Sensor signal fault diagnosis method

Abstract

The embodiment of the invention further provides a storage medium, a computer program is stored on the storage medium, and when the program is executed by a processor, the method for realizing the sensor signal fault diagnosis provided by the embodiment is realized. The storage medium can be implemented by any type of volatile or non-volatile storage devices or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an Erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a memory cell, a memory cell, a memory cell, a memory cell, a memory cell, a memory cell and a memory cell. A flash memory, a disk or an optical disk. It should be explained that the embodiment is only used for explaining the technical scheme of the invention instead of being limited, although the technical scheme is explained in detail by referring to a better embodiment, ordinary technicians in the field should understand that the technical scheme of the invention can be modified or equivalently replaced without being separated from the spirit and range of the technical scheme of the invention, so that the technical scheme of the invention can be modified or equivalently replaced. All the components should be contained in the scope of the claims of the invention.

Description

Technical Field

[0001] The present invention relates to the technical field of signal processing, and in particular to a method for diagnosing faults of sensor signals. Background Art

[0002] In modern industry and equipment management, the monitoring and diagnosis of sensor signals is an important means to ensure the normal operation of equipment. Sensors are widely used in various fields, including manufacturing, energy management, transportation, etc., and can collect real-time data on the working status of equipment. However, as the scale and complexity of equipment increase, traditional fault diagnosis methods face many technical challenges.

[0003] Current fault diagnosis technologies mainly rely on rule-based methods or classic machine learning algorithms. These methods often exhibit problems such as low efficiency, insufficient accuracy, and poor real-time performance when processing complex time series signals. For example, traditional feature extraction methods cannot effectively capture the dynamic changes of signals, resulting in inaccurate judgment of fault types. In addition, with the surge in data volume, the consumption of computing resources has also increased, and traditional computing methods are difficult to meet the needs of real-time monitoring. Summary of the invention

[0004] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0005] In view of the problems existing in the above-mentioned existing method for diagnosing faults using sensor signals, the present invention is proposed.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] A method for diagnosing faults of sensor signals, characterized in that the method comprises the following steps:

[0008] Collect raw data of sensor signals and convert them into time series signal graphs, which can be used as input for deep neural network models;

[0009] Then, a deep neural network model is constructed, and the time evolution characteristics of the signal are provided by the time series signal spectrum, which is used as the input of the deep neural network model. The model will determine the fault type based on the signal changes in the spectrum.

[0010] As a preferred solution of the method for diagnosing faults by sensor signals described in the present invention, the method also includes: introducing quantum computing, accelerating the processing of timing signal graphs and deep neural network operations, and analyzing the status of the equipment in real time to ensure detection and early warning of problems, thereby avoiding the risk of equipment damage and downtime.

[0011] As a preferred solution of the method for diagnosing faults with sensor signals described in the present invention, wherein:

[0012] The conversion process of the timing signal spectrum is as follows:

[0013] S1: Assume the sensor signal is , which represents the signal data collected by the sensor that changes with time t, and then normalized to obtain :

[0014] S2: In order to reduce the influence of noise, the normalized signal is smoothed to obtain :

[0015] S3: Extract features from the signal through Gaussian distribution to capture the changing pattern of the signal in time series:

[0016]

[0017] Which represents the signal The mean of represents the center position of the signal. is the standard deviation of the signal, indicating the fluctuation range of the signal. is the signal feature calculated by Gaussian distribution at time t;

[0018] S4: By summing the features of different signal samples, we have , forming the overall signal spectrum, N represents the number of samples;

[0019] S5: By integrating the signal features, a time series signal spectrum is formed so as to be input into the deep neural network. The final time series signal spectrum is expressed as:

[0020] Among them, I is the final time series signal spectrum, which represents the comprehensive characteristics of the signal within time T, and the calculated spectrum is normalized: , so that its output range remains between [0,1], which is convenient for neural network input.

[0021] As a preferred solution of the method for diagnosing faults with sensor signals described in the present invention, the specific operation of constructing the deep neural network model structure is:

[0022] Input layer: accepts the time series signal graph produced in step 1 As input, let the dimension of the graph be ,in represents the number of features, Indicates the length of the time series;

[0023]

[0024] Indicates that the signal spectrum is a two-dimensional array with a size of ;

[0025] Convolutional layer: In order to extract local features in the time series signal, a convolutional neural network (CNN) is used to perform a convolution operation on the input. The convolution kernel size can be set to: ; ;in, Represents the output result of the kth convolution kernel at time t after the convolution operation. represents the weight of the convolution kernel, Indicates the normalized time series signal spectrum at time The value at the moment, b represents the bias term in the convolution operation; Pooling layer: pooling operation is performed on the output after convolution to retain the most important features in the time series signal spectrum, reducing the amount of data while maintaining key information: ; Among them, max represents the maximum pooling operation, that is, selecting the maximum value from the K convolution results; Recurrent neural network layer: In order to capture the long-term dependency characteristics of the time series signal, the LSTM network is used to process the pooled features. LSTM has a memory unit that can maintain long-term signal change information; ;in and Input weight and hidden layer weight matrix respectively, is the Sigmoid activation function, is the LSTM output at time t.

[0026] As a preferred solution of the method for diagnosing faults of sensor signals described in the present invention, the fault diagnosis process of the deep neural network model is as follows: The input is passed into the input layer of the deep neural network as input; the convolution layer extracts local time series features, the pooling layer reduces the feature dimension, and the LSTM layer captures global time series features; finally, the classification result is output through the fully connected layer, that is, the predicted probability of the fault type: ;in, Represents the output of the model, that is, the prediction result of the fault type. The Softmax function converts the network output into a probability distribution; is the weight matrix of the fully connected layer, which is used to transform the output of the LSTM network Convert to classification results.

[0027] As a preferred solution of the sensor signal fault diagnosis method of the present invention, the introduction process of quantum computing includes:

[0028] The timing signal spectrum Each eigenvalue in is encoded as a quantum bit and represented by a quantum state: ;

[0029] Where, represents the quantum state, N represents the total number of quantum bits or the number of ground states, represents the complex magnitude, represents the ground state;

[0030] Design a quantum computing model, that is, build a quantum circuit suitable for processing timing signals, including quantum gates, to achieve feature extraction and calculation: ;

[0031] Among them, U represents the unit operation of the quantum circuit, H represents the Hadamard gate, which is used to transform the ground state into the superposition state, and CNOT represents the controlled NOT gate, which is used to create the entangled state;

[0032] The real-time collected time series signal spectrum data is converted into quantum states and input into quantum circuits for rapid processing. The quantum computing model outputs fault prediction results based on the input quantum states and predicts the probability distribution of faults: .

[0033] As a preferred solution of the method for diagnosing faults with sensor signals described in the present invention, a fault threshold is set. If the model predicts that the fault probability exceeds the set threshold, an alarm mechanism is triggered to promptly notify the operator to intervene.

[0034] The fault diagnosis system of the sensor signal fault diagnosis method as described above comprises: a data acquisition module, which collects raw data of the sensor and monitors the device status in real time; a data preprocessing module, which cleans and converts the collected raw data and generates a time series signal spectrum;

[0035] The quantum coding module converts the preprocessed time-series signal spectrum into a quantum state for quantum computing. The quantum computing module uses quantum circuits and quantum algorithms to process quantum states for signal analysis and fault prediction.

[0036] The deep learning module classifies faults through deep neural networks based on time series signal graphs and quantum computing results. The prediction and warning module analyzes the calculation results in real time, determines the fault type and triggers the warning mechanism.

[0037] A computer device includes a memory and a processor, wherein the memory stores a computer program, and is characterized in that when the processor executes the computer program, the steps of the above-mentioned method for diagnosing faults with sensor signals are implemented.

[0038] A computer-readable storage medium stores a computer program, wherein the computer program implements the steps of the above-mentioned method for diagnosing faults with sensor signals when executed by a processor.

[0039] Beneficial effects of the present invention:

[0040] 1. The method of the present invention has high accuracy: the deep learning model can accurately determine the fault type by extracting effective features of the time series signal, thereby reducing false alarms and missed alarms; strong adaptability: it can be continuously updated and optimized according to real-time data to improve the generalization ability of the model and adapt to the monitoring needs of different equipment and fault types.

[0041] 2. The real-time performance of the method of the present invention is improved: the high-speed processing capability of quantum computing makes real-time analysis and early warning possible, so as to detect potential faults in time and reduce the risk of equipment downtime. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0043] Figure 1 A schematic diagram of the overall flow of a method for diagnosing faults with sensor signals proposed by the present invention;

[0044] Figure 2 The present invention is a schematic diagram of a method for diagnosing faults with sensor signals. DETAILED DESCRIPTION

[0045] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0046] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0047] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0048] Example 1, reference Figure 1 , as an embodiment of the present invention, provides a method for diagnosing faults of sensor signals, the method comprising:

[0049] Collect the raw data of sensor signals and convert them into time series signal graphs, which can be used as input for deep neural network models.

[0050] The conversion process of the timing signal spectrum is as follows:

[0051] S1: Assume the sensor signal is , which represents the signal data collected by the sensor that changes with time t, and then normalized to obtain :

[0052] S2: In order to reduce the influence of noise, the normalized signal is smoothed to obtain :

[0053] S3: Extract features from the signal through Gaussian distribution to capture the changing pattern of the signal in time series: ; where the signal The mean of represents the center position of the signal. is the standard deviation of the signal, indicating the fluctuation range of the signal. is the signal feature calculated by Gaussian distribution at time t;

[0054] S4: By summing the features of different signal samples, we have , forming the overall signal spectrum, N represents the number of samples;

[0055] S5: By integrating the signal features, a time series signal spectrum is formed so as to be input into the deep neural network. The final time series signal spectrum is expressed as:

[0056] Among them, I is the final time series signal spectrum, which represents the comprehensive characteristics of the signal within time T, and the calculated spectrum is normalized: , so that its output range remains between [0,1], which is convenient for neural network input.

[0057] The time series signal spectrum is like a "dynamic photo" of the signal, allowing us to see how these signals change over time. This helps capture changes before and after a device fails, especially subtle fluctuations in the signal.

[0058] Reference Figure 2 , which reflects the four steps of constructing the time series spectrum. Step 1 shows the original sensor signal, whose fluctuations reflect the changing characteristics of the signal and the influence of random noise. Step 2 shows the signal after normalization, ensuring that the signal is at a uniform level for subsequent analysis. Step 3 shows the signal after logarithmic smoothing, which effectively reduces the interference of noise and highlights the trend of the signal. Step 4 shows the signal features extracted by Gaussian distribution, revealing the statistical characteristics of the signal and providing key feature information for subsequent fault diagnosis.

[0059] Then, a deep neural network model is built, and the time evolution characteristics of the signal are provided by the time series signal spectrum, which is used as the input of the deep neural network model. The model will judge the fault type based on the signal changes in the spectrum. The deep neural network model is like an "intelligent doctor" that can automatically identify different fault types by learning a large amount of historical fault data.

[0060] The specific operations for building a deep neural network model structure are:

[0061] Input layer: accepts the time series signal graph produced in step 1 As input, let the dimension of the graph be ,in represents the number of features, Indicates the length of the time series;

[0062]

[0063] Indicates that the signal spectrum is a two-dimensional array with a size of ;

[0064] Convolutional layer: In order to extract local features in the time series signal, a convolutional neural network (CNN) is used to perform a convolution operation on the input. The convolution kernel size can be set to: ;

[0065] in, Represents the output result of the kth convolution kernel at time t after the convolution operation. represents the weight of the convolution kernel, Indicates the normalized time series signal spectrum at time The value at the moment, b represents the bias term in the convolution operation;

[0066] Pooling layer: Pooling is performed on the output after convolution to retain the most important features in the time series signal spectrum, reducing the amount of data while maintaining key information:

[0067]

[0068] Among them, max represents the maximum pooling operation, that is, selecting the maximum value among the K convolution results;

[0069] Recurrent neural network layer: In order to capture the long-term dependency characteristics of time series signals, the LSTM network is used to process the pooled features. LSTM has memory units that can retain long-term signal change information;

[0070]

[0071] in and Input weight and hidden layer weight matrix respectively, is the Sigmoid activation function, is the LSTM output at time t.

[0072] For fault diagnosis problems, the cross entropy loss function is usually used to evaluate the classification performance. Here, the predicted output of the model is , the actual label is .

[0073] Then the cross entropy loss function is: ;

[0074] The fault diagnosis process of the deep neural network model is:

[0075] The timing signal spectrum Passed as input into the input layer of the deep neural network;

[0076] The convolution layer extracts local temporal features, the pooling layer reduces the feature dimension, and the LSTM layer captures global temporal features;

[0077] Finally, the classification result is output through the fully connected layer, that is, the predicted probability of the fault type:

[0078] in, Represents the output of the model, that is, the prediction result of the fault type. The Softmax function converts the network output into a probability distribution.

[0079] Finally, by comparing the probabilities of all fault types, the fault type with the highest probability is selected as the final prediction result.

[0080] is the weight matrix of the fully connected layer, which is used to transform the output of the LSTM network Convert to classification results.

[0081] It is obtained through the optimization algorithm during the model training process. Specifically, in each iteration, the model calculates the cross entropy loss function based on the training data, and then calculates the loss function relative to Gradient Using an optimization algorithm (such as the Adam optimizer), the model updates .

[0082] Example 2 is different from Example 1 in that, in order to ensure that the entire diagnostic process is fast enough, especially in complex industrial environments, we introduced quantum computing technology. Quantum computing can speed up the processing of time series signal graphs and deep neural network operations. This is like equipping our "smart doctor" with a supercomputer, allowing it to process large amounts of sensor signal data faster and more efficiently, and to obtain diagnostic results in a short time.

[0083] Specifically, quantum computing is introduced to accelerate the processing of timing signal graphs and deep neural network operations, and to analyze the status of equipment in real time, ensuring that problems can be detected and warned in the early stages of their occurrence, thus avoiding the risk of equipment damage and downtime.

[0084] The introduction process of quantum computing includes:

[0085] The timing signal spectrum Each eigenvalue in is encoded as a quantum bit and represented by a quantum state: ;

[0086] Where, represents the quantum state, N represents the total number of quantum bits or the number of ground states, represents the complex magnitude, represents the ground state;

[0087] Design a quantum computing model, that is, build a quantum circuit suitable for processing timing signals, including quantum gates, to achieve feature extraction and calculation: ;

[0088] Among them, U represents the unit operation of the quantum circuit, H represents the Hadamard gate, which is used to transform the ground state into the superposition state, and CNOT represents the controlled NOT gate, which is used to create the entangled state;

[0089] The real-time collected time series signal spectrum data is converted into quantum states and input into quantum circuits for rapid processing. The quantum computing model outputs fault prediction results based on the input quantum states and predicts the probability distribution of faults: .

[0090] Set a fault threshold. If the model predicts that the probability of failure exceeds the set threshold, the alarm mechanism is triggered to promptly notify the operator to intervene.

[0091] In summary, the fault diagnosis system applied to the above-mentioned sensor signal fault diagnosis method comprises: a data acquisition module, which collects the original data of the sensor and monitors the device status in real time; a data preprocessing module, which cleans and converts the collected original data and generates a time series signal spectrum;

[0092] The quantum coding module converts the preprocessed time-series signal spectrum into a quantum state for quantum computing. The quantum computing module uses quantum circuits and quantum algorithms to process quantum states for signal analysis and fault prediction.

[0093] The deep learning module classifies faults through deep neural networks based on time series signal graphs and quantum computing results. The prediction and warning module analyzes the calculation results in real time, determines the fault type and triggers the warning mechanism.

[0094] This embodiment also provides a computer device, which is applicable to a method for diagnosing faults using sensor signals, and includes: a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement a method for diagnosing faults using sensor signals as proposed in the above embodiment.

[0095] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covered on the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0096] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, a method for diagnosing a sensor signal fault as proposed in the above embodiment is implemented. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, a disk or an optical disk.

[0097] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for diagnosing faults of sensor signals, characterized in that: The method includes the following steps: collecting raw data of sensor signals, converting these raw sensor signals into time series signal spectra, and using them as input of a deep neural network model; then constructing a deep neural network model, and using the time series signal spectra to provide the time evolution characteristics of the signal, using them as input of the deep neural network model. The model will determine the fault type based on the signal changes in the spectra.

2. A sensor signal fault diagnosis method according to claim 1, characterized in that: The method also includes: introducing quantum computing to accelerate the processing of time series signal graphs and deep neural network operations, and analyzing the status of equipment in real time to ensure detection and early warning of problems, thereby avoiding the risk of equipment damage and downtime.

3. The method for diagnosing faults of sensor signals according to claim 2, characterized in that: The conversion process of the timing signal spectrum is as follows: S1: Assume that the signal of the sensor is , which represents the signal data collected by the sensor that changes with time t, and then normalized to obtain : ; S2: In order to reduce the influence of noise, the normalized signal is smoothed to obtain : ; S3: Extract features of the signal through Gaussian distribution to capture the change pattern of the signal in time series: in Indicates signal The mean of represents the center position of the signal. is the standard deviation of the signal, indicating the fluctuation range of the signal. is the signal feature calculated by Gaussian distribution at time t; S4: By summing the features of different signal samples, we have , forming the overall signal spectrum, N represents the number of samples; S5: by integrating the signal features, a spectrum of the time series signal is formed so as to be input into the deep neural network, and the final time series signal spectrum is expressed as: Among them, I is the final time series signal spectrum, which represents the comprehensive characteristics of the signal within time T, and the calculated spectrum is normalized: , so that its output range remains between [0,1], which is convenient for neural network input.

4. A sensor signal fault diagnosis method according to claim 3, characterized in that: The specific operations of building a deep neural network model structure are as follows: Input layer: accepts the time series signal graph produced in step 1 As input, let the dimension of the graph be ,in represents the number of features, Indicates the length of the time series; ; Indicates that the signal spectrum is a two-dimensional array with a size of ; Convolutional layer: In order to extract local features in the time series signal, a convolutional neural network (CNN) is used to perform a convolution operation on the input, and the convolution kernel size is set to: ; ;in, Represents the output result of the kth convolution kernel at time t after the convolution operation. represents the weight of the convolution kernel, Indicates the normalized time series signal spectrum at time The value at the moment, b represents the bias term in the convolution operation; Pooling layer: pooling operation is performed on the output after convolution to retain the most important features in the time series signal spectrum, reducing the amount of data while maintaining key information: ; Among them, max represents the maximum pooling operation, that is, selecting the maximum value among the K convolution results; Recurrent neural network layer: In order to capture the long-term dependency characteristics of the time series signal, the LSTM network is used to process the pooled features. LSTM has a memory unit that can maintain long-term signal change information; ;in and Input weight and hidden layer weight matrix respectively, is the Sigmoid activation function, is the LSTM output at time t.

5. The method for diagnosing faults of sensor signals according to claim 4, characterized in that: The fault diagnosis process of the deep neural network model is as follows: The input is passed into the input layer of the deep neural network as input; the convolution layer extracts local time series features, the pooling layer reduces the feature dimension, and the LSTM layer captures global time series features; finally, the classification result is output through the fully connected layer, that is, the predicted probability of the fault type: ;in, Represents the output of the model, that is, the prediction result of the fault type. The Softmax function converts the network output into a probability distribution; is the weight matrix of the fully connected layer, which is used to transform the output of the LSTM network Convert to classification results.

6. The method for diagnosing faults of sensor signals according to claim 2, characterized in that: The introduction process of quantum computing includes: converting the time series signal spectrum Each eigenvalue in is encoded as a quantum bit and represented by a quantum state: ; where, represents the quantum state, N represents the total number of quantum bits or the number of ground states, represents the complex magnitude, Represent the ground state; design a quantum computing model, that is, establish a quantum circuit suitable for processing timing signals, including quantum gates, for feature extraction and calculation: ; Where U represents the unit operation of the quantum circuit, H represents the Hadamard gate, which is used to convert the ground state into the superposition state, and CNOT represents the controlled NOT gate, which is used to create the entangled state; the real-time collected time series signal spectrum data is converted into a quantum state and input into the quantum circuit for rapid processing. The quantum computing model outputs the fault prediction result according to the input quantum state and predicts the probability distribution of the fault: .

7. A method for diagnosing faults using sensor signals according to claim 5 or 6, characterized in that: Set a fault threshold. If the model predicts that the probability of failure exceeds the set threshold, the alarm mechanism is triggered to promptly notify the operator to intervene.

8. The fault diagnosis system of the method for diagnosing faults using sensor signals according to claim 7, characterized in that: The system includes: a data acquisition module, which collects raw data from sensors and monitors the status of equipment in real time; a data preprocessing module, which cleans and converts the collected raw data to generate a time series signal spectrum; a quantum coding module, which converts the preprocessed time series signal spectrum into a quantum state for quantum computing; a quantum computing module, which uses quantum circuits and quantum algorithms to process quantum states for signal analysis and fault prediction; a deep learning module, which classifies faults through a deep neural network based on the time series signal spectrum and quantum computing results; and a prediction and warning module, which analyzes the calculation results in real time, determines the fault type and triggers the warning mechanism.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for diagnosing faults with sensor signals as claimed in claim 7 are implemented. A computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the method for diagnosing faults with sensor signals as described in claim 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for diagnosing faults with sensor signals as claimed in claim 7 are implemented.

Images