CN113988140A - Signal time domain feature analysis method and device, electronic equipment and storage medium - Google Patents

Signal time domain feature analysis method and device, electronic equipment and storage medium Download PDF

Info

Publication number
CN113988140A
CN113988140A CN202111360196.1A CN202111360196A CN113988140A CN 113988140 A CN113988140 A CN 113988140A CN 202111360196 A CN202111360196 A CN 202111360196A CN 113988140 A CN113988140 A CN 113988140A
Authority
CN
China
Prior art keywords
signal
analyzed
segment
value
abnormal
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202111360196.1A
Other languages
Chinese (zh)
Inventor
方晟堃
李平
邓志文
陈启愉
冼荣彬
张华伟
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Institute of Intelligent Manufacturing of Guangdong Academy of Sciences
Original Assignee
Institute of Intelligent Manufacturing of Guangdong Academy of Sciences
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Institute of Intelligent Manufacturing of Guangdong Academy of Sciences filed Critical Institute of Intelligent Manufacturing of Guangdong Academy of Sciences
Priority to CN202111360196.1A priority Critical patent/CN113988140A/en
Publication of CN113988140A publication Critical patent/CN113988140A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/08—Feature extraction
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/02—Preprocessing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/12—Classification; Matching
    • G06F2218/16—Classification; Matching by matching signal segments

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Algebra (AREA)
  • Computational Mathematics (AREA)
  • Signal Processing (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Artificial Intelligence (AREA)
  • Pure & Applied Mathematics (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • Testing And Monitoring For Control Systems (AREA)

Abstract

The invention provides a signal time domain feature analysis method and device, electronic equipment and a storage medium, and relates to the field of signal analysis and processing. The method comprises the steps of firstly obtaining signals to be analyzed with one-dimensional time domain dispersion, dividing the signals to be analyzed to obtain a plurality of signal segments, then determining a convolution kernel of each signal segment according to a signal value of each signal segment, carrying out convolution operation on each signal segment according to the convolution kernel of each signal segment to obtain a characteristic analysis value of each signal segment, and finally analyzing the signals to be analyzed according to the characteristic analysis values of all the signal segments of the signals to be analyzed to determine abnormal signals in the signals to be analyzed. By the method, the extracted signal change characteristics of the signal to be analyzed can be more comprehensive and accurate, and the abnormal signal in the signal to be analyzed can be more accurately determined according to the signal change characteristics.

