A GIL equipment fault detection method, system and device based on a voiceprint dynamic coupling model and a storage medium

Abstract

The application discloses a GIL equipment fault detection method and system based on a voiceprint dynamic coupling model, equipment and a storage medium, relates to the technical field of power system equipment monitoring and operation and maintenance, and comprises the following steps: collecting sound signals inside the GIL equipment in real time, extracting characteristic parameters, and obtaining a characteristic vector; a Gaussian mixture model is established using normal working condition data, the Gaussian mixture model is trained based on the characteristic vector, and an anomaly detection model is obtained; characteristic parameters of real-time signals to be detected are extracted, and the estimated value and residual of the extracted characteristics on the anomaly detection model are calculated; based on the calculated estimated value and residual, an anomaly score is calculated, and corresponding early warning information and fault positioning results are output; the application optimizes the whole process from data acquisition to fault diagnosis, effectively improves the accuracy, timeliness and intelligent level of GIL equipment voiceprint monitoring, and provides a powerful guarantee for safe and stable operation of equipment.

Description

Technical Field

[0001] This invention relates to the field of power system equipment monitoring and maintenance technology, and in particular to a method, system, device and storage medium for GIL equipment fault detection based on a dynamic coupling model of acoustic signatures. Background Technology

[0002] Gas-insulated metal-enclosed transmission lines (GILs) involve numerous manufacturing and installation processes, requiring extremely precise maintenance procedures. Even slight negligence can lead to related quality problems. Partial discharge is a common GIL fault. Although its incidence is relatively low, once it occurs, the consequences are quite severe, causing serious damage to the equipment itself and surrounding equipment, and ultimately affecting the overall operational safety of the power grid.

[0003] Existing GIL condition monitoring devices mainly employ pressure monitoring and UHF partial discharge signal monitoring, which are limited in function and have few installation locations, failing to provide comprehensive real-time monitoring of the GIL system's operational status. When gas-insulated transmission lines are located in vertical shafts with significant elevation differences, monitoring the equipment's daily operation is inconvenient, and after a fault occurs, the fault location can only be determined by individually detecting changes in the gas composition of each chamber, which is time-consuming and labor-intensive. Furthermore, there is a lack of real-time monitoring measures for potential temperature rises, partial discharge noises, and changes in SF6 meter readings. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by this invention is that existing GIL system monitoring devices have limited functions and few installation locations, making it impossible to monitor the operating status in all aspects. The inspection workload is large and there are safety risks. Furthermore, there is a lack of real-time detection, unified analysis, modeling, and early warning methods for various physical indicators.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a GIL (Gas Injection Line) device fault detection method based on a dynamic coupling model of acoustic signatures, comprising:

[0008] Real-time acquisition of sound signals from inside the GIL device, extraction of feature parameters, and generation of feature vectors;

[0009] A Gaussian mixture model is built using normal operating condition data, and the Gaussian mixture model is trained based on feature vectors to obtain an anomaly detection model.

[0010] Feature parameters are extracted from the real-time signal to be detected, and the estimated values ​​and residuals of the extracted features on the anomaly detection model are calculated.

[0011] Based on the calculated estimates and residuals, anomaly scores are calculated and corresponding early warning information and fault location results are output.

[0012] As a preferred embodiment of the GIL (Gas Inertial Communication) device fault detection method based on a dynamic coupling model of acoustic signatures, wherein:

[0013] The real-time acquisition of sound signals from inside the GIL device is followed by feature parameter extraction to obtain a feature vector, which includes:

[0014] The acquired sound signal is decomposed into effective signal and noise components. Wavelet basis is selected to perform wavelet decomposition on the sound signal, thereby obtaining a series of wavelet coefficients.

[0015] As a preferred embodiment of the GIL (Gas Inertial Communication) device fault detection method based on a dynamic coupling model of acoustic signatures, wherein:

[0016] The process of acquiring sound signals from inside the GIL device in real time, extracting feature parameters, and obtaining feature vectors also includes:

[0017] Different processing methods are applied based on the relationship between the absolute value of the wavelet coefficient and the preset threshold: when the absolute value of the wavelet coefficient is greater than or equal to the threshold, it is transformed by a specific function; when it is less than the threshold, it is directly set to zero; where the preset threshold is calculated based on the standard deviation of the noise and the signal length.

[0018] The beneficial effects of this preferred technical solution are as follows: This wavelet coefficient processing method based on a preset threshold can effectively remove noise interference from sound signals. Calculating the threshold based on the standard deviation of the noise and the signal length makes the threshold setting more scientific and reasonable, minimizing the impact of noise on subsequent analysis while preserving effective signal characteristics.

[0019] As a preferred embodiment of the GIL (Gas Inertial Communication) device fault detection method based on a dynamic coupling model of acoustic signatures, wherein:

[0020] The process of acquiring sound signals from inside the GIL device in real time, extracting feature parameters, and obtaining feature vectors also includes:

[0021] The transform coefficients that were determined to be noise after the processing were discarded, and then an inverse wavelet transform was performed to obtain the noise-removed signal.

[0022] The denoised audio signal is pre-emphasized, and the pre-emphasized audio signal is divided into frames with a frame duration between 20ms and 30ms, and the overlap time between different frames is 10ms.

[0023] The beneficial effects of this preferred technical solution are as follows: discarding the transform coefficients corresponding to noise and performing inverse wavelet transform yields a clean and effective signal. Pre-emphasis processing enhances the energy of the high-frequency components of the signal, highlighting the characteristics of the sound signal. Framing and setting overlap time ensures that local feature changes in the signal can be captured during signal analysis while maintaining the continuity of information between frames, thus improving the accuracy of feature extraction.

[0024] As a preferred embodiment of the GIL (Gas Inertial Communication) device fault detection method based on a dynamic coupling model of acoustic signatures, wherein:

[0025] The process of acquiring sound signals from inside the GIL device in real time, extracting feature parameters, and obtaining feature vectors also includes:

[0026] Windowing is performed on the framed audio signal, and the signal of each frame is weighted by a window function;

[0027] The windowed audio signal is converted from the time domain to the frequency domain, and the amplitude at each point in the frequency domain is calculated to obtain the amplitude spectrum of the audio signal.

[0028] As a preferred embodiment of the GIL (Gas Inertial Communication) device fault detection method based on a dynamic coupling model of acoustic signatures, wherein:

[0029] The process of acquiring sound signals from inside the GIL device in real time, extracting feature parameters, and obtaining feature vectors also includes:

[0030] The sound amplitude spectrum is transformed using a triangular bandpass filter bank. The amplitude spectrum is then multiplied by the triangular filter bank to obtain a smoothed sound amplitude spectrum. The signal after spectral smoothing is then subjected to discrete cosine transform to extract the eigenvectors of the Mel frequency cepstral coefficients.

[0031] As a preferred embodiment of the GIL (Gas Inertial Communication) device fault detection method based on a dynamic coupling model of acoustic signatures, wherein:

[0032] The method of establishing a Gaussian mixture model using normal operating condition data, and training the Gaussian mixture model based on feature vectors to obtain an anomaly detection model includes:

[0033] The feature vectors extracted from relevant data under normal equipment operation are used as input data for the Gaussian mixture model. The parameters of the Gaussian mixture model are calculated and optimized using the expectation-maximization algorithm to complete the model training and obtain the anomaly detection model.

[0034] The beneficial effects of this preferred technical solution are as follows: Training a Gaussian mixture model using feature vectors from the equipment under normal operating conditions can accurately characterize the feature distribution during normal operation. The expectation-maximization algorithm can effectively calculate and optimize model parameters, enabling the model to better fit normal data, thereby improving the accuracy and reliability of anomaly detection.

[0035] Secondly, the present invention provides a GIL device fault detection system based on a dynamic coupling model of acoustic signatures, comprising:

[0036] The feature vector extraction module is used to collect the sound signals inside the GIL device in real time, extract feature parameters, and obtain feature vectors.

