Abstract: A virtual metrology method using a convolutional neural network (CNN) is provided. In this method, a dynamic time warping (DTW) algorithm is used to delete unsimilar sets of process data, and adjust the sets of process data to be of the same length, thereby enabling the CNN to be used for virtual metrology. A virtual metrology model of the embodiments of the present invention includes several CNN models and a conjecture model, in which plural inputs of the CNN model are sets of time sequence data of respective parameters, and plural outputs of the CNN models are inputs to the conjecture model.

Abstract: An LED lamp device with fins having longitudinal and transversal channels includes a casing; an upper side of the casing having a cambered shape and a lower side of the casing being formed as a cavity; one end of the cavity being reduced as a neck; a plurality of fins form on an upper surface of the upper side of the casing for increasing heat dissipating areas; the fins being arranged along a longitudinal direction of the casing so as to form with a plurality of longitudinal air channels between the fins; at least one transversal air channel being formed to cut through the fins; by the arrangement of the longitudinal and transversal air paths, air being sufficient filled within the channels between fins; and a copper substrate fixed to the cavity of the casing; a plurality of screw holes being formed on the copper substrate for locking LED lamps.