Method and apparatus for evaluating performance of a maglev control loop based on time-varying fractal analysis

Abstract

The present application relates to a magnetic suspension control loop performance evaluation method and device based on time-varying fractal analysis, the method comprising the following steps: obtaining suspension point information of a magnetic levitation train suspension module, for the suspension point information, based on time-varying fractal analysis, obtaining a time-varying generalized Hurst index estimation value based on a window function matched with a preset index; according to the time-varying generalized Hurst index estimation value of multiple suspension points collected within a period of time, obtaining a comprehensive generalized Hurst index of each suspension point; and according to the comprehensive generalized Hurst index corresponding to each suspension point, obtaining a performance comparison result of a control loop. Compared with the prior art, the present application avoids the great difficulty brought by the establishment of an accurate mathematical model, and can evaluate the loop performance from the horizontal and vertical directions.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of rail transit, in particular to a magnetic suspension control loop performance evaluation method and device based on time-varying fractal analysis. BACKGROUND

[0002] With the rapid development of transportation systems, due to the characteristics of no mechanical contact between the vehicle and the track, and no wheel-rail adhesion, the maglev train has the characteristics of high speed, high stability, good comfort, strong climbing ability, and environmental friendliness, and has broad development prospects. The commercial operation of the maglev train requires excellent reliability and operation performance. Since the suspension system of the maglev train is open-loop unstable and nonlinear, the maglev train must apply active control to achieve stable suspension. As a core component of the maglev train, the suspension controller calculates the output control pulse according to the given value of the suspension gap, the actual gap value fed back by the suspension gap sensor, the acceleration value, and the current value fed back by the current sensor passing through the electromagnetic coil, controls the on-off time of the chopper switch tube, and adjusts the coil current of the suspension electromagnet, so that the suspension gap can always be kept within the rated value and the allowable fluctuation range value, and the stable suspension of the train is realized.

[0003] During the operation of the maglev train, the performance of the suspension control loop will be significantly affected due to external disturbances, sensor sensitivity decline, etc. Therefore, it is necessary to evaluate the performance of the electromagnetic suspension system control loop to ensure the safe and stable operation of the maglev train.

[0004] In recent years, the performance analysis and evaluation method of the suspension control system has attracted attention. Due to the complex structure of the suspension control system in practical application and many influencing factors, it is difficult to establish an accurate mathematical model, and relying on the model to evaluate the performance of the control loop will encounter multiple difficulties such as model parameter estimation.

[0005] In summary, there is currently a lack of a magnetic suspension control loop performance evaluation method, which provides a feasible solution for the performance debugging of the controller before the maglev train is put into commercial operation and the performance monitoring of the controller after the operation. SUMMARY

[0006] The purpose of the present application is to overcome the defects of the prior art and provide a magnetic suspension control loop performance evaluation method and device based on time-varying fractal analysis, which solves or partially solves the problem of difficulty in establishing an accurate mathematical model and difficulty in performance evaluation.

[0007] The purpose of the present application can be achieved by the following technical solutions:

[0008] In one aspect of the present application, a method for evaluating the performance of a magnetic suspension control loop based on time-varying fractal analysis is provided, comprising the following steps:

[0009] Obtaining suspension point information of a magnetic levitation train suspension module, and based on time-varying fractal analysis, obtaining a time-varying generalized Hurst index estimation value based on a window function that matches a preset index for the suspension point information;

[0010] Obtaining a comprehensive generalized Hurst index for each suspension point based on the time-varying generalized Hurst index estimation values of multiple suspension points collected over a period of time;

[0011] Obtaining a performance comparison result of the control loop based on the comprehensive generalized Hurst index corresponding to each suspension point.

[0012] As a preferred technical solution, the suspension point information is obtained from a suspension controller arranged on the suspension module, and the suspension module is arranged on the magnetic levitation train.

[0013] As a preferred technical solution, the magnetic levitation train is provided with multiple suspension modules, and each suspension module is provided with multiple suspension controllers.

[0014] As a preferred technical solution, the preset index includes one or more of a suspension gap value, a control current value, and an acceleration value.

[0015] As a preferred technical solution, the step of obtaining the time-varying generalized Hurst index estimation value includes:

[0016] Based on the suspension point information, a preferred static generalized Hurst index estimation method is used to obtain the time-varying generalized Hurst index estimation value,

[0017] The preferred static generalized Hurst index estimation method includes the following steps:

[0018] A fixed fractal Gaussian noise sequence is set, different static generalized Hurst index estimation methods, different window functions, and parameters of the window functions are used as alternative groups, and the time-varying generalized Hurst index of each alternative group is estimated;

[0019] The mean square error between the estimated value and the actual value of the time-varying generalized Hurst index of each alternative group is calculated, and the alternative group with the smallest mean square error is selected as the preferred static generalized Hurst index estimation method.

[0020] As a preferred technical solution, the step of obtaining the comprehensive generalized Hurst index includes:

[0021] According to the time-varying generalized Hurst index estimation value of the plurality of suspension points collected in a period of time, the weight corresponding to each of the preset indicators in the suspension point information is obtained by an entropy weight method.

[0022] According to the time-varying generalized Hurst index estimation value of the plurality of suspension points collected in a period of time and the weight, the comprehensive generalized Hurst index is obtained.

[0023] As a preferred technical solution, the performance comparison result obtaining step comprises:

[0024] According to the comprehensive generalized Hurst index corresponding to each of the suspension points, the comprehensive generalized Hurst indexes of different suspension points in the same period and the comprehensive generalized Hurst indexes of the same suspension point in different periods are compared, and the performance comparison result is obtained.

[0025] As a preferred technical solution, the performance comparison result obtaining step comprises:

[0026] According to the comparison result, a visualization signal including the comprehensive generalized Hurst index is sent to a visualization device.

[0027] Another aspect of the present application provides an electronic device, comprising one or more processors and a memory, the memory having one or more programs stored therein, the one or more programs comprising instructions for performing the above-mentioned magnetic suspension control loop performance evaluation method based on time-varying fractal analysis.

[0028] Another aspect of the present application provides a computer-readable storage medium comprising one or more programs for execution by one or more processors of an electronic device, the one or more programs comprising instructions for performing the above-mentioned magnetic suspension control loop performance evaluation method based on time-varying fractal analysis.

[0029] Compared with the prior art, the present application has the following advantages:

[0030] (1) Extract each suspension point information from the process data of the closed-loop operation of the suspension control system, and evaluate the performance of the control loop. Since the time-varying generalized Hurst index estimation does not require accurate model information, it can diagnose the control loop by obtaining each suspension point information, avoiding the great difficulty of establishing an accurate mathematical model, and providing a feasible solution for controller performance debugging before the commercial operation of the maglev train and controller performance monitoring after the commercial operation. Since most random signal sequences in reality are non-stationary, some signals have obvious self-similarity, self-similarity is an important feature of fractal theory, which reflects the correlation between the local and the whole, in fractal analysis, time series with self-similarity can be described by Hurst index, which is further used for performance evaluation of the control loop, so that the overall control loop performance can be more intuitively reflected.

[0031] (2) The present application can be calculated offline or run online in real time, and has high efficiency in evaluating the performance of the control loop.

[0032] (3) According to the obtained comprehensive generalized Hurst index, the control performance difference of the same suspension point at different time intervals and speeds can intuitively reflect the different states of the track passed by the train, which is helpful to find and analyze the part of the track with poor state. The heat map of the same time interval and different suspension points can reflect the control performance difference of the whole suspension module, which is conducive to finding and further improving the suspension module with poor performance. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 The flowchart of a kind of maglev control loop performance evaluation method based on time-varying fractal analysis in embodiment 1;

[0034] Figure 2 The schematic diagram of the typical configuration structure of the low-speed maglev vehicle suspension system in embodiment 1;

[0035] Figure 3 The estimated value heat map of the comprehensive generalized Hurst index transverse comparison in embodiment 1;

[0036] Figure 4 The estimated value heat map of the comprehensive generalized Hurst index longitudinal comparison in embodiment 1. DETAILED DESCRIPTION

[0037] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor shall belong to the scope of protection of the present application.

[0038] Embodiment 1

[0039] As Figure 1 described, the embodiment provides a magnetic suspension control loop performance evaluation method based on time-varying fractal analysis, comprising the following steps:

[0040] S1, real environment sensing device data acquisition.

[0041] S2, time-varying generalized Hurst index estimation based on window function. The calculation method of time-varying generalized Hurst index estimation value is: based on window function, the time-varying generalized Hurst index estimation should select appropriate static generalized Hurst index estimation method, appropriate window function and its parameters. Specifically: select FGN sequence with certain generalized Hurst index, select different static generalized Hurst index estimation method, different window function and window function parameter to estimate time-varying generalized Hurst index, calculate the mean square error between the estimated value and the actual value, select the method, window function and its parameter with the minimum error, design the time-varying generalized Hurst index estimation method, and obtain the generalized Hurst index estimation value according to the method. Wherein, fractal Gaussian noise (Fractal Gaussian Noise, FGN) is a kind of autocorrelation function widely used at present, which has strict self-similarity. Static generalized Hurst index estimation method includes R / S analysis method, absolute value method and variance time method; Window function includes rectangular window, triangular window, Hamming window, Hanning window and Blackman window; The parameters of window function include window size, sliding step length, etc.

[0042] S3, calculate the comprehensive generalized Hurst index estimation value by using entropy weight method. The calculation method is: the obtained time-varying generalized Hurst index estimation value is calculated by using entropy weight method to calculate the weight corresponding to the generalized Hurst index estimation value of air gap, current and acceleration signal, and the comprehensive generalized Hurst index is calculated by weighting. Specifically: the time-varying generalized Hurst index estimation value of different detection points is calculated horizontally, the time-varying generalized Hurst index estimation value of different time periods is calculated vertically, and the comprehensive generalized Hurst index estimation value is calculated according to entropy weight method.

[0043] S4, performance evaluation of control loop.

[0044] S5. The performance of the control loops is compared horizontally and vertically. Horizontal comparison allows for the evaluation of the performance of different suspension modules within the same time period, which helps to further improve the performance of weaker control modules. Vertical comparison allows for the evaluation of the performance of the same controller at a constant speed in different time periods, which helps to analyze and compare track conditions, identify tracks with poor performance, and further analyze the causes. Combining the results of horizontal and vertical comparisons with specific situations allows for further analysis of the differences in controller performance within the same time period and the performance of the same controller under different track conditions. Specifically, the horizontal perspective refers to the comparison of the comprehensive generalized Hurst exponent of the suspension points of different suspension modules within the same time period; the vertical perspective refers to the comparison of the comprehensive generalized Hurst exponent of the same suspension point in different time periods.

[0045] S6. Generate a heatmap. A heatmap is generated based on the comprehensive generalized Hurst exponent estimate, and combined with specific circumstances to further analyze the differences in controller performance during the same period and the conditions of the same controller under different track conditions.

[0046] The evaluation method in this embodiment can be summarized as follows: A suitable static Hurst exponent estimation method is selected and combined with a suitable window function to propose a time-varying Hurst exponent estimation method based on the window function. This extends the static Hurst exponent to a generalized Hurst exponent, at which point the time-varying generalized Hurst exponent estimation can be used for performance evaluation. Furthermore, a comprehensive generalized Hurst exponent estimate is calculated using methods such as the entropy weight method. Based on this, horizontal and vertical comparative evaluations are conducted. The horizontal comparison results provide feasible solutions for controller performance debugging before the maglev train is put into commercial operation and for controller performance monitoring after operation. The vertical comparison results provide a basis for track condition analysis during train debugging.

[0047] This embodiment collects data such as the suspension gap value, control current value, and acceleration value of the suspension point of the maglev train's suspension module, performs time-varying generalized Hurst exponent estimation based on window functions, calculates the comprehensive generalized Hurst exponent estimate using entropy weight method, evaluates the performance of the control loop, and finally analyzes the controller's state according to specific operating conditions to achieve state monitoring.

[0048] During the operation of maglev trains, the performance of the levitation control loop is significantly affected over time due to external disturbances, decreased sensor sensitivity, and other factors. Therefore, performance evaluation of the electromagnetic levitation system control loop is essential to ensure the safe and stable operation of maglev trains. The maglev train levitation system control loop performance evaluation method in this embodiment provides a feasible solution for controller performance debugging before maglev trains are put into commercial operation and controller performance monitoring after operation, and has the advantage of high efficiency.

[0049] like Figure 2In this embodiment, each train car is equipped with 5 sets of suspension modules, and each set of modules contains 4 suspension controllers. In practical applications, the number of suspension modules and controllers will vary depending on the specific design. When collecting data using the maglev train's suspension control system sensors, it is necessary to pay attention to the correspondence between the suspension modules and the suspension points. In this embodiment, data needs to be collected from 20 suspension points. The data collected from the suspension points is processed, and the results need to be further analyzed back in the suspension modules.

[0050] like Figure 3 The heatmap of the comprehensive generalized Hurst exponent estimate of the 20 suspension control loops reflects the differences in control performance at different suspension points, and further reflects the differences in the overall control performance of the suspension modules, which is helpful to identify and further improve the poor-performing suspension modules.

[0051] like Figure 4 The aforementioned track state diagram based on the comprehensive generalized Hurst index, which represents the difference in control performance based on the heat map of the same suspension point at a constant speed during different time periods, reflects the different states of the track that the train passes through, and helps to identify and analyze the parts of the track with poor condition.

[0052] Example 2

[0053] This embodiment provides an electronic device, including one or more processors and a memory, in which one or more programs are stored, the one or more programs including instructions for executing the magnetic levitation control loop performance evaluation method based on time-varying fractal analysis as described in Embodiment 1.

[0054] Example 3

[0055] This embodiment provides a computer-readable storage medium including one or more programs executable by one or more processors of an electronic device, the one or more programs including instructions for performing the magnetic levitation control loop performance evaluation method based on time-varying fractal analysis as described in Embodiment 1.

[0056] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for evaluating the performance of a magnetic levitation control loop based on time-varying fractal analysis, characterized in that, The method comprises the following steps: Obtaining suspension point information of a maglev train suspension module, and obtaining a time-varying generalized Hurst index estimation value based on a window function matched with a preset index based on time-varying fractal analysis for the suspension point information; Obtaining a comprehensive generalized Hurst index of each suspension point based on the time-varying generalized Hurst index estimation values of multiple suspension points collected within a period of time; Obtaining a performance comparison result of a control loop based on the comprehensive generalized Hurst index corresponding to each suspension point, The preset index comprises one or more of a suspension gap value, a control current value, and an acceleration value, The performance comparison result comprises the following steps: Comparing the comprehensive generalized Hurst index of different suspension points in the same period of time and the comprehensive generalized Hurst index of the same suspension point in different periods of time based on the comprehensive generalized Hurst index corresponding to each suspension point, and obtaining the performance comparison result.

2. The method for evaluating the performance of a magnetic levitation control loop based on time-varying fractal analysis according to claim 1, characterized in that, The suspension point information is obtained from a suspension controller arranged on the suspension module, and the suspension module is arranged on the maglev train.

3. The method of claim 2, wherein, The maglev train is provided with multiple suspension modules, and each suspension module is provided with multiple suspension controllers.

4. The method of claim 1, wherein, The time-varying generalized Hurst index estimation value comprises the following steps: The time-varying generalized Hurst index estimation value is obtained based on the suspension point information by using a preferred static generalized Hurst index estimation method, The preferred static generalized Hurst index estimation method comprises the following steps: A fixed fractal Gaussian noise sequence is set, different static generalized Hurst index estimation methods, different window functions, and parameters of the window functions are used as alternative groups, and the time-varying generalized Hurst index of each alternative group is estimated; The mean square error between the estimated value and the actual value of the time-varying generalized Hurst index of each alternative group is calculated, and the alternative group with the smallest mean square error is selected as the preferred static generalized Hurst index estimation method.

5. The method for evaluating the performance of a magnetic levitation control loop based on time-varying fractal analysis according to claim 1, characterized in that, The comprehensive generalized Hurst index comprises the following steps: The weight corresponding to each preset index in the suspension point information is obtained by an entropy weight method based on the time-varying generalized Hurst index estimation values of multiple suspension points collected within a period of time; The comprehensive generalized Hurst index is obtained based on the time-varying generalized Hurst index estimation values of multiple suspension points collected within a period of time and the weight.

6. The method of claim 1, wherein, The method further comprises the following steps: A visualization signal comprising the comprehensive generalized Hurst index is sent to a visualization device based on the comparison result.

7. An electronic device, comprising: The method comprises the following steps: One or more processors and a memory, the memory storing one or more programs, the one or more programs comprising instructions for performing the time-varying fractal analysis-based maglev control loop performance evaluation method according to any one of claims 1-6.

8. A computer-readable storage medium, characterized in that, including one or more programs for execution by one or more processors of an electronic device, the one or more programs including instructions for performing the method of claim 1-6 for evaluating performance of a magnetic levitation control loop based on time-varying fractal analysis.

Images