考虑船用汽轮机工作原理与结构特点，基于逆顺序法建立多级单缸汽轮机整机耦合变工况计算模型；结合机组变工况试验数据，提出基于试验数据修正的多参数迭代变工况计算模型，实现高精度汽轮机变工况计算；融合BP神经网络技术，建立基于数据机理融合的高精度BP-NNs变工况计算模型，拓展汽轮机变工况计算模型的使用范围和通用性，并结合试验数据验证BP-NNs汽轮机变工况计算模型有效性，为船用汽轮机复杂变工况的动态运行与调节控制优化提供有力的技术支撑。
Considering the working principle and structural characteristics of marine steam turbines, the coupling variable condition calculation model of the multi-stage single-cylinder steam turbine was established based on the reverse sequence method. Combined with the test data of steam turbine, a multi-parameter iterative variable operating condition corrected based on the test data is proposed, which was used to perform high-precision calculations of steam turbine variable conditions. Then, the BP neural network technology is integrated to establish a high-precision BP-NNs variable-condition calculation model based on data mechanism fusion, which expands the use range and versatility of the steam turbine variable-condition calculation model. Combined with the experimental data, the validity of the BP-NNs steam turbine variable-condition calculation model is verified, and it provides strong technical support for the dynamic operation and adjustment control optimization of marine steam turbines under complex variable conditions.
2023,45(3): 96-100 收稿日期：2022-01-04
DOI：10.3404/j.issn.1672-7649.2023.03.017
分类号：TK212,U664.1
基金项目：国家自然科学基金资助项目(51609251)
作者简介：张磊(1986-),男,博士,讲师,研究方向为舰船动力系统的科学管理
