A dynamic energy-saving central air conditioning control method

Abstract

The application relates to a dynamic energy-saving central air conditioner control method, which comprises the following steps: constructing and training a temperature preference prediction model based on historical indoor and outdoor temperature data; for a current target period, an AI intelligent assistant module acquires a predicted comfortable temperature interval of an indoor area; according to the predicted comfortable temperature interval, a refrigeration system is controlled to regulate the temperature of the indoor area to a primary set temperature; the AI intelligent assistant module pushes an energy-saving optimization scheme, and if the energy-saving optimization scheme is passed, the temperature is regulated to an optimized temperature according to the energy-saving optimization scheme; and the primary set temperature and the optimized temperature are both located in the predicted comfortable temperature interval. Through the temperature preference prediction model trained based on historical data, the system can predict and set a temperature interval conforming to the comfort of most people according to the habits of users in different areas and at different periods, realize on-demand cooling, and improve overall comfort and user satisfaction.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology, and in particular to a dynamic energy-saving central air conditioning control method. Background Technology

[0002] With the continuous increase in building energy consumption, the energy-saving operation of central air conditioning systems, as the main energy-consuming equipment in large public and commercial buildings, has become an important research direction. Traditional central air conditioning systems typically employ fixed temperature settings or control strategies based on simple rules, making it difficult to balance the personalized comfort needs of different users with the overall energy efficiency of the system. Furthermore, most existing systems fail to fully utilize time-of-use pricing policies and energy storage technologies, and cannot perform real-time optimization and control based on dynamic changes in the indoor and outdoor environment and the activity status of personnel, resulting in high energy consumption, high operating costs, and often unsatisfactory user experiences.

[0003] In recent years, the application of artificial intelligence and big data analytics in smart buildings has deepened, enabling personalized and predictive control of air conditioning systems. Simultaneously, the combined operation of ice storage, water storage, and other energy-saving refrigeration technologies with electric refrigeration systems offers the potential for energy conservation and economic benefits through storing cooling during off-peak hours and releasing it during peak hours. However, how to organically combine user temperature preference prediction, multi-regional independent control, time-of-use pricing strategies, and energy storage system scheduling to form an intelligent control scheme that ensures both comfort and dynamic energy saving remains a key technical challenge.

[0004] To address this, the present invention proposes a dynamic energy-saving central air conditioning control method. By training a temperature preference model using historical data, it can predict the comfortable temperature range for users in different areas. Combined with the intelligent switching between reserve cooling and electric cooling systems, as well as energy-saving optimization strategies based on user feedback, it can significantly reduce system operating energy consumption and electricity costs while meeting personalized comfort needs. Summary of the Invention

[0005] (a) Technical problems to be solved The technical problem to be solved by this invention is to address the issue of excessive energy consumption in existing central air conditioning systems and to further optimize the energy consumption of central air conditioning systems.

[0006] (II) Technical Solution To address the above problems, this invention provides a dynamic energy-saving central air conditioning control method, comprising: S100: A temperature preference prediction model is built and trained based on historical internal and external temperature data; S200: For the current target time period, the AI ​​intelligent assistant module inputs the corresponding real-time outdoor temperature into the temperature preference prediction model to obtain the predicted comfortable temperature range for the indoor area; S300: The AI ​​intelligent assistant module controls the cooling system to adjust the temperature of the indoor area to the initial set temperature based on the predicted comfortable temperature range; S400: The AI ​​intelligent assistant module pushes an energy-saving optimization plan. If the energy-saving optimization plan is approved, the temperature is adjusted to the optimized temperature according to the energy-saving optimization plan. Both the initial temperature and the optimized temperature are within the predicted comfortable temperature range.

[0007] The refrigeration system includes a reserve refrigeration system and an electric refrigeration system; The controlled refrigeration system adjusts the temperature of the indoor area to the initial set temperature, including: Start the reserve cooling system to exchange heat with the low-temperature water in the water tank, and deliver cold air to the indoor area according to the initial set temperature until the water temperature in the water tank rises to the high temperature warning value, then stop the reserve cooling system. Start the electric cooling system to deliver cool air to the indoor area according to the initial set temperature.

[0008] The cooling system, which controls the temperature of the indoor area to a preset temperature, also includes: The AI ​​intelligent assistant module obtains the current time in real time. If the current time is during the set off-peak electricity price period T1: Then the reserve refrigeration system is stopped, and the electric refrigeration system is started to deliver cold air to the indoor area according to the initial set temperature; Furthermore, the water storage tank is cooled and stored using an electric refrigeration system until the water temperature in the storage tank drops to a preset temperature threshold.

[0009] S400 includes: When the current outdoor ambient temperature is 30℃ or higher; The AI ​​intelligent assistant module pushes the upper limit of the predicted comfortable temperature range to the user via message; obtains user feedback, and if most agree, the energy-saving optimization plan is approved; the AI ​​intelligent assistant module adjusts the temperature to the upper limit of the predicted comfortable temperature range; if most disagree, the energy-saving optimization plan is canceled; the AI ​​intelligent assistant module raises the temperature of the current indoor area by 0.5℃. When the current outdoor temperature is below 30℃; The AI ​​intelligent assistant module pushes the upper limit of the predicted comfortable temperature range to the user via message; obtains user feedback, and if most agree, the energy-saving optimization plan is approved; the AI ​​intelligent assistant module adjusts the temperature to the upper limit of the predicted comfortable temperature range; if most disagree, the energy-saving optimization plan is canceled.

[0010] This also includes S500: For multiple independent indoor areas within a building, the AI ​​intelligent assistant module executes S200 to S400 in parallel, and generates a personalized predicted comfort temperature range for each area.

[0011] Step S100 includes: Based on historical internal and external temperature data and corresponding user historical adjustment records, a temperature preference prediction model is constructed and trained. The user history adjustment records include the user's actual adjustment data for indoor temperature.

[0012] The determination of the predicted comfort temperature range includes: Obtain information on the current indoor area's population density and activity intensity. Based on the population density information, activity intensity information, and real-time outdoor temperature, the output of the temperature preference prediction model is corrected to obtain the predicted comfortable temperature range.

[0013] In step S400, the energy-saving optimization scheme further includes: adjusting the indoor air supply volume.

[0014] Wherein, obtaining user-side feedback includes, if, in majority opinion, the following: When the message is pushed to the user's mobile terminal application, the application interface provides the option to agree to the optimization or keep the current state. If, within the preset feedback time window, the proportion of users who agree to the optimization exceeds the set threshold, it is determined that the majority agrees, and the energy-saving optimization scheme is approved.

[0015] (III) Beneficial Effects The above-described technical solution of the present invention has the following advantages: By using a temperature preference prediction model trained on historical data, the system can predict and set temperature ranges that meet the comfort level of most people based on user habits in different regions and at different times, thereby achieving on-demand cooling and improving overall comfort and user satisfaction.

[0016] By combining a storage-cooling system with an electric cooling system, and activating electric cooling and storing cold energy during off-peak electricity hours while prioritizing the use of stored cold energy during peak electricity hours, electricity costs are significantly reduced. Furthermore, by using AI to push energy-saving optimization solutions and guide user participation in regulation, energy efficiency can be further improved without compromising comfort. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the execution flow in an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] See Figure 1 As shown in the figure, an embodiment of the present invention provides a dynamic energy-saving central air conditioning control method, characterized in that it includes: S100: Construct and train a temperature preference prediction model based on historical indoor and outdoor temperature data; preferably, construct and train a temperature preference prediction model based on historical indoor and outdoor temperature data and corresponding user historical adjustment records; the user historical adjustment records include the user's actual adjustment data of indoor temperature.

[0020] Specifically, the steps include the following: Data preprocessing: Historical indoor and outdoor temperature data and user historical adjustment records are cleaned, aligned, and normalized. User historical adjustment records can be specified as the target temperature value that the user adjusted the set temperature to under specific outdoor and indoor temperature conditions. The above data serves as labels for model learning.

[0021] Model Selection and Construction: The temperature preference prediction model can be a machine learning model such as Long Short-Term Memory Network or Gradient Boosting Decision Tree. The model's input feature vector is designed as: [time information, outdoor temperature, indoor temperature]. The model's output is a continuous predicted temperature value, representing the user's preferred temperature predicted under these input features.

[0022] Model training: The model is divided into training and validation sets proportionally. Using the training set data, the mean squared error between the user's actual target temperature and the model's predicted temperature is used as the loss function. The model parameters are iteratively updated using optimization methods such as backpropagation until the model's mean absolute error of prediction on the validation set stabilizes below 0.5℃, at which point model training is complete.

[0023] S200: For the current target time period, the AI ​​intelligent assistant module inputs the corresponding real-time outdoor temperature into the temperature preference prediction model to obtain the predicted comfortable temperature range for the indoor area; The determination of the predicted comfortable temperature range includes at least: obtaining the current indoor area's personnel density information and activity intensity information; and, based on the personnel density information, activity intensity information, and the real-time outdoor temperature, correcting the output of the temperature preference prediction model to obtain the predicted comfortable temperature range.

[0024] Basic Prediction: The AI ​​intelligent assistant module obtains the real-time outdoor temperature and current time corresponding to the current target time period, inputs them into the trained temperature preference prediction model, and obtains the basic predicted comfort temperature T for that time period. base .

[0025] Dynamic Environment Correction: The AI ​​intelligent assistant module obtains the current indoor area's population density (people / square meter) and activity intensity through the environmental perception unit. It then calculates based on preset correction rules: Personnel density correction: The preset standard personnel density is D0, the current personnel density is D, and the correction factor is α. Corrected temperature T adj1 =T base- α×(D-D0); Activity Intensity Correction: Let the current activity intensity level be L, and the corresponding temperature compensation value be ΔT. L Corrected temperature T adj2 =T adj1 +ΔT L ΔT L It is usually a negative number.

[0026] Generate predicted comfort temperature range: based on the final corrected temperature T adj2 Centered on a point, a floating value δ is added upwards and downwards; δ can be selected as 1.5℃ to generate the predicted comfort temperature range [T]. adj2 −δ,T adj2 +δ].

[0027] Activity intensity level L and corresponding temperature compensation value ΔT L The correspondence can be shown in the following table. It can also be set according to the actual situation; no restrictions are imposed here.

[0028] Activity Intensity Level (L) describe Metabolic rate (approximately, met) Temperature correction value (ΔTL) 1 Sit quietly and rest 1.0 - 1.2 met 0.0℃ (Baseline) 2 Light activities, such as walking or working. 1.2 - 2.0 met -0.5℃ to -1.0℃ 3 Moderate activities, such as talking and brisk walking 2.0 - 3.0 met -1.0℃ to -2.0℃ 4 Higher activities, such as gymnastics and carrying >3.0 met -2.0℃ Take a 30-square-meter meeting room as an example: D: An anonymous visual counter on the ceiling detects that there are currently 6 people in the room, so the current population density D = 6 / 30 = 0.2 people / square meter.

[0029] Determine D0: The design standard for this conference room is 5 square meters per person, so the standard density D0 = 1 / 5 = 0.2 people / square meter.

[0030] Get ΔT L Analysis using a microphone array and vibration sensors indicates an ambient noise level of 50 dB accompanied by regular low-frequency vibrations. The system determines the activity intensity level to be L=2 (mild activity). Referring to the reference table, the corresponding ΔT... L =-0.5℃ to -1.0℃. In this case, −0.8℃ was selected.

[0031] Substitution and correction: Assume the base predicted temperature T base =24℃, personnel density correction factor α=0.5. Then: Personnel density correction: T adj1 =24 - 0.5 × (0.2 - 0.2) = 24℃. The current density is equal to the standard density, with no correction.

[0032] Activity intensity correction: T adj2 =24+(-0.8)=23.2℃.

[0033] Through the above methods, dynamic and accurate correction of the predicted comfortable temperature range is achieved. Those skilled in the art should understand that the above acquisition methods are merely examples, and other sensing and calculation methods can be used to achieve the same purpose without departing from the core principles of this invention.

[0034] S300: The AI ​​intelligent assistant module controls the cooling system to adjust the temperature of the indoor area to the initial set temperature based on the predicted comfortable temperature range.

[0035] Specifically, The initial temperature is preferably the lower limit of the predicted comfortable temperature range, so as to leave room for subsequent energy-saving optimization while ensuring comfort.

[0036] The refrigeration system includes a reserve refrigeration system and an electric refrigeration system; the reserve refrigeration system has a water tank that stores chilled water. The water tank is preferably placed in a basement to reduce cold loss and facilitate low-temperature storage.

[0037] The control logic is as follows: a) The AI ​​intelligent assistant module obtains the current time in real time and compares it with the preset off-peak electricity price period T1. Generally, the off-peak electricity price period is from 23:00 to 7:00 the next day. The specific time can be set according to different regional policies.

[0038] b) If the current time is during a low electricity price period T1, then execute strategy one: Stop using the reserve refrigeration system for cooling, start the electric refrigeration system, and deliver cool air to the indoor areas according to the initial set temperature.

[0039] At the same time, the electric cooling system, while meeting the indoor cooling demand, uses a portion of its cooling capacity to cool and store the water in the water tank until the water temperature in the tank drops to a preset threshold, which can be set to 4℃.

[0040] c) If the current time is not during a low electricity price period (peak or off-peak hours), then execute strategy two: The reserve cooling system is activated first, allowing the low-temperature water in the storage tank to exchange heat with the air conditioning circulating water, and cool air is delivered to the indoor area according to the initial set temperature.

[0041] The water temperature in the storage tank is monitored in real time. When the water temperature rises to the preset high temperature warning value, such as 12°C, it indicates that the stored cold energy is about to be exhausted.

[0042] Stop the reserve refrigeration system and start the electric refrigeration system to deliver cool air to the indoor area according to the initial set temperature.

[0043] S400: The AI ​​intelligent assistant module pushes an energy-saving optimization plan. If the energy-saving optimization plan is approved, the temperature is adjusted to the optimized temperature according to the energy-saving optimization plan. Both the initial temperature and the optimized temperature are within the predicted comfortable temperature range.

[0044] The energy-saving optimization scheme also includes: adjusting the indoor air supply volume.

[0045] When the current outdoor ambient temperature is 30℃ or higher; The AI ​​intelligent assistant module pushes the upper limit of the predicted comfortable temperature range to the user via message; obtains user feedback, and if most agree, the energy-saving optimization plan is approved; the AI ​​intelligent assistant module adjusts the temperature to the upper limit of the predicted comfortable temperature range; if most disagree, the energy-saving optimization plan is canceled; the AI ​​intelligent assistant module raises the temperature of the current indoor area by 0.5℃. When the current outdoor temperature is below 30℃; The AI ​​intelligent assistant module pushes the upper limit of the predicted comfortable temperature range to the user via message; obtains user feedback, and if most agree, the energy-saving optimization plan is approved; the AI ​​intelligent assistant module adjusts the temperature to the upper limit of the predicted comfortable temperature range; if most disagree, the energy-saving optimization plan is canceled.

[0046] The process of obtaining user feedback, if the majority of consent includes: When the message is pushed to the user's mobile terminal application, the application interface provides the option to agree to the optimization or keep the current state. If, within the preset feedback time window, the proportion of users who agree to the optimization exceeds the set threshold, it is determined that the majority agrees, and the energy-saving optimization scheme is approved.

[0047] Specifically: This step aims to find a more energy-efficient setpoint within the comfort zone by interacting with the user.

[0048] Solution Generation and Push: The AI ​​intelligent assistant module generates an energy-saving optimization solution, the core of which is to suggest adjusting the indoor temperature from the initial setting to the upper limit of the predicted comfortable temperature range, i.e., optimizing the temperature, and may also appropriately increase the air supply volume to maintain human comfort. This solution is pushed to the mobile terminal application of users in the current indoor area through the user interaction unit.

[0049] User feedback collection: The application interface provides options to agree to optimization or maintain the current state. The system sets a feedback time window, such as 15 minutes.

[0050] Decision-making and execution: If, within the feedback time window, the percentage of users who agree to the optimization exceeds the set percentage threshold, such as 50%, it is determined that the majority agrees and the energy-saving optimization plan is approved.

[0051] Once the solution is approved, the AI ​​intelligent assistant module controls the cooling system to adjust the temperature to the optimized temperature and can also adjust the fan speed.

[0052] Handling of proposals that are not approved: If the approval rate does not exceed the threshold, the proposal will not be approved and the optimization proposal will be cancelled.

[0053] Furthermore, to encourage energy-saving behavior, incentive strategies can be set up: if the current outdoor ambient temperature is higher than or equal to 30℃, after the plan is canceled, the AI ​​intelligent assistant module can raise the temperature of the current indoor area by 0.5℃ within the comfort range, and will not raise it further if the upper limit of the range has been reached; if the outdoor ambient temperature is lower than 30℃, the current temperature will remain unchanged.

[0054] S500: For multiple independent indoor areas within a building, the AI ​​intelligent assistant module executes S200 to S400 in parallel and generates a personalized predicted comfort temperature range for each area.

[0055] For multiple areas within the building, such as different rooms and different floors, the AI ​​intelligent assistant module independently executes S200 to S400 for each area. Each area generates a personalized predicted comfort temperature range based on its real-time outdoor temperature, personnel density, and activity intensity, and independently performs initial temperature control and energy-saving optimization interaction, thereby achieving refined zoning energy-saving management.

[0056] To enable those skilled in the art to better understand, the following non-limiting embodiments are provided: Assuming it is 2:00 PM on a summer weekday (excluding off-peak electricity hours), the system detects that the outdoor temperature in a conference room is 35°C, the initial indoor temperature is 28°C, the personnel density is 0.2 people / square meter, and the activity intensity is "sedentary".

[0057] S200: The model predicts the baseline comfort temperature T based on the input. base =24℃. Assuming T is corrected for personnel density. adj1 =23.5℃ and activity intensity correction T adj2 =23.5℃, take δ=0.5℃, and the predicted comfortable temperature range for the conference room is [23℃, 24℃].

[0058] S300: The AI ​​intelligent assistant module selects 23℃ as the lower limit of the temperature range as the initial setting. Because it is not a period of low electricity prices, the system prioritizes the activation of the reserve cooling system, using chilled water from the water tank to cool the conference room until the room temperature reaches 23℃.

[0059] S400: The AI ​​intelligent assistant module pushes a message to the users' mobile apps in the conference room: "To save energy, it is recommended to raise the temperature to 24℃ (the upper limit of comfort) and automatically increase the airflow. Do you agree?" Within 15 minutes, 7 out of 10 online users selected "agree" (70% > 50%), and the proposal was approved. The AI ​​intelligent assistant module then controls the air conditioning terminal in the conference room, adjusting the temperature setpoint to 24℃ and increasing the fan speed from "medium" to "high".

[0060] S500: At the same time, the system independently performs the above process for other areas within the building, generating and executing their respective personalized temperature strategies.

[0061] By incorporating multi-dimensional information such as real-time outdoor temperature and personnel status, the prediction and control strategies are dynamically adjusted, enabling the system to have strong environmental adaptability and learning and evolution capabilities, and realizing the transformation from "passive control" to "active optimization".

[0062] By pushing optimization suggestions to users and obtaining feedback, we not only enhance users' awareness and participation in energy conservation, but also improve the acceptance of temperature settings, thus forming a "human-machine collaboration" energy-saving operation mechanism.

[0063] By rationally scheduling the two refrigeration systems, the electric refrigeration system can be prevented from operating at high load for extended periods, which facilitates equipment rotation and maintenance and improves the overall reliability of the system.

[0064] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that not every embodiment contains only one independent technical solution, and in the absence of conflict between solutions, the various technical features mentioned in each embodiment can be combined in any way to form other implementation methods that can be understood by those skilled in the art.

[0065] Furthermore, without departing from the scope of the present invention, modifications to the technical solutions described in the foregoing embodiments, or equivalent substitutions of some of the technical features, shall not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A dynamic energy-saving central air conditioning control method, characterized in that, include: S100: A temperature preference prediction model is built and trained based on historical internal and external temperature data; S200: For the current target time period, the AI ​​intelligent assistant module inputs the corresponding real-time outdoor temperature into the temperature preference prediction model to obtain the predicted comfortable temperature range for the indoor area; S300: The AI ​​intelligent assistant module controls the cooling system to adjust the temperature of the indoor area to the initial set temperature based on the predicted comfortable temperature range; S400: The AI ​​intelligent assistant module pushes an energy-saving optimization plan. If the energy-saving optimization plan is approved, the temperature is adjusted to the optimized temperature according to the energy-saving optimization plan. Both the initial temperature and the optimized temperature are within the predicted comfortable temperature range.

2. The dynamic energy-saving central air conditioning control method according to claim 1, characterized in that, The refrigeration system includes a reserve refrigeration system and an electric refrigeration system; The controlled refrigeration system adjusts the temperature of the indoor area to the initial set temperature, including: Start the reserve cooling system to exchange heat with the low-temperature water in the water tank, and deliver cold air to the indoor area according to the initial set temperature until the water temperature in the water tank rises to the high temperature warning value, then stop the reserve cooling system. Start the electric cooling system to deliver cool air to the indoor area according to the initial set temperature.

3. The dynamic energy-saving central air conditioning control method according to claim 2, characterized in that, The controlled refrigeration system adjusts the temperature of the indoor area to the initial set temperature, and also includes: The AI ​​intelligent assistant module obtains the current time in real time. If the current time is during the set off-peak electricity price period T1: Then the reserve refrigeration system is stopped, and the electric refrigeration system is started to deliver cold air to the indoor area according to the initial set temperature; Furthermore, the water storage tank is cooled and stored using an electric refrigeration system until the water temperature in the storage tank drops to a preset temperature threshold.

4. The dynamic energy-saving central air conditioning control method according to claim 3, characterized in that, The S400 includes: When the current outdoor ambient temperature is 30℃ or higher; The AI ​​intelligent assistant module pushes the upper limit of the predicted comfortable temperature range to the user via message; obtains user feedback, and if most agree, the energy-saving optimization plan is approved; the AI ​​intelligent assistant module adjusts the temperature to the upper limit of the predicted comfortable temperature range; if most disagree, the energy-saving optimization plan is canceled; the AI ​​intelligent assistant module raises the temperature of the current indoor area by 0.5℃. When the current outdoor temperature is below 30℃; The AI ​​intelligent assistant module pushes the upper limit of the predicted comfortable temperature range to the user via message; obtains user feedback, and if most agree, the energy-saving optimization plan is approved; the AI ​​intelligent assistant module adjusts the temperature to the upper limit of the predicted comfortable temperature range; if most disagree, the energy-saving optimization plan is canceled.

5. The dynamic energy-saving central air conditioning control method according to claim 4, characterized in that, It also includes S500: For multiple independent indoor areas within a building, the AI ​​intelligent assistant module executes S200 to S400 in parallel and generates a personalized predicted comfort temperature range for each area.

6. The dynamic energy-saving central air conditioning control method according to claim 5, characterized in that, Step S100 includes: Based on historical internal and external temperature data and corresponding user historical adjustment records, a temperature preference prediction model is constructed and trained. The user history adjustment records include the user's actual adjustment data for indoor temperature.

7. The dynamic energy-saving central air conditioning control method according to claim 6, characterized in that, The determination of the predicted comfort temperature range includes: Obtain information on the current indoor area's population density and activity intensity. Based on the population density information, activity intensity information, and real-time outdoor temperature, the output of the temperature preference prediction model is corrected to obtain the predicted comfortable temperature range.

8. The dynamic energy-saving central air conditioning control method according to claim 7, characterized in that, In step S400, the energy-saving optimization scheme further includes: adjusting the indoor air supply volume.

9. A dynamic energy-saving central air conditioning control method according to claim 7, characterized in that, The process of obtaining user feedback, if the majority of consent includes: When the message is pushed to the user's mobile terminal application, the application interface provides the option to agree to the optimization or keep the current state. If, within the preset feedback time window, the proportion of users who agree to the optimization exceeds the set threshold, it is determined that the majority agrees, and the energy-saving optimization scheme is approved.

Images