An energy-saving control strategy for air conditioning and ventilation systems in medical buildings based on climate change characteristics and load forecasting

Abstract

This invention presents an energy-saving control strategy for air conditioning and ventilation systems in medical buildings based on climate change characteristics and load forecasting. It utilizes various sensors and actuators in the sensing layer to collect all raw data. The intelligent control layer on the central management server (including modules for comprehensive building cooling and heating load forecasting, calculation of cooling and heating demand in clean operating areas, dynamic water supply temperature optimization, domain-specific optimization using particle swarm optimization (PSO) algorithms, and constraint verification) performs calculations, optimizations, and decisions. The optimization results are then distributed to the execution layer and the executed equipment layer, ultimately achieving energy-saving control. This invention overcomes the shortcomings of existing hospital building load forecasting technologies, which often employ macroscopic and coarse forecasting models, and the limitations of global optimization particle swarm optimization algorithms, such as slow convergence and long solution times when dealing with large medical buildings with multiple types of cooling and heating sources. It is suitable for large hospital projects in hot-summer and warm-winter regions.

Description

Technical Field

[0001] This invention relates to the field of building energy conservation and HVAC control technology, particularly to the air conditioning and ventilation systems of large medical buildings (including clean operating areas and general areas). It proposes an energy-saving control strategy that comprehensively considers climate change characteristics, load forecasting, and regional differences. This strategy is applied to large hospitals in hot-summer and warm-winter climate zones. By combining four-pipe air-cooled heat pump units and high-efficiency variable-frequency centrifugal chillers as heat source equipment, it intelligently optimizes the control of clean and general areas within the hospital. This achieves the lowest overall annual energy consumption for both the building's air conditioning and domestic hot water systems, while ensuring both usability and stable operation. Background Technology

[0002] Large hospitals in hot-summer, warm-winter regions have distinct internal and external air conditioning system zones due to their large size and depth. The external zone requires cooling in summer and heating in winter, while the internal zone requires cooling year-round. The cleanroom air conditioning system in the hospital needs reheating after dehumidification in summer, thus requiring both cooling and heating. Since a large portion of the cleanroom is located within the building's interior, it requires cooling during transitional seasons and even winter, similar to rooms in the non-cleanroom areas. Additionally, hospitals require heating for their domestic hot water systems year-round. Therefore, the cooling and heating needs of different systems, seasons, and areas within a medical building vary, but the combined effect of these needs means that the building requires both cooling and heating throughout the year. Consequently, the building's air conditioning system must be able to provide both cooling and heating simultaneously. While providing cooling and heating, the air conditioning system also discharges waste heat and cooling energy to the outside. If this waste heat could be recovered and directly supplied to rooms requiring heating or the domestic hot water system during cooling, energy waste could be significantly reduced, achieving energy conservation and emission reduction.

[0003] To meet the year-round cooling and heating needs of large hospital buildings and ensure precise control of indoor temperature and humidity, a coupled cold and heat source system consisting of a four-pipe air-cooled heat pump unit and a chiller unit can be used. Because the four-pipe air-cooled heat pump unit can fully recover and utilize condensation heat, its overall system energy efficiency ratio is higher than that of traditional cold and heat source systems throughout the year.

[0004] However, the use of coupled cold and heat source systems increases the types and data of cold and heat source equipment. In addition, to fully utilize the heat recovery advantages of four-pipe air-cooled heat pump units, the preheating system of domestic hot water heat source is incorporated into the overall scheduling, making the control of the overall cold and heat source system more complex than that of traditional cold and heat source systems.

[0005] Traditional air conditioning system control strategies for medical buildings generally have the following shortcomings:

[0006] Firstly, in terms of hospital building load forecasting, existing technologies mostly employ macroscopic and coarse forecasting models. For example, patent application CN201810168265.0 proposes a building load forecasting method, and invention patent application CN202411577622.0 proposes a ventilation and air conditioning control method based on load forecasting. Both categorize hospitals as "public buildings." The former mainly calculates loads using macroscopic indicators such as annual cumulative cooling load, heating load, and electricity load, while the latter uses multiple linear regression analysis as its modeling method, primarily focusing on cooling load forecasting. However, neither patent application explicitly mentions forecasting loads for clean operating room areas and domestic hot water. The load composition of medical buildings is complex, including not only the loads of regular areas such as general wards and outpatient halls, but also the loads of clean operating rooms with extremely high requirements for temperature, humidity, cleanliness, and pressure differential, as well as the 24 / 7, high-intensity domestic hot water load. Simply using macroscopic indicators for forecasting fails to capture the dynamic changes and interrelationships of cooling and heating demands in hospital buildings, especially those with coupled cooling and heating needs in clean operating room areas. It also fails to cover domestic hot water load, resulting in a lack of refined data for system control. Consequently, the heat recovery function of four-pipe air-cooled heat pump units cannot be fully utilized, leading to significant energy waste. Therefore, establishing a comprehensive forecasting model that fully reflects the complex load characteristics of medical buildings is a primary prerequisite for achieving deep energy-saving control.

[0007] Furthermore, in terms of chiller unit operation control, traditional supply water temperature setting strategies are usually fixed values ​​or based on simple outdoor temperature resets. This static or coarse strategy cannot respond to dynamic changes in building terminal loads in real time. When the building's cooling demand decreases, the system still provides excessively cold chilled water, causing the four-pipe air-cooled heat pump unit or chiller unit to operate at a lower evaporation temperature, resulting in a significant reduction in energy efficiency (COP). When the building's heating demand decreases, the system still provides excessively hot water, causing the four-pipe air-cooled heat pump unit to operate at a higher condensation temperature, resulting in a significant reduction in energy efficiency (COP).

