Intelligent proportioning control method for activated carbon and deodorant in air purification system

Abstract

The invention relates to the technical field of proportioning control, and discloses an intelligent proportioning control method for activated carbon and a deodorant in an air purification system, which comprises the following steps of: dividing the pore size distribution of the activated carbon into three types of micropores, mesopores and macropores, and obtaining pollutant response vectors of various typical pollutants and consumption weight vectors of the pollutants to the three types of pore sizes; constructing an agent mapping knowledge base; environmental pollutants are monitored in real time, a current pollutant response vector is obtained, and an activated carbon dynamic ability portrait is generated by matching a knowledge base; identifying a new pollutant event, acquiring a response vector and an aperture consumption weight of the new pollutant event, performing efficiency prediction calculation with the dynamic capability portrait, executing three-section decision logic according to a prediction result, and outputting dynamic ratio data of the activated carbon and the deodorant; accurate prediction of functional saturation of the activated carbon is realized through multi-dimensional capability characterization, purification efficiency reduction caused by'saturation illusion 'is effectively avoided, and the comprehensive purification efficiency of an air purification system is remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of proportioning control technology, and more specifically, to an intelligent proportioning control method for activated carbon and deodorizing agents in air purification systems. Background Technology

[0002] In the field of air purification technology, activated carbon, with its abundant pore structure, has become a core functional material for removing gaseous pollutants such as volatile organic compounds and odors. The stability and adaptability of its adsorption performance directly determine the long-term operational efficiency of air purification systems. Currently, the industry has developed relevant technical solutions for the rapid assessment of activated carbon adsorption performance and the precise control of pore structure. However, in practical applications, there are still technical gaps in the dynamic loss characterization of the activated carbon adsorption process and the coordinated scheduling of purification resources (activated carbon, deodorizers), making it difficult to cope with complex and ever-changing indoor pollution scenarios.

[0003] For example, Chinese patent application CN118039018A discloses a method for evaluating the adsorption performance of activated carbon for trace organic pollutants in water. This method obtains the activated carbon adsorption performance parameters (key parameters for evaluating adsorption capacity) and linear solvation free energy parameters of organic pollutants, constructs a correlation between the two, and then calculates the adsorption performance parameters based on the linear solvation free energy parameters of the pollutant to be tested. This effectively solves the problems of time-consuming and costly traditional experimental methods in determining adsorption characteristics, and provides a feasible path for rapid evaluation of activated carbon adsorption performance. However, this technology focuses on the static performance evaluation of pollutants in "water" and cannot adapt to the dynamic fluctuations in pollutant concentration and the complex and diverse types of pollutants in the air environment. Furthermore, it cannot capture in real time the non-uniform loss state of activated carbon with different pore sizes (micropores, mesopores, macropores) during long-term adsorption, making it difficult to support the accurate judgment of the dynamic adsorption capacity of activated carbon in air purification systems.

[0004] Chinese patent application CN120846948A discloses a method and monitoring device for activated carbon pore control. Based on a probabilistic prediction model, it dynamically adjusts the spectral scanning frequency, extracts the temporal slope and acceleration using an adaptive threshold to generate a change matrix, identifies abnormal spectra, analyzes unexpected signals, and uses the characteristic peak spacing to infer molecular configuration parameters. It calculates the difference in matching degree between pore size and molecular configuration to achieve pore gradient control, and verifies the control effect through real-time monitoring. This technology primarily targets pore structure optimization during activated carbon preparation or pretreatment, focusing on "static pore control." It cannot adjust the purification strategy based on the real-time loss status of activated carbon during system operation, and lacks the ability to identify "new pollutant events" and the synergistic ratio adjustment logic between deodorizers and activated carbon, making it difficult to meet the needs of dynamic operation of purification systems.

[0005] Existing technologies for assessing and regulating the adsorption capacity of activated carbon are limited to scalar descriptions in a single dimension, failing to effectively address the "saturation illusion" problem caused by non-uniform loss of pore size in practical applications. When activated carbon is exposed to small molecule pollutants (such as formaldehyde) for a long time, the micropores are "selectively" filled and blocked by the pollutant and its polymers, forming "memory saturation." At this point, although the total adsorption capacity of the activated carbon is not exhausted, it has lost its ability to treat new pollutants (such as macromolecular oil fumes) that require different pore sizes for effective adsorption. The system cannot recognize this functional saturation state, leading to a sharp decline in the purification effect on new pollutants, as if the activated carbon has become "selectively blind," unable to meet the purification needs of complex indoor pollution scenarios. Summary of the Invention

[0006] This invention is applicable to indoor environments such as newly renovated houses, kitchens, and offices that are frequently exposed to different types of pollutants. It is particularly effective in complex environments requiring the treatment of multiple pollutants, effectively addressing the problem of decreased purification efficiency caused by activated carbon functional saturation. To overcome the aforementioned deficiencies of existing technologies, this invention provides an intelligent ratio control method for activated carbon and deodorizing agents in air purification systems. It establishes a proxy mapping knowledge base containing pollutant response vectors and pore size consumption weight vectors, generates a dynamic capability profile of activated carbon, and combines new pollutant identification and efficiency prediction with a three-stage decision-making logic to achieve intelligent ratio control of activated carbon and deodorizing agents. This method can accurately capture the non-uniform pore size loss state of activated carbon, effectively solving the problem of "selective blindness" caused by "memory saturation," and ensuring the continuous and stable purification efficiency of the air purification system in complex pollution scenarios.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A smart ratio control method for activated carbon and deodorizing agents in air purification systems includes:

[0009] The pore size distribution of activated carbon is divided into three pore size categories: micropores, mesopores, and macropores. Pollutant response vectors of various typical pollutants and pore size consumption weight vectors of each typical pollutant on the micropores, mesopores, and macropores of activated carbon are obtained. The pollutant response vectors and pore size consumption weight vectors of typical pollutants are paired and stored to construct a proxy mapping knowledge base.

[0010] Define a pore capacity spectrum vector containing the available capacity of three pore size categories, monitor environmental pollutants in real time, obtain the current pollutant response vector, use the current pollutant response vector to query and match in the proxy mapping knowledge base, and generate a dynamic capacity profile of activated carbon that characterizes the current capacity state of activated carbon based on the matching results and the pore capacity spectrum vector.

[0011] Identify new pollutant events and obtain the pollutant response vector and pore size consumption weight vector of the new pollutants in the new pollutant events. Perform performance prediction calculations by combining the pore size consumption weight vector of the new pollutants with the dynamic capability profile of activated carbon. Execute a three-stage decision logic based on the performance prediction calculation results and output dynamic ratio data of activated carbon and deodorizer.

