Digital twin data noise reduction method based on convolutional neural network
By designing a digital twin data noise reduction method based on convolutional neural network, the problem of noise affecting data quality in the digital twin system is solved, efficient and accurate data noise reduction is achieved, and the decision-making and prediction capabilities of the model are improved.
Patent Information
- Application Number
- CN202411814085.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-11
- Publication Date
- 2025-04-25
AI Technical Summary
The environmental noise and equipment noise generated during sensor data acquisition in digital twin systems lead to inaccuracy of data, affecting the decision-making and prediction capabilities of the model. Traditional signal processing methods lack the ability to handle complex nonlinear noise.
Design a digital twin data denoising method based on convolutional neural network. By collecting and preprocessing original data, a convolutional neural network model is constructed, and the residual network module is added to optimize and train the model to improve the data denoising effect.
Effectively remove noise in digital twin systems, improve data quality, enhance model decision-making and prediction capabilities, overcome the shortcomings of traditional methods in dealing with complex nonlinear noise, and improve data processing efficiency and accuracy.
Smart Images
Figure CN119885841A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of data denoising, and in particular to a digital twin data denoising method based on convolutional neural network. Background Art
[0002] Digital twin technology refers to a virtual digital model built through physical equipment, sensors, simulation models and data processing technology. It reflects the state changes of physical systems through real-time data. For example, equipment or factories in the manufacturing industry will continuously collect data (such as temperature, pressure, vibration, etc.) through sensors. These data will be transmitted to the digital twin model for monitoring, simulation, prediction and optimization operations. However, in the process of collecting sensor data, noise will inevitably be generated, such as environmental noise, equipment noise, etc., resulting in inaccurate data, which in turn affects the decision-making and prediction capabilities of the digital twin model. Therefore, the digital twin system requires high-quality, denoised data to maintain the consistency of the virtual model and the physical system. Noise reduction technology can help solve this problem, so that the digital twin model can more accurately reflect the state of the real physical system.
[0003] At present, although traditional signal processing methods (such as filters, Fourier transform, etc.) have been used for data denoising, their ability to handle complex nonlinear noise is limited. In recent years, convolutional neural networks (CNNs) have been widely used in image processing, natural language processing and other fields due to their powerful data feature extraction capabilities, and have also shown great potential in data denoising. However, traditional CNN models still have bottlenecks in efficiency and accuracy when processing digital twin big data. Therefore, designing an improved CNN model that can effectively handle the noise problem in the digital twin system has important practical significance. 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 above existing problems, the present invention is proposed. Therefore, the present invention provides a digital twin data denoising method based on convolutional neural network to solve the above problems.
[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:
[0007] In a first aspect, the present invention provides a digital twin data denoising method based on a convolutional neural network, comprising:
[0008] Collect raw data in the digital twin system and perform data preprocessing;
[0009] Construct a convolutional neural network model, use the preprocessed raw data as model input, and output the data denoising result;
[0010] Based on the convolutional neural network model, a residual network module is added to optimize the model;
[0011] The convolutional neural network model is trained, parameters are updated, and the model output results are evaluated.
[0012] As a preferred solution of the digital twin data denoising method based on convolutional neural network described in the present invention, the collecting of raw data in the digital twin system and performing data preprocessing includes:
[0013] Collect the original data X in the digital twin system. Assume that the original data is a noisy time series or multidimensional matrix, expressed as:
[0014] X=S+N
[0015] Among them, S represents the actual signal data, and N represents the noise data;
[0016] The data is standardized and normalized, and the standardized data is defined as:
[0017]
[0018] Among them, μ represents the mean of the data and σ represents the standard deviation.
[0019] As a preferred solution of the digital twin data denoising method based on convolutional neural network described in the present invention, constructing a convolutional neural network model includes:
[0020] Assume that the input standardized data is X norm , after several layers of convolution operations and pooling operations, the local features of the data will be effectively extracted. The convolution operation is defined as:
[0021] F l =f(W l *X l-1 +b l )
[0022] Among them, F l represents the output feature map of the lth layer, W l represents the convolution kernel weight matrix of this layer, b l represents the bias vector of this layer, f is the activation function, and the convolution operation * represents the input data X l-1 With the convolution kernel W l Discrete convolution of
[0023] The pooling operation is defined as:
[0024] P l =pool(F l )
[0025] Among them, P l It is the output feature map after pooling.
