系统工程理论与实践
繫統工程理論與實踐
계통공정이론여실천
Systems Engineering—Theory & Practice
2015年
2期
521~527
,共null页
刘欣 郝晓弘 杨新华 安爱民
劉訢 郝曉弘 楊新華 安愛民
류흔 학효홍 양신화 안애민
板式固体氧化物燃料电池 热能传递 预测控制 鲁棒性
闆式固體氧化物燃料電池 熱能傳遞 預測控製 魯棒性
판식고체양화물연료전지 열능전체 예측공제 로봉성
planar solid oxide fuel cell; thermal energy transfer; predictive control; robustness
固体氧化物燃料电池系统工作过程中燃料与氧化剂的压力差和燃料利用率的变化值是衡量系统运行安全、稳定长效、具有较好鲁棒性的重要指标. 本文提出采用离线计算、在线优化相结合的带有输出反馈的鲁棒模型预测控制方法, 分别将燃料与氧化剂的压力差和燃料利用率作为输入和输出约束, 离线计算目标函数上界及其系列渐近稳定域, 在线时对控制量进行精确定位. 仿真结果表明, 采用了状态反馈的模型预测控制方法能有效克服模型失配问题并迅速获得被控量的预测值, 使系统当负载电流发生波动时能克服变化引起的参数偏差, 提高了响应速度, 增强了系统鲁棒性.
固體氧化物燃料電池繫統工作過程中燃料與氧化劑的壓力差和燃料利用率的變化值是衡量繫統運行安全、穩定長效、具有較好魯棒性的重要指標. 本文提齣採用離線計算、在線優化相結閤的帶有輸齣反饋的魯棒模型預測控製方法, 分彆將燃料與氧化劑的壓力差和燃料利用率作為輸入和輸齣約束, 離線計算目標函數上界及其繫列漸近穩定域, 在線時對控製量進行精確定位. 倣真結果錶明, 採用瞭狀態反饋的模型預測控製方法能有效剋服模型失配問題併迅速穫得被控量的預測值, 使繫統噹負載電流髮生波動時能剋服變化引起的參數偏差, 提高瞭響應速度, 增彊瞭繫統魯棒性.
고체양화물연료전지계통공작과정중연료여양화제적압력차화연료이용솔적변화치시형량계통운행안전、은정장효、구유교호로봉성적중요지표. 본문제출채용리선계산、재선우화상결합적대유수출반궤적로봉모형예측공제방법, 분별장연료여양화제적압력차화연료이용솔작위수입화수출약속, 리선계산목표함수상계급기계렬점근은정역, 재선시대공제량진행정학정위. 방진결과표명, 채용료상태반궤적모형예측공제방법능유효극복모형실배문제병신속획득피공량적예측치, 사계통당부재전류발생파동시능극복변화인기적삼수편차, 제고료향응속도, 증강료계통로봉성.
In the operating process of the solid oxide fuel cell system, the pressure difference between fuel and air and the fuel utility are the important indexes, which evaluate the system's safety, long-term stability and better robustness. It is proposed in this paper that an output feedback robust model predictive control method is applied which is obtained the offline calculation and online optimal. The pressure difference between fuel and air and the fuel utility are used as input and output constraints, respectively. When offline, calculating upper bounds of the objective function and its series asymptotic stable domain, and online, making accurate position of control parameters. The simulation results show that using the state feedback model predictive control method can effectively overcome the model mismatch problem of the system and parameter deviation caused by the loaded current disturbance, improve the response speed, and enhance system robustness.