[0012] The method for obtaining the current pollutant response vector includes:

[0013] Using a sensor array that includes a metal oxide semiconductor sensor, an electrochemical sensor, and a photoionization detector, the response values ​​of environmental pollutants on each sensor are collected in real time and synchronously, and the response values ​​are combined to generate the current pollutant response vector.

[0014] The method for obtaining the pore size consumption weight vector of the typical pollutants on the micropores, mesopores, and macropores of activated carbon includes: using brand-new activated carbon to adsorb each typical pollutant to saturation, measuring the specific surface area changes of the micropores, mesopores, and macropores before and after adsorption; calculating the consumption contribution ratio of each typical pollutant to the micropores, mesopores, and macropores based on the specific surface area change data, and performing normalization processing to generate the pore size consumption weight vector of each typical pollutant on the micropores, mesopores, and macropores of activated carbon.

[0015] The method for generating the dynamic capability profile of activated carbon includes:

[0016] Convert the current pollutant response vector into pollutant concentration and calculate the cumulative adsorption of pollutants.

[0017] Based on the matching results, the cumulative adsorption amount of pollutants, and the pore capacity spectral vector, a dynamic capability profile of activated carbon is generated.

[0018] The method for calculating the cumulative adsorption capacity of the pollutants includes:

[0019] Obtain the current fan airflow parameters of the purification system, set a monitoring time window, multiply the pollutant concentration value with the fan airflow parameters and the monitoring time window, and calculate the amount of pollutants adsorbed within the monitoring time window.

[0020] The matching result is the aperture consumption weight vector corresponding to the current pollutant response vector;

[0021] The method for generating a dynamic capacity profile of activated carbon based on the matching results, cumulative adsorption capacity of pollutants, and pore capacity spectral vector includes:

[0022] Based on the cumulative adsorption amount of pollutants and the pore size consumption weight vector corresponding to the current pollutant response vector, an attenuation operation is performed on the pore size capacity spectrum vector to obtain an updated pore size capacity spectrum vector. The updated pore size capacity spectrum vector is then used as a dynamic capability profile of activated carbon.

[0023] The method for identifying new contaminant events includes:

[0024] Establish the normal fluctuation range of each sensor in the sensor array. When the response value of any sensor exceeds the normal fluctuation range and the duration reaches the set duration, or the rate of change of the response value exceeds the set rate of change threshold, it is determined as a new pollutant event.

[0025] The method for calculating the effectiveness prediction by combining the pore size consumption weight vector of new pollutants with the dynamic capacity profile of activated carbon includes:

[0026] Divide each component of the pore capacity spectral vector in the activated carbon dynamic capacity profile by the corresponding component of the pore size consumption weight vector of the new pollutant, and take the minimum value as the predicted adsorption efficiency scalar E. pred .

[0027] The execution method of the three-stage decision logic includes:

[0028] The predicted adsorption efficiency scalar is compared with the preset multi-level efficiency threshold to determine the treatment mode. Based on the determined treatment mode, a three-stage decision logic is executed to output dynamic ratio data of activated carbon and deodorizer.

[0029] The method for determining the processing mode includes:

[0030] Set high performance threshold T high and low-level performance threshold T low , as a multi-level performance threshold;

[0031] When E pred >T high At that time, determine the activated carbon-only treatment mode;

[0032] When E pred <T low At that time, determine the deodorizing agent as the primary treatment mode;

[0033] When T low ≤E pred ≤T high At that time, the synergistic treatment mode of activated carbon and deodorizing agent was determined.

[0034] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0035] This invention classifies activated carbon pores into micropores, mesopores, and macropores. Combining multi-dimensional pollutant characteristic analysis with pore size consumption correlation modeling, it constructs a proxy mapping knowledge base containing response vectors of typical pollutants and pore size consumption weight vectors. This overcomes the limitations of traditional one-dimensional lifetime estimation based on total adsorption capacity or usage time, achieving precise correspondence analysis between pollutant characteristics and activated carbon pore size consumption. This avoids the shortcomings of single-dimensional assessments that cannot reflect non-uniform pore size loss. Relying on real-time pollutant detection and state quantification technology, combined with the knowledge base, it generates a dynamic capability profile of activated carbon, accurately depicting the actual usable state of each pore size. This effectively avoids misjudgments of capability caused by activated carbon's "memory saturation," providing a reliable state basis for subsequent regulation. Furthermore, through new pollutant identification and performance prediction analysis, it links intelligent adjustment and control logic, executes three-stage decision-making, and outputs dynamic ratio data of activated carbon and deodorizing agent. When functional saturation occurs in a specific pore size of activated carbon, targeted compensation scheduling ensures the system's effective treatment of new pollutants, thus solving the problem of activated carbon's "selective blindness" to new pollutants. Through the synergistic linkage of precise pollutant analysis and intelligent system control, a complete technical closed loop of "detection-characterization-decision-adjustment" is formed, which significantly improves the air purification system's adaptability to complex and ever-changing pollution scenarios and ensures the system's continuous and stable purification efficiency for various pollutants during long-term operation. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 The flowchart illustrates the intelligent ratio control method for activated carbon and deodorizing agent in an air purification system provided in this embodiment of the invention.

[0038] Figure 2 This is a schematic diagram illustrating the classification and function of activated carbon pore size according to an embodiment of the present invention;

[0039] Figure 3 A flowchart illustrating the execution method of a three-stage decision logic provided in an embodiment of the present invention;

[0040] Figure 4 This is a functional block diagram of the intelligent ratio control system for activated carbon and deodorizing agent in the air purification system provided in the embodiments of the present invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Example 1

[0043] Please see Figure 1 As shown, this embodiment provides a method for intelligently controlling the ratio of activated carbon to deodorizing agent in an air purification system, including:

[0044] Step S10: Divide the pore size distribution of activated carbon into three pore size categories: micropores, mesopores and macropores. Obtain the pollutant response vectors of various typical pollutants and the pore size consumption weight vectors of each typical pollutant on the micropores, mesopores and macropores of activated carbon. Pair and store the pollutant response vectors and pore size consumption weight vectors of typical pollutants to construct a proxy mapping knowledge base.

[0045] Further, step S10 includes:

[0046] Step S11: Using a sensor array that includes a metal oxide semiconductor sensor, an electrochemical sensor, and a photoionization detector, the response values ​​of environmental pollutants on each sensor are collected in real time and synchronously, and the response values ​​are combined to generate the current pollutant response vector.

