Abstract: In a continuous kneading apparatus according to an embodiment, for each of a plurality of ring-shaped heaters, a control unit determines a current state and a reward for an action selected in the past based on a control error calculated from an acquired temperature; updates a control condition based on the reward, and determines an optimum action corresponding to the current state under the updated control condition, the control condition being a combination of a state and an action; and controls a target ring-shaped heater based on the optimum action.

Abstract: In a resin film manufacturing device according to an embodiment, a current state and a reward for a previously selected action are determined for each heat bolt based on a control error calculated from a thickness distribution of a resin film acquired from a thickness sensor. Then, control conditions, which are a combination of states and actions, are updated based on the reward, and a most appropriate action corresponding to the current state is selected from the updated control conditions. Then, the heater is controlled based on the most appropriate action.

Abstract: In a resin film manufacturing device according to an embodiment, a current state and a reward for a previously selected action are determined for each heat bolt based on a control error calculated from a thickness distribution of a resin film acquired from a thickness sensor. Then, control conditions, which are a combination of states and actions, are updated based on the reward, and a most appropriate action corresponding to the current state is selected from the updated control conditions. Then, the heater is controlled based on the most appropriate action.

Abstract: In an extrusion molding apparatus according to an embodiment, a control unit configured to perform feedback control for a rotation speed of a pump so as to bring a pressure measured by a pressure sensor closer to a target pressure determines a current state and a reward for an action selected in the past based on a difference between the measured pressure and the target pressure, updates a control condition, which is a combination of a state and an action, based on the reward, selects an optimum action corresponding to the current state under the updated control condition, and controls the rotation speed of the pump based on the optimum action.

Abstract: In a continuous kneading apparatus according to an embodiment, for each of a plurality of ring-shaped heaters, a control unit determines a current state and a reward for an action selected in the past based on a control error calculated from an acquired temperature; updates a control condition based on the reward, and determines an optimum action corresponding to the current state under the updated control condition, the control condition being a combination of a state and an action; and controls a target ring-shaped heater based on the optimum action.

Abstract: In a resin-film manufacturing apparatus of an embodiment, when a lip gap control is started, a difference between a state of a target heat bolt (THB) at a start of current control and a state of each of all the heat bolts at a start of last control is obtained, when the difference between the states of THB at the starts of the current and last control is a smallest one, a learning result of THB is set as an initial value of a control condition for THB, and when the difference between the states of THB is not the smallest one, either one of a learning result of the heat bolt of which the difference between the states is smaller than the difference between the states of THB and the learning result of THB is set as the initial value of the control condition for THB.

Abstract: In a resin film manufacturing device according to an embodiment, a current state and a reward for a previously selected action are determined for each heat bolt based on a control error calculated from a thickness distribution of a resin film acquired from a thickness sensor. Then, control conditions, which are a combination of states and actions, are updated based on the reward, and a most appropriate action corresponding to the current state is selected from the updated control conditions. Then, the heater is controlled based on the most appropriate action.