Machine learning based central air conditioning operation energy efficiency simulation system and method

Abstract

The application relates to the technical field of energy efficiency simulation, and particularly discloses a central air conditioner operation energy efficiency simulation system and method based on machine learning, which comprises the following steps: S1: acquiring annual hourly meteorological data, auxiliary equipment comprehensive power coefficient and cooling equipment performance parameters, and fitting a load rate polynomial; S2: for each hour, inputting a total cooling load, estimating a cooling equipment average energy efficiency coefficient through the load rate polynomial, calculating a condensing heat load, and combining meteorological data and equipment inherent parameters to solve a cooling equipment outlet water temperature; S3: enumerating all cooling equipment start-stop combinations; for each hour, establishing a nonlinear objective function with the maximum instantaneous system total energy efficiency as an objective, and solving an optimal load distribution scheme under each start-stop combination; S4: calculating the instantaneous total energy efficiency of each start-stop combination under the optimal load distribution, selecting a combination corresponding to a maximum value as an optimal combination, and finally calculating the system total energy efficiency in a whole year.

Description

Technical Field

[0001] This invention relates to the field of energy efficiency simulation technology, specifically to a central air conditioning operation energy efficiency simulation system and method based on machine learning. Background Technology

[0002] With the continuous expansion of modern building scale and increasing complexity of functions, the energy efficiency of central air conditioning systems, as an indispensable major energy consumer in large buildings, has received increasing attention.

[0003] In existing technologies, energy efficiency simulation of central air conditioning systems is an important research method. It aims to analyze energy consumption by simulating the system's operating state, thereby optimizing system configuration and operating strategies to achieve energy conservation and consumption reduction. However, current central air conditioning energy efficiency simulation faces many challenges. Because central air conditioning systems contain numerous internal devices, including cooling equipment, water pumps, and fan coil units, these devices are interconnected and interact with each other, forming a complex nonlinear system. To perform a comprehensive and accurate simulation of all devices requires not only substantial computational resources but also involves a lengthy, costly, and inefficient simulation process. Summary of the Invention

[0004] The purpose of this invention is to provide a central air conditioning operation energy efficiency simulation system and method based on machine learning, and to solve the following technical problems.

[0005] The objective of this invention can be achieved through the following technical solutions: A machine learning-based method for simulating the energy efficiency of central air conditioning systems includes the following steps: S1: Obtain hourly meteorological data for the location of the central air conditioning unit within one year, including the hourly wet-bulb temperature; obtain the comprehensive power coefficient of the auxiliary equipment, and obtain the performance data and inherent parameters of each cooling device of the central air conditioning unit; obtain the load rate polynomial of the cooling device based on the performance data. S2: For the i-th hour, obtain the total cooling load of the central air conditioning system for the i-th hour; obtain the average energy efficiency coefficient of the cooling equipment according to the load rate polynomial, and obtain the condensing heat load of the cooling equipment according to the total cooling load and the average energy efficiency coefficient; obtain the outlet water temperature of the cooling equipment for the i-th hour based on the wet-bulb temperature, condensing heat load and inherent parameters. S3: Divide all cooling equipment into several start-stop combinations. For the i-th hour, define the formula for generating the instantaneous total energy efficiency of the central air conditioning in the i-th hour according to the comprehensive power coefficient. Iterate through all start-stop combinations, aim to maximize the instantaneous total energy efficiency, define constraints, construct a nonlinear objective function, solve the nonlinear objective function based on the sequential quadratic programming algorithm, and obtain the load sequence of each start-stop combination. S4: For any start-stop combination, obtain the instantaneous total energy efficiency of the start-stop combination when executing each load sequence, select the start-stop combination with the largest instantaneous total energy efficiency, record it as the optimal combination, and obtain the total energy efficiency of the central air conditioning system when it runs in the optimal combination for one year.

[0006] As a further aspect of the present invention: the process of obtaining the comprehensive power coefficient of the auxiliary equipment includes: Historical operating data of the central air conditioning system is obtained, including the historical total cooling load of the central air conditioning system for each hour within a year and the historical total power of auxiliary equipment. A coordinate system is established with the total cooling load as the abscissa and the total power as the ordinate, and each historical total cooling load and its corresponding historical total power are converted into coordinate points in the coordinate system. Linear regression fitting is performed on each coordinate point based on the least squares method to obtain a linear regression equation. The slope of the linear regression equation is obtained and recorded as the comprehensive power coefficient of the auxiliary equipment.

