微燃机三联供机组变工况动态能效优化多目标非线性经济预测控制

Multi-objective Nonlinear Economic Predictive Controller for Dynamic Energy Efficiency Optimization of MGT-LiBr CCHP Under Variable Working Conditions

针对分布式燃气冷热电联供机组的大范围变负荷跟踪需求和运行过程动态能效提升的迫切性,提出一种基于乌托邦点跟踪框架的微燃机三联供机组多目标非线性经济预测控制方法。首先,根据微燃机冷热电联供机组机理模型深入分析机组运行过程的非线性特征,并以此构建了能够表征机组非线性的人工神经网络非线性预测模型;然后,根据机组㶲效率计算公式,分析影响机组㶲效率的关键影响因素,并以此构建反映机组负荷跟踪性能和动态能效的多目标优化经济预测控制问题。最后,通过仿真实验验证表明所提多目标非线性经济预测控制算法能够在满足微燃机三联供机组冷热电3种负荷大范围变工况跟踪精度的同时,改善机组动态运行过程中的能效和经济性。

To meet the demand for large-range load tracking and dynamic energy efficiency improvement in the micro gas turbine-LiBr double effect absorption refrigerating machine combined cooling,heating and power system(MGTLiBr CCHP),a multi-objective nonlinear economic predictive control method(MON-EMPC)based on the Utopian point tracking framework is proposed.First,the paper analyzes the nonlinear characteristics of operation process based on the MGT-LiBr CCHP mechanistic model and constructs an artificial neural network as predictive model describing the nonlinearity of the MGT-LiBr CCHP.Then,according to the formula of the exergy efficiency,the key factors affecting the exergy efficiency are analyzed,and the multi-objective optimal economic prediction control problem reflecting the load tracking and dynamic energy efficiency is constructed.Finally,the proposed multi-objective nonlinear economic prediction control algorithm is validated by simulation,which shows the algorithm can improve dynamic energy efficiency while satisfying the tracking accuracy of multiple loads in a wide range of variable working conditions.