An energy-saving control algorithm for central air-conditioning energy control system

Abstract

The present invention discloses an energy-saving control algorithm for a central air-conditioning energy control system, comprising: predicting the total power of the central air-conditioning system in a target building space based on historical meteorological data; establishing a heat transfer model for each partitioned space based on the zoning of the target building; establishing a power control model for the central air-conditioning system; establishing individual energy consumption models for each component of the central air-conditioning system using physical models; establishing constraints for each component of the central air-conditioning system, and controlling each component in combination with the constraints of each component of the central air-conditioning system. An updated prediction model is obtained by establishing a heat transfer formula for the partitioned space and an energy consumption model for the central air-conditioning system. The updated model prediction value is compared with the actual interpolated value and introduced into the preliminary prediction model as a correction reference, and the air-conditioning is then controlled according to the constraints.

Description

Technical Field

[0001] The present invention relates to the technical field of energy control, and in particular to an energy-saving control algorithm for a central air-conditioning energy control system. Background Art

[0002] While various technologies are currently available to optimize energy consumption in building central air conditioning systems, most focus on calculating the energy consumption of individual components (such as chillers, pumps, and fans) during operation. This control strategy often overlooks the complexity and diversity of a building's heat distribution, particularly in super-high-rise buildings and specialized industrial structures (such as high-rise industrial pig farms), where heat distribution varies significantly.

[0003] Chinese patent document CN 114508784A describes a multi-source complementary heating system and its optimization control method. The control system uses an optimization control algorithm in a core controller to optimize and calculate sensor data. The calculated data serves as the set value for the underlying controller, which automatically controls the power output of the solar collector, electric boiler, and ground-source heat pump, and the opening of the hot water storage tank outlet valve according to the set value. CN 116734404A describes an energy-saving optimization control system and optimization method for a central air conditioning system. The system includes: a chiller physical model, which reflects the basic operating characteristics of the actual equipment and is used to calculate the COP of the central air conditioning system's upper control unit; a load prediction model, which is constructed based on a neural network algorithm; a cold source energy efficiency model, which is used to obtain energy consumption optimal state point data; and a water pump performance model, which is used to calculate water pump energy consumption under operating conditions.

[0004] The existing technology has the following limitations:

[0005] 1. Ignoring thermal differences between building zones: Most existing energy-saving control technologies fail to fully consider the thermal zoning characteristics of a building. For example, in super-high-rise buildings, rooms at different heights and orientations experience significant differences in heat distribution due to factors such as solar radiation, wind direction and speed. Furthermore, the top and bottom floors of high-rise buildings also experience different thermal environments due to differing meteorological conditions such as atmospheric pressure and temperature stratification. However, existing control systems often employ a unified control strategy and are unable to accurately adjust the water system supply based on the actual heat demand of each zone, resulting in either excess or insufficient cooling or heating capacity.

[0006] 2. Mismatch between target distribution and heat demand: In large public buildings or specialized industrial structures (such as high-rise industrial pig farms), target distribution is often uneven and changes over time. For example, in pig farms, pig density and activity intensity vary across different areas, leading to uneven heat generation and distribution. Existing energy-saving control systems typically operate based on fixed load setpoints and are unable to perceive and respond to these dynamic changes in real time, resulting in wasted energy.

[0007] 3. Uneven energy distribution in the water system: Central air conditioning systems typically consist of both air and water systems, with the water system accounting for a significant portion of energy consumption. However, existing water system control methods often predict and adjust the average load of the entire system, ignoring the varying demands for hot and cold water within each building zone. This results in over- or under-allocation of water to certain areas during operation, resulting in suboptimal cooling or heating performance, and even energy waste and inefficiency.

[0008] 4. Lack of Intelligent Monitoring and Adjustment: Existing energy-saving control systems still have shortcomings in monitoring and adjustment. Many systems lack real-time, accurate sensor networks to monitor internal building parameters such as temperature, humidity, and foot traffic, resulting in the inability to adjust control strategies to actual conditions. Furthermore, some systems have relatively simple adjustment methods, making it impossible to achieve refined control of complex thermal environments.

[0009] In summary, while existing energy-saving control technologies for building central air conditioning systems have achieved some success in reducing energy consumption, they still have many shortcomings in practical application because they ignore the complexity and diversity of heat distribution within building zones. Therefore, it is particularly important to develop an energy-saving control algorithm for central air conditioning based on building zone heat distribution. This algorithm should be able to perceive the heat distribution within each building zone in real time and make precise adjustments based on actual needs, thereby achieving efficient energy utilization and energy conservation and emission reduction goals. Summary of the Invention

[0010] The technical problem to be solved by the present invention is to provide an energy-saving control algorithm for a central air-conditioning energy control system, which overcomes the shortcomings of the existing technology, meets the heat demand of the target building, and improves the utilization efficiency of the energy consumption of the building's central air-conditioning.

