indoor moisture dissipation The calculation formula is: in, This represents the Pearson correlation coefficient between the i-th characteristic parameter and air conditioning energy consumption. This represents the index of the preprocessed operating condition data points. This indicates the total number of preprocessed operating condition data points. This represents the value of the i-th feature parameter of the d-th working condition data point. Let represent the mean of the i-th feature parameter. This represents the air conditioning energy consumption at the d-th operating condition data point. This represents the average energy consumption of the air conditioner. Represents the normalized basic weight coefficients. Indicates the feature parameter index. This represents the Pearson correlation coefficient of the k-th feature parameter; S32. Adjust the basic weight coefficients based on seasonal factors to obtain seasonal weights; S33. The Calinski-Harabasz index is used to determine the optimal number of clusters by traversing different numbers of clusters. S34. Based on the optimal number of clusters and seasonal weights, an iterative clustering method is used to cluster the preprocessed operating data, generating clustered operating conditions, corresponding outdoor temperature and humidity and indoor heat and humidity load data, and the number of hours throughout the year.

[0031] Furthermore, step S32 specifically includes: The season type is determined based on the outdoor temperature, with the following logic: ; in, Represents a seasonal type variable. Indicates the heating season. Indicates the transitional season. Indicates the cooling season. Indicates the outdoor temperature; The seasonal adjustment coefficients for each characteristic parameter are calculated using the following formula: ; in, This represents the seasonal adjustment coefficient for the i-th characteristic parameter. Indicates the feature parameter index; Based on the seasonal type and seasonal adjustment coefficient, the seasonal weight is determined using the following formula: ; in, This represents the seasonal weighting coefficient adjusted for the i-th feature parameter. This represents the basic weight coefficient of the i-th feature parameter.

[0032] In one embodiment of the present invention, the formula for calculating the Calinski-Harabasz index is as follows: ; in, This represents the Calinski-Harabasz index value. Represents the sum of squares between clusters. Represents the sum of squares within a cluster. This represents the total number of clustered operating conditions. This indicates the total number of preprocessed operating condition data points. This represents the adjustment coefficient. This represents the clustering condition index. This represents the average energy consumption value for the j-th cluster condition. This represents the overall average energy consumption value across all data points.

[0033] Understandably, the distance between preprocessed working condition data points is calculated using weighted Euclidean distance based on seasonal weights. For each candidate cluster size K ranging from 2 to 10, the standard K-means clustering algorithm is executed to calculate the Calinski-Harabasz index. By iterating through different K values ​​to calculate the Calinski-Harabasz index, the K that makes the index reach its maximum value is selected as the optimal number of clusters.

[0034] In one embodiment of the present invention, step S34 specifically includes: Step a: Set the control parameters for iterative clustering, including the maximum number of iterations, convergence accuracy threshold, and outlier removal ratio; initialize the cluster centers using the K-means++ algorithm. Step b: Perform outlier detection on the preprocessed operating data. Calculate the weighted distance from each operating data point to its respective cluster center. Determine the outlier threshold based on the statistical characteristics of the distance distribution. Mark operating data points exceeding the outlier threshold as outliers and remove them from the current dataset to form a new dataset. The formula for calculating the outlier threshold is: in, This represents the outlier detection threshold in the m-th iteration. This represents the mean distance within clusters in the m-th iteration. This represents the standard deviation of the distance within a cluster in the m-th iteration; Step c: Update the cluster center locations based on the new dataset, and assign each data point to the nearest cluster center to form the clustered dataset; the calculation formula is: in, This represents the update center of the f-th cluster in the (m+1)-th iteration. This represents the inertia factor, which is 0.3. This represents the center of the f-th cluster in the m-th iteration. This represents the i-th data point. Let f represent the set of data points belonging to the f-th cluster in the m-th iteration. This represents the distance weighting coefficient for the i-th data point. This represents the set of data points belonging to the f-th cluster in the (m+1)-th iteration. This represents the valid data set for the (m+1)th iteration. Represents a clustering index. Representing data points The weighted Euclidean distance to the f-th cluster center; Step d, repeat steps b and c until the maximum number of iterations is reached, to obtain the iteratively optimized cluster centers and the final data allocation. Based on the iteratively optimized cluster centers and the final data allocation, generate clustering conditions, corresponding outdoor temperature and humidity and indoor heat and humidity load data, and the number of hours throughout the year.

[0035] Understandable, weighting coefficient The weighting of cluster centers is determined by the distance from each data point to the cluster center; data points closer to the center have a greater weight in the calculation. During cluster center updates, data points closer to the center typically better represent the typical characteristics of the cluster and should therefore be given higher weight in the center location calculation. For air conditioning system operating condition clustering, this weighting strategy ensures that the cluster centers more accurately reflect the core operating condition characteristics and reduces the interference from edge data points. Introducing an outlier removal mechanism adaptively handles noise and outliers in the data; weighted center updates allow for more precise location of cluster centers; and the introduction of an inertia factor balances convergence speed and stability.

