Intelligent air conditioner energy-saving control system and method based on AI optimization

Abstract

The invention discloses an intelligent air conditioner energy-saving control system and method based on AI optimization, relates to the technical field of energy-saving control, solves the technical problems that state judgment and fault early warning are lagged and an energy-saving strategy lacks pertinence, avoids limitation of a single threshold value by constructing a normal working interval of a sub-scene, and improves the energy-saving control efficiency. The potential anomaly analysis unit reduces the misjudgment rate through trend curve modeling and multi-dimensional mutation environment judgment and correlation parameter verification, accurately distinguishes environment fluctuation and equipment anomaly, achieves accurate tracing of fault causes through construction of an anomaly cause knowledge base and combination of deep learning feature extraction and strong correlation parameter analysis, does not need manual troubleshooting one by one, and improves the detection efficiency. According to the method, the fault response time is shortened, the service life of equipment is prolonged, a scenarized energy-saving strategy is generated through an AI algorithm based on the three-dimensional combination characteristics of indoor and outdoor temperature difference, personnel density and parameter change rate, dynamic linkage of load adaptation-parameter optimization is achieved, system energy consumption is reduced, and indoor comfort can be guaranteed.

Description

Technical Field

[0001] This invention relates to the field of energy-saving control technology, specifically to an AI-optimized intelligent air conditioning energy-saving control system and method. Background Technology

[0002] With the growth of global energy demand and the increasing severity of environmental problems, improving energy efficiency has become a key focus for all industries. As one of the main components of building energy consumption, the energy efficiency optimization of air conditioning systems is of great significance for achieving energy conservation and emission reduction goals.

[0003] According to patent application number CN202411746669.5, an AI-based energy-saving control system and method for water-cooled central air conditioning is disclosed. It uses deep learning-based artificial intelligence technology to comprehensively monitor and analyze the operating data of the water-cooled central air conditioning chiller room to discover the time-series correlation change patterns of multi-source operating parameters, thereby obtaining a comprehensive characterization of the chiller room's operating status. At the same time, combined with the set value of the chilled water outlet temperature of the chiller unit, it conducts cross-modal interactive analysis of the chilled water outlet set temperature and the chiller room's operating status to discover the operating status response characteristics of the chiller room at the current set temperature, and intelligently recommends the chilled water outlet temperature based on this.

[0004] However, existing air conditioning energy-saving control systems mostly rely on manual adjustment or fixed threshold control during use, and cannot dynamically adapt to changes in the environment and equipment operating status, resulting in serious energy waste. Secondly, they only collect a small number of core operating parameters, lack comprehensive capture of air conditioning electrical parameters and multi-dimensional environmental parameters, and do not perform fine preprocessing of data, making it difficult to support accurate decision-making. At the same time, they have not established a dynamic correlation between environmental parameters and air conditioning load, and energy-saving control is based on a single temperature index, which cannot balance comfort and energy-saving effect. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an AI-optimized intelligent air conditioning energy-saving control system and method, which solves the problems of delayed status determination and fault early warning, as well as the lack of targeted energy-saving strategies.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an AI-optimized intelligent air conditioning energy-saving control system, comprising: The intelligent data acquisition unit is used to collect air conditioning parameters and environmental parameters, preprocess the parameters to obtain preprocessed parameters, and transmit them to the parameter change analysis unit. The parameter change analysis unit is used to identify the air conditioner status based on the preprocessed parameters, obtain historical working data of the air conditioner and determine the normal working range, match the air conditioner operating parameters with the normal working range, and generate an energy-saving control analysis signal if the parameters are within the range, and generate a potential anomaly analysis signal if they are not within the range and transmit them respectively. The potential anomaly analysis unit is used to process potential anomaly analysis signals, generate normal trend curves for each parameter based on historical normal data, judge parameter mutations through multidimensional mutation judgment rules, and combine environmental parameter changes and related parameter synergy verification to classify mutations caused by non-environmental factors and generate mutation analysis signals or anomaly cause analysis signals. The abnormal cause analysis and identification unit is used to process abnormal cause analysis signals, build an abnormal cause knowledge base based on historical abnormal data, determine the cause of the fault through parameter feature comparison, deep learning feature extraction and strong correlation parameter analysis, generate abnormal cause information, and transmit it to the intelligent control information output unit. The intelligent control information output unit is used to display information about the cause of the abnormality to the relevant management personnel.

