CN-121978278-A - Real-time monitoring method and system for air quality of steel bar production workshop for high-speed rail

CN121978278ACN 121978278 ACN121978278 ACN 121978278ACN-121978278-A

Abstract

The invention relates to the technical field of environmental monitoring and data processing, and discloses a method and a system for monitoring air quality of a steel bar production workshop for a high-speed railway in real time, wherein the method comprises the steps of constructing a unified state vector comprising a pollutant concentration field, a potential unknown source item, a sensor zero offset and a response gain; the method comprises the steps of obtaining original data of a dense static module and reference values of sparse mobile anchor points as observation vectors, correcting a prediction state set by utilizing an observation operator based on a data assimilation algorithm to obtain an analysis state set, and extracting monitoring results from the analysis state set. According to the scheme, the high-precision information of the sparse mobile anchor point is synchronously transmitted to the sensor parameter estimation by utilizing the statistical correlation in the unified state vector, so that the real-time online calibration of the dense low-cost sensor is realized, the problems of zero drift and sensitivity attenuation are solved, meanwhile, multi-source heterogeneous data are fused, the high spatial resolution and high measurement accuracy are considered, the unknown pollution source can be dynamically compensated, and the robustness of the system is improved.

Inventors

GAO XIAOLIANG
XIE HAIBIN
YU FEI
ZHANG HAIYAN
SHEN ZHIGANG
GUO WENHAO
LI NA

Assignees

河南省鼎鼎实业有限公司

Dates

Publication Date: 20260505
Application Date: 20260204

Claims (10)

1. The method for monitoring the air quality of the steel bar production workshop for the high-speed rail in real time is characterized by comprising the following steps of: constructing a unified state vector, the unified state vector comprising: a contaminant concentration field vector, a potential unknown source term vector, a sensor zero offset vector, and a sensor response gain vector; constructing a prediction state set based on the unified state vector; obtaining an observation vector, wherein the observation vector comprises an original concentration data signal output by a dense static sensing module and a reference concentration value output by a sparse mobile anchor point module; Constructing an observation operator, wherein the observation operator is used for mapping the unified state vector from a state space to an observation space; based on a data assimilation algorithm, correcting the prediction state set by using the observation vector and the observation operator to obtain an analysis state set; monitoring results are extracted from the analysis state set, wherein the monitoring results comprise the pollutant concentration field vector, the sensor zero offset vector and the sensor response gain vector.
2. The method for monitoring air quality in a shop for reinforcing steel bars for high-speed rails according to claim 1, wherein constructing the prediction state set based on the unified state vector comprises: A physical and statistical evolution model is applied, a state set at the current moment is deduced to the next moment, and the prediction state set is generated; the physical and statistical evolution models include an atmospheric diffusion model for evolving the contaminant concentration field vector and a stochastic process model for evolving the potential unknown source term vector, the sensor zero offset vector, and the sensor response gain vector.
3. The method for monitoring the air quality of the steel bar production workshop for the high-speed rail in real time according to claim 1, wherein the observation operator comprises: A first sub-function for calculating a theoretical prediction value of the raw concentration data signal based on the sensor response gain vector, the contaminant concentration field vector, and the sensor zero offset vector; and a second sub-function for extracting a theoretical predicted value of the reference concentration value according to the contaminant concentration field vector.
4. The method for monitoring air quality in a high-speed rail steel bar production workshop in real time according to claim 1, wherein the correcting the prediction state set based on the data assimilation algorithm comprises the following steps: Calculating a sample covariance matrix of the prediction state set, wherein the sample covariance matrix represents the statistical correlation among components in the unified state vector; And synchronously distributing the deviation between the observation vector and the forecasting value of the observation operator to all the components of the unified state vector according to the statistical correlation.
5. The method for real-time monitoring air quality in a steel bar production plant for high-speed rail according to claim 1, wherein extracting the monitoring result from the analysis state set comprises: calculating the statistical mean value of the analysis state set to obtain an optimal estimation state vector; extracting the pollutant concentration field vector from the best estimated state vector as a high-resolution pollutant concentration distribution map.
6. The method for real-time monitoring air quality in a steel bar production plant for high-speed rail according to claim 5, wherein the method for real-time monitoring air quality in a steel bar production plant for high-speed rail further comprises extracting the sensor zero offset vector and the sensor response gain vector from the best estimated state vector as real-time calibration parameters of the dense static sensing module.
7. The method for monitoring air quality in a high-speed rail steel bar production plant in real time according to claim 6, wherein the method further comprises: applying the real-time calibration parameters, and performing inverse operation on the original concentration data signals to obtain corrected concentration values; The inverse operation is subtracting the corresponding component of the sensor zero offset vector from the raw concentration data signal and dividing by the corresponding component of the sensor response gain vector.
8. The method for monitoring air quality in a high-speed rail rebar production plant according to claim 1, further comprising an initialization stage comprising: constructing a unified state vector initial set comprising a plurality of set members; The set members are generated by superposing random disturbance on the initial estimated value; The initial estimate of the sensor response gain vector is 1 and a positive value constraint is imposed.
9. The method for monitoring air quality in a high-speed rail steel bar production workshop in real time according to claim 1, wherein the method further comprises: evaluating an uncertainty of the analysis state set; and generating a feedback instruction for guiding the moving path of the sparse mobile anchor point module based on the uncertainty.
10. Real-time monitoring system of air quality in steel bar production workshop for high-speed railway, its characterized in that includes: The dense static sensing module is used for outputting an original concentration data signal; the sparse mobile anchor point module is used for outputting a reference concentration value; The central fusion and decision module comprises a processor and a memory, and is in communication connection with the dense static sensing module and the sparse mobile anchor point module; the central fusion and decision module is configured to construct a unified state vector comprising: a contaminant concentration field vector, a potential unknown source term vector, a sensor zero offset vector, and a sensor response gain vector; constructing a prediction state set based on the unified state vector; obtaining an observation vector, the observation vector comprising the raw concentration data signal and the reference concentration value; Constructing an observation operator, wherein the observation operator is used for mapping the unified state vector from a state space to an observation space; based on a data assimilation algorithm, correcting the prediction state set by using the observation vector and the observation operator to obtain an analysis state set; monitoring results are extracted from the analysis state set, wherein the monitoring results comprise the pollutant concentration field vector, the sensor zero offset vector and the sensor response gain vector.