[0036] This invention calculates the basic weights of each feature parameter using the Pearson correlation coefficient, adjusts the seasonal weights by combining seasonal factors, automatically determines the optimal number of clusters using the Calinski-Harabasz index, and achieves outlier removal and cluster center optimization through iterative clustering methods. This effectively solves the problems of uneven distribution of operating data, significant seasonal differences, and difficulty in determining clustering parameters in complex air conditioning systems, ensuring steady-state calculations for typical clustering conditions and greatly reducing the difficulty of energy consumption estimation.

[0037] In one embodiment of the present invention, the maximum number of iterations is 10, the convergence accuracy threshold is 0.001, and the outlier extraction ratio is 0.05.

[0038] Specifically, a concrete embodiment is used for illustration. In this embodiment, the characteristic parameters selected are outdoor temperature and humidity and indoor sensible heat load. Only the system operation during the daytime period (8:00-18:00) is considered. The clustering results are shown in Table 1: Table 1 Clustering Results S4. Based on the clustered operating conditions, the corresponding outdoor temperature and humidity and indoor heat and humidity load data of the clustered operating conditions, and the number of hours throughout the year, the operating parameters of the air conditioning system simulation model are optimized in layers to obtain the optimal combination of operating parameters under each clustered operating condition. Specifically, step S4 includes: The adjustable operating parameters of the air conditioning system are divided into primary parameters and secondary parameters according to their impact on system energy consumption. The primary parameters include heat pump frequency, air supply volume and primary and secondary return air ratio, while the secondary parameters include electric heating power and fresh air volume. A mixed-integer nonlinear substitution optimization algorithm is used to perform the first-level coarse optimization of the main parameters, with the objective function being to minimize system energy consumption, to obtain the first-level optimal parameter combination, where the objective function is: ; The constraints are: ; in, This represents the minimum energy consumption target for optimizing the main parameters. This represents the energy consumption function with key parameters as variables and operating conditions as parameters. Indicates the heat pump frequency. Indicates the air supply volume. This indicates the ratio of primary to secondary return air. Indicates the outdoor dry-bulb temperature. Indicates outdoor relative humidity. Indicates the indoor sensible heat load. Indicates indoor moisture dissipation. Represents the set of positive integers; Based on the first-level optimal parameter combination, a second-level fine-tuning is performed on the secondary parameters to obtain the optimal operating parameter combination for each clustering condition. The objective function of the second-level fine-tuning is: The constraints are: in, This represents the minimum energy consumption target for optimizing secondary parameters. This represents the energy consumption function with secondary parameters as decision variables, the first-level optimization result, and operating conditions as given parameters. Indicates the electric heating power. Indicates the fresh air volume. This represents the optimal heat pump frequency obtained from the first layer of optimization. This represents the optimal air supply volume obtained from the first layer of optimization. This represents the optimal primary and secondary return air ratio obtained from the first layer of optimization. This indicates the minimum fresh air volume. This indicates the moisture content of the return air. This indicates the return air temperature.

[0039] This invention employs a mixed-integer nonlinear substitution optimization algorithm to perform a first-level coarse optimization on the main parameters, and then performs a second-level fine optimization on the secondary parameters. This effectively reduces the computational complexity of multi-parameter coupled optimization and the risk of getting trapped in local optima, thereby improving the efficiency of parameter optimization.

[0040] Specifically, a specific embodiment is used for illustration. In this embodiment, the energy consumption of the four clustering conditions is 47.2kW, 32.5kW, 32.7kW, and 14.4kW, respectively. The optimized operating parameter configuration under the clustering conditions is shown in Table 2. Table 2 Optimal operating parameters under clustering conditions S5. Input the optimal combination of operating parameters under each cluster condition into the air conditioning system simulation model, and obtain the optimized energy consumption value under each cluster condition through simulation calculation. Based on the optimized energy consumption value and corresponding number of hours under each cluster condition, calculate the estimated annual energy consumption value of the complex air conditioning system.

[0041] Specifically, the logic for calculating the estimated annual energy consumption of a complex air conditioning system is as follows: ; in, This represents the estimated annual energy consumption of a complex air conditioning system. This represents the clustering condition index. This represents the total number of clustered operating conditions. This represents the optimized energy consumption value for the j-th clustering condition. This represents the total number of hours throughout the year for the j-th cluster condition.

