Air purification method and system fused with AI intelligent big data analysis

Abstract

The invention relates to an air purification method and system fused with AI intelligent big data analysis, and solves the problem that the purification effect is greatly reduced due to the fact that an air purifier cannot adjust operation parameters in time according to upcoming air quality changes in the operation process. The method comprises the following steps: inputting real-time gas related data and gas related data in a historical past time range into a pre-trained air quality prediction model, and outputting an air quality prediction result in a future preset time range; and the mode set by the user about the air purifier and the air quality prediction result in the future preset time range are input into the air purifier operation parameter optimization model, the optimized air purifier operation parameters are output, and the air purifier is indicated to perform corresponding operation according to the optimized air purifier operation parameters. The intelligent air purifier has the following effects that precise and intelligent air purification is achieved, the indoor air quality is effectively improved, and various requirements are met.

Description

Technical Field

[0001] The present invention relates to the field of air purification technology, and in particular to an air purification method and system integrating AI intelligent big data analysis. Background Art

[0002] As people pay more and more attention to indoor air quality, air purifiers are increasingly used in homes, offices and various indoor environments.

[0003] Most air purifiers on the market currently have basic purification functions and several preset working modes. For example, in the fast purification mode, the fan speed is usually adjusted to a higher level to speed up air circulation and purification; the sleep mode will reduce the fan speed and reduce noise to ensure that a certain air purification effect is maintained without affecting the user's rest.

[0004] Due to the lack of advance prediction of air quality changes, the air purifier cannot adjust its operating parameters in time according to the upcoming air quality changes during operation, which greatly reduces the purification effect. For example, before the arrival of smog weather, if the air purifier cannot increase the purification power in advance, the indoor air quality will deteriorate rapidly. Summary of the invention

[0005] In order to achieve accurate and intelligent air purification, effectively improve indoor air quality, and meet diverse needs, the present application provides an air purification method and system integrating AI intelligent big data analysis.

[0006] In the first aspect, the present application provides an air purification method and system integrating AI intelligent big data analysis, which adopts the following technical solutions: An air purification method integrating AI intelligent big data analysis, comprising: Acquire the mode set by the user for the air purifier and the real-time gas-related data collected by the environmental sensors distributed indoors, and perform data preprocessing on the real-time gas-related data, wherein the mode set by the user for the air purifier includes a fast purification mode, a sleep mode, and a smart mode; Input the real-time gas-related data and the gas-related data within the historical past time range into the pre-trained air quality prediction model, and output the air quality prediction results within the future preset time range; The mode set by the user for the air purifier and the air quality prediction results within a preset time range in the future are input into the air purifier operation parameter optimization model, and the optimized air purifier operation parameters are output, and the air purifier is instructed to perform corresponding operations according to the optimized air purifier operation parameters, wherein the air purifier operation parameters include fan speed parameters, filter activation combination parameters, and filter working intensity parameters.

[0007] By adopting the above technical solutions, firstly, according to the user-set mode and sensor data collection, the current status of indoor air quality can be accurately understood. Through data preprocessing and prediction models, the future trend of air quality changes can be known in advance, providing a basis for early response. Secondly, the operation parameter optimization model can generate optimization parameters according to the mode and prediction results, so that the fan speed, filter activation combination and intensity can be adapted to actual needs, and the air can be efficiently purified in different scenarios (such as fast purification, sleep, etc.), which can not only ensure the purification effect, but also reduce energy consumption and noise, improve user experience and comfort, and realize intelligent and personalized air purification control.

[0008] Optionally, the air quality prediction result within a preset time range and the mode set by the user for the air purifier is input into the air purifier operation parameter optimization model, and the optimized air purifier operation parameters are output, including: Analyze whether to obtain the mode set by the user for the air purifier; If not, maintain the original setting; If yes, then obtain the user's health-related indicator data about the ambient air and perform data preprocessing, wherein the health indicator data includes lung function indicators and the frequency of allergic symptom attacks; Based on the user's health-related indicator data on ambient air, a preset time series analysis model is used to perform trend analysis on different health-related indicators and obtain trend analysis results; The trend analysis results of different health-related indicators and the air quality prediction results within a preset time range in the future are input into the air purifier operating parameter optimization model, and the optimized air purifier operating parameters are output.

[0009] By adopting the above technical solutions, the operating parameters of the air purifier are optimized by combining the user setting mode, air quality prediction results and health-related indicator data, which can more accurately adapt to the environment and human needs. Considering health indicators such as lung function and the frequency of allergic symptoms, the trend is grasped with the help of time series analysis models, so that the adjustment of purifier operating parameters can be more in line with individual health conditions. While ensuring good indoor air quality, it can effectively reduce the potential harm to human health caused by poor air quality. Parameter optimization in intelligent mode can improve purification efficiency, reduce energy consumption and noise, realize personalized and intelligent air purification services, and improve user comfort and health protection level.

[0010] Optionally, the method further includes the following steps in parallel with instructing the air purifier to perform corresponding operations according to the optimized air purifier operating parameters: Obtain the difference between the actual operating parameter data and the theoretical operating parameter data of the optimized air purifier; The difference results are input into the problem diagnosis model, and the problem diagnosis results of the air purifier are output, which include the fault types and probabilities of the air purifier.

[0011] By adopting the above technical solution, the difference between the actual operation and theoretical parameters of the air purifier can be monitored in real time and faults can be diagnosed. Its beneficial effects are obvious. On the one hand, the acquisition and analysis of the difference results help to accurately locate the deviations in the operation of the purifier, and can promptly detect problems such as abnormal fan speed and poor filter operation. On the other hand, by outputting the fault type and probability through the problem diagnosis model, potential faults can be warned in advance, which makes it convenient for maintenance personnel to prepare maintenance plans and tools in advance, reduce troubleshooting time, improve maintenance efficiency, and ensure the continuous and stable operation of the air purifier. This not only improves the level of intelligence of equipment management, but also ensures that indoor air purification work is carried out uninterruptedly, providing users with reliable air quality assurance.

[0012] Optionally, the method further includes steps after outputting the problem diagnosis result of the air purifier, which are as follows: According to the diagnosis results of the air purifier problems, analyze and determine the types and probabilities of the faults involved; Analyze and determine the hardware involved in the air purifier based on the fault types that exceed the preset probability; If the hardware is a filter, the pressure difference before and after the filter, the change in air flow rate, and the actual filtering efficiency of pollutants within a preset time range are monitored and obtained, and input into the filter performance value analysis module to analyze and obtain the current filter performance value, and adjust the theoretical operating parameter data related to the filter according to the ratio of the theoretical filter performance value to the current filter performance value; If the hardware is other hardware, the failure type and probability of the relevant hardware will be sent to the terminal held by the person in charge.

[0013] By adopting the above technical solution, for the filter, by monitoring key parameters to analyze performance values ​​and adjusting theoretical parameters according to the ratio, the filter status can be accurately evaluated and its operation can be optimized, thereby extending the filter life, improving purification efficiency and ensuring stable purification effects. For the rest of the hardware, fault information is pushed to the responsible terminal in a timely manner, which can achieve rapid response and accurate maintenance, reduce equipment downtime, and ensure the continued efficient operation of the air purifier. Overall, this technology improves the intelligence level of fault diagnosis and processing, can take targeted measures according to different hardware conditions, effectively reduce maintenance costs and manpower investment, and continuously provide users with high-quality air purification services, enhancing the reliability and practicality of the product.

[0014] Optionally, the method further includes steps after analyzing and obtaining the current filter performance value and before adjusting the theoretical operating parameter data related to the filter according to the ratio of the theoretical filter performance value to the current filter performance value, which are as follows: Start reverse airflow flushing operation; After completing the reverse airflow flushing operation, re-monitor the pressure difference before and after the filter, the air flow change, and the actual filtration efficiency of pollutants within the preset time range, and input them into the filter performance value analysis module to analyze and obtain the filter performance value as the filter performance value this time.

[0015] By adopting the above technical solutions, this technology has great advantages in maintaining filter performance. Starting the reverse airflow flushing operation can effectively remove some blockages on the filter surface and inside, and restore some of its filtering performance. Re-monitoring the relevant parameters after flushing and analyzing the filter performance values ​​can accurately evaluate the flushing effect and provide a more accurate basis for subsequent adjustment of parameters based on the ratio of theoretical and current performance values. On the one hand, this operation can extend the service life of the filter, reduce the frequency of filter replacement, and save costs. On the other hand, it helps to maintain the stable and efficient purification capacity of the air purifier, ensure that the indoor air quality continues to meet the standards, and improve the user experience. Especially in long-term use or high-pollution environments, it can significantly improve the stability and reliability of the overall operation of the equipment.

