Study On The Optimization Of Initial Pressure Of Thermal Power Units Based On Operation Data

With the promotion of China’s "carbon peaking and carbon neutrality" policy and the adjustment of the electricity consumption structure,the peak-to-valley difference of the power grid gradually increases,which requires more and more traditional thermal power plant units to participate in the task of peaking to meet the demand of grid load changes.This has led to many large-capacity,high-parameter units being operated under non-rated conditions for a long time,thus affecting the unit’s economy.As one of the most important equipment in power plants,the turbine may deviate from its design curve due to the long-term variable operating conditions,resulting in lower unit economy.Therefore,in order to improve the economy of the unit under the task of peak shaving,it is necessary to optimize the initial pressure of the unit.The optimization of the initial pressure is of great significance to the economic operation of the power plant and energy saving and consumption reduction.In this paper,a heat consumption rate prediction model is established based on operation data,and the initial pressure is optimized by combining the method of initial pressure model.First,the simulation software is used to run simulation to obtain operation data,and data pre-processing methods such as data cleaning and steady-state detection are used to pre-process data to obtain training data with high quality.Then the input parameters were selected by the heat consumption rate formula and the actual analysis,and the gray correlation was verified for the input parameters and the heat consumption rate,and the results showed that the selected parameters had a high correlation with the heat consumption rate,and the larger influence could be used as the model input parameters.In order to improve the performance of the heat consumption rate model,the data samples are normalized to reduce the distance between the parameters due to different dimensions,and the support vector regression machine type and kernel function type with better regression effect are selected,and the hyperparameters of the model were optimized by grid search and cross-validation.Based on the above modeling optimization preparations,the support vector regression machine was used to train the training data,and the heat consumption rate prediction model was established,and the robustness of the model was analyzed,and the regression verification was performed on the training set and the test set respectively.The results show that the established heat consumption rate prediction model has good regression accuracy and generalization ability.Based on the established heat dissipation rate prediction model,the feasible pressure interval of the main steam pressure is determined,and the initial pressure optimization model is established with the minimum heat dissipation rate as the objective function.There is an unknown parameter in the objective function of the initial pressure model,and this paper proposes a method to determine other parameters using a clustering algorithm.The data samples are divided to obtain the clustering centers,and the clustering center with the smallest Euclidean distance between load and pressure is used as the other parameters,so that the initial pressure model can be solved to obtain the optimization results and the optimized curves under each typical load.The optimized main steam pressure is higher than the actual one and the heat consumption rate is lower,which can improve the economy of the unit and better guide the safe,economic and stable operation of the unit,with certain theoretical significance and application value.