[0011] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0012] An energy-saving control algorithm for a central air-conditioning energy control system includes the following steps:

[0013] Step 1: Partition the target building space based on its structure, including division based on height, location, and lighting;

[0014] Step 2: Set up a weather monitoring station at the target building to monitor the weather parameters in the target building area and save the corresponding data;

[0015] Step 3: Install light intensity sensors in the areas outside the target building where light can reach the target building area. Install temperature and humidity sensors in each area divided in Step 1. Establish communication between each sensor and the central air conditioning control system. Record and store the historical temperature control status of each area and the total power of the central control system.

[0016] Step 4: Preprocess the historical meteorological data in Step 2 and the corresponding historical data of the total power of the central air-conditioning system in Step 3;

[0017] Step 5: Perform correlation analysis on the historical meteorological data preprocessed in Step 4 and the historical data of the total power of the central air-conditioning system, extract meteorological features whose impact on the total power of the central air-conditioning system is greater than the set value, and eliminate invalid meteorological features. Step 6: Decompose the historical data of the total power of the central air-conditioning system into different intrinsic mode functions (IMFs) using variational mode decomposition, and reconstruct all IMFs into high-frequency components and low-frequency components by calculating the approximate entropy (AE) of each IMF.

[0018] Step 7, using a gated neural network to predict the total power of the central air-conditioning system at a future time point based on the historical meteorological data obtained in Step 5 and the historical high-frequency power components and historical low-frequency power components in Step 6;

[0019] Step 8: Establish a heat transfer model for each partitioned space based on the target building partitioning in Step 1.

[0020] Step 9: Establish a power control model for the central air-conditioning system; establish the energy consumption model of each component through the physical model of the central air-conditioning system;

[0021] Step 10: Establish the constraints of each component of the central air-conditioning system and combine them with the heat transfer of each partition in Step 8 to obtain the actual power model of the central air-conditioning system.

[0022] Step 11: Compare the difference between the total power value of the central control system predicted in Step 7 and the actual total power value, and introduce the power model of the central air-conditioning system obtained in Step 10 as a reference for correction to obtain a corrected total power prediction model of the central control system;

[0023] Step 12: Based on the power prediction value obtained from the revised total power prediction model of the central control system, each component of the central air-conditioning system is controlled in combination with the constraints of the components.

[0024] The meteorological parameters in Step 2 above include light intensity, light direction, air pressure, and temperature.

[0025] The preprocessing of Step 4 above includes filling missing values ​​in the collected historical meteorological data and the historical data of the total power of the central air conditioning system with the mean of the records in the set time interval before and after. For outliers, the isolation forest algorithm is used to detect and eliminate them according to the following steps:

[0026] The isolation forest algorithm continuously segments the data samples until each sample space contains only one type of data point. The segmentation process is as follows:

[0027] For the data set X, a binary tree T with N nodes is used to describe it, and each node or The data of is a subset of X, where i represents the number of layers in the tree, j represents the jth node in the previous layer, and r and l are used to distinguish the right and left nodes in the same layer;

[0028] For the dataset contained in a certain layer , randomly select the sample attribute q and its value range space value p to divide and , corresponding to the node set and , data less than or equal to p is divided into Node, others are divided into Node, where j* represents the j*th node in the i+1th layer;

[0029] When the following situation occurs, a complete binary tree is obtained and the partitioning is completed:

[0030] A. The depth of the data tree reaches the set maximum value;

[0031] B. Node Contains only one data point or the data points contained are identical.

[0032] In the above Step 5, the Pearson correlation coefficient is used to analyze the correlation.

[0033] The heat transfer model of a single space in each partition in Step 8 above is:

[0034] = ;

[0035] ;

[0036] ;

[0037] ;

[0038] ;

[0039] Where, i represents the moment, C is the heat capacity of the building space, is the dimensionless distribution coefficient of the air conditioning cooling output in space: ; F is the thermal diffusivity, t is the cooling time, and S is the space area; It is the space between the object space and its neighbors. i The thermal resistance between To open space from the object space to its neighbors k The thermal induction between For space i The temperature of the moment, For k and k 1 The net heat gain during the time between is the length of the time interval, is the thermal resistance between the target space and the outdoor air, M is the number of spaces that communicate with the target space, It is an adjacent space that communicates with the object space k exist i Relative humidity at the time, It is an adjacent space that communicates with the object space k exist i Net heat gain at any given moment.

