Air conditioner and control method and device thereof, storage medium and computer program product

Abstract

The invention discloses an air conditioner control method and device, an air conditioner, a storage medium and a computer program product. The method comprises the steps that a comfort constraint model used for defining the allowable range of health and comfort related multi-dimensional constraint parameters is constructed; a system energy consumption model representing the incidence relation between the air conditioner operation parameters, the equipment characteristic parameters, the environment parameters and the energy consumption is established; based on the comfort constraint model and the system energy consumption model, adopting a preset optimization algorithm to solve a multi-objective optimization problem with constraints, and generating an optimal control sequence containing control actions corresponding to multiple control periods in the future; the targets of the multi-target optimization problem include reduction of air conditioner energy consumption and enabling the indoor environment to conform to the allowable range of the comfort constraint model; and controlling the air conditioner to execute a control action corresponding to the current control period in the optimal control sequence. According to the scheme, the adaptive capacity of the air conditioner to the external environment and the system characteristic change is improved, and the balance effect of health and comfort guarantee and efficient energy saving is improved.

Description

Technical Field

[0001] This invention belongs to the field of air conditioning technology, and specifically relates to an air conditioning control method, device, air conditioner, storage medium, and computer program product. Background Technology

[0002] With the continuous rise in global building energy consumption, air conditioning energy-saving control technology has become a core research direction in the field of building energy conservation. In practical applications, air conditioning not only needs to meet users' basic needs for a healthy and comfortable indoor environment, but also needs to respond to industry requirements for energy conservation and emission reduction. The synergistic achievement of these two aspects has become a key goal of technological research and development. However, existing air conditioning control technologies have always faced a core challenge: it is difficult to achieve high-efficiency energy saving while ensuring health and comfort, and there is an inherent contradiction between the needs for health and comfort and the goals of energy saving.

[0003] Traditional air conditioning energy-saving methods generally adopt fixed operation modes, such as timer switching and simple temperature threshold control. These control logics are only guided by a single energy consumption index, completely ignoring the health and comfort attributes of the indoor environment. They have not established a comprehensive comfort standard that covers multiple dimensions such as temperature, humidity, and air quality, nor have they considered the differences in user preferences and comfort needs in different scenarios (such as sleep and office). This leads to problems such as frequent fluctuations in indoor temperature and humidity deviating from the healthy range, which seriously affects the user's physical experience.

[0004] Meanwhile, existing technologies lack dynamic optimization capabilities and cannot adapt to complex and ever-changing actual operating conditions. On the one hand, they cannot respond in real time to changes in the external environment, such as sudden rises and falls in weather temperature, fluctuations in solar radiation intensity, and dynamic changes in the number of people indoors, and still operate rigidly according to preset programs. On the other hand, they do not fully consider the dynamic differences in the characteristics of the air conditioning system itself, such as changes in the coefficient of performance with operating conditions and the temperature lag effect caused by building thermal inertia, resulting in a disconnect between control strategies and actual needs. For example, in pursuit of rapid cooling or maintaining a constant temperature, existing technologies often allow air conditioners to operate at high power and full load, which can temporarily meet temperature requirements, but consumes extremely high energy. If power is blindly reduced in order to save energy, the indoor environment will exceed the comfort range, making it impossible to simultaneously meet energy-saving requirements and comfort needs.

[0005] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0006] The purpose of this invention is to provide an air conditioner control method, device, air conditioner, storage medium, and computer program product to solve the problems in related solutions where air conditioner energy-saving control adopts a fixed mode and lacks dynamic optimization capabilities, cannot adapt to changes in external environment and system characteristics, and is difficult to achieve high energy efficiency while ensuring health and comfort. This invention improves the air conditioner's adaptability to changes in external environment and system characteristics, enhances the balance between ensuring health and comfort and high energy efficiency, and effectively solves the limitations of traditional fixed mode control.

[0007] This invention provides a method for controlling an air conditioner, comprising: constructing a comfort constraint model, wherein the comfort constraint model is used to define the allowable range of multi-dimensional constraint parameters related to health and comfort; establishing a system energy consumption model, wherein the system energy consumption model is used to characterize the correlation between air conditioner operating parameters, equipment characteristic parameters, environmental parameters and air conditioner energy consumption; based on the comfort constraint model and the system energy consumption model, using a preset optimization algorithm to solve a constrained multi-objective optimization problem, generating an optimal control sequence containing control actions corresponding to multiple future control cycles; the objectives of the multi-objective optimization problem include reducing air conditioner energy consumption and making the indoor environment conform to the allowable range of the comfort constraint model; controlling the air conditioner to execute the control action corresponding to the current control cycle in the optimal control sequence, and collecting real-time feedback data, updating the air conditioner operating status and the parameters of the comfort constraint model and the system energy consumption model based on the real-time feedback data, thereby optimizing the generation process of the subsequent optimal control sequence.

[0008] In some implementations, the preset optimization algorithm is a model predictive control algorithm, which quantifies the multi-objective optimization problem by constructing an objective function; the objective function includes an energy consumption calculation term and a comfort deviation calculation term, and the trade-off between the energy consumption calculation term and the comfort deviation calculation term is adjusted by preset weight coefficients.

[0009] In some embodiments, the method further includes: acquiring environmental prediction data and user behavior prediction data for a future preset time period, and using the environmental prediction data and user behavior prediction data as input to the preset optimization algorithm; the environmental prediction data includes weather change prediction data, and the user behavior prediction data includes personnel entry and exit time prediction data.

[0010] In some embodiments, the method further includes: acquiring user feedback data or identifying the current usage scenario of the air conditioner; adjusting the allowable range of the multi-dimensional constraint parameters based on the user feedback data or the comfort demand data corresponding to the usage scenario; wherein the user feedback data includes explicit feedback data and implicit feedback data, the explicit feedback data being preference instructions input by the user through an interactive terminal, and the implicit feedback data being potential demand data obtained based on the user's behavior analysis of adjusting the air conditioner.

[0011] In some implementations, controlling the air conditioner to perform the control action corresponding to the current control cycle in the optimal control sequence includes: for a variable frequency air conditioner, mapping the control command to the target frequency or target power of the compressor; for a fixed frequency air conditioner, calculating the duty cycle of the compressor operation and executing the control command based on the duty cycle.

[0012] In some embodiments, the method further includes: collecting historical operating data of the air conditioner, the historical operating data including historical energy consumption statistics, historical environmental parameter change data and corresponding air conditioner operating parameter data; and training and calibrating the parameters of the system energy consumption model based on the historical operating data.

[0013] In conjunction with the above method, another aspect of the present invention provides an air conditioner control device, comprising: a modeling unit configured to construct a comfort constraint model, the comfort constraint model being used to define the allowable range of multi-dimensional constraint parameters related to health and comfort; the modeling unit further configured to establish a system energy consumption model, the system energy consumption model being used to characterize the correlation between air conditioner operating parameters, equipment characteristic parameters, environmental parameters and air conditioner energy consumption; an optimization unit configured to solve a constrained multi-objective optimization problem based on the comfort constraint model and the system energy consumption model using a preset optimization algorithm, generating an optimal control sequence containing control actions corresponding to multiple future control cycles; the objectives of the multi-objective optimization problem include reducing air conditioner energy consumption and making the indoor environment conform to the allowable range of the comfort constraint model; and a control unit configured to control the air conditioner to execute the control action corresponding to the current control cycle in the optimal control sequence, and to collect real-time feedback data, update the air conditioner operating status and the parameters of the comfort constraint model and the system energy consumption model based on the real-time feedback data, thereby optimizing the generation process of the subsequent optimal control sequence.

[0014] In some implementations, the preset optimization algorithm is a model predictive control algorithm, which quantifies the multi-objective optimization problem by constructing an objective function; the objective function includes an energy consumption calculation term and a comfort deviation calculation term, and the trade-off between the energy consumption calculation term and the comfort deviation calculation term is adjusted by preset weight coefficients.

[0015] In some implementations, the optimization unit is further configured to: acquire environmental prediction data and user behavior prediction data for a future preset time period, and use the environmental prediction data and user behavior prediction data as input to the preset optimization algorithm; the environmental prediction data includes weather change prediction data, and the user behavior prediction data includes personnel entry and exit time prediction data.

[0016] In some implementations, the modeling unit is further configured to: acquire user feedback data or identify the current usage scenario of the air conditioner; adjust the allowable range of the multi-dimensional constraint parameters according to the user feedback data or the comfort demand data corresponding to the usage scenario; wherein the user feedback data includes explicit feedback data and implicit feedback data, the explicit feedback data being preference instructions input by the user through an interactive terminal, and the implicit feedback data being potential demand data obtained based on the user's behavior analysis of adjusting the air conditioner.

[0017] In some implementations, the control unit controls the air conditioner to perform control actions corresponding to the current control cycle in the optimal control sequence, including: for inverter air conditioners, mapping control commands to the target frequency or target power of the compressor; for fixed-frequency air conditioners, calculating the duty cycle of the compressor operation and executing control commands based on the duty cycle.

[0018] In some implementations, the modeling unit is further configured to: collect historical operating data of the air conditioner, including historical energy consumption statistics, historical environmental parameter change data, and corresponding air conditioner operating parameter data; and train and calibrate the parameters of the system energy consumption model based on the historical operating data.

[0019] In conjunction with the above-described device, the present invention further provides an air conditioner, comprising: the control device for the air conditioner described above.

[0020] In conjunction with the above method, the present invention further provides a storage medium comprising a stored program, wherein, when the program is executed, the device on which the storage medium is located controls the air conditioner control method described above to be performed.

[0021] In conjunction with the above method, the present invention further provides a computer program product comprising a computer program that, when processed and executed, implements the steps of the above-described air conditioner control method.

[0022] The present invention constructs a comfort constraint model to define the allowable range of multi-dimensional constraint parameters related to health and comfort, and establishes a system energy consumption model characterizing the relationship between air conditioning operating parameters, equipment characteristic parameters, environmental parameters, and energy consumption. Based on the above two models, a multi-objective optimization problem aimed at reducing energy consumption and meeting comfort constraints is solved through a preset optimization algorithm, generating an optimal control sequence that includes control actions for multiple future control cycles. The air conditioner is controlled to execute the control actions of the current control cycle, real-time feedback data is collected, the air conditioner operating status and dual model parameters are updated, and the subsequent optimal control sequence is optimized. This effectively solves the limitations of traditional fixed-mode control, improves the adaptability to changes in the external environment and system characteristics, and enhances the balance between health and comfort protection and high-efficiency energy saving.

[0023] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention.

[0024] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0025] Figure 1 This is a flowchart illustrating an embodiment of the air conditioner control method of the present invention;

[0026] Figure 2 This is a schematic diagram of the structure of an embodiment of the air conditioner control device of the present invention;

[0027] Figure 3 A flowchart illustrating another embodiment of the air conditioning control method;

[0028] Figure 4 This is a network topology diagram;

[0029] Figure 5 To control the timing diagram.

