Energy-saving self-adaptive control method and system for heating ventilation air conditioner

Abstract

The invention relates to the technical field of heating ventilation air conditioner control and energy conservation, and discloses an energy-saving self-adaptive control method and system for a heating ventilation air conditioner. The method comprises the following steps: acquiring power consumption data and environment data in a building to obtain initial activity state distribution; air conditioner related energy consumption characteristics are separated according to the activity state distribution; identifying a specific activity state in combination with a user behavior database; determining a target cooling capacity output range according to the activity type and the personnel density; if the range exceeds the current setting, recalculating the operation parameters in combination with the external climate data, and generating an adjustment instruction; dynamic adjustment is executed, the comfort degree is monitored, and operation feedback is obtained; and according to the feedback continuous optimization operation strategy, energy consumption optimization is realized. According to the method, air conditioner operation can be dynamically adjusted according to the real-time activity state and the external environment, and energy consumption is effectively reduced while the comfort level is guaranteed.

Description

Technical Field

[0001] This invention relates to the field of building heating, ventilation and air conditioning control and energy-saving technology, and in particular to a heating, ventilation and air conditioning energy-saving adaptive control method and system. Background Technology

[0002] Currently, HVAC systems are an important component of building energy consumption, and optimizing their energy efficiency is crucial for the promotion and application of energy-saving air conditioners.

[0003] In one existing technology, energy-saving air conditioners typically rely on preset operating parameters or fixed control logic to regulate temperature by setting temperature thresholds, time programs, etc. Specifically, a static upper temperature limit (e.g., 26°C) or operating period (e.g., 9:00-17:00 on weekdays) is first set; the system monitors the indoor temperature through sensors, and if it exceeds the threshold, cooling is activated; during operation, the compressor frequency is adjusted only based on the temperature deviation.

[0004] However, while existing technologies are simple in structure and easy to implement, they struggle to dynamically adapt to the complex changes in the building's internal and external environments during actual operation. This is especially true in scenarios with diverse human activity patterns and frequent climate fluctuations, often exhibiting problems such as control lag and inaccurate response. Therefore, existing technologies cannot dynamically adjust air conditioning operation strategies based on real-time activity status and the external environment, leading to energy waste or decreased comfort. Summary of the Invention

[0005] This invention provides an energy-saving adaptive control method and system for HVAC systems to solve the problem of energy waste or decreased comfort caused by the inability to dynamically adjust the air conditioning operation strategy according to real-time activity status and external environment.

[0006] In a first aspect, to solve the above-mentioned technical problems, the present invention provides an energy-saving adaptive control method for HVAC systems, comprising: Collect power consumption and environmental data inside the building, and monitor power fluctuations and temperature and humidity changes based on the collected data to obtain a preliminary distribution of activity status. Based on the preliminary activity state distribution, the power change portion of air conditioner operation is separated, and the energy consumption characteristics caused by the main activities are determined. The energy consumption characteristics are compared in a pattern, and the population gathering situation and activity type in the current area are determined by combining the pre-established user behavior database to obtain a specific description of the activity status in the current area. Based on the specific activity status description, find the corresponding cooling load value for different activity types and personnel densities, and determine the target cooling output range; If the target cooling capacity output range exceeds the current system operating settings, the air conditioning operating parameters are recalculated by combining external climate fluctuation data with a pre-established environmental change adaptation rule base to obtain the adjusted operating instructions; The adjusted operating command drives the air conditioning system to dynamically adjust the cooling output, monitors the system response speed and comfort maintenance standards in real time, and obtains operating feedback data. Based on the operational feedback data, the criteria for identifying activity status and matching cooling load demand are continuously updated, real-time operation strategies are optimized, and the final energy consumption optimization target is obtained.

[0007] Secondly, the present invention provides an energy-saving adaptive control system for heating, ventilation, and air conditioning, comprising: The multi-source data acquisition and regional status perception module collects power consumption data and environmental data inside the building, and monitors power fluctuations and temperature and humidity changes based on the collected data, and integrates them to obtain a preliminary distribution of activity status. The activity energy consumption feature separation and extraction module separates the power change part of the air conditioner operation based on the preliminary activity state distribution, and determines the energy consumption features caused by the main activities; The activity pattern recognition and status analysis module performs pattern comparison on the energy consumption characteristics, and combines it with a pre-established user behavior database to determine the population gathering situation and activity type in the current area, thereby obtaining a specific activity status description in the current area. The dynamic cooling demand calculation module, based on the specific activity status description, finds the corresponding cooling load value for different activity types and personnel densities, and determines the target cooling output range. If the target cooling capacity output range exceeds the current system operation setting, the climate adaptive operation decision module will recalculate the air conditioning operation parameters by combining external climate fluctuation data with a pre-established environmental change adaptation rule base to obtain the adjusted operation instructions. The comfort tracking and closed-loop control module drives the air conditioning system to dynamically adjust the cooling output through the adjusted operation command, monitors the system response speed and comfort maintenance standard in real time, and obtains operation feedback data. The strategy adaptive optimization and evaluation module continuously updates the basis for activity status identification and cooling load demand matching based on the operational feedback data, optimizes the real-time operation strategy, and obtains the final energy consumption optimization target.

[0008] Compared with the prior art, the present invention has the following beneficial effects: (1) This invention achieves accurate identification of the activity patterns of people in the building through multi-source data fusion and activity status perception. It can distinguish the actual impact of different activity types and personnel density on the air conditioning load, thus providing a reliable basis for the fine calculation of cooling demand and avoiding the problem of excessive or insufficient cooling caused by the inability to perceive real-time activities in the existing technology.

[0009] (2) By introducing a climate adaptive decision-making and closed-loop control mechanism, this invention can dynamically adjust the air conditioning operating parameters in combination with real-time external climate fluctuations when the target cooling demand exceeds the current system setting, and make real-time corrections through comfort tracking, thereby ensuring that the system responds quickly and accurately to changes in the external environment and fluctuations in internal demand, and improving the overall energy efficiency and comfort maintenance level.

[0010] (3) Based on operational feedback data, this invention continuously optimizes the activity identification and demand matching strategies, forming an adaptive optimization closed loop of "perception-decision-execution-evaluation". This mechanism enables the system to continuously learn and adjust, and continuously approach the optimal energy consumption point in long-term operation, achieving dynamic and continuous energy-saving effects and overcoming the defects of fixed strategy control such as lag and poor adaptability. Attached Figure Description

[0011] Figure 1 This is a schematic diagram of a heating, ventilation, and air conditioning energy-saving adaptive control method provided in the first embodiment of the present invention; Figure 2 This is a schematic diagram of a heating, ventilation, air conditioning energy-saving adaptive control system provided in the second embodiment of the present invention. Detailed Implementation

[0012] 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, and 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.

[0013] Reference Figure 1 The first embodiment of the present invention provides an energy-saving adaptive control method for HVAC systems, comprising the following steps: S11 collects power consumption and environmental data inside the building, and monitors power fluctuations and temperature and humidity changes based on the collected data, and integrates them to obtain a preliminary distribution of activity status. S12, Based on the preliminary activity state distribution, separate the power change part of the air conditioner operation and determine the energy consumption characteristics caused by the main activities; S13, perform pattern comparison on the energy consumption characteristics, and combine with the pre-established user behavior database to determine the population gathering situation and activity type in the current area, and obtain a specific activity status description in the current area; S14. Based on the specific activity status description, find the corresponding cooling load value for different activity types and personnel densities, and determine the target cooling output range. S15, if the target cooling capacity output range exceeds the current system operation setting, the air conditioning operation parameters are recalculated by combining external climate fluctuation data and a pre-established environmental change adaptation rule base to obtain the adjusted operation command; S16, through the adjusted operation command, drive the air conditioning system to perform dynamic adjustment of cooling output, monitor the system response speed and comfort maintenance standard in real time, and obtain operation feedback data; S17. Based on the operational feedback data, continuously update the basis for identifying the activity status and matching the cooling load demand, optimize the real-time operation strategy, and obtain the final energy consumption optimization target.

[0014] In step S11, power consumption data and environmental data inside the building are collected, and power fluctuations and temperature and humidity changes are monitored based on the collected data to obtain a preliminary distribution of activity status, including: S1101 acquires raw power consumption data and raw environmental data from inside the building through a sensor network; S1102, calculate the power fluctuation value and temperature and humidity change rate for different regions based on the original power consumption data and the original environmental data. If they exceed the preset power fluctuation threshold or the preset temperature and humidity change rate threshold respectively, they are marked as abnormal power consumption data and abnormal environmental data. S1103, the abnormal fluctuation power consumption data and the abnormal environment data are stored in a pre-established time series database, and the corresponding initial power consumption event records and initial environment state information are extracted. S1104, K-means clustering is used to perform regional clustering on the initial electricity consumption event records and the initial environmental state information to obtain a preliminary activity state distribution.

