Indoor temperature control method and device, electronic equipment and readable storage medium

Abstract

The present disclosure relates to the technical field of temperature control, and provides an indoor temperature control method and device, electronic equipment and a readable storage medium. The method comprises: acquiring running state data of a temperature control device corresponding to a current time according to the current time, and acquiring indoor and outdoor environment characteristic data corresponding to the current time, the temperature control device being a cold station device or a heat station device; predicting a load demand in an indoor environment in a next time period according to the indoor and outdoor environment characteristic data, to obtain a target load in the indoor environment in the next time period; wherein the load demand is a cold demand or a heat demand; predicting a running parameter of the temperature control device according to the target load, the running state data and the indoor and outdoor environment characteristic data, to generate target running state data of the temperature control device; and controlling the temperature control device according to the target running state data, thereby significantly improving the performance of the temperature control device and achieving more refined and personalized temperature control.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of temperature control, and particularly relates to an indoor temperature control method and device, electronic equipment and a readable storage medium. BACKGROUND

[0002] With the development of science and technology, public places such as shopping malls and office buildings build factory area intelligent temperature control equipment to improve energy utilization efficiency and management efficiency, save energy investment cost and reduce carbon emissions. When maintaining indoor temperature, the temperature control equipment usually consumes a large amount of cold or heat energy. The existing temperature control equipment collects indoor and outdoor temperature, humidity and other data through sensors, and automatically adjusts the running state of the temperature control equipment according to the preset control logic, but relies on fixed control strategy, cannot accurately predict the cold and heat demand, lacks deep analysis and prediction ability of real-time data, and cannot realize fine and personalized temperature control of indoor temperature. SUMMARY

[0003] Therefore, the embodiments of the present disclosure provide an indoor temperature control method, device, electronic equipment and readable storage medium to solve the problem that the temperature control equipment in the prior art cannot accurately predict the cold and heat demand, thereby controlling the indoor temperature inaccurately.

[0004] In a first aspect, an embodiment of the present disclosure provides an indoor temperature control method, comprising:

[0005] obtaining running state data of a temperature control equipment corresponding to a current time according to the current time, and obtaining indoor and outdoor environmental feature data corresponding to the current time, the temperature control equipment being a cold station equipment or a heat station equipment;

[0006] predicting a target load in a next time period in the indoor according to the indoor and outdoor environmental feature data, the target load being a cold load or a heat load;

[0007] predicting a target running state data of the temperature control equipment according to the target load, the running state data and the indoor and outdoor environmental feature data;

[0008] controlling the temperature control equipment according to the target running state data.

[0009] In a second aspect, an embodiment of the present disclosure provides an indoor temperature control device, comprising:

[0010] a obtaining module configured to obtain running state data of a temperature control equipment corresponding to a current time according to the current time, and obtain indoor and outdoor environmental feature data corresponding to the current time, the temperature control equipment being a cold station equipment or a heat station equipment;

[0011] The prediction module is configured to predict the load demand in the next time period in the indoor environment according to the indoor and outdoor environment characteristic data, to obtain a target load in the next time period in the indoor environment; wherein the load demand is cold demand or heat demand;

[0012] The strategy generation module is configured to predict the operation parameter of the temperature control device according to the target load, the operation state data and the indoor and outdoor environment characteristic data, to generate target operation state data of the temperature control device;

[0013] The control module is configured to control the temperature control device according to the target operation state data.

[0014] In a third aspect, an electronic device is provided, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor executes the computer program to implement the steps of the above method.

[0015] In a fourth aspect, a readable storage medium is provided, which stores a computer program, and the computer program is executed by a processor to implement the steps of the above method.

[0016] The beneficial effects of this disclosed embodiment compared to the prior art are as follows: By acquiring the current operating status data of temperature control equipment (such as cooling or heating station equipment) and indoor and outdoor environmental characteristic data, an accurate data foundation is provided for subsequent prediction and control. This allows the temperature control equipment to make predictions and decisions based on the environment and equipment status, improving the real-time performance and accuracy of predictions. Through the establishment of a prediction model that can capture the complex relationship between environmental parameters and load demand, the indoor load demand for the next time period is predicted based on indoor and outdoor environmental characteristic data, obtaining the target indoor load for the next time period. The prediction model accurately predicts future cooling or heating load demand. Based on the target load, current equipment status, and environmental characteristics, multiple factors are comprehensively considered to predict the optimal operating state of the temperature control equipment for the next time period, generating target operating state data for the temperature control equipment to achieve more refined and personalized temperature control. The temperature control equipment is controlled according to the predicted target operating state data to reach the predetermined operating state, thereby meeting the indoor temperature demand. The temperature control method disclosed herein utilizes indoor and outdoor environmental characteristic data for load prediction, rather than relying solely on a single model or parameter, thereby improving the accuracy and robustness of prediction. By acquiring real-time operating status data of the temperature control equipment, and dynamically adjusting the control strategy using operating status data, target load, and indoor and outdoor environmental characteristic data, it overcomes the problem in existing technologies where temperature control equipment cannot accurately predict heating and cooling demands, resulting in inaccurate indoor temperature control. By comprehensively considering target load, equipment status, and environmental characteristics, it optimizes equipment operating parameters, minimizing energy consumption while ensuring indoor comfort, significantly improving the performance of temperature control equipment, and achieving more refined and personalized temperature control. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this disclosure, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic flowchart of an indoor temperature control method provided in an embodiment of this disclosure;

[0019] Figure 2 This is a schematic flowchart of another indoor temperature control method provided in this embodiment of the present disclosure;

[0020] Figure 3 This is a schematic flowchart of another indoor temperature control method provided in this embodiment of the present disclosure;

[0021] Figure 4This is a flowchart illustrating a sequential feature selection algorithm provided in an embodiment of this disclosure;

[0022] Figure 5 This is a schematic flowchart of another indoor temperature control method provided in this disclosure embodiment;

[0023] Figure 6 This is a schematic diagram of the structure of an indoor temperature control device provided in an embodiment of this disclosure;

[0024] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this disclosure. Detailed Implementation

[0025] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, so as to provide a thorough understanding of the embodiments of this disclosure. However, those skilled in the art will understand that this disclosure may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this disclosure with unnecessary detail.

[0026] An indoor temperature control method and apparatus according to an embodiment of the present disclosure will now be described in detail with reference to the accompanying drawings.

[0027] Figure 1 This is a schematic flowchart of an indoor temperature control method provided in an embodiment of this disclosure. Figure 1 As shown, the indoor temperature control method includes:

[0028] Step 101: Obtain the operating status data of the temperature control equipment corresponding to the current time, and obtain the indoor and outdoor environmental characteristic data corresponding to the current time. The temperature control equipment is either a cold station equipment or a hot station equipment.

[0029] In some embodiments, different temperature control devices operate at different times. For example, in June, the operating temperature control devices are chiller plant equipment; in December, the operating temperature control devices are heating plant equipment. Specifically, the operating status data of the chiller plant equipment is data obtained after preprocessing the collected raw data (i.e., the current initial operating status data). By preprocessing the raw data (i.e., the current initial operating status data), the raw data is effectively cleaned and transformed to obtain the operating status data of the chiller plant equipment, which can provide reliable input data for subsequent machine learning models, improving the accuracy and generalization ability of the models. The operating status data of the chiller plant equipment may include chilled water supply and return temperature difference, chilled water supply temperature, chilled water return temperature, chilled water supply pressure, chilled water return pressure, flow rate, chilled pump frequency, number of cooling towers, number of chillers, set temperature, and operating status of the chillers, etc. The operating status data of the heating station equipment consists of preprocessed raw data (i.e., current initial operating status data). This data may include the hot water supply and return water temperature difference, hot water supply temperature, hot water return temperature, hot water supply pressure, hot water return pressure, flow rate, hot water pump frequency, number of boilers, number of hot water pumps, set temperature, and boiler operating status. The operating status data corresponding to the current time period is the operating status data for that period, for example, 9:00 to 10:00. The indoor and outdoor environmental characteristic data consists of preprocessed raw data (i.e., initial indoor and outdoor environmental characteristic data). This data may include current time characteristics, indoor wet-bulb temperature, indoor average temperature, indoor humidity, indoor and outdoor temperature difference, indoor and outdoor wet-bulb temperature difference, indoor temperature median, indoor temperature variance, outdoor temperature for the next time period, feature vectors of outdoor weather for the current time period, and feature vectors of outdoor weather for the next time period. The current time period may be 9:00 to 10:00, and the next time period may be 10:00 to 11:00.

[0030] Acquiring operational status data of temperature control equipment (such as chiller or heater equipment) and indoor / outdoor environmental characteristics at the current time is a crucial prerequisite for achieving accurate and efficient temperature control. Real-time data acquisition allows temperature control equipment to promptly understand its operating status and current environmental parameters, providing a basis for subsequent control decisions. The operational status data of the temperature control equipment and the corresponding indoor / outdoor environmental characteristics data provide essential input information for subsequent load forecasting. By utilizing more information sources and making predictions based on the latest actual conditions, temperature control equipment can quickly respond to changes in the current environment and adjust its operating status to adapt to constantly changing demands.

[0031] Step 102: Based on indoor and outdoor environmental characteristic data, predict the indoor load demand for the next time period to obtain the target indoor load for the next time period; where the load demand is either cooling demand or heating demand.

[0032] In some embodiments, the indoor load demand for the next time period is predicted based on indoor and outdoor environmental characteristic data. Advanced prediction algorithms (such as machine learning and deep learning) are used to accurately predict the indoor cooling or heating demand for the next time period, i.e., the target load. Existing temperature control often uses a single prediction model, which struggles to accurately capture the complex relationship between indoor and outdoor environmental parameters and cooling / heating load demand, leading to errors in the prediction of cooling and heating demand and thus hindering more refined and personalized temperature control. The target load obtained from the above prediction can reflect the heat or cooling required to meet the target indoor temperature. Accurately predicting the load demand for the next time period helps to rationally generate optimal operating state data for the next time period. On the one hand, this avoids excessive energy consumption and waste, thereby improving energy efficiency; on the other hand, it makes the indoor temperature more stable, enabling more refined and personalized temperature control to meet user needs and achieve personalized control.

[0033] Step 103: Based on the target load, operating status data and indoor and outdoor environmental characteristic data, predict the operating parameters of the temperature control equipment and generate the target operating status data of the temperature control equipment.

[0034] In some embodiments, by integrating target load, indoor and outdoor environmental characteristic data, and operating status characteristic data, data analysis algorithms or machine learning models can be used to simultaneously consider the stability and efficiency of temperature control, avoid frequent power on / off cycles and over-adjustment leading to increased energy consumption or equipment damage, predict the target operating status data to which temperature control should be adjusted, and the target operating status data is the optimized recommended value. This is a control strategy that meets the needs while achieving optimal energy saving, which helps temperature control to respond quickly and maintain optimal working efficiency when facing future load changes, while ensuring stable indoor temperature.

[0035] Specifically, when the temperature control is for chiller plant equipment, the target operating status data may include: chiller system name: the name or identifier of the chiller plant system; project name: the name of the current project or shopping mall; indoor target temperature: the set target indoor temperature; target number of cooling towers: the set number of cooling towers to be operated; number of chilled pumps: the set number of chilled pumps to be operated; target frequency: the set operating frequency of the chilled pumps; target number of chillers: the set number of chillers to be operated; target set temperature: the set temperature of the chillers, etc.

[0036] Specifically, when the temperature control is for a heating station, the target operating status data may include: heat source system name: the name or identifier of the heating station system; project name: the name of the current project or shopping mall; indoor target temperature: the set target indoor temperature; target number of hot water pumps: the set number of hot water pumps to be operated; target frequency: the set operating frequency of the hot water pumps; target number of boilers: the set number of boilers to be operated; target set temperature: the set temperature of the boilers, etc.

[0037] Step 104: Control the temperature control equipment based on the target operating status data.

[0038] In some embodiments, the predicted target operating state data is applied to the actual temperature control equipment. By adjusting the equipment's operating parameters, the temperature control equipment operates according to the target operating state data, achieving automated control and reducing reliance on manual operation. The temperature control equipment can be chiller or heating plant equipment. Specific operations may include adjusting parameters such as the start-up and shutdown status of the chiller unit, the frequency of the chilled water pump, and the valve opening degree to meet the predicted target cooling load demand. By converting the predicted results (i.e., the target operating state data) into actual control commands, the predicted results are transformed into actual actions, enabling the temperature control equipment to operate as expected. By accurately controlling the operating status of the temperature control equipment, more complex real-world situations can be better addressed, achieving more precise and personalized temperature control. This can better maintain the indoor temperature within the set comfort range, improving the user experience.

[0039] Furthermore, the temperature control equipment continuously collects actual indoor and outdoor environmental characteristic data and operational status data during operation, and feeds this data back to the temperature control equipment for subsequent parameter updates and optimization. Specifically, the feedback adjustment mechanism includes: continuously collecting actual indoor and outdoor environmental characteristic data and operational status data of the temperature control equipment; feeding back the collected indoor and outdoor environmental characteristic data and operational status data of the temperature control equipment to the temperature control equipment for parameter updates and optimization of the temperature control system; updating and adjusting the parameters of the temperature control equipment based on the feedback data to improve its accuracy and predictive capabilities; and adjusting the generated control strategy (i.e., the selected target operational status data) based on the updated temperature control equipment to adapt to current environmental changes and energy consumption demands. By continuously collecting actual indoor and outdoor environmental characteristic data and operational status data of the temperature control equipment and feeding it back to the temperature control equipment, timely adjustments and optimizations are made to the temperature control equipment, thereby improving the accuracy and stability of prediction and control. This feedback adjustment mechanism enables the temperature control equipment to continuously adapt to changes in the indoor and outdoor environment and achieve intelligent management and optimization of the temperature control equipment.

[0040] The indoor temperature control method provided in this disclosure acquires the current operating status data of temperature control equipment (such as chiller or heater equipment) and indoor and outdoor environmental characteristic data, providing an accurate data foundation for subsequent prediction and control. This allows the temperature control equipment to make predictions and decisions based on the environment and equipment status, improving the real-time performance and accuracy of predictions. By establishing a prediction model that captures the complex relationship between environmental parameters and load demand, the indoor load demand for the next time period is predicted based on indoor and outdoor environmental characteristic data, resulting in the target indoor load for the next time period. The prediction model accurately predicts future cooling or heating load demands. Based on the target load, current equipment status, and environmental characteristics, multiple factors are integrated to predict the optimal operating state of the temperature control equipment for the next time period, generating target operating state data for the temperature control equipment to achieve more refined and personalized temperature control. The temperature control equipment is then controlled according to the predicted target operating state data to reach the predetermined operating state, thereby meeting the indoor temperature requirements. The temperature control method disclosed herein utilizes indoor and outdoor environmental characteristic data for load prediction, rather than relying solely on a single model or parameter, thereby improving the accuracy and robustness of prediction. By acquiring real-time operating status data of the temperature control equipment, and dynamically adjusting the control strategy using operating status data, target load, and indoor and outdoor environmental characteristic data, it overcomes the problem in existing technologies where temperature control equipment cannot accurately predict heating and cooling demands, resulting in inaccurate indoor temperature control. By comprehensively considering target load, equipment status, and environmental characteristics, it optimizes equipment operating parameters, minimizing energy consumption while ensuring indoor comfort, significantly improving the performance of temperature control equipment, and achieving more refined and personalized temperature control.

[0041] In some embodiments, when the temperature control device is a chiller plant device, predicting the indoor load demand for the next time period based on indoor and outdoor environmental characteristic data to obtain the target indoor load for the next time period includes: inputting indoor and outdoor environmental characteristic data into a chiller prediction model to obtain the target chiller output by the chiller prediction model for the next time period; and determining the target chiller as the target load.

[0042] Specifically, the aforementioned cooling capacity prediction model can fit the indoor and outdoor environmental parameters of the current time period with the required cooling capacity for the next time period. The model is trained based on historical indoor and outdoor environmental characteristic data and the corresponding historical generated cooling capacity for the next period. The historical generated cooling capacity can be calculated based on the historical operating status data of the chiller plant equipment. For example, the historical operating status data of the chiller plant equipment includes flow rate, chilled water supply temperature, and chilled water return temperature. The corresponding generated cooling capacity can be calculated based on these parameters. Indoor and outdoor environmental characteristic data are input into the cooling capacity prediction model. Based on this data, the indoor cooling capacity demand for a future period is predicted. The target indoor cooling capacity output by the model for the next time period provides a data foundation for subsequent steps. This target cooling capacity output by the model is then determined as the target load for the next time period. Using the predicted target cooling capacity as the target load provides a basis for subsequent control of the chiller plant equipment. This allows the target operating status of the equipment to be controlled based on the target cooling capacity generation temperature, thereby meeting the predicted cooling capacity demand.

[0043] In some embodiments, when constructing the cooling load prediction model, indoor and outdoor environmental feature data are used as the original dataset. This data may include multiple sub-features such as current time features, indoor wet-bulb temperature, indoor average temperature, indoor humidity, indoor-outdoor temperature difference, indoor-outdoor wet-bulb temperature difference, indoor median temperature, indoor temperature variance, outdoor temperature for the next time period, feature vectors of outdoor weather in the current time period, and feature vectors of outdoor weather for the next time period. An optimal feature subset can be obtained by using a Sequential Backward Floating Selection (SBFS) algorithm, as the features in the optimal subset are the most informative for the cooling load prediction model. The cooling load prediction model to be trained can be a Lightweight Gradient Boosting Machine (LightGBM), with a specified hyperparameter search space. Then, random search or grid search is used to determine the optimal hyperparameters to improve model performance. The model is trained using the optimal feature subset and historical generated cooling loads. By fitting the relationship between environmental parameters and the actual cooling load demand in the next time period, the cooling load prediction model can accurately predict the cooling load demand in the next time period. The model is evaluated using metrics such as root mean square error (RMSE) and mean absolute error (MAE) to verify its accuracy and stability, thus obtaining the trained cooling load prediction model.

[0044] In some embodiments, when the temperature control device is a heat station device, predicting the indoor load demand for the next time period based on indoor and outdoor environmental characteristic data to obtain the target indoor load for the next time period includes: inputting indoor and outdoor environmental characteristic data into a heat prediction model to obtain the target indoor heat for the next time period output by the heat prediction model; and determining the target heat as the target load.

[0045] Specifically, the aforementioned heat prediction model can fit the indoor and outdoor environmental parameters of the current time period with the heat demand of the next time period. The heat prediction model is trained based on historical indoor and outdoor environmental characteristic data and the historical heat generation data of the heating station equipment for the next period. The historical heat generation data is calculated based on the historical operating status data of the heating station equipment. For example, the historical operating status data of the heating station equipment includes flow rate, hot water supply temperature, and hot water return temperature. The corresponding heat generation data can be calculated based on these parameters. Indoor and outdoor environmental characteristic data are input into the heat prediction model. Based on this data, the indoor heat demand for a future period is predicted, resulting in the predicted indoor heat demand for the next time period, providing a data foundation for subsequent steps. The target cooling capacity output by the heat prediction model is determined as the target load for the next time period. The predicted target heat demand is used as the target load to provide a basis for subsequent control of the heating station equipment. This allows the target operating status of the equipment to be controlled based on the target heat generation temperature, thus meeting the predicted heat demand.

[0046] In some embodiments, when constructing a heat prediction model, indoor and outdoor environmental feature data are used as the original dataset. This data may include multiple sub-features such as current time features, indoor wet-bulb temperature, indoor average temperature, indoor humidity, indoor-outdoor temperature difference, indoor-outdoor wet-bulb temperature difference, indoor median temperature, indoor temperature variance, outdoor temperature for the next time period, feature vectors of outdoor weather in the current time period, and feature vectors of outdoor weather for the next time period. An optimal feature subset can be obtained through SBFS (Simplified Method Search), as the features in this subset are the most informative for the heat prediction model. The heat prediction model to be trained can be a LightGBM model, with a specified hyperparameter search space. Then, random search or grid search is used to determine the optimal hyperparameters to improve model performance. The model is trained using the optimal feature subset and historical generated heat data. By fitting the relationship between environmental parameters, equipment operating status, and actual heat demand, the heat prediction model can accurately predict the heat demand for the next time period. RMSE (Real-Time Sequence) or MAE (Magnitude Equalization) metrics are used to evaluate the heat prediction model, verifying its accuracy and stability, resulting in a fully trained heat prediction model.

[0047] In some embodiments, reference Figure 2When the temperature control equipment is a chiller plant equipment, step 103 includes steps 201 to 206. The indoor and outdoor environmental characteristic data include the indoor and outdoor wet-bulb temperature difference and the indoor average temperature. The operating status data includes the chilled water supply temperature and the chilled water return temperature.

[0048] Based on target load, operating status data, and indoor and outdoor environmental characteristic data, the operating parameters of the temperature control equipment are predicted, generating target operating status data for the temperature control equipment, including:

[0049] Step 201: Input the indoor and outdoor wet-bulb temperature difference and the target cooling capacity as a set of data into the cooling tower number prediction model to obtain the target number of cooling towers for the next time period output by the cooling tower number prediction model.

[0050] Specifically, the indoor-outdoor wet-bulb temperature difference reflects the humidity conditions of the indoor and outdoor air and serves as an important reference factor for cooling tower efficiency. A higher wet-bulb temperature indicates a higher moisture content in the air, which may affect the cooling effect of the cooling tower. The target cooling capacity is the cooling demand required in the next time period. The cooling towers need to adjust their operating status according to the target cooling capacity to meet this demand. The aforementioned cooling tower quantity prediction model can fit the indoor-outdoor wet-bulb temperature difference, the target cooling capacity, and the target number of cooling towers for the next time period. By inputting the indoor-outdoor wet-bulb temperature difference and the target cooling capacity as a set of data into the cooling tower quantity prediction model, the required number of cooling towers is predicted, resulting in the target number of cooling towers for the chiller plant equipment in the next time period. This helps to ensure sufficient cooling capacity to support the target cooling capacity demand while avoiding overuse or underuse of cooling towers. Cooling towers are important heat dissipation components in chiller plant equipment, and their number directly affects heat dissipation efficiency. Through a comprehensive analysis of the indoor-outdoor wet-bulb temperature difference and the target cooling capacity, the most economical and effective cooling tower configuration is determined, resulting in the target number of cooling towers to meet the expected target cooling capacity demand. The cooling tower quantity prediction model comprehensively considers the impact of environmental conditions on the cooling effect to ensure that the system can effectively dissipate heat under different environments.

[0051] Furthermore, the cooling tower quantity prediction model is trained based on historical indoor and outdoor environmental characteristic data, historical operating status data of chiller plant equipment, target cooling capacity, and the corresponding historical cooling tower quantity. The historical cooling tower quantity refers to the number of cooling towers in the operating status data of the next time period from the aforementioned historical operating status data. Specifically, the historical operating status data of chiller plant equipment may include chilled water supply and return temperature difference, chilled water supply temperature, chilled water return temperature, chilled water supply pressure, chilled water return pressure, flow rate, chilled pump frequency, number of cooling towers, number of chillers, set temperature, and chiller operating status, etc. Historical indoor and outdoor environmental characteristic data may include current time characteristics, indoor wet-bulb temperature, indoor average temperature, indoor humidity, indoor and outdoor temperature difference, indoor and outdoor wet-bulb temperature difference, indoor temperature median, indoor temperature variance, outdoor temperature of the next time period, feature vector of outdoor weather in the current time period, and feature vector of outdoor weather in the next time period, etc. When constructing the cooling tower quantity prediction model, historical indoor and outdoor environmental characteristic data, historical operating status data of the cooling plant equipment, and target cooling capacity are used as the original dataset. The original dataset includes multiple sub-features, and the optimal feature subset can be obtained through SBFS (Simplified Black-Scholes Search). The features in the optimal feature subset are the most informative for the cooling tower quantity prediction model. Specifically, the optimal feature subset can include the indoor and outdoor wet-bulb temperature difference and the target cooling capacity. The cooling tower quantity prediction model to be trained can be a Ridge model, and a hyperparameter search space is specified. Then, random search or grid search is used to determine the optimal hyperparameters to improve the model's performance. The model is trained using the optimal feature subset and historical cooling tower quantities. By fitting the relationship between the input parameters and the actual number of cooling towers, the cooling tower quantity prediction model can accurately predict the number of cooling towers in the next time period. The model is evaluated using metrics such as RMSE or MAE to verify its accuracy and stability, resulting in the trained cooling tower quantity prediction model.

[0052] Step 202: Input the target cooling capacity into the chiller quantity prediction model to obtain the target chiller quantity for the next time period output by the chiller quantity prediction model.

[0053] Specifically, chillers are the main equipment generating cooling capacity. By predicting the number of chillers needed, the load can be rationally allocated, avoiding overload operation of a single chiller or too many chillers being idle. The target cooling capacity is directly input into the chiller quantity prediction model, which then predicts the number of chillers that need to be activated in the next time period based on the target cooling capacity. This helps the chiller plant equipment provide sufficient cooling capacity to meet the target cooling demand. Through the chiller quantity prediction model, the number of chillers can be dynamically adjusted according to the target cooling capacity, avoiding resource waste or insufficient cooling capacity caused by too many or too few chillers.

[0054] Furthermore, the chiller quantity prediction model is trained based on historical generated cooling capacity and the corresponding historical chiller quantity. The trained chiller quantity prediction model can fit the relationship between generated cooling capacity and the number of chillers in operation. The chiller quantity prediction model to be trained can be a Ridge model, with a specified hyperparameter search space. Then, random search or grid search is used to determine the optimal hyperparameters to improve model performance. The model is trained using historical generated cooling capacity and the corresponding historical chiller quantity. By fitting the relationship between the input parameters and the actual chiller quantity, the chiller quantity prediction model can accurately predict the number of chillers in the next time period. The model is evaluated using metrics such as root mean square error or mean absolute error to verify its accuracy and stability, resulting in the completed chiller quantity prediction model.

[0055] Step 203: Input the indoor average temperature and chilled water supply temperature as a set of data into the chiller set temperature prediction model to obtain the target set temperature for the next time period output by the chiller set temperature prediction model.

[0056] Specifically, the indoor average temperature is the average indoor temperature within the current time period, obtained from multiple indoor temperature sensors. The time period can be one hour, and each sensor can transmit temperature data every five minutes. The indoor average temperature is an important reference for the chiller setpoint temperature; a higher indoor temperature may require a lower chiller setpoint temperature for rapid cooling. The chilled water supply temperature reflects the current temperature status of the chilled water in the chiller plant. If the supply water temperature is high, the chiller setpoint temperature needs to be adjusted to lower the chilled water temperature. The chiller setpoint temperature prediction model described above can predict the target temperature that the chiller should be set to in the next time period based on the input indoor average temperature and chilled water supply temperature. This helps optimize the chiller's operating efficiency, ensures that the chilled water reaches the required temperature, and meets the requirements of indoor temperature control.

[0057] Furthermore, the chiller setpoint temperature prediction model is trained based on historical indoor and outdoor environmental characteristic data, historical operating status data of the chiller plant equipment, target cooling capacity, and corresponding historical setpoint temperatures. The historical setpoint temperature is the setpoint temperature in the operating status data of the next time period from the aforementioned historical operating status data. When constructing the chiller setpoint temperature prediction model, the historical indoor and outdoor environmental characteristic data, historical operating status data of the chiller plant equipment, and target cooling capacity are used as the original dataset. The original dataset includes multiple sub-features, and the optimal feature subset can be obtained through SBFS (Simultaneous Bitwise Search). The features in the optimal feature subset are the most informative for the chiller setpoint temperature prediction model. Specifically, the optimal feature subset may include the indoor average temperature and the chilled water supply temperature. The chiller setpoint temperature prediction model to be trained can be a Ridge model, with a specified hyperparameter search space. Then, random search or grid search is used to determine the optimal hyperparameters to improve the model's performance. The model is trained using the optimal feature subset and historical set temperatures. By fitting the relationship between the input parameters and the actual set temperature, the chiller set temperature prediction model can accurately predict the chiller set temperature for the next time period. The model is evaluated using indicators such as root mean square error or mean absolute error to verify its accuracy and stability, thus obtaining the trained chiller set temperature prediction model.

[0058] Step 204: Obtain multiple candidate frequencies of the chilled water pump and the candidate flow rate corresponding to each candidate frequency, and generate candidate heat corresponding to each candidate frequency based on the candidate flow rate corresponding to each candidate frequency, the target set temperature and the chilled water return temperature.

[0059] Step 205: Calculate the absolute value of the difference between the candidate heat and the target cooling amount corresponding to each candidate frequency, and determine the candidate frequency that meets the preset requirements as the target frequency for the next time period. The preset requirements are that the frequency is less than or equal to the first preset value and the absolute value of the difference between the candidate heat and the target cooling amount is the smallest.

[0060] Step 206: Generate target operating status data for the temperature control equipment based on the target frequency, target set temperature, target number of chillers, and target number of cooling towers.

[0061] In some embodiments, multiple candidate frequencies of the chilled pump and corresponding candidate flow rates are obtained, where each candidate frequency represents a different frequency value at which the chilled pump may operate. Each candidate frequency has a corresponding flow rate value. The candidate frequencies are input into a chilled pump frequency model, which uses internal algorithms and parameters to calculate and output the corresponding candidate flow rate. By obtaining multiple candidate frequencies of the chilled pump and corresponding candidate flow rates, several possible options can be provided for subsequent heat calculations and frequency optimization.

[0062] Specifically, the aforementioned chilled pump frequency model can fit the flow rate based on the chilled pump frequency. When training the chilled pump frequency model, the chilled pump frequencies from various historical operating state data are used as training samples, and the corresponding flow rates from each historical operating state data are used as the labels for the samples, thus constructing a training set. The chilled pump frequency model to be trained can be a Ridge model, and Bayesian search is used to determine the model hyperparameters. MAPE is then used to evaluate the model performance, resulting in a trained flow-temperature model. Through the chilled pump frequency model, the flow output under different chilled pump frequency models can be quantified, providing a basis for subsequent selection of the target frequency from multiple candidate frequencies.

[0063] Specifically, the target setpoint temperature is the target temperature of the chiller output chilled water predicted by the chiller setpoint temperature prediction model described above. The chilled water return temperature is the chilled water return temperature for the current time period. The chilled water return temperature reflects the actual cooling effect, and the difference between the chilled water return temperature and the target setpoint temperature affects the actual cooling capacity. For each candidate frequency, the corresponding flow rate, target setpoint temperature, and chilled water return temperature are used to calculate the candidate heat corresponding to each candidate frequency, converting the chiller pump frequency into a specific cooling capacity for subsequent comparison and selection. The absolute value of the difference between the candidate heat corresponding to each candidate frequency and the target cooling capacity is calculated, and a frequency that meets the preset requirements is selected from all candidate frequencies as the target frequency for the next time period. The preset requirements are that the absolute value of the difference between the candidate cooling capacity corresponding to the target frequency and the target cooling capacity is less than or equal to a first preset value, and the absolute value of the difference between the candidate cooling capacity corresponding to the target frequency and the target cooling capacity is minimized. By calculating the difference, the gap between the cooling capacity generated by the chiller pump at different frequencies and the actual required cooling capacity can be quantified. The first preset value is the tolerance range of the difference, ensuring that the operating frequency of the chiller pump does not deviate too far from the target cooling capacity. The candidate frequency with the smallest difference is selected as the target frequency, enabling the chilled pump to operate as efficiently as possible while meeting the target cooling capacity. The determined target frequency, target set temperature, target number of chillers, and target number of cooling towers are integrated to form complete target operating status data for the temperature control equipment, which is used to control the operation of the chiller plant equipment in the next time period.

[0064] In some embodiments, reference Figure 3 Step 103 may include steps 301 to 304, that is, when the temperature control device is a heat station device, the indoor and outdoor environmental characteristic data include the indoor temperature median and indoor temperature variance, and the operating status data includes the hot water supply and return water temperature difference and the hot water supply temperature.

[0065] Based on target load, operating status data, and indoor and outdoor environmental characteristic data, the operating parameters of the temperature control equipment are predicted, generating target operating status data for the temperature control equipment, including:

[0066] Step 301: Obtain the indoor target temperature, the first temperature difference between the indoor target temperature and the median indoor temperature, and input the median indoor temperature, the indoor target temperature, the hot water supply and return temperature difference and the first temperature difference as a set of data into the hot water pump number prediction model to obtain the target number of hot water pumps for the next time period output by the hot water pump number prediction model.

[0067] Specifically, indoor temperature data can be collected within a set time period (e.g., 1 hour) using temperature sensors installed indoors. The sensors collect data every 5 minutes, yielding multiple initial indoor temperatures. The initial indoor temperatures from previous time periods reflect the changes in indoor temperature within the current time period. The collected initial indoor temperature data are sorted, and the median indoor temperature is determined based on the number of data points. The median indoor temperature reflects the overall level of indoor temperature within the current time period. The indoor temperature variance, calculated from multiple initial indoor temperatures, reflects the degree of indoor temperature fluctuation and helps assess indoor temperature stability. The hot water supply and return temperature difference can be calculated using the hot water supply and return temperatures. This difference reflects heat loss or transfer efficiency during hot water circulation, a key parameter for evaluating the efficiency of the heating station equipment. A larger supply and return temperature difference indicates that the heating station equipment provides more heat. The hot water supply temperature reflects the current operating temperature of the heating station equipment and is an important reference for adjusting its operating status. The target indoor temperature is the indoor temperature that the user wishes to achieve or maintain, and can be set by the user. Calculate the first temperature difference between the target indoor temperature and the median indoor temperature. The first temperature difference reflects the gap between the current indoor temperature and the target temperature.

[0068] Specifically, the aforementioned hot water pump quantity prediction model can fit the median indoor temperature, target indoor temperature, hot water supply and return water temperature difference, first temperature difference, and the target number of hot water pumps for the next time period. By inputting the median indoor temperature, target indoor temperature, hot water supply and return water temperature difference, and first temperature difference as a set of data into the hot water pump quantity prediction model, the required number of hot water pumps is predicted, yielding the target number of hot water pumps for the heating station equipment in the next time period. This helps determine if there is sufficient heating capacity to support the target heat demand, while avoiding overuse or underuse of hot water pumps. Hot water pumps are crucial components in heating station equipment, and their number directly affects heat production efficiency. Through comprehensive analysis of the median indoor temperature, target indoor temperature, hot water supply and return water temperature difference, and first temperature difference, the most economical and efficient hot water pump configuration is determined, resulting in the target number of hot water pumps to meet the expected target heat demand.

[0069] Furthermore, the hot water pump quantity prediction model is trained based on historical indoor and outdoor environmental characteristic data, historical operating status data of the heating station equipment, target heat volume, and the corresponding historical hot water pump quantity. The historical hot water pump quantity refers to the number of hot water pumps in the operating status data of the next time period from the aforementioned historical operating status data. Specifically, the historical operating status data of the heating station equipment may include the hot water supply and return water temperature difference, hot water supply temperature, hot water return temperature, hot water supply pressure, hot water return pressure, flow rate, hot water pump frequency, number of boilers, number of hot water pumps, set temperature, boiler operating status, etc. The historical indoor and outdoor environmental characteristic data may include current time characteristics, indoor wet-bulb temperature, indoor average temperature, indoor humidity, indoor and outdoor temperature difference, indoor and outdoor wet-bulb temperature difference, indoor temperature median, indoor temperature variance, outdoor temperature of the next time period, feature vector of outdoor weather in the current time period, and feature vector of outdoor weather in the next time period, etc. When constructing the hot water pump quantity prediction model, historical indoor and outdoor environmental characteristic data, historical operating status data of heating station equipment, and target heat volume are used as the original dataset. The original dataset includes multiple sub-features, and the optimal feature subset can be obtained through SBFS (Simplified Method Search). The features in the optimal feature subset are the most informative for the hot water pump quantity prediction model. Specifically, the optimal feature subset may include the median indoor temperature, the target indoor temperature, the supply and return water temperature difference of the hot water pipe, and the first temperature difference. The hot water pump quantity prediction model to be trained can be a LightGBM model, and a hyperparameter search space is specified. Then, random search or grid search is used to determine the optimal hyperparameters to improve the model's performance. The model is trained using the optimal feature subset and historical hot water pump quantities. By fitting the relationship between the input parameters and the actual number of hot water pumps, the hot water pump quantity prediction model can accurately predict the number of hot water pumps in the next time period. The root mean square error (RMSE) and mean absolute error (MAE) are used to evaluate the hot water pump quantity prediction model to verify its accuracy and stability, resulting in the trained hot water pump quantity prediction model.

[0070] Step 302: Input the median indoor temperature, indoor temperature variance, indoor target temperature, hot water supply and return water temperature difference, first temperature difference, and hot water supply temperature as a set of data into the boiler quantity prediction model to obtain the target number of boilers for the next time period output by the boiler quantity prediction model.

[0071] Specifically, the aforementioned boiler quantity prediction model can fit the median indoor temperature, indoor temperature variance, target indoor temperature, hot water supply and return water temperature difference, first temperature difference, hot water supply temperature, and the target boiler quantity for the next time period. By inputting the median indoor temperature, indoor temperature variance, target indoor temperature, hot water supply and return water temperature difference, first temperature difference, and hot water supply temperature as a set of data into the boiler quantity prediction model, the required number of boilers is predicted, yielding the target boiler quantity for the heating station equipment in the next time period. This helps determine if there is sufficient heating capacity to support the target heat demand, while avoiding overuse or underuse of boilers. Boilers are crucial components in heating station equipment, and their quantity directly affects heat production efficiency. Through comprehensive analysis of the median indoor temperature, indoor temperature variance, target indoor temperature, hot water supply and return water temperature difference, first temperature difference, and hot water supply temperature, the most economical and efficient boiler configuration is determined, resulting in the target boiler quantity to meet the expected target heat demand.

[0072] Furthermore, the boiler count prediction model is trained based on historical indoor and outdoor environmental characteristic data, historical operating status data of the heating station equipment, target heat capacity, and the corresponding historical boiler count. The historical boiler count refers to the number of boilers in the operating status data of the next time period from the aforementioned historical operating status data. Specifically, the historical operating status data of the heating station equipment may include the hot water supply and return water temperature difference, hot water supply temperature, hot water return temperature, hot water supply pressure, hot water return pressure, flow rate, boiler frequency, number of boilers, boiler quantity, set temperature, and boiler operating status, etc. The historical indoor and outdoor environmental characteristic data may include current time characteristics, indoor wet-bulb temperature, indoor average temperature, indoor humidity, indoor and outdoor temperature difference, indoor and outdoor wet-bulb temperature difference, indoor temperature median, indoor temperature variance, outdoor temperature of the next time period, feature vector of outdoor weather in the current time period, and feature vector of outdoor weather in the next time period, etc. When constructing the boiler quantity prediction model, historical indoor and outdoor environmental characteristic data, historical operating status data of heating station equipment, and target heat volume are used as the original dataset. The original dataset includes multiple sub-features, and the optimal feature subset can be obtained through SBFS (Simplified Boiler Flow Search). The features in the optimal feature subset are the most informative for the boiler quantity prediction model. Specifically, the optimal feature subset may include the median indoor temperature, indoor temperature variance, indoor target temperature, hot water supply and return water temperature difference, first temperature difference, and hot water supply temperature. The boiler quantity prediction model to be trained can be a LightGBM model, with a specified hyperparameter search space. Then, random search or grid search is used to determine the optimal hyperparameters to improve the model's performance. The model is trained using the optimal feature subset and historical boiler quantities. By fitting the relationship between the input parameters and the actual boiler quantity, the boiler quantity prediction model can accurately predict the boiler quantity in the next time period. The root mean square error (RMSE) and mean absolute error (MAE) are used to evaluate the boiler quantity prediction model to verify its accuracy and stability, resulting in the trained boiler quantity prediction model.

[0073] Step 303: Input the median indoor temperature and the hot water supply temperature into the boiler set temperature prediction model as a set of data to obtain the target set temperature for the next time period output by the boiler set temperature prediction model.

[0074] Specifically, the aforementioned boiler setpoint temperature prediction model can fit the median indoor temperature, the hot water supply temperature, and the boiler setpoint temperature for the next time period. By inputting the median indoor temperature and the hot water supply temperature as a set of data into the boiler setpoint temperature prediction model, the required target setpoint temperature is predicted, resulting in the target boiler setpoint temperature for the heating station equipment in the next time period. This helps optimize boiler operating efficiency and meet indoor temperature control requirements. Through comprehensive analysis of the median indoor temperature and the hot water supply temperature, the most economical and effective boiler temperature configuration is determined, yielding the target boiler setpoint temperature to meet the anticipated target heat demand.

[0075] Furthermore, the boiler setpoint temperature prediction model is trained based on historical indoor and outdoor environmental feature data, historical operating status data of the heating station equipment, target heat load, and corresponding historical boiler setpoint temperatures. The historical boiler setpoint temperature refers to the historical boiler setpoint temperature in the operating status data of the next time period from the aforementioned historical operating status data. When constructing the boiler setpoint temperature prediction model, the historical indoor and outdoor environmental feature data, historical operating status data of the heating station equipment, and target heat load are used as the original dataset. The original dataset includes multiple sub-features, and the optimal feature subset can be obtained through SBFS (Simplified Boiler-Side Filtering). The features in the optimal feature subset are the most informative for the boiler setpoint temperature prediction model. Specifically, the optimal feature subset may include the median indoor temperature and the hot water supply temperature. The boiler setpoint temperature prediction model to be trained can be a LightGBM model, with a specified hyperparameter search space. Then, random search or grid search is used to determine the optimal hyperparameters to improve model performance. The model is trained using the optimal feature subset and historical boiler set temperatures. By fitting the relationship between the input parameters and the actual boiler set temperature, the boiler set temperature prediction model can accurately predict the target set temperature for the next time period. The model is evaluated using indicators such as root mean square error (RMSE) and mean absolute error (MAE) to verify its accuracy and stability, thus obtaining the trained boiler set temperature prediction model.

[0076] Step 304: Generate target operating status data for the temperature control equipment based on the target set temperature, the target number of boilers, and the target number of hot water pumps.

[0077] Specifically, the target set temperature is the target temperature that the boiler should be set to in the next time period, obtained through a boiler set temperature prediction model. The target number of boilers is the number of boilers that need to be started in the next time period, obtained through a boiler number prediction model. The target number of hot water pumps is the number of hot water pumps that need to be started in the next time period, obtained through a hot water pump number prediction model. The target set temperature, the target number of boilers, and the target number of hot water pumps are integrated to generate target operating status data of the temperature control equipment in the next time period. The target operating status data will be used to control the actual operation of the heating station equipment, so that all relevant equipment of the heating station works together in the same time period to generate the target heat. Through refined prediction and control, the operating efficiency of the heating station system is improved, ensuring the accurate meeting of heating demand.

[0078] In some embodiments, before inputting indoor and outdoor environmental characteristic data into the cooling load prediction model, the method further includes:

[0079] Acquire historical indoor and outdoor environmental feature data sets and historical operating status data of chiller equipment, and determine the corresponding historical cooling capacity based on the historical operating status data. The historical indoor and outdoor environmental feature data sets include multiple sub-features.

[0080] The historical indoor and outdoor environmental feature datasets were used as training samples, and the historical cooling volume was used as the label corresponding to the training samples. The optimal feature subset of the cooling volume prediction model was obtained by using the sequential feature selection algorithm.

[0081] The cold volume prediction model to be trained is trained based on the optimal feature subset and historical cold volume, resulting in a trained cold volume prediction model.

[0082] In some embodiments, the historical indoor and outdoor environmental feature dataset is a set of historical data containing multiple sub-features, such as current time features, indoor wet-bulb temperature, indoor average temperature, indoor humidity, indoor-outdoor temperature difference, indoor-outdoor wet-bulb temperature difference, indoor median temperature, indoor temperature variance, outdoor temperature for the next time period, feature vectors of outdoor weather in the current time period, and feature vectors of outdoor weather for the next time period. The historical operating status data of the chiller plant equipment is a set of data recording the operating status of the chiller plant equipment at different points in the past, such as chilled water supply and return temperature difference, chilled water supply temperature, chilled water return temperature, chilled water supply pressure, chilled water return pressure, flow rate, chilled pump frequency, number of cooling towers, number of chillers, set temperature, and chiller operating status. Based on the historical operating status data, the actual cooling capacity provided in different time periods is calculated. The historical indoor and outdoor environmental feature dataset is used as training samples, and the historical cooling capacity is used as the label corresponding to the training samples.

[0083] The optimal feature subset for the cold volume prediction model is obtained using a sequential feature selection algorithm, which includes forward selection and backward elimination. Sequential feature selection algorithms (such as sequential forward selection, sequential backward selection, or bidirectional selection) evaluate the contribution of each feature to the model performance by progressively adding or removing features. During this process, the optimal feature subset can be selected based on some evaluation metric (such as precision, recall, F1 score, or cross-validation score). The sequential feature selection algorithm determines which sub-features are most important for predicting cold volume; these features constitute the optimal feature subset for subsequent model training. Using the optimal feature subset selected by feature selection as input and historical cold volume as the corresponding label, the cold volume prediction model is trained to obtain the trained cold volume prediction model. The cold volume prediction model to be trained can be a linear regression model, decision tree, random forest, gradient boosting tree, or neural network, etc. The model can be evaluated using metrics such as root mean square error and mean absolute error to verify its accuracy and stability.

[0084] Specifically, the sequential feature selection algorithm is a feature selection algorithm that optimizes model performance by progressively removing features. The algorithm starts with all features and removes the features that contribute the least to model performance at each step until the optimal subset of features is found. The specific steps of the sequential feature selection algorithm are described below. Figure 4The process involves several steps: First, initializing the feature set F. Initially, F contains all available features, ensuring that all potentially useful features are considered. Next, calculating the full-feature model performance M. The model is trained using all features, and its performance is calculated to provide a baseline for comparison in subsequent steps. Then, setting an error threshold E defines the allowable error range for model performance, ensuring that model performance does not significantly decrease while reducing features. Finally, initializing the optimal feature subset B. Initially, the optimal feature subset equals all features, providing an initial "best" feature subset for subsequent steps where the number of features will be reduced without sacrificing too much performance. Next, initializing the current feature set C. The current feature set equals all features and can be used as a temporary set for gradually removing features during subsequent iterations. Finally, initializing the current model performance P. The current model performance equals the full-feature model performance. The algorithm first evaluates the model performance of a feature set, providing an initial performance benchmark for subsequent steps. It then checks if the current feature set is empty; if not, it continues to the next step, ensuring the iteration doesn't go indefinitely. The iteration ends when the feature set is empty. For each feature *f*, it removes it and calculates the model performance of the new set, assessing the impact of each feature on model performance. If the new model performance after feature removal is within the error threshold and the new feature set is smaller than the current best set, it updates the best feature subset, ensuring that the set with fewer features is selected as the current best feature subset without significantly decreasing performance. It then removes features with the least impact on model performance from the current feature set, gradually reducing the number of features while maintaining model performance as much as possible. Finally, it returns the best feature subset *B*, which is the subset with the fewest features while maintaining model performance, even when the current feature set is empty. This sequential feature selection algorithm ensures that the number of features is minimized while maintaining model performance, making the model more concise and efficient.

[0085] In some embodiments, reference Figure 5 After step 101, steps 501 to 506 are also included. After obtaining the indoor and outdoor environmental feature data corresponding to the current time, the following steps are also included:

[0086] Step 501: Obtain multiple historical indoor and outdoor environmental characteristic data and the historical load generated by the indoor temperature control device in the next time period corresponding to each historical indoor and outdoor environmental characteristic data. The historical load is historical cooling or historical heating.

[0087] Step 502: Calculate the similarity between each historical indoor and outdoor environmental feature data and the indoor and outdoor environmental feature data respectively, and obtain the corresponding multiple similarity values;

[0088] Step 503: Based on each similarity value, select a preset number of candidate indoor and outdoor environmental feature data from each historical indoor and outdoor environmental feature data;

[0089] Step 504: Calculate the average load among the historical loads corresponding to the preset candidate indoor and outdoor environmental characteristic data, and determine the average load as the second target load;

[0090] Step 505: Based on the second target load, operating status data and indoor and outdoor environmental characteristic data, predict the operating parameters of the temperature control equipment and generate the second target operating status data of the temperature control equipment;

[0091] Step 506: Control the temperature control device based on the second target's operating status data.

[0092] In some embodiments, the operating status data of the temperature control equipment and indoor and outdoor environmental characteristic data corresponding to the current time are acquired to understand the current real-time operating status of the equipment and external environmental conditions, providing a data foundation for subsequent predictions. Operating status data reflects the current working status of the equipment, providing basic information for subsequent predictions. Indoor and outdoor environmental characteristic data reflects the current environmental conditions and is important input information for predicting load demand. Historical indoor and outdoor environmental characteristic data and historical loads (cooling or heating) for the next time period corresponding to the historical data are acquired. Historical loads reflect the actual load demand under past environmental characteristics. Through a large amount of historical data, the potential relationship between environmental characteristics and the load of the next time period can be explored. The similarity between each historical indoor and outdoor environmental characteristic data point and the corresponding indoor and outdoor environmental characteristic data is calculated, obtaining multiple corresponding similarity values. By calculating the similarity between historical and current environmental data, the time period with the most similar historical environmental conditions is identified. Based on the calculated similarity scores, a preset number of historical indoor and outdoor environmental feature data with the highest similarity are selected as candidate indoor and outdoor environmental feature data. The average historical load corresponding to the preset number of candidate indoor and outdoor environmental feature data is calculated to obtain the second target load. By calculating the average value, a more stable load prediction value is obtained, reducing the impact of individual outliers. Based on the second target load, the current equipment operating status data, and indoor and outdoor environmental feature data, the operating parameters of the temperature control equipment are predicted, i.e., the second target operating status data. By referring to multi-source information, the generated operating parameters can meet the predicted load demand (i.e., the second target load). Based on the predicted second target operating status data, the actual operation of the temperature control equipment is controlled, so that the temperature control equipment operates according to the predicted results, meets the load demand, improves the response speed and control accuracy of the temperature control equipment, and solves the problem that the existing temperature control usually uses a single prediction model, which is difficult to accurately capture the complex relationship between indoor and outdoor environmental parameters and cooling and heating load demand, resulting in errors in the prediction of cooling and heating demand. As a result, the temperature control equipment can better adapt to complex and changing environmental conditions, achieve more refined and personalized temperature control, and thus improve the overall performance of the temperature control equipment.

[0093] All of the above-mentioned optional technical solutions can be combined in any way to form optional embodiments of this disclosure, and will not be described in detail here.

[0094] The following are embodiments of the apparatus disclosed herein, which can be used to execute embodiments of the method disclosed herein. For details not disclosed in the apparatus embodiments of this disclosure, please refer to the embodiments of the method disclosed herein.

[0095] Figure 6 This is a schematic diagram of an indoor temperature control device provided in an embodiment of this disclosure. Figure 6 As shown, the indoor temperature control device includes:

[0096] The acquisition module 601 is configured to acquire the operating status data of the temperature control equipment corresponding to the current time, and acquire the indoor and outdoor environmental characteristic data corresponding to the current time. The temperature control equipment is either a cold station equipment or a hot station equipment.

[0097] The prediction module 602 is configured to predict the indoor load demand for the next time period based on indoor and outdoor environmental characteristic data, and obtain the target indoor load for the next time period; wherein the load demand is cooling demand or heating demand.

[0098] The strategy generation module 603 is configured to predict the operating parameters of the temperature control equipment based on the target load, operating status data and indoor and outdoor environmental characteristic data, and generate the target operating status data of the temperature control equipment.

[0099] The control module 604 is configured to control the temperature control device based on the target operating status data.

[0100] According to the technical solution provided in this disclosure, by acquiring the current operating status data of temperature control equipment (such as cooling or heating station equipment) and indoor and outdoor environmental characteristic data, an accurate data foundation is provided for subsequent prediction and control. This allows the temperature control equipment to make predictions and decisions based on the environment and equipment status, improving the real-time performance and accuracy of predictions. By establishing a prediction model that can capture the complex relationship between environmental parameters and load demand, the indoor load demand for the next time period is predicted based on indoor and outdoor environmental characteristic data, obtaining the target indoor load for the next time period. The prediction model accurately predicts future cooling or heating load demands. Based on the target load, current equipment status, and environmental characteristics, multiple factors are combined to predict the optimal operating state of the temperature control equipment for the next time period, generating target operating state data for the temperature control equipment to achieve more refined and personalized temperature control. The temperature control equipment is controlled according to the predicted target operating state data to reach the predetermined operating state, thereby meeting the indoor temperature requirements. The temperature control method disclosed herein utilizes indoor and outdoor environmental characteristic data for load prediction, rather than relying solely on a single model or parameter, thereby improving the accuracy and robustness of prediction. By acquiring real-time operating status data of the temperature control equipment, and dynamically adjusting the control strategy using operating status data, target load, and indoor and outdoor environmental characteristic data, it overcomes the problem in existing technologies where temperature control equipment cannot accurately predict heating and cooling demands, resulting in inaccurate indoor temperature control. By comprehensively considering target load, equipment status, and environmental characteristics, it optimizes equipment operating parameters, minimizing energy consumption while ensuring indoor comfort, significantly improving the performance of temperature control equipment, and achieving more refined and personalized temperature control.

[0101] In some embodiments, when the temperature control device is a chiller plant device, the prediction module 602 is configured to input indoor and outdoor environmental characteristic data into the chiller load prediction model to obtain the target chiller load for the next time period output by the chiller load prediction model; and to determine the target chiller load as the target load.

[0102] In some embodiments, when the temperature control device is a heating station device, the prediction module 602 is configured to input indoor and outdoor environmental characteristic data into the heat prediction model to obtain the target indoor heat for the next time period output by the heat prediction model; and to determine the target heat as the target load.

[0103] In some embodiments, when the temperature control device is a chiller plant device, the indoor and outdoor environmental characteristic data include the indoor and outdoor wet-bulb temperature difference and the indoor average temperature, and the operating status data includes the chilled water supply temperature and the chilled water return temperature. The strategy generation module 603 is configured to input the indoor and outdoor wet-bulb temperature difference and the target cooling capacity as a set of data into the cooling tower number prediction model to obtain the target number of cooling towers for the next time period output by the cooling tower number prediction model; input the target cooling capacity into the chiller number prediction model to obtain the target number of chillers for the next time period output by the chiller number prediction model; and input the indoor average temperature and the chilled water supply temperature as a set of data into the chiller set temperature prediction model to obtain the chiller set temperature. The model predicts the target set temperature for the next time period; it acquires multiple candidate frequencies of the chilled water pump and the candidate flow rates corresponding to each candidate frequency, and generates candidate heat corresponding to each candidate frequency based on the candidate flow rates, target set temperature, and chilled water return temperature; it calculates the absolute value of the difference between the candidate heat corresponding to each candidate frequency and the target cooling capacity, and determines the candidate frequency that meets the preset requirements as the target frequency for the next time period. The preset requirements are that the frequency is less than or equal to the first preset value and the absolute value of the difference with the target cooling capacity is the smallest; it generates target operating status data of the temperature control equipment based on the target frequency, target set temperature, target number of chillers, and target number of cooling towers.

[0104] In some embodiments, when the temperature control device is a heating station device, the indoor and outdoor environmental characteristic data includes the median indoor temperature and the indoor temperature variance, and the operating status data includes the hot water supply and return water temperature difference and the hot water supply temperature. The strategy generation module 603 is configured to acquire the indoor target temperature, the first temperature difference between the indoor target temperature and the indoor median temperature, and input the indoor median temperature, the indoor target temperature, the hot water supply and return water temperature difference, and the first temperature difference as a set of data into the hot water pump quantity prediction model to obtain the target number of hot water pumps for the next time period output by the hot water pump quantity prediction model; input the indoor median temperature, the indoor temperature variance, the indoor target temperature, the hot water supply and return water temperature difference, the first temperature difference, and the hot water supply temperature as a set of data into the boiler quantity prediction model to obtain the target number of boilers for the next time period output by the boiler quantity prediction model; input the indoor median temperature and the hot water supply temperature as a set of data into the boiler set temperature prediction model to obtain the target set temperature for the next time period output by the boiler set temperature prediction model; and generate target operating status data of the temperature control device based on the target set temperature, the target number of boilers, and the target number of hot water pumps.

[0105] In some embodiments, before inputting indoor and outdoor environmental feature data into the cooling capacity prediction model, the indoor temperature control device is configured to acquire a historical set of indoor and outdoor environmental feature data and historical operating status data of the chiller equipment, and determine the corresponding historical cooling capacity based on the historical operating status data. The historical set of indoor and outdoor environmental feature data includes multiple sub-features. The historical set of indoor and outdoor environmental feature data is used as training samples, and the historical cooling capacity is used as the label corresponding to the training samples. An optimal feature subset of the cooling capacity prediction model is obtained by using a sequential feature selection algorithm. The cooling capacity prediction model to be trained is trained based on the optimal feature subset and the historical cooling capacity to obtain the trained cooling capacity prediction model.

[0106] In some embodiments, after acquiring the indoor and outdoor environmental feature data corresponding to the current time, the indoor temperature control device is configured to acquire multiple historical indoor and outdoor environmental feature data and the historical load generated by the indoor temperature control device in the next time period corresponding to each historical indoor and outdoor environmental feature data, wherein the historical load is historical cooling or historical heating; calculate the similarity between each historical indoor and outdoor environmental feature data and the indoor and outdoor environmental feature data respectively, and obtain multiple corresponding similarity values; based on each similarity value, select a preset number of candidate indoor and outdoor environmental feature data from each historical indoor and outdoor environmental feature data; calculate the average load among the historical loads corresponding to the preset number of candidate indoor and outdoor environmental feature data, and determine the average load as the second target load; predict the operating parameters of the temperature control device according to the second target load, operating status data, and indoor and outdoor environmental feature data, and generate the second target operating status data of the temperature control device; and control the temperature control device according to the second target operating status data.

