| 摘要 |
<p>PURPOSE:To infer a control operation command train based on deep knowledge by generating control operation commands through the use of knowledge on an objective model which describes the construction of a plant, characteristics which equipments have and the like and on an operation principle which describes a restriction conditions at the time of operating the plant. CONSTITUTION:A target input means 1 inputs an output target value train being an operation schedule to an inference controller 6 and a process input means 2 inputs a plant current state to the inference controller 6. It sets plant data in an objective model storage part 8. Then, current time (t) and a sampling period t are set. The target value of time (t)+ t is fetched from the inputted target value column. The fetched target value is given to an operation generating mechanism 4 and the retrieval of an operation command which is necessary for transition to the plant state at the time is started. The operation generating mechanism 4 infers the state of the respective equipments satisfying the output target value through the use of the inputted plant state, the objective model which is stored in the objective model storage part 8 and the operation principle.</p> |
