Data processing method and system for aerosol analyzer

Abstract

The invention provides a data processing method and system for an aerosol analyzer, and belongs to the technical field of data processing of aerosol analyzers, and the method comprises the following steps: collecting light absorption signals of aerosol particles according to a light source system and a detector, and generating an original electric signal; preprocessing the original electric signal to obtain preprocessed data; extracting an absorption coefficient, a multiband light attenuation characteristic and a mass concentration parameter of the black carbon aerosol from the preprocessed data; analyzing the features based on a machine learning model, outputting a real-time prediction result of the concentration of the black carbon aerosol, and preprocessing the original electric signal to obtain preprocessed data, including the step of noise removal. According to the method, baseline drift is dynamically corrected through a polynomial fitting algorithm, wavelet transform denoising and a machine learning model are combined, the accuracy and real-time performance of black carbon aerosol concentration detection are remarkably improved, and the problems that in a traditional method, manual intervention is much, and the model generalization ability is poor are solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of data processing of aerosol analyzers, and in particular relates to a data processing method and system for aerosol analyzers. Background Art

[0002] Black carbon aerosol is a significant light-absorbing particulate matter in the atmosphere, primarily derived from the incomplete combustion of fossil fuels and biomass. Its strong light-absorbing properties significantly impact regional climate, air quality, and human health. Aerosol analyzers are devices specifically designed to monitor and analyze black carbon aerosol particles in the air. They can measure characteristic parameters such as black carbon aerosol concentration and particle size distribution, helping us understand air pollution and its impact on environmental quality and human health. To better process aerosol analyzer data, specialized aerosol analyzer data processing methods and systems are currently commonly used.

[0003] Existing aerosol analyzers usually use traditional signal processing methods and statistical analysis models for data processing. Although they can process data, they suffer from serious noise interference problems. The influence of stray light, temperature fluctuations, circuit noise, etc. in the environment will lead to a low signal-to-noise ratio of the collected signal. Traditional filtering algorithms (such as mean filtering) often have difficulty in effectively separating noise from valid signals, and baseline drift correction relies too much on manual intervention, requiring regular calibration or setting fixed thresholds, and cannot automatically adapt to the dynamically changing detection environment. Summary of the Invention

[0004] In view of this, the present invention provides a data processing method and system for an aerosol analyzer, which can dynamically correct baseline drift through a polynomial fitting algorithm and combine wavelet transform denoising and machine learning models, significantly improving the accuracy and real-time performance of black carbon aerosol concentration detection, and solving the problems of multiple manual interventions and poor model generalization ability in traditional methods.

[0005] To solve the above-mentioned technical problems, in a first aspect, the present invention provides a data processing method for an aerosol analyzer, comprising the following steps: collecting light absorption signals of aerosol particles according to a light source system and a detector to generate an original electrical signal; obtaining preprocessed data by preprocessing the original electrical signal; extracting the absorption coefficient, multi-band light attenuation characteristics, and mass concentration parameters of black carbon aerosol from the preprocessed data; and analyzing the characteristics based on a machine learning model to output a real-time prediction result of the black carbon aerosol concentration.

[0006] The beneficial effects are: dynamically correcting baseline drift through a polynomial fitting algorithm, combined with wavelet transform denoising and machine learning models, significantly improving the accuracy and real-time performance of black carbon aerosol concentration detection, and solving the problems of frequent manual intervention and poor model generalization in traditional methods.

[0007] Optionally, preprocessing the original electrical signal to obtain preprocessed data includes the steps of: noise removal, which uses a wavelet transform algorithm, and the calculation formula is:

[0008]

[0009] Among them, S(t) is the denoised signal, W j (t) is the wavelet coefficient of layer j, V J (t) is the remaining low-frequency component. The noise component is removed by threshold processing and the denoised signal is reconstructed; baseline correction is performed; and spectrum smoothing is performed.

[0010] Optionally, the baseline correction uses a polynomial fitting algorithm, and the calculation formula is:

[0011]

[0012] Where B(t) is the signal after baseline correction, a k is the fitting coefficient, n is the polynomial order; the signal is baseline corrected according to the fitting results.

[0013] The beneficial effects are: the wavelet transform algorithm can effectively separate the noise component and the effective signal in the signal. Compared with traditional methods such as mean filtering, it significantly improves the signal-to-noise ratio and enhances the signal quality, thereby improving the accuracy of subsequent data analysis.

[0014] Optionally, the method for extracting the absorption coefficient includes: calculating the nonlinear relationship between multi-band light attenuation and black carbon mass concentration based on the Lambert-Beer law; obtaining the light attenuation value of each wavelength using the light absorption signals of the main light source 880nm and the auxiliary bands 370nm, 470nm, 520nm, 590nm, 660nm, and 950nm; and calculating the equivalent absorption coefficient of black carbon aerosol through the absorption enhancement model based on the wavelength dependence characteristics.