[0015] In step S1101, raw power consumption data and raw environmental data are acquired from inside the building through a sensor network.

[0016] It should be noted that raw power consumption data and raw environmental data are acquired in real time through a sensor network deployed in multiple areas within the building, such as meeting rooms and offices. Power consumption data is collected by smart meters installed in the power distribution circuits of each area to obtain electrical power values; environmental data is collected by temperature and humidity sensors deployed in the spaces to obtain temperature and relative humidity values. Infrared pyroelectric sensors, cameras, or access control systems deployed in each area acquire data characterizing the presence and gathering of people. The sensors collect data at fixed time intervals and transmit it to the central controller via wired or wireless communication, forming a timestamped raw power consumption data sequence and a raw environmental data sequence.

[0017] In step S1102, power fluctuation values ​​and temperature and humidity change rates are calculated for different regions based on the original power consumption data and the original environmental data. If they exceed the preset power fluctuation threshold or the preset temperature and humidity change rate threshold, they are marked as abnormal power consumption data and abnormal environmental data, respectively.

[0018] It should be noted that the power fluctuation value is obtained by calculating the absolute value of the power difference between adjacent sampling time points; the temperature and humidity change rates are obtained by calculating the change per unit time, such as per hour. The calculated power fluctuation values, temperature change rates, and humidity change rates for each area are compared with their corresponding thresholds. The preset thresholds mentioned above are set based on statistical analysis of historical normal monitoring data of the building. The power fluctuation threshold is calculated by collecting historical power data of the area under baseline conditions (such as no one, only basic equipment operating), and setting its standard deviation as k times the standard deviation. The temperature and humidity change rate thresholds can be determined based on specific percentiles (such as 95%) of the historical data change rates. The historical data sample covers at least one year, different time periods each day, and various activity scenarios, with a sample size of no less than 1000 sets. The initial threshold can also be the statistical average of historical data in a typical period plus twice the standard deviation to ensure the representativeness of the threshold. If the power fluctuation value of a certain area exceeds its preset threshold, the electricity consumption data at that moment is marked as abnormal fluctuation electricity consumption data; if its temperature or humidity change rate exceeds the corresponding preset threshold, the corresponding environmental data is marked as abnormal environmental data. All tagged data is stored together with its timestamp and region identifier.

[0019] In one embodiment, for the aforementioned conference room area, if a significant change in power is detected at adjacent sampling points and exceeds a power fluctuation threshold set based on historical data statistics, the power data at that moment is marked as abnormal power consumption data. Similarly, if the temperature in the office area rises from 22 degrees Celsius to 28 degrees Celsius within one hour, the temperature change rate is 6 degrees Celsius / hour, exceeding a preset threshold of 5 degrees Celsius / hour, and the system will mark the 28-degree Celsius temperature data as abnormal environmental data.

[0020] In step S1103, the abnormal fluctuation power consumption data and the abnormal environment data are stored in a pre-established time-series database, and the corresponding initial power consumption event records and initial environment state information are extracted.

[0021] It should be noted that when the system scans time-series data, if it finds that a certain area has a persistent anomaly marker of the same type within a continuous time window, and the duration exceeds the historical average duration (e.g., 30 minutes), then a valid event is determined to have occurred. The system integrates continuous anomaly data segments that meet this condition to generate an event record. Each initial power consumption event record includes the event start time, duration, associated area, and average power and peak power within that time period; each initial environmental status information record includes the average temperature, average humidity, and their changing trends within the corresponding time period.

[0022] In one embodiment, a time-series database stores tagged abnormal data. The system scans the data chronologically. When it finds an abnormal power consumption flag for a conference room that lasts from 9:00 AM to 9:40 AM, exceeding the 30-minute threshold by 40 minutes, this time period is extracted as an initial power consumption event record. The system records the start time (9:00 AM), duration (40 minutes), area identifier (conference room A), and calculates the average and peak power during this period as features.

[0023] In step S1104, K-means clustering is used to perform regional clustering on the initial electricity consumption event records and the initial environmental state information to obtain a preliminary activity state distribution.

[0024] It should be noted that feature extraction of the extracted event records includes the average power fluctuation value, peak power value, and cumulative duration of electricity events, as well as the average temperature change rate and average humidity change rate of environmental events. A multi-dimensional feature vector is constructed for each region during the analysis period. These feature vectors are then clustered using the K-means clustering algorithm. The number of clusters K can be determined based on prior business knowledge, such as pre-defined "high, medium, and low" activity areas. The algorithm execution process includes: randomly initializing K cluster centers; calculating the Euclidean distance from each feature vector to the cluster center and assigning it to the nearest cluster; recalculating the cluster mean as the new center; and iterating until the centers stabilize or the maximum number of iterations is reached. After clustering, each cluster represents an activity pattern. For example, a cluster containing multiple meeting rooms and exhibiting high power fluctuations and long durations can be identified as a "high activity area." The system outputs the mapping relationship between regions and cluster categories as a preliminary distribution of activity states.

[0025] In step S12, the power variation portion of the air conditioner operation is separated based on the preliminary activity state distribution, and the energy consumption characteristics caused by the main activities are determined, including: S1201, extract the power sequence corresponding to the regional activity from the initial electricity consumption event record according to the preliminary activity state distribution; S1202, The power sequence is decomposed into layers using the empirical mode decomposition method to obtain the power change layers of the air conditioner operation; S1203, the power change layer is smoothed to retain the continuous power change corresponding to the start and stop of the air conditioner. If the duration of the change exceeds the preset first duration threshold, it is marked as an air conditioner operation feature, and the power feature sequence dominated by the air conditioner is obtained. S1204, perform correlation matching between the power feature sequence and the preliminary activity state distribution. If the correlation exceeds the preset first matching degree threshold, determine the air conditioning usage event corresponding to the regional activity and obtain the energy consumption event record caused by the air conditioning operation. S1205, extract the power amplitude and duration features from the energy consumption event records. If they match the activity patterns in the preliminary activity state distribution, they are determined to be energy consumption features caused by the main activity.

[0026] In step S1201, the power sequence corresponding to the regional activity is extracted from the initial electricity consumption event record according to the preliminary activity state distribution.

[0027] It should be noted that, based on the aforementioned preliminary activity state distribution, for areas identified as specific activity states, such as "high activity zones," and their corresponding time periods, all initial electricity consumption event records for those areas within that time period are retrieved from the time-series database. Continuous, timestamped power readings are extracted from these records in chronological order to form a power sequence reflecting the continuous changes in electricity consumption in that area under this activity state.

[0028] In one embodiment, if the preliminary activity status distribution shows that "Meeting Room A" is a "high activity zone" from 9:00 to 11:00 am, the system extracts the power values ​​of all power consumption events recorded in "Meeting Room A" during this period from the time series database, such as one data point every 5 minutes, to form a complete time-power sequence.

[0029] In step S1202, the power sequence is decomposed into layers using the empirical mode decomposition method to obtain the power change layers of the air conditioner operation.

[0030] It should be noted that the empirical mode decomposition method is used to decompose the power sequence into multiple intrinsic mode function components ranging from high frequency to low frequency. Based on the typical start-stop cycle of the air conditioner compressor (usually on the order of tens of minutes to several hours), a frequency threshold such as 0.01Hz is set, corresponding to a cycle of 100 seconds. Intrinsic mode function components with frequencies below this threshold are selected as the power change layer related to the continuous power change of the air conditioner.

[0031] In one embodiment, after decomposing the power sequence of the conference room, a low-frequency component with a frequency of about 0.005 Hz can be obtained. Its slow periodic change pattern is consistent with the start-stop pattern of the compressor, and it is extracted as the power change layer related to air conditioning.

[0032] In step S1203, the power change layer is smoothed to retain the continuous power change corresponding to the start and stop of the air conditioner. If the duration of the change exceeds a preset first duration threshold, it is marked as an air conditioner operation feature, thus obtaining the power feature sequence dominated by the air conditioner.

[0033] It should be noted that the first duration threshold for identifying the continuous power change platform corresponding to air conditioner start-stop in the power change layer is set by analyzing the duration distribution of historical air conditioner operation events (which can be obtained from equipment operation logs). For example, the 25th percentile of the duration can be used as the threshold to distinguish between air conditioner operation and short-term equipment interference. The historical data sample should cover at least one year, different seasons, different time periods of each day, and various typical activity scenarios, with a sample size of no less than 1000 sets. For example, if the P25 value of the duration of a historical operation event is 25 minutes, the first duration threshold can be set to 25 minutes. The specific threshold needs to be dynamically set according to the specific air conditioner system model, capacity, and historical operation data. All marked feature segments are connected in chronological order to form a power feature sequence dominated by the air conditioner.