[0037] The anomaly detection model building module is used to build a Gaussian mixture model using normal operating condition data, train the Gaussian mixture model based on feature vectors, and obtain the anomaly detection model.

[0038] The detection parameter calculation module is used to extract feature parameters from the real-time signal to be detected and calculate the estimated value and residual of the extracted features on the anomaly detection model.

[0039] The fault detection module is used to calculate the anomaly score based on the calculated estimate and residual, and output the corresponding early warning information and fault location results.

[0040] Thirdly, the present invention provides an electronic device, comprising:

[0041] Memory and processor;

[0042] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of a GIL device fault detection method based on a dynamic coupling model of acoustic signatures.

[0043] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement a GIL device fault detection method based on a dynamic coupling model of acoustic signatures.

[0044] The beneficial effects of this invention are as follows: The improved MFCC feature extraction method and Mel filter bank processing provided by this invention can accurately capture acoustic features, while the time-frequency domain fusion analysis integrates time and frequency information, enhancing the detection capability of early weak faults and significantly improving the fault identification rate; the integrated monitoring model, through machine learning and deep learning algorithms, intelligently learns the equipment operation mode and, combined with a dynamic coupling model, can effectively distinguish between environmental noise and electrical fault acoustics, improving the accuracy of anomaly source identification; it can accurately identify abnormal patterns and fault symptoms in real time, issue early warnings, and classify alarm information according to level and pinpoint the component, enabling maintenance personnel to quickly locate faults, reduce fault losses, and improve processing efficiency. Attached Figure Description

[0045] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is an overall flowchart of the GIL device fault detection method based on the acoustic dynamic coupling model provided by the present invention. Detailed Implementation

[0047] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0048] Example 1, referring to Figure 1 This is the first embodiment of the present invention, which provides a GIL device fault detection method based on a dynamic coupling model of acoustic signatures, including:

[0049] S1: Real-time acquisition of sound signals inside the GIL device, extraction of feature parameters, and obtaining feature vectors;

[0050] S2: Use normal operating condition data to build a Gaussian mixture model, train the Gaussian mixture model based on feature vectors, and obtain an anomaly detection model;

[0051] S3: Extract feature parameters from the real-time signal to be detected, and calculate the estimated value and residual of the extracted features on the anomaly detection model;

[0052] S4: Based on the calculated estimated value and residual, calculate the anomaly score and output the corresponding early warning information and fault location results.

[0053] It should be noted that, through steps S1-S4, this embodiment uses high-precision sensors to collect acoustic signals inside the GIL device, and then analyzes the changes in the device's operating status. This monitoring method can effectively identify abnormal sounds inside the device, such as partial discharge and mechanical vibration, thereby promptly detecting potential faults. The advantage of acoustic signature monitoring technology lies in its non-invasiveness and sensitivity, enabling it to capture subtle characteristics of changes in the device's status without affecting its normal operation.

[0054] Example 2, refer to Figure 1 As an embodiment of the present invention, based on the previous embodiment, a GIL device fault detection method based on a dynamic coupling model of acoustic signatures is provided, comprising:

[0055] In this embodiment, the sound signal inside the GIL device is acquired in real time in step S1 above, and feature parameters are extracted to obtain the feature vector, including:

[0056] A high-precision audio sensor is selected to collect the sound signals inside the GIL device. The audio sensor not only has high precision and high sensitivity, but also can work stably in harsh environments, thus ensuring the reliability of data acquisition.

[0057] For the collected audio signal from inside the GIL device, wavelet denoising is used. Assume the signal consists of two parts: valid signal and noise, represented as follows:

[0058] ;

[0059] in, Represents sound signals. , It is the signal length. These are the effective signal components that need to be retained. It is a noise component.

[0060] After selecting a suitable wavelet basis, the signal Wavelet decomposition is performed to obtain a series of wavelet coefficients. After thresholding the wavelet coefficients and discarding the noise transform coefficients, an inverse wavelet transform is performed to obtain the noise-removed signal.