[0008] Furthermore, traditional large-scale air conditioning group control systems often employ experience-based rule control or simple timing control, typically lacking online global optimization for the entire system. While some group control systems introduce Particle Swarm Optimization (PSO) to perform global optimization with the goal of minimizing energy consumption, and can find the optimal operating control strategy under the current conditions with the goal of minimizing system energy consumption, without domain-specific modeling, they suffer from slow convergence and long solution times when dealing with multiple types of cold and heat source equipment (such as chillers and four-pipe air-cooled heat pump units) and cold and heat source systems that simultaneously need to meet cooling and heating demands.

[0009] In summary, existing control methods for air conditioning and ventilation systems in medical buildings suffer from several technical bottlenecks. These include coarse load prediction, failure to include predictions for clean operating room areas and domestic hot water loads, inability of supply water temperature control strategies to respond in real-time to dynamic changes in building terminal loads, and slow convergence and long solution times of the PSO algorithm used in group control technology when dealing with large medical buildings with multiple types of cold and heat source equipment. These limitations prevent them from meeting the ever-increasing energy-saving demands of modern hospitals. This invention aims to fundamentally solve these problems through an integrated and intelligent control method, providing a highly efficient and energy-saving control strategy solution for air conditioning and ventilation systems in medical buildings. Summary of the Invention

[0010] The purpose of this invention is to overcome many problems in existing air conditioning systems for large medical buildings, such as high energy consumption, coarse load prediction in control methods, lack of prediction of clean operating room areas and domestic hot water load, inability of water supply temperature control strategies to respond in real time to dynamic changes in building terminal load, and slow convergence and long solution time of the PSO algorithm used in group control technology when dealing with large medical buildings with multiple types of cold and heat source equipment. For cold and heat source systems with four-pipe air-cooled heat pump units coupled with chillers, this invention proposes an energy-saving control strategy for air conditioning and ventilation systems in medical buildings based on climate change characteristics and load prediction.

[0011] To achieve the above objectives, the technical solution adopted by the present invention is as follows: to construct an energy-saving control strategy for air conditioning and ventilation systems in medical buildings based on climate change characteristics and load forecasting. The perception layer collects all raw data, including physical parameters and operational information, from various sensors and actuators (such as weather stations, temperature and humidity sensors, water valves, and information systems) and inputs them into the intelligent control layer. The intelligent control layer, located on the central management server, performs calculations, optimizations, and decisions, and then sends the optimization results to the execution layer. The intelligent control layer includes modules for predicting overall building heating and cooling loads, calculating cooling and heating demand in clean operating rooms, optimizing dynamic water supply temperature, using particle swarm optimization (PSO) algorithms for domain optimization, and constraint verification. The execution layer receives control commands from the intelligent control layer and converts them into operable electrical signals, which are then transmitted to the controlled equipment layer. The controlled equipment layer contains all major energy-consuming devices in the air conditioning and domestic hot water systems (such as four-pipe air-cooled heat pump units, chillers, water pumps, cooling towers, and air-source heat pump units). These devices receive signals and change their operating states, ultimately achieving energy-saving control. Finally, the operating status data of the controlled equipment is fed back to the perception layer in real time, forming a complete closed-loop control system.

[0012] Specifically, the following steps are included:

[0013] Step 1: Establish a comprehensive building heating and cooling load prediction model. Construct a hierarchical prediction model that includes the heating and cooling load of the general area of ​​the hospital, the indoor heating and cooling load of the clean operating department, and the domestic hot water load. Utilize historical operating load data and real-time meteorological monitoring data to predict the heating and cooling load of different areas and systems of the building hourly.

[0014] Step 2: Calculate the cooling and heating demand of the clean operating department area. Based on the indoor sensible heat load and latent heat load of each room in the clean operating department area predicted in Step a, calculate the cooling and reheat demand of the terminal air conditioning units based on the constant supply air humidity control logic. Then, summarize the total cooling and reheat demand of each room in the clean operating department area, and combine the cooling and heating load of the general area of ​​the hospital and the domestic hot water load predicted in Step a to obtain the overall real-time cooling and heating demand of the hospital.

[0015] Step 3: Perform dynamic water supply temperature optimization, collect feedback signals of the opening of the water valves of the corresponding terminal air conditioning units in the main rooms and temperature and humidity sensors in real time, and dynamically adjust the set value of the chilled / hot water supply temperature of the air conditioning cold and heat source system to improve the unit's operating energy efficiency while meeting the cooling and heating needs. Determine the cooling water inlet temperature of the chiller unit based on real-time meteorological data.

[0016] Step 4: Apply the Particle Swarm Optimization (PSO) algorithm for domain-specific optimization, determine the subdomains for system operation strategy optimization based on the building's cooling and heating needs, call the appropriate PSO sub-algorithm in each optimization subdomain, aim to minimize the total energy consumption of the hospital's cooling and heating source system and the domestic hot water source system, and set comfort level and switching penalty.

[0017] Step 5: Implement multi-unit collaborative control and safety constraints: After the optimization results are checked for constraints, they are sent out for execution. The constraints check includes, but is not limited to, hydraulic feasibility, minimum partial load factor of the unit, minimum start-up and shutdown interval, heat recovery power balance and inlet and outlet water temperature difference, etc.

[0018] Step 6: Continuously collect operating data and user feedback from the control system, analyze system performance, and regularly update the building cooling load model based on new data and feedback to improve prediction accuracy.