[0047] The selection of the sensor array is based on the complementary characteristics of different sensors: metal oxide semiconductor sensors respond to pollutants by changing conductivity due to gas adsorption on the surface of sensitive materials, exhibiting broad-spectrum sensitivity to volatile organic compounds (VOCs); electrochemical sensors utilize the redox reaction between pollutants and electrode surfaces to generate current signals, showing specificity for polar gases such as formaldehyde and ammonia; photoionization detectors generate ion currents by photoionizing pollutant molecules with ultraviolet light, detecting the overall concentration of total volatile organic compounds (TVOCs). These three sensors have fundamentally different detection mechanisms and react differently to pollutants with varying chemical compositions. Combining them creates complementary detection capabilities to cover the chemical characteristics of common indoor pollutants. The sensor array construction requires integrating the three sensors within the airflow channel of the purification system, ensuring that each sensor is exposed to the same pollutant environment and connected to the same data acquisition module for synchronous sampling. The sampling frequency needs to be determined based on the rate of change of pollutant concentration. By simulating abrupt changes in pollutant concentration in a laboratory environment and recording the time it takes for the sensor response to reach a stable value, the sampling period is set to less than one-tenth of this stable time to ensure the capture of the dynamic characteristics of the pollutants. For example, the sampling frequency range can be set to several to tens of times per second. When pollutants appear in the environment, the data acquisition module simultaneously reads the response values ​​of all sensors according to the preset sampling frequency. These response values ​​exist in the form of voltage or current signals, directly reflecting the interaction strength between the pollutants and the sensitive materials of each sensor. This set of response values ​​is arranged in order of sensor type to form an n-dimensional pollutant response vector V=[s1,s2,…,s…]. n ], where n is the number of sensors contained in the sensor array, s1 to s n These represent the response values ​​of each sensor in the sensor array, numbered from 1 to n. This vector form transforms the ambiguous signal from a single sensor into a high-dimensional feature representation. Different pollutants, due to differences in chemical structure and properties, will form unique response patterns on the sensor array, like chemical "fingerprints," thus enabling the differentiation and identification of different pollutants. Existing technologies often use single sensors to detect pollutants, outputting only a single response value, which cannot distinguish the characteristics of different pollutants, making it impossible to establish a correlation between pollutants and activated carbon. The use of sensor arrays, through the combination of multi-dimensional response values, achieves refined extraction of pollutant features, transforming the invisible chemical properties of pollutants into measurable high-dimensional signals. This makes it possible to subsequently establish a mapping relationship between pollutants and pore size consumption, overcoming the limitations of single sensor signals and providing reliable input for accurately inferring the effects of pollutants on activated carbon.

[0048] Step S12: Divide the pore size distribution of activated carbon into three pore size categories: micropores, mesopores, and macropores, and define a pore size capacity spectral vector that includes the available capacity of the three pore size categories.

[0049] The adsorption capacity of activated carbon stems from its abundant pore structure. Different pore sizes exhibit different adsorption mechanisms for pollutants. (See [reference needed]). Figure 2 Activated carbon pore size distribution is classified into three categories according to IUPAC standards: micropores (pore size <2nm), mesopores (pore size 2-50nm), and macropores (pore size >50nm). Micropores mainly adsorb small molecule gaseous pollutants through molecular attraction; mesopores can adsorb medium-sized pollutants and also function as transport channels; macropores mainly serve as transport pathways for pollutant molecules to enter micropores and mesopores, and can also adsorb large molecular clusters of pollutants. This classification method matches the distribution characteristics of pollutant molecule size. Small molecules such as formaldehyde mainly rely on micropore adsorption, medium-sized molecules such as benzene series compounds can be captured by mesopores, and large molecules such as oil fume particles need to be transported through macropores and adsorbed on their surfaces. These three pore size categories basically cover the adsorption needs of common indoor pollutants. Pore size capacity spectral vector C=[c micro ,c meso ,c macro In the middle, c micro c represents the percentage of available capacity of the micropores. meso c represents the percentage of available capacity in the orifice. macro Representing the proportion of usable capacity of macropores, the pore size capacity vector describes the adsorption capacity state of activated carbon through a three-dimensional data structure. For virgin activated carbon, the initial pore size capacity vector is set to C. initial =[1,1,1] indicates that the available capacity of all three pore sizes is at full capacity. This standardized setting provides a unified benchmark for the subsequent quantitative calculation of pore size consumption. Existing technologies typically use a single numerical value to describe the remaining lifetime or total adsorption capacity of activated carbon, which fails to reflect the differences in loss among different pore sizes, leading to an inability to explain the "saturation artifact." The implementation of pore size classification and the definition of the pore capacity spectrum vector transform the complex pore structure of activated carbon into a structured, quantifiable mathematical model, upgrading the single-dimensional capacity description to a multi-dimensional capacity characterization. This makes it possible to track the independent loss of different pore sizes, breaking the limitations of traditional capacity descriptions and providing a basic framework for accurately characterizing the capacity state of activated carbon.

[0050] Step S13: Obtain the pollutant response vectors of various typical pollutants and the pore size consumption weight vectors of each typical pollutant on the micropores, mesopores and macropores of activated carbon. Pair and store the pollutant response vectors and pore size consumption weight vectors of the typical pollutants to construct a proxy mapping knowledge base.

[0051] The method for obtaining the pore size consumption weight vector includes: using brand-new activated carbon to adsorb each typical pollutant to saturation, measuring the changes in the specific surface area of ​​micropores, mesopores, and macropores before and after adsorption; based on the specific surface area change data, calculating the consumption contribution ratio of each typical pollutant to micropores, mesopores, and macropores, and performing normalization processing to generate the pore size consumption weight vector of each typical pollutant to the micropores, mesopores, and macropores of activated carbon.