[0040] The objective function of the power control model of the central air-conditioning system in Step 9 above is:

[0041] ;

[0042] Energy consumption in central air conditioning systems Energy consumption of fan coil units and fresh air units , Chilled water pump energy consumption , refrigeration unit energy consumption , cooling water pump energy consumption and cooling tower energy consumption composition.

[0043] Energy consumption of the above fan coil units and fresh air units It is composed of multiple chillers, and its mathematical model is:

[0044] ;

[0045] in:

[0046] ;

[0047] ;

[0048] The total cooling load of the system is the sum of the cooling loads of each room, and the total cooling load is distributed to each chiller in proportion; the actual cooling capacity of each unit, the number of units, and the unit capacity meet the following formula:

[0049] ;

[0050] In the above formula, is the total energy consumption of multiple chillers, For the i The cooling capacity of the unit, For the i The efficiency of a unit at a certain load rate, For the i The load factor of the unit, For the i The temperature regulation coefficient of each unit, For the i The actual cooling capacity of a unit under a certain load, is the actual cooling load of a circuit in the building, is the chilled water supply temperature of the chiller, is the return water temperature of the chiller cooling water, is the number of refrigeration units, is the number of fans, is the number of air conditioners;

[0051] The mathematical model of energy consumption of chilled water pump is as follows:

[0052] ;

[0053] Where, is the total efficiency of the variable speed pump, which is affected by the water flow rate and the pump head; N2 is the number of chilled water pumps;

[0054] The water flow of the chilled water pump is the sum of the flow rates of the chilled water flowing through each circuit. The distribution relationship between the total chilled water flow and the flow rates of each chilled water pump is determined by the following formula:

[0055] ;

[0056] When the speed of the variable speed water pump is n, the pressure and flow rate satisfy the following curve equation relationship:

[0057] ;

[0058] The speed ratio , is the rated speed, ~ is the model parameter, then the energy consumption of the chilled water pump is Converted into a relationship that is only related to the speed ratio and the chilled water pump flow rate:

[0059] ;

[0060] Energy consumption of fan coil units and fresh air units The mathematical model is as follows:

[0061]

[0062] Where, is the air flow rate, is the air pressure, For variable speed fans, medium efficiency;

[0063] Air volume per circuit Is the air volume supplied to all air-conditioned rooms The relationship between the total air volume and the air volume of each air conditioning unit is determined by the following formula:

[0064] ;

[0065] When the speed of the variable speed fan is n, the pressure and air volume satisfy the following curve equation relationship:

[0066] ;

[0067] The speed ratio , is the rated speed, ~ is the model parameter, then the energy consumption of fan coil unit and fresh air unit is Converted into a relationship that is only related to the speed ratio and the air volume of each fan coil unit:

[0068]

[0069] In the above formula, Represents a constant whose value varies depending on the dimensional unit of the total pressure or flow rate. If the dimensional units of the air flow rate and the water flow rate are the same, and the dimensional units of the pump head and the air pressure are the same, then for the above formulas It will be the same;

[0070] Cooling pump energy consumption and cooling tower energy consumption The mathematical model is expressed as follows:

[0071] ;

[0072] ;

[0073] ;

[0074] ;

[0075] In the system modeling process, parameters , , , , , The value of can be obtained from the sample data of the corresponding manufacturer. The cooling load at the end of each air conditioning circuit Available from DDC Systems.

[0076] The constraints of the power control model of the central air conditioning system in Step 10 above are:

[0077] The range of change of the inlet water temperature and outlet water temperature of the chiller is as follows:

[0078] ;

[0079] ;

[0080] The water flow through the chiller pump and cooling pump is subject to the following constraints:

[0081] ;

[0082] ;

[0083] The air volume of the air conditioning unit and cooling tower is subject to the following constraints:

[0084] ;

[0085] ;

[0086] The pressure that the air conditioning unit and refrigeration pump can provide must be within the operating range of the variable speed control:

[0087] ;

[0088] ;

[0089] The range of variation of fresh air volume of each fan coil unit:

[0090] .

[0091] In the above Step 11, the total power value of the central control system predicted in Step 7 is compared with the actual total power value to obtain the difference between multiple time intervals. The multiple differences are introduced into the power model of the central air-conditioning system obtained in Step 10, and the differences are multiplied by the set weights and added to the power model of the central air-conditioning system to obtain the corrected power model of the central air-conditioning system. The prediction is then repeated until the difference is less than the set threshold.

[0092] In the above Step 1, the area that can receive light is distinguished from the area that cannot receive light.