[0007] As a further aspect of the present invention: the inherent parameters include cooling water temperature difference, design approximation degree and rated heat dissipation, wherein the cooling water temperature difference is the difference between the inlet water temperature and the outlet water temperature of the cooling equipment preset by the manufacturer of the central air conditioning system, the design approximation degree is the difference between the outlet water temperature of the cooling equipment preset by the manufacturer and the preset wet-bulb temperature, and the preset wet-bulb temperature is 38°C.

[0008] As a further aspect of the present invention: the process of obtaining the load rate polynomial of the cooling equipment based on the performance data includes: The performance data includes several historical data pairs, which are the load rate and energy efficiency coefficient of the cooling equipment of the central air conditioning system during one hour of historical operation. A quadratic polynomial is set, and based on the least squares method, the load rate is used as the independent variable and the energy efficiency coefficient is used as the dependent variable. All historical data are substituted into the quadratic polynomial for fitting, and the load rate polynomial of the cooling equipment is output.

[0009] As a further aspect of the present invention: the process of obtaining the total cooling load of the central air conditioning system in the i-th hour includes: The geometric parameters, building envelope parameters, and internal heat source parameters of the building where the central air conditioning is located are obtained. The building envelope parameters are the thermal parameters of the building envelope, which specifically include the insulation strength of the building envelope, the heat transfer coefficient of the building envelope, and the solar radiation transmittance of the doors and windows. The internal heat source parameters include the personnel density, average heat dissipation of personnel, and equipment power density for each hour. The equipment power density is the electrical power consumed per square meter of all equipment that generates heat in the building. Based on the building envelope parameters, the heat transfer and heat gain Qwh of the building envelope is obtained. i =K×A×(Tout i-Tin), where K is the heat transfer coefficient of the building envelope, A is the total area of ​​the building envelope, and Tout i Let be the outdoor temperature at hour i, and Tin be the set indoor target temperature; and obtain the solar radiation heat gain Qsun. i =I i ×A 窗 ×SC×η, where I i Let A be the solar radiation intensity in the i-th hour. 窗 denoted as , where SC is the solar radiation transmittance of the doors and windows, and η is the coefficient by which radiant heat is converted into indoor heat gain. The heat gain of the internal heat source is obtained based on the internal heat source parameters. i =ρpeo i ×A peo ×Q peo +ρ 设备 ×A 设备 ×W, where ρpeo i Let A be the population density in the i-th hour. peo Q represents the area occupied by all personnel. peo For average heat dissipation from personnel, ρ 设备 Let W be the power density of the equipment, and W be the operating coefficient. When the equipment is running in the i-th hour, W=1; when the equipment is not running in the i-th hour, W=0. Based on the heat transfer and gain of the building envelope, solar radiation, and internal heat sources, a heat balance equation is established: Where t is one hour, Qcold i Let be the total cooling load of the central air conditioning system in the i-th hour; the heat balance equation is solved iteratively to obtain the total cooling load of the central air conditioning system in the i-th hour.

[0010] As a further aspect of the present invention: the outlet water temperature of the cooling device in the i-th hour Twb i Let T be the wet-bulb temperature in the i-th hour. c For the cooling water temperature difference, T base To achieve a high degree of approximation, Qrej rate This is the rated heat dissipation.

[0011] As a further aspect of the present invention: the formula for generating the instantaneous total energy efficiency of the central air conditioning system in the i-th hour is as follows: Pch s Let represent the real-time power consumption of the s-th cooling device, where s is the index and s∈[1,N], and , where CAP s For the rated cooling capacity of the s-th cooling device, PLR s For the load rate of the s-th cooling device, COP sLet F(Tow) be the energy efficiency coefficient of the s-th cooling device. i ) is the correction function; where Paux i Let Paux represent the total power consumption of the auxiliary devices in the i-th hour, and Paux i =α×Qclod i , where α is the overall power coefficient of the auxiliary equipment.

[0012] As a further aspect of the present invention: the process of defining constraints includes, Where PLRmin is the minimum load rate within the load rate range, and PLRmax is the maximum load rate within the load rate range.