Description

Signal time domain feature analysis method and device, electronic equipment and storage medium
Technical Field
The invention relates to the field of signal analysis and processing, in particular to a signal time domain feature analysis method and device, electronic equipment and a storage medium.
Background
In the industrial field, the working condition and the state of the equipment can be obtained by analyzing signal data such as real-time working voltage, current and the like of the electric drive equipment, wherein when the equipment normally works, the signal can show regular periodic fluctuation, and when abnormal conditions occur in the working process of the equipment or the equipment, such as motor bearing faults, coil faults, protective gas interruption in the welding process, welding abnormity caused by wire feeding faults and the like, abnormal signals can be generated. Theoretically, the abnormal signal appears as a signal cycle or a signal value changing relative to the normal signal, but in actual industrial production, the abnormal signal and the normal signal are superposed, and the abnormal signal cannot be accurately determined according to the extracted signal characteristic by using the existing signal characteristic analysis method for extracting the variance.
Disclosure of Invention
The embodiment of the invention provides a signal time domain feature analysis method, a signal time domain feature analysis device, terminal equipment and a storage medium, which can overcome the defects of the prior art.
In order to achieve the above purpose, the embodiment of the present invention adopts the following technical solutions:
in a first aspect, an embodiment of the present invention provides a signal time domain feature analysis method, where the signal time domain feature analysis method includes:
acquiring a signal to be analyzed with one-dimensional time domain dispersion;
dividing the signal to be analyzed to obtain a plurality of signal segments;
determining a convolution kernel for each of the signal segments based on the signal value for each of the signal segments;
performing convolution operation on each signal segment according to the convolution core of each signal segment to obtain a characteristic analysis value of each signal segment;
and analyzing the signal to be analyzed according to the characteristic analysis values of all the signal segments of the signal to be analyzed, and determining abnormal signals in the signal to be analyzed.
As a possible implementation manner, each of the signal segments includes a plurality of time instants, each of the time instants corresponds to a signal value, and the step of determining the convolution kernel of each of the signal segments according to the signal value of each of the signal segments includes:
calculating the mean value of the signal values of all the moments in each signal segment to obtain the signal mean value of each signal segment;
obtaining the signal deviation degree of each moment according to the signal mean value of each signal segment and the signal value of each moment;
and constructing a convolution kernel of each signal segment according to the signal deviation degrees of all the moments of each signal segment.
As a possible implementation, the expression of the convolution kernel of any target signal segment is:
Figure BDA0003358921440000021
wherein the starting time of the target signal segment is tau, and the ending time is tau + omega, xtIs the signal value of the target signal segment at time t, mu is the signal mean value of the target signal segment, (x)t-μ)2And a is the signal deviation degree of the target signal segment at the time t, a is a preset constant, and f is the convolution kernel of the target signal segment.
As a possible implementation manner, the step of dividing the signal to be analyzed to obtain a plurality of signal segments includes:
and sliding a preset analysis window from the starting time of the signal to be analyzed according to a preset step length, and sequentially intercepting signal segments from the signal to be analyzed to obtain a plurality of signal segments, wherein the preset analysis window is slid each time to obtain one signal segment.
As a possible implementation manner, the step of performing a convolution operation on each signal segment according to the convolution kernel of each signal segment to obtain a feature analysis value of each signal segment includes:
for any target signal segment, adopting a formula according to the convolution kernel of the target signal segment
Figure BDA0003358921440000031
Performing convolution operation to obtain a feature analysis value of the target signal segment, wherein f is a convolution kernel of the target signal segment, g is the target signal segment, the starting time of the target signal segment is τ, the ending time of the target signal segment is τ + ω, and g (t) ═ xtAnd lambda is the signal value of the target signal segment at the time t, and is the characteristic analysis value of the target signal segment.
As a possible implementation manner, analyzing the signal to be analyzed according to feature analysis values of all signal segments of the signal to be analyzed, and determining an abnormal signal in the signal to be analyzed includes:
generating a characteristic curve of the signal to be analyzed according to the characteristic analysis values of all the signal segments of the signal to be analyzed;
and analyzing the characteristic curve of the signal to be analyzed, and determining an abnormal signal in the signal to be analyzed.
As a possible implementation manner, the abnormal signal includes an abnormal time and an abnormal signal value, the step of analyzing the characteristic curve of the signal to be analyzed and determining the abnormal signal in the signal to be analyzed includes:
taking a signal segment corresponding to a characteristic analysis value which is larger than a preset threshold value in a characteristic curve of the signal to be analyzed as an abnormal signal segment;
taking the starting time of the abnormal signal segment as an abnormal time;
and taking the signal value at the abnormal moment as the abnormal signal value.
In a second aspect, an embodiment of the present invention provides a signal time domain feature analysis device, where the signal time domain feature analysis device includes:
the acquisition module is used for acquiring a signal to be analyzed with one-dimensional time domain discrete;
the segmentation module is used for dividing the signal to be analyzed to obtain a plurality of signal segments;
a determining module, configured to determine a convolution kernel of each signal segment according to a signal value of each signal segment;
the operation module is used for carrying out convolution operation on each signal segment according to the convolution core of each signal segment to obtain a characteristic analysis value of each signal segment;
and the analysis module is used for analyzing the signal to be analyzed according to the characteristic analysis values of all the signal segments of the signal to be analyzed and determining an abnormal signal in the signal to be analyzed.
In a third aspect, an embodiment of the present invention provides an electronic device, which includes a memory and a processor, where the memory is used for storing a computer program; the processor is used for executing the signal time domain feature analysis method provided by the embodiment of the invention when the computer program is called.
In a fourth aspect, an embodiment of the present invention provides a readable storage medium, where a computer program is stored on the readable storage medium, and when the computer program is executed by a processor, the computer program implements the signal time domain feature analysis method provided by the embodiment of the present invention.