[0007] As a further aspect of the present invention, the air conditioning parameters include voltage, current, power and air conditioning operating parameters, including compressor start / stop frequency, fan speed and cooling / heating power; Environmental parameters include outdoor environmental parameters and indoor environmental parameters. Outdoor environmental parameters include temperature and humidity, and light intensity. Indoor environmental parameters include actual temperature and humidity, population density, and CO2 concentration.

[0008] As a further aspect of the present invention, the normal trend curve in the potential anomaly analysis unit is generated in the following way: For parameters with stable fluctuations, a 5-10 minute sliding window is used to calculate the mean and generate a smooth trend curve; for parameters with significant dynamic changes, an LSTM model is used to learn historical time series patterns and generate a predictive trend curve.

[0009] As a further aspect of the present invention, the multidimensional mutation determination rule of the potential anomaly analysis unit includes: When the deviation between the real-time value of the parameter and the normal trend curve exceeds the corresponding threshold and lasts for a duration of t, a preliminary mutation marker is triggered. The change rate of environmental parameters in the same period is extracted, and an environmental impact threshold is set. If the matching degree between the parameter mutation amplitude and the environmental change amplitude is ≥80%, it is marked as a normal fluctuation. For initial mutations not driven by the environment, the parameter correlation matrix is ​​called to calculate the coordination deviation. If the coordination deviation is ≤10%, it is judged as possible normal adjustment, and if the coordination deviation is ≥30%, it is confirmed as abnormal mutation of the equipment.

[0010] As a further aspect of the present invention, the anomaly classification standard of the potential anomaly analysis unit is as follows: Minor anomaly: Deviation exceeds the threshold but is ≤30%, duration t-2 minutes, coordination deviation 10%-30%, generating continuous monitoring information; Suspected fault / abnormality: Deviation 30%-50%, duration [t, t+5] minutes, coordination deviation 30%-50%; High-risk anomalies: deviation ≥ 50%, duration ≥ t + 5 minutes, coordination deviation ≥ 50%; suspected fault anomalies or high-risk anomalies generate anomaly cause analysis signals.

[0011] As a further aspect of the present invention, the fault cause determination logic of the anomaly cause analysis and identification unit includes: The abnormal parameter features are compared with the fault modes in the abnormal cause knowledge base to screen the preliminary fault causes; the feature of parameter change curves before and after the anomaly is extracted by deep learning algorithm, and the dynamic evolution law is compared to narrow down the fault range; for cases where the matching degree does not reach the threshold, the change of correlation coefficient of strongly correlated parameters is analyzed, and the fault cause is determined by combining the operation log and maintenance record.

[0012] As a further aspect of the present invention, it also includes an energy-saving control analysis unit for processing energy-saving control analysis signals, specifically including: The system calculates the temperature difference between indoors and outdoors and classifies it into three levels: high, medium, and low. It also calculates the indoor occupancy density and classifies it into three levels: high, medium, and low. The system calculates the rate of change of parameters and determines the three levels of change: sudden change, gradual change, and stable change. Based on the parameter combination characteristics, the system generates energy-saving control information and transmits it to the intelligent control information output unit.

[0013] As a further aspect of the present invention, the method for generating energy-saving control information based on parameter combination features is as follows: When there are high temperature differences, high density, and stable changes, the high-efficiency cooling + balanced air supply mode is activated, and the cooling power is maintained at 80%-90%. When there is a low temperature difference, low density, or gradual change, switch to constant temperature energy-saving mode, reduce the cooling power to below 50%, or start the ventilation mode; When there is a medium temperature difference, a sudden change in density, or a sudden change in temperature, first increase the cooling power to the peak value, and then reduce it to 60%-70% after the temperature approaches the set value. When there are temperature differences, low density, or sudden changes, reduce the cooling power by 40% and simultaneously increase the set temperature by 1-2℃.

[0014] An AI-optimized intelligent air conditioner energy-saving control method, which specifically includes the following steps: Step 1: Collect air conditioning parameters and environmental parameters, and preprocess the parameters to obtain preprocessed parameters; Step 2: Obtain historical operating data of the air conditioner and determine the normal operating range. Match the air conditioner operating parameters with the normal operating range. If the parameters are within the range, generate an energy-saving control analysis signal; otherwise, generate a potential anomaly analysis signal. Step 3: Generate normal trend curves for each parameter based on historical normal data, judge parameter mutations through multidimensional mutation judgment rules, and combine environmental parameter changes and related parameter synergy verification to classify mutations caused by non-environmental factors and generate mutation analysis signals or abnormal cause analysis signals. Step 4: Construct an anomaly cause knowledge base based on historical anomaly data, determine the cause of the failure through parameter feature comparison, deep learning feature extraction and strong correlation parameter analysis, and generate anomaly cause information; Step 5: Process the energy-saving control analysis signal, determine different levels of change based on indoor and outdoor temperature difference, indoor personnel density, and parameter change rate, and generate energy-saving control information based on parameter combination characteristics.

[0015] This invention provides an AI-optimized intelligent air conditioning energy-saving control system and method. Compared with existing technologies, it has the following advantages: This invention avoids the limitations of a single threshold by constructing normal operating ranges for different scenarios. The potential anomaly analysis unit reduces the false judgment rate and accurately distinguishes between environmental fluctuations and equipment anomalies by modeling trend curves, determining environmental changes through multi-dimensional mutations, and verifying related parameters. By constructing an anomaly cause knowledge base and combining deep learning feature extraction and strong correlation parameter analysis, it achieves accurate tracing of fault causes without the need for manual investigation, shortening fault response time and extending equipment lifespan. Based on the three-dimensional combination of indoor and outdoor temperature difference, personnel density, and parameter change rate, it generates scenario-based energy-saving strategies through AI algorithms, achieving dynamic linkage between load adaptation and parameter optimization, reducing system energy consumption while ensuring indoor comfort. Attached Figure Description

[0016] Figure 1 This is a block diagram of the intelligent air conditioning energy-saving control system of the present invention; Figure 2 This is a flowchart of the steps of the intelligent air conditioner energy-saving control method of the present invention; Figure 3 This is a schematic diagram of the overall architecture of the intelligent air conditioning energy-saving control system of the present invention. Detailed Implementation

[0017] 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.

[0018] First Embodiment Please see Figure 1 and Figure 3 This application provides an AI-optimized intelligent air conditioning energy-saving control system, including: an intelligent data acquisition unit, a parameter change analysis unit, an energy-saving control analysis unit, a potential anomaly analysis unit, an anomaly cause analysis and identification unit, and an intelligent control information output unit, and combined with... Figure 1 It can be seen that the information between the above functional units is transmitted in one direction only.

[0019] The intelligent data acquisition unit is used to collect air conditioning parameters and environmental parameters. The air conditioning parameters include voltage, current, power, and air conditioning operating parameters, including compressor start / stop frequency, fan speed, and cooling / heating power. The environmental parameters include outdoor and indoor environmental parameters. Specifically, the outdoor environmental parameters include temperature and humidity, and light intensity, while the indoor environmental parameters include actual temperature and humidity, population density, and CO2 concentration. The obtained parameters are preprocessed to obtain preprocessed parameters. Preprocessing specifically involves removing abnormal sensor data, such as temperature and humidity values ​​that exceed the physical range, and instantaneous fluctuations in electrical data, and then transmitting the data to the parameter change analysis unit.

