An adaptive micro-milling chatter detection method

Abstract

The application relates to a self-adaptive micro-milling chatter detection method, which belongs to the technical field of self-adaptive micro-milling chatter detection methods. Firstly, acceleration signals at workpiece and machine positions are acquired simultaneously during machining; then, the workpiece acceleration signal and the machine acceleration signal are respectively taken as an expected signal and a reference input signal and are sent into a variable-forgetting-factor recursive least square adaptive filter, so that chatter irrelevant components contained in the workpiece acceleration signal are filtered out; subsequently, a variable mode extraction algorithm is used to process the filtered signal, the initial value of the central frequency is adaptively determined according to the power spectral density corresponding to the filtered signal, and only one eigenmode function is obtained after the variable mode extraction; then, the amplitude of the filtered signal and the amplitude of the marginal spectrum corresponding to the eigenmode function are calculated as chatter characteristics, and the calculated characteristics are compared with a threshold set in advance, so that the occurrence of chatter can be accurately detected.

Description

Technical Field

[0001] This invention belongs to the field of micro-milling technology, and relates to a micro-milling chatter detection method, particularly an adaptive micro-milling chatter detection method. Background Technology

[0002] Existing chatter detection methods are mostly based on variational mode decomposition (VMD), which requires pre-determining the number of mode decomposition layers, often based on empirical rules. In addition, optimization algorithms such as genetic algorithms and particle swarm optimization (PSO) are used to determine the number of decomposition layers, but these methods are often time-consuming. Furthermore, VMD often yields a series of intrinsic mode functions (EMFs), but only specific EMFs are typically used to characterize chatter components, increasing computational costs unnecessarily. To date, there is no clear standard for selecting specific EMFs to characterize chatter components. Therefore, developing an adaptive chatter detection method suitable for micro-milling processes has significant practical application value.

[0003] Existing technology discloses a conventional milling chatter detection method based on variational mode decomposition and energy entropy. This method first uses variational mode decomposition to decompose the acquired force signal into a series of intrinsic mode functions (EMFs), and then calculates the energy entropy corresponding to each EMF. However, only the EMF with the largest energy entropy is used to characterize the chatter component. However, an inappropriate number of mode decomposition layers can lead to the selection of inappropriate EMFs. Furthermore, the calculation of irrelevant EMFs increases the computational cost of the variational mode decomposition method.

[0004] Existing technology also discloses a conventional milling chatter detection method based on variational mode decomposition and power spectral entropy. This method first uses variational mode decomposition to decompose the obtained acceleration signal into a series of intrinsic mode functions (IMFs), but only IMFs whose corresponding frequencies are near the system's natural frequencies are selected as chatter components. However, for micro-milling processes, the natural frequencies are often difficult to obtain. Furthermore, compared to conventional milling processes, micro-milling corresponds to a wider frequency band, making it difficult to determine the appropriate number of mode decomposition layers.

[0005] A key characteristic of the aforementioned literature is that the number of mode decomposition layers and the system's natural frequencies need to be predetermined when using variational mode decomposition methods. However, it is difficult to predetermine an appropriate number of mode decomposition layers for different machining processes. Furthermore, the system's natural frequencies are difficult to obtain for micro-milling processes. Therefore, there is a need to develop an adaptive chatter detection method suitable for micro-milling processes. Summary of the Invention

[0006] The technical problem to be solved by this invention is:

[0007] To address the shortcomings of variational mode decomposition-based chatter detection methods, which require prior determination of the number of mode decomposition layers, this invention provides an adaptive chatter detection method for micro-milling processes.

[0008] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0009] An adaptive micro-milling chatter detection method, characterized by comprising:

[0010] The experiment was conducted using the set cutting parameters, and the workpiece acceleration signal and the machine tool acceleration signal were collected simultaneously during the experiment.

[0011] The workpiece acceleration signal and the machine tool acceleration signal are fed into a recursive least squares adaptive filter with a variable forgetting factor to obtain the filtered signal.