[0042] This invention constructs a simulation model of an air conditioning system, uses time-series preprocessing to eliminate abnormal interference from hourly operating data throughout the year, and employs an iterative clustering algorithm to perform cluster analysis on the preprocessed operating data, combined with hierarchical parameter optimization, thereby improving computational efficiency and the accuracy of air conditioning energy consumption analysis.

[0043] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for estimating the annual energy consumption of a complex air conditioning system based on operating condition clustering and parameter optimization, characterized in that: Includes the following steps: S1. Use the MATLAB-Simulink platform to establish a simulation model of a complex air conditioning system. S2. Obtain the hourly operating data of the complex air conditioning system throughout the year, and perform time-series preprocessing on the hourly operating data throughout the year to obtain the preprocessed operating data, which includes outdoor temperature, outdoor relative humidity, indoor sensible heat load and indoor moisture dissipation data. S3. The preprocessed working condition data is clustered using an iterative clustering algorithm to obtain clustered working conditions, corresponding outdoor temperature and humidity and indoor heat and humidity load data, and the number of hours throughout the year. S4. Based on the clustered operating conditions, the corresponding outdoor temperature and humidity and indoor heat and humidity load data of the clustered operating conditions, and the number of hours throughout the year, the operating parameters of the air conditioning system simulation model are optimized in layers to obtain the optimal combination of operating parameters under each clustered operating condition. S5. Input the optimal combination of operating parameters under each cluster condition into the air conditioning system simulation model, and obtain the optimized energy consumption value under each cluster condition through simulation calculation. Based on the optimized energy consumption value and corresponding number of hours under each cluster condition, calculate the estimated annual energy consumption value of the complex air conditioning system.

2. The method for estimating the annual energy consumption of a complex air conditioning system based on operating condition clustering and parameter optimization as described in claim 1, characterized in that: The simulation model of the complex air conditioning system includes a heat pump system, a dehumidifier impeller, a fan module, an electric heater module, and an air mixing and indoor air parameter calculation module. The heat pump system is modeled using the System-Level Refrigeration Cycle module in Simulink, the fan module is modeled using the Flow Rate Source module in Simulink, the electric heater module is modeled using the Pipe, Heat Flow Rate Source, and Thermal Reference modules in Simulink, the air mixing and indoor air parameter calculation module is modeled using the ConstantVolume Chamber module in Simulink, and the dehumidifier impeller is modeled using the potential function method, with the potential function efficiency corrected based on the structural performance data of the dehumidifier impeller.

3. The method for estimating the annual energy consumption of a complex air conditioning system based on operating condition clustering and parameter optimization as described in claim 1, characterized in that: Step S3 specifically includes: S31. Extract the operating characteristic parameters of the complex air conditioning system from the preprocessed operating condition data, and calculate the Pearson correlation coefficient between each characteristic parameter and the air conditioning energy consumption. Divide the absolute value of the Pearson correlation coefficient of each characteristic parameter by the sum of the absolute values ​​of the Pearson correlation coefficients of all characteristic parameters to obtain the normalized basic weight coefficient of the characteristic parameter. The characteristic parameters include outdoor temperature, outdoor relative humidity, indoor sensible heat load and indoor moisture dissipation. S32. Adjust the basic weight coefficients based on seasonal factors to obtain seasonal weights; S33. The Calinski-Harabasz index is used to determine the optimal number of clusters by traversing different numbers of clusters. S34. Based on the optimal number of clusters and seasonal weights, an iterative clustering method is used to cluster the preprocessed operating data, generating clustered operating conditions, corresponding outdoor temperature and humidity and indoor heat and humidity load data, and the number of hours throughout the year.

4. The method for estimating the annual energy consumption of a complex air conditioning system based on operating condition clustering and parameter optimization as described in claim 3, characterized in that: Step S32 specifically includes: The season type is determined based on the outdoor temperature, with the following logic: ； in, Represents a seasonal type variable. Indicates the heating season. Indicates the transitional season. Indicates the cooling season. Indicates the outdoor temperature; The seasonal adjustment coefficients for each characteristic parameter are calculated using the following formula: ； in, This represents the seasonal adjustment coefficient for the i-th characteristic parameter. Indicates the feature parameter index; Based on the seasonal type and seasonal adjustment coefficient, the seasonal weight is determined using the following formula: ； in, This represents the seasonal weighting coefficient adjusted for the i-th feature parameter. This represents the basic weight coefficient of the i-th feature parameter.

5. The method for estimating the annual energy consumption of a complex air conditioning system based on operating condition clustering and parameter optimization as described in claim 4, characterized in that: The Calinski-Harabasz index is calculated using the following formula: ； in, This represents the Calinski-Harabasz index value. Represents the sum of squares between clusters. Represents the sum of squares within a cluster. This represents the total number of clustered operating conditions. This indicates the total number of preprocessed operating condition data points. This represents the adjustment coefficient. This represents the clustering condition index. This represents the average energy consumption value for the j-th cluster condition. This represents the overall average energy consumption value across all data points.

