Refrigeration method using an artificial neural network
AssigneeArianegroup SAS
InventorsSerge Le Gonidec, Charles-Hubert Bachelet, Sébastien Reichstadt, Anne-Laure Wozniak
Life of this patent
Estimating the term…
Loading patent…
AssigneeArianegroup SAS
InventorsSerge Le Gonidec, Charles-Hubert Bachelet, Sébastien Reichstadt, Anne-Laure Wozniak
Estimating the term…
A method for chilling down a cryogenic fluid supply circuit including a step of acquiring a state value of the cryogenic fluid supply circuit at a current time, a step of computing a corresponding opening time of each out of one or more valves of the cryogenic fluid supply circuit, and a step of opening each valve at the corresponding time to chill down the cryogenic fluid supply circuit. The step of computing is carried out by a computer applying an artificial neural network to compute a sequence of opening of the valves making it possible to reach, starting from the state value of the cryogenic fluid supply circuit at the current time, a state of the cryogenic fluid supply circuit meeting one or more predetermined chill down criteria at a predetermined time in the future.
A method for chilling down a cryogenic fluid supply circuit comprising:
A control unit of a cryogenic fluid supply circuit, configured to be connected to one or more sensors to acquire a state value of the cryogenic fluid supply circuit at a current time and to one or more valves of the cryogenic fluid supply circuit to control their opening, and comprising a computer programmed to apply an artificial neural network, preliminarily trained based on experimental data, to the computing of a corresponding time of opening of each valve to reach, starting from the state value of the cryogenic fluid supply circuit at a current time, a state of the cryogenic fluid supply circuit meeting one or more predetermined chill down criteria at a predetermined time in a future.
A cryogenic fluid supply circuit comprising one or more sensors, one or more valves, and a control unit connected to the sensors to acquire a state value of the cryogenic fluid supply circuit at a current time and to the valves to control opening of the valves, the control unit comprising a computer programmed to apply an artificial neural network, preliminary trained on experimental data, to the computing of a corresponding time of opening of each valve to reach, starting from a state value of the cryogenic fluid supply circuit at a current time, a state of the cryogenic fluid supply circuit meeting one or more predetermined chill down criteria at a predetermined time in a future.
Dependencies are inferred from the claim text.