Rotary Drive Frequency Control Using Autocorrelation Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for controlling rotary drives in track construction machines lack precise and rapid frequency change detection, leading to delayed frequency changes and suboptimal phase stabilization.
Innovation Solution
The method involves forming a series of time-discrete measured values and using autocorrelation and cross-correlation functions to determine frequency and phase shifts in real-time, with iterative calculations and signal filtering to enhance accuracy and reduce computational effort, allowing for precise synchronization of rotary drives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional zero-crossing detection methods are used to determine frequency, then the measurement system is simple, but the frequency change detection is delayed and imprecise
Solution Approach 1:
The patent replaces conventional zero-crossing detection methods with autocorrelation-based frequency analysis. Instead of mechanically detecting zero-crossing points in the signal, the system uses mathematical autocorrelation functions to determine frequency, enabling precise and rapid detection of frequency changes without the delays inherent in zero-crossing methods.
Solution Approach 2:
The patent performs preliminary signal processing by forming time-discrete measured value series and calculating autocorrelation functions in advance. This allows the system to have frequency detection results ready before actual frequency changes occur, reducing detection time and enabling proactive frequency control adjustments.
2Measurement precision
If autocorrelation and cross-correlation calculations are performed in real-time with high accuracy, then frequency and phase detection precision is improved, but computational effort increases
Solution Approach 1:
The patent segments the correlation function calculation into manageable time-discrete measured value series. By processing signals in discrete time steps and using iterative calculation methods, the system breaks down the complex autocorrelation and cross-correlation computations into smaller, more efficient operations that reduce overall computational power requirements.
Solution Approach 2:
The patent implements partial correlation calculations by using a predetermined number of measured values for correlation computations. Instead of calculating correlations over the entire signal history, the system uses a limited window of recent measured values, which provides sufficient accuracy for frequency and phase detection while significantly reducing computational effort.
3Measurement precision
If the cycle time for forming measured value series is reduced to improve detection accuracy, then frequency detection precision is improved, but the computational frequency requirement increases
Solution Approach 1:
The patent uses a predetermined, limited number of measured values for each correlation calculation rather than processing entire signal sequences. This partial action approach maintains high detection accuracy with shorter cycle times while keeping the computational load per cycle manageable, allowing the system to achieve high processing rates without sacrificing precision.
Solution Approach 2:
The patent implements periodic calculation of autocorrelation and cross-correlation functions at defined cycle intervals. This periodic action allows the system to maintain accurate frequency detection by regularly updating measurements while managing computational resources efficiently through structured, periodic processing rather than continuous high-frequency calculations.
Data Source
Figure 1~3
Figure 4~6
Figure 7~8
AI summary
The invention relates to a method for controlling/regulating a rotary drive (13) of a work unit (4) of a track laying machine (1), in which a measurement variable (X) with an approximately periodic history function derived from a rotation of the drive (13) is detected by means of a sensor (19), a frequency (f) or period duration (T) of the history function being determined by means of an evaluation unit (21) and the frequency (f) or period duration (T) being compared with a target variable for predetermining an actuation signal. In this case a series of time-discrete measurement values (xi) is formed for the measurement variable (X), an autocorrelation of these measurement values (xi) for determining the frequency (f) or period duration (T) being carried out by means of a computer unit (22). Thus, in contrast to a conventional method with zero crossing detection, an accurate detection of frequency changes is also possible between two zero crossings.