Detection apparatus and method for steel rail failures based on characteristics of non-negative tensor resolution
A technology of non-negative tensor decomposition and detection method, which is applied in the field of detection devices for railway rail damage, and can solve problems such as difficulty in distinguishing damage characteristics.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
specific Embodiment approach 1
[0022] Specific implementation mode one: combine figure 1 Describe this embodiment, a rail damage detection device based on non-negative tensor decomposition vibration characteristics, which includes a vibration acceleration sensor array 1, a signal conversion circuit 2, a signal feature extraction module 3, a damage matching module 4 and a damage discrimination module 5 , the acquisition signal output end of the vibration acceleration sensor array 1 arranged on the steel rail to be tested is connected to the signal input end of the signal conversion circuit 2, and the signal output end of the signal conversion circuit 2 is connected to the signal input end of the signal feature extraction module 3, and the signal The signal output terminal of the feature extraction module 3 is connected to the signal input terminal of the damage matching module 4 , and the signal output terminal of the damage matching module 4 is connected to the signal input terminal of the damage discriminat...
specific Embodiment approach 2
[0023] Specific implementation mode two: combination figure 2 Describe this embodiment, the specific steps of this embodiment are as follows:
[0024] Step 1: The vibration acceleration sensor array 1 collects the original vibration signal S(t) of the steel rail under test;
[0025] Step 2: the signal conversion circuit 2 converts the original vibration signal S(t) into a vibration signal x(t);
[0026] Step 3: Decompose the vibration signal x(t) by HHT to obtain the internal solid mode function (IMF) and residual of each order, denoted as: c 1 ,c 2 ,... c n And r, n is a natural number, and then the corresponding Hilbert transform is performed to obtain the amplitude-frequency signal h(w) between the frequency and the amplitude;
[0027] Step 4: Utilize HHT to analyze the amplitude-frequency signal h(w) between the obtained frequency amplitudes, and combine the corresponding measurement points and damage types to establish a three-dimensional tensor of damage signals;
...
specific Embodiment approach 3
[0029] Specific implementation mode three: combination image 3 To illustrate this embodiment, the sub-steps included in Step 3 of Embodiment 2 are as follows:
[0030] Sub-step 1, find out the maximum value and minimum value of x(t);
[0031] Sub-step 2, generate the local maximum value and local minimum value envelope of the signal respectively by the cubic spline interpolation method;
[0032] Sub-step 3, adding and averaging the local maximum envelope and the local minimum envelope of the signal to obtain the local envelope mean value m(t);
[0033] Sub-step 4. Subtract the local envelope mean from the signal: h(t)=x(t)-m(t), continue to perform the above screening process on h(t) several times until the two conditions of IMF are satisfied: Condition 1. In the whole function, the number of extreme points is equal to the number of crossing zero points, or the difference is only 1; Condition 2. At any time, the mean value of the envelope defined by the extreme envelope is ...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 