基于历史寻优的火电机组运行优化研究

发布时间：2018-12-16 02:31

【摘要】：目前我国面对十分严峻的人口、恶劣环境和一次能源消耗偏高等一系列问题,要实现经济社会的可持续发展,必须走节约资源的道路。以绿色低碳为主线“十三五”能源规划首次设定这样一个目标,我国将在2020年前把能源消耗计划控制在50亿吨标准煤以内,因此各火力发电企业纷纷考虑降低能耗来增强市场竞争力。在电厂中采取对电站机组的可控运行参数进行合理优化,确保机组在不同的运行条件下能够达到最佳的运行状态,是运行参数优化的根本目的,同时也是实现机组节能降耗的重要手段。由于电厂大量的运行数据受到复杂非线性和噪声污染多等因素出现一些不确定性问题。为保证数据的真实性和有效性,本文首先对历史数据进行预处理。然后提出敏感性因子概念,进行了敏感性分析工作,分析机组热耗率与各边界参数的敏感性因子大小并得出能耗敏感性因子较大的参数。采用模糊C-均值聚类算法对数据进行工况划分。在工况划分之后,本文以机组的运行初压优化为例,进行历史寻优,找出相同工况最优初压,作为后续建模的依据。本文分别采用改进的BP和RBF神经网络建立边界参数与主蒸汽压力之间模型。通过分析两种建模方法的相对误差大小,最终选取RBF神经网络方法进行建模。最后,根据机组整体模型,验证在边界条件改变的情况下,该模型得到的最优主汽压力变化符合理论规律,证明寻优模型的有效性。本论文的研究工作对机组运行参数优化问题的分析,具有一定的理论意义和实用价值。
[Abstract]:At present, our country is faced with a series of problems, such as severe population, bad environment and high primary energy consumption. In order to realize the sustainable development of economy and society, we must take the road of saving resources. Taking green and low carbon as the main line, the 13th Five-Year Plan for Energy Resources has for the first time set such a target. China will limit its energy consumption plan to within 5 billion tons of standard coal by 2020. Therefore, thermal power enterprises have to consider reducing energy consumption to enhance market competitiveness. It is the basic purpose of the operation parameter optimization to adopt the reasonable optimization of the controllable operating parameters of the units in the power plant to ensure that the units can achieve the best operation state under different operating conditions. At the same time, it is also an important means to achieve energy saving and consumption reduction. There are some uncertain problems due to the complex nonlinearity and noise pollution of a large number of operation data in power plants. In order to ensure the authenticity and validity of the data, this paper preprocesses the historical data. Then the concept of sensitivity factor is put forward and sensitivity analysis is carried out. The magnitude of sensitivity factor of unit heat consumption rate and each boundary parameter is analyzed and the parameters with larger energy consumption sensitivity factor are obtained. Fuzzy C-means clustering algorithm is used to partition the data. After the working condition is divided, this paper takes the operation initial pressure optimization of the unit as an example, carries on the historical optimization, and finds out the optimal initial pressure of the same working condition, which is the basis for the subsequent modeling. In this paper, the improved BP and RBF neural networks are used to establish the model between the boundary parameters and the main steam pressure. By analyzing the relative error of the two modeling methods, the RBF neural network method is selected to model the model. Finally, according to the overall model of the unit, it is verified that the optimal main steam pressure variation obtained by the model accords with the theoretical law and proves the effectiveness of the optimization model under the condition of changing the boundary conditions. The research work of this paper has certain theoretical significance and practical value to the analysis of the optimization problem of unit operating parameters.
【学位授予单位】：华北电力大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM621

【参考文献】