铸造
鑄造
주조
FOUNDRY
2010年
3期
231-234
,共4页
蔡增辉%郝启堂%李强%李新雷
蔡增輝%郝啟堂%李彊%李新雷
채증휘%학계당%리강%리신뢰
模糊控制%输入与输出比例因子%分段调整%液面加压系统
模糊控製%輸入與輸齣比例因子%分段調整%液麵加壓繫統
모호공제%수입여수출비례인자%분단조정%액면가압계통
fuzzy controller%input and output scale factor%tune by range%filling system
液面加压系统担负着控制合金液上升、充型、结壳、保压等重要任务,其控制精度对反重力铸造铸件的质量具有重要影响.常规模糊控制仅用一组恒定不变的输入、输出比例因子,难以保证被控全过程处于最佳控制状态,在反重力铸造液面加压控制系统中的应用受到限制.将模糊控制中输入与输出比例因子在不同压力偏差范围内进行分段调整:压力偏差较大时,增大输出比例因子,加快动态响应;压力偏差较小时,增大输入比例因子,减少稳态误差.试验结果表明,该液面加压控制系统具有较好的静态性能和鲁棒性,实际压力曲线的跟踪性能良好,升液、充型阶段压力平稳,满足了反重力铸造的控制要求.
液麵加壓繫統擔負著控製閤金液上升、充型、結殼、保壓等重要任務,其控製精度對反重力鑄造鑄件的質量具有重要影響.常規模糊控製僅用一組恆定不變的輸入、輸齣比例因子,難以保證被控全過程處于最佳控製狀態,在反重力鑄造液麵加壓控製繫統中的應用受到限製.將模糊控製中輸入與輸齣比例因子在不同壓力偏差範圍內進行分段調整:壓力偏差較大時,增大輸齣比例因子,加快動態響應;壓力偏差較小時,增大輸入比例因子,減少穩態誤差.試驗結果錶明,該液麵加壓控製繫統具有較好的靜態性能和魯棒性,實際壓力麯線的跟蹤性能良好,升液、充型階段壓力平穩,滿足瞭反重力鑄造的控製要求.
액면가압계통담부착공제합금액상승、충형、결각、보압등중요임무,기공제정도대반중력주조주건적질량구유중요영향.상규모호공제부용일조항정불변적수입、수출비례인자,난이보증피공전과정처우최가공제상태,재반중력주조액면가압공제계통중적응용수도한제.장모호공제중수입여수출비례인자재불동압력편차범위내진행분단조정:압력편차교대시,증대수출비례인자,가쾌동태향응;압력편차교소시,증대수입비례인자,감소은태오차.시험결과표명,해액면가압공제계통구유교호적정태성능화로봉성,실제압력곡선적근종성능량호,승액、충형계단압력평은,만족료반중력주조적공제요구.
Filling system plays an important role in the process control at the stages of molten metal lifting, filling, freezing and pressure maintenance in counter gravity casting (CGC). Casting quality is affected by control precision of the system to a great extent. Input and output scale factors of conventional fuzzy controller are static and it's hard to ensure good state of the system, so application of the fuzzy technology is restricted on the filling system of CGC .The input and output scale factors of fuzzy controller are tuned by range in different errors in this paper, when the error of pressure is large, the output scale factor is increased to speed up dynamic response; in another condition, the input scaling factor is increased in order to reduce steady state error. Experimental results show that, good steady-state precision and well robust has been got by this method. Lifting and filling stage is steady and tracking property is very. Good, which indicate that the fuzzy controller with self-tuning scaling factors could be used in CGC.