[0030] Referring to the accompanying drawings, the reference numerals in the embodiments of the present invention are as follows:

[0031] 101 - Modeling unit; 102 - Optimization unit; 103 - Control unit. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0033] According to embodiments of the present invention, an air conditioning control method is provided, such as... Figure 1 The flowchart of an embodiment of the method of the present invention is shown. The air conditioner control method may include steps S110 to S140.

[0034] In step S110, a comfort constraint model is constructed, which is used to define the allowable range of multi-dimensional constraint parameters related to health and comfort.

[0035] A comfort constraint model is a model that defines the permissible range of multi-dimensional constraint parameters related to human health and perceived comfort in an indoor environment. Its core function is to set rigid boundaries for air conditioning control to ensure that the indoor environment meets human health needs and comfort standards. Multi-dimensional constraint parameters refer to environmental parameters directly related to human health and comfort, including temperature, humidity, and air quality parameters (such as carbon dioxide concentration). The permissible range of these parameters needs to be determined comprehensively based on health standards, human physiological needs, and usage scenarios.

[0036] Human beings have specific parameter ranges for their health and comfort needs in indoor environments; exceeding these ranges can negatively impact physical health or sensory experience. Constructing a comfort constraint model aims to transform abstract health and comfort needs into quantifiable and actionable constraints, providing clear target boundaries for subsequent optimization control and preventing air conditioning control from solely pursuing energy consumption or a single parameter while neglecting health and comfort.

[0037] Specifically, the first step is to identify the types of multi-dimensional constraint parameters related to health and comfort. Combining health standards (such as relevant industry regulations and human physiological tolerance ranges), usage scenario characteristics, and population characteristics, the allowable ranges for each parameter are determined. Subsequently, these constraints are quantified and mathematically modeled to form a complete comfort constraint model, which can be dynamically adjusted according to actual needs. For example, for a bedroom sleep scenario, considering the slowed metabolism and temperature sensitivity changes during sleep, the allowable range for temperature constraint parameters is determined to be 26℃-28℃, the allowable range for humidity constraint parameters is 40%-60%, and the allowable range for carbon dioxide concentration constraint parameters is below 1000ppm. Integrating these parameter ranges and related constraint logic completes the construction of the comfort constraint model for this scenario.

[0038] In some implementations, the process also includes adjusting the allowable range of multi-dimensional constraint parameters, specifically including: acquiring user feedback data or identifying the current usage scenario of the air conditioner; adjusting the allowable range of the multi-dimensional constraint parameters according to the user feedback data or the comfort demand data corresponding to the usage scenario; wherein, the user feedback data includes explicit feedback data and implicit feedback data, the explicit feedback data being preference instructions input by the user through an interactive terminal, and the implicit feedback data being potential demand data obtained based on the user's behavior analysis of adjusting the air conditioner.

[0039] User feedback data refers to information related to indoor health and comfort needs conveyed by users through direct input or behavioral performance. This includes both explicit and implicit feedback data and serves as a crucial basis for adjusting comfort constraint models. Usage scenarios refer to the specific context in which the air conditioner operates, related to the user's activity level or environmental characteristics. Users' health and comfort needs differ across different scenarios. Comfort requirement data refers to the environmental parameter standards that can meet the user's health and comfort experience within the corresponding usage scenario, including reasonable ranges for parameters such as temperature, humidity, and air quality.

[0040] Users' health and comfort needs are not fixed but dynamically change depending on their feelings, activity levels, and the context in which they are situated. By acquiring user feedback data or identifying the current usage scenario, these dynamic changes in needs can be accurately captured. Adjusting the allowable range of multi-dimensional constraint parameters based on the changed comfort needs data ensures that the comfort constraint model always matches the actual needs of users, thereby making the control strategies generated by subsequent optimization algorithms more aligned with personalized and scenario-based requirements.

[0041] Specifically, on the one hand, explicit feedback data actively input by users is received through user interaction terminals, or implicit feedback data is obtained by analyzing user behavior data of manually adjusting the air conditioner. On the other hand, sensors detect the activity status and time information of people indoors to identify the current usage scenario of the air conditioner (such as a sleep scenario or an office scenario). For the acquired user feedback data, the corresponding comfort preferences are analyzed (e.g., the need for higher temperatures in response to feedback of "too cold"). For the identified usage scenario, preset comfort requirement data for that scenario is retrieved (e.g., the reasonable range of temperature and humidity for an office scenario). Based on the matched comfort requirement data, the allowable range of multi-dimensional constraint parameters in the comfort constraint model is dynamically adjusted to ensure that the adjusted range still meets human health requirements while aligning with the user's current needs. The adjusted comfort constraint model is then used as new input to a preset optimization algorithm, which solves a multi-objective optimization problem based on the updated constraints to generate the optimal control sequence adapted to the current needs.

[0042] In step S120, a system energy consumption model is established, which is used to characterize the relationship between air conditioning operating parameters, equipment characteristic parameters, environmental parameters and air conditioning energy consumption.

[0043] A system energy consumption model is a model that characterizes the relationship between air conditioning operating parameters, equipment characteristic parameters, environmental parameters, and air conditioning energy consumption. It can accurately predict the energy consumption of air conditioning under different operating conditions, providing data support for energy-saving optimization. Air conditioning operating parameters refer to the adjustable parameters during air conditioning operation, including cooling power, heating power, and operating frequency; these are the core parameters for adjusting the air conditioning status. Equipment characteristic parameters refer to the inherent parameters of the air conditioning equipment that affect energy consumption, including the coefficient of performance (COP) and thermal inertia-related parameters. The COP reflects the energy conversion efficiency of the air conditioning, while thermal inertia-related parameters reflect the heat storage effect of the building or equipment. Environmental parameters refer to external and internal environmental data that affect air conditioning operation and the indoor environment, including indoor and outdoor temperatures, solar radiation intensity, and the status of indoor occupants.

[0044] Air conditioning energy consumption is affected by various factors, including air conditioning operating parameters, equipment characteristics, and environmental parameters, and these factors have complex interrelationships. Establishing a system energy consumption model can accurately quantify these relationships, enabling the prediction of air conditioning energy consumption under different operating conditions and avoiding energy waste caused by blind control.

[0045] Specifically, the key parameters affecting air conditioning energy consumption are first identified, including air conditioning operating parameters (such as cooling power and operating frequency), equipment characteristic parameters (such as coefficient of performance and thermal inertia), and environmental parameters (such as indoor and outdoor temperatures and solar radiation intensity). Then, through theoretical analysis (such as deriving energy consumption equations based on thermodynamic principles) and data support (such as collecting historical energy consumption data and environmental parameter data), a mathematical model is constructed that can characterize the relationship between each parameter and energy consumption. Finally, the model parameters are calibrated using historical data to improve the model's prediction accuracy. For example, based on thermodynamic principles, the energy consumption equation E(t) = COP × P(t) + thermal inertia term (where E(t) is energy consumption and P(t) is equipment power) is derived. Historical energy consumption data under different air conditioning operating powers, indoor and outdoor temperatures, and solar radiation intensities over the past month are collected. Regression methods are used to estimate the specific values ​​of COP and thermal inertia-related parameters in the equation, completing the establishment of a system energy consumption model. This model can predict the air conditioning energy consumption under corresponding operating conditions based on the input current operating parameters and environmental parameters.

[0046] In some implementations, the process of training and calibrating the parameters of the system energy consumption model is also included, specifically including: collecting historical operating data of the air conditioner, the historical operating data including historical energy consumption statistics, historical environmental parameter change data and corresponding air conditioner operating parameter data; and training and calibrating the parameters of the system energy consumption model based on the historical operating data.

[0047] The initial parameters of system energy consumption models are usually based on theoretical derivations or general equipment characteristics, which differ from the actual equipment status and environmental conditions during air conditioner operation, resulting in insufficient model prediction accuracy. Historical operating data records the real correlation between energy consumption, environment, and operating parameters during actual air conditioner operation. By training and calibrating the model parameters using this data, the model can accurately match the actual operating characteristics of the air conditioner, thereby improving the accuracy of energy consumption prediction.

[0048] Specifically, the system collects historical operating data of the air conditioner according to a preset period (e.g., daily, weekly), including historical energy consumption statistics (e.g., hourly energy consumption), historical environmental parameter change data (e.g., indoor and outdoor temperature, humidity, and solar radiation intensity during the same period), and corresponding air conditioner operating parameter data (e.g., operating power, frequency), ensuring data integrity and synchronization. The collected historical data is screened and cleaned, removing abnormal data (e.g., energy consumption data during equipment failure, and environmental parameters abnormally collected by sensors), and missing data is appropriately supplemented to ensure data quality. Simultaneously, the data is converted into a format that the model can process, facilitating subsequent parameter calibration. The preprocessed historical data is input into the system's energy consumption model, and appropriate algorithms (e.g., linear regression, neural networks) are used for parameter optimization. By minimizing the deviation between the model's predicted energy consumption and historical actual energy consumption, key parameters in the model, such as performance coefficients and thermal inertia, are adjusted. The calibrated model parameters are validated using some historical data not used in training. If the prediction deviation meets the preset accuracy requirements (e.g., deviation less than 5%), the parameters of the system's energy consumption model are updated; if the deviation does not meet the requirements, the training algorithm is readjusted or supplemented with data and retrained until the model accuracy meets the requirements.

[0049] In step S130, based on the comfort constraint model and the system energy consumption model, a preset optimization algorithm is used to solve the constrained multi-objective optimization problem, generating an optimal control sequence containing control actions corresponding to multiple future control cycles; the objectives of the multi-objective optimization problem include reducing air conditioning energy consumption and making the indoor environment conform to the allowable range of the comfort constraint model.

[0050] A pre-defined optimization algorithm refers to a pre-set algorithm used to solve multi-objective optimization problems. It possesses forward-looking prediction and dynamic optimization capabilities, and can achieve multi-objective balance while satisfying constraints. A multi-objective optimization problem refers to an optimization problem with the core objectives of reducing air conditioning energy consumption and ensuring the indoor environment meets the allowable range of the comfort constraint model. It requires finding the optimal trade-off relationship among multiple objectives. The optimal control sequence refers to the sequence of control actions corresponding to multiple future control cycles, obtained after solving the pre-defined optimization algorithm. Each control action is the optimal choice under the current operating conditions. The control cycle refers to the time interval for updating the air conditioning control strategy. It is set according to the rate of change of the indoor environment and the operating characteristics of the air conditioning system to ensure that the control strategy can respond promptly to changes in the environment and demand.

[0051] The core challenge in air conditioning control is balancing health and comfort with energy consumption. Single-objective control can lead to trade-offs (e.g., pursuing energy saving at the expense of comfort, or pursuing comfort at the expense of high energy consumption). Pre-defined optimization algorithms possess foresight and multi-objective optimization capabilities, enabling them to find the control scheme that minimizes energy consumption while satisfying comfort constraints. The generation of the optimal control sequence provides a clear basis for subsequent phased execution of control actions.