[0013] A machine learning-based central air conditioning energy efficiency simulation system includes: Data input module: acquires hourly meteorological data of the location of the central air conditioning unit within one year, including the wet-bulb temperature of each hour; acquires the comprehensive power coefficient of auxiliary equipment, and acquires the performance data and inherent parameters of each cooling device of the central air conditioning unit, and obtains the load rate polynomial of the cooling device based on the performance data; Hourly simulation module: For the i-th hour, obtain the total cooling load of the central air conditioning system for the i-th hour; obtain the average energy efficiency coefficient of the cooling equipment according to the load rate polynomial, and obtain the condensing heat load of the cooling equipment according to the total cooling load and the average energy efficiency coefficient; obtain the outlet water temperature of the cooling equipment for the i-th hour based on the wet-bulb temperature, condensing heat load and inherent parameters. All cooling equipment is divided into several start-stop combinations. For the i-th hour, the instantaneous total energy efficiency of the central air conditioning in the i-th hour is generated according to the comprehensive power coefficient. All start-stop combinations are traversed to maximize the instantaneous total energy efficiency. Constraints are defined, a nonlinear objective function is constructed, and the nonlinear objective function is solved based on the sequential quadratic programming algorithm to obtain the load sequence of each start-stop combination. Energy efficiency simulation module: For any start-stop combination, obtain the instantaneous total energy efficiency of the start-stop combination when executing each load sequence, select the start-stop combination with the largest instantaneous total energy efficiency, record it as the optimal combination, and obtain the total energy efficiency of the central air conditioning system when it runs in the optimal combination for one year.

[0014] The beneficial effects of this invention are: This invention significantly improves the accuracy of system state simulation by establishing a correlation polynomial between cooling equipment load rate and energy efficiency coefficient, and introducing a dynamic outlet water temperature calculation model based on actual condensing heat load, making the energy efficiency assessment results closer to real operating conditions. Furthermore, by enumerating all possible equipment start-stop combinations and using a sequential quadratic programming algorithm to solve a nonlinear optimization problem aimed at maximizing the system's instantaneous total energy efficiency every hour, it can automatically find the theoretically optimal operating strategy, overcoming the limitations of manual experience or fixed rule strategies, and providing precise guidance for energy-saving system operation. Simultaneously, by dividing all central air conditioning equipment into cooling equipment and auxiliary equipment, it focuses on simulating the high-power cooling equipment while making a general estimate of the power consumption of auxiliary equipment, reducing the problem of low simulation efficiency caused by the complexity and diversity of equipment. Attached Figure Description

[0015] The invention will now be further described with reference to the accompanying drawings.

