Machine tool bus relay module with intelligent identification and automatic connection function

Abstract

The application relates to the technical field of relays, in particular to a machine tool bus relay module with intelligent identification and automatic connection function, which comprises: a relay current signal extraction unit for sampling the current signal in the relay; a relay current noise component analysis unit for obtaining relative noise components based on the discrete degree of all amplitudes corresponding to the frequencies other than the fundamental wave and the harmonic wave in each window and the difference between the amplitudes corresponding to the same frequency in different windows; a relay current signal denoising unit for obtaining a global optimization function and a local optimization function, obtaining the global optimal demand degree of each window based on the time difference between the windows and the amplitude difference between the harmonic wave and the fundamental wave, and obtaining an adaptive function for denoising the current signal; and a relay automatic connection unit for judging whether the machine tool bus relay is disconnected. The application aims to improve the identification accuracy of automatic disconnection of the machine tool bus relay.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of relays, in particular to a machine tool bus relay module with intelligent identification and automatic connection function. BACKGROUND

[0002] In a machine tool control system, the bus relay module serves as an important hub connecting the execution units and the control system. Its technical evolution and intelligent level directly affect the equipment operation efficiency and maintenance cost. The bus relay module has the functions of centralized management and monitoring. The control system can obtain the state information of each execution unit in real time through the bus, which is convenient for fault diagnosis and maintenance. When a fault occurs in an execution unit, the bus relay module can isolate the faulty unit through automatic control to ensure the normal operation of other parts, thereby improving the reliability and stability of the system.

[0003] When using the multiple average adaptive harmonic analysis algorithm to denoise the bus relay current, the selection of the window length needs to balance between noise suppression and dynamic response. If the window is too long, transient harmonic signals may be smoothed, and short-term fluctuations in current changes cannot be captured in time. If the window is too short, random noise may not be effectively filtered out, affecting the denoising effect. This improper selection will affect the timely identification of fault signals by the machine tool bus relay, and further affect the accuracy of automatic connection and the reliability of the system. SUMMARY

[0004] In view of the above, it is necessary to provide a machine tool bus relay module with intelligent identification and automatic connection function to solve the above problems.

[0005] One embodiment of the application provides a machine tool bus relay module with intelligent identification and automatic connection function, which comprises:

[0006] A relay current signal extraction unit for sampling the current signal in the relay;

[0007] A relay current noise component analysis unit for analyzing the frequency spectrum characteristics of the current signal in the preset window length, extracting the THD value of each window, and obtaining the relative noise component of each window based on the discrete degree of all amplitudes corresponding to the frequencies other than the fundamental wave and harmonics in each window, and the difference between the amplitudes corresponding to the same frequency between each window and the remaining windows.

[0008] The relay current signal denoising unit is configured to analyze a window size for denoising the current signal in the history and a difference distribution of THD values corresponding to each window size before and after denoising, to obtain a global optimization function; obtain a local optimization function according to a distribution of relative noise components of all windows; obtain a global optimal demand degree of each window based on a time difference between each window and the remaining windows and an amplitude difference between harmonics and a fundamental wave; assign weights to the global optimization function and the local optimization function according to a distribution of the global optimal demand degrees of all windows, to obtain an adaptive function, and use an optimization algorithm and a filtering algorithm to denoise the current signal.

[0009] The relay automatic connection unit is configured to use a neural network to determine whether the machine tool bus relay is disconnected according to the denoised current signal.

[0010] Preferably, the specific steps for obtaining the relative noise component of each window include:

[0011] The flatness of each window is obtained according to a discrete degree of amplitudes of all frequencies of non-fundamental waves and non-harmonics in each window.

[0012] The ratio between the amplitude corresponding to each same frequency of each window and the remaining windows is calculated and recorded as a first ratio; the mean value of all first ratios obtained by each window and all the remaining windows is calculated, and is positively fused with the flatness to obtain the relative noise component of each window.

[0013] Preferably, the flatness of each window is specifically a negative correlation mapping result of amplitude variance of all frequencies of non-fundamental waves and non-harmonics in each window.

[0014] Preferably, the global optimization function is obtained by:

[0015] The size of all filtering windows in a plurality of denoising processes in the history is obtained, and the THD value difference mean value of each window size is calculated by combining the THD values before and after denoising of each window.

[0016] The window size and the corresponding THD difference mean value are linearly fitted, and the function of the fitted straight line is taken as the global optimization function.

[0017] Preferably, the process of calculating the THD value difference mean value of each window size includes:

[0018] The THD difference before and after denoising of each window is calculated, and the mean value of all THD differences of the same window size is taken as the THD difference mean value of each window size.

[0019] Preferably, the local optimization function is specifically the discrete degree of the relative noise component of all windows.

[0020] Preferably, the global optimal demand degree of each window is obtained, specifically:

[0021] The minimum time interval between window i and window q is obtained, denoted as The second ratio between the amplitude of the xth harmonic wave in window i and window q and the amplitude of the fundamental wave is respectively denoted as 、 The global optimal demand degree of window i is denoted as , and the formula is: ; wherein, exp() represents the exponential function with the natural constant as the base number; represents the number of windows; represents the number of harmonic frequencies of each window.

[0022] Preferably, the adaptive function is obtained, specifically:

[0023] The mean value of the global optimal demand degree of all windows is taken as the weight of the global optimization function; the difference between 1 and the mean value of the global optimal demand degree is taken as the weight of the local optimization function, and the global optimization function with the assigned weight is subtracted from the local optimization function with the assigned weight to obtain the adaptive function of the particle swarm optimization algorithm.

[0024] Preferably, the current signal is denoised by using the optimization algorithm and the filtering algorithm, specifically:

[0025] The optimal window size is obtained by using the particle swarm optimization algorithm combined with the adaptive function; the denoised current signal is obtained by using the multiple average adaptive harmonic analysis algorithm combined with the optimal window size.

[0026] Preferably, the condition for judging the disconnection of the machine tool bus relay is that the denoised current signal is analyzed by using the neural network, the obtained abnormal score is normalized, and when the normalized abnormal score exceeds the preset threshold, the relay disconnection operation is triggered.

[0027] The present application has at least the following beneficial effects:

[0028] The embodiment of the application obtains the relative noise component by analyzing the distribution characteristics of the amplitudes of other frequencies between the fundamental wave and the harmonic wave in the frequency domain, and combining the differences between the amplitudes of the frequencies in different windows at the same frequency, thereby judging the noise component in the corresponding window, and representing the complexity of the noise component in all windows through the distribution of the relative noise component; meanwhile, the fitting linear function of the THD difference value and the window size is obtained by separating the change relationship between the historical window sizes and the THD values, thereby constructing the local optimization function and the global optimization function in the optimization algorithm, which avoids the influence of the changes of the amplitudes of various frequencies in different window sizes on the selection of the optimal window, improves the universality of the optimal window, and makes the subsequent denoising effect better; the weights of the global optimal function and the local optimal function are obtained by analyzing the size of the harmonic wave in different windows, and combining the relationship between the window size change and the noise suppression and the harmonic coverage ability, thereby avoiding the over-coverage or incomplete denoising of the harmonic amplitude at a specific time period and a specific frequency caused by obtaining the optimal window based on only the signal characteristics, improving the accuracy of the optimal window selection during denoising, improving the denoising effect of the current signal, and further increasing the recognition accuracy of the automatic disconnection of the machine tool bus relay, avoiding equipment damage. BRIEF DESCRIPTION OF DRAWINGS

[0029] Figure 1 A machine tool bus relay module with intelligent identification and automatic connection function is provided in the application.

[0030] Figure 2 A specific flowchart of the automatic connection of the machine tool bus relay is provided in the application. DETAILED DESCRIPTION

[0031] In the description of the embodiments of the application, the words "exemplary", "or", "for example" are used to mean as an example, illustration, or description. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of the application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the use of "exemplary", "or", "for example" is intended to present the relevant concept in a specific manner.

[0032] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the application belongs. The terms used in the specification of the application are only for the purpose of describing the specific embodiments and are not intended to limit the application.

[0033] It should be noted that the terms "first", "second" in the present application and its drawings are used to distinguish similar objects, rather than to describe a specific order or sequence. The method disclosed in the embodiments of the present application or the method shown in the flowchart includes one or more steps for implementing the method, and the execution order of the steps can be interchanged with each other without departing from the scope of the present application, and some steps can also be deleted.

[0034] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0035] The specific scheme of the machine tool bus relay module with intelligent identification and automatic connection function provided by the present application will be specifically described below in combination with the drawings.

[0036] Please refer to Figure 1 , which shows a block diagram of a machine tool bus relay module with intelligent identification and automatic connection function provided by an embodiment of the present application, which includes: a relay current signal extraction unit, a relay current noise component analysis unit, a relay current signal denoising unit, and a relay automatic connection unit.

[0037] The embodiment of the present application first proposes a machine tool bus relay module with intelligent identification and automatic connection function, which is applied to the field of relay technology, and the module includes:

[0038] The relay current signal extraction unit is used to sample the current signal in the relay.

[0039] By using the STM32 MPU processor to sample the current signal in the relay in real time, it is convenient to extract the frequency component, harmonic component and other characteristic information in the power grid current signal. In the embodiment, 1 min of current signal data is extracted; the implementer can adjust it according to the actual situation.

[0040] The relay current noise component analysis unit is used to analyze the frequency spectrum characteristics of the current signal in the preset window length, and based on the discrete degree of all amplitudes corresponding to the frequencies other than the fundamental wave and the harmonic wave in each window, and the difference between the amplitudes corresponding to the same frequency between each window and the remaining windows, the relative noise component of each window is obtained.

[0041] Since the window length of multiple averages needs to find a balance between noise suppression and dynamic response, it is very important to choose the appropriate window length. Too long window may cause transient harmonics to be smoothed out, so that these rapidly changing signals cannot be captured; while too short window may not effectively remove random noise, resulting in unstable results.

[0042] Therefore, firstly, whether the current window length is appropriate is determined by analyzing the noise component in the window to ensure that the noise can be effectively suppressed and the response to transient harmonics is not affected.

[0043] The window length is set as D, the current signal of each window is obtained by Fourier transform to obtain the corresponding frequency spectrum, and the total harmonic distortion (THD) of each window is extracted.

[0044] Noise affects the amplitude of multiple frequencies, which is usually manifested as a broadband background noise rise in the frequency spectrum of the current. Since the noise is generally Gaussian noise, its power spectral density is approximately uniformly distributed at different frequencies, so the amplitude fluctuation in the frequency spectrum is relatively stable and does not show obvious peaks. In order to avoid misjudgment of the noise content due to the excessive amplitude of the fundamental wave and its harmonics, the embodiment takes 50Hz as the fundamental frequency, and considers all collected frequencies that are integer multiples of 50Hz as harmonics. The negative correlation mapping result of the amplitude variance of all frequencies corresponding to non-fundamental and non-harmonic frequencies in each window is recorded as the flatness of each window. In this embodiment, the negative correlation mapping result of the variable is calculated by the reciprocal of the variable. It should be noted that, in order to prevent the denominator from being 0, a preset parameter is needed, and the preset parameter is 0.01 in this embodiment. In the frequency spectrum of the current signal, the more flat the amplitude of the frequency other than the fundamental wave and the harmonic is, that is, the greater the value of the flatness is, the more obvious the influence of the noise on the current signal is, and the smaller the value of the flatness is, the more actual current signal of other frequencies exists.

[0045] The influence of noise on the current signal in the local window is randomly changed in different time periods, so the noise content in different windows is different. Since the noise is Gaussian distributed in all sampled current signals, the noise component characteristics in the current window can be obtained by comparing the relationship between the amplitudes of the same frequency in different windows.

[0046] Specifically, the ratio of the amplitude corresponding to each same frequency between each window and the rest of the windows is calculated, which is recorded as the first ratio. The mean value of all first ratios obtained by each window and all other windows is calculated, and is positively fused with the flatness to obtain the relative noise component of each window. In this embodiment, the calculation method of multiplication is used for the positive fusion of multiple variables.

[0047] It should be understood that the greater the flatness is, the more obvious the noise component interference in the current window is, and the higher the reference value of the noise component is. The greater the first ratio is, the more noise components exist in the window corresponding to the frequency compared with other windows. Further, the greater the relative noise component of each window is.

[0048] The relay current signal denoising unit is used for analyzing a window size for denoising the current signal in history and a difference distribution of a THD value corresponding to each window size before and after denoising, obtaining a global optimization function; obtaining a local optimization function according to a distribution of a relative noise component of all windows; obtaining a global optimal demand degree of each window based on a time difference between each window and the remaining windows and an amplitude difference between a harmonic and a fundamental wave; according to a distribution of the global optimal demand degree of all windows, weighting the global optimization function and the local optimization function respectively to obtain an adaptive function, and using an optimization algorithm and a filtering algorithm to denoise the current signal.

[0049] The application adopts a particle swarm optimization algorithm to automatically search for an optimal window size of an optimal filter, obtains sizes of all filter windows in a plurality of adaptive harmonic denoising processes in history, each window size corresponding to a difference between two THD values before and after denoising, in the embodiment, the THD value before denoising is taken as a minuend, the THD value after denoising is taken as a subtrahend, and the sign of the difference value is kept; the greater the difference is, the stronger the corresponding window is in retaining harmonic characteristics.

[0050] It should be noted that since one window size may correspond to a plurality of denoising processes, a plurality of differences between THD values before and after denoising correspond to the window size, therefore, the mean value of the differences between the corresponding THD values before and after denoising under the same window size is obtained; at this time, one window size corresponds to one THD difference mean value.

[0051] The window size is taken as an abscissa, the THD difference mean value is taken as an ordinate, a straight line fitting is performed, a fitting straight line is obtained, and a function of the fitting straight line is taken as a global optimization function; the fitting straight line makes the perpendicular distance from the data points corresponding to each THD difference mean value to the fitting straight line minimum.

[0052] The more complex the noise components in different windows are, the smaller the possibility of local optimization in the window iteration process is, and vice versa, the simpler the noise components are, the greater the possibility of local optimization in the window iteration process is. Therefore, the dispersion of the relative noise components of all windows is obtained to obtain a local optimization function. In the embodiment, the dispersion of a plurality of variables is calculated by using variance.

[0053] Since abnormal components in the current signal exist random fluctuations, at this time, the particle swarm algorithm is used to obtain an optimal window, and the demand for local or global optimization is different.

[0054] When an abnormal current signal exists in the machine tool bus relay module, the harmonic amplitude will significantly increase within a specific time period and at a specific frequency. The higher the harmonic frequency within the time period of the window, the higher the window's requirement for global optimization. Therefore, for any window, the ratio between each harmonic amplitude and the fundamental amplitude is obtained, denoted as the second ratio. Simultaneously, the smaller the time interval between other windows and the current window in the time domain, and the smaller the difference in the ratios of harmonics of the same frequency to the fundamental frequency, the higher the requirement for global optimization. Therefore, the minimum time interval between window i and window q is obtained, denoted as... The second ratio between the x-th harmonic amplitude and the fundamental amplitude in windows i and q is denoted as follows: , The globally optimal demand degree of window i is denoted as Its formula is as follows: ; where exp() represents an exponential function with the natural constant as the base; Indicates the number of windows; Indicates the number of harmonic frequencies.

[0055] It should be noted that when the minimum time interval is 0, a parameter value of 1 needs to be added to avoid the denominator being 0 and unable to be calculated.

[0056] The mean of the global optimal demand for all windows is used as the weight of the global optimization function; the difference between 1 and the mean of the global optimal demand is used as the weight of the local optimization function. The global optimization function with assigned weights is subtracted from the local optimization function with assigned weights to obtain the fitness function of the particle swarm optimization algorithm.

[0057] Initialize the relevant parameters of the particle swarm optimization algorithm. The position of each particle represents the window length value, which is set in the range of [10, 50] in this embodiment. The remaining parameters are the default sizes of the algorithm. Obtain the optimal window size based on the particle swarm optimization algorithm, denoted as d. Further, use the multiple average adaptive harmonic analysis algorithm to process the currently monitored current signal, thereby obtaining the denoised power grid current signal.

[0058] The automatic connection unit is used to train the current signal using a neural network and determine whether the machine tool bus relay is disconnected based on the denoised current signal.

[0059] The machine tool bus relay module monitors the state of the bus in real time. Specifically, the current signals corresponding to various types of faults and normal current signals are trained through a neural network DNN algorithm. Subsequently, the real-time current signals collected are input to score the current signals for abnormalities, and the scores are divided between [0, 1]. In this embodiment, the threshold value is set to 0.3. When the abnormal score is greater than the threshold value, it is determined that there is an abnormality in the current passing through the machine tool bus relay at the current time, indicating that the main station bus signal is lost or the signal check is incorrect. The machine tool bus relay module will immediately clear the output and disconnect to avoid equipment damage due to incorrect data, and an alarm signal is sent through warning lights, buzzers, and other means to alert the operator to handle it in a timely manner. When the bus state returns to normal, i.e., the abnormal value is less than or equal to the threshold value, the module will immediately restore data transmission with the bus master station to ensure the continuity of communication.

[0060] The specific flowchart of the automatic connection of the machine tool bus relay is shown in Figure 2

[0061] In summary, the embodiments of the present application analyze the distribution characteristics of the amplitudes of other frequencies between the fundamental wave and the harmonic wave in the frequency domain, combine the differences between the amplitudes of the frequencies in the same frequency under different windows, obtain the relative noise component, determine the noise component in the corresponding window, and represent the complexity of the noise component in all windows through the distribution of the relative noise component. At the same time, the change relationship between the historical window size and the THD value is separated to obtain the fitting linear function of the THD difference and the window size, thereby constructing the local optimization function and the global optimization function in the optimization algorithm. This operation avoids the influence of the changes of the amplitudes of various frequencies in different window sizes on the selection of the optimal window, improves the universality of the optimal window, and makes the subsequent denoising effect better. By analyzing the size of the harmonic wave in different windows and combining the relationship between the window size change and the noise suppression and harmonic coverage ability, the weights of the global optimal function and the local optimal function are obtained, thereby avoiding the over-coverage of the harmonic amplitude or the incomplete denoising of a specific time period and a specific frequency caused by obtaining the optimal window based only on the signal characteristics, improving the accuracy of the optimal window selection during denoising, improving the denoising effect of the current signal, and further increasing the recognition accuracy of the machine tool bus relay trigger automatic disconnection, avoiding equipment damage.

[0062] ​The computer program product of the present application can be a computer program implemented on one or more computers. The program instructions can be stored on a computer readable medium, such as a hard disk, CD-ROM, optical storage, or any other tangible medium. The program instructions can be downloaded from the Internet or another network. The program instructions can be embodied in a carrier wave traveling over the Internet or other network. The computer readable medium can be a machine readable storage device, a machine readable transmission device, or a combination of both. The computer readable medium can be a computer readable storage device, a computer readable transmission device, or a combination of both.

[0063] The above embodiments are only used to illustrate the technical solutions of the present application, not limit the technical solutions of the present application; although the technical solutions of the present application are described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A machine tool bus relay module having an intelligent recognition automatic connection function, characterized by, The module comprises: A relay current signal extraction unit for sampling a current signal in a relay; A relay current noise component analysis unit for analyzing spectral characteristics of the current signal in a preset window length, extracting a THD value of each window, and obtaining a relative noise component of each window based on a discrete degree of all amplitudes corresponding to frequencies other than a fundamental wave and a harmonic wave in each window and a difference between amplitudes corresponding to the same frequencies between each window and the remaining windows; A relay current signal denoising unit for analyzing a window size in a history in which the current signal is denoised, a difference distribution of THD values corresponding to each window size before and after denoising, obtaining a global optimization function, obtaining a local optimization function based on a distribution of the relative noise components of all the windows, obtaining a global optimal demand degree of each window based on a time difference between each window and the remaining windows and an amplitude difference between the harmonic wave and the fundamental wave, and respectively weighting the global optimization function and the local optimization function based on a distribution of the global optimal demand degrees of all the windows to obtain an adaptive function and denoising the current signal by using an optimization algorithm and a filtering algorithm; A relay automatic connection unit for judging whether a machine tool bus relay is disconnected based on the denoised current signal by using a neural network; The global optimal demand degree of each window is obtained by: The minimum time interval between the acquisition window i and the window q is denoted as The second ratio between the xth harmonic amplitude in the window i and the fundamental amplitude is denoted as , The global optimal demand degree of the window i is denoted as , and the formula is as follows: , wherein exp() represents an exponential function with a natural constant as a base number; represents the number of windows; represents the number of harmonic frequencies of each window. The adaptive function is obtained by: Taking a mean value of the global optimal demand degrees of all the windows as a weight value of the global optimization function, taking a difference between 1 and the mean value of the global optimal demand degrees as a weight value of the local optimization function, subtracting the global optimization function with the weight value from the local optimization function with the weight value to obtain an adaptive function of a particle swarm optimization algorithm.

2. A machine tool bus relay module with intelligent identification and automatic connection function according to claim 1, characterized in that, The relative noise component of each window is obtained by: Obtaining a flatness of each window based on a discrete degree of amplitudes of all frequencies other than the fundamental wave and the harmonic wave in each window; Calculating a ratio between amplitudes corresponding to the same frequency between each window and the remaining windows, denoted as a first ratio, and calculating a mean value of all the first ratios obtained by each window and all the remaining windows, and positively fusing the mean value with the flatness to obtain the relative noise component of each window.

3. A machine tool bus relay module with intelligent identification and automatic connection function according to claim 2, characterized in that, The flatness of each window is a negative correlation mapping result of amplitude variances of all frequencies corresponding to the non-fundamental wave and the non-harmonic wave in each window.

4. The machine tool bus relay module with intelligent identification and automatic connection function according to claim 1, characterized in that, The global optimization function is obtained by: Obtaining sizes of all filtering windows in a history of a plurality of times of denoising processes, combining THD values before and after denoising of each window, and calculating a THD value difference mean value of each window size; Based on the window size and the THD value difference mean value corresponding to the window size, a straight line fitting is performed, and a function of the fitted straight line is taken as the global optimization function.

5. A machine tool bus relay module with intelligent identification and automatic connection function according to claim 4, characterized in that, The process of calculating the THD value difference mean value of each window size is: Calculating THD differences before and after denoising of each window, and taking a mean value of all the THD differences of the same window size as the THD difference mean value of each window size.

6. A machine tool bus relay module with intelligent identification and automatic connection function according to claim 1, characterized in that, The local optimization function is a discrete degree of the relative noise components of all the windows.

7. The machine tool bus relay module with intelligent identification and automatic connection function according to claim 1, characterized in that, The denoising of the current signal by using the optimization algorithm and the filtering algorithm is specifically: The optimal window size is obtained by using particle swarm optimization algorithm combined with adaptive function, and the denoised current signal is obtained by using multiple average adaptive harmonic analysis algorithm combined with the optimal window size.

8. The machine tool bus relay module with intelligent identification and automatic connection function according to claim 1, characterized in that, The condition for judging the disconnection of the machine tool bus relay is that the abnormal score obtained by analyzing the denoised current signal through the neural network is normalized, and when the normalized abnormal score exceeds the preset threshold, the relay disconnection operation is triggered.

Images