Compared with the prior art, the signal time domain feature analysis method, the signal time domain feature analysis device, the electronic device and the storage medium provided by the embodiment of the invention are characterized in that a signal to be analyzed with discrete one-dimensional time domain is firstly obtained, the signal to be analyzed is divided to obtain a plurality of signal segments, then a convolution kernel of each signal segment is determined according to a signal value of each signal segment, then a convolution operation is performed on each signal segment according to the convolution kernel of each signal segment to obtain a feature analysis value of each signal segment, and finally the signal to be analyzed is analyzed according to the feature analysis values of all the signal segments of the signal to be analyzed to determine abnormal signals in the signal to be analyzed. According to the embodiment of the invention, the convolution kernel which is adaptive to the signal change characteristic of each signal segment is constructed for each signal segment, the convolution kernel of each signal segment is utilized to carry out convolution operation on each signal segment, the characteristic analysis value of each signal segment is obtained, the extracted signal change characteristic of the signal to be analyzed is more comprehensive and accurate, and the abnormal signal in the signal to be analyzed is more accurately determined according to the signal change characteristic.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
Fig. 1 is a flowchart of a signal time domain feature analysis method provided in this embodiment;
fig. 2 is a schematic diagram of dividing a signal to be analyzed into a plurality of signal segments according to an embodiment of the present invention;
FIG. 3 is a flow chart of determining a convolution kernel based on signal values of signal segments according to an embodiment of the present invention;
FIG. 4 is a diagram illustrating a convolution operation according to an embodiment of the present invention;
fig. 5 is a flowchart illustrating a method for determining an abnormal signal in a signal to be analyzed based on a feature analysis value of a signal segment according to an embodiment of the present invention;
FIG. 6A is a diagram illustrating normal welding results provided by an embodiment of the present invention;
fig. 6B is a schematic diagram illustrating an application effect according to an embodiment of the present invention;
FIG. 6C is a diagram illustrating an abnormal welding result according to an embodiment of the present invention;
fig. 6D is a schematic diagram illustrating another application effect according to an embodiment of the present invention;
FIG. 6E is a comparison graph of the application effects of two methods provided by the embodiment of the present invention;
fig. 7 is a schematic block diagram of a signal time domain feature analysis apparatus according to an embodiment of the present invention;
fig. 8 is a block diagram schematically illustrating a structure of an electronic device according to an embodiment of the present invention.
Icon: 100-signal time domain feature analysis means; 101-an acquisition module; 102-a segmentation module; 103-a determination module; 104-operation module; 105-an analysis module; 200-an electronic device; 210-a memory; 220-processor.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.
In the description of the present invention, it should be noted that if the terms "upper", "lower", "inside", "outside", etc. indicate an orientation or a positional relationship based on that shown in the drawings or that the product of the present invention is used as it is, this is only for convenience of description and simplification of the description, and it does not indicate or imply that the device or the element referred to must have a specific orientation, be constructed in a specific orientation, and be operated, and thus should not be construed as limiting the present invention.
Furthermore, the appearances of the terms "first," "second," and the like, if any, are used solely to distinguish one from another and are not to be construed as indicating or implying relative importance.
It should be noted that the features of the embodiments of the present invention may be combined with each other without conflict.
In the industrial field, the working voltage or current data of the equipment collected in real time can reflect the working condition and state of the equipment, taking the welding machine equipment as an example, the working state of the welding machine in the welding process is judged by acquiring the working voltage or current data of the welding machine in the welding process, and welding abnormity caused by motor bearing fault, coil fault, shielding gas blocking, wire feeding fault, fusion penetration and the like is found in time. The method comprises the steps that working voltage or current data of a welding machine in the welding process collected in real time can be regarded as one-dimensional time domain discrete signals, when the working state of the welding machine is normal, the signals usually show regular periodic fluctuation, the signals can be regarded as normal signals, when the welding process is abnormal, the period or signal value of the signals can change, the signals can be regarded as abnormal signals, the abnormal signals and the normal signals are overlapped, the existing method for extracting the signal change characteristics through variance is adopted, in the extracted signal change characteristics, the normal signals and the abnormal signals are not obviously different, the time when the abnormal signals appear cannot be accurately positioned, the abnormal signals can not be stripped from the signals, and the factor which causes the abnormal signals to appear in subsequent analysis is not facilitated.
In order to overcome the defects of the prior art, the signal change characteristics can be extracted by performing convolution operation on the signal. The convolution operation process is that a convolution kernel and a signal are subjected to sliding weighted summation, and the change characteristics of the signal are represented according to a series of values obtained by multiple sliding. Although the signal variation features extracted by using different convolution kernels are different, the convolution kernels are fixed in the convolution operation process, so that the variation features of the signal cannot be comprehensively and accurately extracted by using any convolution kernel.
Embodiments of the present invention provide a method and an apparatus for analyzing a time domain feature of a signal, an electronic device, and a storage medium, which are capable of comprehensively and accurately extracting a change feature of a signal and accurately determining an abnormal signal, and will be described in detail below.
Referring to fig. 1, fig. 1 is a flowchart of a signal time domain feature analysis method provided in the present embodiment, including steps S101 to S105.
Step S101, acquiring a signal to be analyzed with one-dimensional time domain discrete.
In the embodiment of the present invention, the signal to be analyzed is a one-dimensional time domain discrete signal, and includes a plurality of discrete time instants, and the time intervals of every two adjacent time instants are equal, and each time instant corresponds to a signal value.
Step S102, dividing the signal to be analyzed to obtain a plurality of signal segments.
In the embodiment of the present invention, it is determined whether a signal value at a certain time is a normal signal or an abnormal signal, and signal values corresponding to a plurality of other times adjacent to the certain time need to be combined, that is, a plurality of signal values corresponding to two adjacent times form a signal segment, and a signal to be analyzed may be divided into at least one signal segment.
Step S103, determining a convolution kernel of each signal segment according to the signal value of each signal segment.
In the embodiment of the present invention, each time in the signal to be analyzed corresponds to a signal value, and the time included in each signal segment is different, so the variation characteristic of the signal in each signal segment is different, and the convolution kernel of each signal segment is constructed based on the signal value corresponding to each time in each signal segment.
Step S104, performing convolution operation on each signal segment according to the convolution core of each signal segment to obtain a characteristic analysis value of each signal segment;
in the embodiment of the invention, each signal segment corresponds to one convolution kernel, each convolution kernel and the signal segment corresponding to the convolution kernel are subjected to convolution operation in sequence according to the sequence of signal segment division, and the obtained convolution operation results are respectively the characteristic analysis values of each signal segment.