[0016] Figure 1 This is a schematic diagram of the central air conditioning operation energy efficiency simulation system and method based on machine learning according to the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1 As shown, this invention is a machine learning-based method for simulating the energy efficiency of central air conditioning operation, comprising the following steps: S1: Obtain hourly meteorological data for the location of the central air conditioning unit within one year, including the hourly wet-bulb temperature; obtain the comprehensive power coefficient of the auxiliary equipment, and obtain the performance data and inherent parameters of each cooling device of the central air conditioning unit; obtain the load rate polynomial of the cooling device based on the performance data. In a preferred embodiment of the present invention, the cooling equipment includes a cooling tower and a cooling unit, and the auxiliary equipment is all the equipment in the central air conditioning system except for the cooling equipment, including fan coil units, combined air conditioning boxes and air valves; In a preferred embodiment of the present invention, the process of obtaining the comprehensive power coefficient of the auxiliary equipment includes: Historical operating data of the central air conditioning system is obtained, including the historical total cooling load of the central air conditioning system for each hour within a year, and the historical total power of auxiliary equipment. A coordinate system is established with the total cooling load as the abscissa and the total power as the ordinate, and each historical total cooling load and its corresponding historical total power are converted into coordinate points in the coordinate system. Linear regression fitting is performed on each coordinate point based on the least squares method to obtain a linear regression equation. The slope of the linear regression equation is obtained and denoted as the comprehensive power coefficient of the auxiliary equipment. In a preferred embodiment of the present invention, the inherent parameters include cooling water temperature difference, design approximation degree, and rated heat dissipation, wherein the cooling water temperature difference is the difference between the inlet water temperature and the outlet water temperature of the cooling equipment preset by the manufacturer of the central air conditioning system, and the design approximation degree is the difference between the outlet water temperature of the cooling equipment preset by the manufacturer and the preset wet-bulb temperature, wherein the preset wet-bulb temperature is 38°C. In a preferred embodiment of the present invention, the process of obtaining the load rate polynomial of the cooling device based on the performance data includes: The performance data includes several historical data pairs, which are the load rate and energy efficiency coefficient of the cooling equipment of the central air conditioning system during one hour of historical operation. A quadratic polynomial is set, and based on the least squares method, the load rate is used as the independent variable and the energy efficiency coefficient is used as the dependent variable. All historical data are substituted into the quadratic polynomial for fitting, and the load rate polynomial of the cooling equipment is output. S2: For the i-th hour, obtain the total cooling load of the central air conditioning system for the i-th hour; obtain the average energy efficiency coefficient of the cooling equipment according to the load rate polynomial, and obtain the condensing heat load of the cooling equipment for the i-th hour according to the total cooling load and the average energy efficiency coefficient; obtain the outlet water temperature of the cooling equipment for the i-th hour based on the wet-bulb temperature, condensing heat load and inherent parameters. In a preferred embodiment of the present invention, the process of obtaining the total cooling load of the central air conditioning system in the i-th hour includes: The geometric parameters, building envelope parameters, and internal heat source parameters of the building where the central air conditioning is located are obtained. The building envelope parameters are the thermal parameters of the building envelope, which specifically include the insulation strength of the building envelope, the heat transfer coefficient of the building envelope, and the solar radiation transmittance of the doors and windows. The internal heat source parameters include the personnel density, average heat dissipation of personnel, and equipment power density for each hour. The equipment power density is the electrical power consumed per square meter of all equipment that generates heat in the building. Based on the building envelope parameters, the heat transfer and heat gain Qwh of the building envelope is obtained. i =K×A×(Tout i -Tin), where K is the heat transfer coefficient of the building envelope, A is the total area of ​​the building envelope, and Tout iLet be the outdoor temperature at hour i, and Tin be the set indoor target temperature; and obtain the solar radiation heat gain Qsun. i =I i ×A 窗 ×SC×η, where I i Let A be the solar radiation intensity in the i-th hour. 窗 denoted as , where SC is the solar radiation transmittance of the doors and windows, and η is the coefficient by which radiant heat is converted into indoor heat gain. The heat gain of the internal heat source is obtained based on the internal heat source parameters. i =ρpeo i ×A peo ×Q peo +ρ 设备 ×A 设备 ×W, where ρpeo i Let A be the population density in the i-th hour. peo Q represents the area occupied by all personnel. peo For average heat dissipation from personnel, ρ 设备 Let W be the power density of the equipment, and W be the operating coefficient. When the equipment is running in the i-th hour, W=1; when the equipment is not running in the i-th hour, W=0. Based on the heat transfer and gain of the building envelope, solar radiation, and internal heat sources, a heat balance equation is established: Where t is one hour, Qcold i Let be the total cooling load of the central air conditioning system in the i-th hour; the heat balance equation is solved iteratively to obtain the total cooling load of the central air conditioning system in the i-th hour; The process of obtaining the power density of the device includes: Obtain the total area occupied by all equipment within the building, and obtain the electrical power consumed by all equipment in one hour. Divide the electrical power by the total area occupied to obtain the equipment power density. It is worth noting that all the equipment that generates heat in the building mentioned in this invention is mainly lighting equipment; In a preferred embodiment of the present invention, the process of obtaining the average energy efficiency coefficient of the cooling device includes: Define the load rate range for the cooling equipment, and select several load rate sampling points {PLR1, PLR2, ..., PLR} at equal intervals within the load rate range. n}, where PLR n Let n represent the nth load rate sampling point, where n is the total number of selected load rate sampling points; substitute all load rate sampling points into the load rate polynomial to obtain the energy efficiency coefficient sampling points corresponding to each load rate sampling point, and obtain the average value of all energy efficiency coefficient sampling points, which is denoted as the average energy efficiency coefficient. In a preferred embodiment of the present invention, the condensation heat load of the cooling equipment in the i-th hour is... COP avg The average energy efficiency coefficient; In a preferred embodiment of the present invention, the outlet water temperature of the cooling device in the i-th hour is... Twb i Let T be the wet-bulb temperature in the i-th hour. c For the cooling water temperature difference, T base To achieve a high degree of approximation, Qrej rate This is the rated heat dissipation. It should be noted that, This refers to the process of correcting the performance of a cooling equipment based on the ratio of its condensing heat load to its rated heat dissipation in the i-th hour. e is a preset empirical correction index, typically set to 0.6, describing the nonlinear relationship between the actual heat dissipation performance of the cooling equipment and the ratio. (T) base -T c ) is a negative constant, representing the portion of temperature rise that cannot achieve the ideal cooling effect due to the existence of cooling limits at the design point; It is a coefficient between 