Study On Air Pollutant Monitoring And Air Quality Prediction System In Thermal Power Plants

As the main form of power generation in China,thermal power generation inevitably causes environmental pollution while driving economic and industrial development.Environmental protection and economic development are dialectical and complementary,and the solution to environmental problems begins with control of the source of pollutants.As an energy hub,thermal power plants are an important source of air pollutant emissions.Monitoring air pollutants in thermal power plants and predicting air quality in the monitoring area can not only understand the real-time status of pollutants and control key emission areas,but also provide data support for production planning and pollution control in thermal power plants.The research is of great significance.To this end,the following research is carried out in this dissertation:1)A thermal power plant air pollutant monitoring system is designed to address the problems of current air pollutant monitoring system such as high monitoring cost,large size of equipment,and difficulty in achieving regionalized and accurate monitoring.Firstly,the Zigbee terminal node and coordination node are used to collect and transmit the data of PM_2.5,PM₁₀,SO₂,CO,NO₂and O₃,then the data format conversion is completed through the gateway module,after that the data uploading is completed with the 4G module,finally the software program is compiled and connected to the cloud platform,and the Web monitoring platform is developed to realize the data storage,management and display.2)To address the problem that the current air quality prediction model is complex and it is difficult to achieve accurate prediction by adjusting the model parameters manually,we propose an optimized random forest model to predict air quality.First,three optimization algorithms,Grid Search（GS）,Bayesian optimization（BO）,and Particle Swarm Optimization（PSO）,are used to optimize the random forests（RF）model parameters n＿estimators,max＿depth,and min＿sample.max＿depth,min＿samples＿split and min＿samples＿leaf,and then three air quality prediction models,GS-RF,BO-RF and PSO-RF,were constructed to predict the Air Quality Index（AQI）,and finally three evaluation indexes,RMSE,MAE and R2,were used to evaluate the performance of the models.Finally,three evaluation indexes,RMSE,MAE and R²,were used to evaluate the performance of the models.Finally,the test session of the system was conducted.The test results showed that the system modules functioned normally and could achieve real-time monitoring of air pollutants in thermal power plants and meet the system design requirements,while the BO-RF air quality prediction model had the best prediction effect for AQI and could meet the air quality prediction requirements in the monitoring area of thermal power plants,and the visual monitoring platform was developed to realize the integration of monitoring and prediction.Figure[82]Table[19]Reference[80]...