CN113358380B - Rail vehicle snaking motion stability detection and evaluation method - Google Patents
Rail vehicle snaking motion stability detection and evaluation method Download PDFInfo
- Publication number
- CN113358380B CN113358380B CN202110584757.XA CN202110584757A CN113358380B CN 113358380 B CN113358380 B CN 113358380B CN 202110584757 A CN202110584757 A CN 202110584757A CN 113358380 B CN113358380 B CN 113358380B
- Authority
- CN
- China
- Prior art keywords
- frequency
- sequence
- snake
- signal sig
- time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M17/00—Testing of vehicles
- G01M17/08—Railway vehicles
Abstract
The invention provides a method for detecting and evaluating the snake movement stability of a railway vehicle, which aims at solving the problems of larger energy loss and inaccurate peak value extraction caused by narrower filter frequency band in the existing snake movement detection method, and firstly filters a sampling signal based on the high-order odd harmonic frequency characteristic of the snake movement to obtain a broadband filter signal; and carrying out snake main frequency identification on the filtered signal within a preset frequency identification range to obtain the snake main frequency; then, filtering the filtered signal again to obtain a snake main frequency signal; then continuous peak extraction in a time domain is carried out on the snake main frequency signal to obtain a first peak sequence and a corresponding time sequence; and extracting continuous peak values of the filtered signals by using a snake main frequency oscillation period proportion rule based on the time sequence so as to obtain a second peak value sequence, and judging the stability of the vehicle based on the second peak value sequence. The method provided by the invention can improve the discrimination sensitivity and the accuracy of peak value extraction.
Description
Technical Field
The invention belongs to the field of monitoring of running safety states of running parts of railway vehicles, and particularly relates to a method for detecting and evaluating snake running motion stability of a railway vehicle.
Background
Wheel-set hunting is one of the inherent characteristics of rail vehicles and is primarily caused by the rail vehicle wheel-rail contact geometry, which cannot be fundamentally eliminated. When the running speed of the vehicle exceeds the critical speed of the snaking motion, the transverse and shaking motion of the bogie and the vehicle can develop into a destabilization form with gradually enlarged amplitude, and great safety risk is brought to the running of the vehicle. Therefore, monitoring the snaking state of the bogie has important significance for guaranteeing and improving the running safety of the vehicle.
At present, the judgment of the snaking motion stability in China is based on the regulation in the Standard GB/T5599-2019 'locomotive vehicle dynamics performance evaluation and test identification Specification', and the transverse vibration acceleration of a bogie frame is adopted to evaluate whether the train bogie generates continuous transverse vibration which can not be rapidly attenuated.
The specific implementation mode is as follows: acceleration of vibration of frame above counter-shaft boxCarrying out real-time continuous sampling, carrying out band-pass filtering by using 0.5-10 Hz, and if the acceleration peak value is continuously reached or exceeded 8m/s for more than 6 times2And judging the transverse instability of the bogie.
And (3) performing band-pass filtering on the transverse vibration acceleration of the bogie frame, and then extracting a signal peak value after filtering as a basis to evaluate whether the bogie generates continuous transverse oscillation which cannot be rapidly attenuated. The method is a common method for evaluating the snaking motion stability of the bogie at home and abroad at present, but researches show that the method can be further perfected on the following two aspects in order to improve the sensitivity of monitoring the snaking motion stability:
first, selection of band pass filtered bands is relevant. Because the snaking frequency of the bogie in the transverse instability is generally low, the bandwidth of a band-pass filtering frequency band is reduced as much as possible in the conventional method at present, so that the frequency characteristics of the bogie after the snaking motion instability can be reflected by the narrow-band filtered signal, but the filtered signal has larger energy loss compared with the original signal along with the narrowing of the band-pass filtering frequency band, and the amplitude of the filtered signal hardly reflects the real snaking vibration energy level of the bogie, so that the sensitivity of the conventional detection method for identifying the small-amplitude snaking motion state is low, and finally the vehicle is caused to continuously run in the small-amplitude snaking state, which brings threat to the running safety of the vehicle.
Second, the rule for acceleration peak extraction. Because the post-filtering signal peak value extraction criterion is not determined in the standard, technicians mostly adopt a method for extracting the maximum value of the acceleration in a fixed-length time window when performing peak value extraction on the acceleration signal at present, the method has the defect that the length of the time window cannot be determined, if the time window is too small, continuous low-frequency high-amplitude noise may bring about misjudgment of a result, and if the time window is too large, the detection system is caused to have slow response to judgment of the snaking motion stability. Meanwhile, when the running speed of the vehicle changes, the snaking frequency of the bogie changes, the transverse vibration of the bogie is a non-stationary random process, and the maximum value of the fixed-length time window extraction signal cannot be quantized and considered. In addition, when the framework vibration signal is filtered, because the filtering is not carried out aiming at the snakelike fundamental frequency, but the band-pass filtering is carried out on the low-frequency band of 0.5-10 Hz, signals except the snakelike fundamental frequency are inevitably introduced, and when the peak value extraction is carried out on the filtered vibration signal, if the peak value is not distinguished, the peak value caused by noise, namely the false peak value, is easily extracted.
In conclusion, the energy loss of the filtered signal is large, or the extracted peak value is not accurate enough, so that the sensitivity of judging the snaking motion stability of the rail vehicle needs to be improved.