[0026] As a preferred solution of the digital twin data denoising method based on convolutional neural network described in the present invention, the residual network module is expressed as:
[0027] F out =f(W l+1 *f(W l *X l +b l )+b l+1 )+X l
[0028] Among them, X l is the input of the residual module, F out is the output feature map.
[0029] As a preferred solution of the digital twin data denoising method based on convolutional neural network described in the present invention, training the convolutional neural network model includes:
[0030] Using a noisy dataset And the corresponding noise-free dataset As target output;
[0031] The loss function uses mean square error, expressed as:
[0032]
[0033] in, represents the denoised data output by the model, represents the target noise-free data, and N is the number of samples in the dataset.
[0034] As a preferred solution of the digital twin data denoising method based on convolutional neural network described in the present invention, the parameter updating includes:
[0035] During the model training process, the parameters are updated through the Adam optimization algorithm, which is expressed as:
[0036]
[0037] Among them, θ t represents the model parameters, η represents the learning rate, Represents the gradient of the loss function with respect to the parameters.
[0038] As a preferred solution of the digital twin data denoising method based on convolutional neural network described in the present invention, the evaluation of the model output result includes:
[0039] For each noisy input X test , the denoised data X is output through the convolutional neural network test , and compared with the actual noise-free data X clean Make a comparison;
[0040] The evaluation index of noise reduction performance is calculated by signal-to-noise ratio and mean square error.
[0041] In a second aspect, the present invention provides a digital twin data denoising system based on a convolutional neural network, comprising:
[0042] The acquisition module is used to collect the raw data in the digital twin system and perform data preprocessing;
[0043] A construction module is used to construct a convolutional neural network model, taking the preprocessed raw data as the model input to output the data denoising result;
[0044] An optimization module, used to add a residual network module based on the convolutional neural network model to optimize the model;
[0045] The output module is used to train the convolutional neural network model, update parameters, and evaluate the model output results.
[0046] In a third aspect, the present invention provides an electronic device, comprising:
[0047] Memory and processor;
[0048] 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, the steps of the digital twin data denoising method based on convolutional neural network are implemented.
[0049] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the digital twin data denoising method based on convolutional neural network.
[0050] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention can effectively deal with the environmental noise and equipment noise generated during the sensor data acquisition process, ensuring that the digital twin system can obtain high-quality, denoised data to improve the decision-making and prediction capabilities of the model; it can overcome the insufficient ability of traditional signal processing methods (such as filters and Fourier transforms) in processing complex nonlinear noise to adapt to the needs of digital twin systems; by designing an improved convolutional neural network model, it can improve the efficiency and accuracy of processing digital twin big data, and enhance the stability and real-time performance of the model; by applying innovative data denoising solutions, it can ensure the efficient operation of the digital twin system and improve its overall performance and application effect. BRIEF DESCRIPTION OF THE DRAWINGS
[0051] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.
[0052] Figure 1 This is a schematic diagram of the overall process of a digital twin data denoising method based on a convolutional neural network according to an embodiment of the present invention;
[0053] Figure 2 A schematic diagram of the structure of a convolutional neural network of a digital twin data denoising method based on a convolutional neural network according to an embodiment of the present invention;
[0054] Figure 3 A noise comparison schematic diagram of a digital twin data denoising method based on a convolutional neural network according to an embodiment of the present invention. DETAILED DESCRIPTION
[0055] In order to make the above-mentioned purposes, 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 drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.
[0056] 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.
[0057] 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.
[0058] The present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.
[0059] At the same time, in the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper, lower, inner and outer" are based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first, second or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.