[0015] The beneficial effect is that the absorption coefficient can be extracted more accurately, providing an important basis for pollution source analysis and health risk assessment.

[0016] Optionally, the machine learning model is a convolutional neural network or a support vector machine, and its training method includes: using a historical black carbon aerosol dataset, which contains 7-band light attenuation data, absorption coefficient time series, environmental temperature and humidity parameters and corresponding concentration labels; normalizing the dataset and dividing it into a training set and a validation set; optimizing the model parameters through a back-propagation algorithm until the prediction error converges to a preset threshold.

[0017] The beneficial effects are: being able to capture the complex nonlinear relationships of aerosol characteristics, greatly improving real-time prediction accuracy and model generalization capabilities, and solving the problem that traditional statistical methods are difficult to deal with complex data patterns.

[0018] Optionally, data storage and output are also included, including the steps of: synchronously storing the original data and analysis structure to a local SD card or cloud platform; and displaying the black carbon aerosol concentration and particle size distribution curve through a mobile terminal.

[0019] The beneficial effects are: it not only facilitates data backup and long-term storage, but also supports multi-terminal access and real-time monitoring, improving the convenience and reliability of data management.

[0020] Optionally, the cloud platform uses an edge computing architecture to aggregate and analyze data from multiple devices to generate regional aerosol pollution maps.

[0021] The beneficial effects are: achieving the upgrade from single-point monitoring to regional joint prevention and control, providing scientific decision-making support for air pollution control.

[0022] In a second aspect, the present application provides a data processing system for an aerosol analyzer, comprising a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a data processing method for an aerosol analyzer according to any one of claims 1-8 is implemented.

[0023] The beneficial effects of the above technical solution of the present invention are as follows:

[0024] 1. Dynamically correcting baseline drift through a polynomial fitting algorithm, combined with wavelet transform denoising and machine learning models, significantly improves the accuracy and real-time performance of black carbon aerosol concentration detection, solving the problems of excessive manual intervention and poor model generalization in traditional methods.

[0025] 2. Supports dual storage modes of local and cloud platforms to ensure data security and accessibility. At the same time, it generates regional pollution maps through edge computing to provide macro-decision-making support for environmental monitoring.

[0026] 3. Real-time display and report generation functions simplify the operation process. Users can obtain key parameters instantly through mobile terminals, improving the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 The present invention is a flow chart of a data processing method for an aerosol analyzer. DETAILED DESCRIPTION

[0028] To make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the following will be combined with the appended drawings of the embodiments of the present invention. Figure 1, clearly and completely describing the technical solutions of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the described embodiments of the present invention, all other embodiments derived by ordinary technicians in this field fall within the scope of protection of the present invention.

[0029] This embodiment provides a remote monitoring method for an aerosol analyzer based on the Internet of Things, as shown in Figure 1, including the following steps:

[0030] S1: Collects light absorption signals of aerosol particles based on the light source system and detector to generate raw electrical signals;

[0031] The light source system uses a multi-band light source array with 7 bands of 370nm, 470nm, 520nm, 590nm, 660nm, 880nm, and 950nm. Based on the Lambert-Beer law, it collects the absorption spectrum signal of black carbon aerosol in real time and generates the original electrical signal.

[0032] S2: Preprocessing the original electrical signal to obtain preprocessed data;

[0033] The preprocessed data is obtained by preprocessing the original electrical signal, including the steps of noise removal. The noise removal adopts the wavelet transform algorithm, and the calculation formula is:

[0034]

[0035] Among them, S(t) is the denoised signal, W j (t) is the wavelet coefficient of layer j, V J (t) is the remaining low-frequency component. The noise component is removed by threshold processing and the denoised signal is reconstructed. This algorithm effectively separates high-frequency noise from low-frequency valid signals. Combined with threshold processing to remove noise components (such as ambient light interference and circuit noise), the signal-to-noise ratio is improved; baseline correction; spectrum smoothing;

[0036] Baseline correction uses a polynomial fitting algorithm, and the calculation formula is:

[0037]

[0038] Where B(t) is the signal after baseline correction, a k is the fitting coefficient, and n is the polynomial order. The signal is baseline corrected according to the fitting result to avoid the subjective error of traditional manual correction. It is suitable for long-term continuous monitoring in complex environments.

[0039] S3: Extract the absorption coefficient, multi-band light attenuation characteristics and mass concentration parameters of black carbon aerosol from the preprocessed data;

[0040] The methods for extracting the absorption coefficient include: inverting the equivalent particle size parameters of black carbon aerosol based on the wavelength dependence of multi-band absorption spectra; optimizing the particle size distribution inversion process through regularization algorithms to reduce data noise interference; and using principal component analysis (PCA) to separate the contributions of black carbon and other aerosol components through differential analysis of multi-wavelength absorption spectra and the wavelength dependence characteristics of black carbon aerosol (the absorption coefficient decays power-law with wavelength). The equivalent black carbon mass concentration is calculated based on the absorption enhancement model (such as the AE33 algorithm).