[0093] The present invention provides an energy-saving control algorithm for a central air-conditioning energy control system. The algorithm divides the building into zones according to its structure, models the meteorological conditions in the area where the current building is located and the historical power data of the building's central air-conditioning to obtain a preliminary prediction model. The algorithm establishes a heat transfer formula for the zoned space and an energy consumption model for the central air-conditioning system to obtain an updated prediction model. The updated model prediction value is compared with the actual interpolation value and introduced into the preliminary prediction model as a correction reference, and the air-conditioning is then controlled according to the constraint conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0094] The present invention will be further described below with reference to the accompanying drawings and examples:

[0095] Figure 1 Schematic diagram of vertical area division of a high-rise building in an embodiment of the present invention;

[0096] Figure 2 Schematic diagram of cross-sectional area division in an embodiment of the present invention. DETAILED DESCRIPTION

[0097] The technical solution of the present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0098] An energy-saving control algorithm for a central air-conditioning energy control system includes the following steps:

[0099] Step 1, such as Figure 1 and 2 As shown in , the target building space is partitioned according to the structure of the target building. The partitioning includes division according to height, location and lighting. In super high-rise buildings, the surrounding area around the core support part is divided into areas that can directly receive sunlight and areas that cannot directly receive sunlight. At the same time, different heights are also divided to cope with changes in meteorological conditions at different heights, such as changes in air pressure, cloud cover and wind speed;

[0100] Similarly, in large-scale building complexes, such as high-rise industrial farm buildings, only the outer layer receives sunlight, while many interior areas cannot directly receive sunlight. These areas should be divided and the heat source in each area, that is, the number of people or farmed animals, should be monitored.

[0101] Step 2: Set up a weather monitoring station at the target building to monitor the weather parameters in the target building area and save the corresponding data;

[0102] Step 3: Install light intensity sensors in the areas outside the target building where light can reach the target building area. Install temperature and humidity sensors in each area divided in Step 1. Establish communication between each sensor and the central air conditioning control system. Record and store the historical temperature control status of each area and the total power of the central control system.

[0103] Step 4: Preprocess the historical meteorological data in Step 2 and the corresponding historical data of the total power of the central air-conditioning system in Step 3;

[0104] Step 5: Perform correlation analysis on the historical meteorological data preprocessed in Step 4 and the historical data of the total power of the central air-conditioning system, extract meteorological features whose impact on the total power of the central air-conditioning system is greater than the set value, and eliminate invalid meteorological features. Step 6: Decompose the historical data of the total power of the central air-conditioning system into different intrinsic mode functions (IMFs) using variational mode decomposition, and reconstruct all IMFs into high-frequency components and low-frequency components by calculating the approximate entropy (AE) of each IMF.

[0105] Step 7, using a gated neural network to predict the total power of the central air-conditioning system at a future time point based on the historical meteorological data obtained in Step 5 and the historical high-frequency power components and historical low-frequency power components in Step 6;

[0106] Step 8: Establish a heat transfer model for each partitioned space based on the target building partitioning in Step 1.

[0107] Step 9: Establish a power control model for the central air-conditioning system; establish the energy consumption model of each component through the physical model of the central air-conditioning system;

[0108] Step 10: Establish the constraints of each component of the central air-conditioning system and combine them with the heat transfer of each partition in Step 8 to obtain the actual power model of the central air-conditioning system.

[0109] Step 11: Compare the difference between the total power value of the central control system predicted in Step 7 and the actual total power value, and introduce the power model of the central air-conditioning system obtained in Step 10 as a reference for correction to obtain a corrected total power prediction model of the central control system;

[0110] Step 12: Based on the power prediction value obtained from the revised total power prediction model of the central control system, each component of the central air-conditioning system is controlled in combination with the constraints of the components.

[0111] The meteorological parameters in Step 2 above include light intensity, light direction, air pressure, and temperature.

[0112] The preprocessing of Step 4 above includes filling missing values ​​in the collected historical meteorological data and the historical data of the total power of the central air conditioning system with the mean of the records in the set time interval before and after. For outliers, the isolation forest algorithm is used to detect and eliminate them according to the following steps:

[0113] The isolation forest algorithm continuously segments the data samples until each sample space contains only one type of data point. The segmentation process is as follows:

[0114] For the data set X, a binary tree T with N nodes is used to describe it, and each node or The data of is a subset of X, where i represents the number of layers in the tree, j represents the jth node in the previous layer, and r and l are used to distinguish the right and left nodes in the same layer;

[0115] For the dataset contained in a certain layer , randomly select the sample attribute q and its value range space value p to divide and , corresponding to the node set and , data less than or equal to p is divided into Node, others are divided into Node, where j* represents the j*th node in the i+1th layer;

[0116] When the following situation occurs, a complete binary tree is obtained and the partitioning is completed:

[0117] A. The depth of the data tree reaches the set maximum value;

[0118] B. Node Contains only one data point or the data points contained are identical.