[0052] Specifically, the comfort constraint model and the system energy consumption model are used as inputs to clarify the core objectives of the multi-objective optimization problem (reducing air conditioning energy consumption and ensuring that the indoor environment meets the allowable range of the comfort constraint model). The system state (such as changes in indoor environmental parameters and energy consumption) in the future multiple control cycles is predicted by a preset optimization algorithm, and an objective function containing energy consumption calculation terms and comfort deviation calculation terms is constructed. The objective function is solved under constraints such as comfort constraints and equipment physical limitations. Finally, the optimal control sequence containing the control actions corresponding to the future multiple control cycles is obtained, and each control action corresponds to the optimal air conditioning operating state under a specific operating condition.

[0053] For example, assuming a control cycle of 5 minutes and a prediction time domain of 24 control cycles (i.e., the next 2 hours), the constructed comfort constraint model (such as temperature 25℃-27℃, humidity 40%-60%) and system energy consumption model are input into the model predictive control algorithm. The algorithm predicts the changes in outdoor temperature and solar radiation intensity at different time points in the next 2 hours, constructs the objective function min (energy consumption calculation term + comfort deviation calculation term), and solves it under constraints such as temperature not exceeding 25℃-27℃ and air conditioning power not exceeding 0kW-3.5kW, generating the air conditioning operating power sequence corresponding to each 5 minutes in the next 2 hours. This sequence is the optimal control sequence.

[0054] In some implementations, the preset optimization algorithm is a model predictive control algorithm, which quantifies the multi-objective optimization problem by constructing an objective function; the objective function includes an energy consumption calculation term and a comfort deviation calculation term, and the trade-off between the energy consumption calculation term and the comfort deviation calculation term is adjusted by preset weight coefficients.

[0055] Model predictive control (MMC) is a model-based closed-loop optimization control algorithm. It predicts the system state over a future period, solves multi-objective optimization problems to generate the optimal control sequence, and continuously corrects the model and control strategy based on real-time feedback data. It is suitable for complex control scenarios with multiple constraints and objectives. The objective function transforms the multi-objective optimization problem into a quantifiable mathematical expression. Solving this expression reveals the optimal trade-off between multiple objectives and is the core computational basis of the multi-objective optimization algorithm. The energy consumption calculation term is a mathematical term in the objective function used to quantify the air conditioning energy consumption level. Its value is related to air conditioning operating parameters, equipment characteristic parameters, and environmental parameters; a lower value indicates lower energy consumption. The comfort deviation calculation term is a mathematical term in the objective function used to quantify the deviation between the indoor environment and the allowable range of the comfort constraint model. Its value is the quantified result of the difference between the actual indoor environmental parameters and the allowable range of the comfort constraint model; a lower value indicates that the indoor environment better meets health and comfort requirements. The weighting coefficient is used to adjust the importance of the energy consumption calculation item and the comfort deviation calculation item in the objective function. By adjusting the weighting coefficient, the emphasis on the two objectives of energy saving and comfort can be controlled.

[0056] Model predictive control algorithms possess forward-looking prediction and multi-constraint optimization capabilities. They can predict the air conditioning operation status and energy consumption level for multiple future control cycles based on comfort constraint models and system energy consumption models. By constructing an objective function that includes energy consumption calculation terms and comfort deviation calculation terms, the dual objectives of "reducing energy consumption" and "meeting comfort constraints" can be transformed into mathematical problems. Combined with the adjustment of preset weight coefficients, the optimal solution that balances both objectives can be finally obtained.

[0057] Specifically, when constructing the objective function, the mathematical expression for the energy consumption calculation item is determined. This expression is based on the system energy consumption model, taking air conditioning operating parameters, equipment characteristic parameters, and environmental parameters as input variables, and outputting the corresponding quantitative energy consumption value. The mathematical expression for the comfort deviation calculation item is also determined. This expression is based on the comfort constraint model, calculating the deviation values ​​between the actual indoor temperature, humidity, air quality, and other parameters and the model's allowable range, and quantifying these deviation values. The energy consumption calculation item and the comfort deviation calculation item are then integrated to form a complete objective function, expressed as: F = w1 × f1 + w2 × f2 (F is the objective function value, w1 and w2 are preset weight coefficients, f1 is the energy consumption calculation item, and f2 is the comfort deviation calculation item). Preset weight coefficients are set according to the usage scenario. For example, in an office scenario, the weight coefficient of the comfort deviation calculation item can be increased to prioritize personnel comfort; in an unmanned scenario, the weight coefficient of the energy consumption calculation item can be increased to prioritize energy saving. The constructed objective function is input into the model predictive control algorithm. The algorithm combines the boundary conditions of the comfort constraint model and the parameter correlation of the system energy consumption model to solve for the minimum value of the objective function and generate the corresponding optimal control sequence.

[0058] In some implementations, the algorithm also includes a process to improve its accuracy, specifically including: acquiring environmental prediction data and user behavior prediction data for a future preset time period, and using the environmental prediction data and user behavior prediction data as input to the preset optimization algorithm; the environmental prediction data includes weather change prediction data, and the user behavior prediction data includes personnel entry and exit time prediction data.

[0059] The future preset time period refers to a pre-defined future time interval used by the algorithm to predict changes in system state. Its duration needs to be determined based on the air conditioning control cycle, environmental change patterns, and optimization requirements; it can be set to 2 hours. Environmental prediction data refers to the prediction results of changes in external environmental parameters affecting air conditioning operation and the indoor environment within the future preset time period. Its core function is to provide the optimization algorithm with forward-looking environmental change information, assisting in generating more adaptive control strategies. Weather change prediction data is a core type of environmental prediction data, including predicted values ​​of weather-related parameters such as outdoor temperature, humidity, solar radiation intensity, and wind speed within the future preset time period. These are key external factors affecting air conditioning load and energy consumption. User behavior prediction data is the prediction result of users' indoor activity status within the future preset time period, used to match actual user needs and avoid a disconnect between control strategies and user behavior. Personnel entry and exit time prediction data is a core type of user behavior prediction data, referring to the predicted values ​​of the time points and number of people entering and leaving the room within the future preset time period, directly affecting indoor heat load and comfort requirements.

[0060] The operating status and energy consumption level of air conditioners are not only affected by the current environment and user status, but also closely related to environmental changes and user behavior changes in the future. By acquiring environmental prediction data and user behavior prediction data for a preset period of time in the future, and using them as input to a preset optimization algorithm, the algorithm can anticipate future system disturbances (such as rising outdoor temperatures or users returning home). Based on this forward-looking information, the algorithm solves multi-objective optimization problems, and the generated optimal control sequence can adapt to future changes in advance, avoiding energy waste or decreased comfort caused by passive adjustments.

[0061] Specifically, the required environmental prediction data and user behavior prediction data types are clearly defined; based on the air conditioning control cycle (e.g., 5 minutes) and optimization requirements, a future preset time period (e.g., the next 2 hours) is set. Weather change prediction data for the future preset time period is acquired; by analyzing historical user activity data and combining it with geofencing and other technologies, the entry and exit times of personnel within the future preset time period are predicted. The acquired weather change prediction data and personnel entry and exit time prediction data are preprocessed to remove outliers and convert the data into a format recognizable and processed by the preset optimization algorithm (e.g., a standardized numerical sequence). The preprocessed environmental prediction data and user behavior prediction data, along with the comfort constraint model and system energy consumption model, are input into the preset optimization algorithm; the algorithm, combined with the current system state (e.g., current indoor temperature, air conditioning operating power), predicts the indoor environmental change trend and energy consumption change trend within the future preset time period, solves the constrained multi-objective optimization problem, and generates the optimal control sequence containing control actions corresponding to multiple future control cycles.

[0062] In step S140, the air conditioner is controlled to execute the control action corresponding to the current control cycle in the optimal control sequence, and real-time feedback data is collected. Based on the real-time feedback data, the air conditioner's operating status and the parameters of the comfort constraint model and the system energy consumption model are updated, thereby optimizing the subsequent optimal control sequence generation process.

[0063] Real-time feedback data refers to control-related data collected in real time during air conditioning operation, including indoor environmental parameters, occupant status data, and air conditioning operating status data. This data reflects the current control effectiveness and system status. Indoor and external conditions are constantly changing; a fixed control sequence cannot adapt to long-term dynamic changes, potentially causing subsequent control actions to deviate from the optimal state. Executing only the control action for the current control cycle ensures that each control cycle's action is the most suitable choice for the current situation. However, by updating the system status and model parameters through real-time feedback data and then optimizing subsequent control sequences, a dynamic closed-loop control can be formed, continuously adapting to changes in the environment and demands, ensuring long-term health, comfort, and energy-saving effects.

[0064] Specifically, the control action corresponding to the current control cycle is extracted from the optimal control sequence, and the control command is sent to the air conditioner to execute the action. During the operation of the air conditioner, real-time feedback data such as indoor environmental parameters, personnel status data, and air conditioner operation status data are collected. Based on this real-time feedback data, it is determined whether the current system state is consistent with the model prediction, the air conditioner operation status parameters are adjusted, and the relevant parameters of the comfort constraint model and the system energy consumption model are calibrated. Based on the updated system state and model parameters, the multi-objective optimization problem is solved again to generate a new optimal control sequence, which provides a basis for the control action of the next control cycle, and so on.

[0065] For example, if the current control cycle is a 5-minute period, the air conditioner operating power corresponding to this period is extracted from the optimal control sequence as 2.0kW, and the air conditioner is controlled to operate at 2.0kW. Within these 5 minutes, data such as indoor temperature, humidity, and actual air conditioner operating power are collected in real time. It is found that the actual rate of change of indoor temperature is slightly higher than the model prediction. Based on this feedback data, the thermal inertia-related parameters in the system energy consumption model are calibrated, while confirming that the temperature allowable range of the comfort constraint model is still 25℃-27℃. Based on the updated model parameters and the current indoor temperature of 26℃ and outdoor temperature of 32℃, the multi-objective optimization problem is solved again to generate the optimal control action for the next 5 minutes (such as adjusting the air conditioner operating power to 1.8kW), ensuring that subsequent control can still meet comfort requirements and achieve optimal energy consumption.

[0066] In some implementations, step S140, controlling the air conditioner to execute the control action corresponding to the current control cycle in the optimal control sequence, includes: for a variable frequency air conditioner, mapping the control command to the target frequency or target power of the compressor; for a fixed frequency air conditioner, calculating the duty cycle of the compressor operation and executing the control command based on the duty cycle.

[0067] Different types of air conditioners (inverter and fixed-frequency) have fundamentally different operating mechanisms. Inverter air conditioners support continuous power adjustment, while fixed-frequency air conditioners can only achieve on / off control. Considering the structural characteristics of these two types of air conditioners, converting a unified control command (target power demand) into specific control parameters adapted to their operating mechanisms ensures that the control actions in the optimal control sequence can be accurately executed, thereby guaranteeing a stable indoor environment that meets comfort constraints while achieving energy-saving goals.

[0068] Specifically, the system identifies the type of air conditioner currently being controlled as either a variable frequency (VFD) or fixed frequency (FQF) air conditioner through device communication or user-preset information. It extracts the control command corresponding to the current control cycle from the optimal control sequence and parses the target power demand corresponding to that command. If it is a VFD air conditioner, the target power demand is converted into the corresponding compressor target frequency based on the air conditioner's power-frequency mapping relationship, or the target power is directly used as the control parameter (if the air conditioner supports direct power control). If it is a FQF air conditioner, the system first obtains the air conditioner's fixed rated power, calculates the ratio of the target power demand to the rated power, and obtains the compressor's duty cycle. The converted control parameters (target frequency, target power, or duty cycle) are sent to the air conditioner, controlling it to operate according to these parameters to achieve the optimal control action for the current control cycle.

[0069] This solution clarifies the quantitative boundaries of health and comfort by constructing a comfort constraint model, avoiding the neglect of comfort requirements during control. By establishing a system energy consumption model, it achieves accurate energy consumption prediction, providing scientific support for energy-saving optimization. Based on a dual-model approach and a preset optimization algorithm, it solves multi-objective optimization problems, minimizing air conditioning energy consumption while ensuring the indoor environment meets health and comfort requirements, effectively resolving the core contradiction between health and comfort and energy conservation. By executing actions in the current control cycle, collecting real-time feedback data, and continuously updating and optimizing, it forms a dynamic closed-loop control that continuously adapts to dynamic changes in the environment and needs, ensuring long-term stable comfort and energy-saving effects.

[0070] In some embodiments, the constraints of the multi-objective optimization problem are determined by the following method: in addition to the limitations of the comfort constraint model, physical limitations and control input change rate limitations of the air conditioning equipment are also included. The physical limitations of the air conditioning equipment include a maximum power limit, and the control input change rate limitation is used to avoid frequent start-stop of the air conditioning equipment and extend the service life of the equipment.

[0071] Solving multi-objective optimization problems requires balancing objective achievement with system stability. Besides the health and comfort boundaries defined by the comfort constraint model, air conditioning equipment itself has physical operating limits, and drastic changes in control input can affect equipment lifespan and operational stability. By incorporating the physical limitations of the air conditioning equipment and the rate of change of control input into the constraints, it is possible to ensure that the air conditioning operates within its own capabilities while meeting health and comfort requirements, and to avoid frequent start-ups, shutdowns, or significant adjustments, thus achieving a balance between optimization objectives and the safe and stable operation of the equipment.

[0072] Specifically, when constructing the multi-objective optimization problem, in addition to the allowable range of multi-dimensional constraint parameters defined by the comfort constraint model, two additional types of constraints are determined: first, the physical limitations of the air conditioning equipment, which corely include the maximum power limit of the compressor, and can also cover other hardware-allowed operating boundaries such as voltage and current; second, the control input change rate limit, which sets the maximum change amplitude of the control input (such as air conditioning power and operating frequency) within adjacent control cycles to avoid abrupt changes. The determined constraints are transformed into calculable mathematical expressions. For example, if the maximum power limit of the air conditioner is set as Pmax and the control input change rate limit is set as ΔPmax, the corresponding constraint expressions are "air conditioning operating power ≤ Pmax" and "power change in adjacent control cycles ≤ ΔPmax". The quantified physical limitations of the equipment and the control input change rate limit, together with the comfort constraint model, are used as constraints for the multi-objective optimization problem and input into the preset optimization algorithm. During the solution process, the optimization algorithm will search for the optimal solution that balances energy consumption reduction and comfort requirements while satisfying all constraints. The generated optimal control sequence will strictly follow the physical limitations of the equipment and the control input change rate limit to ensure the feasibility and safety of the control commands.

[0073] In some embodiments, the control cycle is set by the following methods: a fixed control cycle is set according to the rate of change of the indoor environment and the operating characteristics of the air conditioner, and the process of data acquisition, status update and optimization solution is repeated in each control cycle; or the control cycle duration is dynamically adjusted according to the severity of environmental changes, shortening the cycle when the environment changes rapidly and extending the cycle when the environment is stable.

[0074] The control cycle serves as the time benchmark for updating air conditioning control strategies, and its duration directly impacts control response speed and system resource consumption. The rate of change in the indoor environment and the operating characteristics of the air conditioning system determine the update requirements for the control strategy: a shorter cycle is needed for rapid environmental changes to quickly adapt, while a longer cycle can be used to reduce redundant calculations when the environment is stable. By combining these two key factors to set or dynamically adjust the control cycle, the accuracy of control effects can be ensured while achieving efficient utilization of system resources, guaranteeing that the execution of processes within each control cycle is targeted and reasonable.

[0075] Specifically, the core basis for setting the control cycle is to clearly define the rate of change of the indoor environment (such as the hourly change in temperature and humidity) and the operating characteristics of the air conditioner (such as air conditioner type, response speed, and start-stop constraints), ensuring that the cycle setting matches the actual operating conditions. If a fixed control cycle is selected, a uniform cycle length (such as 5 minutes) is set according to common environmental change patterns and general air conditioner operating characteristics to ensure timely response to changes in most operating conditions without excessive resource consumption. If a dynamic control cycle is selected, an environmental change rate judgment standard is established (such as temperature change exceeding 3°C per hour is considered "rapid change," and below 1°C is considered "stable"), combined with the air conditioner operating status (such as inverter air conditioners supporting shorter cycles, while fixed-frequency air conditioners need to avoid excessively short cycles leading to frequent start-stops), to formulate cycle adjustment rules. Within the set control cycle, the complete process of data acquisition (real-time feedback data collection), status update (air conditioner operating status and model parameter adjustment), and optimization solution (generating a new optimal control sequence) is completed sequentially to ensure that each cycle outputs a control strategy adapted to the current operating conditions. After each control cycle, assess the current rate of change of the indoor environment and the operating status of the air conditioner. If the environment changes rapidly, shorten the duration of the next cycle; if the environment tends to stabilize, extend the duration of the cycle to continuously adapt to changes in operating conditions.

[0076] In some embodiments, when constructing the comfort constraint model, the allowable range of multi-dimensional constraint parameters is defined by the following methods: integrating health standard data or industry standard requirements, and combining room type and user characteristics (such as the elderly and children), the allowable ranges of temperature, humidity, and air quality parameters are determined to ensure that the constraint range takes into account both health and personalized adaptation.

[0077] The core of the comfort constraint model is to define the permissible range of environmental parameters for health and comfort. This range is not set arbitrarily but must be based on authoritative health standards and adapted to the functional attributes of different rooms and the physiological characteristics of the users. By integrating health standard data and combining room type and user characteristics, the permissible range of multi-dimensional constraint parameters can be made more scientific and targeted, ensuring that the environment controlled by air conditioning not only meets the bottom line of health but also adapts to actual usage needs, providing reasonable and rigorous constraint boundaries for subsequent multi-objective optimization.

[0078] Specifically, the process begins by collecting authoritative health standard data or industry regulations to clarify the benchmark values ​​for core parameters related to human health and comfort (such as recommended ranges for temperature, humidity, and carbon dioxide concentration). Simultaneously, the room type where the air conditioner is located (e.g., bedroom, office area) and the characteristics of the main user groups (e.g., whether they are elderly or children) are determined. Based on the health standard data, adjustments are made according to the functional attributes of the room type. For example, bedrooms prioritize sleep comfort, so the permissible temperature range can be slightly higher than in an office setting; children's rooms have higher requirements for air quality, so the permissible upper limit for carbon dioxide concentration can be appropriately tightened. Further optimization of the parameter ranges is then performed based on the physiological characteristics of the user groups. For example, the elderly and children have weaker thermoregulation abilities, so the fluctuation range of the permissible temperature range needs to be smaller; allergy sufferers are more sensitive to humidity, so the permissible humidity range can be adjusted to a range less prone to mold growth. Finally, the adjusted and optimized parameter ranges are integrated to determine the permissible ranges for the multi-dimensional constraint parameters in the comfort constraint model, thus completing the construction of the comfort constraint model.

[0079] In some embodiments, after collecting real-time feedback data, the model parameters are updated by comparing the deviation between the real-time feedback data and the model prediction data, and adjusting the parameters of the comfort constraint model and the system energy consumption model based on the deviation using an iterative algorithm to reduce subsequent prediction errors and improve model adaptability.

[0080] The prediction accuracy of the comfort constraint model and the system energy consumption model directly affects the rationality of the optimal control sequence, while real-time feedback data can accurately reflect the actual operating state of the system. By comparing the deviation between the real-time feedback data and the model prediction data, mismatches between the model parameters and the actual operating conditions can be identified. Then, by using iterative algorithms to adjust the model parameters, the model deviation can be continuously corrected, ensuring that the model always conforms to the actual operating characteristics of the system.

[0081] Specifically, after collecting real-time feedback data (such as actual indoor temperature, humidity, and actual air conditioning energy consumption), this data is compared one-to-one with the model prediction data (such as predicted temperature and predicted energy consumption) from the same period, and the deviation values ​​of each parameter are calculated. A deviation threshold is set, and it is determined whether the calculated deviation value exceeds the threshold. If it does not exceed the threshold, it means that the model parameters are still suitable for the current operating conditions and no adjustment is needed for the time being; if it exceeds the threshold, the parameter adjustment process is initiated. Iterative algorithms (such as least squares method and gradient descent method) are used to gradually adjust the key parameters of the comfort constraint model and the system energy consumption model (such as the parameter allowable range boundary of the comfort constraint model, the performance coefficient of the system energy consumption model, and thermal inertia parameters) with the goal of "reducing the deviation value". After completing one parameter adjustment, the next round of prediction is performed based on the updated model parameters, and the deviation between the new prediction data and the real-time feedback data is compared. If the deviation value is less than the preset threshold, the current model parameters are fixed; if it still exceeds the threshold, the iterative adjustment steps are repeated until the deviation meets the requirements.

[0082] In some embodiments, during the execution of control actions, abnormal operating conditions are handled by the following methods: detecting whether environmental parameters exceed safety thresholds, whether the air conditioning equipment is operating abnormally, or whether there is conflicting user feedback. If any of the above abnormal operating conditions are detected, the current optimized control strategy is paused and the corresponding emergency control strategy is triggered.

[0083] During the execution of the optimal control sequence, air conditioners may experience abnormal operating conditions due to sudden changes in the external environment, equipment aging, or conflicting user needs. If these conditions are not addressed promptly, they may threaten user health and safety, damage equipment, or degrade the user experience. By monitoring key abnormal indicators in real time and triggering targeted emergency control strategies, a rapid response can be achieved when anomalies occur, prioritizing safety and core needs, preventing risks from escalating, and creating conditions for subsequent restoration of normal and optimized control.