Description

Real-time monitoring method and system for air quality of steel bar production workshop for high-speed rail Technical Field The invention relates to the technical field of environmental monitoring and data processing, in particular to a real-time air quality monitoring method and system for a steel bar production workshop for a high-speed rail. Background With the progress of environmental monitoring technology, air quality monitoring networks based on large-scale low-cost sensors have become a mainstream development trend because of the advantages of flexible deployment, low cost, capability of providing high-density space-time data and the like. However, such monitoring networks have significant technical drawbacks in practical applications. First, the manufacturing process and working principle of the low-cost sensor determine that the measurement accuracy and long-term stability of the low-cost sensor are poor, and the low-cost sensor is extremely susceptible to zero drift and response gain attenuation caused by environmental factors and self-aging. This results in a large systematic error between the raw concentration data signal output by the sensor and the true value after long-term operation, which affects the reliability and effectiveness of the monitoring data. The traditional offline calibration or periodic manual calibration mode is low in efficiency, and cannot meet the requirement of a real-time and large-scale monitoring network on continuous online self calibration. The steel bar production workshop for high-speed rail is taken as a typical industrial pollution scene, and the air quality monitoring has clear requirements and specificity. The pollutants in the workshop mainly comprise welding fume, cutting dust (reinforcing steel bar scraps and silicate dust), a small amount of volatile organic compounds and carbon monoxide (generated in the welding process), and have the emission rules of high local concentration, concentrated emission points (such as welding stations and cutting areas), fine dust particle size and easiness in diffusion along with air flow. The pollutants not only affect the health of workshop operators, but also can cause equipment abrasion due to sedimentation, and the pollution prevention and control are supported by a high-precision and real-time monitoring scheme. However, in the scene, the existing monitoring network based on the low-cost sensor has the common problems of zero drift, insufficient data fusion and the like, and also faces the additional challenge of a high-dust environment on the stability of the sensor, so that the application value of the technical scheme of the invention is further highlighted. Secondly, in actual air quality monitoring, there is a situation that high-accuracy point measurement data provided by a high-accuracy reference level instrument (sparse mobile anchor point) coexist with high-density but low-accuracy data provided by a low-cost sensor network. In the prior art, a simple data fusion or interpolation method is generally adopted, and the two heterogeneous data sources are difficult to effectively fuse in a unified mathematical framework, so that the finally output pollutant concentration field spectrum is difficult to simultaneously consider high spatial resolution and high measurement accuracy. Furthermore, pollutant emissions in the environment tend to have uncertainties, with potentially unknown sources of pollution (e.g., sudden leaks or illegal emissions) that are difficult to model. The existing atmospheric diffusion model generally depends on a known emission list, when sudden unknown sources appear, a model prediction result can generate larger errors, and the identification and compensation capability of the system on the dynamically-changed unknown sources is insufficient, so that the real-time response capability and early warning accuracy of the monitoring system are affected. Therefore, a new technical method is needed, which can realize real-time online self-calibration of a low-cost sensor in a unified framework with physical interpretability, and simultaneously, efficiently fuses multi-source heterogeneous data to obtain a high-precision and high-resolution pollutant concentration field, and has robust monitoring capability on potential unknown pollution sources. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a real-time monitoring method and system for the air quality of a steel bar production workshop for a high-speed rail, which solve the problems that a low-cost sensor in the prior art has low data reliability and is difficult to realize real-time online calibration due to zero drift and sensitivity attenuation, and a monitoring network is difficult to effectively fuse high-precision anchor point data to achieve high spatial resolution and high measurement accuracy and has insufficient response capability to sudden unknown pollution sources. The first aspect of the invent