[0012] Variational mode extraction is performed on the filtered signal to obtain the flutter component contained therein;

[0013] The Hilbert-Huang transform of the obtained flutter components yields their corresponding marginal spectra;

[0014] Calculate the amplitude of the filtered signal and the amplitude of its marginal spectrum;

[0015] The amplitude of the filtered signal and the amplitude of the marginal spectrum are compared with their respective preset thresholds. If one of the two values ​​is higher than the preset threshold, it indicates that flutter has occurred during the processing; if both are lower than the threshold, it indicates that the processing is stable.

[0016] A further technical solution of the present invention: the step of sending the workpiece acceleration signal and the machine tool acceleration signal into a recursive least squares adaptive filter with a variable forgetting factor to obtain the filtered signal is specifically as follows:

[0017] S f (t)=S w (t)-h T (t-1)S x (t)

[0018]

[0019] h(t)=h(t-1)+k(t)S f (t)

[0020]

[0021] In the formula, S w (t) is the workpiece acceleration signal, S m (t) is the machine tool acceleration signal, Sf (t) is the filtered signal, h T (t-1) is the weight vector. λ(t) is the input signal vector, L is the filter order, k(t) is the Kalman gain vector, Q(t) is the intermediate variable, and λ(t) is the forgetting factor that varies with time.

[0022] A further technical solution of the present invention: The step of performing variational mode extraction on the filtered signal to obtain the dithering component contained therein; specifically:

[0023]

[0024] In the formula, S′ c (t) represents the flutter component, and α is the penalty factor. The asterisk ('*') represents the partial derivative and convolution symbol, δ(t) is the Dirac function, and S′ c (t) and S r (t) represent the dithering component and residual signal contained in the filtered signal, respectively, ω d is the center frequency of the intrinsic mode function to be determined, j is the imaginary unit, e is the natural constant, β(t) is the impulse response function of the filter used, and γ(t) represents the Lagrange multiplier.

[0025] A further technical solution of the present invention: the Hilbert-Huang transform is performed on the obtained flutter component to obtain its corresponding marginal spectrum; specifically:

[0026] H(f)=∫H(t,f)dt

[0027] in:

[0028]

[0029]

[0030] S a (t)=S c ′(t)+jS h (t)=A(t)e jθ(t)

[0031]

[0032]

[0033]

[0034] In the formula, H(f) is the marginal spectrum, f represents the frequency index, H(t,f) is the intermediate variable, and S h (t) represents the flutter component S′ c Hilbert transform of (t), Sa (t) represents the flutter component S′ c The analysis signal corresponding to (t) is A(t), where A(t) represents the instantaneous amplitude, θ(t) is the instantaneous phase, and f(t) represents the instantaneous frequency.

[0035] A further technical solution of the present invention: the calculation of the amplitude of the filtered signal and the amplitude of the marginal spectrum; specifically:

[0036] F e1 =max(|S f (t)|)

[0037] F e2 =max(H(f))

[0038] A computer system is characterized by comprising: one or more processors, and a computer-readable storage medium for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method described above.

[0039] A computer-readable storage medium is characterized by storing computer-executable instructions, which, when executed, are used to implement the above-described method.

[0040] The beneficial effects of this invention are as follows:

[0041] This invention provides an adaptive chatter detection method for micro-milling processes. It simultaneously uses two accelerometers to obtain acceleration signals at the workpiece and machine tool positions during machining. The machine tool acceleration signal is used as the reference input for an adaptive filter, effectively filtering out chatter-irrelevant components in the workpiece acceleration signal. Therefore, the chatter component dominates the filtered signal. Then, a variational mode extraction algorithm is applied to the filtered signal, with the initial center frequency adaptively determined based on the power spectral density of the filtered signal, effectively extracting the required intrinsic mode functions (IMFs), i.e., the chatter components. Next, a Hilbert-Huang transform is performed on the obtained IMFs to obtain their corresponding marginal spectra. The amplitude of the filtered signal and the amplitude of the corresponding marginal spectra are calculated as chatter features. The calculated features are compared with a pre-set threshold to achieve chatter detection in the micro-milling process. Compared with Reference 1, the present invention obtains only one intrinsic mode function, so there is no need to pre-determine the number of mode decomposition layers; compared with Reference 2, when performing flutter detection, the flutter-irrelevant components of the present invention are effectively filtered out, and the required center frequency can be adaptively determined based on the filtered signal, so there is no need to pre-determine the system's natural frequency. Attached Figure Description

[0042] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts.

