阅读说明：本技术 基于机器学习的热强化sve技术气体排放控制方法 (Machine learning-based heat-strengthened SVE (singular value Environment) technology gas emission control method ) 是由 石岩 芮树旺 杨丽曼 许少峰 王一轩 王娜 孙治博 牛燕霞 蔡茂林 于 2021-10-16 设计创作，主要内容包括：本发明公开了基于机器学习的热强化SVE技术气体排放控制方法,包括：测量土壤在不同温度恒温加热处理下所排放的VOCs气体浓度,以及VOCs气体浓度随土壤温度的变化数据；对VOCs气体浓度进行区间划分,得到多个浓度区间；将多个浓度区间作为不同类,并使用SVM和随机森林方法对所述不同类进行训练,得到最终模型；通过最终模型对后续通过热强化SVE修复土壤所产生的VOCs气体浓度进行预测,并通过调控热强化SVE修复土壤过程中的加热温度,实现控制VOCs气体的排放；通过该方法能够有效控制在使用热强化SVE技术修复土壤过程中所排放的VOCs气体。(The invention discloses a machine learning-based heat-strengthened SVE (singular value extraction) technology gas emission control method, which comprises the following steps: measuring the concentration of VOCs gas discharged by the soil under constant temperature heating treatment at different temperatures and the variation data of the concentration of the VOCs gas along with the soil temperature; carrying out interval division on the concentration of the VOCs gas to obtain a plurality of concentration intervals; taking the concentration intervals as different classes, and training the different classes by using an SVM (support vector machine) and a random forest method to obtain a final model; predicting the concentration of VOCs gas generated by subsequently restoring soil through the heat-enhanced SVE through the final model, and controlling the discharge of the VOCs gas by regulating and controlling the heating temperature in the soil restoring process through the heat-enhanced SVE; by the method, VOCs gas discharged in the process of repairing soil by using the heat-strengthening SVE technology can be effectively controlled.)

1. The heat strengthening SVE technology gas emission control method based on machine learning is characterized by comprising the following steps:

s1, measuring the concentration of VOCs gas discharged by the soil under constant temperature heating treatment at different temperatures and the variation data of the concentration of the VOCs gas along with the soil temperature;

s2, carrying out interval division on the VOCs gas concentration to obtain a plurality of concentration intervals;

s3, taking the concentration intervals as different classes, and training the different classes by using an SVM (support vector machine) and a random forest method to obtain a final model;

and S4, predicting the concentration of VOCs gas generated by subsequently restoring soil through the heat-enhanced SVE through the final model, and controlling the discharge of the VOCs gas by regulating and controlling the heating temperature in the soil restoring process through the heat-enhanced SVE.

2. The machine-learning-based thermally enhanced SVE technique gas emission control method of claim 1, wherein said S1 comprises:

carrying out constant-temperature heating treatment on the soil in a closed device at different temperatures for a preset time;

and recording the soil temperature at the current moment and the VOCs gas concentration corresponding to the soil temperature at the current moment every other preset period from the beginning of heating.

3. The method as claimed in claim 2, wherein the gas emission control method of the heat-enhanced SVE technique based on machine learning is characterized in that the gas concentration of VOCs emitted from the soil gradually becomes stable as the soil temperature is increased, and the gas concentration of VOCs is the highest concentration of VOCs.

4. The machine-learning-based thermally enhanced SVE technique gas emission control method of claim 3, wherein said S2 specifically comprises:

and taking the concentration from 0 to the highest concentration of the VOCs gas as a total concentration interval, and dividing the total concentration interval to obtain a plurality of concentration intervals.

5. The machine-learning-based thermally enhanced SVE technique gas emission control method of claim 4, wherein said S3 specifically comprises:

s31, designing a plurality of corresponding sub-classifiers according to the concentration intervals through an SVM (support vector machine) and a random forest algorithm, and training the sub-classifiers to obtain a plurality of training results;

s32, integrating the multiple training results through a Bagging method to obtain a final training result;

s33, multiplying the final training result by the highest concentration of the VOCs gas to obtain the concentration of the VOCs gas expected to be discharged;

s34, voting is carried out on the plurality of sub-classifiers through a Bagging method, and the relation between the soil moisture content, the organic matter content, the permeability, the heating temperature, the VOCs gas concentration at the previous T moments and the VOCs gas concentration at the current moment is generated, so that a final model is formed.

Technical Field

The invention belongs to the technical field of soil remediation, and particularly relates to a heat-strengthening SVE technology gas emission control method based on machine learning.

Background

The vapor phase extraction (SVE) technology is a soil in-situ remediation technology widely applied, but is often influenced by factors such as pollutant types, soil air permeability, water content, soil layer uniformity and the like in the using process, so that a good remediation effect is difficult to achieve in low-permeability organic substance polluted sites and semi-volatile organic substance polluted sites. In order to improve the repair effect and expand the application range, the SVE technology is often combined with other in-situ repair technologies to form SVE enhancement technologies through complementation, such as heat-strengthening SVE technology. However, a large amount of pollution gas is generated in the soil remediation process through the heat-strengthening SVE technology, so that great harm is generated to human health.

Therefore, how to effectively control the emission of the polluted gases (VOCs) in the soil remediation process by the heat-enhanced SVE technology has become a key issue of current research.

Disclosure of Invention

In view of the above problems, the present invention provides a machine learning-based thermally enhanced SVE technology gas emission control method by which the emission of pollutant gases (VOCs) can be effectively controlled during soil remediation using a thermally enhanced SVE technology.

The embodiment of the invention provides a machine learning-based heat-strengthened SVE (singular value analysis) technology gas emission control method, which comprises the following steps:

s1, measuring the concentration of VOCs gas discharged by the soil under constant temperature heating treatment at different temperatures and the variation data of the concentration of the VOCs gas along with the soil temperature;

s2, carrying out interval division on the VOCs gas concentration to obtain a plurality of concentration intervals;

s3, taking the concentration intervals as different classes, and training the different classes by using an SVM (support vector machine) and a random forest method to obtain a final model;

and S4, predicting the concentration of VOCs gas generated by subsequently restoring soil through the heat-enhanced SVE according to the final model, and controlling the discharge of the VOCs gas by regulating and controlling the heating temperature in the soil restoring process through the heat-enhanced SVE.

Further, the S1 includes:

carrying out constant-temperature heating treatment on the soil in a closed device at different temperatures for a preset time;

and recording the soil temperature at the current moment and the VOCs gas concentration corresponding to the soil temperature at the current moment every other preset period from the beginning of heating.

Further, as the soil temperature continuously increases, the gas concentration of VOCs discharged by the soil gradually tends to be stable, and the gas concentration of VOCs at the moment is the highest concentration of VOCs.

Further, the S2 specifically includes:

and taking the concentration from 0 to the highest concentration of the VOCs gas as a total concentration interval, and dividing the total concentration interval to obtain a plurality of concentration intervals.

Further, the S3 specifically includes:

s31, designing a plurality of corresponding sub-classifiers according to the concentration intervals through an SVM (support vector machine) and a random forest algorithm, and training the sub-classifiers to obtain a plurality of training results;

s32, integrating the multiple training results through a Bagging method to obtain a final training result;

s33, multiplying the final training result by the highest concentration of the VOCs gas to obtain the concentration of the VOCs gas expected to be discharged;

s34, voting is carried out on the plurality of sub-classifiers through a Bagging method, and the relation between the soil moisture content, the organic matter content, the permeability, the heating temperature, the VOCs gas concentration at the previous T moments and the VOCs gas concentration at the current moment is generated, so that a final model is formed.

Compared with the prior art, the heat-strengthened SVE technology gas emission control method based on machine learning, which is disclosed by the invention, can effectively control the polluted gases (VOCs) emitted in the process of repairing soil by using the heat-strengthened SVE technology, and lays a foundation for the subsequent treatment of the emitted polluted gases.

Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

Drawings

The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:

FIG. 1 is a flowchart of a method for controlling gas emissions by a thermally enhanced SVE technique based on machine learning according to an embodiment of the present invention.

Fig. 2 is a graph of a model fitted with kinetics of pollutant volatilization for different initial concentrations of different types of soil according to an embodiment of the present invention.

FIG. 3 is a diagram of a Markov model provided by an embodiment of the present invention.

Fig. 4 is a voting diagram for Bagging provided in the embodiment of the present invention.

Detailed Description

Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

Referring to fig. 1, an embodiment of the present invention provides a method for controlling gas emission in a thermally enhanced SVE technology based on machine learning, which specifically includes the following steps:

s1, measuring the concentration of VOCs gas discharged by the soil under constant temperature heating treatment at different temperatures and the variation data of the concentration of the VOCs gas along with the soil temperature;

s2, carrying out interval division on the VOCs gas concentration to obtain a plurality of concentration intervals;

s3, taking the concentration intervals as different classes, and training the different classes by using an SVM (support vector machine) and a random forest method to obtain a final model;

and S4, predicting the concentration of VOCs gas generated by subsequently restoring soil through the heat-enhanced SVE through the final model, and controlling the discharge of the VOCs gas by regulating and controlling the heating temperature in the soil restoring process through the heat-enhanced SVE.

The above method can be applied to different types of soil.

The above steps will be described in detail below.

In step S1, the soil is first heated in a closed device at a constant temperature, in this embodiment, in order to obtain more experimental data, the soil is heated at a constant temperature for 1 hour; recording the current temperature of the soil and the concentration of VOCs gas discharged by the soil at the current moment every 60s from the beginning of heating; changing the temperature during constant-temperature heating treatment, carrying out constant-temperature heating treatment on the soil again, and recording the concentration of VOCs gas discharged at different moments and the soil temperature at the moment corresponding to the concentration of the VOCs gas; therefore, multiple groups of data are acquired, and in the specific case, reference is made to fig. 2, and it can be seen from fig. 2 that the volatilization rate of the VOCs gas is positively correlated with the temperature;

by comparing the acquired multiple groups of data, the concentration of VOCs gas discharged by the soil gradually tends to be stable along with the continuous increase of the temperature of the soil, and the concentration of the VOCs gas which tends to be stable is the highest; the VOCs gas generated at each moment is only related to the heating temperature at the first K moments and the concentration of the generated VOCs gas, and the process is equivalent to a K-order Markov process; therefore, in the embodiment, the markov model is used for extracting the gas concentration characteristics of the VOCs at k +1 times to generate a characteristic matrix; the markov model is referred to fig. 3.

In the above step S2, the total concentration interval is set to be a concentration range from 0 to the highest concentration of the VOCs gas, and in this embodiment, the total concentration interval is divided into a plurality of concentration intervals every 5%; in the present example, a total of 20 concentration intervals were obtained.

In the step S3, a plurality of corresponding sub-classifiers are designed according to the plurality of concentration intervals through an SVM and a random forest algorithm and are trained respectively by combining the feature matrix obtained in the step S2 through a traditional machine learning method to obtain a plurality of training results; then, integrating the plurality of classification results in a voting mode by a Bagging method; obtaining a final training result, namely an expected exhaust gas concentration interval; with particular reference to fig. 4; multiplying the final training result by the highest concentration of the VOCs gas to obtain the specific concentration of the VOCs gas expected to be discharged;

voting is carried out on the plurality of sub-classifiers by a Bagging method, and the relation between the soil moisture content, the organic matter content, the permeability, the heating temperature, the VOCs gas concentration at the previous T moments and the VOCs gas concentration at the current moment is generated, so that a final model is formed.

In the step S4, predicting the concentration of VOCs gas generated by subsequently restoring soil by using the thermally enhanced SVE according to the model, and controlling the discharge of VOCs gas by adjusting and controlling the heating temperature in the soil restoring process by using the thermally enhanced SVE; when the concentration of the VOCs gas is reduced to a certain degree, the VOCs gas can be subjected to catalytic oxidation treatment, namely catalytic combustion treatment, and the VOCs gas is converted into water and carbon dioxide; when the concentration of the VOCs gas is reduced to a certain degree, the combustion efficiency can be improved; after the generated VOCs gas is detected out, the concentration of the important components is controlled to enable the principle explosion concentration range to effectively avoid the explosion of the combustible gas.

The heat-strengthening SVE technology gas emission control method based on machine learning provided by the embodiment of the invention can be applied to various types of soil.

The embodiment of the invention provides a machine learning-based gas emission control method based on a heat-strengthening SVE (singular value extraction) technology, which adopts an artificial intelligence method to control the concentration of the emitted gas, is more flexible than the traditional negative feedback method, can adapt to various soil environments, and can realize approximate accurate control on the concentration of the gas generated in the soil remediation process; the traditional method needs a large amount of theoretical derivation, and the method provided by the invention can adapt to various mutation conditions, has relatively loose requirements on related theories, and can enable people who are not in the field to understand related problems; in the traditional method, the concentration is classified and predicted only by a machine learning method, but the prediction result is inaccurate due to too few characteristics, and the method effectively solves the problem by expanding the characteristics through Markov assumption, so that the accuracy of the prediction result is improved; the machine learning method provided by the embodiment of the invention is simple, occupies less resources and is easy to popularize.

It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.