Controller
A controller includes: a noise reduction processing unit that acquires a sensor signal, which is based on an output from a sensor that detects time-series data and includes a noise, and that reduces the noise included in the sensor signal on the basis of a recurrent neural network trained so as to learn a correspondence relationship between a first signal including the noise corresponding to the sensor signal and a second signal indicating the first signal from which the noise has been removed; and a control processing unit that controls an actuator on the basis of an output from the noise reduction processing unit.
1. A controller comprising:
a noise reduction processing unit that acquires a sensor signal, which is based on an output from a state quantity sensor that detects time-series data, said time-series data comprising sequential measured values relating to a state quantity of a vehicle and including a noise, and that reduces the noise included in the sensor signal on the basis of a recurrent neural network trained so as to learn a correspondence relationship between a first signal including the noise corresponding to the sensor signal and a second signal indicating the first signal from which the noise has been removed;
wherein the controller is configured to control an actuator on the basis of an output from the noise reduction processing unit.
2. The controller according to claim 1 , wherein
the state quantity sensor includes a displacement sensor that detects the time-series data relating to a vertical displacement of the vehicle as the state quantity of the vehicle, and
the actuator includes a suspension actuator that controls a suspension of the vehicle.
3. The controller according to claim 1 , wherein:
the controller is configured to execute signal processing on the output from the state quantity sensor, and
the noise reduction processing unit acquires the output on which the signal processing has been executed as the sensor signal and reduces at least the noise generated when the state quantity sensor detects the time-series data and the noise generated due to the signal processing included in the sensor signal.
4. The controller according to claim 3 , wherein
the controller executes differential processing or integration processing as the signal processing.
5. The controller according to claim 1 , wherein:
the controller is configured to execute signal processing on the sensor signal which is output from the noise reduction processing unit and from which the noise has been reduced,
the noise reduction processing unit acquires the output from the state quantity sensor as the sensor signal and reduces at least the noise included in the sensor signal and generated when the state quantity sensor detects the time-series data, and
the controller controls the actuator on the basis of the sensor signal on which the signal processing has been executed.
6. The controller according to claim 1 , wherein
the recurrent neural network is configured by a sequence to sequence (Seq2Seq) model based on long short-term memory (LSTM).