0 and 1, used to correct the temperature rise at the design point to the rise under the current actual operating load rate; the lower the load, the smaller this coefficient is, but it is usually not zero, which simulates the phenomenon of reduced efficiency of the cooling tower under low load. S3: Divide all cooling equipment into several start-stop combinations. For the i-th hour, define the formula for generating the instantaneous total energy efficiency of the central air conditioning in the i-th hour according to the comprehensive power coefficient. Iterate through all start-stop combinations, aim to maximize the instantaneous total energy efficiency, define constraints, construct a nonlinear objective function, solve the nonlinear objective function based on the sequential quadratic programming algorithm, and obtain the load sequence of each start-stop combination. In a preferred embodiment of the present invention, the process of dividing all cooling equipment into several start-stop combinations includes: Retrieve all cooling devices within the central air conditioning system and assign a binary variable x to each device. k Where k=1 when the cooling equipment is started and k=0 when the cooling equipment is stopped; arrange the binary variables of all cooling equipment in order to form an N-bit binary number, where N is the total number of cooling equipment; since each binary variable has two choices, we get 2 N Each binary number is represented as a start / stop combination, excluding start / stop combinations that completely stop the cooling equipment; In a preferred embodiment of the present invention, the formula for generating the instantaneous total energy efficiency of the central air conditioning system in the i-th hour is: Pch sLet represent the real-time power consumption of the s-th cooling device, where s is the index and s∈[1,N], and , where CAP s For the rated cooling capacity of the s-th cooling device, PLR s For the load rate of the s-th cooling device, COP s Let F(Tow) be the energy efficiency coefficient of the s-th cooling device. i ) is the correction function; where Paux i Let Paux represent the total power consumption of the auxiliary devices in the i-th hour, and Paux i =α×Qclod i , where α is the overall power coefficient of the auxiliary equipment; It should be noted that the rated cooling capacity CAP s The maximum cooling capacity of the cooling equipment under standard design conditions, obtained from the manufacturer; the correction function F(Tow) i The correction function is used to adjust for the impact of outlet water temperature on the energy efficiency of cooling equipment. The efficiency of cooling equipment is closely related to the condensing temperature, which in turn depends primarily on the outlet water temperature. When the actual outlet water temperature deviates from the unit's rated design temperature, the correction function affects the theoretical coefficient of performance (COP). s Scaling; Tow i When it decreases, F(Tow) i )>1, that is, F(Tow) i The value of Tow increases. i When it increases, F(Tow) i ) < 1, i.e., F(Tow) i The value of ) decreases; Understandably, based on the ratio of the actual operating power of the auxiliary equipment to the output cooling and heating load of the air conditioning system entered in the basic settings module, the real-time operating power of the auxiliary equipment is calculated every hour. Considering that the energy consumption of the auxiliary equipment is closely related to the operating status of the chiller unit, this real-time calculation method can accurately reflect the changes in energy consumption of the auxiliary equipment during the entire operation of the air conditioning system. In a preferred embodiment of the present invention, the process of defining the constraint conditions includes: Where PLRmin is the minimum load rate within the load rate range, and PLRmax is the maximum load rate within the load rate range; In a preferred embodiment of the present invention, the process of solving the nonlinear objective function based on the sequential quadratic programming algorithm includes: To maximize the instantaneous total energy efficiency, we obtain the mathematical form. , And based on the aforementioned constraints, equality constraints are obtained. and inequality constraints , Construct the current Lagrange function Where j is the index, representing the j-th inequality constraint; and the gradient of the Lagrange function is calculated. Hessian matrix ; In x k At this point, the nonlinear objective function is approximated as a quadratic programming problem: , where j∈[1,2N], d is the search direction, and standard algorithms such as the effective set method or interior point method are called to solve this quadratic programming subproblem to obtain the search direction and the corresponding Lagrange multiplier estimate; Perform a one-dimensional search along the search direction, determine the step size ε, and update the iteration point x. k+1 =x k +εd, and update the Hessian matrix; repeat the above steps until the convergence condition is met, at which point the final iteration point is the load sequence of the start-stop combination; In a preferred embodiment of the present invention, the load sequence is the load rate of all cooling devices in the start-stop combination; Understandably, based on the hourly cooling tower outlet water temperature and demand load calculated in the first two steps throughout the year, as well as the energy efficiency curve data of the selected chiller units, an intelligent algorithm generates all possible combinations of chiller unit start-up strategies for each hour's operating conditions. For example, for a project with multiple chiller units, various combinations are considered, including single-unit operation, parallel or series operation of multiple units, and different load distribution conditions. Combined with the energy efficiency performance of the chiller units at corresponding cooling water temperatures and load rates, the energy efficiency value of each start-up strategy combination is calculated. Finally, through comparison and optimization, the chiller unit start-up strategy combination with the highest energy efficiency for each hour is determined to maximize the system's operating energy efficiency. S4: For any start-stop combination, obtain the instantaneous total energy efficiency of the start-stop combination when executing each load sequence, select the start-stop combination with the largest instantaneous total energy efficiency, record it as the optimal combination, and obtain the total energy efficiency of the central air conditioning system when it runs in the optimal combination for one year. In a preferred embodiment of the present invention, the process of obtaining the total energy efficiency includes: Get the total power consumption in the i-th hour Then the total energy efficiency is obtained. Where Y is the total number of hours in a year; Furthermore, the proportion of all cooling equipment and auxiliary equipment in the total annual electricity consumption is calculated, and the energy consumption contribution of each device is displayed through visualization methods such as pie charts and bar charts. This helps users clearly understand the distribution structure of system energy consumption, so as to take targeted energy-saving measures and optimize equipment configuration and operation strategies. Based on the statistical analysis results, a series of rich charts are automatically generated, including but not limited to curves showing energy efficiency changes over time (in hours or months), pie charts showing the electricity consumption of each device, and histograms showing the distribution of chiller unit operation strategies. These charts present the system's operating performance and energy consumption characteristics in an intuitive and visual way, making complex data information easier to understand and interpret, and providing users with a comprehensive and intuitive overview of the project's operation. All analysis results, statistical data, and generated charts are integrated to produce a detailed and complete simulation report. The report covers basic project information, the data and rules on which the calculations were based, detailed calculation results, and related analyses and charts. The report adopts a structured layout and clear text descriptions, making it convenient for users to view and evaluate the energy efficiency of the project's central air conditioning system, and providing strong support and reference for project decision-making, optimization design, and operation management.