[0084] Specifically, throughout the entire process of the air conditioner's control actions, three key indicators are continuously monitored: first, whether indoor and outdoor environmental parameters (such as temperature, humidity, and carbon dioxide concentration) exceed preset health and safety thresholds; second, whether there are any abnormalities in the air conditioning equipment's operating status (such as compressor start-stop frequency, voltage and current, and refrigerant pressure); and third, whether there are any contradictions in user feedback (such as consecutive feedback of "too cold" and "too hot" within a short period of time). If any of the above indicators meets the abnormality judgment conditions (such as temperature exceeding the 32℃ high-temperature threshold, compressor start-stop frequency exceeding 6 times / hour, or receiving contradictory feedback within 5 minutes), it is determined to be an abnormal operating condition, and the currently executing optimized control strategy is immediately suspended. Corresponding emergency measures are initiated based on the type of abnormality: when environmental parameters exceed safety thresholds, the extreme adjustment mode is activated (such as operating at maximum cooling power at high temperatures); when equipment malfunctions, related operating actions are stopped and a self-check is performed, while simultaneously generating an equipment health report; when users provide contradictory feedback, the current control status is maintained and the user is prompted to confirm their true preferences. After the emergency strategy is implemented, relevant indicators are continuously monitored. If environmental parameters return to a safe range, equipment self-test passes, or user preferences are confirmed, and the abnormal factors have been eliminated, then the normal optimized control process should be gradually restored; if the abnormality persists (e.g., equipment failure is not resolved), then the emergency status should be maintained and users or maintenance personnel should be continuously notified.

[0085] Figure 3 This is a flowchart illustrating another embodiment of the air conditioning control method, specifically including steps 1 to 4.

[0086] Step 1: Define comfort constraints and perform comfort constraint modeling. Clearly define the allowable ranges of multi-dimensional constraint parameters related to health and comfort, for example, setting the temperature range to 25-27℃ as the health and comfort boundary for subsequent control; introduce parameter data corresponding to authoritative health standards (such as industry specifications and human physiological tolerance ranges) to provide a basis for the rationality of comfort constraints.

[0087] Step 2: Establish a system energy consumption model. Construct an energy consumption model that includes equipment characteristic parameters such as the air conditioning coefficient of performance (COP) and thermal inertia to characterize the relationship between air conditioning operating parameters, environmental parameters, and energy consumption; import past energy consumption data of the air conditioning system and corresponding indoor and outdoor temperature / humidity and other environmental parameters to provide data support for the calibration of the system energy consumption model.

[0088] Step 3: Solve the multi-objective optimization problem using Model Predictive Control (MPC). Employing the MPC algorithm, based on a comfort constraint model and a system energy consumption model, solve the multi-objective optimization problem of "reducing energy consumption + meeting comfort constraints." The algorithm predicts the indoor environmental state (e.g., temperature changes) and air conditioning energy consumption trends over multiple future control cycles, providing a forward-looking basis for optimization. An objective function containing an "energy consumption calculation term" and a "comfort penalty term" is constructed. By minimizing this function (i.e., min(energy consumption) + comfort penalty), energy saving and comfort requirements are balanced. A sequence of control actions corresponding to multiple future control cycles is generated, with each action representing the optimal choice under the current operating conditions.

[0089] The objective function is: N represents the prediction time domain, α and β are weighting coefficients, and the comfort deviation is the difference between the actual parameters and the constraint range. MPC is solved online using numerical optimization methods (such as quadratic programming (QP) or genetic algorithms). Each step predicts the next N steps based on the current state, generating the optimal control sequence.

[0090] More specifically, in each control cycle, MPC needs to solve the following constrained multi-objective optimization problem:

[0091]

[0092] Where N represents the prediction time domain; α and β are weighting coefficients, and the ratio of α / β determines the trade-off between energy saving and comfort; the larger the ratio, the more the control strategy favors energy saving; T ref The comfort reference temperature is typically set to the midpoint of the comfort range, such as 26°C. The quadratic term (T) is used. in -T ref ) 2 It can penalize any deviation from the reference value, not just violations of the boundary.

[0093] Constraints include: System dynamics constraints: T in (i+1)=A·T in (i)+B·P ac (i) + ... (for i = 0, ..., N-1). Comfort constraint: Tmin ≤ Tin(i) ≤ Tmax (for i = 1, ..., N), for example, [Tmin, Tmax] = [25, 27]℃. Control input constraint: P ac,min ≤P_{ac}(i)≤P ac,max (For i=0,...,N-1), determined by the physical performance of the air conditioning equipment. Control input rate of change constraint (optional): |Pac(i+1)-Pac(i)|≤ΔPac,max, used to prevent frequent start-ups and shutdowns of the equipment and extend its lifespan.

[0094] The MPC loop process is as follows: At the current time k, obtain the real-time measurement value of the system (such as the current indoor temperature T(k)); using the system model, based on the current state x(k) and a set of future candidate control sequences U=[u(k),u(k+1),...,u(k+N-1)], predict the output Y=[y(k+1),y(k+2),...,y(k+N)] of the system in the next N time steps (prediction time domain); solve an optimization problem to find the optimal control sequence U* that minimizes the objective function (such as the lowest total energy consumption and the highest comfort); apply only the first control action u*(k) in the optimal sequence to the actual air conditioning system; at the next time k+1, update the system state with the new measurement value, and repeat the loop.

[0095] Step 4: Execute the optimal control sequence in a rolling fashion. Only the control actions corresponding to the current control cycle in the optimal control sequence are executed, and the actions of subsequent cycles will be dynamically updated based on real-time data. Adjust the air conditioner's operating status according to the instructions of the current control cycle, such as adjusting the compressor frequency to control the temperature. Collect real-time feedback data such as indoor environmental parameters and air conditioner operating status through sensors to reflect the current control effect. Substitute the real-time feedback data into the process, recalibrate the comfort constraint model and system energy consumption model, and start the next round of optimization and control to form a dynamic closed loop.

[0096] To make predictions, a model describing the room's thermodynamic properties and air conditioning energy consumption is needed. To balance accuracy and computational complexity, a first-order equivalent thermal parameter (RC) model is adopted.

[0097] First, a dynamic model of indoor temperature is established, treating the room as a thermodynamic system, with the indoor temperature Tin as the core state variable. The model equations can be simplified to: After discretization: T in (k+1)=A·T in (k)+B·P ac (k)+D·T out (k)+E·Q internal (k). Where T in (k) represents the indoor temperature (state variable) at time k; P ac (k) represents the cooling / heating power (control variable) of the air conditioner at time k, where a positive value indicates cooling and a negative value indicates heating; T out (k) represents the outdoor temperature at time k (measurable disturbance); Q internal (k) represents the internal thermal disturbance at time k, including heat generated by personnel and equipment, solar radiation, etc. (the disturbance can be estimated); A, B, D, and E are model parameters, obtained through system identification (e.g., using historical data to train a linear regression model); A is related to R (thermal resistance) and R (heat capacity), reflecting the thermal inertia of the building; B is related to the air conditioning efficiency η.

[0098] Next, a system energy consumption model was established. The air conditioning energy consumption Eac is related to its power and coefficient of performance (COP). ac (k)=P ac (k)·Δt. A more accurate model can consider the variation of COP with operating conditions: COP(k)=f(T) out (k),T in (k)), P ac,electrical (k)=P ac (k) / COP(k). Where P ac,electrical (k) represents the actual electrical power consumed by the power grid, which is the ultimate goal to minimize.

[0099] Figure 4 The network topology diagram shows a cloud-based AI server used to run complex MPC optimization algorithms, store historical data, and train machine learning models. Edge controllers (such as smart gateways and air conditioner companions) receive cloud commands, execute local control, and handle emergencies. Air conditioning equipment consists of inverter or fixed-frequency units supporting Wi-Fi / Bluetooth control. The sensor suite includes temperature and humidity sensors, CO2 sensors, and human presence sensors. User terminals include mobile apps, voice assistants, and smart panels.

[0100] Figure 5 To control the timing diagram, the specific steps include: steps 11 to 13.

[0101] Step 11, data acquisition stage: Read the current indoor temperature, humidity, and CO2 concentration data to obtain core environmental status information; use infrared, millimeter wave and other technologies to detect whether there are people staying indoors; collect the current outdoor temperature, humidity, sunshine and other weather data, and query peak and off-peak electricity price signals at the same time.

[0102] Step 12, Intelligent Decision-Making Phase: In "Away from Home Mode," the indoor temperature is maintained at 28℃ (summer) or 18℃ (winter) to balance energy conservation and basic environmental protection. If the geofence detects that the user is about to return, the pre-temperature adjustment process is initiated. The air conditioner is turned on 30 minutes in advance for pre-cooling (summer) or pre-heating (winter) to ensure that the environment is comfortable when the user arrives home. If neither of the above two scenarios is met, the system enters the normal operation "Normal Occupied Mode." In "Normal Occupied Mode," the Model Predictive Control (MPC) algorithm is run to calculate the optimal setpoint curve. Based on the calculation results of the MPC algorithm, corresponding air conditioner operation control commands are generated. If an open window is detected indoors, the air conditioner operation is suspended to avoid energy waste. If an alarm is triggered due to excessively high indoor temperature, the emergency cooling mode is activated. If an air conditioner malfunction is detected, the user is immediately notified for repair.

[0103] Step 13, Control Execution Phase: For variable frequency air conditioners, send precise power control commands to adjust the compressor's operating status; for fixed frequency air conditioners, calculate and execute PWM duty cycle control to achieve approximate power regulation; save the execution data of this control (such as energy consumption and environmental changes) for subsequent model updates and optimizations.

[0104] The technical solution of this embodiment constructs a comfort constraint model to define the allowable range of multi-dimensional constraint parameters related to health and comfort, and establishes a system energy consumption model characterizing the relationship between air conditioning operating parameters, equipment characteristic parameters, environmental parameters, and energy consumption. Based on the above dual models, a multi-objective optimization problem aimed at reducing energy consumption and meeting comfort constraints is solved through a preset optimization algorithm, generating an optimal control sequence that includes control actions for multiple future control cycles. The air conditioner is controlled to execute the control actions of the current control cycle, real-time feedback data is collected, the air conditioner operating status and dual model parameters are updated, and the subsequent optimal control sequence is optimized. This effectively solves the limitations of traditional fixed-mode control, improves the adaptability to changes in the external environment and system characteristics, and enhances the balance between health and comfort protection and high-efficiency energy saving.

[0105] According to an embodiment of the present invention, an air conditioner control device corresponding to the air conditioner control method is also provided. See also Figure 2 The schematic diagram shown is a structural diagram of an embodiment of the device of the present invention. The control device of the air conditioner may include: a modeling unit 101, an optimization unit 102, and a control unit 103.