[0034] In one embodiment, the smoothed power variation layer shows a plateau with power stabilizing at around 3kW for 25 minutes. Based on historical air conditioning operation data covering all year round and various activity scenarios, the 25th percentile of the duration of air conditioning operation events in this area is 15 minutes. Since 25 minutes is greater than 15 minutes, this plateau is marked as an air conditioning operation characteristic.

[0035] In step S1204, the power feature sequence is correlated with the preliminary activity state distribution. If the correlation exceeds the preset first matching degree threshold, the air conditioning usage event corresponding to the regional activity is determined, and the energy consumption event record caused by the air conditioning operation is obtained.

[0036] It should be noted that the power characteristic sequence dominated by air conditioning is aligned with the preliminary activity state distribution in time, and the Pearson correlation coefficient between the two in the corresponding time period is calculated as the matching degree. If the calculated matching degree exceeds the preset first matching degree threshold, it is determined that the activity in the area is highly correlated with air conditioning use, thereby identifying an air conditioning use event and generating a corresponding energy consumption event record. The first matching degree threshold is statistically set based on the correlation coefficient distribution of historically confirmed air conditioning use events. Multiple (e.g., 1000) "air conditioning use event" records verified by the system or manually within a period of time (e.g., 6 months) are collected. For each event, the Pearson correlation coefficient between its power characteristic sequence and the activity state distribution sequence is calculated, resulting in a correlation coefficient set {r1,r2,...,rn}. The mean μ and standard deviation σ of this set are calculated, and the first matching degree threshold is set to μ-kσ (where k is a constant set according to the actual misjudgment tolerance, for example, k=0.5).

[0037] In one embodiment, the Pearson correlation coefficient between the power characteristic sequence and the activity status distribution sequence of a conference room during a "high activity" period is 0.8. Based on historical data statistics, the mean μ of the correlation coefficient set is 0.7, the standard deviation σ is 0.2, and k=0.5 is set, then the first matching degree threshold is 0.6. Since 0.8 is greater than 0.6, the match is successful, and the system determines that an air conditioning usage event caused by regional activity occurred during this period.

[0038] In step S1205, power amplitude and duration features are extracted from the energy consumption event records. If they match the activity patterns in the preliminary activity state distribution, they are determined to be energy consumption features caused by major activities.

[0039] It should be noted that from the energy consumption event records, two core features are extracted for each event: the power amplitude (i.e., the average value of the stable phase of the smooth power platform) and the duration (i.e., the complete time span from a significant rise to a significant fall of the platform). These two extracted features are then matched against the preset feature range of the activity mode, such as "high activity," corresponding to the time period in the preliminary activity state distribution. The typical feature ranges of different activity modes are obtained in advance through historical data statistical analysis. When both the power amplitude and duration of an event fall within the preset range of its corresponding activity mode, the match is considered successful, and the event is identified as an energy consumption feature caused by that main activity and included in the final energy consumption features.

[0040] In one embodiment, historical statistics determine that the typical power range of an air conditioning event under the "high activity" mode is 3-8 kW, and the duration range is 1-4 hours. If an air conditioning event has a power range of 5 kW and a duration of 2.5 hours, and its occurrence period is marked as "high activity," then the characteristic values ​​of the event all fall within the preset range of the "high activity" mode, the match is successful, and it is determined to be an energy consumption characteristic caused by the main activity.

[0041] In step S13, the energy consumption characteristics are compared to determine the current population gathering situation and activity type in the current area, based on a pre-established user behavior database, to obtain a specific description of the activity status in the current area, including: S1301, based on the energy consumption characteristics, extract the corresponding personnel gathering patterns and activity type templates from the pre-established user behavior database; S1302, compare the energy consumption characteristics with the activity type template and calculate the matching score to obtain a matching score sequence; S1303, if the score in the matching score sequence exceeds the preset second matching threshold, the corresponding template is marked as a candidate activity type to obtain a set of candidate activity types; S1304, The candidate activity type set is matched with the current population gathering situation in the area. If the range of the number of people gathered is consistent with the activity intensity corresponding to the template, it is determined as the primary activity type description. S1305, combine the energy consumption characteristics to supplement the description of the primary activity type in detail, and obtain a specific activity status description in the current area.

[0042] In step S1301, the corresponding personnel gathering patterns and activity type templates are extracted from the pre-established user behavior database based on the energy consumption characteristics.

[0043] It should be noted that the pre-established user behavior database collects historical datasets, including power consumption sequences, temperature and humidity sequences for each region, time series of personnel numbers obtained through personnel counting sensors such as cameras combined with counting algorithms or access control card swipe records, and activity type labels obtained through a schedule system. The data is then time-aligned and sliced ​​into events, with each event segment containing an activity cycle. Feature vectors are extracted from each event segment, including average power, power standard deviation, event duration, average personnel density, and mean temperature and humidity changes. Using the labeled feature vectors, unsupervised clustering such as K-means is employed to cluster the feature vectors, and each cluster is then manually assigned an activity type label to form a template.

[0044] In one embodiment, the database contains a "large conference" template generated based on historical clustering. This template defines typical air conditioning-related equipment (including air conditioners, projectors, lighting, etc.) with a power range of 3.0-8.0kW (this range is an example value; actual values ​​should be calibrated according to building type and air conditioning system configuration), and a duration range of 2-4 hours. If the power range of a current energy consumption event is 4.5kW and the duration is 3 hours, both values ​​falling within the above range, the system extracts the "large conference" template from the database as the interpretation of that event.

[0045] In step S1302, the energy consumption characteristics are compared with the activity type template to calculate the matching score, and a matching score sequence is obtained.

[0046] It should be noted that, for each energy consumption event, its feature vector is extracted, including power amplitude and duration; the Euclidean distance between this feature vector and the center point of each template feature vector is calculated and converted into basic similarity in each dimension; a weighted method is used to merge the basic similarity of the two dimensions into a comprehensive matching score, where the weights are pre-set based on the distinguishability of various activity features in historical data, usually giving higher weight to power amplitude; the matching score of each event with the most relevant template is calculated and arranged in chronological order to form a sequence.

[0047] In one embodiment, an energy consumption event is characterized by a peak power of 800 watts lasting 2 hours. Its power amplitude similarity to the center point of the "Large Meeting" template (1000 watts, 2.5 hours) is 80%, and its duration similarity is 90%. If the preset weights are power amplitude 0.6 and duration 0.4, the overall matching score of this event is 84%, or 0.84.

[0048] In step S1303, if the score in the matching score sequence exceeds the preset second matching threshold, the corresponding template is marked as a candidate activity type, and a set of candidate activity types is obtained.

[0049] It should be noted that the preset second matching score threshold is set by statistically analyzing the distribution of historical matching scores. This can be achieved through receiver operation characteristic curve analysis, selecting a score that optimally balances the true positive rate (the proportion of identified genuine activities) and the false positive rate (the proportion of false alarms). Historical data samples must cover at least one year, different time periods each day, and various activity scenarios, with a sample size of no less than 1000 groups. For example, a boundary value that effectively distinguishes between target activities and non-target activity clusters can be selected. Each matching score in the sequence is iterated sequentially. If a score exceeds the preset second matching score threshold, its corresponding activity type template is marked as a candidate activity type. After the iteration is complete, all marked templates are collected to form a candidate activity type set.

[0050] In one embodiment, historical data analysis shows that the matching scores for major events such as "large-scale meetings" are mostly above 70%, while the scores for minor events are mostly below 50%. To reliably capture major events, a second matching threshold is set to 60%. If the matching score of a current event is 85%, exceeding the 60% threshold, then its corresponding "large-scale meeting" template is marked as a candidate event type and added to the candidate set.

[0051] In step S1304, the candidate activity type set is matched with the population gathering situation in the current area. If the range of the number of people gathered is consistent with the activity intensity corresponding to the template, it is determined as the primary activity type description.

[0052] It should be noted that the actual number of people in the current area is acquired in real time using sensors such as cameras. This real-time number of people is compared with the preset typical number range for each template in the candidate set. The typical number range is derived from statistical summarization of personnel data from historical similar activities. If the real-time number of people falls within the preset number range of a candidate template, the activity intensity is considered consistent, and that template is selected as the primary activity type description; if multiple templates match, the template with the highest matching score is selected.

[0053] In one embodiment, the real-time number of attendees in the meeting room was determined to be 25 through image recognition statistics. The preset attendee range for the "Large Meeting" template in the candidate set is 15-50 people (based on historical attendance statistics), and the range for the "Group Discussion" template is 3-10 people. The real-time attendee count of 25 falls within the "Large Meeting" range, therefore "Large Meeting" is determined as the primary activity type description.

[0054] In step S1305, the description of the primary activity type is supplemented with details based on the energy consumption characteristics to obtain a specific activity status description within the current area.

[0055] It should be noted that core parameters, including average power level, power fluctuation range, and event duration, are extracted from the energy consumption characteristics of the corresponding time period. These parameters are then compared and analyzed with the preset typical equipment power consumption list and operating modes in the primary activity type template. If the average power level exceeds the power consumption range for maintaining the basic environment in the template, the excess power value is matched with the standard power consumption of various types of equipment in the equipment list to infer the additional equipment in use and its operating status. Finally, details are supplemented by combining the primary activity type, the area where people gather, and the inferred equipment operating status to generate a specific activity status description.

[0056] In one embodiment, the primary activity type is "Large Meeting". The average power consumption extracted from the energy consumption characteristics is 8500 watts. The baseline power consumption of the "Large Meeting" template is 1500 watts, with an excess of 7 kilowatts. In the additional equipment list, "Projector on" consumes approximately 0.2 kilowatts, "Air Conditioner High-Power Cooling Mode" adds approximately 5 kilowatts, and "Multi-person Electronic Devices" consume approximately 1.5 kilowatts, totaling approximately 6.7 kilowatts, which is basically consistent with the excess of 7 kilowatts.

[0057] In step S14, based on the specific activity status description, the corresponding cooling load value is found for different activity types and personnel densities to determine the target cooling output range, including: In S1401, the activity type identifier and personnel density level are obtained from a pre-established activity parameter library according to the specific activity status description. In S1402, the instantaneous total cooling load is obtained by matching the human body heat dissipation pattern and metabolic heat coefficient using a lookup table based on the activity type identifier and the personnel density level. In S1403, if the total instantaneous cooling load exceeds the preset radiative cooling load threshold, the total instantaneous cooling load is adjusted by heat dissipation loss; otherwise, it is maintained, thus determining the instantaneous cooling load curve. In S1404, the instantaneous cooling load curve is analyzed by combining the preset matching accuracy threshold and the historical cooling load prediction sequence to obtain the upper limit of cooling output power and the adjustment response time, and to determine the target cooling output range.

[0058] In step S1401, the activity type identifier and personnel density level are obtained from the pre-established activity parameter library according to the specific activity status description.

[0059] It should be noted that the activity parameter library is pre-established by analyzing historical data on the actual number of participants and the area of ​​the activity. It collects detailed information on various confirmed activities (sourced from the user behavior database) over a historical period, such as six months, including activity type, location, area, and the actual number of participants obtained through personnel sensors or check-in records. Then, for each activity type, the average personnel density (number of participants / area) is calculated for each occurrence. Finally, based on the statistical distribution of personnel density across all activity types, such as the quantiles of density values, density levels are divided into a limited set of categories, such as "high," "medium," and "low," and a most suitable preset density level is assigned to each activity type. The description is parsed to extract standardized activity type names, such as "large conference," and average number of participants; using this activity type name as an index, the activity parameter library is queried to obtain its preset personnel density level.

[0060] In one embodiment, the specific activity status is described as "holding a large meeting, with approximately 20-25 participants." The activity type is parsed as "large meeting," with an average number of participants of 22. The activity parameter database is queried, where the "large meeting" entry has a preset density level of "high." This setting is based on historical statistical data, such as an average density of 0.8 people / square meter, which falls within the highest 30% density range of the sample. Based on this, the system determines the activity type as "large meeting" and the participant density level as "high."

[0061] In step S1402, the instantaneous total cooling load is obtained by matching the human body heat dissipation pattern and metabolic heat coefficient using a lookup table based on the activity type identifier and the personnel density level.

[0062] It should be noted that a predefined cooling load parameter lookup table is constructed, indexed by activity type and density level. Each table cell stores a metabolic cooling load coefficient calibrated according to relevant standards such as ASHRAE standards and combined with local historical data. This coefficient is set at approximately 130-150W per person, referencing the "seated, light activity" type in the ASHRAE standard and calibrated with local data. During calculation, the lookup table is retrieved based on the input activity type and density level to obtain the corresponding metabolic heat coefficient. The metabolic heat coefficient (W / person) is multiplied by the real-time number of people to obtain the instantaneous heat dissipation load per person (W). The total instantaneous cooling load is obtained by adding the instantaneous heat dissipation load per person, the building envelope cooling load, equipment and lighting loads, and other cooling load components.

[0063] In one embodiment, the activity type is identified as "large meeting" and the personnel density level is "high". A lookup table is used to obtain the corresponding metabolic cooling load coefficient of approximately 140W per person (referencing the "seated, light activity" type in the ASHRAE standard, calibrated using local data). If the real-time number of people is 15, the instantaneous heat dissipation load per person is calculated to be 2100W. Combining this with the building envelope cooling load of the meeting room area, for example, 1000W, and the equipment lighting load, for example, 500W, the total instantaneous cooling load is calculated to be 3600W.

[0064] In step S1403, if the total instantaneous cooling load exceeds the preset radiative cooling load threshold, the total instantaneous cooling load is adjusted by conductive heat dissipation; otherwise, it is maintained, thus determining the instantaneous cooling load curve.

[0065] It should be noted that the preset radiative cooling load threshold needs to comprehensively consider the building's inherent thermal characteristics and typical climatic conditions. Its calculation is based on a steady-state heat transfer model of the building envelope, as shown in the following formula.

[0066] in, This represents the radiative cooling load threshold (W). For the first The heat transfer coefficient (W / (m²·K)) of building envelopes such as exterior walls, windows, and roofs shall be determined in accordance with national or local building energy conservation design standards such as the "Energy Conservation Design Standard for Public Buildings" GB50189. The area (m²) of the corresponding enclosure structure; The typical indoor-outdoor design temperature difference (K) is usually selected by referring to local meteorological data, taking the difference between the representative summer indoor design temperature (e.g., 26℃) and the outdoor calculation dry-bulb temperature (e.g., the average temperature over 50 hours). This is achieved by collecting historical building cooling load data for at least one complete cooling season, including the indoor-outdoor temperature difference at various times, the surface temperature of the building envelope, and the cooling capacity of the air conditioning system, selecting an outdoor temperature close to the design temperature difference. For a given time period, calculate the average value and distribution range of the building's natural cooling load during that period, such as the 85th percentile, which will be used to calculate the theoretical value. Compared with historical statistics The data is compared, and if the deviation exceeds a reasonable range (e.g., ±15%), the threshold is adjusted according to the data distribution to determine the final value.

[0067] Here, α is a weighting coefficient (usually between 0.4 and 0.8). High α values ​​(such as 0.6-0.8) are suitable for newly built or high-standard designed buildings. Their building envelope thermal parameters are based on accurate construction drawings and material inspection reports, and the actual building condition closely matches the design model. Low α values, such as 0.4-0.6, are suitable for older buildings or complex renovations. Their actual building envelope performance may deviate significantly from design standards, limiting their reference value.

[0068] If the total instantaneous cooling load exceeds this threshold, the transient heat transfer load (per watt) of the building envelope under the current indoor-outdoor temperature difference is recalculated. The transient heat transfer load is calculated based on the heat transfer coefficient, area, and real-time indoor-outdoor temperature difference of the building envelope. The total instantaneous cooling load is compared with the transient heat transfer load. If the total instantaneous cooling load is greater than the transient heat transfer load, the net cooling load that the air conditioning system needs to bear is equal to the total instantaneous cooling load minus the transient heat transfer load; if the total instantaneous cooling load is less than or equal to the transient heat transfer load, the net cooling load is equal to 0. The instantaneous cooling load curve is obtained by calculating the net cooling load at each time point according to the time series.

[0069] In one embodiment, the product of the total heat transfer coefficient and area of ​​a conference room enclosure is 200 W / K, and the local summer design temperature difference is... =10K, then the theoretical threshold =2000W. Analysis of historical data revealed that the average natural heat dissipation capacity under similar temperature differences is 2200 W. Therefore, α = 0.6 was chosen, and the final threshold was set at 2080W. When the instantaneous cooling load exceeds 2080W, the system determines that the air conditioner needs to be activated to handle the excess cooling load.

[0070] In step S1404, the instantaneous cooling load curve is analyzed by combining the preset matching accuracy threshold and the historical cooling load prediction sequence to obtain the upper limit of cooling output power and the adjustment response time, and to determine the target cooling output range.

[0071] It should be noted that the current instantaneous cooling load curve is compared with the historical cooling load prediction sequence to calculate their morphological matching degree, such as the correlation coefficient. The matching accuracy threshold is preset by analyzing the historical prediction error distribution; for example, an initial threshold of 0.7 can be set, and calibration is performed by monitoring the historical prediction misjudgment rate (e.g., target <5%). The system can periodically and adaptively fine-tune this threshold to optimize performance based on the season and actual control effects. If the current matching degree exceeds the threshold, the prediction is considered reliable. Based on the operating data under historical reliable predictions, the upper limit of cooling output power is statistically obtained, such as by taking the historical 95th percentile value and the typical adjustment response time, such as the historical median, to jointly define the target cooling output range. After each successful execution and verification (comfort level achieved) of a control, the system stores the specific activity status description, instantaneous cooling load curve, and external climate snapshot as a sample in the historical database, forming a prediction sequence.

[0072] In one embodiment, the system initially presets a matching accuracy threshold of 0.8. After a quarter of operation and learning, the system found that during the morning hours on summer weekdays, when the threshold is slightly lowered to 0.75, it can identify reliable upward trends in cooling load demand earlier, thus initiating adjustments in advance, resulting in an overall energy efficiency improvement of 5% without causing a decrease in comfort. Therefore, the system automatically optimizes the matching accuracy threshold for this period (9-12 AM on summer weekdays) to 0.75. Currently, the matching degree between an instantaneous cooling load curve generated at 10 AM on a summer weekday and the historical sequence is 0.78. Since 0.78 exceeds the dynamic threshold of 0.75 for this period, the system determines that this prediction is reliable. A query of the historical database shows that under similar reliable prediction conditions, the statistical upper limit of cooling output power to meet demand is 2400 watts, and the statistical median adjustment response time is 10 minutes. Therefore, the target cooling output range is determined to be an adjustable upper limit of 2400 watts of cooling output power within a 10-minute adjustment response time.

[0073] In step S15, if the target cooling capacity output range exceeds the current system operating settings, the air conditioning operating parameters are recalculated by combining external climate fluctuation data with a pre-established environmental change adaptation rule base to obtain adjusted operating instructions, including: S1501, calculate the deviation between the target cooling output range and the current system operation setting. If it exceeds the preset deviation threshold, obtain external climate fluctuation data from the sensor interface to obtain a climate fluctuation sequence. S1502, Based on the climate fluctuation sequence, a matching rule is searched from the pre-established environmental change adaptation rule base to obtain an adaptation rule set; S1503, combine the adaptation rule set with the current system operation settings to calculate the adjustment amount of the air conditioning operation parameters and obtain the adjustment parameter sequence; S1504, The adjustment parameter sequence is formatted with instructions to obtain the adjusted running instructions.

[0074] In step S1501, the deviation between the target cooling output range and the current system operation setting is calculated. If the deviation exceeds the preset deviation threshold, external climate fluctuation data is obtained from the sensor interface to obtain the climate fluctuation sequence.

[0075] It should be noted that the absolute difference between the center value of the target cooling output range and the currently set cooling output value of the system is calculated, and the percentage of this difference relative to the target center value is used as the relative deviation. The preset deviation threshold is set by analyzing the historical deviation distribution between the system set value and the theoretically calculated demand value under the premise of meeting the comfort standard; for example, the 80th percentile of the absolute value of the deviation is taken. If the calculated relative deviation exceeds the preset threshold, external climate data such as temperature, humidity, and solar radiation intensity are acquired in real time through the sensor interface, collected at fixed time intervals, to form a climate fluctuation sequence.

[0076] In one embodiment, the target cooling output center value is 2400 watts, while the system is currently set to 2000 watts, with a relative deviation of 10%. Historical data analysis shows that, assuming comfort levels are met, the deviation between the theoretical demand value and the system setting is usually less than 10%, so the threshold is set to 10%. Since 16.7% is greater than 10%, climate data acquisition is triggered. The system reads data from the interface indicating that the temperature will rise from 25°C to 30°C and the humidity will rise from 50% to 70% within the next 15 minutes, recording this as a climate fluctuation sequence with timestamps.

[0077] In step S1502, based on the climate fluctuation sequence, matching rules are searched from a pre-established environmental change adaptation rule base to obtain an adaptation rule set.

[0078] It should be noted that the rule base is constructed based on the mapping relationship between historical climate characteristics and verified optimal energy-efficient air conditioning control parameters. This mapping relationship is mainly constructed through the following steps: during the system learning phase, climate data including external temperature, humidity, and solar radiation intensity are continuously collected, and the operating parameters of the air conditioning system and real-time indoor environmental comfort and system energy consumption data are recorded simultaneously for the corresponding time period. For each set of historical climate data collected, the system selects the combination of operating parameters that maximizes the energy efficiency ratio of the air conditioning system or minimizes the energy consumption per unit cooling capacity from all concurrent operating records that meet the preset indoor comfort standards. This combination is marked as the "recommended operating parameters" under that specific climate condition. Then, the system analyzes the change patterns of climate characteristics and associates the above-mentioned climate condition characteristics with the corresponding "recommended operating parameter" adjustment strategies to form rules in the form of "if a specific combination of climate conditions is met, then the corresponding parameter adjustment is executed".

[0079] When a matching rule needs to be found, the system quantifies the real-time climate fluctuation sequence into a feature vector containing key features such as the rate of temperature change. This feature vector is then compared one by one with the climate condition range defined by each rule in the rule base. It determines whether the real-time feature value falls within the numerical range defined by the rule conditions. If the current climate feature meets all or all of the core conditions of a rule, the rule is considered a successful match. The parameter adjustment instructions contained in the conclusions of all successfully matched rules are extracted and collectively form the adaptive rule set for the current moment.

[0080] In one embodiment, the real-time climate fluctuation sequence is quantified as a feature: the temperature rises by 5°C and the humidity increases by 20% within 15 minutes. A rule exists in the rule base whose condition is defined as "temperature increase between 4-6°C and humidity increase between 15-25%", and whose conclusion is "cooling output setting increased by 15%, fan speed adjusted to medium". Since the real-time feature (5°C, 20%) falls entirely within the condition range of this rule, the system determines that the rule is a successful match and includes it in the adaptive rule set.

[0081] In step S1503, the adaptation rule set is combined with the current system operation settings to calculate the adjustment amount of the air conditioning operation parameters and obtain the adjustment parameter sequence.

[0082] It should be noted that the target value of the parameter indicated by the adaptation rule set is read, and the difference between the target value and the current system setting is calculated as the total adjustment amount. Based on the system's dynamic response characteristics, the total adjustment amount is decomposed into multiple smaller adjustment amounts of equal step size, and these are applied sequentially according to a preset adjustment cycle, forming a time-ordered sequence of adjustment parameters. The adjustment step size and cycle are based on the stabilization time and inertia calibration of the historical adjustment process to ensure smooth adjustment. According to the system's inertial time constant and adjustment stability requirements, the adjustment step size is set to 1 / N of the total adjustment amount, where N is the number of adjustments.

[0083] In one embodiment, the adaptation rule set indicates that the cooling output needs to be increased by 15% from 1800 watts to 2070 watts, a total adjustment of 270 watts. Based on historical system response data (stabilization time approximately 5 minutes, single-step adjustments should not exceed 100 watts), the adjustment step size is set to 90 watts, with a cycle of 1.5 minutes. The generated adjustment parameter sequence is adjusted sequentially to 1890 watts, 1980 watts, and 2070 watts, with adjacent settings spaced 1.5 minutes apart.

[0084] In step S1504, the adjustment parameter sequence is formatted to obtain the adjusted execution instruction.

[0085] It should be noted that the adjustment parameter sequence is formatted with instructions. Based on a pre-configured device control mapping table, logical parameters in the sequence, such as target cooling capacity and fan speed, are converted into specific control parameters executable by the air conditioning hardware, such as compressor frequency and fan speed. The mapping table is established based on device performance curves and historical calibration data. After conversion, the control parameters are encapsulated into standardized instruction data packets according to the communication protocol agreed upon with the air conditioning controller.

[0086] In one embodiment, the parameter sequence is adjusted to require a cooling capacity of 2070 watts and a fan speed of "medium". Looking up the mapping table, a cooling capacity of 2070 watts corresponds to a compressor frequency of 52Hz (1Hz for every 40 watts), and "medium" corresponds to a fan speed of 800 rpm. Following the Modbus RTU protocol, these parameters are encapsulated into an instruction frame containing the device address, function code, register address, data, and CRC checksum, resulting in a directly executable operating command.

[0087] In step S16, the adjusted operating command drives the air conditioning system to dynamically adjust the cooling capacity, monitors the system response speed and comfort maintenance standards in real time, and obtains operational feedback data, including: S1601, the adjusted operating command controls the cooling capacity regulating valve and compressor speed, drives the air conditioner to perform dynamic cooling capacity adjustment, and obtains the initial execution response sequence; S1602, if the response speed in the initial execution response sequence exceeds the preset response speed threshold, then real-time environmental data is obtained from the temperature sensor and humidity sensor to obtain the response monitoring data sequence. S1603, Based on the response monitoring data sequence and the preset comfort maintenance standard, calculate the indoor temperature and humidity deviation, adjust the fan speed, and determine the comfort feedback sequence; S1604, the operation status log is collected from the actuator interface through the comfort feedback sequence. When the parameter fluctuation is stable within the preset range, it is determined that the dynamic adjustment is completed, thereby obtaining operation feedback data containing the execution process and the final state.

[0088] In step S1601, the cooling capacity regulating valve and compressor speed are controlled by the adjusted running command to drive the air conditioner to perform dynamic adjustment of cooling capacity and obtain the initial execution response sequence.

[0089] It should be noted that the air conditioning main controller receives and parses the instructions, sending the device control parameters contained in the instructions to the cooling capacity regulating valve driver and the compressor inverter, respectively. The driver linearly adjusts the valve opening according to the target opening value at a preset response speed; the inverter smoothly adjusts the compressor speed to the target value according to the target frequency value. The response speed and the correspondence between the valve opening and the compressor speed are predetermined based on air conditioning system performance test data and historical operation calibration results. During the adjustment process, the system synchronously collects and records timestamps, valve opening feedback values, compressor speed feedback values, and estimated total cooling capacity output values ​​at preset fixed time intervals (sampling period), forming an initial execution response sequence. This sampling period is set based on the system's typical thermal response time constant and data accuracy requirements, typically ranging from 30 seconds to 2 minutes.

[0090] In one embodiment, the system parses instructions aimed at increasing the opening of the cooling capacity regulating valve to 70% and increasing the compressor frequency to 46Hz. The controller drives the valve to linearly open from 50% to 70% within 2 minutes, while simultaneously smoothly increasing the compressor speed from 3000 rpm to 3200 rpm. During this process, the system records the valve position, speed, and calculated real-time cooling capacity every minute for 2 minutes, generating a sequence containing 24 sets of time-state data.

[0091] In step S1602, if the response speed in the initial execution response sequence exceeds a preset response speed threshold, real-time environmental data is obtained from the temperature sensor and humidity sensor to obtain a response monitoring data sequence.

[0092] It should be noted that response speed is defined as the time required for the system's actual cooling output to reach 90% of the target adjustment value. The response speed threshold is set by statistically analyzing the distribution of this time in historical normal adjustment events, for example, using the 95th percentile value. If the measured response time exceeds this threshold, the system synchronously reads data from the temperature and humidity sensor interfaces deployed within the area at a preset fixed sampling frequency, forming a response monitoring data sequence containing timestamps, location identifiers, temperature values, and humidity values. This sampling frequency is set by comprehensively considering the inertia of environmental parameter changes and monitoring requirements, typically using a period of 30 seconds to 2 minutes.

[0093] In one embodiment, historical data shows that 95% of normal adjustments can reach 90% of the target cooling capacity within 2 minutes, so a response speed threshold of 2 minutes is set. The current sequence shows that after 2 minutes, the actual cooling capacity only reaches 70% of the target, and the response times out. The system then triggers monitoring, collecting data from 4 sensors in the conference room every 1 minute for 5 minutes, generating a monitoring sequence containing 240 data points.

[0094] In step S1603, based on the response monitoring data sequence and the preset comfort maintenance standard, the indoor temperature and humidity deviation is calculated, the fan speed is adjusted, and the comfort feedback sequence is determined.

[0095] It should be noted that the comfort maintenance standard, for example, under summer conditions, is set at a temperature of 24-26°C and a relative humidity of 40-60%. This range is determined based on regional design requirements and with reference to the comfort zone provisions in ASHRAE Standard 55, "Thermal Environmental Conditions for Human Occupancy," under the assumption of specific clothing (such as summer clothing) and a light level of activity. The average readings from multiple sensors are extracted from the monitoring sequence to calculate the current overall state, and the deviation is obtained by comparing it with the standard value. A proportional-integral control strategy is used to adjust the fan speed. The base speed increment is calculated according to the current temperature and humidity deviation value using a preset proportional coefficient, and the historical deviation is integrated to calculate the compensation amount. The proportional coefficient and integration time are set based on the regional thermal inertia and historical adjustment data. The system continuously records the timestamp, environmental deviation value, and corresponding fan speed setpoint for each adjustment cycle, forming a comfort feedback sequence.

[0096] In one embodiment, based on ASHRAE Standard 55 and considering local building characteristics, the target comfort parameters are set as follows under typical conditions of summer clothing and light activity: temperature 25°C and relative humidity 50%. The current calculation yields an average temperature of 26°C and humidity of 65%, with deviations of +2°C and +10%, respectively. A base increment is calculated based on preset proportionality coefficients (each 1°C temperature deviation corresponds to a 100 rpm increase in fan speed, and each 5% humidity deviation corresponds to a 50 rpm increase in fan speed). Combined with the deviation integral compensation from the past 10 minutes, the fan speed is increased from 800 rpm to 1100 rpm. The system records the deviation value and fan speed every 30 seconds, forming a feedback sequence.

[0097] In step S1604, the operating status log is collected from the actuator interface through the comfort feedback sequence. When the parameter fluctuation is stable within the preset range, it is determined that the dynamic adjustment is completed, thereby obtaining operating feedback data containing the execution process and the final state.

[0098] It should be noted that the system continuously collects key parameters such as compressor speed, fan speed, valve opening, and real-time power from the actuator interface to form an operation status log. Simultaneously, it monitors the environmental deviation values ​​in the comfort feedback sequence. When the fluctuation amplitudes of all key parameters in the operation status log and the environmental deviation fluctuation amplitudes are stable within their respective preset allowable fluctuation ranges within a consecutive preset monitoring time window, the dynamic adjustment is deemed complete. The monitoring time window and allowable fluctuation range are determined by analyzing the fluctuation characteristics of historical steady-state operation data. After the dynamic adjustment is determined to be complete, the operation status log, comfort feedback sequence, and final stable parameter values ​​throughout the adjustment process are packaged together into the operation feedback data.

[0099] In one embodiment, the preset monitoring window is 5 minutes, the allowable fluctuation range of compressor speed is ±1%, and the allowable fluctuation range of indoor temperature deviation is ±0.2℃. These thresholds are derived based on statistical analysis of historical steady-state data. The adjustment is considered complete when the system detects that all parameters have not exceeded their limits for 5 consecutive minutes. The generated operational feedback data includes a complete time-series record of the entire process and the final state, such as valve opening at 65% and compressor speed at 3450 rpm.

[0100] In step S17, based on the operational feedback data, the basis for identifying the activity status and matching the cooling load demand is continuously updated, the real-time operation strategy is optimized, and the final energy consumption optimization target is obtained, including: S1701, Based on the analysis of the operation feedback data, the activity status identification criteria and the cooling load demand matching criteria are used to calculate the current activity level deviation and cooling load deviation, and a status matching sequence is obtained; S1702, extract the activity level deviation value from the state matching sequence. If it exceeds the preset activity level deviation threshold, obtain real-time cooling load demand data from the operation feedback data and determine the demand optimization sequence. S1703, adjust the operating strategy parameters through the demand optimization sequence, update the compressor frequency and valve opening, and obtain the strategy adjustment sequence; S1704, Collect energy consumption monitoring data after the strategy adjustment sequence is executed. If the energy consumption is reduced to the preset optimization target and remains stable, it is determined that the energy consumption optimization target has been achieved, and the final optimization feedback data is generated.

[0101] In step S1701, based on the analysis of the operation feedback data, the activity status identification criteria and the cooling load demand matching criteria are used to calculate the current activity level deviation and cooling load deviation, thereby obtaining the status matching sequence.

[0102] It should be noted that the activity status identification criteria can be extracted from the personnel sensor readings included in the operational feedback data, such as those of infrared pyroelectric sensors deployed within the area. By analyzing the sensor signals, the number of valid human movement signals detected per unit time, such as per minute, is calculated as the movement frequency. The cooling load demand matching criteria can be extracted from the estimated cooling load values ​​calculated from data such as indoor and outdoor temperatures and equipment power. By comparing the extracted current activity level, such as personnel movement frequency, with the preset standard activity level for the corresponding time period, the activity level deviation is calculated; by comparing the real-time estimated cooling load value with the target cooling load value, the cooling load deviation is calculated. These deviation values ​​are recorded in chronological order to form a status matching sequence.

[0103] In one embodiment, operational feedback data shows that the current human movement frequency in the area, analyzed by infrared sensors, is 15 times per minute, while the preset standard frequency for "low-level activity" during this period is 5 times per minute, indicating a significant positive deviation in activity level. Simultaneously, the actual cooling load is 20% higher than the target value due to sunlight exposure, resulting in a positive cooling load deviation as well. The system records these two deviation values ​​and their corresponding timestamps in the state matching sequence.

[0104] In step S1702, the activity level deviation value is extracted from the state matching sequence. If it exceeds the preset activity level deviation threshold, real-time cooling load demand data is obtained from the operation feedback data to determine the demand optimization sequence.

[0105] It should be noted that the activity level deviation threshold is set by statistically analyzing the distribution of activity level deviation values ​​collected during historical normal operation. During the initial system learning phase or by selecting a representative operating period, such as three consecutive months, activity level deviation values ​​calculated under normal activity scenarios are collected, forming a deviation value set {d1, d2, ..., dn}. A specific percentile of this set is calculated, for example, the 95th percentile (P95), and this percentile value is set as the activity level deviation threshold. When the activity level deviation value extracted in subsequent operations exceeds this threshold, it will be judged as a significant anomaly, thereby triggering the optimization process. If the extracted deviation value exceeds the threshold, the activity status perception is determined to be abnormal, and demand optimization is required. The system extracts the estimated cold load values ​​for the corresponding time period and a subsequent period from the operational feedback data, arranges them in chronological order, and forms a demand optimization sequence.

[0106] In one embodiment, based on three months of historical normal operation data, the 95th percentile of the calculated activity level deviation value set is 28%. The system sets the activity level deviation threshold to 28%. If a deviation value of 35% is extracted from the current state matching sequence at a certain moment, optimization is triggered. The system obtains the cooling load data sequence for the next 15 minutes from that moment from the operation feedback data, such as [145kW, 148kW, 150kW, ...], and combines it with timestamps to form a demand optimization sequence.

[0107] In step S1703, the operating strategy parameters are adjusted through the demand optimization sequence, and the compressor frequency and valve opening are updated to obtain the strategy adjustment sequence.

[0108] It should be noted that the adjustment amount is calculated based on the deviation between the real-time cooling load value and the target value in the demand optimization sequence. The proportional term calculates the basic adjustment amount according to the current deviation using a preset coefficient; the integral term accumulates historical deviations to eliminate steady-state errors; and the derivative term predicts the trend based on the rate of change of the deviation. The proportional coefficient, integral time, and derivative time of the control algorithm are predetermined through system identification and parameter tuning based on the historical step response data of the air conditioning system. The adjustment process outputs a series of compressor frequency setpoints and valve opening setpoints that vary over time, constituting the strategy adjustment sequence.

[0109] In one embodiment, the demand optimization sequence shows that the actual cooling load is consistently higher than the set value of 150kW. After calculation by the PID controller, the proportional term increases the compressor frequency from 50Hz to 53Hz; the integral term further compensates to 55Hz; and the derivative term predicts the increase in cooling load and adjusts the valve opening from 80% to 85%. The system outputs a set of parameters every 30 seconds to form a strategy adjustment sequence, such as [(53Hz, 82%), (54Hz, 83%), (55Hz, 85%).

[0110] In step S1704, energy consumption monitoring data is collected after the strategy adjustment sequence is executed. If the energy consumption reduction reaches the preset optimization target and remains stable, it is determined that the energy consumption optimization target has been achieved, and the final optimization feedback data is generated.

[0111] It should be noted that the real-time operating power of the air conditioning system is collected at fixed intervals via an electricity meter or power meter interface, and the cumulative energy consumption within a preset evaluation period is calculated. This cumulative energy consumption is compared with the historical average cumulative energy consumption for the same period before the strategy adjustment, under the same baseline conditions, to calculate the percentage reduction in energy consumption. The preset optimization target percentage is not a fixed value, but is dynamically set based on in-depth analysis of historical energy consumption data. By analyzing the building's long-term historical operating data, including by time period, season, and activity mode, a baseline energy consumption model is established to determine the "normal" energy consumption level under different typical operating conditions. Then, combined with the design energy efficiency of the air conditioning system, the equipment's service life, and maintenance status, the theoretical energy-saving potential of the current system is assessed. Finally, based on the above analysis, reasonable and differentiated energy consumption reduction target ranges are set for different types of control strategies and different operating conditions. If the percentage reaches or exceeds the preset optimization target percentage and remains stable within a subsequent continuous monitoring time window, the target is considered achieved. The length of the monitoring window is also determined based on the system's thermal inertia and the time characteristics of stable control effects in historical data. The final generated optimization feedback data includes strategy adjustment details, energy consumption comparison data, and target achievement status.

[0112] In one embodiment, after the strategy adjustment, the cumulative energy consumption during the 10-11 AM assessment period was 85 kWh, compared to a historical average of 100 kWh under similar conditions, representing a 15% reduction. Based on analysis, the system set a dynamic optimization target of over 8% for this type of meeting room scenario with high personnel density. Since 15% is greater than 8% and the energy consumption fluctuation was less than the preset ±2% allowable range within the subsequent 10-minute monitoring window, the target was deemed achieved. The system generated feedback data, recording that this optimization reduced energy consumption by 15% and maintained it stably.

[0113] In summary, this invention discloses an energy-saving adaptive control method for HVAC systems. For energy-efficient air conditioning systems, it includes real-time collection and analysis of electricity consumption and environmental data within the building to accurately perceive regional activity states; then, it separates the energy consumption characteristics dominated by air conditioning, and identifies specific activity types and personnel density by combining a pre-set behavioral database; based on this, it calculates dynamic cooling demand, and when the demand exceeds the current setting, it integrates external climate data to generate adaptive adjustment commands; it executes dynamic adjustments through closed-loop control and monitors comfort levels, ultimately continuously optimizing the strategy based on operational feedback. This invention effectively reduces energy consumption while ensuring comfort by dynamically adjusting air conditioning operation according to real-time activity states and the external environment.

[0114] Reference Figure 2 The second embodiment of the present invention provides an energy-saving adaptive control system for heating, ventilation and air conditioning, comprising: The multi-source data acquisition and regional status perception module collects power consumption data and environmental data inside the building, and monitors power fluctuations and temperature and humidity changes based on the collected data, and integrates them to obtain a preliminary distribution of activity status. The activity energy consumption feature separation and extraction module separates the power change part of the air conditioner operation based on the preliminary activity state and determines the energy consumption features caused by the main activities. The activity pattern recognition and status analysis module performs pattern comparison on the energy consumption characteristics, and combines it with a pre-established user behavior database to determine the population gathering situation and activity type in the current area, thereby obtaining a specific activity status description in the current area. The dynamic cooling demand calculation module, based on the specific activity status description, finds the corresponding cooling load value for different activity types and personnel densities, and determines the target cooling output range. If the target cooling capacity output range exceeds the current system operation setting, the climate adaptive operation decision module will recalculate the air conditioning operation parameters by combining external climate fluctuation data with a pre-established environmental change adaptation rule base to obtain the adjusted operation instructions. The comfort tracking and closed-loop control module drives the air conditioning system to dynamically adjust the cooling output through the adjusted operation command, monitors the system response speed and comfort maintenance standard in real time, and obtains operation feedback data. The strategy adaptive optimization and evaluation module continuously updates the basis for activity status identification and cooling load demand matching based on the operational feedback data, optimizes the real-time operation strategy, and obtains the final energy consumption optimization target.

[0115] It should be noted that the HVAC energy-saving adaptive control system provided in this embodiment of the invention is used to execute all the process steps of the HVAC energy-saving adaptive control method in the above embodiment. The working principles and beneficial effects of the two are one-to-one, so they will not be described again.

[0116] This invention also provides an electronic device. The electronic device includes a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a heating, ventilation, and air conditioning (HVAC) energy-saving adaptive control method program. When the processor executes the computer program, it implements the steps described in each of the above-described HVAC energy-saving adaptive control method embodiments, for example... Figure 1 The step S11 shown. Alternatively, when the processor executes the computer program, it implements the functions of each module / unit in the above system embodiments, such as the climate adaptive operation decision module.

[0117] It should be noted that the system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the system embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0118] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for adaptive energy-saving control of HVAC systems, characterized in that, include: Collect power consumption and environmental data inside the building, and monitor power fluctuations and temperature and humidity changes based on the collected data to obtain a preliminary distribution of activity status. Based on the preliminary activity state distribution, the power change portion of air conditioner operation is separated, and the energy consumption characteristics caused by the main activities are determined. The energy consumption characteristics are compared in a pattern, and the population gathering situation and activity type in the current area are determined by combining the pre-established user behavior database to obtain a specific description of the activity status in the current area. Based on the specific activity status description, find the corresponding cooling load value for different activity types and personnel densities, and determine the target cooling output range; If the target cooling capacity output range exceeds the current system operating settings, the air conditioning operating parameters are recalculated by combining external climate fluctuation data with a pre-established environmental change adaptation rule base to obtain the adjusted operating instructions; The adjusted operating command drives the air conditioning system to dynamically adjust the cooling output, monitors the system response speed and comfort maintenance standards in real time, and obtains operating feedback data. Based on the operational feedback data, the criteria for identifying activity status and matching cooling load demand are continuously updated, real-time operation strategies are optimized, and the final energy consumption optimization target is obtained.

2. The HVAC energy-saving adaptive control method according to claim 1, characterized in that, Collect electricity consumption and environmental data inside the building, and monitor power fluctuations and temperature and humidity changes based on the collected data to obtain a preliminary distribution of activity status, including: Raw power consumption and environmental data are acquired from inside the building through sensor networks; Based on the original electricity consumption data and the original environmental data, calculate the power fluctuation value and temperature and humidity change rate for different regions. If they exceed the preset power fluctuation threshold or the preset temperature and humidity change rate threshold, they are marked as abnormal fluctuation electricity consumption data and abnormal environmental data. The abnormal fluctuation power consumption data and the abnormal environment data are stored in a pre-established time-series database, and the corresponding initial power consumption event records and initial environment state information are extracted. K-means clustering was used to perform regional clustering on the initial electricity consumption event records and the initial environmental state information to obtain a preliminary activity state distribution.

3. The HVAC energy-saving adaptive control method according to claim 2, characterized in that, Based on the preliminary activity state distribution, the power variation component of air conditioning operation is separated, and the energy consumption characteristics caused by the main activities are determined, including: Based on the preliminary activity state distribution, extract the power sequence corresponding to the regional activity from the initial electricity consumption event records; The power sequence is decomposed into hierarchical layers using the empirical mode decomposition method to obtain the power variation layers of the air conditioner operation; The power change layer is smoothed to retain the continuous power change corresponding to the start and stop of the air conditioner. If the duration of the change exceeds a preset first duration threshold, it is marked as an air conditioner operation feature, thus obtaining the power feature sequence dominated by the air conditioner. The power feature sequence is correlated with the preliminary activity state distribution. If the correlation exceeds the preset first matching degree threshold, the air conditioning usage event corresponding to the regional activity is determined, and the energy consumption event record caused by the air conditioning operation is obtained. The power amplitude and duration features are extracted from the energy consumption event records. If they match the activity patterns in the preliminary activity state distribution, they are identified as energy consumption features caused by the main activity.

4. The HVAC energy-saving adaptive control method according to claim 1, characterized in that, The energy consumption characteristics are compared using patterns, and combined with a pre-established user behavior database to determine the population gathering situation and activity type in the current area, thereby obtaining a specific description of the activity status in the current area, including: Based on the energy consumption characteristics, corresponding personnel gathering patterns and activity type templates are extracted from a pre-established user behavior database; The energy consumption characteristics are compared with the activity type template to obtain a matching score sequence by calculating the matching score. If the score in the matching score sequence exceeds the preset second matching threshold, the corresponding template is marked as a candidate activity type, and a set of candidate activity types is obtained; The candidate activity type set is matched with the population gathering situation in the current area. If the range of the number of people gathered is consistent with the activity intensity corresponding to the template, it is determined as the primary activity type description. By combining the energy consumption characteristics with the detailed description of the primary activity type, a specific activity status description for the current area is obtained.

5. The HVAC energy-saving adaptive control method according to claim 1, characterized in that, Based on the specific activity status description, the corresponding cooling load value is found for different activity types and personnel densities to determine the target cooling capacity output range, including: Based on the specific activity status description, obtain the activity type identifier and personnel density level from the pre-established activity parameter library; The total instantaneous cooling load is obtained by matching the human body heat dissipation pattern and metabolic heat coefficient using a lookup table based on the activity type identifier and the personnel density level. If the total instantaneous cooling load exceeds the preset radiative cooling load threshold, the total instantaneous cooling load is adjusted by conductive heat dissipation; otherwise, it is maintained, thus determining the instantaneous cooling load curve. By combining the preset matching accuracy threshold and the historical cooling load prediction sequence, the instantaneous cooling load curve is analyzed to obtain the upper limit of cooling output power and the adjustment response time, and to determine the target cooling output range.

6. The HVAC energy-saving adaptive control method according to claim 1, characterized in that, If the target cooling capacity output range exceeds the current system operating settings, the air conditioning operating parameters are recalculated by combining external climate fluctuation data with a pre-established environmental change adaptation rule base to obtain adjusted operating instructions, including: The deviation between the target cooling output range and the current system operation setting is calculated. If the deviation exceeds the preset threshold, external climate fluctuation data is obtained from the sensor interface to obtain the climate fluctuation sequence. Based on the climate fluctuation sequence, matching rules are searched from a pre-established environmental change adaptation rule base to obtain an adaptation rule set; By combining the adaptive rule set with the current system operation settings, the adjustment amount of the air conditioning operation parameters is calculated to obtain the adjustment parameter sequence; The adjusted parameter sequence is formatted to obtain the adjusted execution command.

7. The HVAC energy-saving adaptive control method according to claim 1, characterized in that, The adjusted operating command drives the air conditioning system to dynamically adjust the cooling capacity, monitors the system response speed and comfort maintenance standards in real time, and obtains operational feedback data, including: The adjusted operating command controls the cooling capacity regulating valve and compressor speed, driving the air conditioner to perform dynamic cooling capacity adjustment, thus obtaining the initial execution response sequence. If the response speed in the initial execution response sequence exceeds the preset response speed threshold, real-time environmental data is obtained from the temperature sensor and humidity sensor to obtain the response monitoring data sequence. Based on the response monitoring data sequence and the preset comfort maintenance standard, the indoor temperature and humidity deviation is calculated, the fan speed is adjusted, and the comfort feedback sequence is determined. The comfort feedback sequence is used to collect the operation status log from the actuator interface. When the parameter fluctuation is stable within the preset range, it is determined that the dynamic adjustment is completed, thereby obtaining operation feedback data containing the execution process and the final state.

8. The HVAC energy-saving adaptive control method according to claim 1, characterized in that, Based on the operational feedback data, the criteria for identifying activity status and matching cooling load demand are continuously updated to optimize real-time operation strategies and obtain the final energy consumption optimization target, including: Based on the analysis of the operational feedback data, the activity status identification criteria and the cooling load demand matching criteria are used to calculate the current activity level deviation and cooling load deviation, and a status matching sequence is obtained. The activity level deviation value is extracted from the state matching sequence. If it exceeds the preset activity level deviation threshold, the real-time cooling load demand data is obtained from the operation feedback data to determine the demand optimization sequence. By adjusting the operating strategy parameters through the aforementioned demand optimization sequence, the compressor frequency and valve opening are updated to obtain the strategy adjustment sequence; Collect energy consumption monitoring data after the strategy adjustment sequence is executed. If the energy consumption is reduced to the preset optimization target and remains stable, it is determined that the energy consumption optimization target has been achieved, and the final optimization feedback data is generated.

9. A heating, ventilation, and air conditioning energy-saving adaptive control system, characterized in that, include: The multi-source data acquisition and regional status perception module collects power consumption data and environmental data inside the building, and monitors power fluctuations and temperature and humidity changes based on the collected data, and integrates them to obtain a preliminary distribution of activity status. The activity energy consumption feature separation and extraction module separates the power change part of the air conditioner operation based on the preliminary activity state distribution, and determines the energy consumption features caused by the main activities; The activity pattern recognition and status analysis module performs pattern comparison on the energy consumption characteristics, and combines it with a pre-established user behavior database to determine the population gathering situation and activity type in the current area, thereby obtaining a specific activity status description in the current area. The dynamic cooling demand calculation module, based on the specific activity status description, finds the corresponding cooling load value for different activity types and personnel densities, and determines the target cooling output range. If the target cooling capacity output range exceeds the current system operation setting, the climate adaptive operation decision module will recalculate the air conditioning operation parameters by combining external climate fluctuation data with a pre-established environmental change adaptation rule base to obtain the adjusted operation instructions. The comfort tracking and closed-loop control module drives the air conditioning system to dynamically adjust the cooling output through the adjusted operation command, monitors the system response speed and comfort maintenance standard in real time, and obtains operation feedback data. The strategy adaptive optimization and evaluation module continuously updates the basis for activity status identification and cooling load demand matching based on the operational feedback data, optimizes the real-time operation strategy, and obtains the final energy consumption optimization target.

Images