[0019] A machine learning-based central air conditioning energy efficiency simulation system includes: Data input module: acquires hourly meteorological data of the location of the central air conditioning unit within one year, including the wet-bulb temperature of each hour; acquires the comprehensive power coefficient of auxiliary equipment, and acquires the performance data and inherent parameters of each cooling device of the central air conditioning unit, and obtains the load rate polynomial of the cooling device based on the performance data; Hourly simulation module: For the i-th hour, obtain the total cooling load of the central air conditioning system for the i-th hour; obtain the average energy efficiency coefficient of the cooling equipment according to the load rate polynomial, and obtain the condensing heat load of the cooling equipment according to the total cooling load and the average energy efficiency coefficient; obtain the outlet water temperature of the cooling equipment for the i-th hour based on the wet-bulb temperature, condensing heat load and inherent parameters. All cooling equipment is divided into several start-stop combinations. For the i-th hour, the instantaneous total energy efficiency of the central air conditioning in the i-th hour is generated according to the comprehensive power coefficient. All start-stop combinations are traversed to maximize the instantaneous total energy efficiency. Constraints are defined, a nonlinear objective function is constructed, and the nonlinear objective function is solved based on the sequential quadratic programming algorithm to obtain the load sequence of each start-stop combination. Energy efficiency simulation module: For any start-stop combination, obtain the instantaneous total energy efficiency of the start-stop combination when executing each load sequence, select the start-stop combination with the largest instantaneous total energy efficiency, record it as the optimal combination, and obtain the total energy efficiency of the central air conditioning system when it runs in the optimal combination for one year.