Disclosure of Invention
In order to solve the problems, the invention provides a rail vehicle stability judgment method which can reduce energy loss corresponding to a filtered signal and accurately extract a peak value, and the invention adopts the following technical scheme:
the invention provides a method for detecting and evaluating the snaking motion stability of a rail vehicle, which is used for evaluating the snaking motion stability of the rail vehicle in the running process and is characterized by comprising the following steps: step S1, acquiring a frame transverse vibration acceleration signal of a frame right above two axle boxes at the diagonal angle of the frame in the running process of the rail vehicle in real time by using an acceleration sensor, and sampling the frame transverse vibration acceleration signal to obtain an original sampling signal Sig _ 0; step S2, based on the characteristics of snake motion high-order odd harmonic frequency, broadband band-pass filtering of f1-f2 frequency bands is carried out on an original sampling signal Sig _0 through a preset broadband band-pass filtering method, and a filtered signal Sig _1 is obtained; step S3, carrying out snake main frequency identification on the filtered signal Sig _1 in a preset frequency identification range to obtain a snake main frequency f _ h; step S4, performing broadband band-pass filtering on (f _ h-f _ w) to (f _ h + f _ w) frequency bands on the filtered signal Sig _1 to obtain a frame snake main frequency signal Sig _2, wherein f _ w is a preset bandwidth; step S5, extracting continuous peak values in a time domain from the frame snake main frequency signal Sig _2 to obtain a first peak value sequence PX, and acquiring a corresponding time sequence PT according to the first peak value sequence; step S6, continuous peak value extraction is carried out on the filtered signal Sig _1 by utilizing a preset snake main frequency oscillation period proportion rule and based on the time sequence PT, so as to obtain a second peak value sequence PY; step S7, judging whether N continuous peaks in the second peak sequence PY are not less than a threshold value Ac; step S8, if the judgment in step S7 is no, the bogie of the railway vehicle is judged to be in a state with normal stability in the snaking process, and corresponding stability information of the bogie in the snaking process is output; and step S9, if the judgment in the step S7 is yes, the situation that the stability of the bogie of the railway vehicle is in an abnormal state in the snaking process is judged, and corresponding instability alarm information of the bogie in the snaking process is output.
The method for detecting and evaluating the snaking motion stability of the rail vehicle provided by the invention can also have the technical characteristics that the broadband band-pass filtering method is any one of a filter or a time-frequency decomposition reconstruction method, and the time-frequency decomposition reconstruction method is at least any one of EMD, EEMD and wavelet packet decomposition.
The method for detecting and evaluating the snaking motion stability of the railway vehicle provided by the invention can also have the technical characteristics that the step S2 comprises the following sub-steps: step S2-1, performing time-frequency decomposition on the original sampling signal Sig _0 by using a time-frequency decomposition reconstruction method to obtain sub-signals of the original sampling signal Sig _0 in different frequency bands; step S2-2, performing frequency spectrum analysis on the sub-signals, and identifying to obtain a main frequency position corresponding to each sub-signal; and S2-3, reconstructing the sub-signals with the main frequency positions located at the frequency bands of f1-f2 to obtain the filtered signals Sig _ 1.
According to the method for detecting and evaluating the snake motion stability of the railway vehicle, the technical characteristics can be further provided, wherein the snake motion high-order odd harmonic frequency characteristic is that snake fundamental frequency and high-order odd harmonic frequency of the snake fundamental frequency exist in a framework transverse vibration frequency spectrum, and the frequency bands from f1 to f2 are frequency bands expanded to the high-order odd harmonic frequency.
The method for detecting and evaluating the snaking motion stability of the railway vehicle provided by the invention can also have the technical characteristics that the step S5 comprises the following sub-steps: s5-1, screening out sequence points with the primary difference larger than 0 and the secondary difference smaller than 0 from the sequence points of the frame snake main frequency signal Sig _2 as maximum value points; s5-2, taking the inverse number of the frame snake main frequency signal Sig _2 to obtain an opposite main frequency signal Sig _ 2'; s5-3, screening out sequence points with the primary difference larger than 0 and the secondary difference smaller than 0 from the sequence points of the opposite main frequency signals Sig _2' as minimum value points; s5-4, removing the maximum value points which are negative numbers in all the maximum value points to form a new maximum value point sequence, and removing the minimum value points which are positive numbers in all the minimum value points to form a new minimum value point sequence; step S5-5, judging whether each maximum point in the new maximum point sequence and each minimum point in the new minimum point sequence are adjacent and spaced on the time sequence, and if so, taking the new maximum point sequence as a first peak value sequence PX; s5-6, when the judgment in the step S5-5 is negative, reserving a larger maximum value point of all adjacent two maximum value points in the new maximum value point sequence, and thus obtaining a denoised maximum value sequence as a first peak value sequence PX; step S5-7, a first peak sequence PX is obtained, and the time corresponding to the maximum point in the first peak sequence PX is taken as a time sequence PT.
The method for detecting and evaluating the snaking motion stability of the railway vehicle provided by the invention can also have the technical characteristics that the step S6 comprises the following sub-steps: step S6-1, calibrating the filtered signal Sig _1 based on the time sequence PT to obtain a time calibration point; and step S6-2, extracting the maximum acceleration value of the filtered signal Sig _1 by using a snake dominant frequency oscillation period proportion rule, taking a time calibration point as a center and a half period smaller than the snake movement fundamental frequency as a search time range, so as to obtain all the maximum acceleration values as a peak sequence PY.
Action and Effect of the invention
According to the method for detecting and evaluating the snake motion stability of the railway vehicle, the filtered signal Sig _1 is obtained by filtering the original sampling signal Sig _0 after the filtering frequency band is expanded based on the characteristic of the snake motion high-order odd harmonic frequency, and the energy in the expanded filtering frequency band mainly consists of the odd harmonic frequency caused by the snake motion, so that the filtered signal Sig _1 contains more abundant snake motion information, the problem of large signal energy loss caused by narrow-band filtering is solved, and the sensitivity of the monitoring method for judging the snake motion state is further improved. In addition, because the second peak sequence PY is obtained by carrying out continuous peak extraction on the filtered signal Sig _1 based on the time sequence PT by utilizing the preset snake main frequency oscillation period proportion rule, the possibility that false peaks exist in the peak extraction of the signal containing strong noise can be reduced, the extracted peaks have strong correlation with the snake vibration main frequency, and the accuracy of stability judgment is improved.
