Asynchronous motor fault identification method and system

Abstract

The invention relates to an asynchronous motor fault identification method and system, and the method comprises the steps: collecting a vibration signal generated in the operation process of an asynchronous motor through a sensor device, and carrying out the time-frequency analysis of the collected vibration signal, and obtaining a time-frequency diagram; a plurality of movable points are initialized on the time-frequency graph, the standard deviation of the time-frequency graph in a window with the movable points as the center is calculated, and the horizontal movement direction and the step length of the movable points are determined based on the standard deviation; after horizontal movement, moving upwards or downwards by the step length in the vertical direction at the same probability, and if the value of the movable point in the time-frequency diagram after movement is smaller than that before vertical movement, determining the movement probability and the non-movement probability according to the movable point before vertical movement and two points above and below the movable point; updating the measurement matrix after the movable point moves to a new position; and extracting a ridge line of the time-frequency graph based on the final measurement matrix, and identifying an asynchronous motor fault corresponding to the ridge line.

Description

Technical Field

[0001] The present invention relates to the field of fault detection, and in particular to an asynchronous motor fault identification method and system. Background Art

[0002] Asynchronous motors have been widely used in the industrial field and daily life due to their simple structure, low cost, and reliable operation. However, with the increase of usage time and changes in the working environment, asynchronous motors may have various faults, such as broken rotor bars, bearing damage, stator winding short circuit, etc. These faults will not only affect the performance of the motor and lead to reduced efficiency, but may also cause equipment downtime, resulting in production losses and even safety accidents.

[0003] Timely detection and diagnosis of asynchronous motor faults can prevent unexpected downtime and production line shutdowns and maintain the continuity of the production process. Traditional fault detection methods rely on manual inspection and regular maintenance, but these methods are often time-consuming and labor-intensive, and it is difficult to detect potential problems at an early stage. Summary of the invention

[0004] In order to solve the above problem, the present invention provides an asynchronous motor fault identification method, the method comprising the following steps: The sensor equipment is used to collect the vibration signal generated during the operation of the asynchronous motor, and the collected vibration signal is subjected to time-frequency analysis to obtain a time-frequency diagram; Initialize multiple movable points on the time-frequency graph, calculate the standard deviation of the time-frequency graph in the window centered on the movable point, and determine the horizontal movement direction and step length of the movable point based on the standard deviation; after the horizontal movement, move up or down the step length in the vertical direction with the same probability, if the value of the movable point in the time-frequency graph after the movement is less than the value before the vertical movement, determine the movement probability and the non-movement probability based on the movable point before the vertical movement and the upper and lower points of the movable point; update the metric matrix after the movable point moves to the new position; The ridges of the time-frequency graph are extracted based on the final metric matrix, and the asynchronous motor faults corresponding to the ridges are identified.

[0005] Preferably, the determining of the horizontal moving direction and step length of the movable point based on the standard deviation is specifically as follows: The window is divided into two parts, a left part and a right part, of equal size, and a ratio of a standard deviation on the left side to a standard deviation on the right side is calculated, and a probability of moving to the left and a probability of moving to the right are obtained based on the ratio; A correspondence between the standard deviation interval and the step length is preset, and the step length is determined according to the standard deviation interval in which the standard deviation of the window is located, wherein the average value of the standard deviation interval and the step length are inversely correlated.

[0006] Preferably, after the step length is moved up or down in the vertical direction with the same probability, the method further comprises: If the value of the movable point in the time-frequency diagram after the movement is not less than the value before the vertical movement, the vertical movement direction selected with equal probability is taken as the vertical movement mode.

[0007] Preferably, the moving probability and the non-moving probability are determined according to the movable point before vertical movement and the two points above and below the movable point, specifically: Obtaining the standard deviation of the movable point before vertical movement in the window, as well as the standard deviation of the upper point of the movable point before vertical movement in the window and the standard deviation of the lower point of the movable point in the window; weighting the standard deviation within the window of movable points before vertical movement based on the current temperature setting weight; The probability of not moving, the probability of moving upward and the probability of moving downward are calculated according to the weighted standard deviation of the movable point before the vertical movement, the standard deviation of the upper point and the standard deviation of the lower point.

[0008] Preferably, after determining the moving probability and the non-moving probability according to the movable point before vertical movement and the two points above and below the movable point, the method further comprises: The movable point is determined in a vertical moving direction according to the moving probability and the non-moving probability, and the movable point is moved in the vertical moving direction.

[0009] On the other hand, the present invention provides an asynchronous motor fault identification system, the system comprising the following modules: The vibration acquisition module is used to use sensor equipment to collect vibration signals generated during the operation of the asynchronous motor, and perform time-frequency analysis on the collected vibration signals to obtain a time-frequency graph; The feature extraction module is used to initialize multiple movable points on the time-frequency graph, calculate the standard deviation of the time-frequency graph in the window centered on the movable point, and determine the horizontal movement direction and step length of the movable point based on the standard deviation; after the horizontal movement, move up or down the step length in the vertical direction with the same probability, if the value of the movable point in the time-frequency graph after the movement is less than the value before the vertical movement, then determine the movement probability and the non-movement probability according to the movable point before the vertical movement and the upper and lower points of the movable point; update the metric matrix after the movable point moves to the new position; The fault identification module is used to extract the ridge line of the time-frequency diagram based on the final metric matrix and identify the asynchronous motor fault corresponding to the ridge line.

[0010] Preferably, the determining of the horizontal moving direction and step length of the movable point based on the standard deviation is specifically as follows: The window is divided into two parts, a left part and a right part, of equal size, and a ratio of a standard deviation on the left side to a standard deviation on the right side is calculated, and a probability of moving to the left and a probability of moving to the right are obtained based on the ratio; A correspondence between the standard deviation interval and the step length is preset, and the step length is determined according to the standard deviation interval in which the standard deviation of the window is located, wherein the average value of the standard deviation interval and the step length are inversely correlated.

[0011] Preferably, after the step length is moved up or down in the vertical direction with the same probability, the method further comprises: If the value of the movable point in the time-frequency diagram after the movement is not less than the value before the vertical movement, the vertical movement direction selected with equal probability is taken as the vertical movement mode.

[0012] Preferably, the moving probability and the non-moving probability are determined according to the movable point before vertical movement and the two points above and below the movable point, specifically: Obtaining the standard deviation of the movable point before vertical movement in the window, as well as the standard deviation of the upper point of the movable point before vertical movement in the window and the standard deviation of the lower point of the movable point in the window; weighting the standard deviation within the window of movable points before vertical movement based on the current temperature setting weight; The probability of not moving, the probability of moving upward and the probability of moving downward are calculated according to the weighted standard deviation of the movable point before the vertical movement, the standard deviation of the upper point and the standard deviation of the lower point.

[0013] Preferably, after determining the moving probability and the non-moving probability according to the movable point before vertical movement and the two points above and below the movable point, the method further comprises: The movable point is determined in a vertical moving direction according to the moving probability and the non-moving probability, and the movable point is moved in the vertical moving direction.

[0014] The present invention can automatically adjust the step size of the movable point. Specifically, in areas with drastic changes, the step size is small to improve the search accuracy, and in areas with gentle changes, the step size is large to accelerate the search. Moreover, by comparing the values ​​of the movable point and its upper and lower points to calculate the probability of moving or not moving, the moving process is made more robust, effectively reducing invalid moving operations and improving the reliability of the search. By improving the extraction speed and accuracy of the ridge line, the speed and accuracy of asynchronous motor fault identification are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a flow chart of Embodiment 1; Figure 2 For vertical movement flow chart; Figure 3 This is a structural diagram of the second embodiment. DETAILED DESCRIPTION

[0016] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0017] In the technical solutions disclosed in this disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved are in compliance with the relevant laws and regulations and do not violate public order and good morals. In the technical solutions disclosed in this disclosure, the user's authorization or consent is obtained before obtaining or collecting user personal information.

[0018] In the first embodiment, if Figure 1 As shown, the present invention provides an asynchronous motor fault identification method, the method comprising the following steps: S1, using a sensor device to collect vibration signals generated during the operation of the asynchronous motor, and performing time-frequency analysis on the collected vibration signals to obtain a time-frequency graph; Asynchronous motors generate mechanical vibrations during operation. These vibrations may come from faults such as rotor imbalance, bearing wear, broken rotor bars, and uneven air gaps. These faults will cause the motor to generate vibration signals of different frequencies and amplitudes. By installing vibration sensors, such as accelerometers or velocity sensors, at key parts of the motor, the vibration data of the motor can be collected in real time. Traditional time domain analysis can only reveal the changes in the vibration signal over time, but it is difficult to accurately distinguish different frequency components and their dynamic changes over time. Fault information is often reflected in both time and frequency dimensions. Therefore, through time-frequency analysis, such as short-time Fourier transform (STFT), wavelet transform, etc., the vibration signal can be converted from the time domain to the time-frequency domain to analyze the change law of the signal at different frequencies. The time-frequency diagram is a two-dimensional image, the horizontal axis represents time, the vertical axis represents frequency, and the color or intensity of different areas in the diagram represents the energy intensity of the vibration signal at the corresponding frequency and time. In one embodiment, after obtaining the time-frequency diagram, the time-frequency diagram is also normalized.

[0019] S2, initialize multiple movable points on the time-frequency graph, calculate the standard deviation of the time-frequency graph in the window centered on the movable point, and determine the horizontal movement direction and step size of the movable point based on the standard deviation; after the horizontal movement, move up or down the step size in the vertical direction with the same probability, if the value of the movable point in the time-frequency graph after the movement is less than the value before the vertical movement, determine the movement probability and the non-movement probability based on the movable point before the vertical movement and the upper and lower points of the movable point; update the metric matrix after the movable point moves to the new position; Initialize multiple movable points (Climbers) on the time-frequency graph. The positions of the movable points are randomly determined, but try to ensure uniformity on the time-frequency graph. In one embodiment, the time-frequency graph is divided into multiple grids, and the number of movable points in each grid is determined according to the entropy value of the elements in the grid. Specifically, if the entropy value of the grid is less than the average value, one movable point is allocated to the grid, and if the entropy value of the grid is not less than the average value, two movable points are allocated to the grid. For example, there are 5 grids, 2 of which are less than the average value of the entropy value, then these two grids are each allocated one movable point, and the other 3 grids are allocated 2 movable points.

[0020] There are multiple movable points on the time-frequency graph. For each movable point, first calculate the standard deviation of the time-frequency graph in the window centered on the movable point, and the window size is 3×3 or 5×5. The larger the standard deviation, the greater the change in the position of the movable point in the time-frequency graph, which means that there are significant features. Continuing to explore in this area can capture important information; the smaller the standard deviation, the smaller the change around the movable point in the time-frequency graph, and it is necessary to move in the direction of greater change. At this time, the step size should be increased.

[0021] In one embodiment, the horizontal moving direction and step length of the movable point are determined based on the standard deviation, specifically: The window is divided into two parts, a left part and a right part, of equal size, and a ratio of a standard deviation on the left side to a standard deviation on the right side is calculated, and a probability of moving to the left and a probability of moving to the right are obtained based on the ratio; By dividing the window into two parts, the standard deviations of the left and right sides are calculated respectively, and the probability of moving to the left or right is determined based on their ratio. The direction with a larger standard deviation means that the change in that area is drastic, so the probability of moving in that direction will be higher. For example, if the standard deviation on the left is 2 and the standard deviation on the right is 3, the ratio is 2:3, the probability of moving to the left is 0.4, and the probability of moving to the right is 0.6.

[0022] A correspondence between the standard deviation interval and the step length is preset, and the step length is determined according to the standard deviation interval in which the standard deviation of the window is located, wherein the average value of the standard deviation interval and the step length are inversely correlated.

[0023] The larger the standard deviation of the movable point in the window, the smaller the step length, and vice versa. The corresponding relationship between the standard deviation interval and the step length is set. For example, if the standard deviation is between the interval [1, 2), the step length is 3, and if the standard deviation is between the interval [2, 4), the step length is 2, etc. In another alternative embodiment, the standard deviation of the time-frequency graph is calculated, and then the ratio of the standard deviation of the time-frequency graph to the standard deviation in the window is calculated, and the ratio is rounded up as the step length of the movement.

[0024] For vertical movement, first choose to move up or down with equal probability. If the value of the movable point in the time-frequency diagram after vertical movement is greater than the value of the movable point in the time-frequency diagram after vertical movement, then the vertical movement mode at this time is used as the vertical movement mode of the movable point. Otherwise, the probability of moving and the probability of not moving are determined according to the movable point before vertical movement and the two points above and below the movable point, such as Figure 2 As shown. The vertical movement mode at this time is used as the vertical movement mode of the movable point. Specifically, if the probability of moving up or down is equal, and after the movement, the value of the position of the movable point is greater than the value of the position before the vertical movement, then this movement mode is determined as the vertical movement mode. However, after the movement, the value of the position of the movable point is not greater than the value of the position before the vertical movement. Then, the probability of moving and the probability of not moving are determined according to the movable point before the vertical movement and the upper and lower points of the movable point, and then the vertical movement mode is determined according to the probability of moving and the probability of not moving.

[0025] In one embodiment, the upper and lower points of the movable point refer to a point above the movable point at a distance of the step length from the movable point and a point below the movable point at a distance of the step length from the movable point when the horizontal movement is completed but no vertical movement is performed.

[0026] In one embodiment, the moving probability and the non-moving probability are determined based on the movable point before vertical movement and the two points above and below the movable point, specifically: Obtaining the standard deviation of the movable point before vertical movement in the window, as well as the standard deviation of the upper point of the movable point before vertical movement in the window and the standard deviation of the lower point of the movable point in the window; weighting the standard deviation within the window of movable points before vertical movement based on the current temperature setting weight; The probability of not moving, the probability of moving upward and the probability of moving downward are calculated according to the weighted standard deviation of the movable point before the vertical movement, the standard deviation of the upper point and the standard deviation of the lower point.

[0027] The present invention also controls vertical movement by temperature. Specifically, the weight is set based on the current temperature. In one embodiment, specifically, the weight is set based on the current temperature. Determine the weight. As the movement or time progresses, T becomes smaller and the weight increases. The larger the weighted standard deviation of the movable point in the window before the vertical movement is, the greater the probability of non-movement is. exp represents exponential operation. The ratio of the non-movement probability a, the upward movement probability b and the downward movement probability c is the same as the ratio of the weighted standard deviation, the standard deviation of the upper point and the standard deviation of the lower point, and a+b+c=1.

[0028] After determining the probability of vertical upward movement, the probability of downward movement, and the probability of no movement, the vertical movement mode is further determined according to the above three probabilities, and then the step length is moved vertically upward, or the step length is moved vertically downward, or no vertical movement is performed. After the movable point moves, the metric matrix is ​​updated according to the metric matrix update method in the Crazy Climber algorithm. A simple update method is that if the movable point moves to a new position, the value at the new position in the metric matrix is ​​increased by 1.

[0029] S3, extract the ridge of the time-frequency graph based on the final metric matrix, and identify the asynchronous motor fault corresponding to the ridge.

[0030] In the time-frequency diagram, the ridge represents the energy concentration area of ​​a specific frequency component in the signal, which is usually manifested as one or several obvious high-energy curves in the time-frequency diagram. The ridge represents the change of the signal at a specific frequency over time, which can reflect the frequency characteristics of the asynchronous motor during operation. Some ridges are related to specific faults of the asynchronous motor. The ridge is usually the continuous part with the largest energy in the time-frequency diagram, which can be realized by local maximum detection or peak tracking algorithm. It is preferably extracted by using the ridge extraction method of the Crazy Climber algorithm.

[0031] Asynchronous motor faults usually manifest as abnormalities in specific frequency components, such as broken rotor bars, bearing faults, etc. These faults will produce obvious characteristic ridges on the time-frequency graph. By analyzing the ridges, these frequency features can be identified and matched with known fault ridges to determine the fault. In one embodiment, curve matching is used to identify whether a fault ridge exists. In another embodiment, machine learning is used to identify whether a ridge is a fault ridge.

[0032] In the second embodiment, if Figure 3 As shown, the present invention provides an asynchronous motor fault identification system, the system comprising the following modules: The vibration acquisition module is used to use sensor equipment to collect vibration signals generated during the operation of the asynchronous motor, and perform time-frequency analysis on the collected vibration signals to obtain a time-frequency graph; The feature extraction module is used to initialize multiple movable points on the time-frequency graph, calculate the standard deviation of the time-frequency graph in the window centered on the movable point, and determine the horizontal movement direction and step length of the movable point based on the standard deviation; after the horizontal movement, move up or down the step length in the vertical direction with the same probability, if the value of the movable point in the time-frequency graph after the movement is less than the value before the vertical movement, then determine the movement probability and the non-movement probability according to the movable point before the vertical movement and the upper and lower points of the movable point; update the metric matrix after the movable point moves to the new position; The fault identification module is used to extract the ridge line of the time-frequency diagram based on the final metric matrix and identify the asynchronous motor fault corresponding to the ridge line.

[0033] Preferably, the determining of the horizontal moving direction and step length of the movable point based on the standard deviation is specifically as follows: The window is divided into two parts, a left part and a right part, of equal size, and a ratio of a standard deviation on the left side to a standard deviation on the right side is calculated, and a probability of moving to the left and a probability of moving to the right are obtained based on the ratio; A correspondence between the standard deviation interval and the step length is preset, and the step length is determined according to the standard deviation interval in which the standard deviation of the window is located, wherein the average value of the standard deviation interval and the step length are inversely correlated.

[0034] Preferably, after the step length is moved up or down in the vertical direction with the same probability, the method further comprises: If the value of the movable point in the time-frequency diagram after the movement is not less than the value before the vertical movement, the vertical movement direction selected with equal probability is taken as the vertical movement mode.

[0035] Preferably, the moving probability and the non-moving probability are determined according to the movable point before vertical movement and the two points above and below the movable point, specifically: Obtaining the standard deviation of the movable point before vertical movement in the window, as well as the standard deviation of the upper point of the movable point before vertical movement in the window and the standard deviation of the lower point of the movable point in the window; weighting the standard deviation within the window of movable points before vertical movement based on the current temperature setting weight; The probability of not moving, the probability of moving upward and the probability of moving downward are calculated according to the weighted standard deviation of the movable point before the vertical movement, the standard deviation of the upper point and the standard deviation of the lower point.

[0036] Preferably, after determining the moving probability and the non-moving probability according to the movable point before vertical movement and the two points above and below the movable point, the method further comprises: The movable point is determined in a vertical moving direction according to the moving probability and the non-moving probability, and the movable point is moved in the vertical moving direction.

[0037] In a third embodiment, the present invention also provides an electronic device comprising at least one processor and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the asynchronous motor fault identification method based on time-frequency analysis.

[0038] In a fourth embodiment, the present disclosure further provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute the asynchronous motor fault identification method based on time-frequency analysis.

[0039] Through the description of the above implementation methods, technical personnel in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0040] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0041] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0042] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for identifying an asynchronous motor fault, characterized in that: The method comprises the following steps: The sensor equipment is used to collect the vibration signal generated during the operation of the asynchronous motor, and the collected vibration signal is subjected to time-frequency analysis to obtain a time-frequency diagram; Initialize multiple movable points on the time-frequency graph, calculate the standard deviation of the time-frequency graph in the window centered on the movable point, and determine the horizontal movement direction and step length of the movable point based on the standard deviation; after the horizontal movement, move up or down the step length in the vertical direction with the same probability, if the value of the movable point in the time-frequency graph after the movement is less than the value before the vertical movement, determine the movement probability and the non-movement probability based on the movable point before the vertical movement and the upper and lower points of the movable point; update the metric matrix after the movable point moves to the new position; The ridges of the time-frequency graph are extracted based on the final metric matrix, and the asynchronous motor faults corresponding to the ridges are identified.

2. The method according to claim 1, characterized in that The determining of the horizontal moving direction and step length of the movable point based on the standard deviation is specifically as follows: The window is divided into two parts, a left part and a right part, of equal size, and a ratio of a standard deviation on the left side to a standard deviation on the right side is calculated, and a probability of moving to the left and a probability of moving to the right are obtained based on the ratio; A correspondence between the standard deviation interval and the step length is preset, and the step length is determined according to the standard deviation interval in which the standard deviation of the window is located, wherein the average value of the standard deviation interval and the step length are inversely correlated.

3. The method according to claim 1, characterized in that After the step length is moved up or down in the vertical direction with the same probability, the method further includes: If the value of the movable point in the time-frequency diagram after the movement is not less than the value before the vertical movement, the vertical movement direction selected with equal probability is taken as the vertical movement mode.

4. The method according to claim 1, characterized in that The moving probability and the non-moving probability are determined according to the movable point before vertical movement and the two points above and below the movable point, specifically: Obtaining the standard deviation of the movable point before vertical movement in the window, as well as the standard deviation of the upper point of the movable point before vertical movement in the window and the standard deviation of the lower point of the movable point in the window; weighting the standard deviation within the window of movable points before vertical movement based on the current temperature setting weight; The probability of not moving, the probability of moving upward and the probability of moving downward are calculated according to the weighted standard deviation of the movable point before the vertical movement, the standard deviation of the upper point and the standard deviation of the lower point.

5. The method according to claim 4, characterized in that After determining the moving probability and the non-moving probability according to the movable point before vertical movement and the two points above and below the movable point, the method further includes: The movable point is determined in a vertical moving direction according to the moving probability and the non-moving probability, and the movable point is moved in the vertical moving direction.

6. An asynchronous motor fault identification system, characterized in that: The system includes the following modules: The vibration acquisition module is used to use sensor equipment to collect vibration signals generated during the operation of the asynchronous motor, and perform time-frequency analysis on the collected vibration signals to obtain a time-frequency graph; A feature extraction module is used to initialize multiple movable points on the time-frequency graph, calculate the standard deviation of the time-frequency graph in a window centered on the movable point, and determine the horizontal movement direction and step length of the movable point based on the standard deviation; after the horizontal movement, move up or down the step length in the vertical direction with the same probability, and if the value of the movable point in the time-frequency graph after the movement is less than the value before the vertical movement, determine the movement probability and the non-movement probability based on the movable point before the vertical movement and the upper and lower points of the movable point; Update the metric matrix after the movable point moves to a new position; The fault identification module is used to extract the ridge line of the time-frequency diagram based on the final metric matrix and identify the asynchronous motor fault corresponding to the ridge line.

7. The system according to claim 6, characterized in that The determining of the horizontal moving direction and step length of the movable point based on the standard deviation is specifically as follows: The window is divided into two parts, a left part and a right part, of equal size, and a ratio of a standard deviation on the left side to a standard deviation on the right side is calculated, and a probability of moving to the left and a probability of moving to the right are obtained based on the ratio; A correspondence between the standard deviation interval and the step length is preset, and the step length is determined according to the standard deviation interval in which the standard deviation of the window is located, wherein the average value of the standard deviation interval and the step length are inversely correlated.

8. The system according to claim 6, characterized in that After the step length is moved up or down in the vertical direction with the same probability, the method further includes: If the value of the movable point in the time-frequency diagram after the movement is not less than the value before the vertical movement, the vertical movement direction selected with equal probability is taken as the vertical movement mode.

9. The method according to claim 6, characterized in that The moving probability and the non-moving probability are determined according to the movable point before vertical movement and the two points above and below the movable point, specifically: Obtaining the standard deviation of the movable point before vertical movement in the window, as well as the standard deviation of the upper point of the movable point before vertical movement in the window and the standard deviation of the lower point of the movable point in the window; weighting the standard deviation within the window of movable points before vertical movement based on the current temperature setting weight; The probability of not moving, the probability of moving upward and the probability of moving downward are calculated according to the weighted standard deviation of the movable point before the vertical movement, the standard deviation of the upper point and the standard deviation of the lower point.

10. The system according to claim 9, characterized in that After determining the moving probability and the non-moving probability according to the movable point before vertical movement and the two points above and below the movable point, the method further includes: The movable point is determined in a vertical moving direction according to the moving probability and the non-moving probability, and the movable point is moved in the vertical moving direction.

Images