[0020] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the present invention should still fall within the scope of the present invention.

Claims

1. A machine learning-based method for simulating the energy efficiency of central air conditioning operation, characterized in that, Includes the following steps: S1: Obtain hourly meteorological data for the location of the central air conditioning unit within one year, including the hourly wet-bulb temperature; obtain the comprehensive power coefficient of the auxiliary equipment, and obtain the performance data and inherent parameters of each cooling device of the central air conditioning unit; obtain the load rate polynomial of the cooling device based on the performance data. S2: For the i-th hour, obtain the total cooling load of the central air conditioning system for the i-th hour; obtain the average energy efficiency coefficient of the cooling equipment according to the load rate polynomial, and obtain the condensing heat load of the cooling equipment according to the total cooling load and the average energy efficiency coefficient; obtain the outlet water temperature of the cooling equipment for the i-th hour based on the wet-bulb temperature, condensing heat load and inherent parameters. S3: Divide all cooling equipment into several start-stop combinations. For the i-th hour, define the formula for generating the instantaneous total energy efficiency of the central air conditioning in the i-th hour according to the comprehensive power coefficient. Iterate through all start-stop combinations, aim to maximize the instantaneous total energy efficiency, define constraints, construct a nonlinear objective function, solve the nonlinear objective function based on the sequential quadratic programming algorithm, and obtain the load sequence of each start-stop combination. S4: For any start-stop combination, obtain the instantaneous total energy efficiency of the start-stop combination when executing each load sequence, select the start-stop combination with the largest instantaneous total energy efficiency, record it as the optimal combination, and obtain the total energy efficiency of the central air conditioning system when it runs in the optimal combination for one year.

2. The machine learning-based central air conditioning energy efficiency simulation method according to claim 1, characterized in that, The process of obtaining the overall power coefficient of the auxiliary equipment includes: Historical operating data of the central air conditioning system is obtained, including the historical total cooling load of the central air conditioning system for each hour within a year and the historical total power of auxiliary equipment. A coordinate system is established with the total cooling load as the abscissa and the total power as the ordinate, and each historical total cooling load and its corresponding historical total power are converted into coordinate points in the coordinate system. Linear regression fitting is performed on each coordinate point based on the least squares method to obtain a linear regression equation. The slope of the linear regression equation is obtained and recorded as the comprehensive power coefficient of the auxiliary equipment.

3. The machine learning-based central air conditioning energy efficiency simulation method according to claim 1, characterized in that, The inherent parameters include cooling water temperature difference, design approximation, and rated heat dissipation. The cooling water temperature difference is the difference between the inlet and outlet water temperatures of the cooling equipment preset by the central air conditioning manufacturer. The design approximation is the difference between the outlet water temperature of the cooling equipment preset by the manufacturer and the preset wet-bulb temperature, where the preset wet-bulb temperature is 38°C.

4. The machine learning-based central air conditioning energy efficiency simulation method according to claim 1, characterized in that, The process of obtaining the load rate polynomial of the cooling equipment based on the performance data includes: The performance data includes several historical data pairs, which are the load rate and energy efficiency coefficient of the cooling equipment of the central air conditioning system during one hour of historical operation. A quadratic polynomial is set, and based on the least squares method, the load rate is used as the independent variable and the energy efficiency coefficient is used as the dependent variable. All historical data are substituted into the quadratic polynomial for fitting, and the load rate polynomial of the cooling equipment is output.

