Ambient temperature estimating device, ambient temperature estimating method, program and system
Provided is an ambient temperature estimating device, ambient temperature estimating method, program, and system that are able to realize both high robustness and high ambient temperature estimation accuracy. An ambient temperature estimating device includes a neural network, a temperature acquisition unit configured to acquire one or more temperature values inside the ambient temperature estimating device, and a neural network calculator configured to estimate an ambient temperature around the ambient temperature estimating device using the neural network. Input values inputted to the neural network by the neural network calculator include the temperature values acquired by the temperature acquisition unit and a heat source control value for controlling a heat source inside the ambient temperature estimating device.
1. An ambient temperature estimating device comprising:
a neural network;
a temperature acquisition unit configured to acquire one or more temperature values inside the ambient temperature estimating device;
a neural network calculator configured to estimate an ambient temperature around the ambient temperature estimating device using the neural network; and
a cooling controller configured to control an inside of the ambient temperature estimating device based on a cooling control value,
wherein the cooling controller is configured to dynamically control the cooling control value,
input values inputted to the neural network by the neural network calculator include the temperature values acquired by the temperature acquisition unit, a heat source control value for controlling a heat source inside the ambient temperature estimating device, and the cooling control value, and
the cooling control value includes a set value between a minimum set value and a maximum set value.
2. The ambient temperature estimating device of claim 1 , wherein the input values include the amount of change in the temperature values in a predetermined period.
3. The ambient temperature estimating device of claim 1 , wherein the heat source is a backlight or an internal circuit.
4. The ambient temperature estimating device of claim 1 , wherein the input values include an energizing time of at least one of the ambient temperature estimating device and the heat source.
5. The ambient temperature estimating device of claim 1 , wherein
the neural network includes a plurality of calculation nodes,
predetermined weights are set for the respective calculation nodes, and
the weights are set through machine learning previously performed by another information processing device or through machine learning performed by the ambient temperature estimating device.
6. An ambient temperature estimating method comprising:
a temperature acquisition step of acquiring, by a temperature acquisition unit, one or more temperature values inside an ambient temperature estimating device;
a neural network calculation step of estimating, by a neural network calculator, an ambient temperature around the ambient temperature estimating device using a neural network; and
a cooling controlling step of controlling, by a cooling controller, an inside of the ambient temperature estimating device based on a cooling control value, wherein
in cooling controlling step, the cooling control value is dynamically controlled,
input values inputted to the neural network include
the temperature values acquired by the temperature acquisition unit,
a heat source control value for controlling a heat source inside the ambient temperature estimating device, and
the cooling control value,
the cooling control value includes a set value between a minimum set value and a maximum set value.
7. A non-transitory computer readable medium that stores a program for causing a computer to function as:
a neural network;
a temperature acquisition unit configured to acquire one or more temperature values inside an ambient temperature estimating device;
a neural network calculator configured to estimate an ambient temperature around the ambient temperature estimating device using the neural network
a cooling controller configured to control an inside of the ambient temperature estimating device based on a cooling control value, wherein
the cooling controller is configured to dynamically control the cooling control value,
input values inputted to the neural network by the neural network calculator include
the temperature values acquired by the temperature acquisition unit,
a heat source control value for controlling a heat source inside the ambient temperature estimating device, and
the cooling control value,
the cooling control value includes a set value between a minimum set value and a maximum set value.
8. A system comprising:
the ambient temperature estimating device of claim 1 ; and
an information processing device, wherein
the ambient temperature estimating device and the information processing device include communication units configured to be able to communicate data to each other and neural networks,
the neural networks each include a plurality of calculation nodes,
predetermined weights are set for the respective calculation nodes,
the information processing device is configured to acquire, through the communication unit, temperature values acquired by the temperature acquisition unit of the ambient temperature estimating device, a heat source control value for controlling a heat source inside the ambient temperature estimating device, and a cooling control value for cooling control an inside of the ambient temperature estimating device,
the calculation nodes and the weights are determined by machine leaning of the temperature values, the heat source control value, and the cooling control value using the neural network of the information processing device, the temperature values from the ambient temperature estimating device, the heat source control value, and the cooling control value being acquired by the communication unit of the information processing device, and
the ambient temperature estimating device is configured to acquire the determined weights through the communication unit.