Research On Initial Steam Pressure Optimization Method Of Power Plant Based On Data Analysis

Under the background of energy saving and emission reduction,more and more enterprises pay attention to the saving and utilization of the resources of environment.With the improvement of the large data system,a large number of historical data is stored in the power station SIS system,which can transform to useful information.Data mining technology is gradually innovating and perfected in the data analysis of thermal power plant playing a very good role in guiding the optimal operation of the power plant.In this paper,the data analysis technology is used to optimize the operating parameters of the steam turbine unit in thermal power plant,thus achieving the aim of saving energy and reducing consumption.The main contents of the paper are as follows:1.The data based machine learning problem and the basic computation process of SVM algorithm widely applied in soft sensing technology were introduced,than the least squares support vector machine was proposed.In the process of establishing LSSVM model,the penalty coefficient C and kernel function parameter need to be determined.After comparison,particle swarm optimization algorithm has better effect.In this paper,the optimization principle and process of the algorithm were discussed.Data pre-processing of data mining data were removed with large noise and large errors,so as to improve the accuracy of the experimental results.2.Through detailed analysis of the whole operation of the steam turbine,the operation parameters related to the heat consumption rate were obtained.Three commonly used kernel functions were introduced,than three kinds of soft sensor models were built respectively,at last the error of the final resultd were compared to select basis function.Using the data of the working condition of the power plant as the experimental data,the value of the fitness and the model C and G is calculated.Compared with the prediction result of PSO-SVM model,it was verified that the PSO-LSSVM model has good generalization,high convergence speed and strong robustness.3.The improved Apriori algorithm was proposed based on association rules.In order to make the interval not too stiff,K-Means clustering partitioning algorithm was put forward,and a polynomial curve that can guide operation is synthesized.The coal consumption rate and heat consumption rate curve before and after optimization are compared and calculated.The comparison results demonstrated that the heat consumption rate and coal consumption rate of the optimized unit are significantly lower than those before optimization,which greatly improves the energy utilization rate.