Drawings
FIG. 1 is a flow chart of a method for detecting and assessing hunting stability of a rail vehicle according to an embodiment of the present invention;
FIG. 2 is a diagram of simulation results for a wheel-set model according to an embodiment of the present invention;
FIG. 3 is a time domain plot of wheel-track linear and non-linear contact lateral acceleration for a wheel-set model according to an embodiment of the present invention;
fig. 4 is a time domain diagram and a frequency spectrum diagram of an original sampling signal Sig _0 of the entire vehicle model according to the embodiment of the present invention;
FIG. 5 is a flowchart of the substep of step S2 according to an embodiment of the present invention;
FIG. 6 is a time-frequency decomposition result diagram of EEMD performed on the original sampling signal Sig _0 according to the embodiment of the present invention;
FIG. 7 is a time domain diagram and a frequency spectrum diagram of the filtered signal Sig _1 according to the embodiment of the present invention;
FIG. 8 is a flowchart of the substep of step S5 according to an embodiment of the present invention;
fig. 9 is a time domain diagram and a frequency spectrum diagram of a framework meandering main frequency signal Sig _2 according to an embodiment of the present invention;
FIG. 10 is a flowchart of the substep of step S6 according to an embodiment of the present invention;
fig. 11 is a diagram illustrating a second peak sequence PY according to an embodiment of the invention.
FIG. 12 is an analysis graph of the range of the filter band and the mean of the extracted peaks according to an embodiment of the present invention;
FIG. 13 is an analysis graph of the filter band range and the peak increase factor according to an embodiment of the present invention.
Detailed Description
In order to make the technical means, the creation characteristics, the achievement purposes and the effects of the invention easy to understand, the following embodiment and the accompanying drawings are used to specifically describe the detection and evaluation method for the snaking motion stability of the rail vehicle.
< example >
In order to illustrate the method for detecting and evaluating the snaking motion stability of the rail vehicle, the embodiment specifically demonstrates based on a dynamic model of a high-speed vehicle of a certain type in China, and the dynamic model of the vehicle (referred to as a whole vehicle model for short) is established by simulation software Simpack based on the dynamics of the rail vehicle.
In the embodiment, the parameters of the secondary anti-snaking damper are adjusted, so that the vehicle is in a snaking motion instability state at the speed of 300 km/h.
Fig. 1 is a flowchart of a method for detecting and evaluating the hunting stability of a rail vehicle according to an embodiment of the present invention.
Referring to fig. 1, a method for detecting and evaluating the snaking motion stability of a rail vehicle includes the following steps:
step S1, acquiring a frame transverse vibration acceleration signal of the position right above two axle boxes at the frame diagonal angle in the running process of the rail vehicle in real time by using the acceleration sensor, and sampling the frame transverse vibration acceleration signal to obtain an original sampling signal Sig _ 0.
Wherein the sampling frequency for sampling the transverse vibration acceleration signal of the framework is fs. In this embodiment, the sampling frequency fs is 256 Hz.
And step S2, based on the characteristics of the snake motion high-order odd harmonic frequency, carrying out broadband band-pass filtering on the original sampling signal Sig _0 in the frequency ranges of f1-f2 by a preset broadband band-pass filtering method to obtain a filtered signal Sig _ 1.
The broadband band-pass filtering method is any one of a filter or a time-frequency decomposition reconstruction method. In this embodiment, the wideband band-pass filtering method is a time-frequency decomposition reconstruction method.
The time-frequency decomposition reconstruction method is at least any one of EMD, EEMD and wavelet packet decomposition. In this embodiment, the time-frequency decomposition reconstruction method is EEMD.
The snake motion high-order odd harmonic frequency characteristic is that a snake fundamental frequency and high-order odd harmonic frequencies of the snake fundamental frequency exist in a framework transverse vibration frequency spectrum, and the frequency band from f1 to f2 is a frequency band expanded to the high-order odd harmonic frequencies. Specifically, the method comprises the following steps:
the characteristics of the snake motion high-order odd harmonic frequency are specifically analyzed through simulation aiming at a model and comparison of wheel-track linear contact and non-linear contact:
first, a simulation of the dynamics of a free wheel set with non-linear wheel-rail contact.
Firstly, establishing a free wheel set dynamic model (wheel set model for short) based on an LMA abrasion type tread and a 60kg/m type track.
Fig. 2 is a simulation result diagram of the wheel set model according to the embodiment of the present invention.
Secondly, the running speed simulation is carried out on the wheel model at the speed of 60km/h, and the simulation result is shown in figure 2.
Fig. 2(a) is a time domain diagram of wheel set lateral shift, fig. 2(b) is a frequency spectrum diagram of wheel set lateral shift, fig. 2(c) is an acceleration time domain diagram, and fig. 2(d) is an acceleration frequency spectrum diagram.
Although the traversing amount waveform in fig. 2(a) is similar to a sinusoidal waveform, since the track-and-wheel contact is nonlinear, higher harmonic frequencies other than the fundamental meandering frequency are included in the traversing signal (as shown in fig. 2 (b)), except that the amplitude of the higher harmonic frequencies is small compared to the amplitude of the fundamental frequency.
In addition, referring to fig. 2(a) and 2(c), when the time point of fig. 2(a) at which the traverse amount is the largest corresponds to the same time point as fig. 2(c), the acceleration waveform has a rising peak, and at this time, the traverse of the wheel set reaches a strong nonlinear region, and lateral shock is generated between the wheel rails due to hunting instability, and the amplitude of the traverse of the wheel set is not dispersed due to the rim restraint.
From fig. 2(d), it can be seen that the higher harmonic frequency exists in addition to the snake-like fundamental frequency, the amplitude of the higher harmonic frequency is obviously increased compared with the transverse frequency spectrum, and the higher harmonic frequency value only exists in odd multiples of the fundamental frequency.
In summary, for the nonlinear wheel-track contact, when the wheel set generates the hunting instability, the transverse vibration of the wheel set includes a higher order odd harmonic frequency besides the fundamental frequency of the hunting, and the higher order odd harmonic frequency is smaller in the traverse signal but larger in the transverse vibration acceleration signal. At present, a monitoring method for the snake movement stability of a train mainly extracts and analyzes snake fundamental frequency, and attention to high-order odd harmonic frequency is lacked.
Second, the wheel-rail linear versus nonlinear contact lateral acceleration for the wheelset model is compared.
It is proposed to filter the snake fundamental frequency signal in fig. 2(c), which corresponds to the wheel-track linear contact hypothesis.