[0106] Modeling unit 101 is configured to construct a comfort constraint model, which defines the allowable range of multi-dimensional constraint parameters related to health and comfort. For the specific functions and processing of this unit, please refer to step S110.

[0107] A comfort constraint model is a model that defines the permissible range of multi-dimensional constraint parameters related to human health and perceived comfort in an indoor environment. Its core function is to set rigid boundaries for air conditioning control to ensure that the indoor environment meets human health needs and comfort standards. Multi-dimensional constraint parameters refer to environmental parameters directly related to human health and comfort, including temperature, humidity, and air quality parameters (such as carbon dioxide concentration). The permissible range of these parameters needs to be determined comprehensively based on health standards, human physiological needs, and usage scenarios.

[0108] Human beings have specific parameter ranges for their health and comfort needs in indoor environments; exceeding these ranges can negatively impact physical health or sensory experience. Constructing a comfort constraint model aims to transform abstract health and comfort needs into quantifiable and actionable constraints, providing clear target boundaries for subsequent optimization control and preventing air conditioning control from solely pursuing energy consumption or a single parameter while neglecting health and comfort.

[0109] Specifically, the first step is to identify the types of multi-dimensional constraint parameters related to health and comfort. Combining health standards (such as relevant industry regulations and human physiological tolerance ranges), usage scenario characteristics, and population characteristics, the allowable ranges for each parameter are determined. Subsequently, these constraints are quantified and mathematically modeled to form a complete comfort constraint model, which can be dynamically adjusted according to actual needs. For example, for a bedroom sleep scenario, considering the slowed metabolism and temperature sensitivity changes during sleep, the allowable range for temperature constraint parameters is determined to be 26℃-28℃, the allowable range for humidity constraint parameters is 40%-60%, and the allowable range for carbon dioxide concentration constraint parameters is below 1000ppm. Integrating these parameter ranges and related constraint logic completes the construction of the comfort constraint model for this scenario.

[0110] In some implementations, the modeling unit 101 is further configured to: acquire user feedback data or identify the current usage scenario of the air conditioner; adjust the allowable range of the multi-dimensional constraint parameters according to the user feedback data or the comfort demand data corresponding to the usage scenario; wherein the user feedback data includes explicit feedback data and implicit feedback data, the explicit feedback data being preference instructions input by the user through an interactive terminal, and the implicit feedback data being potential demand data obtained based on the user's behavior analysis of adjusting the air conditioner.

[0111] User feedback data refers to information related to indoor health and comfort needs conveyed by users through direct input or behavioral performance. This includes both explicit and implicit feedback data and serves as a crucial basis for adjusting comfort constraint models. Usage scenarios refer to the specific context in which the air conditioner operates, related to the user's activity level or environmental characteristics. Users' health and comfort needs differ across different scenarios. Comfort requirement data refers to the environmental parameter standards that can meet the user's health and comfort experience within the corresponding usage scenario, including reasonable ranges for parameters such as temperature, humidity, and air quality.

[0112] Users' health and comfort needs are not fixed but dynamically change depending on their feelings, activity levels, and the context in which they are situated. By acquiring user feedback data or identifying the current usage scenario, these dynamic changes in needs can be accurately captured. Adjusting the allowable range of multi-dimensional constraint parameters based on the changed comfort needs data ensures that the comfort constraint model always matches the actual needs of users, thereby making the control strategies generated by subsequent optimization algorithms more aligned with personalized and scenario-based requirements.

[0113] Specifically, on the one hand, explicit feedback data actively input by users is received through user interaction terminals, or implicit feedback data is obtained by analyzing user behavior data of manually adjusting the air conditioner. On the other hand, sensors detect the activity status and time information of people indoors to identify the current usage scenario of the air conditioner (such as a sleep scenario or an office scenario). For the acquired user feedback data, the corresponding comfort preferences are analyzed (e.g., the need for higher temperatures in response to feedback of "too cold"). For the identified usage scenario, preset comfort requirement data for that scenario is retrieved (e.g., the reasonable range of temperature and humidity for an office scenario). Based on the matched comfort requirement data, the allowable range of multi-dimensional constraint parameters in the comfort constraint model is dynamically adjusted to ensure that the adjusted range still meets human health requirements while aligning with the user's current needs. The adjusted comfort constraint model is then used as new input to a preset optimization algorithm, which solves a multi-objective optimization problem based on the updated constraints to generate the optimal control sequence adapted to the current needs.

[0114] The modeling unit 101 is also configured to establish a system energy consumption model, which is used to characterize the correlation between air conditioning operating parameters, equipment characteristic parameters, environmental parameters and air conditioning energy consumption. For the specific functions and processing of this unit, please refer to step S120.

[0115] A system energy consumption model is a model that characterizes the relationship between air conditioning operating parameters, equipment characteristic parameters, environmental parameters, and air conditioning energy consumption. It can accurately predict the energy consumption of air conditioning under different operating conditions, providing data support for energy-saving optimization. Air conditioning operating parameters refer to the adjustable parameters during air conditioning operation, including cooling power, heating power, and operating frequency; these are the core parameters for adjusting the air conditioning status. Equipment characteristic parameters refer to the inherent parameters of the air conditioning equipment that affect energy consumption, including the coefficient of performance (COP) and thermal inertia-related parameters. The COP reflects the energy conversion efficiency of the air conditioning, while thermal inertia-related parameters reflect the heat storage effect of the building or equipment. Environmental parameters refer to external and internal environmental data that affect air conditioning operation and the indoor environment, including indoor and outdoor temperatures, solar radiation intensity, and the status of indoor occupants.

[0116] Air conditioning energy consumption is affected by various factors, including air conditioning operating parameters, equipment characteristics, and environmental parameters, and these factors have complex interrelationships. Establishing a system energy consumption model can accurately quantify these relationships, enabling the prediction of air conditioning energy consumption under different operating conditions and avoiding energy waste caused by blind control.

[0117] Specifically, the key parameters affecting air conditioning energy consumption are first identified, including air conditioning operating parameters (such as cooling power and operating frequency), equipment characteristic parameters (such as coefficient of performance and thermal inertia), and environmental parameters (such as indoor and outdoor temperatures and solar radiation intensity). Then, through theoretical analysis (such as deriving energy consumption equations based on thermodynamic principles) and data support (such as collecting historical energy consumption data and environmental parameter data), a mathematical model is constructed that can characterize the relationship between each parameter and energy consumption. Finally, the model parameters are calibrated using historical data to improve the model's prediction accuracy. For example, based on thermodynamic principles, the energy consumption equation E(t) = COP × P(t) + thermal inertia term (where E(t) is energy consumption and P(t) is equipment power) is derived. Historical energy consumption data under different air conditioning operating powers, indoor and outdoor temperatures, and solar radiation intensities over the past month are collected. Regression methods are used to estimate the specific values ​​of COP and thermal inertia-related parameters in the equation, completing the establishment of a system energy consumption model. This model can predict the air conditioning energy consumption under corresponding operating conditions based on the input current operating parameters and environmental parameters.

[0118] In some implementations, the modeling unit 101 is further configured to: collect historical operating data of the air conditioner, the historical operating data including historical energy consumption statistics, historical environmental parameter change data and corresponding air conditioner operating parameter data; and train and calibrate the parameters of the system energy consumption model based on the historical operating data.

[0119] The initial parameters of system energy consumption models are usually based on theoretical derivations or general equipment characteristics, which differ from the actual equipment status and environmental conditions during air conditioner operation, resulting in insufficient model prediction accuracy. Historical operating data records the real correlation between energy consumption, environment, and operating parameters during actual air conditioner operation. By training and calibrating the model parameters using this data, the model can accurately match the actual operating characteristics of the air conditioner, thereby improving the accuracy of energy consumption prediction.

[0120] Specifically, the system collects historical operating data of the air conditioner according to a preset period (e.g., daily, weekly), including historical energy consumption statistics (e.g., hourly energy consumption), historical environmental parameter change data (e.g., indoor and outdoor temperature, humidity, and solar radiation intensity during the same period), and corresponding air conditioner operating parameter data (e.g., operating power, frequency), ensuring data integrity and synchronization. The collected historical data is screened and cleaned, removing abnormal data (e.g., energy consumption data during equipment failure, and environmental parameters abnormally collected by sensors), and missing data is appropriately supplemented to ensure data quality. Simultaneously, the data is converted into a format that the model can process, facilitating subsequent parameter calibration. The preprocessed historical data is input into the system's energy consumption model, and appropriate algorithms (e.g., linear regression, neural networks) are used for parameter optimization. By minimizing the deviation between the model's predicted energy consumption and historical actual energy consumption, key parameters in the model, such as performance coefficients and thermal inertia, are adjusted. The calibrated model parameters are validated using some historical data not used in training. If the prediction deviation meets the preset accuracy requirements (e.g., deviation less than 5%), the parameters of the system's energy consumption model are updated; if the deviation does not meet the requirements, the training algorithm is readjusted or supplemented with data and retrained until the model accuracy meets the requirements.

[0121] The optimization unit 102 is configured to solve a constrained multi-objective optimization problem based on the comfort constraint model and the system energy consumption model using a preset optimization algorithm, generating an optimal control sequence containing control actions corresponding to multiple future control cycles. The objectives of the multi-objective optimization problem include reducing air conditioning energy consumption and ensuring that the indoor environment meets the allowable range of the comfort constraint model. For the specific functions and processing of this unit, please refer to step S130.

[0122] A pre-defined optimization algorithm refers to a pre-set algorithm used to solve multi-objective optimization problems. It possesses forward-looking prediction and dynamic optimization capabilities, and can achieve multi-objective balance while satisfying constraints. A multi-objective optimization problem refers to an optimization problem with the core objectives of reducing air conditioning energy consumption and ensuring the indoor environment meets the allowable range of the comfort constraint model. It requires finding the optimal trade-off relationship among multiple objectives. The optimal control sequence refers to the sequence of control actions corresponding to multiple future control cycles, obtained after solving the pre-defined optimization algorithm. Each control action is the optimal choice under the current operating conditions. The control cycle refers to the time interval for updating the air conditioning control strategy. It is set according to the rate of change of the indoor environment and the operating characteristics of the air conditioning system to ensure that the control strategy can respond promptly to changes in the environment and demand.

[0123] The core challenge in air conditioning control is balancing health and comfort with energy consumption. Single-objective control can lead to trade-offs (e.g., pursuing energy saving at the expense of comfort, or pursuing comfort at the expense of high energy consumption). Pre-defined optimization algorithms possess foresight and multi-objective optimization capabilities, enabling them to find the control scheme that minimizes energy consumption while satisfying comfort constraints. The generation of the optimal control sequence provides a clear basis for subsequent phased execution of control actions.

[0124] Specifically, the comfort constraint model and the system energy consumption model are used as inputs to clarify the core objectives of the multi-objective optimization problem (reducing air conditioning energy consumption and ensuring that the indoor environment meets the allowable range of the comfort constraint model). The system state (such as changes in indoor environmental parameters and energy consumption) in the future multiple control cycles is predicted by a preset optimization algorithm, and an objective function containing energy consumption calculation terms and comfort deviation calculation terms is constructed. The objective function is solved under constraints such as comfort constraints and equipment physical limitations. Finally, the optimal control sequence containing the control actions corresponding to the future multiple control cycles is obtained, and each control action corresponds to the optimal air conditioning operating state under a specific operating condition.

[0125] For example, assuming a control cycle of 5 minutes and a prediction time domain of 24 control cycles (i.e., the next 2 hours), the constructed comfort constraint model (such as temperature 25℃-27℃, humidity 40%-60%) and system energy consumption model are input into the model predictive control algorithm. The algorithm predicts the changes in outdoor temperature and solar radiation intensity at different time points in the next 2 hours, constructs the objective function min (energy consumption calculation term + comfort deviation calculation term), and solves it under constraints such as temperature not exceeding 25℃-27℃ and air conditioning power not exceeding 0kW-3.5kW, generating the air conditioning operating power sequence corresponding to each 5 minutes in the next 2 hours. This sequence is the optimal control sequence.

[0126] In some implementations, the preset optimization algorithm is a model predictive control algorithm, which quantifies the multi-objective optimization problem by constructing an objective function; the objective function includes an energy consumption calculation term and a comfort deviation calculation term, and the trade-off between the energy consumption calculation term and the comfort deviation calculation term is adjusted by preset weight coefficients.

[0127] Model predictive control (MMC) is a model-based closed-loop optimization control algorithm. It predicts the system state over a future period, solves multi-objective optimization problems to generate the optimal control sequence, and continuously corrects the model and control strategy based on real-time feedback data. It is suitable for complex control scenarios with multiple constraints and objectives. The objective function transforms the multi-objective optimization problem into a quantifiable mathematical expression. Solving this expression reveals the optimal trade-off between multiple objectives and is the core computational basis of the multi-objective optimization algorithm. The energy consumption calculation term is a mathematical term in the objective function used to quantify the air conditioning energy consumption level. Its value is related to air conditioning operating parameters, equipment characteristic parameters, and environmental parameters; a lower value indicates lower energy consumption. The comfort deviation calculation term is a mathematical term in the objective function used to quantify the deviation between the indoor environment and the allowable range of the comfort constraint model. Its value is the quantified result of the difference between the actual indoor environmental parameters and the allowable range of the comfort constraint model; a lower value indicates that the indoor environment better meets health and comfort requirements. The weighting coefficient is used to adjust the importance of the energy consumption calculation item and the comfort deviation calculation item in the objective function. By adjusting the weighting coefficient, the emphasis on the two objectives of energy saving and comfort can be controlled.

[0128] Model predictive control algorithms possess forward-looking prediction and multi-constraint optimization capabilities. They can predict the air conditioning operation status and energy consumption level for multiple future control cycles based on comfort constraint models and system energy consumption models. By constructing an objective function that includes energy consumption calculation terms and comfort deviation calculation terms, the dual objectives of "reducing energy consumption" and "meeting comfort constraints" can be transformed into mathematical problems. Combined with the adjustment of preset weight coefficients, the optimal solution that balances both objectives can be finally obtained.

[0129] Specifically, when constructing the objective function, the mathematical expression for the energy consumption calculation item is determined. This expression is based on the system energy consumption model, taking air conditioning operating parameters, equipment characteristic parameters, and environmental parameters as input variables, and outputting the corresponding quantitative energy consumption value. The mathematical expression for the comfort deviation calculation item is also determined. This expression is based on the comfort constraint model, calculating the deviation values ​​between the actual indoor temperature, humidity, air quality, and other parameters and the model's allowable range, and quantifying these deviation values. The energy consumption calculation item and the comfort deviation calculation item are then integrated to form a complete objective function, expressed as: F = w1 × f1 + w2 × f2 (F is the objective function value, w1 and w2 are preset weight coefficients, f1 is the energy consumption calculation item, and f2 is the comfort deviation calculation item). Preset weight coefficients are set according to the usage scenario. For example, in an office scenario, the weight coefficient of the comfort deviation calculation item can be increased to prioritize personnel comfort; in an unmanned scenario, the weight coefficient of the energy consumption calculation item can be increased to prioritize energy saving. The constructed objective function is input into the model predictive control algorithm. The algorithm combines the boundary conditions of the comfort constraint model and the parameter correlation of the system energy consumption model to solve for the minimum value of the objective function and generate the corresponding optimal control sequence.

[0130] In some implementations, the optimization unit 102 is further configured to: acquire environmental prediction data and user behavior prediction data for a future preset time period, and use the environmental prediction data and user behavior prediction data as input to the preset optimization algorithm; the environmental prediction data includes weather change prediction data, and the user behavior prediction data includes personnel entry and exit time prediction data.

[0131] The future preset time period refers to a pre-defined future time interval used by the algorithm to predict changes in system state. Its duration needs to be determined based on the air conditioning control cycle, environmental change patterns, and optimization requirements; it can be set to 2 hours. Environmental prediction data refers to the prediction results of changes in external environmental parameters affecting air conditioning operation and the indoor environment within the future preset time period. Its core function is to provide the optimization algorithm with forward-looking environmental change information, assisting in generating more adaptive control strategies. Weather change prediction data is a core type of environmental prediction data, including predicted values ​​of weather-related parameters such as outdoor temperature, humidity, solar radiation intensity, and wind speed within the future preset time period. These are key external factors affecting air conditioning load and energy consumption. User behavior prediction data is the prediction result of users' indoor activity status within the future preset time period, used to match actual user needs and avoid a disconnect between control strategies and user behavior. Personnel entry and exit time prediction data is a core type of user behavior prediction data, referring to the predicted values ​​of the time points and number of people entering and leaving the room within the future preset time period, directly affecting indoor heat load and comfort requirements.

[0132] The operating status and energy consumption level of air conditioners are not only affected by the current environment and user status, but also closely related to environmental changes and user behavior changes in the future. By acquiring environmental prediction data and user behavior prediction data for a preset period of time in the future, and using them as input to a preset optimization algorithm, the algorithm can anticipate future system disturbances (such as rising outdoor temperatures or users returning home). Based on this forward-looking information, the algorithm solves multi-objective optimization problems, and the generated optimal control sequence can adapt to future changes in advance, avoiding energy waste or decreased comfort caused by passive adjustments.

[0133] Specifically, the required environmental prediction data and user behavior prediction data types are clearly defined; based on the air conditioning control cycle (e.g., 5 minutes) and optimization requirements, a future preset time period (e.g., the next 2 hours) is set. Weather change prediction data for the future preset time period is acquired; by analyzing historical user activity data and combining it with geofencing and other technologies, the entry and exit times of personnel within the future preset time period are predicted. The acquired weather change prediction data and personnel entry and exit time prediction data are preprocessed to remove outliers and convert the data into a format recognizable and processed by the preset optimization algorithm (e.g., a standardized numerical sequence). The preprocessed environmental prediction data and user behavior prediction data, along with the comfort constraint model and system energy consumption model, are input into the preset optimization algorithm; the algorithm, combined with the current system state (e.g., current indoor temperature, air conditioning operating power), predicts the indoor environmental change trend and energy consumption change trend within the future preset time period, solves the constrained multi-objective optimization problem, and generates the optimal control sequence containing control actions corresponding to multiple future control cycles.

[0134] The control unit 103 is configured to control the air conditioner to execute the control action corresponding to the current control cycle in the optimal control sequence, and to collect real-time feedback data. Based on the real-time feedback data, it updates the air conditioner's operating status and the parameters of the comfort constraint model and the system energy consumption model, thereby optimizing the generation process of the subsequent optimal control sequence. The specific functions and processing of this unit are described in step S140.

[0135] Real-time feedback data refers to control-related data collected in real time during air conditioning operation, including indoor environmental parameters, occupant status data, and air conditioning operating status data. This data reflects the current control effectiveness and system status. Indoor and external conditions are constantly changing; a fixed control sequence cannot adapt to long-term dynamic changes, potentially causing subsequent control actions to deviate from the optimal state. Executing only the control action for the current control cycle ensures that each control cycle's action is the most suitable choice for the current situation. However, by updating the system status and model parameters through real-time feedback data and then optimizing subsequent control sequences, a dynamic closed-loop control can be formed, continuously adapting to changes in the environment and demands, ensuring long-term health, comfort, and energy-saving effects.

[0136] Specifically, the control action corresponding to the current control cycle is extracted from the optimal control sequence, and the control command is sent to the air conditioner to execute the action. During the operation of the air conditioner, real-time feedback data such as indoor environmental parameters, personnel status data, and air conditioner operation status data are collected. Based on this real-time feedback data, it is determined whether the current system state is consistent with the model prediction, the air conditioner operation status parameters are adjusted, and the relevant parameters of the comfort constraint model and the system energy consumption model are calibrated. Based on the updated system state and model parameters, the multi-objective optimization problem is solved again to generate a new optimal control sequence, which provides a basis for the control action of the next control cycle, and so on.

[0137] For example, if the current control cycle is a 5-minute period, the air conditioner operating power corresponding to this period is extracted from the optimal control sequence as 2.0kW, and the air conditioner is controlled to operate at 2.0kW. Within these 5 minutes, data such as indoor temperature, humidity, and actual air conditioner operating power are collected in real time. It is found that the actual rate of change of indoor temperature is slightly higher than the model prediction. Based on this feedback data, the thermal inertia-related parameters in the system energy consumption model are calibrated, while confirming that the temperature allowable range of the comfort constraint model is still 25℃-27℃. Based on the updated model parameters and the current indoor temperature of 26℃ and outdoor temperature of 32℃, the multi-objective optimization problem is solved again to generate the optimal control action for the next 5 minutes (such as adjusting the air conditioner operating power to 1.8kW), ensuring that subsequent control can still meet comfort requirements and achieve optimal energy consumption.

[0138] In some implementations, the control unit 103 controls the air conditioner to perform control actions corresponding to the current control cycle in the optimal control sequence, including: for a variable frequency air conditioner, mapping the control command to the target frequency or target power of the compressor; for a fixed frequency air conditioner, calculating the duty cycle of the compressor operation and executing the control command based on the duty cycle.

[0139] Different types of air conditioners (inverter and fixed-frequency) have fundamentally different operating mechanisms. Inverter air conditioners support continuous power adjustment, while fixed-frequency air conditioners can only achieve on / off control. Considering the structural characteristics of these two types of air conditioners, converting a unified control command (target power demand) into specific control parameters adapted to their operating mechanisms ensures that the control actions in the optimal control sequence can be accurately executed, thereby guaranteeing a stable indoor environment that meets comfort constraints while achieving energy-saving goals.

[0140] Specifically, the system identifies the type of air conditioner currently being controlled as either a variable frequency (VFD) or fixed frequency (FQF) air conditioner through device communication or user-preset information. It extracts the control command corresponding to the current control cycle from the optimal control sequence and parses the target power demand corresponding to that command. If it is a VFD air conditioner, the target power demand is converted into the corresponding compressor target frequency based on the air conditioner's power-frequency mapping relationship, or the target power is directly used as the control parameter (if the air conditioner supports direct power control). If it is a FQF air conditioner, the system first obtains the air conditioner's fixed rated power, calculates the ratio of the target power demand to the rated power, and obtains the compressor's duty cycle. The converted control parameters (target frequency, target power, or duty cycle) are sent to the air conditioner, controlling it to operate according to these parameters to achieve the optimal control action for the current control cycle.

[0141] This solution clarifies the quantitative boundaries of health and comfort by constructing a comfort constraint model, avoiding the neglect of comfort requirements during control. By establishing a system energy consumption model, it achieves accurate energy consumption prediction, providing scientific support for energy-saving optimization. Based on a dual-model approach and a preset optimization algorithm, it solves multi-objective optimization problems, minimizing air conditioning energy consumption while ensuring the indoor environment meets health and comfort requirements, effectively resolving the core contradiction between health and comfort and energy conservation. By executing actions in the current control cycle, collecting real-time feedback data, and continuously updating and optimizing, it forms a dynamic closed-loop control that continuously adapts to dynamic changes in the environment and needs, ensuring long-term stable comfort and energy-saving effects.

[0142] Since the processing and functions implemented by the device in this embodiment are basically the same as the embodiments, principles and examples of the aforementioned methods, any details not covered in the description of this embodiment can be found in the relevant descriptions in the aforementioned embodiments, and will not be repeated here.

[0143] The technical solution of this invention constructs a comfort constraint model to define the allowable range of multi-dimensional constraint parameters related to health and comfort, and establishes a system energy consumption model characterizing the relationship between air conditioning operating parameters, equipment characteristic parameters, environmental parameters, and energy consumption. Based on the above dual models, a multi-objective optimization problem aimed at reducing energy consumption and meeting comfort constraints is solved through a preset optimization algorithm, generating an optimal control sequence that includes control actions for multiple future control cycles. The air conditioner is controlled to execute the control actions of the current control cycle, real-time feedback data is collected, the air conditioner operating status and dual model parameters are updated, and the subsequent optimal control sequence is optimized. This effectively solves the limitations of traditional fixed-mode control, improves the adaptability to changes in the external environment and system characteristics, and enhances the balance between health and comfort protection and high-efficiency energy saving.

[0144] According to an embodiment of the present invention, an air conditioner corresponding to an air conditioner control device is also provided. This air conditioner may include the air conditioner control device described above.

[0145] Since the processing and functions implemented by the air conditioner in this embodiment are basically the same as the embodiments, principles and examples of the aforementioned device, any details not covered in the description of this embodiment can be found in the relevant descriptions in the aforementioned embodiments, and will not be repeated here.

[0146] The technical solution of this invention constructs a comfort constraint model to define the allowable range of multi-dimensional constraint parameters related to health and comfort, and establishes a system energy consumption model characterizing the relationship between air conditioning operating parameters, equipment characteristic parameters, environmental parameters, and energy consumption. Based on the above dual models, a multi-objective optimization problem aimed at reducing energy consumption and meeting comfort constraints is solved through a preset optimization algorithm, generating an optimal control sequence that includes control actions for multiple future control cycles. The air conditioner is controlled to execute the control actions of the current control cycle, real-time feedback data is collected, the air conditioner operating status and dual model parameters are updated, and the subsequent optimal control sequence is optimized. This effectively solves the limitations of traditional fixed-mode control, improves the adaptability to changes in the external environment and system characteristics, and enhances the balance between health and comfort protection and high-efficiency energy saving.

[0147] According to an embodiment of the present invention, a storage medium corresponding to an air conditioner control method is also provided, the storage medium including a stored program, wherein the program controls the device where the storage medium is located to execute the air conditioner control method described above when it is executed.

[0148] Since the processing and functions implemented by the storage medium in this embodiment are basically the same as the embodiments, principles and examples of the aforementioned methods, any details not covered in this embodiment can be found in the relevant descriptions in the aforementioned embodiments, and will not be repeated here.

[0149] The technical solution of this invention constructs a comfort constraint model to define the allowable range of multi-dimensional constraint parameters related to health and comfort, and establishes a system energy consumption model characterizing the relationship between air conditioning operating parameters, equipment characteristic parameters, environmental parameters, and energy consumption. Based on the above dual models, a multi-objective optimization problem aimed at reducing energy consumption and meeting comfort constraints is solved through a preset optimization algorithm, generating an optimal control sequence that includes control actions for multiple future control cycles. The air conditioner is controlled to execute the control actions of the current control cycle, real-time feedback data is collected, the air conditioner operating status and dual model parameters are updated, and the subsequent optimal control sequence is optimized. This effectively solves the limitations of traditional fixed-mode control, improves the adaptability to changes in the external environment and system characteristics, and enhances the balance between health and comfort protection and high-efficiency energy saving.

[0150] According to an embodiment of the present invention, a computer program product corresponding to the control method for an air conditioner is also provided. The computer program product includes a computer program that, when processed and executed, implements the steps of the control method for the air conditioner described above.

[0151] Since the processing and functions implemented by the computer program product in this embodiment are basically corresponding to the embodiments, principles and examples of the aforementioned methods, any details not covered in the description of this embodiment can be found in the relevant descriptions in the aforementioned embodiments, and will not be repeated here.

[0152] The technical solution of this invention constructs a comfort constraint model to define the allowable range of multi-dimensional constraint parameters related to health and comfort, and establishes a system energy consumption model characterizing the relationship between air conditioning operating parameters, equipment characteristic parameters, environmental parameters, and energy consumption. Based on the above dual models, a multi-objective optimization problem aimed at reducing energy consumption and meeting comfort constraints is solved through a preset optimization algorithm, generating an optimal control sequence that includes control actions for multiple future control cycles. The air conditioner is controlled to execute the control actions of the current control cycle, real-time feedback data is collected, the air conditioner operating status and dual model parameters are updated, and the subsequent optimal control sequence is optimized. This effectively solves the limitations of traditional fixed-mode control, improves the adaptability to changes in the external environment and system characteristics, and enhances the balance between health and comfort protection and high-efficiency energy saving.

[0153] In summary, it is readily understood by those skilled in the art that, without conflict, the aforementioned advantageous methods can be freely combined and superimposed.

[0154] The above description is merely an embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A method for controlling an air conditioner, characterized in that, include: A comfort constraint model is constructed, which is used to define the allowable range of multi-dimensional constraint parameters related to health and comfort; A system energy consumption model is established, which is used to characterize the relationship between air conditioning operating parameters, equipment characteristic parameters, environmental parameters and air conditioning energy consumption; Based on the comfort constraint model and the system energy consumption model, a preset optimization algorithm is used to solve the constrained multi-objective optimization problem, generating an optimal control sequence containing control actions corresponding to multiple future control cycles; the objectives of the multi-objective optimization problem include reducing air conditioning energy consumption and making the indoor environment conform to the allowable range of the comfort constraint model. The system controls the air conditioner to execute the control action corresponding to the current control cycle in the optimal control sequence, and collects real-time feedback data. Based on the real-time feedback data, the system updates the air conditioner's operating status and the parameters of the comfort constraint model and the system energy consumption model, thereby optimizing the generation process of the subsequent optimal control sequence.

2. The air conditioning control method according to claim 1, characterized in that, The preset optimization algorithm is a model predictive control algorithm. The model predictive control algorithm quantifies the multi-objective optimization problem by constructing an objective function. The objective function includes an energy consumption calculation term and a comfort deviation calculation term. The trade-off between the energy consumption calculation term and the comfort deviation calculation term is adjusted by preset weight coefficients.

3. The air conditioning control method according to claim 1 or 2, characterized in that, The method further includes: Obtain environmental prediction data and user behavior prediction data for a future preset time period, and use the environmental prediction data and user behavior prediction data as input to the preset optimization algorithm; the environmental prediction data includes weather change prediction data, and the user behavior prediction data includes personnel entry and exit time prediction data.

4. The air conditioning control method according to claim 1, characterized in that, The method further includes: Obtain user feedback data or identify the current usage scenario of the air conditioner; Based on the user feedback data or the comfort requirement data corresponding to the usage scenario, adjust the allowable range of the multi-dimensional constraint parameters; The user feedback data includes explicit feedback data and implicit feedback data. The explicit feedback data is the preference instructions input by the user through the interactive terminal, and the implicit feedback data is the potential demand data obtained based on the user's behavior of adjusting the air conditioner.

5. The air conditioning control method according to claim 1, characterized in that, Controlling the air conditioner to execute the control action corresponding to the current control cycle in the optimal control sequence includes: For inverter air conditioners, control commands are mapped to the compressor's target frequency or target power. For fixed-frequency air conditioners, the duty cycle of the compressor is calculated, and control commands are executed based on the duty cycle.

6. The air conditioning control method according to claim 1, characterized in that, The method further includes: Collect historical operating data of the air conditioner, including historical energy consumption statistics, historical environmental parameter change data, and corresponding air conditioner operating parameter data; The parameters of the system energy consumption model are trained and calibrated based on the historical operating data.

7. A control device for an air conditioner, characterized in that, include: The modeling unit is configured to construct a comfort constraint model, which is used to define the allowable range of multi-dimensional constraint parameters related to health and comfort; The modeling unit is also configured to establish a system energy consumption model, which is used to characterize the relationship between air conditioning operating parameters, equipment characteristic parameters, environmental parameters and air conditioning energy consumption. The optimization unit is configured to solve a constrained multi-objective optimization problem based on the comfort constraint model and the system energy consumption model using a preset optimization algorithm, and generate an optimal control sequence containing control actions corresponding to multiple future control cycles; the objectives of the multi-objective optimization problem include reducing air conditioning energy consumption and making the indoor environment conform to the allowable range of the comfort constraint model; The control unit is configured to control the air conditioner to perform the control action corresponding to the current control cycle in the optimal control sequence, and to collect real-time feedback data, update the air conditioner operating status and the parameters of the comfort constraint model and the system energy consumption model based on the real-time feedback data, thereby optimizing the generation process of the subsequent optimal control sequence.

8. An air conditioner, characterized in that, include: The air conditioning control device as described in claim 7.

9. A storage medium, characterized in that, The storage medium includes a stored program, wherein, when the program is executed, it controls the device containing the storage medium to perform the air conditioning control method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the air conditioning control method according to any one of claims 1 to 6.

Images