[0119] In the above Step 5, the Pearson correlation coefficient is used to analyze the correlation. The specific process is as follows:

[0120] The Pearson correlation coefficient is used to calculate the correlation coefficient between each meteorological characteristic and the power of the central air conditioning system. The formula is:

[0121] ;

[0122] in the formula Indicates meteorological characteristics. represents photovoltaic power generation power, for and The covariance of and They are and variance;

[0123] According to the results of correlation analysis, invalid meteorological features with correlation coefficients lower than the set value or negative correlation are eliminated, and the remaining meteorological features are retained; the multidimensional meteorological features are reduced in dimension to obtain multiple components and corresponding contribution rates, and each component is weighted and summed according to its contribution rate to obtain a comprehensive meteorological factor and its contribution rate.

[0124] In Step 6, variational mode decomposition (VMD) is used to decompose the photovoltaic power into a series of intrinsic mode functions (IMFs) with specific bandwidths. By calculating the approximate entropy AE of each IMF, all IMFs are reconstructed into high-frequency components and low-frequency components. The high-frequency components reflect the fluctuation characteristics of the power, while the low-frequency components reflect the changing trend of the power.

[0125] In Step 7, a gated recurrent neural network is used to add an attention mechanism to obtain different prediction models. Different prediction models receive the high-frequency component and low-frequency component in Step 6, and the prediction results under the two frequency modes are superimposed and reconstructed to obtain the final prediction result.

[0126] The heat transfer model of a single space in each partition in Step 8 above is:

[0127] = ;

[0128] ;

[0129] ;

[0130] ;

[0131] ;

[0132] Where, i represents the moment, C is the heat capacity of the building space, is the dimensionless distribution coefficient of the air conditioning cooling output in space: ; F is the thermal diffusivity, t is the cooling time, and S is the space area; It is the space between the object space and its neighbors. i The thermal resistance between To open space from the object space to its neighbors k The thermal induction between For space i The temperature of the moment, For k and k 1 The net heat gain during the time between is the length of the time interval, is the thermal resistance between the target space and the outdoor air, M is the number of spaces that communicate with the target space, It is an adjacent space that communicates with the object space k exist i Relative humidity at the time, It is an adjacent space that communicates with the object space k exist i Net heat gain at any given moment.

[0133] The spatial heat transfer here includes the heat transfer between areas that can directly receive light and those that cannot. At the same time, the longitude and latitude information can be used to obtain the light direction of different areas that can directly receive light at different times and when the light is lost (as the earth rotates, the light direction changes, and the light conditions are lost).

[0134] The objective function of the power control model of the central air-conditioning system in Step 9 above is:

[0135] ;

[0136] Energy consumption in central air conditioning systems Energy consumption of fan coil units and fresh air units , Chilled water pump energy consumption , refrigeration unit energy consumption , cooling water pump energy consumption and cooling tower energy consumption composition.

[0137] Energy consumption of the above fan coil units and fresh air units It is composed of multiple chillers, and its mathematical model is:

[0138] ;

[0139] in:

[0140] ;

[0141] ;

[0142] The total cooling load of the system is the sum of the cooling loads of each room, and the total cooling load is distributed to each chiller in proportion; the actual cooling capacity of each unit, the number of units, and the unit capacity meet the following formula:

[0143] ;

[0144] In the above formula, is the total energy consumption of multiple chillers, For the i The cooling capacity of the unit, For the i The efficiency of a unit at a certain load rate, For the i The load factor of the unit, For the i The temperature regulation coefficient of each unit, For the i The actual cooling capacity of a unit under a certain load, is the actual cooling load of a circuit in the building, is the chilled water supply temperature of the chiller, is the return water temperature of the chiller cooling water, is the number of refrigeration units, is the number of fans, is the number of air conditioners;

[0145] The mathematical model of energy consumption of chilled water pump is as follows:

[0146] ;

[0147] Where, is the total efficiency of the variable speed pump, which is affected by the water flow rate and the pump head; N2 is the number of chilled water pumps;

[0148] The water flow of the chilled water pump is the sum of the flow rates of the chilled water flowing through each circuit. The distribution relationship between the total chilled water flow and the flow rates of each chilled water pump is determined by the following formula:

[0149] ;

[0150] When the speed of the variable speed water pump is n, the pressure and flow rate satisfy the following curve equation relationship:

[0151] ;

