摘要 |
PROBLEM TO BE SOLVED: To improve accuracy of recognition of speech which is input to a control device of a plant such as power generation. SOLUTION: A distributed speech recognition system includes a plurality of control devices 2 including: a control section in which input speech from a speech output section 1 is converted to a speech signal and recognized, and if instructions to an auxiliary device 9 are included in the speech, a data for control is output to the auxiliary device 9 according to the instructions; and a storage section for storing the instructions to the auxiliary device 9, and algorithm for determining the data for control, by relating them. A time data for showing time when the speech is input to the speech signal are added in each control device 2. One control device 2 decides whether the input speech signal is normal, by comparing the speech signal to which the time data included in one control device 2, with the speech signal which is obtained from the other control device 2, and in which the time indicated by the time data approximately matches. COPYRIGHT: (C)2010,JPO&INPIT |