| 摘要 |
<P>PROBLEM TO BE SOLVED: To quicken prediction computation in response to an operation, and to stabilize control in response to a control object. <P>SOLUTION: An optimum control input determination part 3 of this model prediction controller 10 predicts a controlled variable as to an optimum control input candidate, using a dynamic characteristic model of the control object 20, and determines the optimum control input. An optimum control input precision setting part 4 sets precision of the optimum control input in every operation, and limits a searching frequency of the optimum control input. When a time constant change of the control object 20 is observed by a time constant change observation part 6, a time and section setting part 8 changes at least one out of a sampling time, a prediction section and a control section in response to the time constant change to stabilize the control. <P>COPYRIGHT: (C)2006,JPO&NCIPI |