[0020] The parameter change analysis unit is used to identify and analyze the air conditioner status based on the acquired preprocessed parameters. It obtains historical operating data of the air conditioner, uses the K-means algorithm to cluster the sample set of each scenario, automatically identifies the parameter distribution pattern under the same operating state, initially divides the normal parameter clusters, calculates the 95% confidence interval for the clustered normal parameter clusters, determines the upper and lower thresholds of each parameter, generates the normal operating range, and matches the air conditioner operating parameters with the normal operating range. If the air conditioner operating parameters are within the normal operating range, it indicates that the air conditioner is operating normally, and an energy-saving control analysis signal is generated and transmitted to the energy-saving control analysis unit. Conversely, if the air conditioner operating parameters are not within the normal operating range, it indicates that the air conditioner is operating abnormally, and a potential anomaly analysis signal is generated and transmitted to the potential anomaly analysis unit.

[0021] The potential anomaly analysis unit is used to process the acquired potential anomaly analysis signals, extract the core information from the potential anomaly analysis signals, including the anomaly parameter type, real-time value, corresponding spatiotemporal environmental parameters and data acquisition timestamp, and use the Kalman filter algorithm to remove transient interference noise in the signal and retain effective anomaly features. Based on historical normal operating data of air conditioners, the normal trend curves of each parameter are generated using the moving average method or LSTM model. For parameters with stable fluctuations, the mean is calculated using a 5-10 minute sliding window to generate a smooth normal trend curve. For parameters with significant dynamic changes, the time series variation pattern of historical data is learned to generate a predictive normal trend curve. A dedicated normal trend curve is generated for each core parameter to clarify the reasonable fluctuation range for each time period. Next, multidimensional mutation judgment rules are set. When the deviation between the real-time value of a parameter and the normal trend curve exceeds the corresponding threshold and lasts for a duration of t (where t is set by the operator), a preliminary mutation marker is triggered. Simultaneously, verification is performed by combining changes in environmental parameters and the synergy of related parameters. Specifically, the rate of change of environmental parameters during the same period is extracted, and an environmental impact threshold is set. If the match between the parameter mutation amplitude and the environmental change amplitude is ≥80%, it is marked as a normal fluctuation driven by the environment, and no processing is performed. For preliminary mutations not driven by the environment, cross-validation of strongly correlated parameters is used to confirm whether it is an equipment anomaly. The parameter correlation matrix is ​​called to calculate the synergistic deviation between the mutated parameter and the related parameters. If the synergistic deviation is ≤10%, it may be a normal adjustment; if the synergistic deviation is ≥30%, it is confirmed as an abnormal equipment mutation, and a mutation analysis signal is generated. The generated mutation analysis signals are processed, and anomalies are classified into three levels based on the mutation magnitude, duration, and synergy verification results: A deviation exceeding the threshold but ≤30%, duration t-2 minutes, and synergy deviation 10%-30% are marked as minor anomalies, generating continuous monitoring information and transmitting it to the intelligent control information output unit; a deviation of 30%-50%, duration [t, t+5] minutes, and synergy deviation 30%-50% are marked as suspected fault anomalies; and a deviation ≥50%, duration ≥t+5 minutes, and synergy deviation ≥50% are marked as high-risk anomalies. For suspected fault anomalies or high-risk anomalies, an anomaly cause analysis signal is generated and transmitted to the anomaly cause analysis and identification unit.

[0022] The anomaly cause analysis and identification unit processes the acquired anomaly cause analysis signals and constructs an anomaly cause knowledge base based on historical anomaly data. This knowledge base covers various common fault modes of air conditioning systems and their corresponding characteristic parameters. The unit compares and matches the parameter features in the received anomaly cause analysis signals with the fault modes in the anomaly cause knowledge base to screen preliminary fault causes. Then, it uses deep learning algorithms to extract and analyze the parameter change curves over a period of time before and after the anomaly occurs. By comparing the parameter change features with the dynamic evolution of fault modes in the anomaly cause knowledge base, the range of fault causes is further narrowed. If the matching degree exceeds a preset threshold, the confirmed fault type is directly output, and anomaly cause information is generated. If the matching degree does not reach the threshold, strong correlation parameter analysis is performed. First, strong correlation parameter pairs are identified. For the strong correlation parameters involved in the preliminary fault causes, the change data of these parameters before and after the anomaly occurs are extracted, and the correlation coefficient between them is calculated. If the correlation coefficient between strongly correlated parameters changes before and after the anomaly occurs, and this change is consistent with the change pattern of correlated parameters in the anomaly cause knowledge base, then the fault mode is considered as one of the possible causes of the fault. At the same time, combined with auxiliary information such as the air conditioning system's operation log and maintenance records, the possible causes of the fault are comprehensively evaluated, and the most likely cause of the fault is finally determined. Detailed anomaly cause information is generated and transmitted to the intelligent control information output unit.

[0023] The intelligent control information output unit is used to display the acquired continuous monitoring information and abnormal cause information to the corresponding management personnel.

[0024] Second Embodiment As a second embodiment of the present invention, it is implemented based on the first embodiment, and the difference from the first embodiment is as follows: The parameter change analysis unit transmits the generated energy-saving control analysis signal to the energy-saving control analysis unit, and processes the signal. It collects outdoor temperature data using distributed temperature sensors and calculates the temperature difference between indoor and outdoor temperatures. Based on this difference, it categorizes the temperature difference into high, medium, and low levels. A difference greater than 8°C is considered a high temperature difference; a difference between 3°C and 8°C is considered a medium temperature difference; and a difference less than 3°C is considered a low temperature difference. Secondly, it captures human activity signals using human body sensors and calculates the indoor occupant density, converting it to high / medium / low levels. If the indoor occupant density is greater than 5 people / 10m², the difference is considered low. 2 For high density, if it is 2-5 people / 10m 2 Medium density, less than 5 people / 10m² 2For low density, the intelligent controller calculates the parameter difference between the previous time t1 and the current time in real time. The parameter difference, such as the rate of change of indoor temperature and the rate of change of personnel density, determines the three levels of change: sudden change, gradual change, and stable. For example, if the temperature changes by more than 3°C within 1 minute, it is a sudden change; if it is between 1°C and 3°C, it is a gradual change; and if it is less than 1°C, it is a stable change. Abnormal data is removed in real time. The three parameters are integrated with the air conditioner operation data to generate parameter combination characteristics. At the same time, the air conditioner energy-saving control is performed in response to the parameter combination characteristics to generate energy-saving control information. The specific control method is as follows: When there is a high temperature difference, high density, and stable change, activate the high-efficiency cooling + balanced air supply mode, maintain the cooling power at 80%-90%, and adjust the air outlet direction to cover the entire area; when there is a low temperature difference, low density, and gradual change, switch to constant temperature energy-saving mode, reduce the cooling power to below 50%, or activate the ventilation mode to reduce energy consumption by utilizing outdoor natural cold sources; when there is a medium temperature difference, sudden density change, or sudden temperature change, first quickly increase the cooling power to the peak value, and then reduce it to 60%-70% power after the temperature approaches the set value; when there is a high temperature difference, low density, or sudden change, immediately reduce the cooling power by 40% and simultaneously raise the set temperature by 1-2℃.

[0025] At the same time, energy-saving control information is transmitted to the intelligent control information output unit.

[0026] The intelligent control information output unit is used to display the acquired energy-saving control information to the corresponding management personnel.

[0027] Third Embodiment As a third embodiment of the present invention, the focus is on combining the implementation processes of the first and second embodiments.

[0028] Fourth embodiment Please see Figure 2 This application provides an AI-optimized intelligent air conditioner energy-saving control method, which specifically includes the following steps: Step 1: Collect air conditioning parameters and environmental parameters, and preprocess the parameters to obtain preprocessed parameters. The specific processing method is the same as the processing process of the intelligent data acquisition unit. Step 2: Obtain historical operating data of the air conditioner and determine the normal operating range. Match the air conditioner operating parameters with the normal operating range. If the parameters are within the range, generate an energy-saving control analysis signal; otherwise, generate a potential anomaly analysis signal. The specific processing method is the same as that of the parameter change analysis unit. Step 3: Generate normal trend curves for each parameter based on historical normal data, judge parameter mutations through multidimensional mutation judgment rules, and combine environmental parameter changes and related parameter synergy verification to classify mutations caused by non-environmental factors and generate mutation analysis signals or abnormal cause analysis signals. The specific processing method is the same as the processing process of the potential anomaly analysis unit. Step 4: Construct an anomaly cause knowledge base based on historical anomaly data. Determine the cause of the failure through parameter feature comparison, deep learning feature extraction, and strong correlation parameter analysis, and generate anomaly cause information. The specific processing method is the same as the processing process of the anomaly cause analysis unit. Step 5: Process the energy-saving control analysis signal, determine different levels of change based on indoor and outdoor temperature difference, indoor personnel density, and parameter change rate, and generate energy-saving control information based on parameter combination characteristics. The specific processing method is the same as the processing process of the energy-saving control analysis unit.

[0029] Some of the data in the above formulas are numerical calculations with dimensions removed, and the contents not described in detail in this specification are all prior art known to those skilled in the art.

[0030] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. An AI-optimized intelligent air conditioning energy-saving control system, characterized in that, include: The intelligent data acquisition unit is used to collect air conditioning parameters and environmental parameters, preprocess the parameters to obtain preprocessed parameters, and transmit them to the parameter change analysis unit. The parameter change analysis unit is used to identify the air conditioner status based on the preprocessed parameters, obtain historical working data of the air conditioner and determine the normal working range, match the air conditioner operating parameters with the normal working range, and generate an energy-saving control analysis signal if the parameters are within the range, and generate a potential anomaly analysis signal if they are not within the range and transmit them respectively. The potential anomaly analysis unit is used to process potential anomaly analysis signals, generate normal trend curves for each parameter based on historical normal data, judge parameter mutations through multidimensional mutation judgment rules, and combine environmental parameter changes and related parameter synergy verification to classify mutations caused by non-environmental factors and generate mutation analysis signals or anomaly cause analysis signals. The abnormal cause analysis and identification unit is used to process abnormal cause analysis signals, build an abnormal cause knowledge base based on historical abnormal data, determine the cause of the fault through parameter feature comparison, deep learning feature extraction and strong correlation parameter analysis, generate abnormal cause information, and transmit it to the intelligent control information output unit. The intelligent control information output unit is used to display information about the cause of the abnormality to the relevant management personnel.

2. The AI-optimized intelligent air conditioning energy-saving control system according to claim 1, characterized in that, Air conditioner parameters include voltage, current, power, and air conditioner operating parameters, including compressor start / stop frequency, fan speed, and cooling / heating power. Environmental parameters include outdoor environmental parameters and indoor environmental parameters. Outdoor environmental parameters include temperature and humidity, and light intensity. Indoor environmental parameters include actual temperature and humidity, population density, and CO2 concentration.

3. The AI-optimized intelligent air conditioning energy-saving control system according to claim 1, characterized in that, In the potential anomaly analysis unit, the normal trend curve is generated as follows: For parameters with stable fluctuations, a 5-10 minute sliding window is used to calculate the mean and generate a smooth trend curve; for parameters with significant dynamic changes, an LSTM model is used to learn historical time series patterns and generate a predictive trend curve.

4. The AI-optimized intelligent air conditioning energy-saving control system according to claim 1, characterized in that, The multidimensional mutation determination rules for the potential anomaly analysis unit include: When the deviation between the real-time value of the parameter and the normal trend curve exceeds the corresponding threshold and lasts for a duration of t, a preliminary mutation marker is triggered. The change rate of environmental parameters in the same period is extracted, and an environmental impact threshold is set. If the matching degree between the parameter mutation amplitude and the environmental change amplitude is ≥80%, it is marked as a normal fluctuation. For initial mutations not driven by the environment, the parameter correlation matrix is ​​called to calculate the coordination deviation. If the coordination deviation is ≤10%, it is judged as possible normal adjustment, and if the coordination deviation is ≥30%, it is confirmed as abnormal mutation of the equipment.

5. The AI-optimized intelligent air conditioning energy-saving control system according to claim 1, characterized in that, The anomaly classification criteria for the potential anomaly analysis unit are as follows: Minor anomaly: Deviation exceeds the threshold but is ≤30%, duration t-2 minutes, coordination deviation 10%-30%, generating continuous monitoring information; Suspected fault / abnormality: Deviation 30%-50%, duration [t, t+5] minutes, coordination deviation 30%-50%; High-risk anomalies: deviation ≥ 50%, duration ≥ t + 5 minutes, coordination deviation ≥ 50%; suspected fault anomalies or high-risk anomalies generate anomaly cause analysis signals.

6. The AI-optimized intelligent air conditioning energy-saving control system according to claim 1, characterized in that, The fault cause determination logic of the anomaly cause analysis and identification unit includes: By comparing the abnormal parameter features with the fault modes in the abnormal cause knowledge base, preliminary fault causes are screened. Deep learning algorithms are used to extract the feature curves of parameter changes before and after the anomaly, and the dynamic evolution pattern is compared to narrow down the fault range. For cases where the matching degree does not reach the threshold, the changes in the correlation coefficients of strongly correlated parameters are analyzed, and the fault cause is determined by combining the operation logs and maintenance records.

7. The AI-optimized intelligent air conditioning energy-saving control system according to claim 1, characterized in that, It also includes an energy-saving control analysis unit for processing energy-saving control analysis signals, specifically including: The system calculates the temperature difference between indoors and outdoors and classifies it into three levels: high, medium, and low. It also calculates the indoor occupancy density and classifies it into three levels: high, medium, and low. The system calculates the rate of change of parameters and determines the three levels of change: sudden change, gradual change, and stable change. Based on the parameter combination characteristics, the system generates energy-saving control information and transmits it to the intelligent control information output unit.

8. The AI-optimized intelligent air conditioning energy-saving control system according to claim 7, characterized in that, The method for generating energy-saving control information based on parameter combination features is as follows: When there are high temperature differences, high density, and stable changes, the high-efficiency cooling + balanced air supply mode is activated, and the cooling power is maintained at 80%-90%. When there is a low temperature difference, low density, or gradual change, switch to constant temperature energy-saving mode, reduce the cooling power to below 50%, or start the ventilation mode; When there is a medium temperature difference, a sudden change in density, or a sudden change in temperature, first increase the cooling power to the peak value, and then reduce it to 60%-70% after the temperature approaches the set value. When there are temperature differences, low density, or sudden changes, reduce the cooling power by 40% and simultaneously increase the set temperature by 1-2℃.

9. An AI-optimized intelligent air conditioning energy-saving control method, executed by the intelligent air conditioning energy-saving control system according to any one of claims 1-8, characterized in that, The method specifically includes the following steps: Step 1: Collect air conditioning parameters and environmental parameters, and preprocess the parameters to obtain preprocessed parameters; Step 2: Obtain historical operating data of the air conditioner and determine the normal operating range. Match the air conditioner operating parameters with the normal operating range. If the parameters are within the range, generate an energy-saving control analysis signal; otherwise, generate a potential anomaly analysis signal. Step 3: Generate normal trend curves for each parameter based on historical normal data, judge parameter mutations through multidimensional mutation judgment rules, and combine environmental parameter changes and related parameter synergy verification to classify mutations caused by non-environmental factors and generate mutation analysis signals or abnormal cause analysis signals. Step 4: Construct an anomaly cause knowledge base based on historical anomaly data, determine the cause of the failure through parameter feature comparison, deep learning feature extraction and strong correlation parameter analysis, and generate anomaly cause information; Step 5: Process the energy-saving control analysis signal, determine different levels of change based on indoor and outdoor temperature difference, indoor personnel density, and parameter change rate, and generate energy-saving control information based on parameter combination characteristics.

Images