[0107] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this disclosure.

[0108] Figure 7 This is a schematic diagram of the electronic device 7 provided in an embodiment of this disclosure. Figure 7 As shown, the electronic device 7 of this embodiment includes a processor 701, a memory 702, and a computer program 703 stored in the memory 702 and executable on the processor 701. When the processor 701 executes the computer program 703, it implements the steps in the various method embodiments described above. Alternatively, when the processor 701 executes the computer program 703, it implements the functions of each module / unit in the various device embodiments described above.

[0109] Electronic device 7 can be a desktop computer, laptop, handheld computer, cloud server, or other electronic device. Electronic device 7 may include, but is not limited to, processor 701 and memory 702. Those skilled in the art will understand that...Figure 7 This is merely an example of electronic device 7 and does not constitute a limitation on electronic device 7. It may include more or fewer components than shown, or different components.

[0110] The processor 701 can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0111] The memory 702 can be an internal storage unit of the electronic device 7, such as a hard disk or RAM of the electronic device 7. The memory 702 can also be an external storage device of the electronic device 7, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, etc., equipped on the electronic device 7. The memory 702 can also include both internal and external storage units of the electronic device 7. The memory 702 is used to store computer programs and other programs and data required by the electronic device.

[0112] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0113] If integrated modules / units are implemented as software functional units and sold or used as independent products, they can be stored in a readable storage medium (e.g., a computer-readable storage medium). Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program may include computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. A computer-readable storage medium may include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0114] The above embodiments are only used to illustrate the technical solutions of this disclosure, and are not intended to limit it. Although this disclosure has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this disclosure, and should all be included within the protection scope of this disclosure.

Claims

1. An indoor temperature control method, characterized in that, include: Based on the current time, obtain the operating status data of the temperature control equipment corresponding to the current time, and obtain the indoor and outdoor environmental characteristic data corresponding to the current time. The temperature control equipment is a cold station equipment or a hot station equipment. Based on the indoor and outdoor environmental characteristic data, the indoor load demand for the next time period is predicted to obtain the target indoor load for the next time period, wherein the load demand is cooling demand or heating demand. When the temperature control device is the chiller plant equipment, the step of predicting the indoor load demand for the next time period based on the indoor and outdoor environmental characteristic data to obtain the target indoor load for the next time period includes: The indoor and outdoor environmental characteristic data are input into the cooling capacity prediction model to obtain the target indoor cooling capacity for the next time period output by the cooling capacity prediction model. The target cooling capacity is defined as the target load; The indoor and outdoor environmental characteristic data includes the indoor and outdoor wet-bulb temperature difference and the indoor average temperature; the operating status data includes the chilled water supply temperature and the chilled water return temperature; based on the target load, the operating status data, and the indoor and outdoor environmental characteristic data, the operating parameters of the temperature control equipment are predicted to generate target operating status data for the temperature control equipment, including: The indoor and outdoor wet-bulb temperature difference and the target cooling capacity are input as a set of data into the cooling tower number prediction model to obtain the target number of cooling towers for the next time period output by the cooling tower number prediction model. The target cooling capacity is input into the chiller quantity prediction model to obtain the target chiller quantity for the next time period output by the chiller quantity prediction model. The indoor average temperature and the chilled water supply temperature are input as a set of data into the chiller set temperature prediction model to obtain the target set temperature for the next time period output by the chiller set temperature prediction model. Multiple candidate frequencies of the chilled water pump and the candidate flow rate corresponding to each candidate frequency are obtained, and candidate heat corresponding to each candidate frequency is generated based on the candidate flow rate corresponding to each candidate frequency, the target set temperature and the chilled water return temperature. Calculate the absolute value of the difference between the candidate heat and the target cooling amount corresponding to each candidate frequency, and determine the candidate frequency that meets the preset requirement as the target frequency of the next time period. The preset requirement is that it is less than or equal to a first preset value and the absolute value of the difference between it and the target cooling amount is the smallest. The target operating status data of the temperature control equipment is generated based on the target frequency, the target set temperature, the target number of chillers, and the target number of cooling towers; When the temperature control device is the heating station equipment, the step of predicting the indoor load demand for the next time period based on the indoor and outdoor environmental characteristic data to obtain the target indoor load for the next time period includes: The indoor and outdoor environmental characteristic data are input into the heat prediction model to obtain the target indoor heat for the next time period output by the heat prediction model. The target heat is defined as the target load; The indoor and outdoor environmental characteristic data includes the median indoor temperature and the indoor temperature variance; the operating status data includes the hot water supply and return water temperature difference and the hot water supply temperature; based on the target load, the operating status data, and the indoor and outdoor environmental characteristic data, the operating parameters of the temperature control equipment are predicted to generate target operating status data for the temperature control equipment, including: Obtain the indoor target temperature, the first temperature difference between the indoor target temperature and the median indoor temperature, and input the median indoor temperature, the indoor target temperature, the hot water supply and return temperature difference and the first temperature difference as a set of data into the hot water pump number prediction model to obtain the target number of hot water pumps for the next time period output by the hot water pump number prediction model. The median indoor temperature, the variance of indoor temperature, the target indoor temperature, the temperature difference between the supply and return water of the hot water pipe, the first temperature difference, and the supply water temperature of the hot water pipe are input as a set of data into the boiler quantity prediction model to obtain the target boiler quantity for the next time period output by the boiler quantity prediction model. The median indoor temperature and the hot water supply temperature are input as a set of data into the boiler set temperature prediction model to obtain the target set temperature for the next time period output by the boiler set temperature prediction model. The target operating status data of the temperature control equipment is generated based on the target set temperature, the target number of boilers, and the target number of hot water pumps; The temperature control device is controlled based on the target operating status data.

2. The method according to claim 1, characterized in that, Before inputting the indoor and outdoor environmental characteristic data into the cooling capacity prediction model, the method further includes: Acquire historical indoor and outdoor environmental feature data sets and historical operating status data of the chiller equipment, and determine the corresponding historical cooling capacity based on the historical operating status data. The historical indoor and outdoor environmental feature data sets include multiple sub-features. The historical indoor and outdoor environmental feature datasets were used as training samples, and the historical cooling volume was used as the label corresponding to the training samples. The sequential feature selection algorithm was used to obtain the optimal feature subset of the cooling volume prediction model. The cold volume prediction model to be trained is trained based on the optimal feature subset and the historical cold volume, and the trained cold volume prediction model is obtained.

3. The method according to claim 1, characterized in that, After obtaining the indoor and outdoor environmental feature data corresponding to the current time, the method further includes: Acquire multiple historical indoor and outdoor environmental characteristic data and the historical load generated by the indoor temperature control device in the next time period corresponding to each historical indoor and outdoor environmental characteristic data, wherein the historical load is historical cooling or historical heating. Calculate the similarity between each of the historical indoor and outdoor environmental feature data and the indoor and outdoor environmental feature data respectively to obtain multiple corresponding similarity values; Based on the similarity values, a preset number of candidate indoor and outdoor environmental feature data are selected from the historical indoor and outdoor environmental feature data. Calculate the mean load among the historical loads corresponding to the preset candidate indoor and outdoor environmental characteristic data, and determine the mean load as the second target load; Based on the second target load, the operating status data, and the indoor and outdoor environmental characteristic data, the operating parameters of the temperature control device are predicted to generate the second target operating status data of the temperature control device; The temperature control device is controlled based on the second target operating status data.

4. An indoor temperature control device for executing the indoor temperature control method according to any one of claims 1-3, characterized in that, include: The acquisition module is configured to acquire the operating status data of the temperature control equipment corresponding to the current time, and acquire the indoor and outdoor environmental characteristic data corresponding to the current time, wherein the temperature control equipment is a cold station equipment or a hot station equipment; The prediction module is configured to predict the indoor load demand for the next time period based on the indoor and outdoor environmental characteristic data, and obtain the target indoor load for the next time period. The load demand mentioned above refers to either cooling demand or heating demand. The strategy generation module is configured to predict the operating parameters of the temperature control device based on the target load, the operating status data, and the indoor and outdoor environmental characteristic data, and generate the target operating status data of the temperature control device. The control module is configured to control the temperature control device based on the target operating status data.

5. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 3.

6. A readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 3.

Images