[0060] In the present invention, unless otherwise clearly specified and limited, the terms "install, connect, connect" should be understood in a broad sense, for example: it can be a fixed connection, a detachable connection or an integral connection; it can also be a mechanical connection, an electrical connection or a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0061] Reference Figure 1-Figure 3 , is an embodiment of the present invention, and provides a digital twin data denoising method based on convolutional neural network, such as Figure 1 As shown, including:
[0062] S101, collect raw data in the digital twin system and perform data preprocessing;
[0063] S102, constructing a convolutional neural network model, using the preprocessed raw data as model input to output a data denoising result;
[0064] S103, based on the convolutional neural network model, adding a residual network module to optimize the model;
[0065] S104, training the convolutional neural network model, updating parameters, and evaluating the model output results.
[0066] In a preferred implementation of step S101, collecting raw data in the digital twin system and performing data preprocessing includes:
[0067] Collect the original data X in the digital twin system. Assume that the original data is a noisy time series or multidimensional matrix, expressed as:
[0068] X=S+N
[0069] Among them, S represents the actual signal data, and N represents the noise data;
[0070] In order to facilitate subsequent model training, the data is standardized and normalized, and the standardized data is defined as:
[0071]
[0072] Among them, μ represents the mean of the data and σ represents the standard deviation.
[0073] It should be noted that this step helps to reduce the impact of noisy data, allowing the convolutional neural network to focus more on extracting effective features.
[0074] In step S102, a convolutional neural network structure is designed. In order to process the complex noise in the data, a multi-layer convolutional neural network structure is designed in this embodiment. The convolutional neural network structure is shown in FIG. Figure 2 shown.
[0075] In a preferred embodiment, constructing a convolutional neural network model includes:
[0076] Assume that the input standardized data is X norm , after several layers of convolution operations and pooling operations, the local features of the data are effectively extracted. The convolution operation is defined as:
[0077] F l =f(W l *X l-1 +b l )
[0078] Among them, F l represents the output feature map of the lth layer, W l represents the convolution kernel weight matrix of this layer, b l represents the bias vector of the layer, f is the activation function (such as ReLU function), and the convolution operation * represents the input data X l-1 With the convolution kernel W l Discrete convolution of
[0079] Through the operation of multiple convolutional layers, the convolutional neural network can capture the multi-level features of the input data, including the spectral distribution of noise and potential characteristic patterns.
[0080] The convolution layer is used to extract local features of the data, while the pooling layer reduces the dimension of the feature map by downsampling (such as maximum pooling, average pooling). The pooling operation is defined as:
[0081] P l =pool(F l )
[0082] Among them, P l It is the output feature map after pooling.
[0083] It should be noted that through the operation of multiple convolutional layers, the convolutional neural network can capture the multi-level features of the input data, including the spectral distribution of noise and potential feature patterns. The convolutional layer is used to extract local features of the data, while the pooling layer reduces the dimension of the feature map by downsampling (such as maximum pooling and average pooling). Through the pooling operation, the redundant information of the data can be effectively reduced and the computational efficiency of the model can be improved.
[0084] In a preferred implementation of step S103, the residual network module is represented as:
[0085] F out =f(W l+1 *f(W l *X l +b l )+b l+1 )+X l
[0086] Among them, X l is the input of the residual module, F out is the output feature map.
[0087] It should be noted that in order to improve the denoising effect of the convolutional neural network on complex noise, this embodiment introduces a residual network module (ResNet) to solve the problem of gradient disappearance in the deep network and enable the input signal to directly participate in the calculation of the output layer. Through this structure, the network can learn deeper features while retaining important information in the input signal and reducing the loss of useful information in the denoising process.
[0088] In a preferred implementation of step S104, training the convolutional neural network model includes:
[0089] Using a noisy dataset And the corresponding noise-free dataset As target output;
[0090] The loss function uses mean square error, expressed as:
[0091]
[0092] Among them, X i represents the denoised data output by the model, represents the target noise-free data, and N is the number of samples in the dataset.
[0093] Parameter updates include:
[0094] During the model training process, the parameters are updated through the Adam optimization algorithm, which is expressed as:
[0095]
[0096] Among them, θ t represents the model parameters, η represents the learning rate, Represents the gradient of the loss function with respect to the parameters.
[0097] In a preferred implementation of step S104, evaluating the model output result includes:
[0098] For each noisy input X test , the denoised data X is output through the convolutional neural network test , and compared with the actual noise-free data X clean Make a comparison;
[0099] The evaluation index of noise reduction performance is calculated by signal-to-noise ratio and mean square error.
[0100] Among them, the signal-to-noise ratio (SNR) and the mean square error (MSE) are expressed as:
[0101]
[0102] It should be noted that after the training is completed, the model is applied to the real-time data denoising in the digital twin system. The noise comparison diagram of the application model is shown in Figure 3 As shown in Table 1, the above evaluation indicators are used to verify the denoising effect of the model under different noise conditions to ensure that the model can be applied to actual digital twin scenarios. The comparison of noise processing effects is shown in Table 1.
[0103] Table 1: Comparison of noise processing effects
[0104]
[0105] It should be noted that the present invention can effectively deal with the environmental noise and equipment noise generated during the sensor data acquisition process, ensuring that the digital twin system can obtain high-quality, denoised data to improve the decision-making and prediction capabilities of the model; it can overcome the insufficient ability of traditional signal processing methods (such as filters and Fourier transforms) in processing complex nonlinear noise to adapt to the needs of digital twin systems; by designing an improved convolutional neural network model, it can improve the efficiency and accuracy of processing digital twin big data and enhance the stability and real-time performance of the model; by applying innovative data denoising solutions, it can ensure the efficient operation of the digital twin system and improve its overall performance and application effect.
[0106] First, the present invention effectively removes noise from sensor data by applying improved convolutional neural network technology, ensuring that the digital twin system obtains higher quality input data, thereby improving the accuracy of decision-making and prediction. At the same time, the improved convolutional neural network model can better handle complex nonlinear noise, thereby improving the stability of the digital twin system and reducing fluctuations and errors caused by inaccurate data. In addition, when processing digital twin big data, the newly designed model can significantly improve data processing efficiency compared to traditional convolutional neural networks, ensure that real-time requirements are met, and enable the system to respond quickly to changes.
[0107] Through optimized data noise reduction technology, the digital twin system can more accurately reflect the state changes of the physical system, enhance real-time monitoring capabilities, and improve decision-making capabilities. It is worth mentioning that the technical solution of the present invention is not only applicable to traditional fields such as manufacturing, but can also be extended to other industries that require real-time data monitoring and analysis, and has strong application prospects. Ultimately, by solving the deficiencies in the prior art, the present invention will promote the innovation and development of digital twin technology in various industries and improve its application capabilities in complex environments.
[0108] The above is a schematic scheme of a digital twin data denoising method based on a convolutional neural network in this embodiment. It should be noted that the technical scheme of the digital twin data denoising system based on a convolutional neural network and the technical scheme of the digital twin data denoising method based on a convolutional neural network mentioned above belong to the same concept. The details of the technical scheme of the digital twin data denoising system based on a convolutional neural network in this embodiment that are not described in detail can all be referred to the description of the technical scheme of the digital twin data denoising method based on a convolutional neural network mentioned above.
[0109] The digital twin data denoising system based on convolutional neural network in this embodiment includes:
[0110] The acquisition module is used to collect the raw data in the digital twin system and perform data preprocessing;
[0111] A construction module is used to construct a convolutional neural network model, taking the preprocessed raw data as the model input to output the data denoising result;
[0112] The optimization module is used to optimize the model by adding a residual network module based on the convolutional neural network model;
[0113] The output module is used to train the convolutional neural network model, update its parameters, and evaluate the model output results.
[0114] This embodiment also provides an electronic device, which is applicable to the case of digital twin data noise reduction based on convolutional neural network, including:
[0115] Memory and processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the digital twin data denoising method based on convolutional neural network as proposed in the above embodiment.
[0116] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the digital twin data denoising method based on a convolutional neural network proposed in the above embodiment is implemented.
[0117] The storage medium proposed in this embodiment and the method for implementing digital twin data denoising based on convolutional neural network proposed in the above embodiment belong to the same inventive concept. For technical details not fully described in this embodiment, please refer to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.
[0118] Through the above description of the implementation methods, the technicians in the relevant field can clearly understand that the present invention can be implemented by means of software and necessary general hardware, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ReadOnly, Memory, ROM), random access memory (RandomAccess Memory, RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform the methods of various embodiments of the present invention.
[0119] 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 the present invention. 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 digital twin data denoising method based on convolutional neural network, characterized in that: include: Collect raw data in the digital twin system and perform data preprocessing; Construct a convolutional neural network model, use the preprocessed raw data as model input, and output the data denoising result; Based on the convolutional neural network model, a residual network module is added to optimize the model; The convolutional neural network model is trained, parameters are updated, and the model output results are evaluated.
2. The digital twin data denoising method based on convolutional neural network according to claim 1, characterized in that: The collecting of raw data in the digital twin system and the data preprocessing include: Collect the original data X in the digital twin system. Assume that the original data is a noisy time series or multidimensional matrix, expressed as: X=S+N Among them, S represents the actual signal data, and N represents the noise data; The data is standardized and normalized, and the standardized data is defined as: Among them, μ represents the mean of the data and σ represents the standard deviation.
3. The digital twin data denoising method based on convolutional neural network according to claim 1 or 2, characterized in that: Building a convolutional neural network model includes: Assume that the input standardized data is X norm , after several layers of convolution operations and pooling operations, the local features of the data will be effectively extracted. The convolution operation is defined as: F l =f(W l *X l-1 +b l ) Among them, F l represents the output feature map of the lth layer, W l represents the convolution kernel weight matrix of this layer, b l represents the bias vector of this layer, f is the activation function, and the convolution operation * represents the input data X l-1 With the convolution kernel W l Discrete convolution of The pooling operation is defined as: P l =pool(F l ) Among them, P l It is the output feature map after pooling.
4. The digital twin data denoising method based on convolutional neural network according to claim 3, characterized in that: The residual network module is expressed as: F out =f(W l+1 *f(W l *X l +b l )+b l+1 )+X l Among them, X l is the input of the residual module, F out is the output feature map.
5. The digital twin data denoising method based on convolutional neural network according to claim 1, characterized in that: Training the convolutional neural network model includes: Using a noisy dataset And the corresponding noise-free dataset As target output; The loss function uses mean square error, expressed as: in, represents the denoised data output by the model, represents the target noise-free data, and N is the number of samples in the dataset.
6. The digital twin data denoising method based on convolutional neural network according to claim 5, characterized in that: Parameter updates include: During the model training process, the parameters are updated through the Adam optimization algorithm, which is expressed as: Among them, θ t represents the model parameters, η represents the learning rate, Represents the gradient of the loss function with respect to the parameters.
7. The digital twin data denoising method based on convolutional neural network according to claim 6, characterized in that: Evaluation of model outputs includes: For each noisy input X test , the denoised data X is output through the convolutional neural network test , and compared with the actual noise-free data X clean Make a comparison; The evaluation index of noise reduction performance is calculated by signal-to-noise ratio and mean square error.
8. A digital twin data denoising system based on convolutional neural network, characterized in that: include, The acquisition module is used to collect the raw data in the digital twin system and perform data preprocessing; A construction module is used to construct a convolutional neural network model, taking the preprocessed raw data as the model input to output the data denoising result; An optimization module, used to add a residual network module based on the convolutional neural network model to optimize the model; The output module is used to train the convolutional neural network model, update parameters, and evaluate the model output results.
9. An electronic device, characterized in that: include: Memory and processor; 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, the steps of the digital twin data denoising method based on convolutional neural network described in any one of claims 1 to 7 are implemented.
10. A computer-readable storage medium, characterized in that: It stores computer executable instructions, which, when executed by a processor, can implement the steps of the digital twin data denoising method based on a convolutional neural network as described in any one of claims 1 to 7.