CAJ | 학술논문

对于磨煤机负荷对象，当采用常规广义预测控制算法设计控制器时，发现存在上升时间过长的现象；当调节参数进而减少上升时间时，则会出现控制量振荡的现象。为了使预测控制计算的控制量逼近实际的最优控制量，提出一种改进的广义预测控制算法。通过在控制律中引入过去输入加权因子γ，调整过去输入对未来输出的影响，在不增加计算量的情况下，实现了对实际最优控制量的逼近。仿真结果表明，改进的广义预测控制算法，克服了常规广义预测控制响应时间较长，控制量振荡的问题，并且控制效果要优于常规广义预测控制、阶梯式广义预测、β增量型广义预测控制。
대우마매궤부하대상，당채용상규엄의예측공제산법설계공제기시，발현존재상승시간과장적현상；당조절삼수진이감소상승시간시，칙회출현공제량진탕적현상。위료사예측공제계산적공제량핍근실제적최우공제량，제출일충개진적엄의예측공제산법。통과재공제률중인입과거수입가권인자γ，조정과거수입대미래수출적영향，재불증가계산량적정황하，실현료대실제최우공제량적핍근。방진결과표명，개진적엄의예측공제산법，극복료상규엄의예측공제향응시간교장，공제량진탕적문제，병차공제효과요우우상규엄의예측공제、계제식엄의예측、β증량형엄의예측공제。
For the ball mill load, the response speed is slow with traditional generalized predictive control strategy. By tuning related parameters, response speed is faster with manipulated variables oscillating. To approach the optimal manipulated variable, an improved generalized predictive control is proposed. A past input weight factor γ is introduced to regulate the effect of past input on future output, which brings the optimal manipulated variable of approximation without extra calculation. Simulation results show that, slow response speed and oscillation of the manipulated variable are overcome in the improved generalized predictive control method. In addition, its control performance is better than that of traditional predictive control, stair-like predictive control and β type incremental generalized predictive control.