5. The machine learning-based central air conditioning energy efficiency simulation method according to claim 1, characterized in that, The process of obtaining the total cooling load of the central air conditioning system in the i-th hour includes: The geometric parameters, building envelope parameters, and internal heat source parameters of the building where the central air conditioning is located are obtained. The building envelope parameters are the thermal parameters of the building envelope, which specifically include the insulation strength of the building envelope, the heat transfer coefficient of the building envelope, and the solar radiation transmittance of the doors and windows. The internal heat source parameters include the personnel density, average heat dissipation of personnel, and equipment power density for each hour. The equipment power density is the electrical power consumed per square meter of all equipment that generates heat in the building. Based on the building envelope parameters, the heat transfer and heat gain Qwh of the building envelope is obtained. i =K×A×(Tout i -Tin), where K is the heat transfer coefficient of the building envelope, A is the total area of ​​the building envelope, and Tout i Let be the outdoor temperature at hour i, and Tin be the set indoor target temperature; and obtain the solar radiation heat gain Qsun. i =I i ×A 窗 ×SC×η, where I i Let A be the solar radiation intensity in the i-th hour. 窗 denoted as , where SC is the solar radiation transmittance of the doors and windows, and η is the coefficient by which radiant heat is converted into indoor heat gain. The heat gain of the internal heat source is obtained based on the internal heat source parameters. i =ρpeo i ×A peo ×Q peo +ρ 设备 ×A 设备 ×W, where ρpeo i Let A be the population density in the i-th hour. peo Q represents the area occupied by all personnel. peo For average heat dissipation from personnel, ρ 设备 Let W be the power density of the equipment, and W be the operating coefficient. When the equipment is running in the i-th hour, W=1; when the equipment is not running in the i-th hour, W=0. Based on the heat transfer and gain of the building envelope, solar radiation, and internal heat sources, a heat balance equation is established: Where t is one hour, Qcold i Let be the total cooling load of the central air conditioning system in the i-th hour; the heat balance equation is solved iteratively to obtain the total cooling load of the central air conditioning system in the i-th hour.

6. The machine learning-based central air conditioning energy efficiency simulation method according to claim 1, characterized in that, The condensing heat load of the cooling equipment in the i-th hour COP avg The average energy efficiency coefficient.

7. The machine learning-based central air conditioning energy efficiency simulation method according to claim 1, characterized in that, The outlet water temperature of the cooling equipment in the i-th hour Twb i Let T be the wet-bulb temperature in the i-th hour. c For the cooling water temperature difference, T base To achieve a high degree of approximation, Qrej rate This is the rated heat dissipation.

8. The machine learning-based central air conditioning energy efficiency simulation method according to claim 1, characterized in that, The formula for calculating the instantaneous total energy efficiency of the central air conditioning system in the i-th hour is: Pch s Let represent the real-time power consumption of the s-th cooling device, where s is the index and s∈[1,N], and , where CAP s For the rated cooling capacity of the s-th cooling device, PLR s For the load rate of the s-th cooling device, COP s Let F(Tow) be the energy efficiency coefficient of the s-th cooling device. i ) is the correction function; where Paux i Let Paux represent the total power consumption of the auxiliary devices in the i-th hour, and Paux i =α×Qclod i , where α is the overall power coefficient of the auxiliary equipment.

9. The machine learning-based central air conditioning energy efficiency simulation method according to claim 1, characterized in that, The process of defining constraints includes, Where PLRmin is the minimum load rate within the load rate range, and PLRmax is the maximum load rate within the load rate range.

10. A central air conditioning operation energy efficiency simulation system based on machine learning, characterized in that, include: Data input module: acquires hourly meteorological data of the location of the central air conditioning unit within one year, including the wet-bulb temperature of each hour; acquires the comprehensive power coefficient of auxiliary equipment, and acquires the performance data and inherent parameters of each cooling device of the central air conditioning unit, and obtains the load rate polynomial of the cooling device based on the performance data; Hourly simulation module: For the i-th hour, obtain the total cooling load of the central air conditioning system for the i-th hour; obtain the average energy efficiency coefficient of the cooling equipment according to the load rate polynomial, and obtain the condensing heat load of the cooling equipment according to the total cooling load and the average energy efficiency coefficient; obtain the outlet water temperature of the cooling equipment for the i-th hour based on the wet-bulb temperature, condensing heat load and inherent parameters. All cooling equipment is divided into several start-stop combinations. For the i-th hour, the instantaneous total energy efficiency of the central air conditioning in the i-th hour is generated according to the comprehensive power coefficient. All start-stop combinations are traversed to maximize the instantaneous total energy efficiency. Constraints are defined, a nonlinear objective function is constructed, and the nonlinear objective function is solved based on the sequential quadratic programming algorithm to obtain the load sequence of each start-stop combination. Energy efficiency simulation module: For any start-stop combination, obtain the instantaneous total energy efficiency of the start-stop combination when executing each load sequence, select the start-stop combination with the largest instantaneous total energy efficiency, record it as the optimal combination, and obtain the total energy efficiency of the central air conditioning system when it runs in the optimal combination for one year.

Images