[0152] The speed ratio , is the rated speed, ~ is the model parameter, then the energy consumption of the chilled water pump is Converted into a relationship that is only related to the speed ratio and the chilled water pump flow rate:

[0153] ;

[0154] Energy consumption of fan coil units and fresh air units The mathematical model is as follows:

[0155]

[0156] Where, is the air flow rate, is the air pressure, For variable speed fans, medium efficiency;

[0157] Air volume per circuit Is the air volume supplied to all air-conditioned rooms The relationship between the total air volume and the air volume of each air conditioning unit is determined by the following formula:

[0158] ;

[0159] When the speed of the variable speed fan is n, the pressure and air volume satisfy the following curve equation relationship:

[0160] ;

[0161] The speed ratio , is the rated speed, ~ is the model parameter, then the energy consumption of fan coil unit and fresh air unit is Converted into a relationship that is only related to the speed ratio and the air volume of each fan coil unit:

[0162]

[0163] In the above formula, Represents a constant whose value varies depending on the dimensional unit of the total pressure or flow rate. If the dimensional units of the air flow rate and the water flow rate are the same, and the dimensional units of the pump head and the air pressure are the same, then for the above formulas It will be the same;

[0164] Cooling pump energy consumption and cooling tower energy consumption The mathematical model is expressed as follows:

[0165] ;

[0166] ;

[0167] ;

[0168] ;

[0169] In the system modeling process, parameters , , , , , The value of can be obtained from the sample data of the corresponding manufacturer. The cooling load at the end of each air conditioning circuit Available from DDC Systems.

[0170] The constraints of the power control model of the central air conditioning system in Step 10 above are:

[0171] The range of change of the inlet water temperature and outlet water temperature of the chiller is as follows:

[0172] ;

[0173] ;

[0174] Chilled water outlet temperature Cannot be too low, otherwise it will cause ice and blockage of the unit; chilled water outlet temperature The upper limit of the cooling water return temperature is mainly determined by two factors: whether the maximum temperature can meet the cooling load requirements of the air-conditioning terminal; and the comfort requirements of the air-conditioning place. Generally speaking, its lower limit is the wet-bulb temperature of the outdoor environment, which is also the minimum limit for realizing the heat conduction process; its upper limit is mainly from the perspective of product manufacturers, to ensure that the pressure is maintained within a certain acceptable range while ensuring the safety standard requirements to achieve safe and reliable operation of the equipment.

[0175] The water flow through the chiller pump and cooling pump is subject to the following constraints:

[0176] ;

[0177] ;

[0178] If the water flow rate is too low, the cooling load of the chilled water and cooling water circuits will not be met promptly. Rising water temperatures will exacerbate the impact of the aforementioned constraints. Therefore, the minimum flow rate setting should ensure that the water temperatures in the chilled water and cooling water circuits remain within normal ranges. The maximum water flow rate is primarily limited by the motor capacity of the pump.

[0179] The air volume of the air conditioning unit and cooling tower is subject to the following constraints:

[0180] ;

[0181] ;

[0182] The minimum airflow for air conditioning units and cooling towers is primarily determined by the end user or operator. Otherwise, the load will not be met promptly. Rising water temperatures will exacerbate the impact of the aforementioned constraints. Similar to constraint 2, the maximum airflow is primarily limited by the fan motor capacity.

[0183] The pressure that the air conditioning unit and refrigeration pump can provide must be within the operating range of the variable speed control:

[0184] ;

[0185] ;

[0186] Theoretically, the lowest pressure limit that the air conditioning unit and chiller pump can provide occurs when the variable speed control is operating at the minimum constraint and the hypothetical losses through the ventilation ducts and water pump network are at their maximum. The maximum pressure limit occurs when the variable speed control is operating at full speed and the dampers and valves in the network are open.

[0187] The range of variation of fresh air volume for each fan coil unit:

[0188] .

[0189] In the above Step 11, the total power value of the central control system predicted in Step 7 is compared with the actual total power value to obtain the difference between multiple time intervals. The multiple differences are introduced into the power model of the central air-conditioning system obtained in Step 10, and the differences are multiplied by the set weights and added to the power model of the central air-conditioning system to obtain the corrected power model of the central air-conditioning system. The prediction is then repeated until the difference is less than the set threshold.

[0190] After obtaining the above control model, a stable basic power part can be provided for the building to use based on the prediction, which can meet the basic needs of the building. In case of emergency, such as drastic changes in weather conditions causing rapid changes in demand, the power can be increased or decreased based on the basic power, and fuzzy PID can be used for adjustment.

[0191] The above embodiments are merely preferred technical solutions of the present invention and should not be construed as limiting the present invention. The embodiments and features in the embodiments of this application may be arbitrarily combined with each other unless they conflict. The scope of protection of the present invention shall be the technical solutions described in the claims, including equivalent alternatives to the technical features of the technical solutions described in the claims. Equivalent alternatives and improvements within this scope are also within the scope of protection of the present invention.

Claims

1. An energy-saving control algorithm for a central air-conditioning energy control system, characterized in that: The following steps are involved: Step 1: Partition the target building space based on its structure, including division based on height, location, and lighting; Step 2: Set up a weather monitoring station at the target building to monitor the weather parameters in the target building area and save the corresponding data; Step 3: Install light intensity sensors in the areas outside the target building where light can reach the target building area. Install temperature and humidity sensors in each area divided in Step 1. Establish communication between each sensor and the central air conditioning control system. Record and store the historical temperature control status of each area and the total power of the central control system. Step 4: Preprocess the historical meteorological data in Step 2 and the corresponding historical data of the total power of the central air-conditioning system in Step 3; Step 5: Perform correlation analysis on the historical meteorological data preprocessed in Step 4 and the historical data of the total power of the central air-conditioning system, extract meteorological features whose impact on the total power of the central air-conditioning system is greater than the set value, and eliminate invalid meteorological features. Step 6: Decompose the historical data of the total power of the central air-conditioning system into different intrinsic mode functions (IMFs) using variational mode decomposition, and reconstruct all IMFs into high-frequency components and low-frequency components by calculating the approximate entropy (AE) of each IMF. Step 7, using a gated neural network to predict the total power of the central air-conditioning system at a future time point based on the historical meteorological data obtained in Step 5 and the historical high-frequency power components and historical low-frequency power components in Step 6; Step 8: Establish a heat transfer model for each partitioned space based on the target building partitioning in Step 1. Step 9: Establish a power control model for the central air-conditioning system; establish the energy consumption model of each component through the physical model of the central air-conditioning system; Step 10: Establish the constraints of each component of the central air-conditioning system and combine them with the heat transfer of each partition in Step 8 to obtain the actual power model of the central air-conditioning system. Step 11: Compare the difference between the total power value of the central control system predicted in Step 7 and the actual total power value, and introduce the power model of the central air-conditioning system obtained in Step 10 as a reference for correction to obtain a corrected total power prediction model of the central control system; Step 12: Based on the power prediction value obtained from the revised total power prediction model of the central control system, each component of the central air-conditioning system is controlled in combination with the constraints of the components.

2. According to the energy-saving control algorithm of a central air-conditioning energy control system described in claim 1, it is characterized in that: The meteorological parameters in Step 2 include light intensity, light direction, air pressure and temperature.

3. The energy-saving control algorithm of a central air-conditioning energy control system according to claim 2 is characterized in that: The preprocessing of Step 4 includes filling missing values ​​in the collected historical meteorological data and the historical data of the total power of the central air-conditioning system with the mean of the records in the set time interval before and after. For outliers, the isolation forest algorithm is used to detect and eliminate them according to the following steps: The isolation forest algorithm continuously segments the data samples until each sample space contains only one type of data point. The segmentation process is as follows: For the data set X, a binary tree T with N nodes is used to describe it, and each node or The data of is a subset of X, where i represents the number of layers in the tree, j represents the jth node in the previous layer, and r and l are used to distinguish the right and left nodes in the same layer; For the dataset contained in a certain layer , randomly select the sample attribute q and its value range space value p to divide and , corresponding to the node set and , data less than or equal to p is divided into Node, others are divided into Node, where j* represents the j*th node in the i+1th layer; When the following situation occurs, a complete binary tree is obtained and the partitioning is completed: A. The depth of the data tree reaches the set maximum value; B. Node Contains only one data point or the data points contained are identical.

4. The energy-saving control algorithm for a central air-conditioning energy control system according to claim 3 is characterized in that: In Step 5, the Pearson correlation coefficient is used to analyze the correlation.

5. The energy-saving control algorithm of a central air-conditioning energy control system according to claim 4 is characterized in that: The heat transfer model of a single space in each partition in Step 8 is as follows: = ； ； ； ； ； Where, i represents the moment, C is the heat capacity of the building space, is the dimensionless distribution coefficient of the air conditioning cooling output in space: ; F is the thermal diffusivity, t is the cooling time, and S is the space area; It is the space between the object space and its neighbors. i The thermal resistance between To open space from the object space to its neighbors k The thermal induction between For space i The temperature of the moment, For k and k 1 The net heat gain during the time between is the length of the time interval, is the thermal resistance between the target space and the outdoor air, M is the number of spaces that communicate with the target space, It is an adjacent space that communicates with the object space k exist i Relative humidity at the time, It is an adjacent space that communicates with the object space k exist i Net heat gain at any given moment.