[0061] Furthermore, wavelet coefficient thresholding includes various methods such as hard thresholding and soft thresholding. To address the discontinuous transitions caused by thresholding, an improved thresholding method is adopted. Specifically, when the absolute value of a wavelet coefficient is greater than or equal to the threshold, a specific function transformation is applied; when it is less than the threshold, it is directly set to zero, as shown below:

[0062] ;

[0063] Where w represents the wavelet coefficients obtained after wavelet decomposition. It is a sign function used to determine the sign of w, and its definition is as follows:

[0064] ;

[0065] threshold The calculation method is as follows:

[0066] ;

[0067] in, The standard deviation of noise Indicates the signal length.

[0068] It should be noted that combining the above algorithms with adaptive filtering techniques can improve signal quality.

[0069] For the denoised audio signal, pre-emphasis is applied, and the calculation formula is as follows:

[0070] ;

[0071] in This represents the sample value of the pre-emphasis processed audio signal at the nth sampling time. This represents the sample value of the denoised audio signal at the nth sampling time. This represents the sample value of the denoised audio signal at the (n-1)th sampling time. It is the pre-weighting coefficient. It takes a value between 0.9 and 1.

[0072] The pre-emphasized audio signal is divided into frames, with each frame lasting between 20ms and 30ms, and the overlap time between different frames is 10ms.

[0073] After framing, spectral leakage is suppressed by windowing the audio signal. A Hamming window is applied to the framed audio signal, as shown below:

[0074] ;

[0075] ;

[0076] in, This represents the value of the windowed audio signal at the nth sampling point. This represents the value of the audio signal at the nth sampling point after frame division. This represents the value of the window function at the nth sampling point, where n is a discrete time index representing the position of the current sampling point in the window function sequence. The value of n ranges from 0 to N−1. This represents the number of sampling points in each frame of the audio signal, that is, the number of samples contained in a frame of the signal.

[0077] After the windowing operation is completed, the sound is converted from the time domain to the frequency domain by fast Fourier transform and the amplitude spectrum is calculated. That is, the amplitude is calculated for each point in the frequency domain, and then the sound spectrum is converted by a triangular bandpass filter bank.

[0078] Specifically, the formula for calculating a triangular bandpass filter is as follows:

[0079] ;

[0080] in This represents the filter coefficient of the m-th triangular bandpass filter at the k-th frequency point. This is the filter number, with values ​​ranging from m=1,2,⋯,M, where M is the total number of filters. These are the weighting coefficients of the filtered amplitude spectrum. It is the center frequency of the m-th Mel bandpass filter.

[0081] The conversion relationship between Mel frequency and physical frequency is as follows:

[0082] ;

[0083] in, This represents the function value that converts the physical frequency f to the Mel frequency.

[0084] The formula for calculating the center frequency is as follows:

[0085] ;

[0086] ;

[0087] in It is the physical frequency, where and These are the lower and upper limits of the filter frequency. It is the number of sampling points. It is the sampling frequency.

[0088] A smooth sound amplitude spectrum was obtained by multiplying the amplitude spectrum with a triangular filter bank.

[0089] The discrete cosine transform is defined as follows:

[0090] ;

[0091] in It is the first The logarithmic magnitude of the output of each Mel filter bank, It is the number of Mel filters. It is the eigenvector of MFCC (Mel Frequency Cepstral Coefficients).

[0092] In this embodiment, step S2 above uses normal operating condition data to establish a Gaussian mixture model, and trains the Gaussian mixture model based on feature vectors to obtain an anomaly detection model, including:

[0093] A Gaussian mixture model (GMM) was established using normal operating data to obtain the characteristic distribution of the conveyor during normal operation.

[0094] The model is defined as follows:

[0095] ;

[0096] in, Let K represent the probability density function of the Gaussian mixture model, where K is the number of Gaussian components in the Gaussian mixture model. These are the weights of each Gaussian distribution. and They are the first The mean vector and covariance matrix of a Gaussian distribution. Let i represent the i-th Gaussian distribution (normal distribution).

[0097] Collect sound data of the GIL operating normally, calculate MFCC features as input to GMM, and use the EM algorithm to calculate the parameters of the GMM model.

[0098] In this embodiment, step S3 above involves extracting feature parameters from the real-time signal to be detected and calculating the estimated values ​​and residuals of the extracted features on the anomaly detection model, including:

[0099] For the sound sample to be tested, the sound is first converted into MFCC features, and then the estimated value is calculated on the normal GMM model. The specific calculation formula is as follows:

[0100] ;

[0101] in, The MFCC feature to be tested is in the first... Posterior probabilities on a Gaussian distribution, characteristic residuals for:

[0102] ;

[0103] in, This represents the MFCC feature obtained by converting the sound sample to be tested.

[0104] In this embodiment, step S4 above, which calculates the anomaly score and outputs the corresponding early warning information and fault location results based on the calculated estimate and residual, includes:

[0105] Define anomaly score As shown below:

[0106] ;

[0107] Where D is the dimension of the MFCC feature, It is a variable that measures the degree of abnormality of the sound sample being tested.

[0108] In another possible implementation, multiple experiments can be conducted on normal operating condition data, and a suitable anomaly threshold T can be determined based on the calculated distribution of anomaly scores. For example, the threshold can be set as a certain quantile (such as the 95th quantile) of the anomaly scores of normal samples by statistically analyzing the anomaly scores of normal samples. When the anomaly score exceeds this threshold, the sample is considered to be abnormal.

[0109] For example, the calculated anomaly score F is compared with a set anomaly threshold T.

[0110] If F>T, the GIL device corresponding to the sound sample being tested is determined to be malfunctioning, and a corresponding warning message is output, such as "The GIL device may be faulty; please check it promptly." The warning message can be sent to relevant personnel via system interface display, SMS, email, etc.

[0111] If F≤T, the GIL device is determined to be operating normally, and a normal operating prompt message, such as "GIL device is operating normally", is output.

[0112] Analyze the MFCC characteristics of abnormal samples to identify which feature dimensions have larger residuals. Combine this with the correlation between known fault modes and MFCC characteristics to preliminarily determine the possible fault types. For example, abnormal MFCC characteristics within certain frequency ranges may correspond to partial discharge faults, while other abnormal characteristics may be related to mechanical vibration faults. Based on the analysis results, output specific fault location information, such as "GIL equipment [specific location] may have [fault type] fault," providing clear maintenance guidance for equipment maintenance personnel.

[0113] Example 3: The above is an illustrative scheme of the GIL device fault detection method based on the acoustic print dynamic coupling model in this embodiment. It should be noted that the technical solution of the GIL device fault detection system based on the acoustic print dynamic coupling model and the technical solution of the GIL device fault detection method based on the acoustic print dynamic coupling model described above belong to the same concept. Details not described in detail in the technical solution of the GIL device fault detection system based on the acoustic print dynamic coupling model in this embodiment can be found in the description of the technical solution of the GIL device fault detection method based on the acoustic print dynamic coupling model described above.

[0114] This embodiment also provides a GIL device fault detection system based on a dynamic coupling model of acoustic signatures, including:

[0115] The feature vector extraction module is used to collect the sound signals inside the GIL device in real time, extract feature parameters, and obtain feature vectors.

[0116] The anomaly detection model building module is used to build a Gaussian mixture model using normal operating condition data, train the Gaussian mixture model based on feature vectors, and obtain the anomaly detection model.

[0117] The detection parameter calculation module is used to extract feature parameters from the real-time signal to be detected and calculate the estimated value and residual of the extracted features on the anomaly detection model.

[0118] The fault detection module is used to calculate the anomaly score based on the calculated estimate and residual, and output the corresponding early warning information and fault location results.

[0119] This embodiment also provides an electronic device applicable to the GIL device fault detection method based on the acoustic signature dynamic coupling model, including:

[0120] The system includes a memory and a processor. The memory stores computer-executable instructions, and the processor executes these instructions to implement the GIL device fault detection method based on the dynamic coupling model of acoustic signatures as proposed in the above embodiments.

[0121] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the GIL device fault detection method based on the acoustic dynamic coupling model proposed in the above embodiments.

[0122] The storage medium proposed in this embodiment belongs to the same inventive concept as the GIL device fault detection method based on the acoustic dynamic coupling model proposed in the above embodiments. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0123] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A GIL device fault detection method based on a dynamic coupling model of voiceprints, characterized in that, The method comprises the following steps: Real-time acquisition of sound signals inside the GIL device, feature parameter extraction, and obtaining of a feature vector; Establishment of a Gaussian mixture model using normal operating condition data, training of the Gaussian mixture model based on the feature vector, and obtaining of an anomaly detection model; Feature parameter extraction of real-time signals to be detected, calculation of an estimated value and a residual of the extracted features on the anomaly detection model; Based on the calculated estimated value and residual, an anomaly score is calculated, and corresponding warning information and fault location results are outputted. The real-time acquisition of sound signals inside the GIL device, feature parameter extraction, and obtaining of a feature vector comprise the following steps: The collected sound signals are decomposed into effective signals and noise components, and wavelet decomposition is performed on the sound signals by selecting a wavelet basis, thereby obtaining a series of wavelet coefficients; Different processing is performed according to the size relationship between the absolute value of the wavelet coefficient and the preset threshold value: when the absolute value of the wavelet coefficient is greater than or equal to the threshold value, a specific function transformation is performed thereon; when it is less than the threshold value, it is directly set to zero; wherein the preset threshold value is calculated according to the standard deviation of the noise and the signal length; The transformed coefficients determined as noise after processing are discarded, and inverse wavelet transformation is performed to obtain the signal after noise removal; The sound signal after noise removal is pre-emphasized, and the pre-emphasized sound signal is segmented into frames, the duration of the frame being between 20 ms and 30 ms, and the overlapping time between different frames being 10 ms; Windowing operation is performed on the sound signal after segmentation, and each frame of signal is weighted by a window function; The windowed sound signal is converted from the time domain to the frequency domain, and the amplitude of each point in the frequency domain is calculated to obtain the amplitude spectrum of the sound signal; The triangular bandpass filter bank is used to convert the sound amplitude spectrum, the amplitude spectrum is multiplied by the triangular filter bank to obtain the smoothed sound amplitude spectrum, and the discrete cosine transform is performed on the signal after the spectrum smoothing processing to extract the mel frequency cepstral coefficient feature vector.

2. The GIL device fault detection method based on a dynamic coupling model of voiceprints according to claim 1, wherein, The Gaussian mixture model is established using normal operating condition data, the Gaussian mixture model is trained based on the feature vector, and the anomaly detection model is obtained, which comprises the following steps: The feature vector extracted from the relevant data under the normal operating condition of the device is used as the input data of the Gaussian mixture model, the expectation maximization algorithm is used to calculate and optimize the parameters of the Gaussian mixture model, the model training is completed, and the anomaly detection model is obtained.

3. A GIL equipment fault detection system based on a dynamic coupling model of voiceprint, applying the method of any one of claims 1-2, characterized in that, The method comprises the following steps: A feature vector extraction module is configured to real-time acquisition of sound signals inside the GIL device, feature parameter extraction, and obtaining of a feature vector; An anomaly detection model establishment module is configured to establishment of a Gaussian mixture model using normal operating condition data, training of the Gaussian mixture model based on the feature vector, and obtaining of an anomaly detection model; A detection parameter calculation module is configured to feature parameter extraction of real-time signals to be detected, calculation of an estimated value and a residual of the extracted features on the anomaly detection model; A fault detection module is configured to calculation of an anomaly score based on the calculated estimated value and residual, and outputting of corresponding warning information and fault location results.

4. An electronic device, comprising: The method comprises the following steps: A memory and a processor; The memory is configured to store computer-executable instructions, and the processor is configured to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the method of any one of claims 1 to 2.

5. A computer readable storage medium, characterized in that, The computer program product has stored computer-executable instructions, which, when executed by the processor, implement the steps of the method of any one of claims 1 to 2.

Images