6. The method for estimating the annual energy consumption of a complex air conditioning system based on operating condition clustering and parameter optimization as described in claim 5, characterized in that: Step S34 specifically includes: Step a: Set the control parameters for iterative clustering, including the maximum number of iterations, convergence accuracy threshold, and outlier removal ratio; initialize the cluster centers using the K-means++ algorithm. Step b: Perform outlier detection on the preprocessed working condition data, calculate the weighted distance from each working condition data point to its respective cluster center, determine the outlier threshold based on the statistical characteristics of the distance distribution, mark the working condition data points that exceed the outlier threshold as outliers and remove them from the current dataset to form a new dataset; Step c: Update the cluster center locations based on the new dataset, and assign each data point to the nearest cluster center to form the dataset after cluster assignment; Step d, repeat steps b and c until the maximum number of iterations is reached, to obtain the iteratively optimized cluster centers and the final data allocation. Based on the iteratively optimized cluster centers and the final data allocation, generate clustering conditions, corresponding outdoor temperature and humidity and indoor heat and humidity load data, and the number of hours throughout the year.

7. The method for estimating the annual energy consumption of a complex air conditioning system based on operating condition clustering and parameter optimization as described in claim 1, characterized in that: Step S4 specifically includes: The adjustable operating parameters of the air conditioning system are divided into primary parameters and secondary parameters according to their impact on system energy consumption. The primary parameters include heat pump frequency, air supply volume and primary and secondary return air ratio, while the secondary parameters include electric heating power and fresh air volume. A mixed-integer nonlinear substitution optimization algorithm is used to perform the first-level coarse optimization of the main parameters, with the objective function being to minimize system energy consumption, to obtain the first-level optimal parameter combination, where the objective function is: ； The constraints are: ; in, This represents the minimum energy consumption target for optimizing the main parameters. This represents the energy consumption function with key parameters as variables and operating conditions as parameters. Indicates the heat pump frequency. Indicates the air supply volume. This indicates the ratio of primary to secondary return air. Indicates the outdoor dry-bulb temperature. Indicates outdoor relative humidity. Indicates the indoor sensible heat load. Indicates indoor moisture dissipation. Represents the set of positive integers; Based on the first-level optimal parameter combination, a second-level fine-tuning is performed on the secondary parameters to obtain the optimal operating parameter combination for each clustering condition. The objective function of the second-level fine-tuning is: The constraints are: in, This represents the minimum energy consumption target for optimizing secondary parameters. This represents the energy consumption function with secondary parameters as decision variables, the first-level optimization result, and operating conditions as given parameters. Indicates the electric heating power. Indicates the fresh air volume. This represents the optimal heat pump frequency obtained from the first layer of optimization. This represents the optimal air supply volume obtained from the first layer of optimization. This represents the optimal primary and secondary return air ratio obtained from the first layer of optimization. This indicates the minimum fresh air volume. This indicates the moisture content of the return air. This indicates the return air temperature.

8. The method for estimating the annual energy consumption of a complex air conditioning system based on operating condition clustering and parameter optimization as described in claim 1, characterized in that: The logic for calculating the estimated annual energy consumption of a complex air conditioning system is as follows: ； in, This represents the estimated annual energy consumption of a complex air conditioning system. This represents the clustering condition index. This represents the total number of clustered operating conditions. This represents the optimized energy consumption value for the j-th clustering condition. This represents the total number of hours throughout the year for the j-th cluster condition.

9. The method for estimating the annual energy consumption of a complex air conditioning system based on operating condition clustering and parameter optimization as described in claim 1, characterized in that: The calculation formula for establishing the dehumidification rotor model using the potential function method is as follows: ； ； in, This represents the first potential function of the dehumidifier rotor. This indicates the temperature of the air entering the dehumidifier rotor. This indicates the moisture content of the air entering the dehumidification rotor. This represents the second potential function of the dehumidification rotor.

10. The method for estimating the annual energy consumption of a complex air conditioning system based on operating condition clustering and parameter optimization as described in claim 1, characterized in that: Step S2 specifically includes: Calculate the statistical characteristics of the data distribution for hourly operating data throughout the year, including the first quartile, the third quartile, and the interquartile range; Based on the statistical principle of box plots, the criteria for identifying outlier data are determined, and data points that exceed the normal data distribution range are marked as outliers. The identified abnormal data is marked and temporarily removed, and an abnormal data index list is established, which includes the time location, original value and abnormal type of the abnormal data. The data points marked as anomalous are numerically reconstructed using cubic spline interpolation, and smoothed by combining the correlation between time series data to obtain preprocessed operating condition data.

Images