6. The energy-saving control algorithm of a central air-conditioning energy control system according to claim 5 is characterized in that: The objective function of the power control model of the central air-conditioning system in Step 9 is: ； Energy consumption in central air conditioning systems Energy consumption of fan coil units and fresh air units , Chilled water pump energy consumption , refrigeration unit energy consumption , cooling water pump energy consumption and cooling tower energy consumption composition.

7. The energy-saving control algorithm for a central air-conditioning energy control system according to claim 6 is characterized in that: The fan coil unit and fresh air unit energy consumption It is composed of multiple chillers, and its mathematical model is: ； in: ； ； The total cooling load of the system is the sum of the cooling loads of each room, and the total cooling load is distributed to each chiller in proportion; the actual cooling capacity of each unit, the number of units, and the unit capacity meet the following formula: ; In the above formula, is the total energy consumption of multiple chillers, For the i The cooling capacity of the unit, For the i The efficiency of a unit at a certain load rate, For the i The load factor of the unit, For the i The temperature regulation coefficient of each unit, For the i The actual cooling capacity of a unit under a certain load, is the actual cooling load of a circuit in the building, is the chilled water supply temperature of the chiller, is the return water temperature of the chiller cooling water, is the number of refrigeration units, is the number of fans, is the number of air conditioners; The mathematical model of energy consumption of chilled water pump is as follows: ； Where, is the total efficiency of the variable speed pump, which is affected by the water flow rate and the pump head; N2 is the number of chilled water pumps; The water flow of the chilled water pump is the sum of the flow rates of the chilled water flowing through each circuit. The distribution relationship between the total chilled water flow and the flow rates of each chilled water pump is determined by the following formula: ； When the speed of the variable speed water pump is n, the pressure and flow rate satisfy the following curve equation relationship: ； The speed ratio , is the rated speed, ~ is the model parameter, then the energy consumption of the chilled water pump is Converted into a relationship that is only related to the speed ratio and the chilled water pump flow rate: ； Energy consumption of fan coil units and fresh air units The mathematical model is as follows: Where, is the air flow rate, is the air pressure, For variable speed fans, medium efficiency; Air volume per circuit Is the air volume supplied to all air-conditioned rooms The relationship between the total air volume and the air volume of each air conditioning unit is determined by the following formula: ； When the speed of the variable speed fan is n, the pressure and air volume satisfy the following curve equation relationship: ； The speed ratio , is the rated speed, ~ is the model parameter, then the energy consumption of fan coil unit and fresh air unit is Converted into a relationship that is only related to the speed ratio and the air volume of each fan coil unit: In the above formula, Represents a constant whose value varies depending on the dimensional unit of the total pressure or flow rate. If the dimensional units of the air flow rate and the water flow rate are the same, and the dimensional units of the pump head and the air pressure are the same, then for the above formulas It will be the same; Cooling pump energy consumption and cooling tower energy consumption The mathematical model is expressed as follows: ； ； ； ； In the system modeling process, parameters , , , , , The value of can be obtained from the sample data of the corresponding manufacturer. The cooling load at the end of each air conditioning circuit Available from DDC Systems.

8. The energy-saving control algorithm for a central air-conditioning energy control system according to claim 7 is characterized in that: The constraints of the power control model of the central air conditioning system in Step 10 are: The range of change of the inlet water temperature and outlet water temperature of the chiller is as follows: ； ； The water flow through the chiller pump and cooling pump is subject to the following constraints: ； ； The air volume of the air conditioning unit and cooling tower is subject to the following constraints: ； ； The pressure that the air conditioning unit and refrigeration pump can provide must be within the operating range of the variable speed control: ； ； The range of variation of fresh air volume of each fan coil unit: 。 9. The energy-saving control algorithm for a central air-conditioning energy control system according to claim 8 is characterized in that: In the Step 11, the total power value of the central control system predicted in Step 7 is compared with the actual total power value to obtain the difference between multiple time intervals. The multiple differences are introduced into the power model of the central air-conditioning system obtained in Step 10, and the differences are multiplied by the set weights and added to the power model of the central air-conditioning system to obtain a corrected power model of the central air-conditioning system. The prediction is then repeated until the difference is less than the set threshold.

10. The energy-saving control algorithm of a central air-conditioning energy control system according to claim 9 is characterized in that: In the aforementioned Step 1, the area that can receive light is distinguished from the area that cannot receive light.

Images