[0043] Figure 1 This is the workpiece acceleration signal obtained when the spindle speed is 24,000 rpm and the axial cutting depth is 20 micrometers in this embodiment of the invention.

[0044] Figure 2 The machine tool acceleration signal is obtained when the spindle speed is 24,000 rpm and the axial cutting depth is 20 micrometers in this embodiment of the invention.

[0045] Figure 3 This is the filtered signal obtained by using a recursive least squares adaptive filter with a variable forgetting factor when the spindle speed is 24,000 rpm and the axial cutting depth is 20 micrometers in this embodiment of the invention.

[0046] Figure 4 The intrinsic mode function is obtained by using the variational mode extraction algorithm when the spindle speed is 24,000 rpm and the axial cutting depth is 20 micrometers in this embodiment of the invention.

[0047] Figure 5 The marginal spectrum is obtained by performing a Hilbert-Huang transform on the extracted intrinsic mode functions when the spindle speed is 24,000 rpm and the axial cutting depth is 20 micrometers in this embodiment of the invention.

[0048] Figure 6 This is the workpiece acceleration signal obtained when the spindle speed is 24,000 rpm and the axial cutting depth is 180 micrometers in an embodiment of the present invention.

[0049] Figure 7 The machine tool acceleration signal is obtained when the spindle speed is 24,000 rpm and the axial cutting depth is 180 micrometers in this embodiment of the invention.

[0050] Figure 8 The filtered signal is obtained by using a recursive least squares adaptive filter with a variable forgetting factor when the spindle speed is 24,000 rpm and the axial cutting depth is 180 micrometers in this embodiment of the invention.

[0051] Figure 9 The intrinsic mode function is obtained by using the variational mode extraction algorithm when the spindle speed is 24,000 rpm and the axial cutting depth is 180 micrometers in this embodiment of the invention.

[0052] Figure 10 The marginal spectrum is obtained by performing a Hilbert-Huang transform on the extracted intrinsic mode functions when the spindle speed is 24,000 rpm and the axial cutting depth is 180 micrometers in this embodiment of the invention.

[0053] Figure 11 This is a flowchart of the method of the present invention. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0055] This invention provides an adaptive chatter detection method for micro-milling. The method first uses two accelerometers to simultaneously acquire acceleration signals at the workpiece and machine tool positions during machining. Then, the workpiece acceleration signal and the machine tool acceleration signal are respectively fed into a recursive least squares adaptive filter with a variable forgetting factor, thereby filtering out chatter-irrelevant components in the workpiece acceleration signal. Next, a variational mode extraction algorithm is used to process the filtered signal. The initial value of the center frequency is adaptively determined based on the power spectral density corresponding to the filtered signal, and only one intrinsic mode function (EMF) is obtained after variational mode extraction. Then, the amplitude of the filtered signal and the amplitude of the marginal spectrum corresponding to the EMF are calculated as chatter features, and the calculated features are compared with a pre-set threshold to accurately detect chatter occurrence. Figure 11 As shown, it includes the following steps:

[0056] Step 1: Conduct the experiment using the set cutting parameters, and simultaneously collect acceleration signals at the workpiece and machine tool positions during the experiment. The workpiece acceleration signal is represented as S. w (t), the machine tool acceleration signal is represented as S m (t), where t represents the time index.

[0057] Step 2: Transfer the workpiece acceleration signal S w (t) and machine tool acceleration signal S m (t) is fed into a recursive least squares adaptive filter with a variable forgetting factor to obtain the filtered signal S. f The specific calculation process for (t) can be represented as follows:

[0058] S f (t)=S w (t)-h T (t-1)S x (t)

[0059]

[0060] h(t)=h(t-1)+k(t)S f(t)

[0061]

[0062] In the formula, h T (t-1) is the weight vector. λ(t) is the input signal vector, L is the filter order, k(t) is the Kalman gain vector, Q(t) is the intermediate variable, and λ(t) is the forgetting factor that varies with time.

[0063] Step 3: Filter the signal S f (t) Perform variational mode extraction to obtain the flutter component S′ contained therein. c (t):

[0064]

[0065] In the formula, α is the penalty factor. The asterisk ('*') represents the partial derivative and convolution symbol, δ(t) is the Dirac function, and S′ c (t) and S r (t) represent the dithering component and residual signal contained in the filtered signal, respectively, ω d is the center frequency of the intrinsic mode function to be determined, j is the imaginary unit, e is the natural constant, β(t) is the impulse response function of the filter used, and γ(t) represents the Lagrange multiplier.

[0066] Step 4: Process the obtained flutter component S′ c (t) is subjected to Hilbert-Huang transform to obtain its corresponding marginal spectrum H(f):

[0067] H(f)=∫H(t,f)dt

[0068] in:

[0069]

[0070]

[0071] S a (t)=S′ c (t)+jS h (t)=A(t)e jθ(t)

[0072]

[0073]

[0074]

[0075] In the formula, f represents the frequency index, H(t,f) is an intermediate variable, and S...h (t) represents the flutter component S′ c Hilbert transform of (t), S a (t) represents the flutter component S′ c The analysis signal corresponding to (t) is A(t), where A(t) represents the instantaneous amplitude, θ(t) is the instantaneous phase, and f(t) represents the instantaneous frequency.

[0076] Step 5: Calculate the amplitude F of the filtered signal using the following formula. e1 and the amplitude F of the marginal spectrum e2 As a flutter characteristic:

[0077] F e1 =max(|S f (t)|)

[0078] F e2 =max(H(f))

[0079] Step 6: Compare the result calculated in Step 5 with the set threshold. If one of the two values ​​is higher than the set threshold, it indicates that chatter has occurred during the processing; if both are lower than the threshold, it indicates that the processing is stable.

[0080] To enable those skilled in the art to better understand the present invention, the present invention will be described in detail below with reference to specific embodiments.

[0081] Example 1 corresponds to a stable micro-milling process, while Example 2 corresponds to a micro-milling process that experiences chatter.

[0082] Example 1:

[0083] The experiment used a flat-bottomed carbide end mill with N=2 teeth and R=0.5 mm radius to conduct micro-milling experiments on aluminum alloys on a GR50 cradle-type five-axis CNC machining center using climb milling. The cutting parameter used was radial depth of cut a. e =0.5 mm, feed per tooth c = 0.005 mm / tooth, spindle speed 24000 rpm, axial depth of cut a p = 20 micrometers.

[0084] Step 1, refer to the appendix Figure 1-2 An experiment was conducted using the set cutting parameters, and acceleration signals at both the workpiece and machine tool positions were simultaneously acquired during the experiment. The workpiece acceleration signal is represented as S. w (t), the machine tool acceleration signal is represented as S m (t), where t = 1, 2, ..., 256000.

[0085] Step 2: Transfer the workpiece acceleration signal S w (t) and machine tool acceleration signal Sm (t) is fed into a recursive least squares adaptive filter with a variable forgetting factor to obtain the filtered signal S. f (t), see appendix Figure 3 The specific calculation process can be represented as follows:

[0086] S f (t)=S w (t)-h T (t-1)S x (t)

[0087]

[0088] h(t)=h(t-1)+k(t)S f (t)

[0089]

[0090] In the formula, h T (t-1) is the weight vector. λ(t) is the input signal vector, L = 5, k(t) is the Kalman gain vector, Q(t) is the intermediate variable, and λ(t) is the forgetting factor that varies with time.

[0091] Step 3: Filter the signal S f (t) Perform variational mode extraction to obtain the flutter component S′ contained therein. c (t), see appendix Figure 4 :

[0092]

[0093] In the formula, α = 2000, The asterisk ('*') represents the partial derivative and convolution symbol, δ(t) is the Dirac function, and S′ c (t) and S r (t) represent the dithering component and residual signal contained in the filtered signal, respectively, ω d =2072.06Hz, j is the imaginary unit, e =2.7183, β(t) is the impulse response function of the filter used, and γ(t) represents the Lagrange multiplier.