[0016] Optionally, according to the ratio of the theoretical filter performance value to the current filter performance value, adjusting the theoretical operating parameter data related to the filter includes: Analyze and determine the parameter adjustment plan based on the interval range and parameter adjustment plan of the ratio of the theoretical filter performance value to the current filter performance value; Adjust the theoretical operating parameter data related to the filter according to the parameter adjustment plan.

[0017] By adopting the above technical solutions, this technology has achieved remarkable results in the filter parameter adjustment link. According to the interval of the ratio of the theoretical and current filter performance values, an accurate parameter adjustment plan is determined to achieve refined and scientific adjustment. According to the actual performance of the filter, the parameters can be fine-tuned when the performance drops slightly, and larger adjustments can be made when the performance drops significantly, so that the filter always operates in a suitable working state. This can not only optimize the air purifier's filtration efficiency for air pollutants and ensure the quality of indoor air purification, but also avoid energy waste or excessive filter loss caused by unreasonable parameters. It improves the intelligent adaptive ability of the air purifier, extends the service life of the filter, reduces the operating cost of the equipment, and provides users with continuous, efficient and energy-saving air purification services.

[0018] Optionally, the method further includes the following steps in parallel with instructing the air purifier to perform corresponding operations according to the optimized air purifier operating parameters: Obtain specific feedback from users about the air purifier and analyze whether the feedback involves purification effects; If yes, then based on the user's feedback on the air purifier, a semantic analysis algorithm is used to determine the purification effect that the user prefers; The purification effect preferred by the user is input into the filter type matching model, and the matching filter type is output, and the matching filter type is sent to the terminal held by the person in charge.

[0019] By adopting the above technical solutions, the user adaptability of air purifiers can be effectively improved. By obtaining user feedback and analyzing whether it involves purification effects, the core concerns of users can be accurately located. The semantic analysis algorithm is used to determine the purification effect preferred by the user, and then the filter type matching model is used to output the adapted filter type. This process realizes the personalized customization of the purification function. It can meet the differentiated needs of different users for the purification of specific pollutants, such as providing a better formaldehyde filtration solution for formaldehyde-sensitive users. It not only improves user satisfaction, but also enables air purifiers to operate efficiently in different environments and user preferences, optimizes resource allocation, avoids poor purification effects or waste of resources due to filter mismatch, and enhances product market competitiveness and user loyalty.

[0020] Optionally, based on user feedback on air purifiers, a semantic analysis algorithm is used to analyze and determine the purification effects preferred by the user, including: Clean the original text of user feedback and perform original text analysis, including syntactic analysis, entity recognition and topic extraction, and semantic role labeling; Based on the original text analysis results, extract the features that reflect the user's preference for purification effects and construct a feature vector; The current user feedback data constructed by the feature vector is input into the trained classification model, and the corresponding purification effect category prediction result is output.

[0021] By adopting the above technical solutions, we have outstanding advantages in understanding user purification needs. Cleaning and analyzing the original text of user feedback can remove interference and accurately extract key information. Syntactic analysis can clarify the text structure and semantic relationship, entity recognition and other aspects can focus on the core content, and semantic role labeling can deepen understanding and lay the foundation for feature extraction. Constructing feature vectors to quantify needs is convenient for model processing. The trained classification model is input to output the purification effect category prediction results, so that the air purifier can be adjusted according to the user's personalized needs. This not only improves the accuracy and efficiency of user opinion analysis, but also can quickly respond to and meet user expectations for purification effects.

[0022] In the second aspect, the present application provides an air purification system integrating AI intelligent big data analysis, which adopts the following technical solutions: An air purification system integrating AI intelligent big data analysis includes a memory, a processor, and a program stored in the memory and executable on the processor. The program can be loaded and executed by the processor to implement the air purification method integrating AI intelligent big data analysis as described in the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a flow chart of an air purification method integrating AI intelligent big data analysis according to an embodiment of the present application.

[0024] Figure 2 Another embodiment of the present application is to input the mode set by the user for the air purifier and the air quality prediction results within a preset time range in the future into the air purifier operating parameter optimization model, and output a flow chart of the optimized air purifier operating parameters.

[0025] Figure 3 It is a flowchart diagram of another embodiment of the present application and the steps of instructing the air purifier to perform corresponding operations in parallel according to the optimized air purifier operating parameters.

[0026] Figure 4 It is a flowchart diagram of the steps after outputting the problem diagnosis result of the air purifier in another embodiment of the present application. DETAILED DESCRIPTION

[0027] The present application is further described in detail below in conjunction with the accompanying drawings.

[0028] Reference Figure 1 , is an air purification method integrating AI intelligent big data analysis disclosed in this application, comprising: Step S100, obtaining the mode set by the user for the air purifier and the real-time gas-related data collected by the environmental sensors distributed in the room, and performing data preprocessing on the real-time gas-related data.

[0029] Among them, the mode set by the air purifier refers to the different working modes selected and set by the user on the air purifier according to his or her own needs, mainly including fast purification mode (in this mode, the air purifier will run at a higher power, aiming to quickly improve the indoor air quality in a short time), sleep mode (in this mode, the air purifier will usually reduce the operating noise, adjust the power appropriately, etc., to ensure that a certain air purification effect is maintained without affecting the user's sleep) and smart mode (the air purifier can automatically adjust the operating parameters according to the data collected by the environmental sensor, and intelligently adapt to changes in indoor air quality to perform purification work).

[0030] Environmental sensor: a device that can detect gas-related parameters in the indoor environment. Common sensors include those that can detect the concentration of harmful gases in the air (such as formaldehyde, benzene, etc.), particulate matter concentration (such as PM2.5, PM10, etc.), temperature and humidity, etc. They are distributed in different locations indoors to comprehensively and accurately obtain environmental information everywhere in the room.

[0031] Real-time gas-related data refers to data related to indoor air gas conditions collected at the current moment by environmental sensors distributed indoors.

[0032] Data preprocessing refers to a series of preliminary processing operations on the collected raw real-time gas-related data, such as data cleaning (removing outliers, duplicate data, etc.), data standardization (unifying data of different magnitudes to a certain standard range for subsequent analysis and processing), and data encoding (converting some classified data into a format that can be recognized and processed by computers, etc.). The purpose is to improve data quality and make it more suitable for subsequent analysis and model input.

[0033] The method for users to obtain the mode set for the air purifier is as follows: users set the corresponding mode through the air purifier's own operation panel (generally located on the surface of the air purifier body, with corresponding buttons, knobs or touch areas for mode selection operations) or the corresponding mobile phone APP (if the air purifier supports the smart interconnection function, users can download the corresponding APP on the mobile phone, log in to the account and pair it with the air purifier, and then they can easily choose to set the working mode of the air purifier on the APP interface) to set the corresponding mode, and the system can record and obtain the mode information selected by the user.

[0034] The real-time gas-related data is obtained as follows: it is collected by environmental sensors installed in advance at different locations in the room. Under normal working conditions, these environmental sensors will monitor the surrounding air conditions in real time, and transmit the collected data to the connected data receiving terminal (which can be the built-in data processing module of the air purifier, or the control terminal for centralized data processing, etc.) through wired (such as connecting to the data acquisition terminal through a network cable) or wireless (commonly using wireless communication protocols such as Bluetooth and Wi-Fi) at set time intervals (such as every few minutes, etc., the specific interval can be determined according to the setting parameters of the sensor) to obtain real-time gas-related data.

[0035] Step S200, input the real-time gas-related data and the gas-related data within the historical past time range into a pre-trained air quality prediction model, and output the air quality prediction result within a future preset time range.

[0036] The air quality prediction model is a model that is built based on artificial intelligence algorithms (such as regression algorithms in machine learning) and pre-trained with a large amount of labeled data (including historical gas-related data and corresponding actual air quality conditions, etc.). Its core function is to predict the air quality conditions in a certain period of time in the future based on the input related gas data and output the corresponding prediction results, such as predicting whether the concentration of harmful gases in indoor air will exceed the standard in a certain period of time in the future, and the approximate range of particulate matter concentration, etc.

[0037] First, the real-time gas-related data obtained in step S100 and pre-processed are sorted together with the gas-related data within the stored historical past time range, and organized according to the input data format required by the air quality prediction model (for example, the data is sorted into a vector or matrix form of a specific dimension, etc.). Then, these organized data are input into the pre-trained air quality prediction model. The model will analyze and process the input data based on the laws and characteristics learned internally, and then output the air quality prediction results within the future preset time range (this preset time range is also pre-set, such as the next half hour, two hours, etc.). These results can provide key basis for further optimizing the operating parameters of the air purifier, so that it can make reasonable operating adjustments in advance and better cope with the upcoming air quality changes.

[0038] Step S300, input the user's set mode for the air purifier and the air quality prediction results within a preset time range in the future into the air purifier operation parameter optimization model, output the optimized air purifier operation parameters, and instruct the air purifier to perform corresponding operations according to the optimized air purifier operation parameters.

[0039] Among them, the air purifier operating parameters mainly include the following: Fan speed parameter: refers to the speed-related parameters of the fan inside the air purifier. The speed of the fan directly affects the flow rate of air in the purifier and the efficiency of air passing through purification components such as filters. The higher the speed, the faster the air circulation and purification speed, but it may also be accompanied by greater noise, etc., and needs to be optimized and adjusted according to actual conditions.

[0040] Filter activation combination parameters: Some air purifiers are equipped with filters of multiple functions or materials (for example, filters specifically for removing formaldehyde, filters for removing particulate matter, etc.). This parameter determines which filters need to be activated in specific situations and how they work together. Through reasonable combination, different air quality issues can be dealt with more accurately.

[0041] Filter working intensity parameters: related to the degree of performance of the filter when working, such as the filter's adsorption capacity, filtration efficiency and other adjustment parameters. Reasonable setting of this parameter can ensure that the filter can effectively purify the air while also extending its service life.

[0042] The specific operation process is as follows: First, collect and organize the mode information set by the user about the air purifier and the air quality prediction results within the future preset time range obtained through step S200, and prepare these two aspects of data according to the input format required by the air purifier operation parameter optimization model (such as specific data structure, data type, etc.). Then, input the prepared data into the air purifier operation parameter optimization model. The model will perform complex analysis, calculation and other processing processes on the input data based on the logic and rules learned in advance by itself, and finally output the optimized air purifier operation parameters. These optimized parameters will be used to instruct the air purifier to operate according to the corresponding requirements, so that the air purifier can adaptively adjust its own operating status according to the current and future air quality conditions, better complete the air purification task, and ensure that the indoor air quality is at a good level.

[0043] Reference Figure 2 , the user's set mode for the air purifier and the air quality prediction results within the future preset time range are input into the air purifier operation parameter optimization model, and the optimized air purifier operation parameters are output including: Step S310, analyzing whether the mode set by the user for the air purifier is obtained. If yes, executing step S320; if no, executing step S330.

[0044] Specifically, in this step, the system will detect and confirm whether the relevant information about the mode set by the user for the air purifier has been successfully obtained. This detection and judgment process can be understood as the system using the pre-set program code logic to check whether there is data of the user-set mode in the corresponding storage area or data interface. For example, if the system searches for valid mode data records in the corresponding fields of the database table that records user operation settings in accordance with the established communication protocol and data storage rules, if the corresponding valid records can be found, it means that the mode set by the user has been obtained, and step S330 will be executed according to the "yes" situation; and if no relevant valid mode data records are found in the specified location after searching, it is determined that the mode set by the user has not been obtained, and step S320 will be executed in this case.

[0045] Step S320, maintaining the original settings.

[0046] Step S330, obtaining the user's health-related indicator data about the ambient air and performing data preprocessing.

[0047] Among them, users' health-related indicator data on ambient air mainly include lung function indicators (such as vital capacity, forced expiratory volume in one second, etc., which can reflect the functional state of the respiratory system, and poor air quality may lead to changes in lung function. For example, long-term exposure to polluted air environment may reduce lung capacity, etc.) and the frequency of allergic symptoms (for example, the frequency of allergic symptoms such as sneezing, runny nose, and itchy skin in specific people in indoor environments, because air quality factors such as the content of allergens in the air can affect the onset of allergic symptoms). These data are of great significance for comprehensively evaluating the impact of air on human health.

[0048] Data preprocessing: Preliminary processing of the health-related indicator data of the user about the ambient air. Similar to the preprocessing of gas-related data mentioned above, it also involves data cleaning (removing those abnormal values ​​that do not conform to the actual situation and are obviously wrong, such as incorrect lung function indicator values ​​entered, etc.), data standardization (standardizing data of different magnitudes and units into a standard range that is convenient for subsequent analysis and comparison, such as converting lung function indicator data obtained by different testing equipment into a unified standard unit and value range) and data coding (if it involves some classified data, convert it into a code form that the computer can better recognize and process, etc.) and other operations. Through these preprocessing operations, the data quality is improved and the accuracy and effectiveness of subsequent analysis are guaranteed.

[0049] Step S340, based on the user's health-related indicator data about ambient air, a preset time series analysis model is used to perform trend analysis on different health-related indicators and obtain trend analysis results.

[0050] Specifically, the preset time series analysis model is a model built on the principles of mathematical statistics and data analysis, and is specifically used to process data sequences arranged in chronological order. It can mine the laws, trends, and periodicity of data changes over time, such as the autoregressive moving average model (ARMA) and the autoregressive integrated moving average model (ARIMA). The model is applied here to analyze the changing trends at different time points based on the user's health-related indicator data on ambient air, so as to provide a basis for optimizing the operating parameters of the air purifier that is more in line with human health needs.

[0051] Trend analysis results refer to the summary of the trend of various indicators over time obtained by processing the health-related indicator data using a time series analysis model. For example, whether the lung function indicators are gradually getting better, worse, or relatively stable, whether the frequency of allergic symptoms is increasing, decreasing, or maintaining a stable level, etc. These results can clearly show the dynamic changes in the relationship between air quality and human health, and help to optimize the air purifier operating parameters in a targeted manner.

[0052] In this step, the pre-processed health-related indicator data will be input into the preset time series analysis model. Specifically, the system will organize and arrange the data according to the input format required by the time series analysis model (for example, arranging the health indicator values ​​corresponding to each time point in chronological order, etc.) to ensure that the data can be accurately identified and processed by the model.

[0053] Then, the time series analysis model will perform operations and analysis on the input data based on its own built-in algorithm logic, such as calculating the rate of change of various health-related indicator data in different time spans, identifying the periodic characteristics in the data, and judging the overall increase or decrease trend. After these complex analysis processes, the trend analysis results corresponding to various health-related indicators are finally output.

[0054] Step S350, input the trend analysis results of different health-related indicators and the air quality prediction results within a preset time range in the future into the air purifier operating parameter optimization model, and output the optimized air purifier operating parameters.

[0055] The air quality prediction result within the preset time range in the future is output in step S200 with the help of the air quality prediction model. It estimates the indoor air quality conditions within a specific time period in the future (such as the next 1 hour, 2 hours, etc., the specific duration is based on the actual setting), such as the numerical range that indicators such as the concentration of harmful gases and particulate matter in the air are expected to reach. It can present the changing trend of air quality in advance and assist in the reasonable adjustment of the air purifier operating parameters.

[0056] The air purifier operating parameter optimization model is based on a specific algorithm (such as the neural network algorithm, support vector machine algorithm, etc. in machine learning, or a fusion of multiple algorithms), and is pre-trained using a large amount of sample data containing trends of various health-related indicators, different air quality predictions, and corresponding ideal air purifier operating parameters.

[0057] In step S350, the trend analysis results of different health-related indicators and the air quality prediction results within the future preset time range are first sorted and summarized. The input data is prepared according to the input format required by the air purifier operation parameter optimization model (for example, the data is sorted into a vector form of a specific dimension to ensure that the type and order of each data meet the model requirements).

[0058] Next, input the prepared data accurately into the air purifier operating parameter optimization model. At this point, the air purifier operating parameter optimization model will conduct comprehensive and complex analysis, calculation, reasoning and other processing operations on the input data based on its own pre-trained rules, logic and algorithm mechanisms. For example, the model may weigh the purification effect and health impact of adjusting different operating parameters (such as fan speed, filter activation combination, filter working intensity, etc.) based on the impact of air quality on human health reflected by the trend of health-related indicators, combined with the trend of air composition changes in future air quality forecast results, and then after a series of internal calculations, finally output the optimized air purifier operating parameters.

[0059] Reference Figure 3 , an air purification method integrating AI intelligent big data analysis also includes the steps of instructing the air purifier to operate accordingly according to the optimized air purifier operating parameters, as follows: Step SA00, obtaining the difference between the actual operating parameter data and the theoretical operating parameter data of the optimized air purifier.

[0060] The actual operating parameter data of the optimized air purifier refers to the relevant parameter data that truly reflects the working status of each operating component of the air purifier during the actual operation according to the optimized operating parameters output in step S300. For example, the actual rotation speed value of the fan, the combination of the actual activation of the filter, and the specific working intensity of the filter when it is actually working. These data reflect the actual operating status of the air purifier.

[0061] Theoretical operating parameter data: that is, the operating parameter settings that the air purifier should theoretically reach based on the optimized operating parameters output by the air purifier operating parameter optimization model in step S300. For example, the theoretical speed that the fan should maintain, the ideal combination of filters that should be enabled, and the corresponding ideal working intensity, etc. It is used as a standard reference to measure whether the actual operation meets expectations.

[0062] Difference result: It is obtained by comparing the actual operating parameter data of the optimized air purifier with the theoretical operating parameter data. It is usually reflected as the difference between the two (such as the difference between the actual fan speed and the theoretical speed, the difference between the actual working intensity of the filter and the theoretical working intensity, etc.) or the difference ratio (the percentage of the actual parameter relative to the theoretical parameter deviation, etc.). This result can intuitively reflect the degree of deviation between the actual operating state of the air purifier and the expected state, and provide a basis for subsequent analysis to determine whether there is a problem.

[0063] In this step, the system will obtain the actual operating parameter data of the optimized air purifier through corresponding monitoring means and data collection mechanisms. These monitoring means may include sensors installed at key components inside the air purifier (for example, speed sensors installed on the fan motor to detect the actual speed, pressure sensors and flow sensors installed around the filter to monitor parameters related to its working status, etc.), and the sensors will collect the corresponding data in real time and transmit it to the data processing module.

[0064] At the same time, the system will call out the previously saved theoretical operating parameter data of the air purifier (the data is output and recorded by the air purifier operating parameter optimization model in step S300).

[0065] Then, using a special data comparison program or algorithm, the actual operating parameter data obtained are compared and analyzed with the theoretical operating parameter data according to the corresponding parameter categories (such as fan speed parameters, filter related parameters, etc.), and the difference or difference ratio between the two is calculated, and finally the difference result is obtained. This result will be input into the problem diagnosis model as a subsequent step.

[0066] Step SB00, input the difference result into the problem diagnosis model, and output the problem diagnosis result of the air purifier, which includes the fault type and probability of the air purifier.

[0067] Among them, the problem diagnosis model is a model built based on machine learning and artificial intelligence algorithms (such as decision tree algorithms, neural network algorithms, etc.), and pre-trained using a large amount of labeled sample data (the sample data contains different parameter differences and the corresponding air purifier's actual fault types, fault probability and other information). Its function is to analyze and evaluate the current operating status of the air purifier based on the input difference results, and then output the problem diagnosis results of the air purifier, helping to quickly and accurately determine whether the air purifier has a fault and the specific situation of the fault.

[0068] The problem diagnosis result of the air purifier refers to the output after analysis and processing by the problem diagnosis model, which is about the current fault type of the air purifier (such as fan failure, which may be motor damage, bearing wear, etc.; filter failure, such as filter blockage, filter damage, etc.; or other possible fault types such as circuit failure, sensor failure, etc.) and the probability of the corresponding fault occurring (expressed in the form of percentage, such as the probability of fan motor damage is 10%, etc.). This result can intuitively present the current health status of the air purifier so that targeted countermeasures can be taken later.

[0069] Reference Figure 4 , an air purification method integrating AI intelligent big data analysis also includes steps after outputting the problem diagnosis result of the air purifier, which are as follows: Step SC00, based on the problem diagnosis result of the air purifier, analyze and determine the fault type and probability involved.

[0070] This is the content output by the problem diagnosis model in step SB00, which includes the fault types and corresponding probabilities of the air purifier. For example, it may be diagnosed that the fan has a fault, the fault type is unstable motor speed, and the probability of occurrence is 30%; or the filter has a clogging fault, and the probability is 20%. This information directly reflects the specific aspects and probability of the current problems of the air purifier, which is an important basis for further analysis and measures.

[0071] Step SD00, analyzing and determining the hardware involved in the air purifier according to the fault type exceeding the preset probability.

[0072] Fault types exceeding the preset probability: refers to the fault types analyzed and determined in step SC00, those fault types whose occurrence probability is higher than the preset value (for example, if the preset probability is 30%, then the fault types with a probability higher than 30% belong to this category). These fault types often require more attention due to their higher probability of occurrence, and are likely to have a substantial impact on the normal operation of the air purifier, and are the key basis for further investigation and determination of the specific hardware involved.

[0073] An air purifier is a device composed of multiple hardware components, including a fan (responsible for driving air flow and causing air to circulate and purify through filters, etc.), filters (used to filter various pollutants in the air, such as particulate matter, harmful gases, etc., and different types of filters have different functions), sensors (such as environmental sensors for detecting air quality, speed sensors for monitoring fan speed, etc., used to collect relevant data to assist equipment operation and regulation), circuit boards (play a key role in controlling the power supply of various components and coordinating the coordinated work of various parts), and structural components such as the casing. The hardware involved in the air purifier mentioned here refers to the specific hardware components that are associated with the failure situation and may have problems.

[0074] In this step, the system will conduct further analysis on those fault types that exceed the preset probability, with the aim of accurately identifying the hardware involved in the corresponding air purifier.

[0075] For example, if analysis determines that the probability of the "unstable fan speed" fault type exceeds the preset probability, the system will search for the corresponding hardware that may be involved based on the pre-established knowledge base of fault types and hardware associations (this knowledge base records the hardware problems that usually cause various common fault types, such as unstable fan speed may be related to hardware-related problems such as fan motor failure, aging or loosening of the drive belt, and damage to the fan blades). In other words, it determines that the fan and its related motors, belts, blades and other components may be the hardware where the problem occurs.

[0076] For another example, if the probability of diagnosing the fault type of "reduced filter efficiency" is high, then according to the knowledge base information, the hardware involved may be the filter itself. Perhaps the filter is clogged, damaged, or adsorbed saturated, affecting its normal filtering efficiency.

[0077] By analyzing and determining the hardware involved based on the fault types that exceed the preset probability, the specific hardware components in the air purifier that may fail can be accurately located, thereby providing a clear target for subsequent targeted repair, replacement or debugging measures, ensuring that the air purifier can return to normal operation as soon as possible and effectively perform its air purification function.

[0078] Step SE00, if the hardware is a filter, the pressure difference before and after the filter, the change in air flow and the actual filtering efficiency of pollutants within a preset time range are monitored and input into the filter performance value analysis module, the current filter performance value is analyzed and obtained, and according to the ratio of the theoretical filter performance value to the current filter performance value, the theoretical operating parameter data related to the filter is adjusted.

[0079] Among the many hardware components of an air purifier, the filter is the component that performs the key function of filtering out pollutants in the air. Common types include primary filters, high-efficiency filters, activated carbon filters, etc., which filter and intercept particulate matter of different sizes and harmful gases to ensure that the air passing through the air purifier can be purified. When it is determined that the hardware involved in the fault is the filter, specific monitoring and handling measures need to be taken according to its characteristics and problems.

[0080] Within the preset time range: This is a specific pre-set time interval, such as 10 minutes, 30 minutes or 1 hour, during which the filter-related parameters are monitored to obtain a set of relatively stable and representative data that can reflect the current performance status of the filter, so as to facilitate subsequent accurate analysis of the filter performance.

[0081] The pressure difference before and after the filter refers to the pressure difference between the air before it enters the filter and after it passes through the filter. Under normal circumstances, when the filter is not blocked and the performance is good, the pressure difference is in a reasonable range. When the filter is blocked, the resistance to air passage increases and the pressure difference increases accordingly. Therefore, it is an important indicator to measure whether the filter is working properly and its performance.

[0082] Air flow change refers to the change in the volume of air flowing through the filter per unit time. If the filter is normal, the air flow is relatively stable and meets the equipment design standards. Once the filter fails, such as being blocked or damaged, the air flow will change abnormally, either becoming smaller (when blocked) or possibly unstable. It is also one of the key parameters for evaluating filter performance.

[0083] The actual filtration efficiency for pollutants represents the proportion of various pollutants in the air (such as particulate matter, harmful gases, etc.) that can be effectively filtered and removed when the filter is actually working. It is calculated by comparing the changes in pollutant content before the air enters the filter and after passing through the filter. Its value can intuitively reflect the actual purification capacity of the filter at present, and it is one of the core indicators for judging the performance of the filter.

[0084] The filter performance value analysis module is a functional module specially used to analyze and process data related to filter performance. It receives data such as the pressure difference before and after the filter, air flow changes, and the actual filtration efficiency of pollutants, and then performs comprehensive calculations and analyses on these data based on built-in algorithms, calculation formulas, and related performance evaluation standards. It finally outputs the current filter performance value to quantitatively display the current working status and performance level of the filter.

[0085] The ratio of the theoretical filter performance value to the current filter performance value: The theoretical filter performance value is the performance value under ideal conditions determined by factors such as the filter's design specifications and initial performance, while the current filter performance value is the actual performance value obtained through the above-mentioned monitoring and analysis. The ratio of the two can clearly reflect the degree of deviation of the actual filter performance compared with the theoretical performance. For example, a ratio close to 1 indicates that the filter performance is good and close to the theoretical performance, while a ratio far less than 1 indicates that the filter performance has declined significantly. This ratio result is an important basis for subsequent adjustment of the filter-related theoretical operating parameter data.

[0086] First, the corresponding monitoring equipment will be activated to monitor the pressure difference before and after the filter, the change in air flow, and the actual filtration efficiency of pollutants within a preset time range. These monitoring equipment usually include pressure sensors installed at different positions before and after the filter (for measuring pressure difference), air flow sensors (for monitoring air flow changes), and related detection instruments for detecting the content of air pollutants (to calculate the actual filtration efficiency of pollutants), etc. These sensors and instruments will continuously collect data within a preset time range and transmit the collected data to the data processing unit.

[0087] Then, the obtained data such as the pressure difference before and after the filter, the change in air flow, and the actual filtration efficiency of pollutants are sorted according to the input format required by the filter performance value analysis module. Generally, the accuracy and completeness of the data must be ensured, and it must comply with the corresponding unit and data type requirements.

[0088] Next, the sorted data is input into the filter performance value analysis module, which will perform comprehensive calculation and analysis on the input data based on built-in algorithms and performance evaluation standards. For example, the current filter performance value is calculated according to a specific mathematical model based on multiple data such as pressure difference, air flow and filtration efficiency.

[0089] Finally, the theoretical operating parameter data related to the filter is adjusted according to the ratio of the theoretical filter performance value to the current filter performance value. The ratio of the two will be calculated first, and then the specific parameter adjustment plan will be determined according to the pre-set rules that different ratio ranges correspond to different parameter adjustment plans (for example, when the ratio is between 0.8-1, the plan of fine-tuning the filter working intensity parameter is adopted; when the ratio is less than 0.8, in addition to adjusting the working intensity parameter, a more comprehensive adjustment plan such as replacing the filter may be required), and then the theoretical operating parameter data related to the filter will be adjusted accordingly according to the plan, so that the filter can better play its air purification function in the future and ensure the overall operation effect of the air purifier.

[0090] Step SF00: If the hardware is other hardware, the failure type and probability of the relevant hardware will be sent to the terminal held by the person in charge.

[0091] In the composition of air purifiers, in addition to filters, there are many other important hardware components, such as fans (responsible for driving the air to flow inside the purifier, which is the key power source for air circulation), sensors (including environmental sensors for detecting air quality, sensors for monitoring the operating status of fan speed, etc., used to collect various data to assist equipment operation and regulation), circuit boards (play a core role in controlling the power supply of various components, coordinating the collaborative work of various hardware, and executing various control instructions), and structural components such as shells. The "other hardware" mentioned here refers to these related hardware components, except for the filters, which are determined to have the possibility of failure and the probability exceeds the preset value through step SD00 analysis.

[0092] An air purification method integrating AI intelligent big data analysis also includes steps after analyzing and obtaining the current filter performance value and before adjusting the theoretical operating parameter data related to the filter according to the ratio result of the theoretical filter performance value and the current filter performance value, which are as follows: Step Sa00, start the reverse airflow flushing operation.

[0093] Reverse airflow flushing is a maintenance operation for air purifier filters. Normally, air enters the air purifier in a specific forward flow direction, enters from the air inlet, passes through the filter and other purification components in sequence, and is discharged from the air outlet. The reverse airflow flushing operation changes the direction of the airflow, so that the airflow flows from the original outlet direction to the air inlet direction, and the airflow impacts the filter in the opposite direction. Its purpose is to use the impact force of the airflow to disperse and remove part of the dust, particulate matter and other pollutants attached to or blocked on the surface or inside of the filter, thereby restoring some of the filtering performance of the filter to a certain extent, extending the service life of the filter, and improving its purification effect.

[0094] First, the air purifier system will start the relevant devices or modules specifically used to control the reverse airflow (these devices are generally composed of control circuits, valves, etc. After the control circuit receives the instruction to start the reverse airflow flushing operation, it will drive the corresponding valve to change the opening and closing state, so as to adjust the airflow channel and achieve the reversal of the airflow direction).

[0095] Next, the system will perform the flushing operation according to the pre-set reverse airflow parameters. These parameters include the flow rate of the reverse airflow (the flow rate needs to be controlled within an appropriate range to ensure that the pollutants in the filter can be effectively dispersed without damaging the filter due to excessive flow rate. For example, the flow rate can be set to a certain proportion of the normal intake air flow rate, such as about 1.2 times, etc.), flushing duration (determined based on factors such as the type of filter, area size, and the degree of pollution accumulated during normal times. For example, a small filter takes about 30 seconds to flush, and a large filter may take 1-2 minutes, etc.), etc.

[0096] During the entire reverse airflow flushing operation, the airflow will impact the filter in the reverse direction according to the set parameters, so that some blocked particles and other pollutants on the surface and inside of the filter will be loosened and separated from the filter under the action of the airflow, and discharged from the air purifier with the reverse airflow or collected in a specific dust collection device (if equipped), so as to prepare for the subsequent re-monitoring of the filter performance and further adjustment of the filter-related parameters, so as to better maintain the good performance of the filter and ensure the overall air purification efficiency and quality of the air purifier.

[0097] Step Sb00, after completing the reverse airflow flushing operation, re-monitor and obtain the pressure difference before and after the filter, the air flow change and the actual filtration efficiency of pollutants within the preset time range, and input them into the filter performance value analysis module to analyze and obtain the filter performance value as the filter performance value this time.

[0098] First, the corresponding monitoring equipment will be activated to monitor the pressure difference before and after the filter, the change in air flow, and the actual filtration efficiency of pollutants within the preset time range. These monitoring equipment include pressure sensors installed at different positions before and after the filter (for real-time measurement of pressure difference), air flow sensors (continuous monitoring of air flow), and professional testing instruments for detecting the content of air pollutants (in order to accurately calculate the actual filtration efficiency of pollutants), etc. They will continuously collect data at set time intervals and transmit the collected data to the data processing unit.

[0099] Then, the acquired data on the pressure difference before and after the filter, the change in air flow, and the actual filtration efficiency of pollutants are sorted according to the input format required by the filter performance value analysis module, such as unifying the data units, ensuring the accuracy and completeness of the data, and arranging them in the prescribed order, so that they can be accurately identified and received by the filter performance value analysis module.

[0100] Then, the sorted data is input into the filter performance value analysis module, which will perform comprehensive analysis and calculation on the input data according to the pre-set algorithm and performance evaluation standard. For example, it may calculate a value that can represent the current filter performance status based on multiple dimensions of data such as pressure difference, air flow and filtration efficiency through a specific mathematical model. This value is the filter performance value obtained after the reverse airflow flushing operation.

[0101] Finally, this filter performance value will be used for further analysis and processing in subsequent steps, such as comparison with the theoretical filter performance value, so as to determine the actual recovery of the filter performance, and provide an accurate data basis for adjusting the theoretical operating parameter data related to the filter according to the ratio of the theoretical filter performance value to the current filter performance value, to ensure that the air purifier filter can maintain a good operating state and purification effect.

[0102] According to the ratio of the theoretical filter performance value to the current filter performance value, the theoretical operating parameter data related to adjusting the filter include: Step SE10: Analyze and determine the parameter adjustment plan according to the interval range in which the ratio of the theoretical filter performance value to the current filter performance value falls and the parameter adjustment plan.

[0103] The ratio of the theoretical filter performance value to the current filter performance value: The theoretical filter performance value is determined based on the initial design standard of the filter, the ideal working state and other factors, and represents the performance level that the filter should achieve. The current filter performance value is obtained through actual monitoring and analysis (such as the detection and calculation of the filter-related parameters in the previous steps), and reflects the actual performance status of the filter at the moment. The ratio of the two can intuitively show the degree of deviation of the actual performance of the filter compared to the theoretical performance. For example, a ratio of 1 means that the actual performance is consistent with the theoretical performance, and a ratio close to 1 means that the actual performance is good and close to the theoretical performance. If the ratio is much less than 1, it means that the actual performance of the filter has declined significantly. This ratio result is the key basis for subsequent judgment on how to adjust the filter-related parameters.

[0104] The parameter adjustment plan is a set of pre-defined specific methods for adjusting the theoretical operating parameter data of the filter for different ratio ranges. For example, for different ratio ranges, it may involve making different degrees of adjustments to the working intensity parameters of the filter (such as the parameters related to the adsorption and filtering capacity of the filter), the activation time parameters (the time setting for the filter to start working within a certain period), etc., such as increasing the working intensity, appropriately extending the activation time, or reducing or shortening it accordingly. The purpose is to make the filter play a better role and restore or optimize its air purification effect based on the comparison between the actual performance of the filter and the theoretical performance.

[0105] The interval range and parameter adjustment plan divides the ratio of the theoretical filter performance value to the current filter performance value into multiple different numerical intervals, and each interval corresponds to a specific parameter adjustment plan. For example, the ratio between 0.8-1 is one interval, and the corresponding parameter adjustment plan may be to only make fine adjustments to the filter working intensity parameter; the ratio between 0.6-0.8 is another interval, and the corresponding adjustment plan may be to moderately increase the filter working intensity parameter while shortening its single activation time, etc. Through such a clear correspondence, it can be accurately determined according to the actual ratio results How to adjust the filter-related parameters.

[0106] In step SE10, firstly, the ratio result of the theoretical filter performance value and the current filter performance value is obtained. This ratio result is usually the data obtained by the previous steps (such as the relevant operations involving calculating the ratio of the two).

[0107] Then, according to the pre-set corresponding rules for the interval range of the ratio result and the parameter adjustment scheme, the obtained ratio result is judged to belong to the interval range. For example, if the ratio result is 0.75, it is necessary to check the set interval range to determine which specific interval it falls into (assuming that the rule corresponding to the 0.6-0.8 interval is to moderately increase the filter working intensity parameter and shorten the single activation time, and 0.75 is exactly in this interval).

[0108] Finally, according to the interval range to which the determined ratio result belongs, a specific parameter adjustment plan is determined accordingly, that is, it is clear which theoretical operating parameter data related to the filter should be adjusted to what extent, so as to provide an accurate basis for the actual execution of these parameter adjustment operations in the subsequent step SE20, so that the operating parameters of the filter can be reasonably optimized according to its actual performance, thereby ensuring that the air purifier can achieve good air purification function through the filter.

[0109] Step SE20, adjusting the theoretical operating parameter data related to the filter according to the parameter adjustment scheme.

[0110] After entering step SE20, the parameter adjustment solution determined in step SE10 is first extracted. For example, if the parameter adjustment solution determined is to appropriately increase the filter working intensity parameter and appropriately extend the single activation time of the filter, then the specific parameters to be adjusted and the general direction and amplitude of the adjustment are clarified.

[0111] Then, according to the requirements of the parameter adjustment plan, the control system of the air purifier will perform actual adjustments to the theoretical operating parameter data related to the corresponding filter. For the filter working intensity parameter, the control system will send instructions to the corresponding module responsible for regulating the filter working intensity (such as the electronic control unit that controls the filter adsorption and filtering function), so that it can change the corresponding setting value according to the established adjustment range (such as increasing the working intensity from the original 50% to 60%, etc., the specific value is determined according to the actual plan). For the filter activation time parameter, instructions will also be sent to the timing control-related modules to adjust their timing rules to achieve an appropriate extension of the single activation time of the filter (for example, from the original 1 hour to 1.2 hours, etc.).

[0112] During the entire adjustment process, the control system will ensure the accuracy and effectiveness of the adjustment of each parameter, and after the adjustment is completed, the adjusted parameters will be recorded and saved so that the subsequent air purifier can operate according to the new filter-related theoretical operating parameter data, so that the filter can work under a parameter configuration that is more in line with its actual performance, thereby ensuring that the overall air purification effect of the air purifier is optimized and improved, and maintaining good indoor air quality.

[0113] An air purification method integrating AI intelligent big data analysis also includes the steps of instructing the air purifier to operate accordingly according to the optimized air purifier operating parameters, as follows: Step S400, obtaining specific opinions from users regarding the air purifier, and analyzing whether the opinions involve purification effects.

[0114] The specific opinions about air purifiers reported by users refer to the various views, comments and suggestions put forward by users based on their actual experience and feelings when using air purifiers. These opinions can cover many aspects, such as the feeling of the purification effect of the air purifier (whether the purification speed is fast or slow, whether the purified air is fresh, etc.), whether the noise level generated during operation is acceptable, the convenience of operation (such as whether the operation panel is easy to use, whether the mobile phone APP control function is easy to use, etc.), whether the appearance design is satisfactory, etc., which are important sources of information reflecting the actual use of air purifiers and user needs.

[0115] Purification effect is a key indicator to measure the working efficiency of air purifiers, which is mainly reflected in the ability to remove various pollutants in indoor air. Specifically, it includes the ability to adsorb and decompose harmful gases (such as formaldehyde, benzene, sulfur dioxide, etc.), so that the concentration of harmful gases in the room is reduced to a safe and comfortable range; the filtering and interception efficiency of particulate matter (such as PM2.5, PM10, etc.), which can effectively reduce the tiny particles suspended in the air and make the air cleaner; and the degree of elimination of odors (such as smoke from smoking, kitchen fumes, etc.), which comprehensively reflect the actual results of air purifiers in improving indoor air quality.

[0116] Step S500: If yes, based on the user's feedback on the air purifier, a semantic analysis algorithm is used to analyze and determine the purification effect preferred by the user.

[0117] User feedback on air purifiers refers to various evaluations, opinions, suggestions, etc. given by users based on their actual experience of using air purifiers. The scope covered is relatively broad, and may involve multiple dimensions such as the purification effect, operating noise, ease of operation, and appearance design of the air purifier.

[0118] The semantic analysis algorithm is an algorithm based on natural language processing technology, which aims to understand the true meaning and intention expressed in the text. It mines deep semantic information from the text opinions fed back by users through a series of operations such as lexical analysis (identifying the word part, structure, etc.), syntactic analysis (analyzing the grammatical structure of the sentence and determining the relationship between the various components), semantic role labeling (clarifying the semantic roles played by each entity in the text, such as who is the executor of the action and who is the recipient), and semantic understanding (integrating the results of the previous analyses to infer the semantic connotation of the entire sentence), and then accurately analyzes the purification effect-related content that users prefer.

[0119] The purification effect preferred by users refers to the specific effect that users express through their feedback and expect the air purifier to achieve in the process of air purification. For example, users may hope that the air purifier can quickly reduce the indoor formaldehyde concentration, or focus on improving the filtration efficiency of PM2.5 particles, or pay more attention to eliminating indoor odors, etc. These different tendencies reflect the user's personalized needs for the optimization direction of air purifier functions, and are an important reference for subsequent operations such as filter type matching.

[0120] First, the user feedback text content related to the purification effect will be extracted and input into the semantic analysis algorithm as the data to be analyzed. For example, if the user feedback "I feel that the air purifier is a bit slow in removing formaldehyde, I hope it can reduce the formaldehyde content faster", this text will be sent to the semantic analysis algorithm for processing.

[0121] Next, the semantic analysis algorithm begins to perform multi-faceted analysis operations on the input text. First, lexical analysis is performed to identify nouns (such as "air purifier" and "formaldehyde"), verbs (such as "remove" and "reduce"), adjectives (such as "slow" and "fast"), etc. Then, syntactic analysis is performed to determine the grammatical structure of the sentence, such as the subject, predicate, and object, and to clarify that "air purifier" is the executor of the action, "formaldehyde" is the object of the action, and "reduce" is the main action. After that, semantic role annotation is performed to further clarify the role of each word in the semantic scenario of expressing the purification effect. Finally, through semantic understanding, the results of the previous analysis steps are integrated to infer that the user's core intention is to hope that the air purifier can improve the purification speed of formaldehyde, that is, it is more inclined to improve the purification effect of harmful gas formaldehyde.

[0122] After such a processing flow based on semantic analysis algorithm, the purification effect preferred by the user is finally determined.

[0123] Step S600: input the purification effect preferred by the user into the filter type matching model, output the matching filter type, and send the matching filter type to the terminal held by the person in charge.

[0124] The filter type matching model is a model built based on machine learning algorithms (such as decision trees, neural networks, etc.) or professional air purification knowledge rules. It is pre-trained or rule-set with a large number of data samples containing different purification effect requirements and corresponding filter types (for example, it is known that for the need to quickly reduce indoor formaldehyde concentration, activated carbon filters are more effective; for efficient interception of PM2.5 particles, high-efficiency HEPA filters are more suitable, etc.). Its main function is to analyze and output the filter type that best matches the user's preferred purification effect based on the input, and provide scientific reference suggestions for the optimization of the air purifier's filter configuration.

[0125] The matching filter type refers to the specific filter type that can best meet the user's preferred purification effect requirements after analysis by the filter type matching model. Common ones include primary filters (mainly used to filter large particles of dust, hair, etc., for preliminary filtration of the air), medium-efficiency filters (can filter medium-sized particles, and improve the degree of air purification to a certain extent), high-efficiency HEPA filters (have high filtration efficiency for tiny particles such as PM2.5), activated carbon filters (good at adsorbing harmful gases and odors in the air) and photocatalyst filters (can decompose harmful gases under light conditions), etc. Different filter types have their own expertise in purifying the air. The output matching filter type is intended to enable the air purifier to specifically improve the corresponding purification effect and better meet user needs.

[0126] Step S700: If no, maintain the original setting.

[0127] Based on user feedback on air purifiers, a semantic analysis algorithm is used to determine the purification effects that users prefer, including: Step S510, cleaning the original text of the user feedback and performing original text analysis, wherein the original text analysis includes syntactic analysis, entity recognition and topic extraction, and semantic role labeling.

[0128] The original text of user feedback is the initial, unprocessed text content input or expressed by the user when giving feedback to relevant parties (such as through the feedback function of the air purifier, customer service channels, etc.) based on their actual experience of using the air purifier. It may contain various natural language expressions, such as smooth or not smooth evaluations, detailed or brief descriptions of user experience, etc. Texts that directly reflect user ideas, such as "I have used this air purifier for a while, and I feel that the odor removal effect is not very good. I hope it can be improved" and "I think the speed of this purifier to purify PM2.5 is a bit slow", are the basic materials for a series of subsequent analysis and processing.

[0129] Raw text analysis is the process of deeply exploring the internal language structure and semantic information of the cleaned user feedback text. It mainly covers specific operations such as syntactic analysis, entity recognition and topic extraction, and semantic role labeling. Through these analyses, we can more comprehensively and accurately grasp the core meaning and true intention of the user through the text. It is an important link in understanding user feedback and determining their inclination towards purification effect.

[0130] Syntactic analysis is a technology in natural language processing. It mainly analyzes the structure of sentences based on grammatical rules and determines the various components in the sentence (such as subject, predicate, object, attributive, adverbial, complement, etc.) and their relationships. For example, for the sentence "the purification speed of the air purifier is very slow", through syntactic analysis, it can be made clear that "air purifier" is the subject, indicating the object being described; "purification speed" is the attributive of the subject, modifying a certain attribute of the object; "very slow" is the predicate, describing the state of the attribute, so as to clarify the grammatical structure of the sentence and pave the way for the subsequent understanding of semantics.

[0131] Entity recognition and topic extraction: Entity recognition refers to finding entity objects with specific meanings from text. In the user feedback text about air purifiers, "air purifier", "formaldehyde", "PM2.5" and so on are all entities. Topic extraction is to summarize the core topic content around the text. For example, if the text focuses on the purification speed of air purifiers, then "purification speed" is the extracted topic. Through this operation, we can focus on the key information of the text and grasp the core concerns of user feedback.

[0132] Semantic role labeling is also a means of natural language processing, which aims to clarify the semantic role of each entity in the text in the event or situation described by the sentence. For example, in the sentence "The air purifier quickly reduces the formaldehyde concentration in the room", "air purifier" acts as the executor of the action, "formaldehyde concentration" is the object of the action, "reduce" is the action itself, and "quickly" is the modification of the action. Through such labeling, we can have a deeper understanding of the semantic relationship expressed by the text and assist in judging the user's intention.

[0133] In step S510, data cleaning is first performed on the original text fed back by the user. The system uses the corresponding text processing algorithms and rules to scan and filter the original text and remove the parts that do not meet the standards. For example, by identifying consecutive spaces and merging them into one space, correcting obvious typos (using the built-in common typos correction dictionary, etc.), deleting those repeated, redundant, and meaningless words, etc., the text becomes clean and tidy, meeting the basic requirements of subsequent analysis.

[0134] Next, the cleaned text is analyzed for the original text, starting with syntactic analysis. The system will use a syntactic analysis algorithm (such as a syntactic analysis algorithm based on a statistical model or a deep learning model) to analyze the structure of each sentence based on a pre-set natural language grammar rule system (such as the Chinese grammar rule set), identify the different components in the sentence and the grammatical relationship between them, such as determining the subject, predicate, and object structure, and construct a visual representation such as a grammatical tree diagram of the sentence (if necessary) to clearly present the structure of the sentence.

[0135] Then, entity recognition and topic extraction operations are performed. Entity recognition algorithms (which may be based on named entity recognition models in machine learning, etc.) are used to find meaningful entity objects in the text, such as air purifiers and names of various pollutants. The core topics surrounding the entire text are summarized through topic extraction algorithms (such as those based on word frequency statistics, text clustering, etc.). For example, the topic content about a certain aspect of the purification effect (such as the purification ability of a certain gas, the filtering speed of particulate matter, etc.) is extracted to extract these key information.

[0136] Finally, semantic role labeling is carried out. With the help of a semantic role labeling algorithm (which can also be built based on deep learning and other related technologies), according to the existing syntactic analysis structure and entity recognition results, the semantic role of each entity in the text in the situation it describes is determined, and the relationship between who is the subject of the behavior, who is the object of the behavior, and the way of behavior is clarified. This will prepare for the subsequent extraction of features that reflect the user's preference for purification effects based on the original text analysis results and the construction of feature vectors, thereby more accurately analyzing the purification effect expected by the user.

[0137] Step S520: based on the original text analysis result, extract features reflecting the user's preference for purification effect and construct a feature vector.

[0138] The original text analysis result is the result obtained after a series of operations such as data cleaning, syntactic analysis, entity recognition and topic extraction, and semantic role labeling on the original text of user feedback in step S510. It contains detailed information on text structure, key entities, core topics, and semantic roles of each entity. For example, it clarifies the grammatical relationship between each component in the sentence, identifies key entities such as "air purifier", "formaldehyde", "purification efficiency" and their roles in the context described in the text, etc. These results are the basic basis for the subsequent construction of feature vectors, which can help to deeply explore the real tendency of purification effect reflected in user feedback.

[0139] A feature vector is a vector form that mathematically represents features related to purification effects in user feedback. The key features extracted from the original text analysis results are quantified or encoded and combined into vectors in a certain order. For example, if the degree of concern for the purification effects of different pollutants is set as a feature, numbers may be used to represent different degrees of concern (0 means no concern, 1 means general concern, and 2 means high concern), and then the values ​​corresponding to these features are arranged in a specific order to form a vector form such as [concern for formaldehyde purification, concern for PM2.5 purification, concern for odor purification...], which is convenient for subsequent analysis and processing using mathematical models, and intuitively shows the user's inclination towards various aspects of purification effects.

[0140] Feature extraction is an operation based on the original text analysis results, which aims to filter and extract key feature information that can directly reflect the user's preference for purification effect from complex analysis results. For example, from the analysis result of the sentence "I hope the air purifier can remove formaldehyde faster", the feature "concern for formaldehyde purification efficiency" is extracted, and many such key features related to purification effect are found to prepare for the construction of feature vectors, ensuring that the feature vectors can accurately reflect the focus of the user's expected purification effect.

[0141] First, feature extraction is performed based on the original text analysis results obtained in step S510. The staff or the corresponding system module will carefully sort out these analysis results, and find the corresponding key features from the text analysis results according to the pre-set feature classification standards related to the purification effect (for example, divided into several major categories such as purification of different harmful gases, purification of various types of particulate matter, and elimination of odors). For example, if the original text analysis results show that the user has repeatedly mentioned the expectation of reducing the indoor formaldehyde concentration, the feature of "highly concerned about the formaldehyde purification efficiency" is extracted; if it is mentioned that the air purifier is expected to better filter PM2.5 particles, then the feature of "paying more attention to the PM2.5 purification effect" is extracted, and all such features that can reflect the user's purification effect tendency are screened out one by one.

[0142] Next, these extracted key features are quantified or encoded so that they can be represented in the form of appropriate numerical values ​​or codes. For example, for the feature of "high attention to formaldehyde purification efficiency", the value "2" can be set to represent it; "paying more attention to PM2.5 purification effect" can be represented by "1", etc. The specific quantification or encoding rules will be determined in advance according to the actual application scenarios and analysis requirements to ensure that different features can be presented in a unified and distinguishable way.

[0143] Then, these quantified or encoded features are combined in a predetermined order to construct a feature vector. For example, assuming that the order of the feature vector is the concern for formaldehyde purification, the concern for PM2.5 purification, the concern for odor purification, etc., if the corresponding feature values ​​extracted and quantified above are "2", "1", and "0", respectively, then the constructed feature vector is [2,1,0]. Of course, the actual feature vector dimension and the features included will be more complex and comprehensive according to the specific situation. This feature vector fully reflects the user's preference for various aspects of the purification effect, and provides key data support for the subsequent input into relevant models (such as semantic analysis models, etc.) for further analysis, so as to accurately determine the purification effect that the user prefers.

[0144] Step S530, input the current user feedback data constructed by the feature vector into the trained classification model, and output the corresponding purification effect category prediction result.

[0145] First, the feature vector data constructed in step S520 must be obtained to ensure that its data format is accurate and complete and can be normally recognized and received by the semantic analysis model.

[0146] Then, the feature vector is sorted and input according to the input format required by the semantic analysis model. For example, the semantic analysis model may require data to be presented in a specific data structure (such as a two-dimensional array, etc.) or a specific data type (such as a floating-point value, etc.). In this case, the feature vector needs to be converted and adjusted accordingly to meet the input requirements of the model.

[0147] Next, the sorted feature vectors are input into the semantic analysis model, and the model will conduct in-depth analysis and processing on the input feature vectors based on the algorithm logic learned in advance through its own training, the mapping relationship between the internally constructed features and the purification effect tendency, etc. For example, the model may use multi-layer calculations of neural networks, branch judgments of decision trees, etc., and comprehensively weigh and analyze the degree of attention of users to different purification effects represented by each eigenvalue in the feature vector, to determine which specific purification effect dimension the user most expects the air purifier to focus on, and then output the analysis results of the purification effect preferred by the corresponding user.

[0148] Finally, the output of the purification effect preferred by the user will be used in subsequent steps (such as filter type matching, etc.), providing a key basis for further optimizing the configuration of the air purifier to make it more in line with user expectations and improve user satisfaction, ensuring that the air purifier can be improved and adjusted in the direction of meeting the user's personalized needs.

[0149] Based on the same inventive concept, an embodiment of the present invention provides an air purification system integrating AI intelligent big data analysis, including a memory and a processor, wherein the memory stores data that can be executed on the processor to implement the following Figures 1 to 4 Procedure for either method.

[0150] The embodiments of this specific implementation method are all preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. Therefore, all equivalent changes made based on the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. An air purification method integrating AI intelligent big data analysis, characterized in that: include: Acquire the mode set by the user for the air purifier and the real-time gas-related data collected by the environmental sensors distributed indoors, and perform data preprocessing on the real-time gas-related data, wherein the mode set by the user for the air purifier includes a fast purification mode, a sleep mode, and a smart mode; Input the real-time gas-related data and the gas-related data within the historical past time range into the pre-trained air quality prediction model, and output the air quality prediction results within the future preset time range; The mode set by the user for the air purifier and the air quality prediction results within a preset time range in the future are input into the air purifier operation parameter optimization model, and the optimized air purifier operation parameters are output, and the air purifier is instructed to perform corresponding operations according to the optimized air purifier operation parameters, wherein the air purifier operation parameters include fan speed parameters, filter activation combination parameters, and filter working intensity parameters.

2. The air purification method integrating AI intelligent big data analysis according to claim 1 is characterized in that: The user's set mode for the air purifier and the air quality prediction results within the future preset time range are input into the air purifier operation parameter optimization model, and the optimized air purifier operation parameters are output including: Analyze whether to obtain the mode set by the user for the air purifier; If not, maintain the original setting; If yes, then obtain the user's health-related indicator data about the ambient air and perform data preprocessing, wherein the health indicator data includes lung function indicators and the frequency of allergic symptom attacks; Based on the user's health-related indicator data on ambient air, a preset time series analysis model is used to perform trend analysis on different health-related indicators and obtain trend analysis results; The trend analysis results of different health-related indicators and the air quality prediction results within a preset time range in the future are input into the air purifier operating parameter optimization model, and the optimized air purifier operating parameters are output.

3. The air purification method integrating AI intelligent big data analysis according to claim 2 is characterized in that: The steps of instructing the air purifier to operate accordingly according to the optimized air purifier operating parameters are also included, which are as follows: Obtain the difference between the actual operating parameter data and the theoretical operating parameter data of the optimized air purifier; The difference results are input into the problem diagnosis model, and the problem diagnosis results of the air purifier are output, which include the fault types and probabilities of the air purifier.

4. The air purification method integrating AI intelligent big data analysis according to claim 3 is characterized in that: The steps after outputting the problem diagnosis result of the air purifier are also included, which are as follows: According to the diagnosis results of the air purifier problems, analyze and determine the types and probabilities of the faults involved; Analyze and determine the hardware involved in the air purifier based on the fault types that exceed the preset probability; If the hardware is a filter, the pressure difference before and after the filter, the change in air flow rate, and the actual filtering efficiency of pollutants within a preset time range are monitored and obtained, and input into the filter performance value analysis module to analyze and obtain the current filter performance value, and adjust the theoretical operating parameter data related to the filter according to the ratio of the theoretical filter performance value to the current filter performance value; If the hardware is other hardware, the failure type and probability of the relevant hardware will be sent to the terminal held by the person in charge.

5. The air purification method integrating AI intelligent big data analysis according to claim 4 is characterized in that: The method further includes steps after analyzing and obtaining the current filter performance value and before adjusting the theoretical operating parameter data related to the filter according to the ratio of the theoretical filter performance value to the current filter performance value, which are as follows: Start reverse airflow flushing operation; After completing the reverse airflow flushing operation, re-monitor the pressure difference before and after the filter, the air flow change, and the actual filtration efficiency of pollutants within the preset time range, and input them into the filter performance value analysis module to analyze and obtain the filter performance value as the filter performance value this time.

6. The air purification method integrating AI intelligent big data analysis according to claim 5 is characterized in that: According to the ratio of the theoretical filter performance value to the current filter performance value, the theoretical operating parameter data related to adjusting the filter include: Analyze and determine the parameter adjustment plan based on the interval range and parameter adjustment plan of the ratio of the theoretical filter performance value to the current filter performance value; Adjust the theoretical operating parameter data related to the filter according to the parameter adjustment plan.

7. The air purification method integrating AI intelligent big data analysis according to claim 6 is characterized in that: The steps of instructing the air purifier to operate accordingly according to the optimized air purifier operating parameters are also included, which are as follows: Obtain specific feedback from users about the air purifier and analyze whether the feedback involves purification effects; If yes, then based on the user's feedback on the air purifier, a semantic analysis algorithm is used to determine the purification effect that the user prefers; The purification effect preferred by the user is input into the filter type matching model, and the matching filter type is output, and the matching filter type is sent to the terminal held by the person in charge.

8. The air purification method integrating AI intelligent big data analysis according to claim 7 is characterized in that: Based on user feedback on air purifiers, a semantic analysis algorithm is used to determine the purification effects that users prefer, including: Clean the original text of user feedback and perform original text analysis, including syntactic analysis, entity recognition and topic extraction, and semantic role labeling; Based on the original text analysis results, extract the features that reflect the user's preference for purification effects and construct a feature vector; The current user feedback data constructed by the feature vector is input into the trained classification model, and the corresponding purification effect category prediction result is output.

9. An air purification system integrating AI intelligent big data analysis, characterized in that: It includes a memory, a processor, and a program stored in the memory and executable on the processor, wherein the program can be loaded and executed by the processor to implement an air purification method integrating AI intelligent big data analysis as described in any one of claims 1 to 8.

Images