[0052] The selection of typical pollutants needs to cover common indoor pollution types, including pollutants of different molecular sizes and chemical properties. For example, typical pollutants may include small molecule polar gases, small molecule non-polar gases, medium molecule volatile organic compounds, alkaline gases, and large molecule particulate pollutants, to ensure that the knowledge base can cover the main pollution scenarios in actual applications. The pore size consumption weight vector was obtained through controlled laboratory calibration: First, using virgin activated carbon, a typical pollutant was adsorbed to saturation in an environment isolated from other pollutants. This process ensured that pore size consumption was solely caused by the target pollutant, eliminating interfering factors. Then, the BET nitrogen adsorption method was used to measure the changes in the specific surface area of ​​the activated carbon's micropores, mesopores, and macropores before and after adsorption. BET nitrogen adsorption, a mature technology in materials characterization, can accurately measure the specific surface area and pore volume of different pore size ranges, providing reliable data support for pore size consumption assessment. Based on the specific surface area change data, the pollutant's contribution ratio to the consumption of the three pore sizes was calculated, i.e., the ratio of the reduction in the specific surface area of ​​a certain pore size to the total reduction in the specific surface area of ​​the three pore sizes. This ratio was then normalized so that the sum of the consumption contribution ratios of the three pore sizes was 1. The normalized consumption contribution ratios of the three pore sizes were combined to form the pore size consumption weight vector for the pollutant. Simultaneously, the stable pollutant response vector of the typical pollutant was recorded according to the method in step S11, ensuring that the acquisition conditions of the response vector were consistent with the actual application scenario. The pollutant response vector and corresponding pore size consumption weight vector for each typical pollutant are stored as key-value pairs in a data structure, forming a surrogate mapping knowledge base. This knowledge base is essentially a dictionary relating the chemical characteristics and physical adsorption properties of pollutants, providing prior knowledge for online real-time inference of pollutant pore size consumption preferences. Existing technologies lack a direct correlation between pollutant characteristics and activated carbon pore size consumption, making it impossible to infer the effect of pollutants on activated carbon from detected pollutant signals. The calibration of typical pollutants, the calculation of pore size consumption weight vectors, and the construction of the knowledge base establish a stable mapping relationship between measurable pollutant characteristic signals and indirectly measurable pore size consumption characteristics, making online real-time inference of the impact of pollutants on activated carbon possible.

[0053] Step S10 establishes a cross-dimensional mapping mechanism between pollutant characteristics and activated carbon pore size consumption. Through feature extraction from the sensor array, dimensional division of the pore structure, and the construction of a knowledge base, the invisible physical interaction process is transformed into a measurable and computable model. This solves the problem in existing technologies where the impact of pollutants on activated carbon pore size loss cannot be inferred from pollutant detection signals, making the assessment of activated carbon's capacity no longer dependent on a single total capacity or usage time. It provides a core mapping basis for the subsequent generation of a dynamic capacity profile of activated carbon. Without this mapping relationship, dynamic updating of the pore size capacity spectral vector cannot be achieved, thus making it impossible to accurately predict the functional saturation of activated carbon. The proxy mapping knowledge base is scalable and can be updated by adding calibration data of new pollutants, enabling the system to adapt to scenarios where new pollutants appear and extending the system's technical life cycle. It achieves a precise correlation between features and loss, providing a basis for advance judgment for the intelligent scheduling of subsequent deodorizing agents, and can predict demand before activated carbon exhibits functional shortcomings, avoiding a decline in purification effect.

[0054] Step S20: Define a pore capacity spectrum vector containing the available capacity of three pore size categories; monitor environmental pollutants in real time to obtain the current pollutant response vector; use the current pollutant response vector to query and match in the proxy mapping knowledge base; and generate a dynamic capacity profile of activated carbon that characterizes the current capacity state of activated carbon based on the matching results and the pore capacity spectrum vector.

[0055] Further, step S20 includes:

[0056] Step S21: Convert the current pollutant response vector into pollutant concentration and calculate the cumulative adsorption amount of pollutants;

[0057] Obtain the current fan airflow parameters of the purification system, set a monitoring time window, multiply the pollutant concentration value with the fan airflow parameters and the monitoring time window, and calculate the amount of pollutants adsorbed within the monitoring time window.

[0058] An offline-calibrated sensor response-concentration conversion model is used to convert the real-time acquired current pollutant response vector into the corresponding pollutant concentration. The method for constructing the sensor response-concentration conversion model includes: selecting typical pollutants covered in step S13, setting a series of gradient concentrations in a laboratory environment, recording the stable response vector of the sensor array at each concentration, establishing a mapping relationship between the response vector and the corresponding concentration, and using a multivariate regression algorithm to fit the conversion model. This model can eliminate cross-interference between different sensor responses and achieve accurate mapping from high-dimensional response vectors to a single concentration value. During the conversion process, the real-time acquired current pollutant response vector is input into the model, and the approximate concentration P of the corresponding pollutant is output. i(t) represents the pollutant concentration, where i is the index of the pollutant. The pollutant concentration directly quantifies the content of pollutants in the environment, providing basic data for subsequent adsorption calculations. The fan airflow parameter is acquired in real-time through the fan controller of the purification system. The fan controller generates corresponding operating parameters when adjusting the airflow speed, and the system can directly read these parameters to obtain the current airflow F. Real-time acquisition of the airflow parameter ensures accurate calculation of the airflow volume passing through the activated carbon per unit time, thereby determining the total amount of pollutants carried in the airflow and avoiding calculation deviations caused by using a fixed airflow. This method ensures consistency between the data and the actual operating state, especially when the system automatically adjusts the airflow speed according to the pollutant concentration. The setting of the monitoring time window ∆t must be based on the sensor response characteristics: Simulate a sudden change in pollutant concentration in a laboratory environment, record the time from sensor contact with the pollutant to the stable response value, and set the monitoring time window to less than one-tenth of this stable time. This setting ensures that the dynamic changes in pollutant concentration can be captured, avoiding the masking of instantaneous peak values ​​due to concentration averaging caused by an excessively large window, while also controlling the calculation frequency to avoid excessive computational burden on the system.

[0059] Pollutant cumulative adsorption capacity Q i The calculation formula is: Q i This represents the total amount of pollutants passing through activated carbon within a time window ∆t, i.e., the approximate total amount of pollutants adsorbed by activated carbon during that period. The design logic of this formula closely aligns with the actual scenario of activated carbon adsorption, integrating three key influencing factors—concentration, airflow, and time—into a quantifiable adsorption capacity index, providing a direct basis for subsequent pore size loss calculations. When P... i When (t) increases, Q under the same time and air volume i The increase in F indicates an increase in the total amount of pollutants and a faster consumption of activated carbon pore size; when F increases, even if P... i (t) remains unchanged, Q i The increase also reflects the enhanced impact of pollutants brought about by airflow acceleration. The formula can accurately capture the impact of these changes on pore size loss. Existing technologies often directly estimate activated carbon lifespan based on usage time or total adsorption capacity, failing to distinguish between the concentration differences and adsorption intensities of different pollutants, leading to distorted loss assessments. This step achieves accurate quantification of pollutant content through response vector-concentration conversion. By combining airflow and time parameters to calculate the cumulative adsorption capacity, a quantitative basis for the consumption of activated carbon pore size by pollutants is obtained. This provides accurate input for subsequent calculations of pore size capacity loss, solving the problem of inaccurate capacity assessment caused by fuzzy adsorption capacity estimation in traditional technologies, and providing data support for refined updates of pore size capacity decay.