[0041] S4: Analyze the features based on the machine learning model and output the real-time prediction results of black carbon aerosol concentration;

[0042] The machine learning model is a convolutional neural network or support vector machine, and its training method includes: using a historical black carbon aerosol dataset, which contains 7-band light attenuation data, absorption coefficient time series, environmental temperature and humidity parameters, and corresponding concentration labels; normalizing the dataset and dividing it into a training set and a validation set; optimizing the model parameters through the backpropagation algorithm until the prediction error converges to a preset threshold, automatically extracting local features of the spectral data through the convolution kernel, designing the hidden layer as a multi-layer perception structure, and mapping the output layer to the concentration value. The kernel function is used to process nonlinear relationships, and it performs excellently on small datasets, shortening the prediction response time.

[0043] It also includes data storage and output, including the following steps: synchronously storing the original data and analysis structure to the local SD card or cloud platform; displaying the black carbon aerosol concentration and particle size distribution curve through the mobile terminal. The cloud platform adopts an edge computing architecture to aggregate and analyze multi-device data to generate a regional aerosol pollution map. It supports local and cloud platform dual storage modes to ensure data security and accessibility. At the same time, regional pollution maps are generated through edge computing to provide macro-decision-making support for environmental monitoring. The real-time display and report generation functions simplify the operation process. Users can obtain key parameters instantly through the mobile terminal, which improves the user experience.

[0044] An embodiment of the present application also discloses a data processing system for an aerosol analyzer, including a processor and a memory, wherein the memory stores computer program instructions. When the computer program instructions are executed by the processor, a remote monitoring method for a weighing sensor based on the Internet of Things according to the present application is implemented.

[0045] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface. The configuration and functions of these components are known in the art and will not be described in detail here.

[0046] In this application, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, such as resistive random access memory, dynamic random access memory, static random access memory, enhanced dynamic random access memory, etc., or any other medium that can be used to store the required information and can be accessed by an application, module, or both. Any such computer storage medium may be part of, accessible to, or connectable to a device.

[0047] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A data processing method for an aerosol analyzer, characterized in that: Including steps: The light absorption signal of aerosol particles is collected by the light source system and the detector to generate the original electrical signal; Preprocessing the original electrical signal to obtain preprocessed data; Extract the absorption coefficient, multi-band light attenuation characteristics and mass concentration parameters of black carbon aerosol from the preprocessed data; The features are analyzed based on a machine learning model to output real-time prediction results of black carbon aerosol concentration.

2. The data processing method for an aerosol analyzer according to claim 1, wherein: Preprocessing the original electrical signal to obtain preprocessed data includes the following steps: Noise removal, noise removal uses wavelet transform algorithm, the calculation formula is: Among them, S(t) is the denoised signal, E j (t) is the wavelet coefficient of layer j, V J (t) is the remaining low-frequency component. The noise component is removed by threshold processing and the denoised signal is reconstructed; Baseline correction; Spectral smoothing.

3. The data processing method for an aerosol analyzer according to claim 2, wherein: The baseline correction adopts a polynomial fitting algorithm, and the calculation formula is: Where B(t) is the signal after baseline correction, a k is the fitting coefficient, n is the polynomial order; The signals were baseline corrected according to the fitting results.

4. The data processing method for an aerosol analyzer according to claim 1, wherein: The method for extracting the absorption coefficient includes: calculating the nonlinear relationship between multi-band light attenuation and black carbon mass concentration based on the Lambert-Beer law; The light attenuation value of each wavelength is obtained by using the light absorption signals of the main light source 880nm and the auxiliary bands 370nm, 470nm, 520nm, 590nm, 660nm, and 950nm; According to the wavelength-dependent characteristics, the equivalent absorption coefficient of black carbon aerosol is calculated using the absorption enhancement model.

5. The data processing method for an aerosol analyzer according to claim 1, wherein: The machine learning model is a convolutional neural network or a support vector machine, and its training method includes: Use a historical black carbon aerosol dataset, which includes 7-band light attenuation data, absorption coefficient time series, ambient temperature and humidity parameters, and corresponding concentration labels; Normalize the dataset and divide it into training set and validation set; The model parameters are optimized through the back propagation algorithm until the prediction error converges to the preset threshold.

6. The data processing method for an aerosol analyzer according to claim 1, wherein: It also includes data storage and output, including the steps: Synchronously store raw data and analysis structures to a local SD card or cloud platform; The black carbon aerosol concentration and particle size distribution curve are displayed on the mobile terminal.

7. The data processing method for an aerosol analyzer according to claim 6, wherein: The cloud platform adopts an edge computing architecture to aggregate and analyze data from multiple devices to generate regional aerosol pollution maps.

8. A data processing system for an aerosol analyzer, characterized in that: The device comprises a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a data processing method for an aerosol analyzer according to any one of claims 1 to 7 is implemented.

Images