Abstract:
An anomaly detection apparatus estimates time series data of a first signal of each appliance by disaggregating an aggregate signal that is a sum of first signals of a plurality of appliances into each first signal of an individual appliance, calculates a residual by subtracting sum of the estimated first signals of the plurality of appliances from the aggregate signal, and obtains value of the second signal at a time point at which the residual indicates presence of anomaly, and checks if there is match of pair of value of the estimated first signal and estimated state of the appliance, with any one of the one or more pairs of value of the first signal and state of the appliance, stored in a table, in association with the value of the second signal at the time point at which the residual indicates presence of anomaly to identify appliance with anomaly.
Abstract:
Concerning a thermoelectric conversion element, it is desired to provide a new spin current to charge current conversion material. A thermoelectric conversion element includes a magnetic layer possessing in-plane magnetization, and an electromotive layer magnetically coupled to the magnetic layer. The electromotive layer is formed of a carbon material, possesses anisotropy of electric conductivity, and further includes an additive.
Abstract:
A thermoelectric conversion element includes a magnetic layer which has a component magnetized in an in-plane direction, an electromotive layer which includes a material with spin orbit coupling, and a spin injection layer. The spin injection layer is provided between the magnetic layer and the electromotive layer and magnetically coupled to both the magnetic layer and the electromotive layer. A magnetic moment per unit volume of the spin injection layer is smaller than the magnetic moment per unit volume of the magnetic layer.