[0019] Furthermore, step 1, establishing a comprehensive building heating and cooling load prediction model, includes the following steps:

[0020] Step 1.1: Raw data access, including but not limited to historical operating data, meteorological data, and operational data. Historical operating data affecting the cooling and heating load of the air conditioning system is captured through sensors and data acquisition equipment, including but not limited to indoor ambient temperature, relative humidity, cooling and heating output from the cold and heat source system, domestic hot water consumption, and operating parameters of various equipment in the air conditioning and domestic hot water systems. Real-time meteorological data is collected through a rooftop weather station, including but not limited to outdoor dry-bulb temperature, relative humidity, radiation intensity, and wind speed. Operational data, including but not limited to surgical scheduling, outpatient and emergency room visitor flow, and inpatient bed capacity, is obtained after anonymization through integration with the hospital information system.

[0021] Step 1.2: Clean and time-series aligned data collected in Step 1.1: Outlier identification and missing value imputation are employed, sampling granularity and time zone are standardized, feature derivation and label consistency verification are constructed, and the training set, validation set, and test set are divided chronologically.

[0022] Step 1.3: Establish a hierarchical / domain-specific hybrid prediction model: This includes the following steps:

[0023] Process 1: For ordinary areas, a time series forecasting model (ARIMAX / ARX) is used. The formula for predicting the cooling / heating load at time t is as follows: ,in, For constant terms, This is the autoregression order (the number of historical load lag steps used). Let i be the i-th autoregressive coefficient. The moving average order is... This is a random perturbation at this moment. Let k be the order of the moving average. For exogenous feature vectors, This is the regression coefficient vector corresponding to the exogenous features; where the exogenous feature vector includes operational feature parameters, outdoor meteorological feature parameters, and operational feature parameters;

[0024] Procedure 2: The clean operating room uses sensible heat / latent heat decomposition modeling to determine the indoor sensible heat load. Latent heat load ,in, For the heat dissipation of medical electronic equipment, For heat dissipation from lighting, For personnel sensible heat load, The latent heat of vaporization of water, For the mass flow rate of personnel moisture dissipation, Mass flow rate for dehumidification in the operating room;

[0025] Process 3: Domestic Hot Water Heat Load In the formula, The density of water, The specific heat capacity of water at constant pressure. Set the outlet water temperature for domestic hot water. This refers to the inlet / outlet water temperature. To predict volumetric flow rate, a dynamic regression model was used with bed occupancy, number of surgeries, and other exogenous disturbances as parameters.

[0026] Furthermore, the method for calculating the required cooling and heating for the clean operating room described in step 2 includes the following steps:

[0027] Step 2.1: Based on the air volume of the room purification air conditioning system Predicted indoor latent heat load and indoor temperature measurement value Indoor relative humidity monitoring value Dynamically adjust the supply air dew point temperature setpoint. Calculate the supply air dew point ratio enthalpy. ;

[0028] Step 2.2: Based on the air volume of the room purification air conditioning system Predicted indoor sensible heat load Indoor dry bulb temperature monitoring value Determine the supply air dry bulb temperature at the supply air state point. Calculate the specific enthalpy of the supply air state point. ;

[0029] Step 2.3: Based on the dry bulb temperature monitored at the fresh air and return air mixing point. and relative humidity value Calculate the specific enthalpy at the mixed air state point. ;

[0030] Step 2.4: Based on the system air supply volume Enthalpy value compared to the air supply state point Supply air dew point enthalpy value Calculate the reheat required by the air conditioning unit = ( According to the system air supply volume Supply air dew point enthalpy value Enthalpy value of the mixed state point of fresh air and return air Calculate the cooling capacity required for the air conditioning unit = ( );

[0031] Step 2.5: Summarize the total cooling demand of each system to obtain the total cooling capacity required for the clean operating room area. and reheat Combined forecast of heating and cooling loads in general hospital areas Domestic hot water load The real-time cooling demand of the hospital as a whole was calculated. and heating ,in, = + , = + + .

[0032] Furthermore, the dynamic water supply temperature optimization strategy described in step 3 includes the following:

[0033] a. Real-time acquisition of feedback signals from the temperature and humidity sensors in the main rooms and the water valve opening of the corresponding terminal air conditioning units;

[0034] b. When the maximum opening of the electric regulating valve of the chilled water of all monitored terminal air conditioning units is lower than a preset threshold, the controller gradually increases the set point of the chilled water supply temperature; when the maximum opening of the electric regulating valve of one terminal air conditioning unit reaches or exceeds the preset threshold, the controller gradually decreases the set point of the chilled water supply temperature.

[0035] c. When the maximum opening of the hot water electric regulating valve of all monitored terminal air conditioning units is lower than a preset threshold, the controller gradually lowers the hot water supply temperature setpoint; when the maximum opening of the water valve of one terminal air conditioning unit reaches or exceeds the preset threshold, the controller gradually raises the hot water supply temperature setpoint.

[0036] d. When the feedback signal of the electric regulating valve fails, the temperature and humidity of the controlled room are used as the control value, and whether the room temperature and humidity meet the standard is used instead of the preset threshold of the electric regulating valve position.

[0037] Furthermore, the particle swarm optimization (PSO) algorithm described in step 4 includes the following steps:

[0038] Step 4.1: Based on the performance parameters in the equipment sample file provided by the manufacturer, establish mathematical models for cold and heat source equipment and domestic hot water heat source equipment using a semi-empirical model. Among them, the four-pipe air-cooled heat pump unit and the chiller unit establish equipment correction curve clusters using the mode of cooling / heating capacity correction and operating power correction, which are used to correct the parameters under rated operating conditions to any actual operating conditions.

[0039] Step 4.2: The equipment correction curve cluster mentioned in Step 4.1 includes the capacity correction curve CAPFT with respect to temperature, the power consumption to heating / cooling capacity ratio correction curve EIRFT with respect to temperature, and the power consumption to heating / cooling capacity ratio correction curve EIRFPLR with respect to partial load, i.e., the cooling / heating capacity under the current operating conditions. Equipment operating power under current operating conditions In the formula and These are the cooling / heating capacity under rated operating conditions and the operating power under rated operating conditions, respectively.

[0040] In the equipment correction curve cluster described in steps 4.3, 4.1, and 4.2, the correction curve for cooling / heating capacity with respect to temperature uses the chilled / hot water outlet temperature. Outdoor ambient air temperature The empirical formula for a quadratic equation in two variables is given. The power consumption to cooling capacity ratio correction curve with respect to temperature uses the chilled water / hot water outlet temperature. Outdoor ambient air temperature The empirical formula for a quadratic equation in two variables is given. The ratio of power consumption to heating / cooling capacity is calculated using the partial load factor on the partial load correction curve. The empirical formula for cubic variables is given. ;

[0041] In steps 4.4 and 4.3, the correction curves in step 3 are all set to a value of 1 under rated operating conditions to maintain consistency with the rated operating conditions.

[0042] Step 4.5: In the above equipment correction curve clusters, the four-pipe air-cooled heat pump unit establishes equipment correction curve clusters according to five operating modes: single cooling, single heating, complete heat recovery, cooling-mainly heat recovery, and heating-mainly heat recovery. The chiller unit only establishes equipment correction curves for the cooling mode.

[0043] Step 4.6: Based on the equipment performance and the relationship between the cooling and heating demand of the building, develop typical operating condition strategy optimization subdomains, and set the particle swarm optimization algorithm (PSO) in each optimization domain.

[0044] Step 4.7: Based on the collected outdoor dry-bulb temperature, the determined chilled / hot water supply temperature setpoints of the air conditioning cold / heat source system, and the chiller unit cooling water inlet temperature setpoint, calculate the cooling and heating capacities of the four-pipe air-cooled heat pump unit and the chiller unit under the current operating conditions. Using the temperature correction factor (CAPFT) and the energy input ratio with temperature correction factor (EIRFT), determine the relationships between the cooling and heating capacities and power input and the partial load rate (PLR) of the four-pipe air-cooled heat pump unit and the chiller unit under the current operating conditions.

[0045] Step 4.8: Based on the equipment performance under the current operating conditions and the calculated real-time cooling and heating demand of the hospital, determine the optimization subdomain of the operating strategy adopted by the system, and call the PSO algorithm in each optimization subdomain to find the minimum total energy consumption of the hospital's cold and heat source system and domestic hot water heat source system. The optimization variables include, but are not limited to, the operating mode (cooling / heating / heat recovery), number or frequency of four-pipe air-cooled heat pump units, number or frequency of chiller units, number or frequency of domestic hot water heat source equipment, chilled water and hot water supply temperature settings, number and frequency of water pumps, and cooling tower fan and water pump speed.

[0046] The overall design concept of this invention is as follows: A hierarchical prediction model is established, encompassing the cooling and heating loads of general hospital areas, the indoor cooling and heating loads of clean operating rooms, and the domestic hot water load. Based on the predicted sensible and latent heat loads of each room in the clean operating room area, the required cooling and reheat capacity of the terminal air conditioning units is calculated, resulting in a comprehensive real-time cooling and heating demand for the entire hospital. A dynamic water supply temperature optimization strategy based on terminal water valve opening feedback is adopted. This strategy dynamically adjusts the chiller's water supply temperature setpoint by real-time monitoring of the terminal water valve opening, ensuring that the chiller's energy efficiency is maximized while meeting room comfort and dehumidification requirements. A domain-specific particle swarm optimization (PSO) algorithm is introduced, decomposing the complex global optimization problem into sub-optimization problems for different typical load conditions. The algorithm switches optimization modes based on load prediction results to improve computational efficiency, convergence speed, and operational stability. The four-pipe air-cooled heat pump unit, chiller unit, and domestic hot water heat source equipment are integrated and coordinated for control. Intelligent scheduling enables complementary advantages and tiered energy utilization among different units, thereby achieving optimal overall energy consumption at the system level.

[0047] The advantages of this invention are as follows:

[0048] 1. Compared with existing technologies that mostly use macroscopic and coarse load forecasting models, this paper constructs a hierarchical forecasting model that includes the heating and cooling loads of general areas of the hospital, the heating and cooling loads of clean operating rooms, and the domestic hot water load. This improves the forecasting accuracy, adapts to the complex heating and cooling load characteristics of large medical buildings, and provides a data foundation for the on-demand allocation and efficient operation of the hospital's energy system.

[0049] 2. Compared to traditional air conditioning chilled / hot water supply temperature setting strategies that use fixed supply water temperatures or are based on simple outdoor temperature resets, this system dynamically adjusts the chilled / hot water supply temperature setpoints of the air conditioning chiller / heater system by collecting feedback signals from the temperature and humidity sensors in the main rooms and the opening of the corresponding terminal air conditioning unit water valves in real time. This prioritizes meeting the cooling and heating needs of the terminals, ensuring the temperature and humidity control requirements of each room in the hospital's functional areas, especially the clean operating room area. At the same time, it can significantly improve the operating energy efficiency of the chiller / heater equipment.

[0050] 3. Compared to the global optimization Particle Swarm Optimization (PSO) algorithm, applying the domain-specific optimization PSO algorithm by pre-setting typical operating conditions can improve computational efficiency, convergence speed, and operational stability.

[0051] 4. By jointly optimizing the total energy consumption of the hospital's cold and heat source system and the domestic hot water source system, the overall cooling and heating needs of the building can be taken into account. This can reduce the cooling and heating imbalance when the four-pipe air-cooled heat pump unit is operating in heat recovery mode, give full play to the advantages of the four-pipe air-cooled heat pump unit's high comprehensive energy efficiency in providing both cooling and heating, and improve the overall energy utilization efficiency of the building. Attached Figure Description

[0052] Figure 1 A schematic diagram of the steps involved in an energy-saving control strategy for air conditioning and ventilation systems in medical buildings based on climate change characteristics and load forecasting;

[0053] Figure 2 A schematic diagram illustrating the execution process of an energy-saving control strategy for air conditioning and ventilation systems in medical buildings based on climate change characteristics and load forecasting. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for explaining this invention and are not intended to limit this invention.

[0055] This invention provides an embodiment of an energy-saving control strategy for air conditioning and ventilation systems in medical buildings based on climate change characteristics and load forecasting. This strategy is applied to a large general hospital located in Fuzhou City. The hospital's cold and heat source system consists of four four-pipe air-cooled heat pump units (rated cooling capacity 1398kW in single cooling mode, rated heating capacity 1422kW in single heating mode; rated cooling capacity 1348kW and rated heating capacity 1678kW in simultaneous cooling and heating mode) coupled with four high-efficiency centrifugal chillers (rated cooling capacity 3516kW), simultaneously serving the air conditioning system and the domestic hot water system (using air-source heat pump units as the heat source). The control strategy of this invention is implemented through an intelligent closed-loop control system, which includes a sensing layer, an intelligent control layer, an execution layer, and a controlled equipment layer.

[0056] The specific execution steps of this embodiment are as follows:

[0057] Step 1: Establish a comprehensive building heating and cooling load prediction model

[0058] This step aims to accurately predict the hospital's future heating and cooling needs, providing a basis for optimized system scheduling. Specifically, it includes the following steps:

[0059] Step 1.1: Raw Data Access, including but not limited to historical operating data, meteorological data, and operational data. Historical operating data affecting the cooling and heating load of the air conditioning system will be captured using sensors and data acquisition equipment. This includes, but is not limited to, indoor ambient temperature, relative humidity, cooling and heating output from the cooling and heating source systems, domestic hot water consumption, and operating parameters of various equipment in the air conditioning and domestic hot water systems. Real-time meteorological data will be collected using a rooftop weather station, including but not limited to outdoor dry-bulb temperature, relative humidity, radiation intensity, and wind speed. Real-time collection of meteorological data, equipment operating data, and operational data will be conducted, including but not limited to surgical scheduling, outpatient and emergency room visitor flow, and inpatient bed capacity.

[0060] Step 1.2: Clean and time-series aligned data collected in Step 1.1: outlier identification and missing value imputation are used, sampling granularity and time zone are unified, and training set, validation set and test set are divided in chronological order.

[0061] c. Establishing a hierarchical / domain-specific hybrid prediction model includes the following steps:

[0062] Process 1: Normal area, using ARIMAX / ARX model:

[0063] To balance interpretability and short-term forecast accuracy, ARIMAX (degenerates to ARX when q=0) is used for general areas. The formula for predicting the cooling / heating load at time t is as follows: .in, For constant terms, This is the autoregression order (the number of historical load lag steps used). Let i be the i-th autoregressive coefficient. The moving average order is... This is a random perturbation at this moment. Let k be the order of the moving average. It is an exogenous feature vector (including operational feature parameters, outdoor meteorological feature parameters, and operational feature parameters, etc.). This represents the regression coefficient vector corresponding to the exogenous features. During training... The target is q=0, which yields ARX: .

[0064] Procedure 2: The clean operating room uses sensible heat / latent heat decomposition modeling to determine the indoor sensible heat load. Latent heat load .in, For the heat dissipation of medical electronic equipment, For heat dissipation from lighting, For personnel sensible heat load, The latent heat of vaporization of water, For the mass flow rate of personnel moisture dissipation, The mass flow rate for dehumidification in the operating room.

[0065] Step 3: Based on the law of conservation of energy, the heat load of domestic hot water In the formula, The density of water, The specific heat capacity of water at constant pressure. Set the outlet water temperature for domestic hot water. This refers to the inlet temperature of tap water / return water. To predict volumetric flow rate, a dynamic regression model was used with bed occupancy, number of surgeries, and other exogenous disturbances as parameters.

[0066] Step 2, the calculation method for the required cooling and heating in the clean operating room, includes the following steps:

[0067] Step 2.1: Based on the air volume of the room purification air conditioning system Predicted indoor latent heat load and indoor temperature measurement value Indoor relative humidity monitoring value The specific moisture content of the supply air dew point was calculated. = - / ( ),in This refers to the latent heat of vaporization. =f( , Dynamically adjust the supply air dew point temperature setpoint. (Assuming machine dew point is 95%), calculate the supply air dew point enthalpy. .

[0068] Step 2.2: Based on the air volume of the room purification air conditioning system Predicted indoor sensible heat load Indoor dry bulb temperature monitoring value Determine the supply air dry bulb temperature at the supply air state point. = / - Calculate the specific enthalpy of the supply air state point. =f( , );

[0069] Step 2.3: Based on the dry bulb temperature monitored at the fresh air and return air mixing point. and relative humidity value Calculate the specific enthalpy at the mixed air state point. .

[0070] Step 2.4: Based on the system air supply volume Enthalpy value compared to the air supply state point Supply air dew point enthalpy value Calculate the reheat required by the air conditioning unit = ( According to the system air supply volume Supply air dew point enthalpy value Enthalpy value of the mixed state point of fresh air and return air Calculate the cooling capacity required for the air conditioning unit = ( ).

[0071] Step 2.5: Summarize the total cooling demand of each system to obtain the total cooling capacity required for the clean operating room area. and reheat Combined forecast of heating and cooling loads in general hospital areas Domestic hot water load The real-time cooling demand of the hospital as a whole was calculated. and heating ,in, = + , = + + .

[0072] Step 3: Dynamic water supply temperature optimization strategy

[0073] This step aims to maximize the energy efficiency of the cooling and heating unit while ensuring end-user demand. Specifically, it includes the following steps:

[0074] a. Real-time acquisition of feedback signals from the temperature and humidity sensors in the main rooms and the water valve opening of the corresponding terminal air conditioning units.

[0075] b. When the maximum opening of the electric regulating valve of the chilled water of all monitored terminal air conditioning units is lower than a preset threshold (95% in this embodiment), the controller gradually increases the set point of the chilled water supply temperature; when the maximum opening of the electric regulating valve of one terminal air conditioning unit reaches or exceeds the preset threshold, the controller gradually decreases the set point of the chilled water supply temperature.

[0076] c. When the maximum opening of the hot water electric regulating valve of all monitored terminal air conditioning units is lower than a preset threshold (95% in this embodiment), the controller gradually lowers the hot water supply temperature set point; when the maximum opening of the water valve of one terminal air conditioning unit reaches or exceeds the preset threshold, the controller gradually raises the hot water supply temperature set point.

[0077] d. When the feedback signal of the electric regulating valve fails, the temperature and humidity of the controlled room are used as the control value, and whether the room temperature and humidity meet the standard is used instead of the preset threshold of the electric regulating valve position.

[0078] Step 4, the Particle Swarm Optimization (PSO) algorithm for domain-specific optimization, includes the following steps:

[0079] This step is the brain of the system, determining how each device should operate to minimize overall energy consumption. It specifically includes the following steps:

[0080] Step 4.1: Based on the performance parameters in the equipment sample files provided by the manufacturer, establish mathematical models for the cold and heat source equipment and the domestic hot water heat source equipment using a semi-empirical model. Among them, the four-pipe air-cooled heat pump unit and the chiller unit establish equipment correction curve clusters using the mode of cooling / heating capacity correction and operating power correction, which are used to correct the parameters under rated operating conditions to any actual operating conditions.

[0081] Step 4.2: The equipment correction curve cluster mentioned in Step 4.1 includes the capacity correction curve CAPFT with respect to temperature, the power consumption to heating / cooling capacity ratio correction curve EIRFT with respect to temperature, and the power consumption to heating / cooling capacity ratio correction curve EIRFPLR with respect to partial load, i.e., the cooling / heating capacity under the current operating conditions. Equipment operating power under current operating conditions In the formula and These refer to the cooling / heating capacity under rated operating conditions and the operating power under rated operating conditions, respectively.

[0082] In the equipment correction curve cluster described in steps 4.3, 4.1, and 4.2, the correction curve for cooling / heating capacity with respect to temperature uses the chilled / hot water outlet temperature. Outdoor ambient air temperature The empirical formula for a quadratic equation in two variables is given. The power consumption to cooling capacity ratio correction curve with respect to temperature uses the chilled water / hot water outlet temperature. Outdoor ambient air temperature The empirical formula for a quadratic equation in two variables is given. The ratio of power consumption to heating / cooling capacity is calculated using the partial load factor on the partial load correction curve. The empirical formula for cubic variables is given. .

[0083] In steps 4.4 and 4.3, the correction curves in step 3 are all set to a value of 1 under rated operating conditions to maintain consistency with the rated operating conditions.

[0084] Step 4.5: In the above equipment correction curve clusters, the four-pipe air-cooled heat pump unit establishes equipment correction curve clusters according to the five operating modes: single cooling, single heating, complete heat recovery, cooling as the main heat recovery, and heating as the main heat recovery. The chiller unit only establishes equipment correction curves for the cooling mode.

[0085] Step 4.6: Based on the equipment performance and the relationship between the cooling and heating demand of the building, develop typical operating condition strategy optimization subdomains, and set the Particle Swarm Optimization (PSO) algorithm in each optimization domain. When the heating demand is less than the cooling demand plus the operating power of the four-pipe air-cooled heat pump unit in heat recovery mode, it is subdomain 1, which uses a combination of four-pipe air-cooled heat pump, chiller, and domestic hot water heat source equipment for cooling and heating; when the heating demand is greater than or equal to the cooling demand plus the operating power of the four-pipe air-cooled heat pump unit in heat recovery mode, it is subdomain 2, which uses a combination of four-pipe air-cooled heat pump and domestic hot water heat source equipment for cooling and heating.

[0086] Step 4.7: Based on the collected outdoor dry-bulb temperature, the determined chilled / hot water supply temperature setpoints of the air conditioning cold and heat source system, and the setpoints of the chiller unit's cooling water inlet temperature, calculate the cooling and heating capacities of the four-pipe air-cooled heat pump unit and the chiller unit under the current operating conditions. Based on the temperature correction factor (CAPFT) and the energy input ratio with temperature correction factor (EIRFT), determine the relationship between the cooling and heating capacities and power input and the partial load rate (PLR) of the four-pipe air-cooled heat pump unit and the chiller unit under the current operating conditions.

[0087] Step 4.8: Based on the equipment performance under the current operating conditions and the calculated real-time cooling and heating demand of the hospital, determine the optimization subdomain of the operating strategy adopted by the system, and call the PSO algorithm in each optimization subdomain to find the minimum total energy consumption of the hospital's cold and heat source system and domestic hot water heat source system. The optimization variables include, but are not limited to, the operating mode (cooling / heating / heat recovery), number or frequency of four-pipe air-cooled heat pump units, number or frequency of chiller units, number or frequency of domestic hot water heat source equipment, chilled water and hot water supply temperature settings, number and frequency of water pumps, and cooling tower fan and water pump speed.

[0088] Step 5: Implement multi-unit coordinated control and safety constraints: Before issuing the optimal command derived from the PSO algorithm, it must be verified by the safety constraint verification module to ensure the feasibility and safety of the command. Constraint verification includes, but is not limited to, hydraulic feasibility, minimum partial load factor of the unit, minimum start-up and shutdown interval, heat recovery power balance, and inlet and outlet water temperature difference.

[0089] Step 6, Continuous Learning and Model Updates: Continuously collect operational data and user feedback from the control system, analyze system performance, and regularly update the building cooling load model based on new data and feedback to improve prediction accuracy.

[0090] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments. Within the scope of knowledge possessed by those skilled in the art, various changes can be made without departing from the spirit of the present invention.

Claims

1. An energy-saving control strategy for air conditioning and ventilation systems in medical buildings based on climate change characteristics and load forecasting, characterized in that, Includes the following steps: Step 1: Establish a comprehensive building heating and cooling load prediction model. Construct a hierarchical / regional hybrid prediction model that includes the heating and cooling load of the general area of ​​the hospital, the indoor heating and cooling load of the clean operating department, and the domestic hot water load. Utilize historical operating load data and real-time meteorological monitoring data to predict the heating and cooling load of different areas and systems of the building hourly. The process of establishing a hierarchical / domain hybrid prediction model includes the following steps: Process 1: For ordinary areas, a time series forecasting model is used. The formula for predicting the cooling / heating load at time t is as follows: ,in, For constant terms, Let the order be the autoregressive order. Let i be the i-th autoregressive coefficient. The moving average order is... This is a random perturbation at this moment. Let k be the order of the moving average. For exogenous feature vectors, This is the vector of regression coefficients corresponding to the exogenous features; Procedure 2: The clean operating room uses sensible heat / latent heat decomposition modeling to determine the indoor sensible heat load. Latent heat load ,in, For the heat dissipation of medical electronic equipment, For heat dissipation from lighting, For personnel sensible heat load, The latent heat of vaporization of water, For the mass flow rate of personnel moisture dissipation, Mass flow rate for dehumidification in the operating room; Process 3: Domestic Hot Water Heat Load In the formula, The density of water, The specific heat capacity of water at constant pressure. Set the outlet water temperature for domestic hot water. This refers to the inlet / outlet water temperature. To predict volumetric flow rate; Step 2: Calculate the cooling and heating demand of the clean operating department area. Based on the indoor sensible heat load and latent heat load of each room in the clean operating department area predicted in Step 1, calculate the cooling and reheat demand of the terminal air conditioning units based on the constant supply air humidity control logic. Then, summarize the total cooling and reheat demand of each room in the clean operating department area, and combine the cooling and heating load of the general area of ​​the hospital and the domestic hot water load predicted in Step 1 to obtain the overall real-time cooling and heating demand of the hospital. Step 3: Perform dynamic water supply temperature optimization, collect feedback signals of the opening of the water valves of the corresponding terminal air conditioning units in the main rooms and temperature and humidity sensors in real time, and dynamically adjust the set value of the chilled / hot water supply temperature of the air conditioning cold and heat source system to improve the unit's operating energy efficiency while meeting the cooling and heating needs. Determine the cooling water inlet temperature of the chiller unit based on real-time meteorological data. Step 4: Apply the domain-based optimization PSO algorithm to determine the system operation strategy optimization subdomain based on the building's cooling and heating demand. In each optimization subdomain, call the targeted PSO sub-algorithm to minimize the total energy consumption of the hospital's cold and heat source system and the domestic hot water source system, and set comfort level and switching penalty. Step 5: Implement multi-unit collaborative control and safety constraints: After the optimization results are checked for constraints, they are sent out for execution. The constraints check includes hydraulic feasibility, minimum partial load factor of the unit, minimum start-up and shutdown interval, heat recovery power balance and inlet and outlet water temperature difference. Step 6: Continuously collect operating data and user feedback from the control system, analyze system performance, and regularly update the building cooling load model based on new data and feedback to improve prediction accuracy.

2. The energy-saving control strategy for air conditioning and ventilation systems in medical buildings based on climate change characteristics and load forecasting as described in claim 1, characterized in that, In process 1, the exogenous feature vector includes operational feature parameters, outdoor meteorological feature parameters, and operational feature parameters; in process 3, the predicted volumetric flow rate is obtained using a dynamic regression model with bed occupancy and number of surgeries as exogenous disturbances.

3. The energy-saving control strategy for air conditioning and ventilation systems in medical buildings based on climate change characteristics and load forecasting as described in claim 1, characterized in that, Step 1, establishing a comprehensive building heating and cooling load prediction model, includes the following steps: Step 1.1: Raw data access, including historical operating data, meteorological data, and operational data. Historical operating data affecting the cooling and heating load of the air conditioning system is captured through sensors and data acquisition equipment, including indoor ambient temperature, relative humidity, cooling and heating output of the cooling and heating source system, domestic hot water consumption, and operating parameters of each device in the air conditioning system and domestic hot water system. Meteorological data is collected in real time through the roof weather station, including outdoor dry-bulb temperature, relative humidity, radiation intensity, and wind speed. Operational data is obtained after anonymization by connecting to the hospital information system, including surgical scheduling, outpatient and emergency patient flow, and number of inpatient beds. Step 1.2: Clean and time-series aligned data collected in Step 1.1: outlier identification and missing value imputation are used, sampling granularity and time zone are unified, feature derivation and label consistency verification are constructed, and training set, validation set and test set are divided in chronological order.

4. The energy-saving control strategy for air conditioning and ventilation systems in medical buildings based on climate change characteristics and load forecasting as described in claim 1, characterized in that, Step 2 describes the calculation method for the required cooling and heating for the clean operating room, which includes the following steps: Step 2.1: Based on the air volume of the room purification air conditioning system Predicted indoor latent heat load and indoor temperature measurement value Indoor relative humidity monitoring value Dynamically adjust the supply air dew point temperature setpoint. Calculate the supply air dew point ratio enthalpy. ; Step 2.2: Based on the air volume of the room purification air conditioning system Predicted indoor sensible heat load Indoor dry bulb temperature monitoring value Determine the supply air dry bulb temperature at the supply air state point. Calculate the specific enthalpy of the supply air state point. ; Step 2.3: Based on the dry bulb temperature monitored at the fresh air and return air mixing point. and relative humidity value Calculate the specific enthalpy at the mixed air state point. ; Step 2.4: Based on the system air supply volume Enthalpy value compared to the air supply state point Supply air dew point enthalpy value Calculate the reheat required by the air conditioning unit = ( According to the system air supply volume Supply air dew point enthalpy value Enthalpy value of the mixed state point of fresh air and return air Calculate the cooling capacity required for the air conditioning unit = ( ); Step 2.5: Summarize the total cooling demand of each system to obtain the total cooling capacity required for the clean operating room area. and reheat Combined forecast of heating and cooling loads in general hospital areas Domestic hot water load The real-time cooling demand of the hospital as a whole was calculated. and heating ,in, = + , = + + .

5. The energy-saving control strategy for air conditioning and ventilation systems in medical buildings based on climate change characteristics and load forecasting as described in claim 1, characterized in that, The dynamic water supply temperature optimization strategy described in step 3 includes the following: a. Real-time acquisition of feedback signals from the temperature and humidity sensors in the main rooms and the water valve opening of the corresponding terminal air conditioning units; b. When the maximum opening of the electric regulating valve of the chilled water of all monitored terminal air conditioning units is lower than a preset threshold, the controller gradually increases the set point of the chilled water supply temperature; when the maximum opening of the electric regulating valve of one terminal air conditioning unit reaches or exceeds the preset threshold, the controller gradually decreases the set point of the chilled water supply temperature. c. When the maximum opening of the hot water electric regulating valve of all monitored terminal air conditioning units is lower than a preset threshold, the controller gradually lowers the hot water supply temperature setpoint; when the maximum opening of the water valve of one terminal air conditioning unit reaches or exceeds the preset threshold, the controller gradually raises the hot water supply temperature setpoint. d. When the feedback signal of the electric regulating valve fails, the temperature and humidity of the controlled room are used as the control value, and whether the room temperature and humidity meet the standard is used instead of the preset threshold of the electric regulating valve position.

6. The energy-saving control strategy for air conditioning and ventilation systems in medical buildings based on climate change characteristics and load forecasting as described in claim 1, characterized in that, Step 4 of the domain-specific optimization PSO algorithm includes the following steps: Step 4.1: Based on the performance parameters in the equipment sample file provided by the manufacturer, establish mathematical models for cold and heat source equipment and domestic hot water heat source equipment using a semi-empirical model. Among them, the four-pipe air-cooled heat pump unit and the chiller unit establish equipment correction curve clusters using the mode of cooling / heating capacity correction and operating power correction, which are used to correct the parameters under rated operating conditions to any actual operating conditions. Step 4.2: The equipment correction curve cluster mentioned in Step 4.1 includes the capacity correction curve CAPFT with respect to temperature, the power consumption to heating / cooling capacity ratio correction curve EIRFT with respect to temperature, and the power consumption to heating / cooling capacity ratio correction curve EIRFPLR with respect to partial load, i.e., the cooling / heating capacity under the current operating conditions. Equipment operating power under current operating conditions In the formula and These are the cooling / heating capacity under rated operating conditions and the operating power under rated operating conditions, respectively. In the equipment correction curve cluster described in steps 4.3, 4.1, and 4.2, the correction curve for cooling / heating capacity with respect to temperature uses the chilled / hot water outlet temperature. Outdoor ambient air temperature The empirical formula for a quadratic equation in two variables is given. The power consumption to cooling capacity ratio correction curve with respect to temperature uses the chilled water / hot water outlet temperature. Outdoor ambient air temperature The empirical formula for a quadratic equation in two variables is given. The ratio of power consumption to heating / cooling capacity is calculated using the partial load factor on the partial load correction curve. The empirical formula for cubic variables is given. ; The correction curves mentioned in steps 4.4 and 4.3 are all set to a value of 1 under rated operating conditions to maintain consistency with rated operating conditions. Step 4.5: In the above equipment correction curve clusters, the four-pipe air-cooled heat pump unit establishes equipment correction curve clusters according to five operating modes: single cooling, single heating, complete heat recovery, cooling-mainly heat recovery, and heating-mainly heat recovery. The chiller unit only establishes equipment correction curves for the cooling mode. Step 4.6: Based on equipment performance and the relationship between cooling and heating requirements of the building, develop typical operating condition strategy optimization subdomains, and set the PSO optimization algorithm within each optimization domain. Step 4.7: Based on the collected outdoor dry-bulb temperature, the determined chilled / hot water supply temperature setpoints of the air conditioning cold / heat source system, and the setpoints for the chiller unit's cooling water inlet temperature, calculate the cooling and heating capacities of the four-pipe air-cooled heat pump unit and the chiller unit under the current operating conditions. Using the temperature correction coefficient and the energy input ratio with temperature correction coefficient, determine the relationships between the cooling and heating capacities and power input and the partial load rate of the four-pipe air-cooled heat pump unit and the chiller unit under the current operating conditions. Step 4.8: Based on the equipment performance under the current operating conditions and the calculated real-time cooling and heating demand of the hospital, determine the optimization subdomain of the system's operating strategy, and call the PSO algorithm in each optimization subdomain to find the minimum total energy consumption of the hospital's cold and heat source system and domestic hot water heat source system. The optimization variables include the operating mode, number or frequency of the four-pipe air-cooled heat pump unit, the number or frequency of the chiller unit, the number or frequency of the domestic hot water heat source equipment, the set temperature of chilled water and hot water supply, the number and frequency of water pumps, and the speed of the cooling tower fan and water pump.

Images