Method for steady-state detection and related equipment

Abstract

The embodiment of the invention discloses a method for steady-state detection and related equipment. The method and equipment are used for carrying out the steady-state detection on time series data based on a Gaussian function, can overcome noise influences, reduce complexity of the steady-state detection, improve accuracy of the steady-state detection, and improve efficiency of the steady-statedetection. The method comprises the steps that object time series data is acquired; the object time series data is divided into data sections of preset quantity, and each data section in the data sections of the preset quantity comprises an overlapping part with an adjacent data section; each data section in the data sections of the preset quantity acquires an object fitting curve through the Gaussian function; it is judged whether or not an object moment point in each object fitting curve satisfies a preset condition; if yes, the object moment point is determined to be in the steady state.

Description

technical field [0001] The present application relates to the field of computers, in particular to a steady-state detection method and related equipment. Background technique [0002] When performing data-driven system modeling, process identification, and performance evaluation, it is an important task to check whether the target object is in a steady state based on time series data. For example, in order to optimize the feed-forward strategy in the thermal power unit control system, it is necessary to establish a static model of the relevant variables, and the static model is based on the steady-state time series data, that is, the time series data in the steady state, so steady-state detection is required . For another example, in order to optimize the feedback strategy of the thermal power unit control system, a dynamic model in the identification process is required. Although the dynamic model in the identification process is based on dynamic time series data, it is re...

AI Technical Summary

Problems solved by technology

The steady-state detection methods of time series data include various methods based on statistical analysis, wavelet transform or fuzzy set, etc. However, these methods are either too complicated or too sensitive to signal noise, thus limiting the scope of application of these methods

Images