Fig. 3 is a time domain comparison graph of the wheel-track linear and non-linear contact lateral acceleration of the wheel-set model according to the embodiment of the invention.
As can be seen from fig. 3, the meandering fundamental frequency waveform shows an approximate sine wave characteristic in the whole time history, i.e. the energy of the meandering fundamental frequency is widely distributed in the whole time domain; the high-order harmonic frequency signal waveform can generate sharp rise of a peak value when traversing approaches to a maximum value, and at the moment, the wheel-rail contact reaches a strong nonlinear contact area, namely, the snake high-order harmonic frequency signal waveform is high in energy concentration and appears at a time point with a large traversing amount, and the high-order harmonic frequency signal waveform is an important expression of strong nonlinear impact between wheel-rails after the wheel set generates snake instability. Because the transverse vibration response of the framework is mainly caused by the snaking motion of the wheel pair, when the transverse vibration acceleration signal of the framework is filtered, the filtering frequency band can be widened to a certain order of odd harmonic frequency except the snaking fundamental frequency, so that the signal extracted by broadband filtering contains richer snaking motion information, and the sensitivity of subsequent snaking state judgment is improved.
Fig. 4 is a time domain diagram and a frequency spectrum diagram of an original sampling signal Sig _0 of the entire vehicle model according to the embodiment of the present invention.
Fig. 4(a) is a time domain diagram of the original sampling signal Sig _0, and fig. 4(b) is a spectrum diagram of the original sampling signal Sig _ 0.
As can be seen from FIG. 4(b), the salient energy of 0-10 Hz corresponds to the impact of the framework snake base frequency, the salient energy of 10-20 Hz corresponds to the impact of snake 3-th harmonic, the natural frequency of the framework exists at 30-40 Hz, the salient energy corresponds to the response of snake 7-th harmonic and the broadband track excitation impact, and in order to reduce the interference of the track excitation on the feature extraction, only the snake base frequency and the 3-th harmonic frequency are considered during the filtering, and the filtering frequency band is determined to be 0.5 Hz-25 Hz (i.e. the frequency band of f1-f 2).
Fig. 5 is a flowchart of the substep of step S2 according to an embodiment of the present invention.
As shown in fig. 5, step S2 includes the following sub-steps:
and step S2-1, performing time-frequency decomposition on the original sampling signal Sig _0 by using a time-frequency decomposition reconstruction method to obtain sub-signals of the original sampling signal Sig _0 in different frequency bands.
In this embodiment, an EEMD time-frequency decomposition reconstruction method is used to perform time-frequency decomposition on the original sampling signal Sig _0 to obtain a plurality of inherent modal function components (i.e., sub-signals).
FIG. 6 is a time-frequency decomposition result diagram of EEMD performed on the original sampling signal Sig _0 according to the embodiment of the present invention.
As shown in fig. 6, in the first 8 Intrinsic Mode Functions (IMFs) obtained by decomposing the original sampling signal Sig _0, the 1 st IMF corresponds to 30-40 Hz impulse response, the energy is caused by excitation of the framework transverse intrinsic frequency by snake 7 harmonic frequency and orbit excitation, the 2 nd and 3 rd IMFs correspond to snake 3 harmonic frequency impulse, the 4 th and 5 th IMFs correspond to snake fundamental frequency impulse, and the 8 th IMF has a main frequency less than 0.5Hz and is related to transverse vibration of the vehicle body.
And step S2-2, performing spectrum analysis on the sub-signals, and identifying to obtain a main frequency position corresponding to each sub-signal.
And S2-3, reconstructing the sub-signals with the main frequency positions located at the frequency bands of f1-f2 to obtain the filtered signals Sig _ 1.
Wherein, the frequency range of f1-f2 is 0.5 Hz-25 Hz, including snake-like fundamental frequency and 3 times of harmonic frequency.
In this embodiment, when filtering the original sampling signal Sig _0, the filtering frequency band is expanded to the 3 rd order snake odd harmonic frequency in addition to considering the snake motion fundamental frequency, so that the energy loss of the filtered signal is small, and the filtered signal contains richer snake motion information.
Fig. 7 is a time domain diagram and a frequency spectrum diagram of the filtered signal Sig _1 according to the embodiment of the present invention.
In this embodiment, the intrinsic mode function (i.e., the 2 nd to 7 th IMFs) between 0.5Hz and 25Hz is reconstructed to obtain the filtered signal Sig _1 (as shown in fig. 7).
Fig. 7(a) is a time domain diagram of the filtered signal Sig _1, and fig. 7(b) is a frequency domain diagram of the filtered signal Sig _ 1.
Compared with the original sampling signal Sig _0 (fig. 4 a), the peak value of the time domain waveform of Sig _1 after filtering Sig _1 (fig. 7 a) is lower than the peak value of the time domain waveform of Sig _0, but the impact amplitude and time domain phase are unchanged.
In addition, as seen in FIG. 7(b), the fundamental framework snake frequencies and the 3 harmonic frequencies are filtered out, while signals above 25Hz are substantially filtered out.
Step S3, performing snake main frequency identification on the filtered signal Sig _1 within a predetermined frequency identification range to obtain a snake main frequency f _ h.
In the embodiment, the frequency identification range is from f3 to f4, specifically from 0.5Hz to 10 Hz.
The filtered signal Sig _1 contains a snake fundamental frequency and 3 harmonic frequencies, which are both strongly related to snake motion, but due to the introduction of frequencies other than the snake fundamental frequency after the frequency band is expanded, the time domain waveform vibration is more complex, and many false peaks (peaks unrelated to snake dominant frequency impact) exist in the signal, so that a peak point with higher correlation with the snake dominant frequency signal needs to be extracted from the filtered signal Sig _ 1.
In this embodiment, as the hunting main frequency of the bogie generally appears at 0.5 to 10Hz, 0.5 to 10Hz hunting main frequency extraction is performed on the filtered signal Sig _1, and a frequency peak point is found at a frequency of 5.3Hz, which is the hunting main frequency f _ h.