[0094] Step 4: Process the obtained flutter component S c Performing a Hilbert-Huang transform on ′(t) yields its corresponding marginal spectrum H(f), see Appendix. Figure 5 :

[0095] H(f)=∫H(t,f)dt

[0096] in:

[0097]

[0098]

[0099] S a (t)=S′ c (t)+jS h (t)=A(t)e jθ(t)

[0100]

[0101]

[0102]

[0103] In the formula, f represents the frequency index, H(t,f) is an intermediate variable, and S... h (t) represents the flutter component S′ c Hilbert transform of (t), S a (t) represents the flutter component S′ c The analysis signal corresponding to (t) is A(t), where A(t) represents the instantaneous amplitude, θ(t) is the instantaneous phase, and f(t) represents the instantaneous frequency.

[0104] Step 5: Calculate the amplitude F of the filtered signal using the following formula. e1 and the amplitude F of the marginal spectrum e2 As a flutter characteristic:

[0105] F e1 =max(|S f (t)|)

[0106] F e2 =max(H(f))

[0107] The amplitude F of the filtered signal is calculated. e1 =0.747, the amplitude F of the marginal spectrum e2 =0.032.

[0108] Step Six: Compare the results calculated in Step Five with the pre-set thresholds. Here, the thresholds corresponding to the amplitude of the filtered signal and the amplitude of the marginal spectrum are set to 2 and 0.1, respectively. The comparison shows that the amplitude of the filtered signal and the amplitude of the marginal spectrum are both less than the corresponding thresholds. Therefore, the processing is stable.

[0109] Example 2:

[0110] The experiment used a flat-bottomed carbide end mill with N=2 teeth and R=0.5 mm radius to conduct micro-milling experiments on aluminum alloys on a GR50 cradle-type five-axis CNC machining center using climb milling. The cutting parameter used was radial depth of cut a. e=0.005 mm, feed per tooth c = 0.005 mm / tooth, spindle speed 24000 rpm, axial depth of cut a p = 20 micrometers.

[0111] Step 1, refer to the appendix Figure 6-7 An experiment was conducted using the set cutting parameters, and acceleration signals at both the workpiece and machine tool positions were simultaneously acquired during the experiment. The workpiece acceleration signal is represented as S. w (t), the machine tool acceleration signal is represented as S m (t), where t = 1, 2, ..., 256000.

[0112] Step 2: Transfer the workpiece acceleration signal S w (t) and machine tool acceleration signal S m (t) is fed into a recursive least squares adaptive filter with a variable forgetting factor to obtain the filtered signal S. f (t), see appendix Figure 8 The specific calculation process can be represented as follows:

[0113] S f (t)=S w (t)-h T (t-1)S x (t)

[0114]

[0115] h(t)=h(t-1)+k(t)S f (t)

[0116]

[0117] In the formula, h T (t-1) is the weight vector. λ(t) is the input signal vector, L = 5, k(t) is the Kalman gain vector, Q(t) is the intermediate variable, and λ(t) is the forgetting factor that varies with time.

[0118] Step 3: Filter the signal S f (t) Perform variational mode extraction to obtain the flutter component S′ contained therein. c (t), see appendix Figure 9 :

[0119]

[0120] In the formula, α = 2000, The asterisk ('*') represents the partial derivative and convolution symbol, δ(t) is the Dirac function, and S′ c (t) and S r(t) represent the dithering component and residual signal contained in the filtered signal, respectively, ω d =4309.58Hz, j is the imaginary unit, e =2.7183, β(t) is the impulse response function of the filter used, and γ(t) represents the Lagrange multiplier.

[0121] Step 4: Process the obtained flutter component S c Performing a Hilbert-Huang transform on ′(t) yields its corresponding marginal spectrum H(f), see Appendix. Figure 10 :

[0122] H(f)=∫H(t,f)dt

[0123] in:

[0124]

[0125]

[0126] S a (t)=S c ′(t)+jS h (t)=A(t)e jθ(t)

[0127]

[0128]

[0129]

[0130] In the formula, f represents the frequency index, H(t,f) is an intermediate variable, and S... h (t) represents the flutter component S′ c Hilbert transform of (t), S a (t) represents the flutter component S′ c The analysis signal corresponding to (t) is A(t), where A(t) represents the instantaneous amplitude, θ(t) is the instantaneous phase, and f(t) represents the instantaneous frequency.

[0131] Step 5: Calculate the amplitude F of the filtered signal using the following formula. e1 and the amplitude F of the marginal spectrum e2 As a flutter characteristic:

[0132] F e1 =max(|S f (t)|)

[0133] F e2 =max(H(f))

[0134] The amplitude F of the filtered signal is calculated. e1=6.064, the amplitude F of the marginal spectrum e2 =0.266.

[0135] Step Six: Compare the results calculated in Step Five with the pre-set thresholds. Here, the thresholds corresponding to the amplitude of the filtered signal and the amplitude of the marginal spectrum are set to 2 and 0.1, respectively. The comparison shows that the amplitude of the filtered signal and the amplitude of the marginal spectrum are both greater than the corresponding thresholds. Therefore, flutter occurred during the processing.

[0136] As can be seen from the above examples, the proposed method can effectively filter out chatter-irrelevant components contained in the workpiece acceleration signal and extract chatter components from the filtered signal. The chatter characteristics calculated based on the filtered signal and marginal spectrum have significant differences under different processing states. By comparing the calculated chatter characteristics with the corresponding threshold, accurate detection of micro-milling chatter is achieved.

[0137] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the scope of the technology disclosed in the present invention, and such modifications or substitutions should all be covered within the scope of protection of the present invention.

Claims

1. An adaptive micro-milling chatter detection method, characterized in that, include: The experiment was conducted using the set cutting parameters, and the workpiece acceleration signal and the machine tool acceleration signal were collected simultaneously during the experiment. The workpiece acceleration signal and the machine tool acceleration signal are fed into a recursive least squares adaptive filter with a variable forgetting factor to obtain the filtered signal. Variational mode extraction is performed on the filtered signal to obtain the dizziness component contained therein; specifically, the variational mode extraction is performed on the filtered signal to obtain the dizziness component contained therein; In the formula, For flutter components, As a penalty factor, 'and' ' represents the partial derivative and convolution symbols, For the Dirac function, and These represent the dithering component and the residual signal contained in the filtered signal, respectively. It is the center frequency of the eigenmode function being sought. It is the imaginary unit. It is a natural constant. It is the impulse response function of the filter used. Represents the Lagrange multipliers; The Hilbert-Huang transform of the obtained flutter components yields their corresponding marginal spectra; Calculate the amplitude of the filtered signal and the amplitude of its marginal spectrum; The amplitude of the filtered signal and the amplitude of the marginal spectrum are compared with their respective preset thresholds. If one of the two values ​​is higher than the preset threshold, it indicates that flutter has occurred during the processing; if both are lower than the threshold, it indicates that the processing is stable.

2. The adaptive micro-milling chatter detection method according to claim 1, characterized in that, The process of feeding the workpiece acceleration signal and the machine tool acceleration signal into a recursive least squares adaptive filter with a variable forgetting factor to obtain the filtered signal is as follows: In the formula, It is the workpiece acceleration signal. It is a machine tool acceleration signal. It is the filtered signal. It is a weight vector. It is the input signal vector. It is the filter order. It is the Kalman gain vector. It is an intermediate variable. It is a forgetting factor that changes over time.

3. The adaptive micro-milling chatter detection method according to claim 1, characterized in that, The Hilbert-Huang transform is performed on the obtained flutter components to obtain their corresponding marginal spectra; specifically: in: In the formula, For marginal spectrum, Indicates frequency index, As an intermediate variable, Flutter component Hilbert transform, Flutter component The corresponding analysis signal, Indicates instantaneous amplitude. For instantaneous phase, Indicates instantaneous frequency.

4. The adaptive micro-milling chatter detection method according to claim 3, characterized in that, The calculation of the amplitude of the filtered signal and the amplitude of its marginal spectrum; specifically: 。 5. A computer system, characterized in that... include: One or more processors, a computer-readable storage medium for storing one or more programs, wherein, when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to perform the method of any one of claims 1-4.

6. A computer-readable storage medium, characterized in that... The device stores computer-executable instructions, which, when executed, are used to implement the method described in any one of claims 1-4.

Images