[0060] Step S22: Use the current pollutant response vector to query and match in the proxy mapping knowledge base. Based on the matching results, the cumulative adsorption amount of pollutants and the pore capacity spectrum vector, generate a dynamic capacity profile of activated carbon that characterizes the current capacity state of activated carbon.

[0061] The matching result is the pore size consumption weight vector corresponding to the current pollutant response vector. Based on the cumulative adsorption amount of pollutants and the pore size consumption weight vector corresponding to the current pollutant response vector, an attenuation operation is performed on the pore size capacity spectrum vector to obtain the updated pore size capacity spectrum vector. The updated pore size capacity spectrum vector is used as the dynamic capacity profile of activated carbon.

[0062] The matching process employs vector similarity calculation: the Euclidean distance between the current pollutant response vector and the pollutant response vectors of all typical pollutants in the proxy mapping knowledge base is calculated. Euclidean distance quantifies the similarity between high-dimensional vectors; a smaller distance indicates closer chemical characteristics. The pore size consumption weight vector corresponding to the pollutant response vector of the typical pollutant with the smallest distance is selected as the matching result. The matched pore size consumption weight vector W... k =[w micro,k ,w meso,k ,w macro,k ], where w micro,k w meso,k w macro,k These correspond to the consumption weights of pollutants on micropores, mesopores, and macropores, respectively, directly linking the relationship between pollutants and pore size loss. k is the index of a typical pollutant. The attenuation calculation of the pore size capacity spectral vector follows the vector subtraction rule, and the update formula is C. new =C old -(Q i / Q max )×W k Its component form is c micro,new =c micro,old -(Q i / Q max )×w micro,k c meso,new =c meso,old -(Q i / Q max )×w meso,k c macro,new =c macro,old -(Q i / Q max )×w macro,k , where C old C is the aperture capacity spectral vector before the update. new For the updated aperture capacity spectral vector, c micro,old c meso,old c macro,oldThese represent the percentage of available capacity for micropores, mesopores, and macropores before the update, respectively. micro,new c meso,new c macro,new These represent the updated percentages of usable capacity for micropores, mesopores, and macropores. Q max The theoretical maximum adsorption capacity of activated carbon was determined through fully saturated adsorption experiments conducted in the laboratory under a pure toluene or n-butanol vapor atmosphere, combined with calculations based on the BET total pore volume. A typical value is 25%–45% of the activated carbon mass. The attenuation calculation formula is designed based on the physical property that "adsorption capacity is directly proportional to pore size consumption," Q. i The total adsorption capacity of pollutants, Q, was quantified. max This represents the upper limit of the total adsorption capacity of activated carbon, (Q) i / Q max The actual adsorption amount is normalized to a dimensionless capacity consumption ratio (between 0 and 1) to ensure consistency with the dimensions of the pore capacity spectral vector; W k This refers to the dimensionless consumption allowance for each pore size. Subtracting this allowance from the available capacity percentage of the corresponding pore size achieves precise decay of pore capacity. During decay, if the calculated available capacity percentage of a certain type of pore size is less than 0, it is automatically corrected to 0, indicating that this type of pore size is completely saturated and can no longer adsorb the corresponding pollutants, ensuring the physical rationality of the vector data. The activated carbon dynamic capacity profile is a continuously updated pore capacity spectrum vector. This vector uses three dimensions to represent the available capacity percentage of micropores, mesopores, and macropores, respectively, transforming the complex pore structure of activated carbon into a structured mathematical representation. Unlike traditional single-value lifetime indicators, this profile can intuitively present the differences in loss among various pore sizes. For example, when the available capacity percentage of micropores is close to zero, even if mesopores and macropores still have high available capacity, it can be determined that the activated carbon has lost its ability to treat pollutants that rely on micropore adsorption, accurately revealing the essence of the "saturation illusion."

[0063] Step S22 establishes a correlation between real-time pollution and pore size consumption characteristics through vector matching, and dynamically updates the pore size capacity state through attenuation calculation. This transforms discrete pollution monitoring events into continuous capacity status tracking, and the generated dynamic capacity profile provides the core basis for efficiency prediction and ratio decision-making in step S30. Without this step, the system cannot know the actual loss of each pore size of activated carbon and can only rely on the total adsorption amount or usage time for judgment, which will inevitably lead to misjudgment of "saturation illusion" and subsequent deodorizing agent scheduling will also lose accuracy. The multivariate calibration of the sensor response-concentration conversion model ensures the accuracy of concentration data and makes the calculation of cumulative adsorption amount more reliable; real-time air volume acquisition and dynamic monitoring window setting further improve the accuracy of adsorption amount calculation, providing high-quality input for loss assessment; Euclidean distance matching ensures the applicability of pore size consumption weights, making loss allocation consistent with actual pollution characteristics; vector attenuation calculation realizes independent quantification of loss of different pore sizes, breaking the limitations of traditional single-dimensional assessment. These technologies work together to enable dynamic capability profiling to accurately reflect the multi-dimensional capability status of activated carbon. This not only solves the problem of inaccurate capability assessment under the "saturation illusion," but also allows the system to detect the loss trend of specific pore sizes in advance. The purification system has shifted from passively responding to pollution to actively sensing its own processing capacity, making it possible to make subsequent forward-looking scheduling possible and promoting the upgrade of the system operation and maintenance mode from "post-event replacement" to "pre-event optimization."

[0064] Step S30: Identify new pollutant events and obtain the pollutant response vector and pore size consumption weight vector of the new pollutants in the new pollutant events. Perform performance prediction calculations by combining the pore size consumption weight vector of the new pollutants with the dynamic capability profile of activated carbon. Execute a three-stage decision logic based on the performance prediction calculation results and output the dynamic ratio data of activated carbon and deodorizer.

[0065] The core of step S30 lies in accurately matching the multi-dimensional capabilities of activated carbon with the adsorption requirements of new pollutants, thereby achieving forward-looking and intelligent scheduling of purification resources and fundamentally avoiding the decline in purification efficiency caused by the "saturation illusion".

[0066] Further, step S30 includes:

[0067] Step S31: Identify new pollutant events, generate pollutant response vectors for new pollutants in new pollutant events, and obtain pore size consumption weight vectors for new pollutants by querying the proxy mapping knowledge base;

[0068] The identification of new pollutant events is achieved through baseline fluctuation monitoring of a sensor array: Response values ​​of the sensor array are continuously collected in an environment without significant pollution to establish baseline values ​​and normal fluctuation ranges for each sensor in the array. These baseline values ​​are updated using a moving average algorithm to ensure adaptation to slow changes in environmental factors such as temperature and humidity. A new pollutant event is identified when the response value of any sensor exceeds the normal fluctuation range for a set duration, or when the rate of change of the response value exceeds a set rate of change threshold. Establishing baseline values ​​requires simulating different temperature and humidity conditions in a laboratory environment, recording stable sensor response values, and determining the normal range covering common environmental fluctuations. The setting of the duration and rate of change thresholds is based on the dynamic characteristics of pollutant diffusion and sensor response. By simulating pollutant scenarios with different release rates, the time and rate of sensor response from baseline to stability are recorded to ensure that the thresholds can capture real pollution events while avoiding false events caused by environmental noise. After a new pollutant event is identified, the sensor array synchronous acquisition process in step S11 is immediately initiated to obtain the pollutant response vector V of the new pollutant. new Subsequently, the proxy mapping knowledge base is invoked, and the Euclidean distance matching method from step S22 is used to find the match with V. new The pollutant response vector of the most similar typical pollutant is used to obtain the corresponding pore size consumption weight vector W. new =[w micro,new ,w meso,new ,w macro,new ], where w micro,new ,w meso,new ,w macro,new These represent the consumption weights of new pollutants on micropores, mesopores, and macropores, respectively. Existing technologies often employ fixed purification modes for newly emerging pollutants, failing to adjust strategies based on pollutant characteristics and activated carbon status, resulting in poor purification effects. This step utilizes an event recognition method combining baseline monitoring and fluctuation thresholds to rapidly detect the emergence of new pollutants, ensuring timely response. It also employs a collaborative approach using sensor arrays and a proxy mapping knowledge base to ensure consistency between new pollutant feature extraction and pore size consumption preference inference, providing accurate input for subsequent performance prediction. This addresses the issues of delayed response and ambiguous characteristic judgment in traditional technologies, achieving rapid correlation between new pollutants and activated carbon capacity status.

[0069] Step S32: Perform efficiency prediction calculations by combining the pore size consumption weight vector of the new pollutant with the dynamic capacity profile of activated carbon to obtain a scalar of predicted adsorption efficiency for quantifying the activated carbon's ability to treat new pollutants.

[0070] Divide each component of the pore capacity spectral vector in the activated carbon dynamic capacity profile by the corresponding component of the pore size consumption weight vector of the new pollutant, and take the minimum value as the predicted adsorption efficiency scalar E. pred .

[0071] The dynamic capacity profile of activated carbon is the updated pore capacity vector C generated in step S20. new =[c micro,new ,c meso,new ,c macro,new ], where c micro,new c meso,new c macro,new These represent the updated usable capacity percentages of micropores, mesopores, and macropores, directly reflecting the remaining adsorption capacity of activated carbon at each pore size; W new Each component in the formula represents the adsorption demand intensity of a new pollutant for the corresponding pore size, i.e., the proportion of that pore size consumed per unit adsorption amount. The scalar value for predicting adsorption efficiency, E, is also included. pred The core formula is E pred =min(c x,new / w x,new ), where min is the minimum value, x is the index of micro, meso, and macro pores, and x iterates through micro, meso, and macro pores. The calculation process only occurs in w. x,new The calculation is performed on dimensions greater than 0, meaning only the pore size dimension actually consumed by new pollutants is considered. The logical relationship between the parameters in the formula is: c x,new The "supply capacity" representing the aperture size, w x,new The ratio of the two represents the "unit requirement" of the new pollutant for that pore size. It characterizes the maximum adsorption ratio of the new pollutant that the pore size can support, i.e., the upper limit of the activated carbon's treatment capacity for the new pollutant at that pore size dimension. Since the overall treatment capacity of activated carbon for new pollutants is limited by the weakest pore size dimension, the minimum value of the ratio of each effective dimension is taken as the predicted adsorption efficiency scalar E. pred This scalar quantifies the maximum feasible capacity of activated carbon to treat new pollutants alone; a higher value indicates a stronger treatment capacity, while a lower value indicates a risk of functional saturation. For example, the calculation process is as follows: if C... new =[0.2,0.8,0.9],W new Given [0.9, 0.1, 0], only the micropore and mesopore dimensions are calculated. The micropore dimension ratio is 0.2 / 0.9 ≈ 0.22, and the mesopore dimension ratio is 0.8 / 0.1 = 8. The minimum value of 0.22 is taken as E. predThis indicates that although the available capacity of mesopores and macropores is sufficient, severe loss in micropores results in extremely weak treatment capacity of activated carbon for this new pollutant, accurately predicting the "saturation illusion." Existing technologies cannot quantify the remaining treatment capacity of activated carbon for specific new pollutants; they can only infer it from actual purification effects, which introduces a lag. This step, through quantitative calculations of "supply-demand" matching, transforms the abstract adsorption capacity into a comparable scalar indicator, achieving a forward-looking prediction of treatment capacity. This solves the problems of lagging treatment capacity assessment and inability to anticipate efficiency decline in traditional technologies, providing a quantitative basis for the precise allocation of subsequent purification resources.

[0072] Step S33: Compare the predicted adsorption efficiency scalar value with the preset multi-level efficiency threshold to determine the processing mode. Based on the determined processing mode, execute the three-stage decision logic and output the dynamic ratio data of activated carbon and deodorizer.

[0073] Further, see Figure 3 Step S33 includes:

[0074] Step S331: Set the high-performance threshold T high and low-level performance threshold T low , as a multi-level performance threshold;

[0075] Step S332, when E pred >T high At that time, determine the activated carbon-only treatment mode;

[0076] Step S333, when E pred <T low At that time, determine the deodorizing agent as the primary treatment mode;

[0077] Step S334, when T low ≤E pred ≤T high At that time, the synergistic treatment mode of activated carbon and deodorizing agent was determined, and the scalar value of the predicted adsorption efficiency was within the threshold range [T]. low ,T high The proportion of activated carbon borne by the relative position within the [ ] is calculated.

[0078] Step S335: Based on the processing mode, generate dynamic ratio data of activated carbon and deodorizing agent.

[0079] Multi-level performance thresholds include the high-level performance threshold T high and low-level performance threshold T low Its setting needs to be achieved through laboratory purification efficiency testing: Select typical pollutants from step S13, and test them at different E... predUnder the corresponding activated carbon conditions, the purification efficiency of the activated carbon for this pollutant is tested, which is the ratio of the difference in pollutant concentration before and after purification to the initial concentration. The minimum Ep when the purification efficiency is stably maintained at a high level (e.g., above 90%) is recorded. pred Value, set this value to T high Record the maximum E when the purification efficiency drops to a low level (e.g., below 50%). pred Value, set this value to T low For example, if the test finds E pred When the value is ≥0.7, the purification efficiency is above 90%. pred When the value is ≤0.3, the purification efficiency is less than 50%, so T is set. high =0.7, T low =0.3. This setting method will set E pred It is directly related to the actual purification effect, ensuring that the threshold has a clear physical meaning and avoiding decision-making bias caused by subjective setting.

[0080] The three-stage design of the decision-making logic is based on the principle of "dynamic adaptation of capabilities and needs": when E pred >T high When the activated carbon is selected as the sole treatment mode, its processing capacity fully meets the needs of the new pollutants. The decision is made to allocate all purification tasks to the activated carbon, while the deodorizing agent remains on standby. Maintaining normal fan speed ensures the adsorption efficiency of the activated carbon, while avoiding ineffective consumption of the deodorizing agent. When E pred <T low In this scenario, the deodorizing agent is determined to be the primary treatment mode. At this point, activated carbon is unable to effectively treat new pollutants due to functional saturation. The decision is to allocate the main weight of the purification task to the deodorizing agent, while reducing the air velocity passing through the activated carbon to reduce further clogging of the already damaged pores by new pollutants and protect the remaining pore capacity. The deodorizing agent's release amount is adjusted through a controllable release device to ensure that the preset purification effect is achieved. This mode solves the problem in existing technologies where activated carbon is "selectively blind" but is still forcibly used. By having the deodorizing agent lead the purification, the overall purification effect of the system is ensured, while protecting the remaining capacity of the activated carbon and extending its overall service life.

[0081] When T low ≤E pred ≤T high When determining the co-processing mode, the formula for calculating the proportion of activated carbon is: α=(E pred -T low ) / (T high -T low In this formula, α represents the proportion of purification work undertaken by activated carbon, and the remaining proportion (1-α) is undertaken by deodorizing agents. The logic of this formula is that the closer α is to T... highThe stronger the activated carbon's processing capacity, the higher its proportion of processing; conversely, the weaker the activated carbon, the higher the proportion of deodorizing agent, achieving a smooth transition between the two purification resources. The scheduling of the deodorizing agent is achieved through the coordinated adjustment of its release rate and the activated carbon's wind speed: the release rate is positively correlated with the deodorizing agent's proportion, and the wind speed is positively correlated with the activated carbon's proportion. The adjustment parameters are determined through laboratory calibration to ensure stable purification effects under different ratios. In existing technologies, activated carbon and deodorizing agents mostly operate in a fixed synergistic mode, unable to be dynamically adjusted according to actual capacity, leading to resource waste or insufficient purification. This step, through multi-level threshold division and proportion calculation formulas, directly achieves precise allocation of purification tasks, enabling the synergy of activated carbon and deodorizing agents to adapt to actual treatment needs. This solves the resource waste and inefficiency problems of traditional fixed modes, providing decision support for the optimized utilization of purification resources.

[0082] In step S30, the new pollutant event identification mechanism ensures the accuracy of "demand information," avoiding scheduling deviations caused by information errors; the efficiency prediction formula accurately quantifies the activated carbon's processing capacity, revealing the essence of the "saturation illusion" and providing an objective basis for decision-making; the three-stage decision-making logic realizes dynamic resource allocation, balancing purification effect and consumable costs. These technical means work together to solve the core problem of inaccurate purification resource scheduling under the "saturation illusion" in existing technologies, enabling the system to proactively avoid purification failures and maintain continuous and stable purification capabilities. Simultaneously, the dynamic consumption of the deodorizing agent reduces unnecessary replacement frequency, further lowering long-term operating costs for users.

[0083] Example 2

[0084] This embodiment, based on Embodiment 1, provides an intelligent ratio control system for activated carbon and deodorizing agents in an air purification system, such as... Figure 4 As shown, it includes:

[0085] The knowledge base construction module is used to classify the pore size distribution of activated carbon into three pore size categories: micropores, mesopores, and macropores. It obtains the pollutant response vectors of various typical pollutants and the pore size consumption weight vectors of each typical pollutant on the micropores, mesopores, and macropores of activated carbon. The pollutant response vectors and pore size consumption weight vectors of typical pollutants are paired and stored to construct a proxy mapping knowledge base.

[0086] Capability profile building module: used to define a pore capacity spectrum vector containing the available capacity of three pore size categories, monitor environmental pollutants in real time to obtain the current pollutant response vector, use the current pollutant response vector to query and match in the proxy mapping knowledge base, and generate a dynamic capability profile of activated carbon that characterizes the current capability state of activated carbon based on the matching results and the pore capacity spectrum vector.

[0087] The ratio output module identifies new pollutant events and obtains the pollutant response vector and pore size consumption weight vector of the new pollutants in the new pollutant events. It performs performance prediction calculations by combining the pore size consumption weight vector of the new pollutants with the dynamic capability profile of activated carbon. Based on the performance prediction calculation results, it executes a three-stage decision logic and outputs the dynamic ratio data of activated carbon and deodorizer.

[0088] Furthermore, in the knowledge base construction module, the method for constructing the proxy mapping knowledge base includes: obtaining the pollutant response vectors of various typical pollutants and the pore size consumption weight vectors of each typical pollutant on the micropores, mesopores, and macropores of activated carbon; pairing and storing the pollutant response vectors and pore size consumption weight vectors of the typical pollutants to construct the proxy mapping knowledge base.

[0089] The method for obtaining the pore size consumption weight vector includes: using brand-new activated carbon to adsorb each typical pollutant to saturation, measuring the specific surface area changes of micropores, mesopores, and macropores before and after adsorption; calculating the consumption contribution ratio of each typical pollutant to micropores, mesopores, and macropores based on the specific surface area change data, and performing normalization processing to generate the pore size consumption weight vector of each typical pollutant to the micropores, mesopores, and macropores of activated carbon.

[0090] Furthermore, in the proportioning output module, the execution method of the three-stage decision logic includes:

[0091] Step S331: Set the high-performance threshold T high and low-level performance threshold T low , as a multi-level performance threshold;

[0092] Step S332, when E pred >T high At that time, determine the activated carbon-only treatment mode;

[0093] Step S333, when E pred <T low At that time, determine the deodorizing agent as the primary treatment mode;

[0094] Step S334, when T low ≤E pred ≤T high At that time, the synergistic treatment mode of activated carbon and deodorizing agent was determined, and the scalar value of the predicted adsorption efficiency was within the threshold range [T]. low ,T high The proportion of activated carbon borne by the relative position within the [ ] is calculated.

[0095] Step S335: Based on the processing mode, generate dynamic ratio data of activated carbon and deodorizing agent.

[0096] The methods and systems of this application may be implemented in many ways. For example, they may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above-described order of steps for the method is for illustrative purposes only, and the steps of the method of this application are not limited to the order specifically described above, unless otherwise specifically stated.

[0097] In addition, the parts of the technical solutions provided in the embodiments of this application that are consistent with the implementation principles of the corresponding technical solutions in the prior art have not been described in detail, so as to avoid excessive elaboration.

[0098] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for intelligently controlling the ratio of activated carbon to deodorizing agent in an air purification system, characterized in that, The method includes: The pore size distribution of activated carbon is divided into three pore size categories: micropores, mesopores, and macropores. Pollutant response vectors of various typical pollutants and pore size consumption weight vectors of each typical pollutant on the micropores, mesopores, and macropores of activated carbon are obtained. The pollutant response vectors and pore size consumption weight vectors of typical pollutants are paired and stored to construct a proxy mapping knowledge base. Define a pore capacity spectrum vector containing the available capacity of three pore size categories, monitor environmental pollutants in real time, obtain the current pollutant response vector, use the current pollutant response vector to query and match in the proxy mapping knowledge base, and generate a dynamic capacity profile of activated carbon that characterizes the current capacity state of activated carbon based on the matching results and the pore capacity spectrum vector. Identify new pollutant events and obtain the pollutant response vector and pore size consumption weight vector of the new pollutants in the new pollutant events. Perform performance prediction calculations by combining the pore size consumption weight vector of the new pollutants with the dynamic capability profile of activated carbon. Execute a three-stage decision logic based on the performance prediction calculation results and output dynamic ratio data of activated carbon and deodorizer.

2. The intelligent ratio control method for activated carbon and deodorizing agent in the air purification system according to claim 1, characterized in that, The method for obtaining the current pollutant response vector includes: Using a sensor array that includes a metal oxide semiconductor sensor, an electrochemical sensor, and a photoionization detector, the response values ​​of environmental pollutants on each sensor are collected in real time and synchronously, and the response values ​​are combined to generate the current pollutant response vector.

3. The intelligent ratio control method for activated carbon and deodorizing agent in the air purification system according to claim 2, characterized in that, The method for obtaining the pore size consumption weight vector of the typical pollutants on the micropores, mesopores, and macropores of activated carbon includes: using brand-new activated carbon to adsorb each typical pollutant to saturation, measuring the specific surface area changes of the micropores, mesopores, and macropores before and after adsorption; calculating the consumption contribution ratio of each typical pollutant to the micropores, mesopores, and macropores based on the specific surface area change data, and performing normalization processing to generate the pore size consumption weight vector of each typical pollutant on the micropores, mesopores, and macropores of activated carbon.

4. The intelligent ratio control method for activated carbon and deodorizing agent in the air purification system according to claim 3, characterized in that, The method for generating the dynamic capability profile of activated carbon includes: Convert the current pollutant response vector into pollutant concentration and calculate the cumulative adsorption of pollutants. Based on the matching results, the cumulative adsorption amount of pollutants, and the pore capacity spectral vector, a dynamic capability profile of activated carbon is generated.

5. The intelligent ratio control method for activated carbon and deodorizing agent in the air purification system according to claim 4, characterized in that, The method for calculating the cumulative adsorption capacity of the pollutants includes: Obtain the current fan airflow parameters of the purification system, set a monitoring time window, multiply the pollutant concentration value with the fan airflow parameters and the monitoring time window, and calculate the amount of pollutants adsorbed within the monitoring time window.

6. The intelligent ratio control method for activated carbon and deodorizing agent in the air purification system according to claim 5, characterized in that, The matching result is the aperture consumption weight vector corresponding to the current pollutant response vector; The method for generating a dynamic capacity profile of activated carbon based on the matching results, cumulative adsorption capacity of pollutants, and pore capacity spectral vector includes: Based on the cumulative adsorption amount of pollutants and the pore size consumption weight vector corresponding to the current pollutant response vector, an attenuation operation is performed on the pore size capacity spectrum vector to obtain an updated pore size capacity spectrum vector. The updated pore size capacity spectrum vector is then used as a dynamic capability profile of activated carbon.

7. The intelligent ratio control method for activated carbon and deodorizing agent in the air purification system according to claim 6, characterized in that, The method for identifying new contaminant events includes: Establish the normal fluctuation range of each sensor in the sensor array. When the response value of any sensor exceeds the normal fluctuation range and the duration reaches the set duration, or the rate of change of the response value exceeds the set rate of change threshold, it is determined as a new pollutant event.

8. The intelligent ratio control method for activated carbon and deodorizing agent in the air purification system according to claim 7, characterized in that, The method for calculating the effectiveness prediction by combining the pore size consumption weight vector of new pollutants with the dynamic capacity profile of activated carbon includes: Divide each component of the pore capacity spectral vector in the activated carbon dynamic capacity profile by the corresponding component of the pore size consumption weight vector of the new pollutant, and take the minimum value as the predicted adsorption efficiency scalar E. pred .

9. The intelligent ratio control method for activated carbon and deodorizing agent in the air purification system according to claim 8, characterized in that, The execution method of the three-stage decision logic includes: The predicted adsorption efficiency scalar is compared with the preset multi-level efficiency threshold to determine the treatment mode. Based on the determined treatment mode, a three-stage decision logic is executed to output dynamic ratio data of activated carbon and deodorizer.

10. The intelligent ratio control method for activated carbon and deodorizing agent in the air purification system according to claim 9, characterized in that, The method for determining the processing mode includes: Set high performance threshold T high and low-level performance threshold T low , as a multi-level performance threshold; When E pred >T high At that time, determine the activated carbon-only treatment mode; When E pred <T low At that time, determine the deodorizing agent as the primary treatment mode; When T low ≤E pred ≤T high At that time, the synergistic treatment mode of activated carbon and deodorizing agent was determined.

Images