Step S4, performing broadband band-pass filtering on (f _ h-f _ w) to (f _ h + f _ w) bands on the filtered signal Sig _1 to obtain a frame snake main frequency signal Sig _2, where f _ w is a predetermined bandwidth.
In this embodiment, f _ h is 5.3Hz, and f _ w is 1 Hz.
In addition, the wideband band-pass filtering method for the filtered signal Sig _1 is one of a filter and a time-frequency decomposition reconstruction method, and may be the same as or different from the wideband band-pass filtering method in step S2.
In this embodiment, the method of filtering the filtered signal Sig _1 with a broadband bandpass is a common filter.
Step S5, extracting consecutive peaks in the time domain from the frame snake main frequency signal Sig _2 to obtain a first peak sequence PX, and obtaining a corresponding time sequence PT according to the first peak sequence.
Fig. 8 is a flowchart of the substep of step S5 according to an embodiment of the present invention.
As shown in fig. 8, step S5 includes the following sub-steps:
and S5-1, screening out sequence points with the primary difference larger than 0 and the secondary difference smaller than 0 from the sequence points of the frame snake main frequency signal Sig _2 as maximum value points. Specifically, the method comprises the following steps:
suppose that the frame snake main frequency signal Sig _2 is X ═ X1,x2,…,xi,…,xn]Wherein n is the sequence length, [1, 2 …, i, …, n]Are corresponding time series.
Firstly, carrying out first-order forward difference on X to obtain a first difference sequence X':
X'＝[x2-x1,x3-x2,…,xi+1-xi,…,xn-xn-1]。
then judging whether the point in the difference sequence X' is greater than 0 or not, thereby obtaining a first difference judgment sequence J with each point having a judgment result1：
J1＝[a1,a2,…,ai,…,an-1]Wherein
In this example, ai1 is expressed as a point in the difference sequence X' greater than 0, ai0 denotes that the point in the difference sequence X' is not more than 0.
Then, the sequence J is differentially selected once1Then, a first-order forward difference is carried out to obtainSecond order difference sequence J1'：
J1'＝[a2-a1,a3-a2,…,ai+1-ai,…,an-1-an-2]。
Then, the second order difference sequence J is judged1' whether the point in is less than 0, thereby obtaining a second order difference sequence J1' two-time differential judgment sequence J having judgment result at each point2：
J2＝[b1,b2,…,bi,…,bn-2]Wherein
In this example, biExpressed as differential sequence J when 11Point of' less than 0, biExpressed as differential sequence J when 01The point in' is not less than 0.
Finally, a second difference judgment sequence J is obtained2In (b)iThe next sequence position i +1 corresponds to the value x in the frame snake main frequency signal Sig _2i+1The value xi+1I.e. the maximum value point, and i +1 is the corresponding point of the maximum value point in the time series (i.e. the maximum value corresponding point).
Step S5-2, the frame snake main frequency signal Sig _2 is inverted to obtain an opposite main frequency signal Sig _ 2'.
And step S5-3, screening out sequence points with the primary difference larger than 0 and the secondary difference smaller than 0 from the sequence points of the opposite main frequency signals Sig _2' as minimum value points.
And step S5-4, removing the negative maximum points from all the maximum points to form a new maximum point sequence, and removing the positive minimum points from all the minimum points to form a new minimum point sequence. Specifically, the method comprises the following steps:
firstly, judging whether all the maximum value points are positive numbers or not, removing the non-positive number maximum value points when the maximum value points are not positive numbers, and taking all the remaining maximum value points as a new maximum value point sequence. If yes, the flow proceeds to step S5-5.
And simultaneously, judging whether all the minimum value points are negative numbers, if not, removing the non-negative minimum value points, and taking all the remaining minimum value points as a new minimum value point sequence. If yes, the flow proceeds to step S5-5.
Step S5-5, determining whether each maximum point in the new maximum point sequence and each minimum point in the new minimum point sequence are adjacent in time sequence, and if yes, taking the new maximum point sequence as the first peak sequence PX.
And S5-6, when the judgment in the step S5-5 is negative, reserving the larger maximum value point of all the two adjacent maximum value points in the new maximum value point sequence, and thus obtaining the denoised maximum value sequence as a first peak value sequence PX.
Step S5-7, a first peak sequence PX is obtained ═ xi,xj,xk,…]And the time corresponding to the maximum point in the first peak sequence PX is taken as the time sequence PT ═ i, j, k, …]。
Fig. 9 is a time domain diagram and a frequency spectrum diagram of a framework meandering main frequency signal Sig _2 according to an embodiment of the present invention.
Fig. 9(a) is a time domain diagram of the framework meandering main frequency signal Sig _2, and fig. 9(b) is a spectrum diagram of the framework meandering main frequency signal Sig _ 2.
To describe the first peak sequence PX and the time sequence PT specifically, the amplitude of a random certain time period in the time domain diagram of the framework snake main frequency signal Sig _2 in fig. 9(a) is amplified locally, and from the local amplification of the time domain diagram of Sig _2, it can be seen that the maximum value in the first peak sequence PX corresponds to the time point in the time sequence PT.
As can be seen from the frequency spectrum of fig. 9(b), the framework snake main frequency signal is only related to the snake fundamental frequency, and no other frequency exists in the frequency spectrum, while the time domain signal of fig. 9(a) can also be seen, the signal oscillation period is obvious, and the noise contained in the signal is small, which indicates that Sig _2 has strong correlation with the snake main frequency; due to the narrow filtering band, the amplitude of the signal in fig. 9(a) is much lower than that of the original signal (fig. 4(a)), i.e., the vibration energy of the narrow-band filtered signal is difficult to reflect the vibration energy level of the original signal.
Step S6, using a predetermined snake main frequency oscillation period proportion rule and based on the time series PT, performs continuous peak extraction on the filtered signal Sig _1, thereby obtaining a second peak sequence PY.
Fig. 10 is a flowchart of the substep of step S6 according to an embodiment of the present invention.
As shown in fig. 10, step S6 includes the following sub-steps:
and step S6-1, calibrating the filtered signal Sig _1 based on the time sequence PT to obtain a time calibration point.
Since the frame crawling main frequency signal Sig _2 is obtained by band-pass filtering the filtered signal Sig _1, and the signals have no time delay during filtering, the vibration characteristics of the acceleration signals of the two signals have the same time phase, the impact of the corresponding frame crawling motion at each time point is the same, and the frame crawling main frequency signal Sig _2 is responded to the time histories of the signals Sig _1 and Sig _2 together.
In summary, the time sequence PT obtained by constructing the snake main frequency signal Sig _2 is used to perform peak value search in the time domain of the filtered signal Sig _1, so that the obtained second peak value sequence PY can better reflect the snake impact energy, and has strong correlation with the snake main frequency, and in addition, false peaks caused by noise are also removed.
Step S6-2, using a snake main frequency oscillation period proportion rule, taking a time calibration point as a center, taking a half period smaller than the snake movement fundamental frequency as a search time range, and extracting the maximum acceleration value of the filtered signal Sig _1, thereby obtaining all the maximum acceleration values as a peak sequence PY
The preset range is 2 × N +1 sampling points, wherein N is the number of sampling points corresponding to the range of occurrence of the snake main frequency oscillation peak value.
For example: suppose that the filtered signal Sig _1 is Y ═ Y1,y2,…,yi,…,yn]Wherein n is the sequence length, and n is equal to the sequence length corresponding to the framework snake main frequency signal Sig _ 2.
Due to the fact that the snaking motion of the framework is affected by non-stationary excitation and noise exists in collected signals, the snaking oscillation peak value appears in a certain proportion range of the oscillation period (namely, the snaking main frequency oscillation period proportion rule). Meanwhile, given a signal sampling frequency fs and a snake main frequency f _ h, the number of sampling points Nt of each period of the snake main frequency oscillation is as follows:
let (fs/f _ h), where ceil is an rounding-up function.
If the ratio of the peak value possibly appearing in the snake-walking period is Sw (Sw should be less than 0.5), the range of the snake-walking dominant frequency oscillation peak value corresponding to the number of sampling points N is:
Further, in the filtered signal Sig _1, the second peak sequence PY corresponding to the snake main frequency is:
where max is a function of the maximum.
Fig. 11 is a diagram illustrating a second peak sequence PY according to an embodiment of the invention.
To describe the second peak sequence PY and the time series PT in detail, the amplitude of a random certain time period in the diagram of the second peak sequence PY of fig. 11 is partially enlarged, and from the partially enlarged view, it can be seen that the maximum value in the second peak sequence PY corresponds to the time point in the time series PT, thereby illustrating that the second peak sequence PY is strongly correlated with the snake main frequencies and the false peak caused by the noise is removed.
In step S7, it is determined whether N consecutive peaks in the second peak sequence PY are not less than the threshold Ac.
And step S8, if the judgment in the step S7 is negative, the rail vehicle bogie is judged to be in a state with normal stability in the snaking process, and corresponding stability information of the bogie in the snaking process is output.
And step S9, if the judgment in the step S7 is yes, the situation that the stability of the bogie of the railway vehicle is in an abnormal state in the snaking process is judged, and corresponding instability alarm information of the bogie in the snaking process is output.
In order to analyze the influence of the expanded frequency band on the recognition result of the snaking motion state based on the snaking odd harmonic frequency characteristic, three different frequency band selections are carried out on the filtering frequency bands f1-f2 in the step S2-3, wherein the frequency band selections are divided into 0.5-10 Hz filtering considering the snaking fundamental frequency, 0.5-25 Hz filtering considering the snaking fundamental frequency and 3-order harmonic frequency, 0.5-50 Hz filtering considering the snaking fundamental frequency and harmonic frequencies of 3, 5 and 7 orders, and the influence of the expanded frequency band is illustrated by analyzing the second peak sequence PY mean value extracted from each frequency band.
FIG. 12 is an analysis diagram of the range of the filter band and the mean of the extracted peaks according to an embodiment of the present invention.
As can be seen from fig. 12, with the widening of the filtering frequency band, the extracted acceleration peak value increases faster with the increase of the vehicle running speed, taking the second peak sequence PY mean value corresponding to the train running speed of 350km/h as an example, the second peak sequence PY mean values corresponding to different frequency band ranges are specifically:
1. the frequency range is 0.5-10 Hz (only the snake base frequency is considered): the mean value of the second peak value sequence PY is 6.97m/s2。
2. The frequency range is 0.5-25 Hz (considering snake base frequency and 3 harmonic frequencies): the mean value of the second peak value sequence PY is 12.41m/s2。
Because the frequency range is 12.41m/s corresponding to 0.5-25 Hz2The specific frequency range is 6.97m/s corresponding to 0.5-10 Hz2Increase by 5.44m/s2(corresponding to 3 harmonic energies) to show that the 3 harmonic energies account for 78% of the energy of the fundamental snake frequency at 350km/h, and such high energy caused by instability can be effectively utilized by widening the frequency band, thereby serving as a basis for judging the vehicle snake instability.
FIG. 13 is an analysis graph of the filter band range and the peak increase factor according to an embodiment of the present invention.
In each filtering frequency range, assuming that 15 times of the extracted vibration amplitude reaches 100km/h as a unified standard for detecting the hunting instability, fig. 13 shows that the hunting instability can be detected at 234km/h by the 0.5-50 Hz filtering, at 261km/h by the 0.5-25 Hz filtering, and at 317km/h by the traditional 0.5-10 Hz filtering. Therefore, by widening the frequency band, harmonic frequency energy except the snake-like fundamental frequency is extracted, and the sensitivity of judging the snake-like motion stability can be improved, namely, the snake-like motion stability of the vehicle can be judged in advance under the condition that the train running speed is low.
Examples effects and effects
According to the method for detecting and evaluating the snake motion stability of the railway vehicle, the filtered signal Sig _1 is obtained by filtering the original sampling signal Sig _0 after the filtering frequency band is expanded based on the characteristic of the snake motion high-order odd harmonic frequency, and the energy in the expanded filtering frequency band mainly consists of the odd harmonic frequency caused by the snake motion, so that the filtered signal Sig _1 contains richer snake motion information, the problem of large signal energy loss caused by narrow-band filtering is solved, and the sensitivity of the monitoring method for judging the snake motion state is further improved. In addition, because the second peak sequence PY is obtained by carrying out continuous peak extraction on the filtered signal Sig _1 based on the time sequence PT by utilizing the preset snake main frequency oscillation period proportion rule, the possibility that false peaks exist in the peak extraction of the signal containing strong noise can be reduced, the extracted peaks have strong correlation with the snake vibration main frequency, and the accuracy of stability judgment is improved.
The above-described embodiments are merely illustrative of specific embodiments of the present invention, and the present invention is not limited to the description of the above-described embodiments.
Claims (4)
1. A rail vehicle snaking motion stability detection and evaluation method is used for evaluating the snaking motion stability of a rail vehicle in the running process, and is characterized by comprising the following steps:
step S1, acquiring a frame transverse vibration acceleration signal of the rail vehicle at a position right above two axle boxes at the diagonal angle of the frame in the running process in real time by using an acceleration sensor, and sampling the frame transverse vibration acceleration signal to obtain an original sampling signal Sig _ 0;
step S2, performing broadband band-pass filtering on the f1-f2 frequency band on the original sampling signal Sig _0 through a preset broadband band-pass filtering method based on the snake motion high-order odd harmonic frequency characteristic to obtain a filtered signal Sig _1, wherein the snake motion high-order odd harmonic frequency characteristic is that a snake fundamental frequency and high-order odd harmonic frequencies of the snake fundamental frequency exist in a framework transverse vibration frequency spectrum, and the f1-f2 frequency band is a frequency band expanded to the high-order odd harmonic frequencies;
step S3, carrying out snake main frequency identification on the filtered signal Sig _1 in a preset frequency identification range to obtain a snake main frequency f _ h;
step S4, performing broadband band-pass filtering on (f _ h-f _ w) to (f _ h + f _ w) frequency bands on the filtered signal Sig _1 to obtain a frame snake main frequency signal Sig _2, wherein f _ w is a preset bandwidth;
step S5, extracting continuous peak values in a time domain from the frame snake main frequency signal Sig _2 to obtain a first peak value sequence PX, and acquiring a corresponding time sequence PT according to the first peak value sequence;
step S6, using a preset snake main frequency oscillation period proportion rule and based on the time sequence PT, carrying out continuous peak value extraction on the filtered signal Sig _1 so as to obtain a second peak value sequence PY;
step S7, judging whether N continuous peaks in the second peak sequence PY are not less than a threshold value Ac;
step S8, when the judgment in the step S7 is negative, the bogie of the railway vehicle is judged to be in a state with normal stability in the snaking process, and corresponding stability information of the bogie in the snaking process is output;
step S9, if the judgment of the step S7 is yes, the state that the stability of the bogie of the railway vehicle is abnormal in the snaking process is judged, and corresponding instability alarm information of the bogie in the snaking process is output,
wherein the step S6 includes the following sub-steps:
step S6-1, calibrating the filtered signal Sig _1 based on the time sequence PT to obtain a time calibration point;
and step S6-2, taking the time calibration point as the center and the half period smaller than the fundamental frequency of the snaking motion as the search time range by utilizing the snaking dominant frequency oscillation period proportion rule, and extracting the maximum acceleration value of the filtered signal Sig _1 to obtain all the maximum acceleration values as the second peak value sequence PY.
2. The rail vehicle hunting motion stability detection and assessment method according to claim 1, wherein:
wherein the broadband band-pass filtering method is any one of a filter or a time-frequency decomposition reconstruction method,
the time-frequency decomposition reconstruction method is at least any one of EMD, EEMD and wavelet packet decomposition.
3. The rail vehicle hunting motion stability detection and assessment method according to claim 2, wherein:
wherein the step S2 includes the following sub-steps:
step S2-1, performing time-frequency decomposition on the original sampling signal Sig _0 by using the time-frequency decomposition reconstruction method to obtain sub-signals of the original sampling signal Sig _0 in different frequency bands;
step S2-2, performing spectrum analysis on the sub-signals, and identifying to obtain a main frequency position corresponding to each sub-signal;
and S2-3, reconstructing the sub-signals with the main frequency positions located at the f1-f2 frequency bands to obtain the filtered signals Sig _ 1.
4. The rail vehicle hunting motion stability detection and assessment method according to claim 1, wherein:
wherein the step S5 includes the following sub-steps:
s5-1, screening out sequence points with the primary difference larger than 0 and the secondary difference smaller than 0 from the sequence points of the framework snake main frequency signal Sig _2 as maximum value points;
step S5-2, taking the inverse number of the frame snake main frequency signal Sig _2 to obtain an opposite main frequency signal Sig _ 2';
s5-3, screening out sequence points with the primary difference larger than 0 and the secondary difference smaller than 0 from the sequence points of the opposite main frequency signals Sig _2' as minimum value points;
step S5-4, removing the maximum value points which are negative numbers from all the maximum value points to form a new maximum value point sequence, and removing the minimum value points which are positive numbers from all the minimum value points to form a new minimum value point sequence;
step S5-5, determining whether each maximum point in the new maximum point sequence and each minimum point in the new minimum point sequence are adjacent in time sequence, and if yes, taking the new maximum point sequence as a first peak value sequence PX;
step S5-6, when the judgment in the step S5-5 is negative, reserving the larger maximum value point of all the two adjacent maximum value points in the new maximum value point sequence, and thus obtaining the denoised maximum value sequence as a first peak value sequence PX;
step S5-7, acquiring the first peak sequence PX, and taking the time corresponding to the maximum point in the first peak sequence PX as the time sequence PT.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110584757.XA CN113358380B (en) | 2021-05-27 | 2021-05-27 | Rail vehicle snaking motion stability detection and evaluation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110584757.XA CN113358380B (en) | 2021-05-27 | 2021-05-27 | Rail vehicle snaking motion stability detection and evaluation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113358380A CN113358380A (en) | 2021-09-07 |
| CN113358380B true CN113358380B (en) | 2022-04-08 |
Family
ID=77527920
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110584757.XA Active CN113358380B (en) | 2021-05-27 | 2021-05-27 | Rail vehicle snaking motion stability detection and evaluation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113358380B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114013475B (en) * | 2021-11-30 | 2023-08-08 | 中国铁道科学研究院集团有限公司 | Train transverse movement stability detection method and device based on framework transverse movement signals |
| CN114861741B (en) * | 2022-07-11 | 2022-09-13 | 西南交通大学 | Snake state identification method based on wheel set transverse displacement |
| CN116223075B (en) * | 2023-05-05 | 2023-08-11 | 昆明轨道交通四号线土建项目建设管理有限公司 | Vibration stability detection system and method for rail transit vehicle |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6026311A (en) * | 1993-05-28 | 2000-02-15 | Superconductor Technologies, Inc. | High temperature superconducting structures and methods for high Q, reduced intermodulation resonators and filters |
| EP2233379A2 (en) * | 2009-03-26 | 2010-09-29 | Siemens Aktiengesellschaft | Method for monitoring the run stability of railway vehicles |
| EP2436574A1 (en) * | 2010-10-01 | 2012-04-04 | Hitachi, Ltd. | State monitoring apparatus and state monitoring method of railway car, and railway car |
| CN110411766A (en) * | 2019-07-30 | 2019-11-05 | 中国神华能源股份有限公司神朔铁路分公司 | The snakelike unstability detection method of train bogie, device, system and storage medium |
| CN112528403A (en) * | 2020-12-03 | 2021-03-19 | 同济大学 | Economic optimization turning repair method for wheel tread of railway vehicle |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2007070927A1 (en) * | 2005-12-20 | 2007-06-28 | Minelab Electronics Pty Limited | Real-time rectangular-wave transmitting metal detector platform with user selectable transmission and reception properties |
-
2021
- 2021-05-27 CN CN202110584757.XA patent/CN113358380B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6026311A (en) * | 1993-05-28 | 2000-02-15 | Superconductor Technologies, Inc. | High temperature superconducting structures and methods for high Q, reduced intermodulation resonators and filters |
| EP2233379A2 (en) * | 2009-03-26 | 2010-09-29 | Siemens Aktiengesellschaft | Method for monitoring the run stability of railway vehicles |
| EP2436574A1 (en) * | 2010-10-01 | 2012-04-04 | Hitachi, Ltd. | State monitoring apparatus and state monitoring method of railway car, and railway car |
| CN110411766A (en) * | 2019-07-30 | 2019-11-05 | 中国神华能源股份有限公司神朔铁路分公司 | The snakelike unstability detection method of train bogie, device, system and storage medium |
| CN112528403A (en) * | 2020-12-03 | 2021-03-19 | 同济大学 | Economic optimization turning repair method for wheel tread of railway vehicle |
Non-Patent Citations (3)
| Title |
|---|
| Effect of discrete track support by sleepers on rail corrugation at a curved track;X.S.Jin等;《Journal of Sound and Vibration》;20080831;279-300 * |
| 兰新客专高速列车转向架蛇行运动稳定性及影响研究;王安国等;《铁道机车车辆》;20190228;第39卷(第1期);27-31 * |
| 基于自适应IMM算法的蛇形机动目标加速度估计研究;吴新宏等;《上海航天(中英文)》;20201231;第37卷(第1期);18-23 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113358380A (en) | 2021-09-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113358380B (en) | Rail vehicle snaking motion stability detection and evaluation method | |
| CN110987348B (en) | Catenary hard spot determination method and device based on pantograph-catenary dynamic response | |
| CN106845429B (en) | Vibration signal classification judgment and identification method, rockfall energy scale calculation method and rockfall risk early warning method | |
| CN101259016A (en) | Method for real time automatically detecting epileptic character wave | |
| CN106441893A (en) | Train rolling bearing fault and impurity vibration distinguishing method | |
| CN106092015A (en) | A kind of raceway surface depression length detecting method | |
| CN107063339A (en) | Falling rocks judges recognition methods with train signal classification classification along railway | |
| CN106093701A (en) | A kind of cable fault signal detecting method based on empirical mode decomposition filtering | |
| CN109059839B (en) | Method, device and system for diagnosing out-of-roundness fault of wheel tread | |
| Azami et al. | A novel signal segmentation method based on standard deviation and variable threshold | |
| Christos et al. | Multicomponent decomposition of a time-varying acoustic Doppler signal generated by a passing railway vehicle using Complex Shifted Morlet Wavelets | |
| CN107521511B (en) | A kind of method and system of wheel derailment alarm | |
| CN107563403B (en) | Working condition identification method for high-speed train operation | |
| CN113281414B (en) | Method and device for identifying short-wave disease types of steel rails and electronic equipment | |
| Wei et al. | Bridge static influence line identification based on structural dynamic responses under high-speed trains | |
| JP3885297B2 (en) | Abnormal sound determination device and abnormal sound determination method | |
| CN116252820B (en) | Polygonal quantitative detection method for high-speed train wheels driven by improved frequency domain integration method | |
| CN113435281A (en) | Ripple compensator fault diagnosis method based on improved HHT conversion | |
| CN106096199B (en) | A kind of WT of rolling bearing, spectrum kurtosis and smooth iteration envelope Analysis Method | |
| CN114580460A (en) | Railway vehicle wheel rail fault diagnosis method based on morphological filtering and HHT conversion | |
| Xing et al. | A feature extraction method of wheelset-bearing fault based on wavelet sparse representation with adaptive local iterative filtering | |
| CN115358088A (en) | Bridge influence line identification method based on high-speed train excitation power response | |
| CN113343919B (en) | Method and device for detecting continuous equidistant rubbing damage of steel rail and computer equipment | |
| CN109975025B (en) | Locomotive wheel pair bearing quantitative diagnosis method based on adaptive filtering demodulation | |
| Yin et al. | Multiscale detrended cross-correlation analysis of traffic time series based on empirical mode decomposition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |