CN111368713B - Vehicle network system harmonic time-frequency analysis method based on synchronous compression wavelet transform - Google Patents

Abstract

The invention discloses a vehicle network system harmonic time-frequency analysis method based on synchronous compression wavelet transformation, which comprises the steps of observing network side current and voltage signals by setting different switching frequencies and traction network equivalent parameters under the cascade operation of an equivalent traction network and a CRH5 type vehicle rectifier under the condition of considering time delay generated by a digital control circuit, and analyzing the signals by utilizing the synchronous compression wavelet transformation; the analysis result of the method has good time-frequency resolution, and the amplitude and the frequency of harmonic amplification can be effectively identified when the system has harmonic instability; the invention has convenient implementation and higher adaptability.

Description

A Harmonic Time-Frequency Analysis Method for Vehicle-Network System Based on Synchronous Compressed Wavelet Transform

technical field

The invention belongs to the technical field of time-frequency analysis and processing of non-stationary data of high-speed railways, and in particular relates to a time-frequency analysis method for harmonics of a vehicle network system based on synchronous compression wavelet transform.

Background technique

In recent years, with the widespread use of HXD and CRH series electric locomotives, a large number of power electronic devices have been introduced into the traction system. Due to the nonlinear characteristics of power electronic equipment, when the locomotive is put into the traction network, it may interact with the traction network and bring new challenges to the stability of the traction system. In addition, the switching frequencies of the converters in different models are quite different, so the delays caused by the use of digital controllers are also different.

In order to study the influence of the switching frequency of the EMU converter on the vehicle network system, Assefa H Y et al. pointed out the influence of the switching frequency on the stability of the vehicle network system in Reference 1. However, this document only describes this phenomenon, and does not analyze the time-frequency spectrum characteristics of the electrical waveform of the vehicle network under different switching frequencies in detail. References 2 and 3 pointed out that the delay will cause the equivalent admittance of the converter to have a negative damping effect in the high frequency band, thereby reducing the system stability. Reference 4 points out that due to the large traction power of the EMU, the switching frequency of the converter is set low, which will cause a serious control delay. Literature 5 points out that this phenomenon is a characteristic sub-harmonic amplification phenomenon by performing Fourier transform on the harmonically unstable waveform. In addition, for non-stationary signals, wavelet transform is a commonly used time-frequency analysis tool. By constructing a suitable wavelet basis, non-periodic signals and signals containing singularities can be effectively analyzed. However, the resolution of the wavelet analysis results is often very dependent on the selection of the wavelet basis, and the choice of the wavelet basis function has no fixed rules, and the selection is often based on experience. Based on the above research methods:

1) In the research, the analysis methods of Fourier transform and wavelet transform are used for the harmonic instability problem in the vehicle network system. The time-frequency resolution is limited by the selection of window function or wavelet base, and traditional time-frequency analysis is used. The method cannot achieve the ideal analysis effect.

2) The stability of the vehicle network system is a strong coupling system, which is more sensitive to the change of the switching frequency of the converter and the parameters of the traction network.

Therefore, it is necessary to introduce a method with high time-frequency resolution into the analysis of harmonic instability. The synchronous compression wavelet transform does not depend too much on the selection of the wavelet base, and can accurately analyze the law of the amplitude and frequency of non-stationary signals changing with time, and effectively judge the influence of the switching frequency of the converter and the parameters of the traction network on the stability of the vehicle network system.

references:

[1]: Assefa H Y, Danielsen S, Molinas M. Impact of PWM switching onmodeling of low frequency power oscillation in electrical rail vehicle [C]. 13th European Conference on Power Electronics and Applications. IEEE, 2009: 1-9.

[2]: Harnefors L, Finger R, Wang X, et al.VSC input-admittance modeling and analysis above the Nyquist frequency for passivity-based stabilityassessment[J].IEEE Transactions on Industrial Electronics,2017,64(8):6362- 6370.

[3]: Harnefors L, Wang X, Yepes A G, et al. Passivity-based stability assessment of grid-connected VSCs—An overview [J]. IEEE Journal of emerging and selected topics in Power Electronics, 2015, 4(1):116 -125.

[4]: Mollerstedt E, Bernhardsson B. Out of control because of harmonics-an analysis of the harmonic response of an inverter locomotive[J]. IEEE Control Systems Magazine, 2000, 20(4): 70-81.

[5]: Tao Haidong, Hu Haitao, Zhu Xiaojuan, et al.Resonance instability mechanism analysis of traction power supply system under vehicle-to-grid coupling[J].Chinese Journal of Electrical Engineering,2019,39(8):2315-2324.

SUMMARY OF THE INVENTION

The purpose of the present invention is to: carry out time-frequency analysis when harmonic instability occurs in the vehicle network system, analyze the influence of the switching frequency of the EMU converter and traction network parameters on the stability of the vehicle network system, in order to solve the harmonics of the vehicle network system. Instability problems provide new ideas. To this end, the present invention provides a time-frequency analysis method for harmonics of a vehicle network system based on synchronous compressed wavelet transform.

The present invention provides a time-frequency analysis method for vehicle network system harmonics based on synchronous compression wavelet transform, under the cascade operation of equivalent traction network and CRH5 vehicle rectifier, by setting different switching frequencies and equivalent parameters of the traction network, considering Under the delay generated by the digital control circuit, the current and voltage signals on the grid side are observed, and the signals are analyzed by synchronous compression wavelet transform, which specifically includes the following steps:

Step 1: Given the initial value of the switching frequency of the converter, calculate the delay introduced by the digital controller into the vehicle network system when the PWM modulation circuit is single frequency sampling;

Step 2: Add the delay generated in the control circuit in the simulation process, observe and collect the voltage and current of the traction network;

Step 3: Analyze the voltage and current signals of the traction network through synchronous compression wavelet transform, and extract their time-frequency variation laws;

Step 4: Extract and track the time-frequency ridge of the synchronous compression transformation result, and perform modal extraction on the signal.

Among them, the specific process of step 3 is:

Step a: Select the analytical wavelet basis, perform wavelet transformation on the voltage and current of the traction network, and record the obtained result as WT _f (a,b), the process is as follows:

where f(t) is the signal to be analyzed, a is the scale factor in the wavelet basis ψ _a,b (t), and b is the translation factor;

Step b: Calculate the partial derivative of the obtained wavelet transform result WT _f (a, b) with respect to the translation factor b to get

Step c: Divide the obtained partial derivative result by the product of wavelet transform and i2π to obtain the instantaneous frequency value of the signal:

Step d: Redistribute the obtained instantaneous frequency value in the time-frequency region obtained by wavelet transform to improve the time-frequency resolution:

where T _f is the synchronous compressed wavelet transform matrix, a _k is the calculation interval of the wavelet transform coefficient matrix WT _f for the compression transformation, which can be obtained according to the formula ω( _ak ,b)-ω _l |≤Δω/2; The result of the frequency distribution is distributed around the center frequency _ωl .

Compared with the prior art, the beneficial effects of the present invention are:

1. The present invention can accurately analyze the time-varying law of the resonance frequency and amplitude of the traction network when harmonic instability occurs in the vehicle network system through the synchronous compression wavelet transform.

2. The present invention is convenient to implement, does not need to strictly select the wavelet basis to analyze the harmonic instability phenomenon, and the adaptability of the time-frequency analysis method is high.

Description of drawings

FIG. 1 is an equivalent block diagram of the vehicle network system of the present invention.

FIG. 2 is an equivalent circuit of the vehicle network system of the present invention.

FIG. 3 is the traction network voltage and current waveforms when the switching frequency of the on-board rectifier of the present invention is set to 1250 Hz.

FIG. 4 shows the voltage and current waveforms of the traction network when the switching frequency of the on-board rectifier of the present invention is set to 800 Hz.

FIG. 5 shows the voltage and current waveforms of the traction network when the switching frequency of the on-board rectifier of the present invention is set to 500 Hz.

Figure 6 is a time-frequency analysis diagram of the traction network voltage when the switching frequency is set to 1250Hz.

Figure 7 is a time-frequency analysis diagram of the traction grid voltage when the switching frequency is set to 800Hz.

Figure 8 is a time-frequency analysis diagram of the traction grid voltage when the switching frequency is set to 500Hz.

Figure 9 is a modal exploded view of the traction grid voltage when the switching frequency is set to 1250Hz.

Figure 10 is a modal exploded view of the traction grid voltage when the switching frequency is set to 800Hz.

Figure 11 is a modal exploded view of the traction grid voltage when the switching frequency is set to 500Hz.

Figure 12 is the oscilloscope diagram of the traction network voltage and current when the switching frequency is set to 1250Hz under the hardware-in-the-loop platform.

Figure 13 is an oscilloscope diagram of the traction network voltage and current when the switching frequency is set to 800Hz under the hardware-in-the-loop platform.

Figure 14 is the oscilloscope diagram of the traction network voltage and current when the switching frequency is set to 500Hz under the hardware-in-the-loop platform.

Detailed ways

The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

This example takes the traction network and the CRH5 EMU cascade system as an example. Figure 1 shows the equivalent model of the vehicle network system under the AT power supply mode of the traction power supply system. The traction power supply system is composed of traction substation and traction network, and the electric locomotive is composed of on-board rectifier, inverter and motor. Figure 2 is the equivalent circuit diagram of the vehicle network system, in which U _s is the traction grid voltage, and L _s , R _s , and C _s are the equivalent inductance, resistance and capacitance of the traction grid, respectively. L _g is the equivalent inductance of the on-board transformer, C _d is the DC side filter capacitor, and R _d is the equivalent load of the locomotive. The control loop of the simplified on-board rectifier consists of a sampling link, a control circuit, a PWM modulation circuit and a delay circuit. The delay is mainly generated in the PWM modulation process. When single frequency modulation is used, the delay time is the same as the switching frequency.

The spectrum analysis when the harmonics of the synchronous compressed wavelet transform vehicle-to-network system are unstable in this embodiment consists of the following steps.

1. By setting the switching frequency of different on-board converters in Matlab/Simulink, and considering the delay generated by the digital controller, observe and collect the waveforms of the voltage and current of the traction network;

Figures 3, 4, and 5 are the voltage and current waveforms of the traction network when the switching frequencies are 1250Hz, 800Hz, and 500Hz, respectively. When the switching frequency is 1250Hz, the delay caused by the control circuit is considered in the 5th second, after that, the voltage and current on the traction network appear harmonic amplification. In actual operation, when the traction network voltage exceeds 29kV, the locomotive protection device will be triggered to lock the locomotive and the locomotive will stop running. When the switching frequency is 800Hz, after adding the delay, there is no harmonic instability in the vehicle network system, and the voltage and current of the traction network return to normal operation after 0.15 seconds. When the switching frequency is 500Hz, after adding the delay, the vehicle network system cannot be stabilized, and the voltage and current are out of synchronization.

2. The time-frequency distribution diagram of traction network voltage under different switching frequencies using synchronous compression wavelet transform;

Figures 6, 7, and 8 are respectively the results of wavelet transform and synchronous compression wavelet transform on the voltage waveforms in Figures 3, 4, and 5 between 5 and 5.1 seconds. It can be clearly seen from Fig. 6 that the analysis result of synchronous compression wavelet transform has better time-frequency resolution than wavelet transform. The wavelet transform cannot accurately analyze the harmonic frequency and harmonic amplitude, but the synchronous compression wavelet transform can achieve precise positioning. From the analysis results of synchronous compression wavelet transform, it can be seen that after adding the delay in the 5th second, the 461Hz harmonic voltage amplitude amplification phenomenon appears in the traction network voltage, and this result is also consistent with Figure 3. It can be known from the data in Table 1 that 461Hz is also the parallel resonance frequency of the traction grid. Therefore, when harmonic instability occurs, the amplitude of the voltage at the natural resonant frequency of the traction network will be amplified.

Figures 7 and 8 are the time-frequency analysis results of the traction network voltage waveform when the switching frequency is 800Hz and 500Hz. It can be seen that at this time, there is no specific harmonic amplification phenomenon in the voltage of the vehicle network system.

3. Extract and track the time-frequency ridges of the synchronous compression and transformation results of harmonically unstable waveforms, and perform modal extraction on the signals;

In order to further analyze the waveform of the traction network voltage when harmonic instability occurs, the modal extraction of the voltage waveform in Figure 3 is carried out to analyze the composition of the waveform and the law of the frequency and amplitude of each component changing with time.

Figure 9 shows the modal decomposition result of the voltage waveform when the switching frequency is 1250 Hz, where S(t) is the harmonic unstable voltage signal intercepted within 5 to 5.1 seconds, and IMF is the natural modal component of the signal S(t). It can be found that when harmonic instability occurs, the waveform is composed of 50Hz fundamental wave and harmonic components with increasing amplitude. Figures 10 and 11 show the modal decomposition results of the voltage waveform when the switching frequency is 800Hz and 500Hz, respectively. It can be found that when the switching frequency is 800Hz, the voltage waveform is composed of the 50Hz fundamental wave and the harmonic components whose amplitude varies within a small range, and there is no instability due to harmonic amplification. When the switching frequency is reduced to 500Hz, the voltage waveform is composed of the 50Hz fundamental wave and the harmonic components with no obvious change law. At this time, comparing Fig. 5, it can be found that due to the low switching frequency, the voltage and current signals of the traction network are out of synchronization after considering the delay.

4. Carry out the test on the semi-physical platform to observe the influence of switching frequency on the stability of the vehicle network system;

In order to verify the influence of switching frequency on the stability of the vehicle network system, a hardware-in-the-loop vehicle network system hardware-in-the-loop experimental platform is built. The controller is loaded with the control circuit, the simulator is loaded with the main circuit, and the StarSim software is run on the PC side to form a closed-loop system. Figures 12, 13, and 14 are the waveforms of the voltage and current of the traction network intercepted from the oscilloscope when the switching frequency is set to 1250, 800, and 500 Hz. It can be found that the experimental results on the semi-physical platform are consistent with the simulation results of Matlab\Simulink.

The parameters of the vehicle network cascade system are shown in Table 1.

Table Ι vehicle network system model parameters

The synchronous compression wavelet transform is used to analyze the voltage and current signals of the traction network, and the analysis results have good time-frequency resolution, which can effectively identify the amplitude and frequency of harmonic amplification when harmonic instability occurs in the system.

Claims (1)

1. a kind of vehicle network system harmonic time-frequency analysis method based on synchronous compression wavelet transformation, it is characterized in that, under equivalent traction network and CRH5 type vehicle rectifier cascade operation, by setting different switching frequencies and traction network equivalent parameters , considering the delay generated by the digital control circuit, observe the current and voltage signal on the grid side, and analyze the signal by using the synchronous compression wavelet transform, which specifically includes the following steps: Step 1: Given the initial value of the switching frequency of the converter, calculate the delay introduced by the digital controller into the vehicle network system when the PWM modulation circuit is single frequency sampling; Step 2: Add the delay generated in the control circuit in the simulation process, observe and collect the voltage and current of the traction network; set different switching frequencies of the on-board converter in Matlab/Simulink, and consider the delay generated by the digital controller. , observe and collect the voltage and current waveforms of the traction network; Step 3: Analyze the voltage and current signals of the traction network through synchronous compression wavelet transform, and extract their time-frequency variation laws; Step a: Select the analytical wavelet basis, perform wavelet transformation on the voltage and current of the traction network, and record the obtained result as WT _f (a,b), the process is as follows:

where f(t) is the signal to be analyzed, a is the scale factor in the wavelet basis ψ _a,b (t), and b is the translation factor; Step b: Calculate the partial derivative of the obtained wavelet transform result WT _f (a, b) with respect to the translation factor b to get

Step c: Divide the obtained partial derivative result by the product of wavelet transform and i2π to obtain the instantaneous frequency value of the signal:

Step d: Redistribute the obtained instantaneous frequency value in the time-frequency region obtained by wavelet transform to improve the time-frequency resolution:

where T _f is the synchronous compressed wavelet transform matrix, a _k is the calculation interval of the wavelet transform coefficient matrix WT _f for the compression transformation, which can be obtained according to the formula |ω( _ak ,b)-ω _l |≤Δω/2; The results of the time-frequency distribution are distributed near the center frequency ω _l ; Step 4: Extract and track the time-frequency ridge of the synchronous compression transformation result, and perform modal extraction on the signal.

Images