Step S105, analyzing the signal to be analyzed according to the characteristic analysis values of all the signal segments of the signal to be analyzed, and determining an abnormal signal in the signal to be analyzed.
In the embodiment of the invention, the feature analysis value of each signal segment represents the change feature of the signal in the signal segment, the change feature of the signal to be analyzed is obtained according to the change features of the signals of all the signal segments, and the abnormal signal is determined according to the change feature of the signal to be analyzed.
The method provided by the embodiment of the invention has the advantages that a convolution kernel which is adaptive to the signal change characteristic of each signal segment is constructed for each signal segment, the convolution kernel of each signal segment is utilized to carry out convolution operation on each signal segment to obtain the characteristic analysis value of each signal segment, the signal change characteristic of the extracted signal to be analyzed is more comprehensive and accurate, and the abnormal signal in the signal to be analyzed is more accurately determined according to the signal change characteristic.
Based on fig. 1, the embodiment of the present invention further provides a specific implementation manner for dividing a signal to be analyzed, and step S102 may be implemented in the following manner:
and sliding a preset analysis window according to a preset step length from the initial moment of the signal to be analyzed, and sequentially intercepting signal segments from the signal to be analyzed to obtain a plurality of signal segments, wherein the preset analysis window is slid each time to obtain one signal segment.
In the embodiment of the invention, the signal segment is intercepted by using the preset analysis window, the size of the preset analysis window represents the number of discrete moments contained in the intercepted signal segment, the size of the signal segment can be determined by the period of the signal to be analyzed, and if the period of the signal to be analyzed is not easy to determine, the signal segment can be set according to experience. The preset step length is used for representing the number of discrete moments spaced by sliding the preset analysis window once.
In order to more intuitively show the process of dividing the signal to be analyzed into a plurality of signal segments, the embodiment of the invention uses the signal x (t) xtWhere t is 0, 1, 2, 3, …, n is the signal to be analyzed, and the size of the preset analysis window is 4, i.e. the preset analysis window is used to interceptFor example, please refer to fig. 2, and fig. 2 is a schematic diagram illustrating dividing a signal to be analyzed into a plurality of signal segments according to an embodiment of the present invention, where the number of discrete moments included in the signal segment is 4, and the preset step size is 1, that is, the number of discrete moments spaced by sliding the preset analysis window once is 1.
Firstly, aligning the start position of a preset analysis window with the time t of a signal x (t) to be analyzed being 0, and aligning the end position of the preset analysis window with the time t of the signal x (t) to be analyzed being 3;
then, the signal values at the time t-0, t-1, t-2 and t-3 of the signal x (t) to be analyzed within the preset analysis window are extracted as a signal segment g0(t)＝xt，t＝0，1，2，3；
Then, sliding a preset analysis window once to the right according to a preset step length, at this time, aligning the start position of the preset analysis window with the time t of the signal x (t) to be analyzed being 1, aligning the end position of the preset analysis window with the time t of the signal x (t) to be analyzed being 4, and then intercepting the signal values of the signal x (t) to be analyzed in the preset analysis window at the time t of 1, t of 2, t of 3 and t of 4 as a signal segment g1(t)＝xt，t＝1，2，3，4；
Sliding a preset analysis window to the right once according to a preset step length, at this time, aligning the start position of the preset analysis window with the time t of the signal x (t) to be analyzed being 2, aligning the end position of the preset analysis window with the time t of the signal x (t) to be analyzed being 5, and intercepting the signal values of the signal x (t) to be analyzed in the preset analysis window at the time t of 2, t of 3, t of 4 and t of 5 as a signal segment g2(t)＝t，t＝2，3，4，5；
And intercepting a signal segment every time the preset analysis window slides to the right, and ending the sliding of the preset analysis window when the ending position of the preset analysis window is aligned with the time t ═ n of the signal x (t) to be analyzed, so as to finish the division of the signal x (t) to be analyzed.
The dividing manner is a specific implementation manner for dividing the signal acquired in advance, and in fact, in the embodiment of the present invention, the signal to be analyzed acquired in real time may be divided in a manner of acquiring and dividing at the same time.
Referring to fig. 3, fig. 3 is a flowchart illustrating a method for determining a convolution kernel based on a signal value of a signal segment according to an embodiment of the present invention, where step S103 further includes sub-steps S103-1 to S103-3.
And a substep S103-1 of calculating the mean value of the signal values at all the moments in each signal segment to obtain the signal mean value of each signal segment.
In the embodiment of the present invention, the signal mean value of each signal segment is the sum of the signal values of all time instants of the signal segment divided by the number of time instants included in the signal segment, and reflects the concentration trend of the signal values of the signal segment.
And a substep S103-2 of obtaining the signal deviation degree at each moment according to the signal mean value of each signal segment and the signal value at each moment.
In the embodiment of the present invention, for any signal segment, the signal deviation degree at each time point is a square value of a difference between the signal value at each time point and a mean value of the signal segment, and reflects a deviation degree of the signal value at each time point and the mean value of the signal segment.
And a substep S103-3, constructing a convolution kernel of each signal segment according to the signal deviation degrees of all the time instants of each signal segment.
In the embodiment of the invention, the variation characteristics of the signals in each signal segment are different, and the convolution kernel of each signal segment is constructed based on the signal deviation degree corresponding to each time in each signal segment.
As a specific implementation, the convolution kernel expression of any target signal segment is:
Figure BDA0003358921440000101
in the above expression, the start time of the target signal segment is τ, and the end time is τ + ω, xtIs a target messageThe signal value of the signal segment at the time t, mu, is the signal mean of the target signal segment, (x)t-μ)2And f is the signal deviation degree of the target signal segment at the time t, a is a preset constant, and f is the convolution kernel of the target signal segment.
The convolution kernel of the target signal segment is the ratio of the exponential function of the signal deviation degree at each time and the linear function of the signal deviation degree, and when the signal value of each time in the target signal segment is normal signal, the signal deviation degree (x) of each time in the target signal segment ist-μ)2Close to a constant, the degree of signal deviation (x) at the moment of abnormality when an abnormal signal is present in the target signal segmentt-μ)2The difference between the normal signal and the abnormal signal is amplified by calculating the ratio of an exponential function about the signal deviation degree and a linear function about the signal deviation degree at each moment.
With signal segment g in FIG. 21(t)＝xtWhile t is 1, 2, 3, 4, which is an example of a target signal segment, it is understood that the signal mean of the signal segment is the average of the signal
Figure BDA0003358921440000102
The expression of the convolution kernel corresponding to the signal segment is:
Figure BDA0003358921440000103
based on fig. 1, the embodiment of the present invention further provides a specific implementation manner of step S104.
Specifically, for any target signal segment, the following formula is adopted to perform convolution operation according to the convolution kernel of the target signal segment.
Figure BDA0003358921440000111
In the above formula, f is the convolution kernel of the target signal segment, g is the target signal segment, the starting time of the target signal segment is τ, and the ending time is τIs marked by τ + ω, g (t) ═ xtThe signal value of the target signal segment at the time t is shown, and lambda is the characteristic analysis value of the target signal segment.
In order to more intuitively show the process of performing convolution operation on the target signal segment according to the convolution kernel of the target signal segment, in the embodiment of the present invention, the signal segment signal g in fig. 2 is used as the signal segment signal g1(t)＝xtPlease refer to fig. 4, wherein t is 1, 2, 3, and 4, which are exemplary target signal segments, and fig. 4 is a schematic diagram of a convolution operation according to an embodiment of the present invention.
Firstly, g is mixed1(1) Alignment f1(0)，g1(2) Alignment f1(1)，g1(3) Alignment f1(2)，g1(4) Alignment f1(3) (ii) a Then, g is calculated in turn1(1)×f1(0)、g1(2)×f1(1)、g1(3)×f1(2) And g1(4)×f1(3) (ii) a Finally, g is mixed1(1)×f1(0)、g1(2)×f1(1)、g1(3)×f1(2) And g1(4)×f1(3) The result of (a) is summed, i.e. the result of the convolution operation is g1(1)×f1(0)+g1(2)×f1(1)+g1(3)×f1(2)+g1(4)×f1(3) And taking the convolution operation result as a characteristic analysis value of the target signal segment.
Referring to fig. 5, fig. 5 is a flowchart illustrating a process of determining an abnormal signal in a signal to be analyzed based on a feature analysis value of a signal segment according to an embodiment of the present invention, where step S105 includes sub-step S105-1 and sub-step S105-2.
And a substep S105-1 of generating a characteristic curve of the signal to be analyzed according to the characteristic analysis values of all the signal segments of the signal to be analyzed.
In the embodiment of the present invention, the feature analysis values of all the signal segments are arranged in the extraction order of the signal segments, and a feature curve of the signal to be analyzed is generated using statistical software such as EXCEL.
And a substep S105-2, analyzing the characteristic curve of the signal to be analyzed, and determining an abnormal signal in the signal to be analyzed.
In the embodiment of the present invention, the size of the feature analysis value reflects the number of abnormal signals contained in the corresponding signal segment, and the larger the feature analysis value is, the more abnormal signals contained in the representative signal segment are, and the smaller the feature analysis value is, the fewer abnormal signals contained in the representative signal segment are.
As a specific implementation manner, the step of determining an abnormal signal in the signal to be analyzed is as follows:
firstly, taking a signal segment corresponding to a characteristic analysis value which is larger than a preset threshold value in a characteristic curve of a signal to be analyzed as an abnormal signal segment;
secondly, taking the initial time of the abnormal signal segment as the abnormal time;
and finally, taking the signal value at the abnormal moment as an abnormal signal value.
In the embodiment of the present invention, the preset threshold is an upper limit value of the characteristic analysis value, considering that many factors affecting generation of abnormal signals in industrial production and differences exist in sizes of the abnormal signals caused by different factors, so that any signal segment in the actually acquired signals contains more or less abnormal signals, when the characteristic analysis value is smaller than the preset value, the abnormal signals contained in the corresponding signal segment are weak, within an allowable range, when the characteristic analysis value is larger than the preset value, the abnormal signals contained in the corresponding signal segment exceed the allowable range, the starting time of the signal segment is taken as an abnormal time, and the signal value at the time is taken as an abnormal signal value.
In order to better show the application effect of the method, the embodiment of the invention applies the method to the welding current data in the actual industrial production, wherein the current data without abnormal conditions in the welding process is the normal welding current data, and the current data with abnormal conditions in the welding process is the abnormal welding current data. In the embodiment of the invention, 89 welding current data are selected, wherein 66 welding current data correspond to a normal welding result, 23 welding current data correspond to an abnormal welding result, the size of a preset analysis window is set to be 60, a preset step length is set to be 1, and a preset threshold value is set to be 60.
Fig. 6A is a diagram showing a normal welding result provided by the embodiment of the present invention, and as can be seen from fig. 6A, no region of the welded material is melted through. As shown in fig. 6B, it can be seen from fig. 6B that the welding current data corresponding to the normal welding result shows regular periodic fluctuation, and the characteristic curve extracted by the signal time domain characteristic analysis method provided in the embodiment of the present invention can be approximately regarded as a horizontal straight line and is much smaller than the preset threshold.
Fig. 6C is a diagram illustrating an abnormal welding result according to an embodiment of the present invention, and it can be seen from fig. 6C that a plurality of fusion-penetrated areas with different degrees appear on the welded material. As shown in fig. 6D, it can be seen from fig. 6D that the characteristic curve extracted by the signal time domain characteristic analysis method provided in the embodiment of the present invention includes multiple sections of steady change regions and multiple peaks with different sizes, and the values of some of the peaks exceed a preset threshold, where the data segment corresponding to the characteristic analysis value of the steady change region is a normal data segment, the data segment corresponding to the characteristic analysis value of the peak region is an abnormal data segment, and the characteristic analysis values corresponding to the normal data segment and the abnormal data segment have clear boundaries on the characteristic curve.
As for the welding current data corresponding to the abnormal welding result shown in fig. 6C, the signal time domain feature analysis method provided by the embodiment of the present invention and the existing method for extracting the signal change feature through the variance are applied, and as for the result, please refer to fig. 6E, it can be seen from fig. 6E that the variance curve extracted by the existing method is only the process of amplifying and slowing down the change of the current value of the welding current data, and the variance values corresponding to the normal data segment and the abnormal data segment have no clear boundary on the variance curve, and the abnormal data segment in the welding current data cannot be accurately located.
The signal time domain characteristic analysis method provided by the embodiment of the invention is respectively applied to the 89 selected welding current data to extract a characteristic curve, and the welding results are detected and distinguished, wherein the distinguishing principle is that for any characteristic curve of the welding current data, if all characteristic analysis values on the characteristic curve are smaller than a preset threshold value, the welding current data correspond to a normal welding result, and if the characteristic analysis values on the characteristic curve are larger than the preset threshold value, the welding current data correspond to an abnormal welding result. The comparison of the test case with the actual case is shown in the following table:
Figure BDA0003358921440000131
among the 66 welding current data corresponding to the normal welding result, 64 detection conditions are in accordance with the actual conditions, the accuracy rate is 96.97%, the detection conditions of the 23 welding current data corresponding to the abnormal welding result are all in accordance with the actual conditions, the accuracy rate is 100%, namely, among the 89 welding current data, 87 detection conditions are in accordance with the actual conditions, and the accuracy rate is 97.75%.
In order to perform the corresponding steps in the above embodiments and various possible implementations, an implementation of the signal time domain feature analysis apparatus 100 is given below. Referring to fig. 7, fig. 7 is a block diagram illustrating a signal time domain feature analysis apparatus 100 according to an embodiment of the invention. It should be noted that the signal time domain feature analysis device 100 provided by the embodiment of the present invention has the same basic principle and technical effect as those of the above embodiments, and for the sake of brief description, no reference is made to the embodiment of the present invention.
The signal time domain feature analysis device 100 comprises an acquisition module 101, a segmentation module 102, a determination module 103, an operation module 104 and an analysis module 105.
An obtaining module 101, configured to obtain a signal to be analyzed with one-dimensional time domain dispersion.
The segmentation module 102 is configured to divide a signal to be analyzed to obtain a plurality of signal segments.
As a specific implementation manner, the segmenting module 102 is specifically configured to slide a preset analysis window according to a preset step length from a start time of a signal to be analyzed, and sequentially intercept signal segments from the signal to be analyzed to obtain a plurality of signal segments, where sliding the preset analysis window each time obtains one signal segment.
A determining module 103, configured to determine a convolution kernel for each signal segment according to the signal value of each signal segment.
As a specific embodiment, the determining module 103 is specifically configured to calculate a mean value of signal values at all time instants in each signal segment, so as to obtain a signal mean value of each signal segment; obtaining the signal deviation degree of each moment according to the signal mean value of each signal segment and the signal value of each moment; and constructing a convolution kernel of each signal segment according to the signal deviation degrees of all the moments of each signal segment.
And the operation module 104 is configured to perform convolution operation on each signal segment according to the convolution kernel of each signal segment to obtain a feature analysis value of each signal segment.
As a specific implementation, the operation module 104 is specifically configured to apply a formula to any target signal segment according to a convolution kernel of the target signal segment
Figure BDA0003358921440000141
Performing convolution operation to obtain a characteristic analysis value of the target signal segment, wherein f is a convolution kernel of the target signal segment, g is the target signal segment, the starting time of the target signal segment is tau, the ending time of the target signal segment is tau + omega, and g (t) ═ xtThe signal value of a target signal segment at the time t is shown, and lambda is a characteristic analysis value of the target signal segment.
The analysis module 105 is configured to analyze the signal to be analyzed according to the feature analysis values of all signal segments of the signal to be analyzed, and determine an abnormal signal in the signal to be analyzed.
As a specific embodiment, the analysis module 105 is specifically configured to generate a characteristic curve of the signal to be analyzed according to the characteristic analysis values of all signal segments of the signal to be analyzed; and analyzing the characteristic curve of the signal to be analyzed, and determining an abnormal signal in the signal to be analyzed.
As a specific embodiment, the analysis module 105, when configured to analyze the characteristic curve of the signal to be analyzed and determine an abnormal signal in the signal to be analyzed, is specifically configured to: taking a signal segment corresponding to a characteristic analysis value which is greater than a preset value in a characteristic curve of a signal to be analyzed as an abnormal signal segment; taking the starting time of the abnormal signal segment as the abnormal time; and taking the signal value at the abnormal time as an abnormal signal value.
Further, referring to fig. 8, fig. 8 is a schematic block diagram illustrating a structure of an electronic device 200 according to an embodiment of the present invention, where the electronic device 200 may include a memory 210 and a processor 220.
The processor 220 may be a general-purpose Central Processing Unit (CPU), a microprocessor, an Application-Specific Integrated Circuit (ASIC), or one or more Integrated circuits for controlling the execution of the signal temporal feature analysis method provided by the method embodiments described below.
The MEMory 210 may be, but is not limited to, a ROM or other type of static storage device that can store static information and instructions, a RAM or other type of dynamic storage device that can store information and instructions, an Electrically Erasable programmable Read-Only MEMory (EEPROM), a compact disc Read-Only MEMory (CD-ROM) or other optical disk storage, optical disk storage (including compact disc, laser disc, optical disc, digital versatile disc, blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. The memory 210 may be self-contained and coupled to the processor 220 via a communication bus. Memory 210 may also be integrated with processor 220. Memory 210 is used to store, among other things, machine-executable instructions for performing aspects of the present application. Processor 220 is operative to execute machine executable instructions stored in memory 210 to implement the foregoing method embodiments.
Since the electronic device 200 provided in the embodiment of the present invention is another implementation form of the signal time domain feature analysis method provided in the foregoing method embodiment, reference may be made to the above method embodiment for technical effects that can be obtained by the electronic device 200, and details are not described here again.
Embodiments of the present invention also provide a readable storage medium containing computer-executable instructions, which when executed, can be used to perform relevant operations in the signal time domain feature analysis method provided in the foregoing method embodiments.
To sum up, in the signal time domain feature analysis method, apparatus, electronic device and storage medium provided in the embodiments of the present invention, a signal to be analyzed with a discrete one-dimensional time domain is obtained, the signal to be analyzed is divided into a plurality of signal segments, a convolution kernel of each signal segment is determined according to a signal value of each signal segment, a convolution operation is performed on each signal segment according to the convolution kernel of each signal segment to obtain a feature analysis value of each signal segment, the signal to be analyzed is analyzed according to the feature analysis values of all signal segments of the signal to be analyzed to determine an abnormal signal in the signal to be analyzed, compared with the prior art, a convolution kernel adapted to a signal change feature of each signal segment is constructed for each signal segment, the convolution kernel of each signal segment is used to perform a convolution operation, and obtaining the characteristic analysis value of each signal segment, further enabling the extracted signal change characteristics of the signal to be analyzed to be more comprehensive and accurate, and more accurately determining the abnormal signal in the signal to be analyzed according to the signal change characteristics.
The above description is only for the specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.

Claims (10)

1. A method for time domain signature analysis of a signal, the method comprising the steps of:
acquiring a signal to be analyzed with one-dimensional time domain dispersion;
dividing the signal to be analyzed to obtain a plurality of signal segments;
determining a convolution kernel for each of the signal segments based on the signal value for each of the signal segments;
performing convolution operation on each signal segment according to the convolution core of each signal segment to obtain a characteristic analysis value of each signal segment;
and analyzing the signal to be analyzed according to the characteristic analysis values of all the signal segments of the signal to be analyzed, and determining abnormal signals in the signal to be analyzed.
2. The method of claim 1, wherein each of the signal segments comprises a plurality of time instants, one signal value for each of the time instants, and wherein the step of determining the convolution kernel for each of the signal segments based on the signal value for each of the signal segments comprises:
calculating the mean value of the signal values of all the moments in each signal segment to obtain the signal mean value of each signal segment;
obtaining the signal deviation degree of each moment according to the signal mean value of each signal segment and the signal value of each moment;
and constructing a convolution kernel of each signal segment according to the signal deviation degrees of all the moments of each signal segment.
3. The method of claim 2, wherein the convolution kernel for any target signal segment is expressed as:
Figure FDA0003358921430000021
wherein the starting time of the target signal segment is tau,the end time is tau + omega, xtIs the signal value of the target signal segment at time t, mu is the signal mean value of the target signal segment, (x)t-μ)2And a is the signal deviation degree of the target signal segment at the time t, a is a preset constant, and f is the convolution kernel of the target signal segment.
4. The method of claim 1, wherein the step of dividing the signal to be analyzed into a plurality of signal segments comprises:
and sliding a preset analysis window from the starting time of the signal to be analyzed according to a preset step length, and sequentially intercepting signal segments from the signal to be analyzed to obtain a plurality of signal segments, wherein the preset analysis window is slid each time to obtain one signal segment.
5. The method of claim 1, wherein the step of performing a convolution operation on each of the signal segments according to the convolution kernel of each of the signal segments to obtain a feature analysis value of each of the signal segments comprises:
for any target signal segment, adopting a formula according to the convolution kernel of the target signal segment
Figure FDA0003358921430000022
Performing convolution operation to obtain a feature analysis value of the target signal segment, wherein f is a convolution kernel of the target signal segment, g is the target signal segment, the starting time of the target signal segment is τ, the ending time of the target signal segment is τ + ω, and g (t) ═ xtAnd lambda is the signal value of the target signal segment at the time t, and is the characteristic analysis value of the target signal segment.
6. The method according to claim 1, wherein the step of analyzing the signal to be analyzed according to the characteristic analysis values of all signal segments of the signal to be analyzed to determine an abnormal signal in the signal to be analyzed comprises:
generating a characteristic curve of the signal to be analyzed according to the characteristic analysis values of all the signal segments of the signal to be analyzed;
and analyzing the characteristic curve of the signal to be analyzed, and determining an abnormal signal in the signal to be analyzed.
7. The method according to claim 6, wherein the abnormal signal comprises an abnormal time and an abnormal signal value, and the step of analyzing the characteristic curve of the signal to be analyzed to determine the abnormal signal in the signal to be analyzed comprises:
taking a signal segment corresponding to a characteristic analysis value which is larger than a preset threshold value in a characteristic curve of the signal to be analyzed as an abnormal signal segment;
taking the starting time of the abnormal signal segment as an abnormal time;
and taking the signal value at the abnormal moment as the abnormal signal value.
8. An apparatus for time domain signature analysis of a signal, the apparatus comprising:
the acquisition module is used for acquiring a signal to be analyzed with one-dimensional time domain discrete;
the segmentation module is used for dividing the signal to be analyzed to obtain a plurality of signal segments;
a determining module, configured to determine a convolution kernel of each signal segment according to a signal value of each signal segment;
the operation module is used for carrying out convolution operation on each signal segment according to the convolution core of each signal segment to obtain a characteristic analysis value of each signal segment;
and the analysis module is used for analyzing the signal to be analyzed according to the characteristic analysis values of all the signal segments of the signal to be analyzed and determining an abnormal signal in the signal to be analyzed.
9. An electronic device, comprising: a memory for storing a computer program and a processor; the processor is adapted to perform the method of any of claims 1-7 when the computer program is invoked.
10. A readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the method according to any one of claims 1-7.
CN202111360196.1A 2021-11-17 2021-11-17 Signal time domain feature analysis method and device, electronic equipment and storage medium Pending CN113988140A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111360196.1A CN113988140A (en) 2021-11-17 2021-11-17 Signal time domain feature analysis method and device, electronic equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111360196.1A CN113988140A (en) 2021-11-17 2021-11-17 Signal time domain feature analysis method and device, electronic equipment and storage medium

Publications (1)

Publication Number Publication Date
CN113988140A true CN113988140A (en) 2022-01-28

Family

ID=79748942

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111360196.1A Pending CN113988140A (en) 2021-11-17 2021-11-17 Signal time domain feature analysis method and device, electronic equipment and storage medium

Country Status (1)

Country Link
CN (1) CN113988140A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116548928A (en) * 2023-07-11 2023-08-08 西安浩阳志德医疗科技有限公司 Nursing service system based on internet

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116548928A (en) * 2023-07-11 2023-08-08 西安浩阳志德医疗科技有限公司 Nursing service system based on internet
CN116548928B (en) * 2023-07-11 2023-09-08 西安浩阳志德医疗科技有限公司 Nursing service system based on internet

Similar Documents

Publication Publication Date Title
US10956451B2 (en) Time-series data processing method, recording medium having recorded thereon time-series data processing program, and time-series data processing device
CN109213654B (en) Anomaly detection method and device
CN107025174B (en) Method, device and readable storage medium for user interface anomaly test of equipment
JP6409983B2 (en) Chromatograph mass spectrometry data processing method and processing apparatus
CN106611023B (en) Method and device for detecting website access abnormality
CN114166988B (en) Chromatograph cluster fault detection method and device
CN113988140A (en) Signal time domain feature analysis method and device, electronic equipment and storage medium
US7949497B2 (en) Machine condition monitoring using discontinuity detection
CN111912521A (en) Frequency detection method of non-stationary signal and storage medium
EP4026134A1 (en) Techniques for exception-based validation of analytical information
JP6874864B2 (en) Image processing equipment, image processing methods and programs
CN108604142B (en) Touch screen device operation method and touch screen device
CN112268696A (en) Clutch detection method, device, storage medium and device
CN109976576B (en) Method and device for preventing mistaken touch of mobile terminal screen
CN113608956B (en) Method, system and related device for testing hard disk power consumption
JP6747161B2 (en) Software quality judgment method
US20130191071A1 (en) System and method for automatic modal parameter extraction in structural dynamics analysis
CN112543897B (en) Analysis device, analysis method, and storage medium
WO2018037487A1 (en) Data processing method and device
JP2016066227A (en) Incompatibility detecting device, incompatibility detecting method and program
JP2006105714A (en) Method for determining cause of accident in power transmission line and distribution line
US20230259447A1 (en) Defect Analysis Apparatus, Program, and Defect Analysis Method
US20230243789A1 (en) Analysis device and analysis method
US20230280318A1 (en) Learning data producing method, waveform analysis device, waveform analysis method, and recording medium
CN113407455B (en) Automatic evaluation method and device for intelligent white electric products

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination