A method and system for controlling the absorption capacity of a radiant air conditioning microgrid

Abstract

This invention proposes a method and system for controlling the absorption capacity of a radiant air conditioning microgrid. The method determines the allowable thermal and humidity dynamic elastic range for the radiant terminal surface based on collected data, combined with a human thermal comfort model and a dew point temperature dynamic evolution model. This dynamic elastic range is then transformed into equivalent virtual energy storage capacity, maximum absorbable charging / discharging power, and maximum sustainable discharge duration, based on building thermal capacity characteristics and the variable operating condition performance curves of the heat pump unit. According to the received power scheduling command, within the constraints of the thermal and humidity dynamic elastic range, the optimal control sequence for each device parameter in the radiant air conditioning system is solved using a rolling optimization algorithm. Based on this optimal control sequence, the system controls each device in the radiant air conditioning system, achieving a shift from passive response to active guidance. This improves the safety and reliability of source-load interaction and enhances the reliability of the radiant air conditioning microgrid absorption capacity control.

Description

Technical Field

[0001] This invention relates to the field of synergistic control between local renewable energy consumption and building environmental comfort and safety, and in particular to a method and system for controlling the absorption capacity of a radiant air conditioning microgrid. Background Technology

[0002] Against the backdrop of global efforts to address climate change and promote energy structure transformation, national strategies have placed unprecedented demands on energy efficiency. In this context, the penetration rate of distributed renewable energy, primarily in the form of photovoltaics and wind power, in buildings, industrial parks, and urban areas is growing at an unprecedented rate. This trend is prompting traditional electricity consumers, such as large public buildings and industrial parks, to gradually evolve into prosumers integrating energy production, consumption, storage, and management, forming microgrids with a certain degree of autonomy. However, the inherent intermittency, volatility, and randomness of distributed renewable energy pose severe challenges to the power balance, power quality, and stable operation of microgrids. When photovoltaic power generation surges dramatically in a short period or drops sharply due to cloud cover, without effective regulation, it will lead to severe fluctuations in microgrid frequency and voltage, and may even cause system instability.

[0003] To address this challenge, the industry commonly adopts electrochemical energy storage systems, such as lithium-ion battery energy storage power stations. Electrochemical energy storage boasts rapid charge and discharge response capabilities, effectively smoothing out output fluctuations from renewable energy sources and achieving peak shaving and valley filling. However, its drawbacks are also significant: First, the construction cost is high, accounting for a considerable proportion of the total investment in microgrids, thus limiting its economic viability; second, the cycle life of batteries is limited, with capacity decay issues, and the cost-effectiveness over the entire life cycle needs further improvement; third, the operation of large-scale battery arrays is also accompanied by potential thermal runaway and fire safety risks. Therefore, seeking and developing more economical and inherently flexible resources besides electrochemical energy storage has become a key bottleneck driving the development of microgrid technology.

[0004] Buildings, as one of the primary energy-consuming terminals in society, possess enormous, yet largely untapped, potential for flexible regulation. In particular, the heating, ventilation, and air conditioning (HVAC) system, which accounts for nearly half of building energy consumption, is considered an ideal demand-side response resource due to its interruptible and adjustable operation. Radiant air conditioning systems, including underfloor heating / cooling systems, ceiling radiant panels, and capillary networks, regulate the indoor environment by embedding hot and cold water circulation pipes within the building's concrete floor slabs, walls, or ceilings, utilizing the large-area radiant heat exchange of these building envelope surfaces. This working principle gives radiant air conditioning systems a unique advantage over traditional convection air conditioning: significant thermal inertia. Concrete and other building materials have extremely high heat capacity, giving radiant systems a natural, large-scale capacity for heat or cold storage, much like a thermal battery. In theory, this characteristic can be utilized to allow the radiant system to operate ahead of time when there is a surplus of renewable energy generation in the microgrid, such as when solar power is generating a lot of electricity at noon. This allows the excess electricity to be converted into cooling or heating and stored in the building structure. When electricity prices are high or renewable energy output is insufficient, the operating power of the air conditioning unit can be stopped or reduced, and the stored energy can be slowly released to maintain the comfort of the indoor environment. This can achieve peak shaving and valley filling of the power grid and efficiently absorb fluctuating energy.

[0005] However, putting this theoretical concept into practice faces a series of complex technical challenges that existing technical solutions cannot effectively address.

[0006] First, there's the issue of control logic mismatch and response lag. Currently, most microgrids interact with loads using a grid-dominated, one-way command model. This means the microgrid energy management system (EMS) directly issues coarse control commands to the load side based on its own power balance needs, such as forced start / stop and power limit settings. This control method completely ignores the temperature response lag characteristic of radiant air conditioning systems, which can last for several hours. A dispatch command from the grid side based on a current power deficit measured in minutes might, several hours later, cause indoor temperatures to deviate significantly from the comfort zone, leading to user complaints and rendering the function practically useless in real-world applications.

[0007] Secondly, and most critically, is the unavoidable risk of condensation. During summer cooling operations, the surface temperature of the radiant terminals must be strictly controlled above the indoor air dew point temperature. Otherwise, water vapor in the air will condense on the radiant surface, potentially damaging floor or wall finishes, causing mold growth, harming human health, and even leading to electrical safety accidents. The indoor air dew point temperature is a dynamically changing parameter, closely related to indoor dry-bulb temperature and relative humidity, which are in turn affected in real-time by various factors such as outdoor weather conditions, indoor occupancy, and the operating status of the fresh air system. Existing demand-side response strategies often focus only on the economic benefits or power targets of the grid, lacking precise quantification and forward-looking constraints on the safety margin between the radiant terminal surface temperature and the dynamic dew point temperature. This blind scheduling easily triggers the condensation safety red line, leading maintenance providers to prioritize system safety over the economic benefits of grid interaction, thus completely rendering the enormous flexibility potential of radiant air conditioning systems idle.

[0008] Finally, there is the barrier to source-load information exchange. Microgrid energy management systems use electricity as their language for dispatching, focusing on how many kilowatts of regulating power the load side can provide, for how long it can sustain this power, or how many kilowatt-hours of electricity it can absorb. In contrast, the operating status of radiant air conditioning systems is described by thermophysical quantities such as temperature, humidity, and heat flux. A translation gap exists between these two. Currently, there is a lack of a standardized method to accurately and dynamically translate the controllable potential of radiant air conditioning systems at specific times, under the dual constraints of comfort and safety, into electrical parameters that the microgrid energy management system can understand and directly dispatch. This information asymmetry prevents microgrids from accurately assessing and optimizing the flexibility of building loads, resulting in poor source-load matching.

[0009] In summary, in order to solve at least one of the technical problems in the prior art, this solution is proposed to improve the reliability of the absorption capacity control of the radiant air conditioning microgrid. Summary of the Invention

[0010] To address the problems existing in the prior art, this invention innovatively proposes a method and system for controlling the absorption capacity of radiant air conditioning microgrids. This effectively solves the problem of low reliability in the control of the absorption capacity of radiant air conditioning microgrids caused by the prior art, and effectively improves the reliability of the control of the absorption capacity of radiant air conditioning microgrids.

[0011] The first aspect of this invention provides a method for controlling the absorption capacity of a radiant air conditioning microgrid, comprising: Real-time data collection of indoor and outdoor temperature and humidity, radiant terminal surface temperature, water supply and return status, and power flow data at the microgrid common connection point; Based on the collected data, combined with the human thermal comfort model and the dew point temperature dynamic evolution model, the upper and lower limits of the dynamic temperature that the radiant terminal surface can operate at are determined, forming a dynamic elastic range of thermal and humidity. The dynamic elastic range of heat and humidity is transformed into equivalent virtual energy storage capacity, maximum absorbable charging / discharging power and maximum sustainable discharge duration based on building heat capacity characteristics and heat pump unit variable operating condition performance curves, generating a dynamic map of microgrid absorption capacity. The dynamic map of microgrid absorption capacity is reported to the microgrid energy management system, and power dispatch instructions or frequency response instructions issued by the microgrid energy management system based on the dynamic map of microgrid absorption capacity are received. Based on the received power scheduling command or frequency response command, within the constraints of the thermal and humidity dynamic elastic range, the optimal control sequence of each device parameter in the radiant air conditioner is solved by the rolling optimization algorithm, and the device in the radiant air conditioner is controlled based on the optimal control sequence of each device parameter.

[0012] A second aspect of the present invention provides a radiant air conditioning microgrid absorption capacity control system, comprising: The environmental and operating condition multi-dimensional sensing subsystem collects real-time data on indoor and outdoor temperature and humidity, surface temperature of radiant terminals, water supply and return status, and power flow data at the microgrid common connection point. The thermal and humidity dynamic elastic range calculation unit, based on the collected data and combined with the human thermal comfort model and the dew point temperature dynamic evolution model, determines the upper and lower limits of the dynamic temperature that the radiant terminal surface can operate at, thus forming the thermal and humidity dynamic elastic range. The virtual energy storage mapping and conversion unit transforms the dynamic elastic range of heat and humidity based on the building's thermal capacity characteristics and the variable operating condition performance curve of the heat pump unit into an equivalent virtual energy storage capacity, maximum absorbable charging power, and maximum sustainable discharge duration, thereby generating a dynamic map of the microgrid's absorption capacity. The source-load two-way handshake interaction interface reports the dynamic map of microgrid absorption capacity to the microgrid energy management system and receives power dispatch instructions or frequency response instructions issued by the microgrid energy management system based on the dynamic map of microgrid absorption capacity. The model predictive control execution unit, based on the received power scheduling command or frequency response command, solves the optimal control sequence of each device parameter in the radiant air conditioner through a rolling optimization algorithm within the constraint range of the thermal and humidity dynamic elastic range, and controls each device in the radiant air conditioner based on the optimal control sequence of each device parameter.

[0013] The technical solution adopted in this invention has the following technical effects: 1. In the technical solution of this invention, based on the collected data and combined with the human thermal comfort model and the dew point temperature dynamic evolution model, the upper and lower limits of the dynamic temperature that the surface of the radiant terminal can operate at are determined, forming a dynamic elastic range of thermal and humidity. The dynamic elastic range of thermal and humidity is transformed into equivalent virtual energy storage capacity, maximum absorbable charging / discharging power, and maximum sustainable discharge time based on the building thermal capacity characteristics and the variable operating condition performance curve of the heat pump unit. According to the received power scheduling command or frequency response command, within the constraints of the dynamic elastic range of thermal and humidity, the optimal control sequence of each equipment parameter in the radiant air conditioner is solved by the rolling optimization algorithm. Based on the optimal control sequence of each equipment parameter in the radiant air conditioner, the equipment of the radiant air conditioner is controlled, realizing the mode transformation from passive response to active guidance, improving the safety and reliability of source-load interaction, effectively solving the problem of low reliability of radiant air conditioner microgrid absorption capacity control caused by existing technology, and effectively improving the reliability of radiant air conditioner microgrid absorption capacity control.

[0014] 2. In this invention, the current air dew point temperature is calculated using the indoor air dry-bulb temperature and relative humidity; the dynamic cold-side limit temperature of the radiant terminal surface is determined based on the current air dew point temperature and the anti-condensation safety margin coefficient; an indoor operating temperature range that meets the requirements of human thermal comfort level is set according to the expected average thermal sensation index model, and the dynamic hot-side limit temperature of the radiant terminal surface is back-calculated by combining the radiant heat transfer equation; the union region between the dynamic cold-side limit temperature and the dynamic hot-side limit temperature is taken as the current dynamic elastic range of heat and humidity, thus defining a clear anti-condensation safety and comfort boundary for the flexible adjustment of radiant air conditioning.

[0015] 3. In this invention, the technical solution calculates the heat that can be absorbed or released when the current radiant terminal is driven from the average temperature to the boundary of the elastic range. The heat value that can be absorbed or released is converted into the virtual electrical energy storage capacity on the grid side through the real-time comprehensive performance coefficient of the heat pump unit. Based on the real-time comprehensive performance coefficient of the heat pump unit, the maximum heat exchange that the water system circulation flow can support under the current operating conditions is mapped to the maximum absorbable charging power. The current total operating power of the radiant air conditioning system is used as the maximum absorbable discharge power. Using the building heat attenuation model, the boundary value of the thermal and humidity dynamic elastic range is converted into the maximum sustainable discharge duration. On the one hand, through the inverse mapping of the virtual energy pool, the complex physical... The constraints are translated into a language understandable to the power grid, enabling proactive reporting of load-side absorption capacity. On the other hand, it provides a unified and standardized way to describe the flexibility of radiant air-conditioned buildings of different types and sizes. This standardized interface greatly simplifies the development and deployment of microgrid energy management systems, allowing them to connect to and manage large numbers of flexible building loads like plug-and-play. This lays a solid technical foundation for building larger-scale virtual power plants and regional energy internet, with broad application prospects and promotional value. Furthermore, it equates the building's huge heat capacity to a virtual energy pool, providing microgrids with a large-capacity energy storage option at almost zero cost. During periods of high photovoltaic power generation, the system can accurately calculate the amount of electricity that can be safely absorbed and guide radiant air conditioning systems to charge, for example, through pre-cooling or pre-heating, converting potentially wasted green electricity into useful thermal energy for storage. This not only effectively mitigates the fluctuations in renewable energy and improves its utilization rate but also significantly reduces the need for and dependence on expensive electrochemical energy storage systems, bringing significant economic benefits to the construction and operation of microgrids.

[0016] 4. The technical solution of this invention establishes a nonlinear building thermal dynamic model. Using state data, equipment control sequences, and disturbance data as inputs, and the power curve issued by the microgrid as the target tracking trajectory, and the thermal-humidity dynamic elastic range as the state constraint, the optimal control sequence of each equipment parameter in the radiant air conditioning system within the future control time domain is solved by minimizing the objective function. The state data includes the average indoor air temperature and the average surface temperature of the radiant terminals. The disturbance data includes the outdoor air temperature, solar radiation heat gain entering the room through windows, and internal heat sources generated by indoor occupants and equipment. This approach decomposes the macroscopic grid dispatch objective into a refined and forward-looking control sequence for underlying equipment such as heat pumps, water pumps, and valves. This multi-objective optimization control, while fulfilling grid tasks, maximizes the stability and comfort of the indoor environment, and also considers the economy and lifespan of equipment operation, achieving the best balance between grid efficiency, user comfort, and equipment health.

[0017] 5. The technical solution of this invention adopts differentiated control for fluctuation signals of microgrids at different time scales. For frequency fluctuations or smoothing needs at the second to minute level, the flow rate of the variable frequency water pump and the opening of the mixing valve in the optimal control sequence are adjusted first, and the heat capacity of the water in the pipe network is used for rapid response. For energy peak shaving and valley filling or new energy consumption needs at the hour level, the water supply temperature setpoint of the heat pump host in the optimal control sequence is adjusted first, and the sensible heat capacity of the building concrete floor slab is used for energy storage. This hierarchical response strategy can not only meet the grid's rapid response needs at the second and minute levels by adjusting the hydraulic system and provide auxiliary services such as frequency regulation, but also meet the energy time shift needs at the hour level by adjusting the heat source host.

[0018] 6. The technical solution of this invention, when the power dispatch command issued by the microgrid energy management system causes the predicted surface temperature of the radiant terminal to exceed the boundary of the dynamic elastic range of heat and humidity, or causes the expected average thermal sensation index to deviate from the preset threshold of the comfort zone, will forcibly correct the execution power to the maximum safe value that can be allowed by the pre-simulation, and feed back the current actual executable power limit. The dispatch decision of the microgrid is based on the active commitment of the load, and there is a safety pre-simulation and rejection mechanism before execution, which solves the condensation risk and comfort sacrifice problem caused by traditional blind dispatch, and makes it possible for the radiant air conditioning system to participate in grid interaction on a large scale and in a normalized manner.

[0019] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit the invention. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart illustrating the method of Embodiment 1 in the present invention; Figure 2 This is a schematic diagram of the system corresponding to the method in Embodiment 1 of the present invention (in Embodiment 2); Figure 3 This is a flowchart illustrating step S2 in the method of Embodiment 1 of the present invention; Figure 4 This is a schematic diagram illustrating the implementation of step S3 in the method of Embodiment 1 of the present invention; Figure 5 This is a schematic diagram of the dynamic map of microgrid absorption capacity being reported to the microgrid energy management system in step S4 of the method of Embodiment 1 of the present invention. Figure 6This is a schematic diagram illustrating the implementation of step S5 in the method of Embodiment 1 of the present invention. Detailed Implementation

[0022] To clearly illustrate the technical features of this solution, the invention will be described in detail below through specific embodiments and in conjunction with the accompanying drawings. The following disclosure provides many different embodiments or examples for implementing different structures of the invention. To simplify the disclosure of the invention, components and arrangements of specific examples are described below. Furthermore, reference numerals and / or letters may be repeated in different examples. This repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed. It should be noted that the components illustrated in the drawings are not necessarily drawn to scale. Descriptions of well-known components, processing techniques, and processes are omitted in this invention to avoid unnecessarily limiting the invention.

[0023] Example 1 This invention relates to the intersection of building energy management systems, distributed energy microgrid control, and HVAC automatic control technology. Specifically, it relates to a method for controlling the absorption capacity of a radiant air conditioning microgrid based on the dynamic elastic range of building thermal and humidity. This method is mainly applied to zero-energy buildings, campus microgrids, and regional energy internet systems employing radiant cooling / heating systems, and is used to address the coordinated control problem between on-site renewable energy absorption and building environmental comfort and safety. This invention deeply integrates building thermophysics, ergonomics, automatic control theory, and power system dispatching technology, aiming to provide a standardized and quantifiable interactive solution for flexible building loads in the next-generation smart grid.

[0024] like Figures 1-2 As shown, the present invention provides a method for controlling the absorption capacity of a radiant air conditioning microgrid, comprising: S1 collects real-time data on indoor and outdoor temperature and humidity, surface temperature of radiant terminals, water supply and return status, and power flow at the microgrid common connection point. S2. Based on the collected data, combined with the human thermal comfort model and the dew point temperature dynamic evolution model, the upper and lower limits of the dynamic temperature that the radiant terminal surface can operate at are determined, forming a dynamic elastic range of thermal and humidity. S3 transforms the dynamic elastic range of heat and humidity into an equivalent virtual energy storage capacity, maximum absorbable charging / discharging power, and maximum sustainable discharge duration based on the building's heat capacity characteristics and the variable operating condition performance curve of the heat pump unit, thereby generating a dynamic map of the microgrid's absorption capacity. S4, report the dynamic map of microgrid absorption capacity to the microgrid energy management system, and receive power dispatch instructions or frequency response instructions issued by the microgrid energy management system based on the dynamic map of microgrid absorption capacity; S5, based on the received power scheduling command or frequency response command, within the constraint range of the thermal and humidity dynamic elastic range, solves the optimal control sequence of each equipment parameter in the radiant air conditioner through the rolling optimization algorithm, and controls each equipment in the radiant air conditioner based on the optimal control sequence of each equipment parameter in the radiant air conditioner.

[0025] In step S1, this can be achieved through the multi-dimensional environmental and operating condition sensing subsystem 110. First, a multi-dimensional environmental and operating condition sensing subsystem can be set up as the data foundation for the entire control decision, responsible for comprehensively and accurately collecting various physical quantities related to system operation. Its sensor network is meticulously designed, specifically including: meteorological sensors deployed outdoors to measure outdoor dry-bulb temperature, relative humidity, solar radiation intensity, etc., providing boundary conditions for predicting building heat load and evaluating heat pump unit efficiency; composite environmental sensors deployed in typical areas indoors to measure indoor air dry-bulb temperature, relative humidity, and black-bulb temperature, with the black-bulb temperature used to calculate average radiant temperature, thereby more accurately assessing human thermal comfort; high-precision temperature sensor arrays pre-embedded or surface-mounted at key locations in the radiant air conditioning terminals (floors or walls), especially near exterior walls, windows, and other areas prone to low temperatures, to monitor the temperature distribution of the radiant surface in real time and identify local minimum temperature points; temperature and flow sensors installed at key nodes such as the heat pump unit inlet / outlet and manifold in the hydraulic loop of the radiant air conditioning system to monitor supply and return water temperatures, temperature differences, and water flow rates, calculating the real-time cooling / heating power provided by the system; and bidirectional power metering devices configured at the common connection point of the building to the microgrid to monitor the power exchange and flow direction between the building and the microgrid in real time. All sensor data is collected at a high frequency, such as once per minute, from the system's central processor via wired (e.g., Modbus, BACnet) or wireless (e.g., Zigbee, LoRa) communication methods.

[0026] Specifically, a small weather station can be installed on the building roof to collect outdoor dry-bulb temperature, relative humidity, and total solar radiation in real time. In typical functional areas such as open-plan offices, private offices, and meeting rooms on each floor, indoor environmental monitoring terminals integrating temperature, humidity, and black bulb thermometers are installed. During concrete floor slab pouring, at least three representative measuring points are selected on each floor, such as near the south-facing glass curtain wall, the central area of ​​the floor, and near the north-facing exterior wall. PT1000 high-precision temperature sensors are pre-embedded between the radiant coil circuits, with the probes positioned approximately 2 cm from the upper surface of the floor slab. Electromagnetic flow meters and temperature sensors are installed on the supply and return water mains of the air-source heat pump units in the radiant air conditioning system's chiller / heater room. A smart meter supporting bidirectional metering is installed at the common connection point between the microgrid and the main power grid in the building's main power distribution room. All sensors upload data to the system's central controller via the building automation bus (BACnet / IP) at a frequency of 1 minute.

[0027] Preferably, the multi-dimensional sensing subsystem for environment and operating conditions is also equipped with a temperature sensor array embedded in the radiation layer to construct a three-dimensional temperature field distribution map, which facilitates the identification of local low temperature points based on the three-dimensional temperature field distribution map, and uses the local lowest temperature rather than the average temperature as the reference input for anti-condensation control.

[0028] Among them, such as Figure 3 As shown, step S2 (which can be implemented by the thermal-humid dynamic elastic range solution unit 120) specifically includes: S21, calculate the current air dew point temperature using indoor air dry-bulb temperature and relative humidity; Specifically, based on real-time data collected by the multi-dimensional environmental and operating condition sensing subsystem, a dynamic, safe, and comfortable operating range is defined for the surface temperature of the radiating terminal. The calculation process for this range integrates physical safety constraints and human subjective perception constraints.

[0029] On the one hand, to ensure physical safety, namely to prevent condensation, the unit first uses the real-time collected indoor air dry-bulb temperature (T) a The dew point temperature (To) of the current indoor air is calculated using physical formulas (such as a simplified Magnus-Tyton equation) based on the relative humidity (RH) and the relative humidity (RH). dp The formula can be expressed as: T dp ≈T a (100-RH) / 5.

[0030] S22, determine the dynamic cold-side limit temperature of the radiant terminal surface based on the current air dew point temperature and the anti-condensation safety margin factor; Specifically, taking into account measurement errors, uneven airflow, and the buffer time required for system response, a settable anti-condensation safety margin factor (ΔT) is introduced. safe For example, 1.5℃), thus determining the dynamic cold-side limit temperature that the radiating terminal surface must not fall below at any time. This dynamic cold-side limit temperature is a hard constraint to ensure the safe operation of the system. The specific calculation method is as follows: T surf,min (t1)=T dp (t1)+ΔT safe (t1); Among them, T surf,min (t1) represents the dynamic cold-side limit temperature of the radiation terminal surface during the current control cycle t1. dp (t1) represents the air dew point temperature at the current control cycle t1, ΔT safe (t1) is the safety margin coefficient for preventing condensation at the current control cycle t1; Safety margin factor for condensation prevention ΔT safe It can be a fixed empirical value or a dynamic variable composed of multiple sub-items that is calculated in real time. The higher the risk of the system, the greater the margin; the more stable the system, the smaller the margin, thus maximizing the release of the system's cooling potential while ensuring safety.

[0031] ΔT, the safety margin factor for preventing condensation at the current control period t1 safe The specific method for determining (t1) is as follows: ΔT safe (t1)=ΔT err +ΔT dist (t1)+ΔT buf (t1); Where, ΔT err This represents the static measurement error compensation margin; ΔT dist (t1) represents the dynamic spatial non-uniformity compensation margin at the current control period t1; ΔT buf (t1) represents the dynamic response buffer compensation margin at the current control cycle t1.

[0032] Specifically, ΔT err The static measurement error compensation margin is set to the maximum possible calculation error, which has been analyzed and experimentally calibrated and covers a confidence interval of over 99%. This is a static baseline value that is set once and remains unchanged throughout the system's lifetime; for example, after calibration, it can be set to ΔT. err =0.5℃.

[0033] ΔT dist (t1) represents the dynamic spatial non-uniformity compensation margin, calculated using the formula ΔT. dist(t1) = k1 σ Tdp (t). The readings of multiple indoor environmental sensors deployed in different areas are used to quantify the spatial non-uniformity. First, calculate the standard deviation σ of the dew point temperature samples calculated by all sensors at the current time t1 Tdp (t), and then multiply it by the spatial non-uniformity risk coefficient k1. The value of k1 ranges from 2 to 3, representing the degree to which the local dew point deviation follows a normal distribution.

[0034] ΔT buf (t1) is the dynamic response buffer compensation margin, and the calculation formula is ΔT buf (t1) = (dT dp (t1) / dt1), considering the current situation that when the indoor humidity increases rapidly due to certain reasons (such as a large number of people entering, turning on humidification equipment, etc.), the dew point temperature will rise rapidly accordingly. is the thermal response time constant of the radiation system, which characterizes the response speed of the system temperature to changes in control instructions. For a concrete floor radiation system, it takes 1 to 3 hours. dT dp (t1) / dt1 is the change rate of the indoor average dew point temperature, which is obtained by differentiating or fitting the T dp data in the recent period.

[0035] S23. According to the predicted mean thermal sensation index model, set the indoor operating temperature range that meets the requirements of the human thermal comfort level, and inversely deduce the dynamic thermal side limit temperature of the radiation terminal surface in combination with the radiation heat transfer equation; Specifically, to ensure human thermal comfort, the predicted mean vote (PMV) model is introduced. The PMV model combines six parameters, including environmental factors (air temperature, mean radiant temperature, air velocity, relative humidity) and individual factors (human metabolic rate, clothing thermal resistance), to predict the average voting value of a group's perception of the warmth or coldness of a specific thermal environment. Preset the activity level and clothing situation of the personnel according to the building use (such as office, residence), and use the collected data to maintain the PMV index within an acceptable comfort range, such as -0.5 < PMV < +0.5 recommended by the ASHRAE standard. Through this PMV equation, when other parameters are fixed, the indoor operating temperature range corresponding to maintaining the comfortable PMV range can be solved inversely. Due to the characteristics of the radiant air conditioning, there is a strong correlation between the indoor operating temperature and the temperature of the radiation terminal surface. Further, through the basic radiation heat transfer equation, combined with the temperatures of other indoor surfaces, the temperature range of the radiation terminal surface that can maintain the comfortable operating temperature range is inversely calculated. The upper and lower limits of this range constitute the dynamic comfort upper limit temperature or dynamic thermal side limit temperature (T surf,max_comfort ) and the dynamic comfort lower limit temperature (Tsurf,min_comfort ).

[0036] Specifically, the reverse calculation steps are as follows: S231. Based on the expected average thermal sensation index model and the requirements of human thermal comfort level, calculate the upper and lower limits of the operating temperature corresponding to the operating temperature range. First, parameters are input and set. Based on the building's purpose (e.g., office), the metabolic rate of personnel (M, 1.0met) and the thermal resistance of clothing (I) are preset. cl (0.5clo in summer). Indoor air temperature (T) is measured in real-time by sensors. a ), relative humidity (RH) and air velocity (v) ar At the same time, a PMV target range conforming to international standards (ASHRAE 55) is set. target ∈[-0.5, +0.5]. Then, the PMV calculation formula is solved using the bisection numerical iteration method, and the above M and I... cl T a RH and v ar Substituting these known conditions into the PMV equation, and solving iteratively, the corresponding upper and lower operating temperature limits, T, are obtained based on the PMV upper limit of 0.5 and lower limit of -0.5. op,max and T op,min .

[0037] S232, determine the average radiation temperature range based on the operating temperature range; Specifically, at the operating temperature T op In the influence relationship, the average radiation temperature T, which is directly related to the radiation surface, is separated. r In general, indoor environments with low wind speeds, the operating temperature can be accurately approximated as the arithmetic mean of the air temperature and the average radiant temperature, i.e., T. op ≈ (T a +T r ) / 2. Transforming the above formula, we get: T r ≈2 T op T a The T obtained in step S231 op Substituting the upper and lower limits of the interval respectively, we can obtain the dynamic range that the average radiation temperature must satisfy: Upper limit of average radiation temperature: T r,max (t1) =2 T op,max (t1) T a (t1); Lower limit of mean radiation temperature: T r,min (t1) =2 T op,min (t1) T a (t1). Due to T a (t1) changes in real time, so the interval of Tr is also dynamically updated.

[0038] S233, Solving for the radiation terminal surface temperature range based on the average radiation temperature range: Inversely solving for the operating temperature range [T] from the PMV target [T] op,min , T op,max ].

[0039] Mean radiant temperature (T) r The temperature of all indoor surfaces as seen by the human body is determined by its proportion within the human field of vision (i.e., the visual field factor F). p-i A weighted average is then applied. To simplify the calculation, the controllable radiating end surface is defined as T. surf Uncontrollable surfaces such as walls or windows are T uncon_avg T r It can be represented as: Tr ≈ F p-surf T surf +(1 F p-surf ) T uncon_avg ; Among them, F p-surf T is the viewing angle coefficient of the human body at the radiation terminal, a constant determined by the geometry of the room, typically with a radiation surface area of ​​0.3~0.45. uncon_avg This is the average temperature of the uncontrolled surface. In practical engineering, it can be reasonably approximated as equal to the indoor air temperature, i.e., T. uncon_avg ≈T a .

[0040] Then, the above model is subjected to algebraic transformation to solve for the required control variable T. surf ,Right now: T surf ≈ [T r (1 F p-surf ) T a (t1)] / F p-surf ; Finally, T obtained in step S232 r Substituting the range into the above formula, we can obtain the final temperature range that the radiating terminal surface needs to maintain comfort: Comfortable upper limit temperature: T surf,max_comfort (t1)=[T r,max (t1) (1 Fp-surf ) T a (t1)] / F p-surf ; Comfortable lower limit temperature: T surf,min_comfort (t1) =[T r,min (t1) (1 F p-surf ) T a (t1)] / F p-surf ; S24, take the union region between the dynamic cold-side limit temperature and the dynamic hot-side limit temperature as the current dynamic elastic range of temperature and humidity.

[0041] Finally, the previously independently calculated safety and comfort constraints are intelligently integrated logically to precisely define the unique and executable thermo-humidity dynamic elastic range [T] at each control cycle t1. surf,lower (t1),T surf,upper (t1)]. This integration process employs different strategies depending on the operating mode of the radiant air conditioning system (heating or cooling) to ensure optimal decision-making under all operating conditions.

[0042] During the critical cooling season, radiant air conditioning systems must simultaneously mitigate the risks of condensation and compromised thermal comfort. The upper boundary of the comfort zone is set relatively directly, primarily to prevent discomfort caused by overheating; therefore, it is directly taken from the comfort upper limit temperature calculated from the PMV model, i.e., T. surf,upper (t1)=T surf,max_comfort (t1). Determining the lower boundary involves a more precise, double-insurance decision-making process. It is essential to simultaneously satisfy the physical safety baseline—that is, the temperature must not be lower than the anti-condensation constraint T. surf,min (t1), and the physiological comfort baseline—that is, the temperature should not be lower than the comfort lower limit T. surf,min_comfort (t1). To simultaneously satisfy these two mutually restrictive requirements, the more stringent constraint (i.e., higher temperature) must be adopted. This decision logic is perfectly implemented through a maximum value function: T surf,lower (t)=max{T surf,min (t1), T surf,min_comfort (t1)}. This function ensures that when the indoor environment is dry, the decision is dominated by comfort to prevent excessive cold; while when the environment is humid, the decision is dominated by safety, forcibly raising the operating baseline to eliminate the risk of condensation.

[0043] During the heating season, since there is no risk of condensation, the logic is simplified, and the entire focus of the system is on maintaining comfort. At this time, the upper and lower boundaries of the flexibility range are entirely defined by comfort requirements, i.e., [Tsurf,lower (t1),T surf,upper (t1)]=[T surf,min_comfort (t1), T surf,max_comfort (t1)].

[0044] In summary, through this rigorous seasonal integration, the system constructs a dynamic range that adapts to the environment in real time. This range is not fixed; its width and position are frequently adjusted based on fluctuations in the indoor and outdoor environment (personnel load, solar radiation, weather changes). It is this intelligent adaptability based on physical and physiological models that enables the technical solution of this invention to maximize the flexible energy storage potential of buildings while ensuring absolute safety and high comfort, providing a solid and reliable control foundation for the efficient and economical absorption of intermittent renewable energy.

[0045] Among them, such as Figure 4 As shown, the thermo-humid dynamic elastic range defined in the thermophysical domain is translated into electrical domain parameters that the microgrid energy management system can understand and schedule, i.e., an equivalent virtual energy pool model. Step S3 (which can be implemented through the virtual energy pool mapping and conversion unit 130) specifically includes: S31, calculate the amount of heat that can be absorbed or released when the current radiant terminal surface is driven from the average temperature to the boundary of the thermal-moisture dynamic elastic range, and convert the amount of heat that can be absorbed or released into the virtual electrical energy storage capacity on the grid side through the real-time comprehensive performance coefficient of the heat pump unit. Specifically, when calculating the equivalent virtual electrical energy storage capacity, the integral method is used to calculate the sensible heat storage energy corresponding to the temperature change of the building radiant layer material within the dynamic elastic range of heat and humidity. Combined with the real-time energy efficiency ratio (COP) of the heat pump unit under the current outdoor meteorological parameters, the heat value is divided by the real-time energy efficiency ratio to derive the corresponding grid-side dispatchable electrical energy value (kWh). First, calculate the temperature of the current radiating terminal surface from the average temperature T. surf,current Driven to the boundary of the dynamic elastic range of temperature and humidity (e.g., T during the cooling season) surf,min The heat that can be absorbed or released. This heat is mainly stored as sensible heat in the large heat capacity of the radiant layer (such as a concrete floor slab). Its calculation formula is Q. storage =M1 C (T surf,current T surf,minM1 is the effective mass of the radiant layer participating in heat exchange, and C is its specific heat capacity. This is an integral process, and accurate calculation requires consideration of the non-uniform distribution of the temperature field. Then, this heat value (joules or kilowatt-hours of heat) is converted into electrical energy on the grid side. The bridge for this conversion is the real-time comprehensive coefficient of performance (COP) of the heat pump unit. Based on the current outdoor temperature and the system supply and return water temperatures, the current COP value is obtained by consulting the performance curve provided by the heat pump manufacturer or by calling the built-in performance model. Finally, the equivalent absorbable electrical energy E is calculated. electric =Q storage / COP. This E electric The value represents how many kilowatt-hours of electricity the radiant air conditioning system can currently consume, which is the remaining capacity of the virtual energy storage capacity or virtual power pool on the grid side.

[0046] S32, based on the real-time comprehensive performance coefficient of the heat pump unit, maps the maximum heat exchange that the water system circulation flow can support under the current operating conditions to the maximum absorbable charging power; and uses the current total operating power of the radiant air conditioning system as the maximum absorbable discharge power; Specifically, the mapping of maximum charge / discharge power (kW) represents the power throughput capacity of the radiant air conditioning system (i.e., the virtual energy pool). Its upper limit is primarily determined by the physical performance of the heat pump unit, including the compressor's rated maximum power, the upper limit of the inverter's adjustment range, and the maximum heat exchange supported by the water system's circulation flow rate under current operating conditions. This maximum power P max It is the maximum capability of the system to respond instantly to microgrid dispatch commands.

[0047] This invention equates the radiant air conditioning system to a virtual power pool that can be dispatched by a microgrid. The key lies in dynamically mapping the system's maximum heat exchange capacity to the maximum charge / discharge power (P) that the power grid can understand. max This power value is not a fixed parameter on the equipment nameplate, but is determined in real time by the thermal limitations of the system's source, input, and output terminals. Specifically, the maximum charging power (i.e., absorption capacity) of a radiant air conditioning system directly depends on its instantaneous maximum heat exchange capacity Q. max_charge The heat exchange capacity is limited by the minimum of three factors: the maximum heat / cooling capacity of the heat pump, the maximum delivery capacity of the water system, and the current maximum absorption capacity of the building structure. Once this instantaneous Q, representing the physical limit, is determined... max_charge The system then converts this into electrical power using the real-time coefficient of performance (COP), calculated as follows: P charge_max = Qmax_charge / COP. This calculation process ensures that the absorption capacity reported to the microgrid is physically achievable. On the other hand, the system's maximum discharge power (i.e., load shedding capacity) is more directly calculated; it equals the current total operating power P of the radiant air conditioning system. currentThis is because it represents the maximum power that a radiant air conditioning system can instantly release to the power grid by immediately shutting down the equipment.

[0048] S33 utilizes a building thermal decay model to transform the boundary values ​​of the thermal and moisture dynamic elastic range into the maximum sustainable discharge duration.

[0049] Specifically, the duration of inertia maintenance T inertia The mapping represents the endurance of the virtual energy pool. When the radiant air conditioning system is completely shut down by the microgrid command, the cooling or heating energy stored in the building structure will gradually dissipate due to natural heat exchange with the indoor and outdoor environments. A simplified building heat decay model (such as a first-order RC model) is used for mapping: T(t2)=(T initial T ambient ) exp( t2 / )+T ambient ; Where t2 is the independent variable, representing the time elapsed after the system shuts down and begins discharging, i.e., the discharge moment; T(t2) represents the instantaneous temperature of the surface of the radiating terminal (such as a floor slab) after the discharge time t2 has elapsed since the system shut down; T initial It is the temperature at the radiant terminal at the moment of shutdown, and it is the starting point of attenuation; T ambient This is the natural equilibrium temperature that the indoor environment will eventually reach without air conditioning, representing the end point of decay. The instantaneous temperature T(t) of the building's interior environment will change from T... initial infinitely close to T ambient .

[0050] It is the building’s overall thermal time constant, used to predict the time required for the surface temperature of the radiant terminal to decay from its current value to the boundary of the elastic range without active power supply.

[0051] T inertia This refers to the inertial maintenance duration (in hours). Specifically, a target temperature is set, which is the boundary of the thermo-humid dynamic elastic range (e.g., the highest acceptable floor temperature in summer). Then, the value is substituted into the above formula to calculate the time t2 required for the floor temperature to decay from Tinitial to this boundary value. This calculated time t2 is the defined T. inertia It precisely answers the question of how long a room can remain comfortable after the air conditioner is turned off, thus quantifying the virtual battery's endurance. In other words, this time T... inertia It refers to the longest time the system can maintain comfort and safety in discharge mode.

[0052] like Figure 5As shown, through the above three-dimensional mapping, a dynamic map of the microgrid's absorption capacity is generated in real time. This map clearly shows the microgrid the flexibility potential of the building load at the current moment: for example, its ability to absorb P max A total of kilowatts of power can absorb E electric The kilowatt-hours of electricity, if the power supply were to stop now, could still support T inertia Hour.

[0053] In step S4, the source-load bidirectional handshake interaction interface 140 can be used to report the dynamic map of the microgrid absorption capacity to the microgrid energy management system and receive the power scheduling command or frequency response command issued by the microgrid energy management system based on the dynamic map of the microgrid absorption capacity.

[0054] Specifically, a source-load bidirectional handshake interface can be configured. This interface serves as a bridge for information exchange between the system and the microgrid energy management system (EMS), implemented using standardized industrial communication protocols (such as IEC 61850, Modbus TCP / IP, or MQTT). Its operating mode is bidirectional and negotiation-based. On one hand, it periodically and proactively reports the generated dynamic absorption capacity map to the microgrid EMS every 5 minutes. This allows the EMS to accurately understand the real-time state (SOC, P) of the radiant air conditioning virtual energy pool, just as it would a real battery storage unit, when formulating a globally optimal scheduling strategy. max (etc.), thereby making more reasonable power allocation decisions. On the other hand, it is responsible for receiving power dispatch instructions issued by EMS based on this map, such as, please consume energy at a power of 20kW in the next hour or participate in a frequency regulation in the next 10 minutes, with a power fluctuation range of ±5kW.

[0055] More importantly, this interface incorporates an active rejection and correction mechanism. Upon receiving a command from the EMS, the command is not executed blindly but is first sent to the model predictive control execution unit for pre-simulation. If the pre-simulation results show that executing the command will cause the surface temperature of the radiating terminal to exceed the safety boundary of the thermal-humidity dynamic elastic range at some point in the future prediction time domain, or cause the PMV value to exceed the comfort range, this mechanism will be triggered. At this time, the system will not execute the original command but will autonomously correct the execution power to the maximum safe value allowed by the pre-simulation and immediately report back to the EMS through the interactive interface: the command could not be fully executed, a certain amount of power has actually been executed, because the safety / comfort boundary has been touched. This handshake-style safety verification mechanism fundamentally eliminates the possibility of microgrid scheduling causing damage to the building itself and establishes a trust relationship between the source and load.

[0056] In step S5, this can be achieved through the model prediction control execution unit 150, such as... Figure 6 As shown, it is responsible for converting the macroscopic power commands, which have been verified for safety, from the upper layer of the microgrid into a refined and optimized sequence of control actions for the lower-level equipment (heat pumps, water pumps, valves). The core of Model Predictive Control (MPC) lies in its accurate predictive capability based on a nonlinear building thermal dynamics model. By constructing a nonlinear state-space model based on the RC network method, it accurately characterizes the thermal dynamic behavior of the building and radiation system. It can comprehensively deduce the system's state evolution trajectory, typically within a predicted time domain of 2 to 4 hours, by combining the real-time states such as the current temperature of various parts of the system with predicted values ​​of future disturbances such as outdoor temperature and solar radiation.

[0057] The control process of MPC is an iterative rolling optimization process. In each control cycle, such as every 15 minutes, MPC solves a multi-objective optimization problem. Its objective function typically includes several weighted terms: first, to track the power target curve issued by the microgrid, so that the actual power consumed by the system is as close as possible to the commanded value; second, to minimize the variation in control actions, avoiding frequent start-ups and drastic changes in equipment to extend equipment lifespan; and third, to maintain indoor comfort, so that the PMV value is as close as possible to the ideal value of 0. During the optimization process, a series of strict constraints must be met: first, state constraints, meaning that all predicted radiant terminal surface temperatures must be within the thermal-humid dynamic elastic range throughout the entire prediction time domain; second, control constraints, meaning that the output values ​​of control variables such as heat pump compressor frequency, water pump flow rate, and mixing valve opening must be within their physically adjustable range.

[0058] By solving this constrained optimization problem, MPC can obtain the optimal control sequence for a future period, such as the compressor frequency and water pump speed setpoints every 15 minutes within the next hour. However, MPC only executes the first control action in this sequence, and then repeats the entire prediction and optimization process in the next control cycle based on the latest system measurements. This rolling optimization and feedback correction mechanism enables the MPC controller to effectively cope with model mismatch and unforeseen disturbances, exhibiting excellent robustness and control accuracy.

[0059] Specifically, the nonlinear building thermal dynamics model takes state data, equipment control sequences, and disturbance data as inputs, uses the power curve issued by the microgrid as the target tracking trajectory, and uses the thermal and humidity dynamic elastic range as the state constraint condition. It solves the optimal control sequence of each equipment parameter in the radiant air conditioning system in the future control time domain by minimizing the objective function. Among them, the state data includes the average indoor air temperature and the average surface temperature of the radiant terminal, and the disturbance data includes the outdoor air temperature, the solar radiation heat gain entering the room through the windows, and the internal heat sources generated by indoor personnel and equipment.

[0060] Input data for nonlinear building thermal dynamics models may include: 1. State data x(k), which is the starting point of the nonlinear building thermal dynamics model, comes from real-time measurements of the environmental and operational condition multi-dimensional sensing subsystem 110. At discrete time point k (e.g., 11:00 AM), x(k) includes: T a (k): The average temperature of the indoor air.

[0061] T s (k): Average surface temperature at the radiating end.

[0062] Therefore, the state vector is x(k)=[T a (k), T s (k)] T .

[0063] 2. Equipment control action u(k), this is the variable to be optimized, representing the cooling / heating power input from the radiant air conditioning system to the building. In cooling mode, it is determined by the electrical power P of the heat pump unit. elec (k) and real-time energy efficiency ratio COP(k) determine: u(k) = Q hvac (k)= P elec (k) COP(k) (negative for cooling); In the formula, P elec (k) comes from the power command data, that is, the power command data issued by the EMS.

[0064] 3. Environmental / disturbance data d(k): These are uncontrollable external influences that the model needs to consider when predicting the future. This data comes partly from real-time sensing and partly from weather forecasts or preset models. T out (k): Outdoor air temperature.

[0065] Q sol (k): Heat gained from solar radiation entering the room through the window.

[0066] Q int (k): Internal heat sources generated by indoor personnel, equipment, etc.

[0067] The disturbance vector is d(k)=[T out (k),Q sol (k),Q int (k)] T .

[0068] The nonlinear building thermal dynamics model is as follows: x(k+1)=A x(k)+B u u(k)+B d d(k); Where k represents the current time; i+1 represents the next time; x(i+1) is the predicted state data for the next time; x(i) is the state data for the current time; A is the internal state matrix; B u For external control matrix; B d The external perturbation matrix; These matrices are determined by the building's physical parameters (thermal resistance R and heat capacity C), and in this invention, they are calibrated and corrected through initial building information modeling and subsequent adaptive learning. Matrix A describes how floor and air temperatures interact and naturally decay in the absence of external forces. Matrix B u The matrix describes how the input cooling / heating power u(k) of the air conditioning system changes the temperature of the floor and the air. B d The matrix describes how external disturbances d(k), such as outdoor temperature and solar radiation, affect the temperature of the floor and the air.

[0069] The objective function is as follows: ; Where J is the objective function value, i is the prediction step number, k is the initial prediction time, and N is the number of time steps in the prediction time domain; Let i be the actual electrical power consumed by the radiant air conditioning system at prediction step i. The power command issued by the microgrid at prediction step i; w p It is the power tracking weight; T op (i) is the predicted indoor operating temperature at prediction step i; T set It is the ideal comfort temperature set by the user; w c It is a comfort weight; For the control action at prediction step i, For the control action at prediction step i-1; w u It controls the smoothing weights.

[0070] When the MPC execution unit 150 performs rolling optimization to solve for the optimal control sequence, its performance is judged by the objective function J. The goal of MPC is to find a set of future control action sequences {u(k), u(k+1), ..., u(k+N)}. 1)}, which minimizes the value of the objective function J. According to the technical solution and embodiments of the present invention, this objective function needs to comprehensively consider three aspects: power grid command tracking, user comfort, and stable equipment operation.

[0071] The three components of the objective function will be explained in detail below: 1. Power Tracking Item: w p (P elec (i) P ref (i)) 2 ; The purpose of this is to ensure that the actual electrical power P consumed by the system is within the acceptable range. elec Accurately track the power command P issued by the microgrid EMS. ref This directly corresponds to the objective of tracking the power target curve issued by the microgrid in the invention.

[0072] P elec (i) is the predicted power consumption of the system at prediction step i, which is directly related to the control variable u(i): P elec (i)=|u(i)| / COP(i).

[0073] P ref (i) is the i-th value in the power command sequence issued by the microgrid EMS, that is, the power command issued by the microgrid at prediction step i. For example, in the embodiment, it is a constant 40kW.

[0074] w p It refers to the power tracking weight. In scenarios where priority must be given to grid dispatching, w p It will be set to a larger value.

[0075] 2. Indoor environmental comfort maintenance item: w c (T op (i) T set ) 2 ; The purpose of this invention is to maximize the thermal comfort of indoor occupants while performing power grid tasks. This corresponds to the invention's goal of maintaining indoor comfort, i.e., keeping the PMV value as close as possible to the ideal value of 0.

[0076] T op (i) is the predicted indoor operating temperature, which is determined by the state variable T predicted by the model. a (i) and T s (i) Calculation yields (Top≈(T) a +T s ) / 2).

[0077] T set It is the ideal comfort temperature set by the user (for example, the operating temperature corresponding to PMV=0 calculated by the PMV model, such as 25.5℃).

[0078] w c This is a comfort-based weighting. If the user has extremely high comfort requirements, w c The value will be very large, which makes MPC willing to sacrifice some power tracking accuracy in order to ensure comfort.

[0079] 3. Control motion smoothness: w u (u(i) u(i 1)) 2 ; The purpose of this is to avoid excessively frequent and drastic control of equipment (such as heat pumps), thereby reducing equipment wear and extending its service life. This corresponds to the objective of minimizing the variation in control actions as stated in the invention.

[0080] u(i) u(i 1) represents the change in the cooling / heating power output of the air conditioning system between two adjacent control steps. MPC tends to select a smoother, less variable control sequence by minimizing this term. u It controls the smoothing weights.

[0081] Constraints: During the minimization of the objective function J, the optimization solver must always satisfy the constraint of the thermo-hygroscopic dynamic elastic range calculated in the embodiment. This is a hard constraint that has a veto power. T surf_lower (i)≤T s (i)≤T surf_upper (i) For all prediction steps i.

[0082] In summary, the MPC execution unit 150 finds a control sequence that minimizes the weighted total cost J by solving this constrained optimization problem. This process is like a smart housekeeper making the most comprehensive decision after weighing the requirements of the power grid, the user's experience, and the wear and tear on home appliances. This is precisely the core intelligence behind the safe, efficient, and comfortable collaborative control achieved in this invention.

[0083] In step S5, the control of each device in the radiant air conditioner based on the optimal control sequence of each device parameter is specifically as follows: Differentiated control is adopted for fluctuation signals of microgrid at different time scales. For frequency fluctuations or smoothing needs at the second to minute level, the flow rate of variable frequency water pump and the opening of mixing valve in the optimal control sequence are adjusted first, and the heat capacity of water in the pipeline network is used for rapid response. For energy peak shaving and valley filling or new energy consumption needs at the hour level, the water supply temperature setpoint of heat pump host in the optimal control sequence is adjusted first, and the sensible heat capacity of building concrete floor slab is used for energy storage.

[0084] Preferably, to further improve the system's response speed and precision, a hierarchical response strategy can be implemented. That is, different regulatory resources are allocated based on the time scale of the microgrid signal. For rapid frequency fluctuations or power stabilization needs at the second to minute level, i.e., primary frequency regulation or rapid secondary frequency regulation services, the system prioritizes adjusting the speed of the variable frequency circulating water pump or the opening of the bypass mixing valve in the optimal control sequence. This is because changing the water flow rate can almost instantaneously change the flow rate of the heat medium flowing through the radiant coils, thereby quickly fine-tuning the system's heat exchange power and power consumption. This utilizes the small heat capacity of the water itself in the pipe network, resulting in a fast response. For hourly-level, large-scale peak shaving and valley filling or large-scale renewable energy consumption needs, the system prioritizes adjusting the operating status of the heat pump unit in the optimal control sequence, such as changing its outlet water temperature setpoint or starting / stopping the unit, to deeply utilize the huge sensible heat storage capacity of the building floor structure, achieving large-scale energy time shifting. This hierarchical strategy achieves decoupling and synergy between rapid response and deep regulation.

[0085] Furthermore, the present invention also provides a method for controlling the absorption capacity of a radiant air conditioning microgrid, which further includes: S6, when the power dispatch command issued by the microgrid energy management system causes the predicted surface temperature of the radiating terminal to exceed the boundary of the thermal and humidity dynamic elastic range, or causes the expected average thermal sensation index to deviate from the preset threshold of the comfort zone, the execution power will be forcibly corrected to the maximum safe value that can be allowed by the pre-simulation, and the current actual executable power limit will be fed back upward.

[0086] Specifically, the active rejection mechanism of the source-load two-way handshake interaction interface 140 is triggered, determining that the original command poses a risk and cannot be fully executed. It will autonomously correct the command. The MPC re-optimizes and calculates the maximum average power absorption that can be achieved under the boundary constraints of the dynamic elastic range of floor temperature and humidity, or under the constraint of not being lower than the preset threshold that would cause the expected average thermal sensation index to deviate from the comfort zone. This corrected and safe control strategy is then sent to the underlying equipment for execution. At the same time, a message is fed back to the EMS through the source-load two-way handshake interaction interface 140: the original command has a comfort risk, has been corrected and executed, and the current actual average power is [not specified]. In this way, a safe, closed-loop, and negotiated source-load interaction is completed.

[0087] Preferably, by deploying a temperature sensor array embedded inside the radiation layer in the multi-dimensional environmental and operating condition sensing subsystem 110, the discrete point measurement can be upgraded to the reconstruction and monitoring of the three-dimensional temperature field inside the radiation layer. Step S2 will no longer use the average surface temperature as the basis for anti-condensation judgment, but will use the local lowest temperature point identified in the real-time constructed temperature field cloud map as the reference input. This greatly improves the reliability and safety of anti-condensation control, especially at the location of cold bridges in buildings.

[0088] Finally, adaptive learning functionality can be integrated. By recording and analyzing historical operating data over a long period, machine learning algorithms, such as regression analysis and neural networks, can be used to identify and correct key model parameters online. For example, by comparing the deviation between the microgrid's commanded power and the actual executed power, the true efficiency degradation of the heat pump under different operating conditions can be learned; by analyzing the system's temperature response curve to a step input, the building's thermal time constant (τ2) and heat capacity (M2) can be more accurately determined. C). This self-learning capability enables the mapping relationship of the virtual energy pool and the prediction accuracy of the MPC model to continuously improve over time, thereby allowing the evaluation and control performance of the entire system to achieve adaptive optimization.

[0089] To illustrate this solution more clearly, the following example uses an office building employing a floor radiant cooling system and a rooftop photovoltaic power generation system to explain in detail the operation process of the radiant air conditioning microgrid absorption capacity control described in this invention. This building constitutes a small microgrid, including photovoltaic power generation units, radiant air conditioning load units, conventional office loads, and connection points to the main power grid.

[0090] A small weather station is installed on the building roof to collect real-time outdoor dry-bulb temperature, relative humidity, and total solar irradiance. In typical functional areas such as open-plan offices, private offices, and meeting rooms on each floor, indoor environmental monitoring terminals integrating temperature, humidity, and black bulb thermometers are installed. During concrete floor slab pouring, at least three representative measuring points are selected on each floor, such as near the south-facing glass curtain wall, the central area of ​​the floor, and near the north-facing exterior wall. PT1000 high-precision temperature sensors are pre-embedded between the radiant coil circuits, with the probes positioned approximately 2 cm from the upper surface of the floor slab. Electromagnetic flow meters and temperature sensors are installed on the supply and return water mains of the air-source heat pump units in the radiant air conditioning system's chiller / heater room. In the building's main power distribution room, a smart meter supporting bidirectional metering is installed at the common connection point between the microgrid and the main power grid. All sensors upload data to the central controller via the building automation bus (BACnet / IP) at a frequency of 1 minute.

[0091] Taking 9:00 AM on a summer weekday as an example, the collected data shows an average indoor air temperature of 26℃ and an average relative humidity of 60%. First, the safety boundary for preventing condensation is calculated. Based on 26℃ and 60%RH, the dew point temperature T of the indoor air at this moment is calculated. dp Approximately 17.9℃. Set an anti-condensation safety margin ΔT. safe If the temperature is 2℃, then the dynamic cold-side limit temperature T of the radiant floor slab surface is obtained. surf,min (t) = 17.9℃ + 2℃ = 19.9℃. This means that, in order to absolutely prevent condensation, the surface temperature of the floor slab must not be lower than 19.9℃ under any circumstances.

[0092] Simultaneously, the comfort boundary was calculated. The preset metabolic rate of office occupants was 1.2 MET, the thermal resistance of summer clothing was 0.5 clo, and the indoor air velocity was below 0.1 m / s. To maintain the PMV index at […] Within the comfort range of [0.5, +0.5], iterative calculations using the PMV equation yielded an indoor operating temperature range of [24.5℃, 27.0℃]. Considering that the indoor operating temperature is approximately equal to the average of the air temperature and the mean radiant temperature, and the mean radiant temperature is primarily determined by the floor surface temperature, a simplified heat balance equation can be used to calculate that to achieve this operating temperature range, the floor surface temperature should be maintained between approximately [21.5℃, 24.0℃].

[0093] Reference Figure 3 The logic integrates the two constraints mentioned above. The lower limit for preventing condensation is 19.9℃, while the lower limit for comfort is 21.5℃. To satisfy both simultaneously, the larger value must be taken as the final operating lower limit. Therefore, at 9:00 AM, the system's determined dynamic temperature and humidity range is [21.5℃, 24.0℃]. This range will serve as the fundamental basis for all subsequent evaluations and controls.

[0094] Next, refer to Figure 4 The illustration assumes that at this moment, the average surface temperature T of the floor slab is measured. surf,current It is 23.5℃.

[0095] 1. Capacity Mapping: The current temperature is 23.5℃, with a cooling space of 2.0℃ below the lower limit of the elastic range (21.5℃). The effective cold storage layer of the office building's radiant floor is a 10cm thick concrete layer with a total area of ​​2000 square meters. The density of concrete is approximately 2400 kg / m³, and its specific heat capacity is approximately 0.92 kJ / (kg·K). Therefore, the amount of cold that the building's floor can store within this temperature difference is: Q storage =(2000m² 0.1m 2400kg / m³) 0.92 kJ / (kg·K) 2.0K≈883,200kJ≈245.3kWh; At this moment, the outdoor temperature is 32℃. Under this operating condition, the heat pump unit's supply water temperature is set to 18℃, and the COP value, obtained from the equipment performance curve, is assumed to be 3.5. Therefore, converting this heat value into electrical energy on the grid side, the equivalent absorbing electricity E... electric =245.3kWh / 3.5≈70.1kWh. This means that the building can also absorb about 70.1 kWh of electricity to reduce the floor temperature to a safe and comfortable lower limit.

[0096] 2. Power Mapping: The air-source heat pump configured in this building has a maximum cooling input power of 50kW and a frequency conversion range of 30%-100%. Therefore, the maximum absorbable charging power P max =50kW.

[0097] 3. Time Mapping: Based on historical data and a simplified building model, the building's thermal time constant τ2 is calculated to be approximately 10 hours. After cooling is stopped, it takes approximately 5 hours for the floor temperature to recover from 21.5℃ to the comfort upper limit of 24.0℃. Therefore, the inertial maintenance duration T... inertia ≈5 hours.

[0098] These three key parameters {P max 50kW, E electric 70.1 kWh, T inertia The data is packaged into a single data object, forming a real-time dynamic graph of the microgrid's absorption capacity, such as... Figure 5 As shown.

[0099] Then, the source-load bidirectional handshake interaction interface 140 is used to report the dynamic map of the microgrid's absorption capacity to the microgrid energy management system. When the handshake interaction is conducted via the MQTT protocol, this data object is published to the microgrid energy management system (EMS). At 11:00 AM, the photovoltaic power generation reaches its peak, far exceeding the building's regular office load, resulting in a power surplus in the microgrid. The EMS needs to find a way to absorb the surplus. The EMS subscribes to the building load capacity topic and receives the aforementioned data. It discovers that the building's radiant air conditioning is an available virtual battery with a remaining capacity of 70.1 kWh and can be charged at a maximum power of 50 kW. Therefore, the EMS issues a scheduling instruction to system 100: to operate at a constant power of 40 kW from 11:00 AM to 12:00 PM to absorb the photovoltaic power.

[0100] Upon receiving the instruction, the model predictive control execution unit 150 first performs a safety rehearsal of the instruction. (Refer to...) Figure 6The building thermal model built into the MPC unit simulates the consequences of continuous cooling at 40kW for one hour, with a prediction time domain of 2 hours and a control step size of 15 minutes. The model predicts that in about 50 minutes, around 11:50, the local lowest temperature point of the floor slab, such as the predicted temperature near the south-facing curtain wall, will reach 21.4℃, which is lower than the elastic lower limit of 21.5℃.

[0101] A typical area of ​​an office building is simplified into a second-order (2R2C) model that includes two key thermal capacity nodes: indoor air and radiant floor.

[0102] 1. Model input data: 1) State data x(k), which is the starting point of the model, comes from real-time measurements of the environmental and operating condition multi-dimensional sensing subsystem 110. At discrete time point k (e.g., 11:00 AM), x(k) includes: T a (k): The average temperature of the indoor air.

[0103] T s (k): Average surface temperature at the radiating end.

[0104] Therefore, the state vector is x(k)=[T a (k), T s (k)] T .

[0105] 2) Equipment control action u(k), this is the variable to be optimized, representing the cooling / heating power input from the radiant air conditioning system to the building. In cooling mode, it is determined by the electrical power P of the heat pump unit. elec (k) and real-time energy efficiency ratio COP(k) determine: u(k) = Q hvac (k)= P elec (k) COP(k) (negative for cooling); In the formula P elec (k) comes from the power command data, namely the 40kW issued by the EMS.

[0106] 3) Environmental / disturbance data d(k): This represents the uncontrollable external influences that the model needs to consider when predicting the future. This data comes partly from real-time sensing and partly from weather forecasts or preset models. T out (k): Outdoor air temperature.

[0107] Q sol (k): Heat gained from solar radiation entering the room through the window.

[0108] Qint (k): Internal heat sources generated by indoor personnel, equipment, etc.

[0109] The disturbance vector is d(k)=[T out (k),Q sol (k),Q int (k)] T .

[0110] 2. Detailed description of the model formula: Based on the law of conservation of energy, differential equations can be established for the two nodes, and these equations can be discretized to obtain the following discrete-time state-space equations: x(k+1)=A x(k)+B u u(k)+B d d(k); In the formula, k represents the current time step, and k+1 represents the next time step (e.g., 15 minutes later). x(k+1) are the indoor air temperature and floor temperature predicted by the model for the next moment. A, B u and B d These are system matrices. These matrices are determined by the building's physical parameters (thermal resistance R and heat capacity C), and in this invention, they are calibrated and corrected through initial building information modeling and subsequent adaptive learning. Matrix A describes how floor and air temperatures interact and naturally decay in the absence of external forces. Matrix B... u The matrix describes how the input cooling / heating power u(k) of the air conditioning system changes the temperature of the floor and the air. B d The matrix describes how external disturbances d(k), such as outdoor temperature and solar radiation, affect the temperature of the floor and the air.

[0111] 3. The relationship between the final output and the input data: The goal of the MPC execution unit 150 is to predict the state in a future time domain (e.g., N=8 steps, or 2 hours). It calculates this through an iterative autoregressive approach using the aforementioned state-space equations: The first step is to predict (k->k+1), and the calculation formula is as follows: x(k+1|k)=A x(k)+B u u(k)+B d d(k); In the formula, x(k) is the measured value, and u(k) and d(k) are the control and disturbance values ​​for the current step. x(k+1|k) represents the prediction of the state at time k+1 at time k.

[0112] The second step, prediction (k+1->k+2), is calculated using the following formula: x(k+2|k)=A x(k+1|k)+B u u(k+1)+B d d(k+1); In the formula, x(k+1|k) is the prediction result of the previous step, and u(k+1) and d(k+1) are the predicted values ​​of control and disturbance for the first step in the future. ... Continue in this manner until the Nth step.

[0114] Finally, the output of the MPC unit is a state prediction sequence containing N future time steps, calculated as follows: {x(k+1|k), x(k+2|k), .., x(k+N|k)}; In the formula, each term x(k+i|k) in the sequence contains the predicted future floor temperature T. s (k+i|k) and air temperature T a (k+i|k). The relationship between this prediction sequence and the input data is as follows: it begins with the current state data x(k) and evolves in response to a series of future power command data {u(k)...u(k+N-1)} and environmental data {d(k)...d(k+N-1)}.

[0115] In this embodiment, when the prediction shows that approximately 50 minutes later, around 11:50, the local minimum temperature point of the floor slab will reach 21.4℃, this 21.4℃ is the predicted floor slab temperature T corresponding to k+3 (the fourth 15-minute step, approximately 50 minutes later) in the above prediction sequence. s (k+3|k). This prediction clearly demonstrates the consequences of executing a 40kW power command, thus providing a basis for subsequent decision-making.

[0116] When the MPC execution unit 150 performs rolling optimization to solve for the optimal control sequence, the criterion for judging its performance is the objective function J. The goal of MPC is to find a set of future control action sequences {u(k), u(k+1), ..., u(k+N-1)} that minimizes the value of the objective function J. According to the technical solution and embodiments of the present invention, this objective function needs to comprehensively consider three aspects: grid command tracking, user comfort, and stable equipment operation.

[0117] In summary, the MPC execution unit 150 finds a control sequence that minimizes the weighted total cost J by solving this constrained optimization problem. This process is like a smart housekeeper making the most comprehensive decision after weighing the requirements of the power grid, the user's experience, and the wear and tear on home appliances. This is precisely the core intelligence behind the safe, efficient, and comfortable collaborative control achieved in this invention.

[0118] At this point, the active rejection mechanism of the source-load bidirectional handshake interaction interface 140 is triggered. It is determined that the original command poses a risk and cannot be fully executed. It will autonomously correct the command. The MPC recalculates to determine the maximum average power absorption under the constraint of a floor temperature not lower than 21.5℃. The calculation result may be 40kW operation for the first 45 minutes, decreasing to 15kW for the next 15 minutes, or a smoother power reduction curve. System 100 sends this corrected, safe control strategy to the underlying equipment for execution. Simultaneously, a message is fed back to the EMS through the source-load bidirectional handshake interaction interface 140: the original command (40kW for 1h) poses a comfort risk and has been corrected. The current actual average power is 34kW. This completes a safe, closed-loop, negotiated source-load interaction.

[0119] At the execution level, the calculated optimal power target (e.g., an average power of 38kW over the next 15 minutes) is broken down into specific control actions for the heat pump compressor and circulating water pump. It might set the heat pump compressor to operate at 85% of its frequency while simultaneously increasing the speed of the secondary-side variable frequency water pump to 90% to ensure efficient delivery of cooling capacity to the floor slab, thus precisely matching the 38kW power consumption target. After 15 minutes, the MPC will perform a new round of rolling optimization based on the latest building condition measurements, continuing to execute control actions for the next 15 minutes.

[0120] Furthermore, if the microgrid experiences a frequency drop due to sudden cloud cover changes, a rapid response from the load side is required. The hierarchical response strategy module will then activate. It will temporarily suspend hourly energy dispatch targets and prioritize responding to second-level frequency support commands. It will instantly reduce the speed of the circulating water pumps (e.g., from 90% to 70%), causing the total system power consumption to drop rapidly within seconds, providing the necessary frequency regulation support to the grid. Once the frequency stabilizes, the original energy dispatch tasks will resume.

[0121] Through the aforementioned series of precise and interconnected steps, a static building floor slab in the form of a radiant air conditioning system can be transformed into a dynamic virtual energy storage unit that can intelligently interact with the power grid. While ensuring its own absolute safety and comfort, it makes the greatest contribution to absorbing unstable renewable energy.

[0122] In this invention, based on collected data and combined with a human thermal comfort model and a dew point temperature dynamic evolution model, the upper and lower limits of the dynamic temperature that the radiant terminal surface can operate at are determined, forming a dynamic elastic range of thermal and humidity. This dynamic elastic range is then transformed into equivalent virtual energy storage capacity, maximum absorbable charging / discharging power, and maximum sustainable discharge duration, based on building thermal capacity characteristics and the variable operating condition performance curve of the heat pump unit. According to the received power scheduling command or frequency response command, within the constraints of the dynamic elastic range of thermal and humidity, the optimal control sequence of each device parameter in the radiant air conditioner is solved using a rolling optimization algorithm. Based on this optimal control sequence, each device in the radiant air conditioner is controlled, realizing a shift from passive response to active guidance. This improves the safety and reliability of source-load interaction, effectively solving the problem of low reliability in the control of the absorption capacity of the radiant air conditioner microgrid caused by existing technologies, and effectively improving the reliability of the absorption capacity control of the radiant air conditioner microgrid.

[0123] In this invention, the current air dew point temperature is calculated using indoor air dry-bulb temperature and relative humidity. Based on the current air dew point temperature and the anti-condensation safety margin coefficient, the dynamic cold-side limit temperature of the radiant terminal surface is determined. According to the predicted average thermal sensation index model, an indoor operating temperature range that meets the requirements of human thermal comfort level is set, and the dynamic hot-side limit temperature of the radiant terminal surface is inversely calculated by combining the radiant heat transfer equation. The union region between the dynamic cold-side limit temperature and the dynamic hot-side limit temperature is taken as the current dynamic elastic range of heat and humidity, thus defining a clear anti-condensation safety and comfort boundary for the flexible adjustment of radiant air conditioning.

[0124] In this invention, the technical solution calculates the heat that can be absorbed or released when the current radiant terminal is driven from the average temperature to the boundary of the elastic range. The heat value that can be absorbed or released is converted into the virtual electrical energy storage capacity on the grid side through the real-time comprehensive performance coefficient of the heat pump unit. Based on the real-time comprehensive performance coefficient of the heat pump unit, the maximum heat exchange that the water system circulation flow can support under the current operating conditions is mapped to the maximum absorbable charging power. The current total operating power of the radiant air conditioning system is used as the maximum absorbable discharge power. Using the building heat attenuation model, the boundary value of the thermal and humidity dynamic elastic range is converted into the maximum sustainable discharge duration. On the one hand, through the inverse mapping of the virtual energy pool, the complex physical constraints are transformed into a more sustainable discharge duration. The system translates loads into a language understandable to the power grid, enabling proactive reporting of load-side absorption capacity. Furthermore, it provides a unified and standardized way to describe the flexibility of radiant air-conditioned buildings of different types and sizes. This standardized interface greatly simplifies the development and deployment of microgrid energy management systems, allowing them to connect to and manage large numbers of flexible building loads like plug-and-play. This lays a solid technical foundation for building larger-scale virtual power plants and regional energy internet, with broad application prospects and promotional value. Moreover, it equates the building's huge heat capacity to a virtual energy pool, providing microgrids with a large-capacity energy storage option at almost zero cost. During periods of high photovoltaic power generation, the system can accurately calculate the amount of electricity that can be safely absorbed and guide radiant air conditioning systems to charge, for example, through pre-cooling or pre-heating, converting potentially wasted green electricity into useful heat energy for storage. This not only effectively mitigates the fluctuations in renewable energy and improves its utilization rate but also significantly reduces the need for and dependence on expensive electrochemical energy storage systems, bringing significant economic benefits to the construction and operation of microgrids.

[0125] This invention establishes a nonlinear building thermal dynamics model. Using state data, equipment control sequences, and disturbance data as inputs, and the power curve from the microgrid as the target trajectory, and the thermal-humidity dynamic elastic range as the state constraint, the optimal control sequence for each device parameter in the radiant air conditioning system within the future control time domain is solved by minimizing the objective function. The state data includes the average indoor air temperature and the average surface temperature of the radiant terminals. The disturbance data includes the outdoor air temperature, solar radiation heat gain entering the room through windows, and internal heat sources generated by indoor occupants and equipment. This approach decomposes the macroscopic grid dispatch objective into refined and forward-looking control sequences for underlying equipment such as heat pumps, water pumps, and valves. This multi-objective optimization control, while fulfilling grid tasks, maximizes the stability and comfort of the indoor environment, and considers the economy and lifespan of equipment operation, achieving the best balance between grid efficiency, user comfort, and equipment health.

[0126] The technical solution of this invention adopts differentiated control for fluctuation signals of microgrids at different time scales. For frequency fluctuations or smoothing needs at the second to minute level, the flow rate of the variable frequency water pump and the opening of the mixing valve in the optimal control sequence are adjusted first, and the heat capacity of the water in the pipeline network is used for rapid response. For energy peak shaving and valley filling or new energy consumption needs at the hour level, the water supply temperature setpoint of the heat pump host in the optimal control sequence is adjusted first, and the sensible heat capacity of the building concrete floor slab is used for energy storage. This hierarchical response strategy can not only meet the grid's rapid response needs at the second and minute levels by adjusting the hydraulic system and provide auxiliary services such as frequency regulation, but also meet the energy time shift needs at the hour level by adjusting the heat source host.

[0127] When the power dispatch command issued by the microgrid energy management system causes the predicted surface temperature of the radiant terminal to exceed the boundary of the dynamic elastic range of heat and humidity, or causes the expected average thermal sensation index to deviate from the preset threshold of the comfort zone, the technical solution of this invention will forcibly correct the execution power to the maximum safe value that can be allowed by the pre-simulation, and feed back the current actual executable power limit. The dispatch decision of the microgrid is based on the active commitment of the load, and there is a safety pre-simulation and rejection mechanism before execution. This solves the condensation risk and comfort sacrifice problem caused by traditional blind dispatch, and makes it possible for the radiant air conditioning system to participate in grid interaction on a large scale and in a normalized manner.

[0128] Example 2 like Figure 2 As shown, the present invention also provides a radiant air conditioning microgrid absorption capacity control system, comprising: The environmental and operating condition multi-dimensional sensing subsystem 110 collects real-time data on indoor and outdoor temperature and humidity, surface temperature of radiant terminals, water supply and return status, and power flow data at the microgrid common connection point. The thermal and humidity dynamic elastic range calculation unit 120, based on the collected data and combined with the human thermal comfort model and the dew point temperature dynamic evolution model, determines the upper and lower limits of the dynamic temperature that the radiant terminal surface can operate at, thus forming the thermal and humidity dynamic elastic range. The virtual electric energy storage mapping and conversion unit 130 converts the thermal and humidity dynamic elastic range based on the building thermal capacity characteristics and the variable operating condition performance curve of the heat pump unit into an equivalent virtual electric energy storage capacity, maximum absorbable charging power and maximum sustainable discharge duration, thereby generating a dynamic map of the microgrid absorption capacity. The source-load bidirectional handshake interaction interface 140 reports the dynamic map of microgrid absorption capacity to the microgrid energy management system and receives power dispatch instructions or frequency response instructions issued by the microgrid energy management system based on the dynamic map of microgrid absorption capacity. The model predictive control execution unit 150, based on the received power scheduling command or frequency response command, solves the optimal control sequence of each device parameter in the radiant air conditioner through a rolling optimization algorithm within the constraint range of the thermal and humidity dynamic elastic range, and controls each device in the radiant air conditioner based on the optimal control sequence of each device parameter.

[0129] An environmental and operating condition multidimensional sensing subsystem 110 is the data foundation for the entire assessment and control decision-making process, responsible for comprehensively and accurately collecting various physical quantities related to system operation. Its sensor network is meticulously designed, specifically including: meteorological sensors deployed outdoors to measure outdoor dry-bulb temperature, relative humidity, solar radiation intensity, etc., providing boundary conditions for predicting building heat load and evaluating heat pump unit efficiency; composite environmental sensors deployed in typical areas indoors to measure indoor air dry-bulb temperature, relative humidity, and black-bulb temperature, with the black-bulb temperature used to calculate average radiant temperature, thereby more accurately assessing human thermal comfort; high-precision temperature sensor arrays pre-embedded or surface-mounted at key locations in the radiant air conditioning terminals (floors or walls), especially near exterior walls, windows, and other areas prone to low temperatures, to monitor the temperature distribution of the radiant surface in real time and identify local minimum temperature points; temperature and flow sensors installed at key nodes such as the heat pump unit inlet / outlet and manifold in the hydraulic loop of the radiant air conditioning system to monitor supply and return water temperatures, temperature differences, and water flow rates, calculating the real-time cooling / heating power provided by the system; and bidirectional power metering devices configured at the common connection point of the building to the microgrid to monitor the power exchange and flow direction between the building and the microgrid in real time. All sensor data is collected at a high frequency, such as once per minute, from the system's central processor via wired (e.g., Modbus, BACnet) or wireless (e.g., Zigbee, LoRa) communication methods.

[0130] Secondly, the thermal and humidity dynamic elastic range calculation unit 120 is one of the key technologies of this invention. Its function is to define a dynamic, safe, and comfortable operating range for the surface temperature of the radiant terminal based on real-time data collected by the sensing subsystem. The calculation process of this range integrates physical safety constraints and human subjective perception constraints.

[0131] The core task of the thermal and humidity dynamic elastic range calculation unit 120 is to intelligently and logically integrate the previously independently calculated safety constraints and comfort constraints, thereby accurately defining the unique and executable thermal and humidity dynamic elastic range [T] at each control cycle t. surf,lower (t), T surf,upper (t)]. This integration process employs different strategies depending on the system's operating mode (heating or cooling) to ensure optimal decision-making under all operating conditions.

[0132] During the critical cooling season, the system needs to simultaneously mitigate the risks of condensation and compromised thermal comfort. The upper boundary of the temperature range is set relatively directly, primarily to prevent discomfort caused by overheating; therefore, it is directly taken from the upper comfort limit temperature calculated from the PMV model, i.e., T. surf,upper (t)=T surf,max_comfort (t). Determining the lower boundary involves a more precise double-insurance decision-making process. The system must simultaneously meet the physical safety baseline—that is, the temperature must not fall below the anti-condensation constraint T. surf,min (t), and the physiological comfort baseline—that is, the temperature should not be lower than the comfort lower limit T. surf,min_comfort (t). To simultaneously satisfy these two mutually restrictive requirements, the control unit must adopt the more stringent one (i.e., the higher temperature). This decision logic is perfectly implemented through a maximum value function: T surf,lower (t)=max{T surf,min (t), T surf,min_comfort (t)}. This function ensures that when the indoor environment is dry, the decision is dominated by comfort to prevent excessive cold; while when the environment is humid, the decision is dominated by safety, forcibly raising the operating baseline to eliminate the risk of condensation.

[0133] During the heating season, since there is no risk of condensation, the logic is simplified, and the entire focus of the system is on maintaining comfort. At this time, the upper and lower boundaries of the flexibility range are entirely defined by comfort requirements, i.e., [T surf,lower (t),T surf,upper (t)]=[T surf,min_comfort (t), T surf,max_comfort (t)].

[0134] In summary, through this rigorous seasonal integration, the system constructs a dynamic range that adapts to the environment in real time. This range is not fixed; its width and position are frequently adjusted based on fluctuations in the indoor and outdoor environment (personnel load, solar radiation, weather changes). It is this intelligent adaptability based on physical and physiological models that enables this invention to maximize the building's flexible energy storage potential while ensuring absolute safety and high comfort, providing a solid and reliable control foundation for the efficient and economical absorption of intermittent renewable energy.

[0135] Furthermore, this system includes a virtual battery mapping and conversion unit 130. The task of this unit is to translate the thermo-humid dynamic elastic range defined in the thermophysical domain into electrical domain parameters that the microgrid energy management system can understand and schedule—that is, an equivalent virtual battery model. This conversion process involves mapping three key dimensions: First, mapping of equivalent energy storage capacity (kWh). Second, mapping of maximum charge / discharge power (kW). Third, inertial sustaining time T.inertia The mapping of (h). Through the above three-dimensional mapping, the unit generates a dynamic map of the microgrid's absorption capacity that is updated in real time. This map clearly shows the microgrid the flexibility potential of the building load at the current moment: for example, the ability to absorb P. max A total of kilowatts of power can absorb E electric The kilowatt-hours of electricity, if the power supply were to stop now, could still support T inertia Hour.

[0136] Next, the system is configured with a source-load bidirectional handshake interface 140. This interface 140 serves as a bridge for information exchange between the system and the microgrid energy management system (EMS), employing standardized industrial communication protocols (such as IEC 61850, Modbus TCP / IP, or MQTT). Its operating mode is bidirectional and negotiation-based. On one hand, it periodically and proactively reports the dynamic absorption capacity map generated by the virtual energy pool mapping and conversion unit to the microgrid EMS every 5 minutes. This allows the EMS to accurately know the real-time state (SOC, P) of the virtual battery (radiant air conditioning unit) when formulating a globally optimal scheduling strategy, just as it would with a real battery storage unit. max (etc.), thereby making more reasonable power allocation decisions. On the other hand, it is responsible for receiving power dispatch instructions issued by EMS based on this map, such as, please consume energy at a power of 20kW in the next hour or participate in a frequency regulation in the next 10 minutes, with a power fluctuation range of ±5kW.

[0137] More importantly, this interface incorporates an active rejection and correction mechanism. Upon receiving a command from the EMS, the command is not executed blindly but is first sent to the model predictive control execution unit for pre-simulation. If the pre-simulation results show that executing the command will cause the surface temperature of the radiating terminal to exceed the safety boundary of the thermal-humidity dynamic elastic range at some point in the future prediction time domain, or cause the PMV value to exceed the comfort range, this mechanism will be triggered. At this time, the system will not execute the original command but will autonomously correct the execution power to the maximum safe value allowed by the pre-simulation and immediately report back to the EMS through the source-load bidirectional handshake interaction interface 140: the command could not be fully executed, a certain amount of power has actually been executed, because the safety / comfort boundary has been touched. This handshake-style safety verification mechanism fundamentally eliminates the possibility of microgrid dispatch causing damage to the building itself and establishes a trust relationship between the source and load.

[0138] Finally, this system includes an advanced Model Predictive Control (MPC) execution unit 150. This unit is the brain of the system 100, responsible for translating the macroscopic power commands from the upper layer, which have undergone safety verification, into a refined and optimized sequence of control actions for the underlying equipment (heat pumps, water pumps, valves). The core of Model Predictive Control (MPC) lies in its model-based accurate prediction capability. This unit internally constructs a nonlinear state-space model based on the RC network method to accurately characterize the thermal dynamic behavior of the building and radiation system. This model can integrate the current real-time states of various parts of the system, such as temperature, combined with predicted values ​​of future disturbances such as outdoor temperature and solar radiation, and a preset control input sequence, to completely deduce the system's state evolution trajectory over a predicted time domain, typically 2 to 4 hours.

[0139] The control process of MPC is an iterative rolling optimization process. In each control cycle, such as every 15 minutes, MPC solves a multi-objective optimization problem. Its objective function typically includes several weighted terms: first, to track the power target curve issued by the microgrid, so that the actual power consumed by the system is as close as possible to the commanded value; second, to minimize the variation in control actions, avoiding frequent start-ups and drastic changes in equipment to extend equipment lifespan; and third, to maintain indoor comfort, so that the PMV value is as close as possible to the ideal value of 0. During the optimization process, a series of strict constraints must be met: first, state constraints, meaning that all predicted radiant terminal surface temperatures must be within the thermal-humid dynamic elastic range throughout the entire prediction time domain; second, control constraints, meaning that the output values ​​of control variables such as heat pump compressor frequency, water pump flow rate, and mixing valve opening must be within their physically adjustable range.

[0140] By solving this constrained optimization problem, MPC can obtain the optimal control sequence for a future period, such as the compressor frequency and water pump speed setpoints every 15 minutes within the next hour. However, MPC only executes the first control action in this sequence, and then repeats the entire prediction and optimization process in the next control cycle based on the latest system measurements. This rolling optimization and feedback correction mechanism enables the MPC controller to effectively cope with model mismatch and unforeseen disturbances, exhibiting excellent robustness and control accuracy.

[0141] In this invention, based on collected data and combined with a human thermal comfort model and a dew point temperature dynamic evolution model, the upper and lower limits of the dynamic temperature that the radiant terminal surface can operate at are determined, forming a dynamic elastic range of thermal and humidity. This dynamic elastic range is then transformed into equivalent virtual energy storage capacity, maximum absorbable charging / discharging power, and maximum sustainable discharge duration, based on building thermal capacity characteristics and the variable operating condition performance curve of the heat pump unit. According to the received power scheduling command or frequency response command, within the constraints of the dynamic elastic range of thermal and humidity, the optimal control sequence of each device parameter in the radiant air conditioner is solved using a rolling optimization algorithm. Based on this optimal control sequence, each device in the radiant air conditioner is controlled, realizing a shift from passive response to active guidance. This improves the safety and reliability of source-load interaction, effectively solving the problem of low reliability in the control of the absorption capacity of the radiant air conditioner microgrid caused by existing technologies, and effectively improving the reliability of the absorption capacity control of the radiant air conditioner microgrid.

[0142] In this invention, the current air dew point temperature is calculated using indoor air dry-bulb temperature and relative humidity. Based on the current air dew point temperature and the anti-condensation safety margin coefficient, the dynamic cold-side limit temperature of the radiant terminal surface is determined. According to the predicted average thermal sensation index model, an indoor operating temperature range that meets the requirements of human thermal comfort level is set, and the dynamic hot-side limit temperature of the radiant terminal surface is inversely calculated by combining the radiant heat transfer equation. The union region between the dynamic cold-side limit temperature and the dynamic hot-side limit temperature is taken as the current dynamic elastic range of heat and humidity, thus defining a clear anti-condensation safety and comfort boundary for the flexible adjustment of radiant air conditioning.

[0143] In this invention, the technical solution calculates the heat that can be absorbed or released when the current radiant terminal is driven from the average temperature to the boundary of the elastic range. The heat value that can be absorbed or released is converted into the virtual electrical energy storage capacity on the grid side through the real-time comprehensive performance coefficient of the heat pump unit. Based on the real-time comprehensive performance coefficient of the heat pump unit, the maximum heat exchange that the water system circulation flow can support under the current operating conditions is mapped to the maximum absorbable charging power. The current total operating power of the radiant air conditioning system is used as the maximum absorbable discharge power. Using the building heat attenuation model, the boundary value of the thermal and humidity dynamic elastic range is converted into the maximum sustainable discharge duration. On the one hand, through the inverse mapping of the virtual energy pool, the complex physical constraints are transformed into a more sustainable discharge duration. The system translates loads into a language understandable to the power grid, enabling proactive reporting of load-side absorption capacity. Furthermore, it provides a unified and standardized way to describe the flexibility of radiant air-conditioned buildings of different types and sizes. This standardized interface greatly simplifies the development and deployment of microgrid energy management systems, allowing them to connect to and manage large numbers of flexible building loads like plug-and-play. This lays a solid technical foundation for building larger-scale virtual power plants and regional energy internet, with broad application prospects and promotional value. Moreover, it equates the building's huge heat capacity to a virtual energy pool, providing microgrids with a large-capacity energy storage option at almost zero cost. During periods of high photovoltaic power generation, the system can accurately calculate the amount of electricity that can be safely absorbed and guide radiant air conditioning systems to charge, for example, through pre-cooling or pre-heating, converting potentially wasted green electricity into useful heat energy for storage. This not only effectively mitigates the fluctuations in renewable energy and improves its utilization rate but also significantly reduces the need for and dependence on expensive electrochemical energy storage systems, bringing significant economic benefits to the construction and operation of microgrids.

[0144] This invention establishes a nonlinear building thermal dynamics model. Using state data, equipment control sequences, and disturbance data as inputs, and the power curve from the microgrid as the target trajectory, and the thermal-humidity dynamic elastic range as the state constraint, the optimal control sequence for each device parameter in the radiant air conditioning system within the future control time domain is solved by minimizing the objective function. The state data includes the average indoor air temperature and the average surface temperature of the radiant terminals. The disturbance data includes the outdoor air temperature, solar radiation heat gain entering the room through windows, and internal heat sources generated by indoor occupants and equipment. This approach decomposes the macroscopic grid dispatch objective into refined and forward-looking control sequences for underlying equipment such as heat pumps, water pumps, and valves. This multi-objective optimization control, while fulfilling grid tasks, maximizes the stability and comfort of the indoor environment, and considers the economy and lifespan of equipment operation, achieving the best balance between grid efficiency, user comfort, and equipment health.

[0145] The technical solution of this invention adopts differentiated control for fluctuation signals of microgrids at different time scales. For frequency fluctuations or smoothing needs at the second to minute level, the flow rate of the variable frequency water pump and the opening of the mixing valve in the optimal control sequence are adjusted first, and the heat capacity of the water in the pipeline network is used for rapid response. For energy peak shaving and valley filling or new energy consumption needs at the hour level, the water supply temperature setpoint of the heat pump host in the optimal control sequence is adjusted first, and the sensible heat capacity of the building concrete floor slab is used for energy storage. This hierarchical response strategy can not only meet the grid's rapid response needs at the second and minute levels by adjusting the hydraulic system and provide auxiliary services such as frequency regulation, but also meet the energy time shift needs at the hour level by adjusting the heat source host.

[0146] When the power dispatch command issued by the microgrid energy management system causes the predicted surface temperature of the radiant terminal to exceed the boundary of the dynamic elastic range of heat and humidity, or causes the expected average thermal sensation index to deviate from the preset threshold of the comfort zone, the technical solution of this invention will forcibly correct the execution power to the maximum safe value that can be allowed by the pre-simulation, and feed back the current actual executable power limit. The dispatch decision of the microgrid is based on the active commitment of the load, and there is a safety pre-simulation and rejection mechanism before execution. This solves the condensation risk and comfort sacrifice problem caused by traditional blind dispatch, and makes it possible for the radiant air conditioning system to participate in grid interaction on a large scale and in a normalized manner.

[0147] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for controlling the absorption capacity of a radiant air conditioning microgrid, characterized in that, include: Real-time data collection of indoor and outdoor temperature and humidity, radiant terminal surface temperature, water supply and return status, and power flow data at the microgrid common connection point; Based on the collected data, combined with the human thermal comfort model and the dew point temperature dynamic evolution model, the upper and lower limits of the dynamic temperature that the radiant terminal surface can operate at are determined, forming a dynamic elastic range of thermal and humidity. The dynamic elastic range of heat and humidity is transformed into equivalent virtual energy storage capacity, maximum absorbable charging / discharging power and maximum sustainable discharge duration based on building heat capacity characteristics and heat pump unit variable operating condition performance curves, generating a dynamic map of microgrid absorption capacity. The dynamic map of microgrid absorption capacity is reported to the microgrid energy management system, and power dispatch instructions or frequency response instructions issued by the microgrid energy management system based on the dynamic map of microgrid absorption capacity are received. Based on the received power scheduling command or frequency response command, within the constraints of the thermal and humidity dynamic elastic range, the optimal control sequence of each device parameter in the radiant air conditioner is solved by the rolling optimization algorithm, and the device in the radiant air conditioner is controlled based on the optimal control sequence of each device parameter.

2. The method for controlling the absorption capacity of a radiant air conditioning microgrid according to claim 1, characterized in that, Based on the collected data, combined with the human thermal comfort model and the dew point temperature dynamic evolution model, the upper and lower limits of the dynamic temperature that the radiant terminal surface can operate at are determined, forming a dynamic elastic range of thermal and humidity. Specifically, this includes: Calculate the current air dew point temperature using the indoor air dry-bulb temperature and relative humidity; Based on the current air dew point temperature and the anti-condensation safety margin factor, determine the dynamic cold-side limit temperature of the radiant terminal surface; Based on the expected average thermal sensation index model, an indoor operating temperature range that meets the requirements of human thermal comfort level is set, and the dynamic thermal limit temperature of the radiant terminal surface is inversely calculated by combining the radiative heat transfer equation. The union region between the dynamic cold-side limit temperature and the dynamic hot-side limit temperature is taken as the dynamic elastic range of heat and humidity at the current moment.

3. The method for controlling the absorption capacity of a radiant air conditioning microgrid according to claim 2, characterized in that, The determination of the dynamic cold-side limit temperature of the radiant terminal surface based on the current air dew point temperature and the anti-condensation safety margin coefficient is specifically as follows: T surf,min (t1)=T dp (t1)+ΔT safe (t1); Among them, T surf,min (t1) represents the dynamic cold-side limit temperature of the radiation terminal surface during the current control cycle t1. dp (t1) represents the air dew point temperature at the current control cycle t1, ΔT safe (t1) is the anti-condensation safety margin coefficient at the current control cycle t1; ΔT is the anti-condensation safety margin coefficient at the current control cycle t1. safe The specific method for determining (t1) is as follows: ΔT safe (t1)=ΔT err +ΔT dist (t1)+ΔT buf (t1); Where, ΔT err This represents the static measurement error compensation margin; ΔT dist (t1) represents the dynamic spatial non-uniformity compensation margin at the current control period t1; ΔT buf (t1) represents the dynamic response buffer compensation margin at the current control cycle t1.

4. The method for controlling the absorption capacity of a radiant air conditioning microgrid according to claim 2, characterized in that, Based on the predicted average thermal sensation index model, an indoor operating temperature range that meets the requirements of human thermal comfort level is set. The dynamic thermal-side limit temperature of the radiative terminal surface is then calculated using the radiative heat transfer equation, specifically including: Based on the expected average thermal sensation index model and the requirements for human thermal comfort levels, the upper and lower limits of the operating temperature corresponding to the operating temperature range are determined. Determine the average radiation temperature range based on the operating temperature range; The surface temperature range of the radiation terminal is determined based on the average radiation temperature range.

5. The method for controlling the absorption capacity of a radiant air conditioning microgrid according to claim 1, characterized in that, The dynamic elastic range of thermal and humidity characteristics, based on the building's heat capacity and the variable operating condition performance curve of the heat pump unit, is transformed into an equivalent virtual electrical energy storage capacity, maximum absorbable charging / discharging power, and maximum sustainable discharge duration, specifically including: The calculation shows the amount of heat that can be absorbed or released when the current radiant terminal surface is driven from the average temperature to the boundary of the thermo-moisture dynamic elastic range. The value of the heat that can be absorbed or released is converted into the virtual electrical energy storage capacity on the grid side through the real-time comprehensive performance coefficient of the heat pump unit. Based on the real-time comprehensive performance coefficient of the heat pump unit, the maximum heat exchange that the water system circulation flow can support under the current operating conditions is mapped to the maximum absorbable charging power; the current total operating power of the radiant air conditioning system is taken as the maximum absorbable discharge power. By using a building thermal decay model, the boundary value of the thermal and moisture dynamic elastic interval is transformed into the maximum sustainable discharge duration.

6. The method for controlling the absorption capacity of a radiant air conditioning microgrid according to claim 1, characterized in that, The specific steps of finding the optimal control sequence for each device parameter in a radiant air conditioner using the rolling optimization algorithm include: A nonlinear building thermal dynamics model is established, using state data, equipment control sequences, and disturbance data as inputs, the power curve issued by the microgrid as the target tracking trajectory, and the thermal and humidity dynamic elastic range as the state constraint. The optimal control sequence of each equipment parameter in the radiant air conditioning system in the future control time domain is solved by minimizing the objective function. The state data includes the average indoor air temperature and the average surface temperature of the radiant terminals, while the disturbance data includes the outdoor air temperature, the solar radiation heat gain entering the room through the windows, and the internal heat sources generated by indoor personnel and equipment.

7. A method for controlling the absorption capacity of a radiant air conditioning microgrid according to claim 6, characterized in that, The nonlinear building thermal dynamics model is specifically as follows: x(k+1)=A x(k)+B u u(k)+B d d(k) Where k represents the current time; i+1 represents the next time; x(i+1) is the predicted state data for the next time; x(i) is the state data for the current time; A is the internal state matrix; B u For external control matrix; B d The external perturbation matrix; The objective function is specifically: ； Where J is the objective function value, i is the prediction step number, k is the initial prediction time, and N is the number of time steps in the prediction time domain; Let i be the actual electrical power consumed by the radiant air conditioning system at prediction step i. This refers to the power command issued by the microgrid at prediction step i; w p It is the power tracking weight; T op (i) is the predicted indoor operating temperature at prediction step i; T set It is the ideal comfort temperature set by the user; w c It is a comfort weight; For the control action at prediction step i, For the control action at prediction step i-1; w u It controls the smoothing weights.

8. The method for controlling the absorption capacity of a radiant air conditioning microgrid according to claim 1, characterized in that, The control of each device in a radiant air conditioner is specifically based on the optimal control sequence of each device parameter: Differentiated control is adopted for fluctuation signals of microgrid at different time scales. For frequency fluctuations or smoothing needs at the second to minute level, the flow rate of variable frequency water pump and the opening of mixing valve in the optimal control sequence are adjusted first, and the heat capacity of water in the pipeline network is used for rapid response. For energy peak shaving and valley filling or new energy consumption needs at the hour level, the water supply temperature setpoint of heat pump host in the optimal control sequence is adjusted first, and the sensible heat capacity of building concrete floor slab is used for energy storage.

9. A method for controlling the absorption capacity of a radiant air conditioning microgrid according to any one of claims 1-8, characterized in that, Also includes: When the power dispatch command issued by the microgrid energy management system causes the predicted surface temperature of the radiating terminal to exceed the boundary of the thermal and humidity dynamic elastic range, or causes the expected average thermal sensation index to deviate from the preset threshold of the comfort zone, the execution power will be forcibly corrected to the maximum safe value that can be allowed by the pre-simulation, and the current actual executable power limit will be fed back upwards.

10. A control system for the absorption capacity of a radiant air conditioning microgrid, characterized in that, include: The environmental and operating condition multi-dimensional sensing subsystem collects real-time data on indoor and outdoor temperature and humidity, surface temperature of radiant terminals, water supply and return status, and power flow data at the microgrid common connection point. The thermal and humidity dynamic elastic range calculation unit, based on the collected data and combined with the human thermal comfort model and the dew point temperature dynamic evolution model, determines the upper and lower limits of the dynamic temperature that the radiant terminal surface can operate at, thus forming the thermal and humidity dynamic elastic range. The virtual energy storage mapping and conversion unit transforms the dynamic elastic range of heat and humidity based on the building's thermal capacity characteristics and the variable operating condition performance curve of the heat pump unit into an equivalent virtual energy storage capacity, maximum absorbable charging power, and maximum sustainable discharge duration, thereby generating a dynamic map of the microgrid's absorption capacity. The source-load two-way handshake interaction interface reports the dynamic map of microgrid absorption capacity to the microgrid energy management system and receives power dispatch instructions or frequency response instructions issued by the microgrid energy management system based on the dynamic map of microgrid absorption capacity. The model predictive control execution unit, based on the received power scheduling command or frequency response command, solves the optimal control sequence of each device parameter in the radiant air conditioner through a rolling optimization algorithm within the constraint range of the thermal and humidity dynamic elastic range, and controls each device in the radiant air conditioner based on the optimal control sequence of each device parameter.

Images