Abstract: In a rotary machine mounted on a rotary shaft journaled in journal bearings, a source of unusual vibration of the rotary machine is identified on the basis of the results of the steps of (1) subjecting a shaft vibration signal to frequency analysis for analyzing its frequency distribution, and (2) analyzing the amplitude of the shaft vibration signal and the variation rate of a phase signal of a component related to the shaft rotation frequency among the frequency components of the shaft vibration signal. The present invention is usefully applicable to diagnosis of vibration of a rotary shaft of a large-size turbine or a generator in a power plant by continuously monitoring the shaft vibration for immediately automatically detecting unusual vibration of the rotary machine in the event of occurrence of such unusual vibration.

Abstract: A system and method for the adaptive control of a process in which a feed forward control signal corresponding to a process demand is calculated according to a predetermined algebraic function, while a feedback correction signal is calculated on the basis of an error of a process feedback signal indicative of an error of a controlled variable from a predetermined setting, and the controlled variable of the process is controlled on the basis of the sum of these two signals. The adaptive control is such that, when a set point of the function deviates from the actual process demand, a value corresponding to the error appears in the feedback correction signal, and this value is used for automatically modifying the function itself to ensure the adaptive control of the process. A determination is made whether or not the process is in the steady state and when steady state operation is determined, the function of the feed forward control signal is modified.

Abstract: The present invention relates to a method of controlling the operation of large rotary machines such as steam turbines and generators in a thermal power plant or nuclear power plant. More specifically, the invention relates to a method of controlling such operation as increasing or decreasing the speed of the rotary machines or stopping the operation of the rotary machines, by detecting vibration at the time of starting the operation or during the steady-speed operation, by analyzing the detected vibration signals to determine whether the operation is in normal condition or abnormal condition, and by detecting or forecasting the cause in case the operation is in abnormal condition.

Abstract: In an automatic control system, controls receive an input indicative of a deviation between feedback and demand signals and produce an output which in turn is applied to a plant to control it. A diagnostic device includes a model having an equivalent characteristic to the controls and receives the deviation signal. When a noise coming into the automatic control system is detected or when a deviation between the outputs of the model and controls is sufficiently small, no model operation is performed, and the output of the controls is determined as that of the model. However, when the deviation between the outputs of the model and the controls is large, the model is supplied with an input thereby performing a model operation. A resultant output of the model is then compared with the output of the controls to produce an alarm indicative of an abnormal state of the controls in accordance with a deviation between these outputs.

Abstract: A fault-detecting apparatus comprises a model simulating the input-output characteristics of an object for fault detection. The model includes statics and dynamics and is impressed with the same input as the object for fault detection. The output of the model is compared with the output of the object for fault detection, and if the deviation is great, it is indicated that a fault has generated in the object for fault detection. In order to determine whether or not correction of the model is possible, the degree of the deviation is determined from an average value of the time-series signals produced from various parts of the fault-detecting apparatus during a certain period of time. For this purpose, a variation is calculated, for example; and if it is smaller than a predetermined value V.sub.L, a normal condition is judged and the statics correction is considered possible, while if the variation is larger than V.sub.H (which in turn is larger than V.sub.